WAINESS PHD QUALIFYING EXAM
Qualifying Examination
Richard Wainess
Rossier School of Education
University of Southern California
to
Dr. Harold O’Neil (Chair)
Dr. Richard Clark
Dr. Edward Kazlauskas
Dr. Janice Schafrik
Dr. Yanis Yortsos (Outside member)
14009 Barner Ave.
Sylmar, CA 91342
Home Phone: (818) 364-9419
E-Mail: wainess@usc.edu
In partial fulfillment of the requirement for the Degree
Doctor of Philosophy in Education in
Educational Psychology and Technology
1
WAINESS PHD QUALIFYING EXAM
1.
2
Review the theoretical and empirical literature on the impact of games on learning and
motivation. Please, focus on training of adults and include a discussion of various game
characteristics, such as fun, competition, fantasy, and challenge.
The purpose of this review is to describe: the differences between games and simulations;
the motivational aspects of games; how game are currently being used; possible learning
outcomes attributed to games; and some issues pertaining to individual differences, such as
gender. While instructional games have been touted as the next great educational tool, research
hasn’t supported these hopes or claims. Richard Clark’s famous remarks that media, such as
video games, do not improve learning, they are simply a means of delivering content (CITE)
continue to prove true. Therefore, the question becomes, “If instructional games do not improve
learning, then why bother creating them?” The answer is relatively simple. As Clark (YEAR)
states, media do not improve learning. That means books and lectures, which are both mediums
(i.e., delivery vehicles for instruction), do not improve learning either. And that’s correct. A
lecture, in and of itself, does not improve learning. Neither does a book. It’s the instructional
methods that are incorporated into the book or into the lecturer that cause learning (Schacter &
Fagnano, 1999).
According to Cobb (1997), while there may be no unique medium for any job, that does
not mean that one medium isn’t better than another or that determinining which is better is not a
worthwhile empirical question. We use books and lectures because they’re practical—they
achieve some form of efficiency. Books are portable, can be read just about anywhere, and
provide a convenient way to reaccess large amounts of information. Lectures provide an
opportunity for spontaneous changes to presentations and for engaging in real-time dialog with
WAINESS PHD QUALIFYING EXAM
3
the instructor and among classmates. Video games also have the potential to work well for
specific situations. As Cobb (1997) argued, no medium is instrinsically better than another, but
one may be better than another depending upon instructional needs.
If computers can provide opportunities for learning, they should be considered, and not
patently rejected. However, they should also not be patently accepted. Computers are not
teaching devices. The instructional content they deliver and the instructional methods embedded
in that delivery are the teaching devices (CITE). Computers are simply powerful instruments
with robust capabilities for delivering experiences that are either unique from other media or, in
certain circumstances, more practical than using other media. For example, computers can
deliver games with a combination of features and capabilities not available with other electronic
media.
Researchers generally agree that games provide motivational outcomes (CITE). What is
questioned is learning outcomes, as measured by retention, and more importantly, transfer. While
there is inconsistent evidence of learning outcomes (Lee, 1999; CITE), agreement on the role of
motivation in learning (CITE) is relatively stable. Many, if not most, of the current models of
learning and problem solving include affective components, including motivation, as critical to
learning (e.g., the CANE model: Clark, 1999; O’Neil’s Problem Solving Model: O’Neil, YEAR;
GIVE MODEL NAME: Waller, Knapp, & Hunt, 2001). While it can easily be argued that
motivation does not mean learning will occur (CITE), it can equally be argued that motivation
supports learning by encouraging mental effort and persistence (CITE).
Researchers have cited a large number of benefits of computers and, particularly, of
computer games. Taylor, Renshaw, and Jensen (1997) commented that technology, in the form
of computer-assisted instruction (CAI), incorporates a wealth of techniques that promote
WAINESS PHD QUALIFYING EXAM
4
motivation, and has the potential to tremendously improve educational effectiveness. It has also
been argued that some empirical evidence exists that games can be effective tools for enhancing
learning and understanding of complex subject matter (Cordova, & Lepper, 1996; Ricci, Salas, &
Cannon-Bowers, 1996). And LeemKull, de Jong, de Hoog, and Christoph (2003) argued that
games and simulations provide students with a framework of rules and roles through which they
can learn interactively through a live experience, the students can tackle situations they might not
be prepared to risk in reality, and they can experiment with new ideas and strategies. Mayer,
Mautone, and Prothero (2002) also commented that when learning by doing in a physical
environment is not feasible, learning by doing can be implemented using computer simulations.
Mayer, Mautone, Prothero further commented that, in learning by doing in virtual environments,
students actively work in realistic situations that simulate authentic tasks for a particular
domain.According to Cross (1993), experiential learning is the process of gaining knowledge
through experience and behavior, and games are commonly used tools for experiential learning.
However, for games to be effective, they must embed sound instructional strategies and
appropriate content. According to Garris, Ahlers, & Driskell (2002), recent research has begun to
establish links between instructional strategies, motivational processes, and learning outcomes.
The researcher argued that people learn from active engagement with the environment and this
experience coupled with instructional support (i.e., debriefing, scaffolding) can provide an
effective learning environment. GIVE OTHER EXAMPLES?
In the early 1970s, Duke (1995) developed a series of game for the United Nations
Educational, Scientific, and Cultural Organization (UNESCO) for use in underdeveloped
countries. The games showed promise as a way to quickly provide a cogent model for urgent
problems, such as nutrition planning and economic planning. Also in the 1970s, a new type of
WAINESS PHD QUALIFYING EXAM
5
client for gaming began to emerge as, increasing, leadership of large public and private
organizations sought to locate new methods for developing strategic vision. These clients
included international banks, railroads, pharmaceutical companies and chemical companies
(Duke, 1995). Resnik and Sherer (1994) commented that computerized games and simulations
can be used like any other professional tool to deal with clients’ conflicts, their current troubled
situation, or with future dilemmas, and Rieber (1996) argued that games offer an organizational
function based on cognitive, social and cultural factors all related to play. According to Salas,
Bowers, & Rodenzer (1998), Simulation is a way of life in many aviation training environments.
For example, military, commercial, and general aviation all use simulations to train a variety of
tasks. And in the past decade, there has been considerable interest in using computer-simulated
(virtual) environments (VEs) for training spatial knowledge (Waller, 2000).
DEFINITIONS
One of the first problems areas with research into game and simulations is terminology.
Many studies that claim to have examined the use of games, did not use a game (CITE). At best,
they used an interactive multimedia that exhibits some of the features of a game, but not enough
features to actually be called a game. A similar problem occurs with simulations. A large number
of research studies use simulations but call them games (CITE). Because the goals and features
of games and simulations differ, it is important when examining the potential effects of the two
media to be clear about which one is being examined. However, there is little consensus in the
education and training literature on how games and simulations are defined.
According to Ricci, Salas, and Cannon-Bowers (1996), computer-based educational
games generally fall into one of two categories: simulation games and video games. Simulation
games model a process or mechanism relating task-relevant input changes to outcomes in a
simplified reality that may not have a definite endpoint. They often depend on learners reaching
conclusions through exploration of the relation between input changes and subsequent outcomes.
Video games, on the other hand, are competitive interactions bound by rules to achieve specified
WAINESS PHD QUALIFYING EXAM
6
goals that are dependent on skill or knowledge and that often involve chance and imaginary
settings (Randel, Morris, Wetzel, & Whitehill, 1992).
Games
According to Garris, Ahlers, and Driskell (2002) early work in defining games suggested
that there are no properties that are common to all games and that games belong to the same
semantic category only because they bear a family resemblance to one another (Garris, Ahlers, &
Driskell, 2002). Betz (1995) argued that a game is being played when the actions of individuals
are determined by both their own actions and the actions of one or more actors. Dempsey,
Haynes, Lucassen, and Casey (2002) commented that a game is a set of activities involving one
or more players. It has goals, constraints, payoffs, and consequences.
A number of researchers agree that games have rules (Crookall, Oxford, and Saunders,
1987; Dempsey, Haynes, Lucassen, and Casey, 2002; Garris, Ahlers, & Driskell, 2002; Ricci,
1994). Researchers also agree that games have goals and strategies to achieve those goals
(Crookall & Arai, 1995; Crookall, Oxford, and Saunders, 1987; Garris, Ahlers, & Driskell, 2002;
Ricci, 1994). Many researchers also agree that games have competition (Dempsey, Haynes,
Lucassen, and Casey; 2002) and consequences such as winning or losing (Crookall, Oxford, and
Saunders, 1987).
Betz (1995) further argued that games simulate whole systems, not parts, forcing players
to organize and integrate many skills. Students will learn from whole systems by their individual
actions, individual action being the student’s game moves. Crookall, Oxfodr, and Saunders also
noted that a game does not intend to represent any real-world system; it is a “real” system in its
own right. According to Duke (1995), games are situation specific. If well designed for a specific
client, the same game should not be expected to perform well in a different environment.
WAINESS PHD QUALIFYING EXAM
7
Simulations
In contrast to games, Crookall and Saunders (1989) viewed a simulation as a
representation of some real-world system that can also take on some aspects of reality for
participants or users. Similarly, Garris, Ahlers, & Driskell (2002) wrote that key features of
simulations are that they represent real-world systems, and Henderson, Klemes, and Eshet (2000)
commented that a simulation attempts to faithfully mimic an imaginary or real environment and
content that cannot be experienced directly, for such reasons as cost, danger, accessibility, or
time (Henderson, Klemes, & Eshet, 2000). Berson (1996) also argued that Simulations allow
students to engage in activities that would otherwise be too expensive, dangerous, or impractical
to conduct in the classroom.
Lee (1999) added that a simulation is defined as a computer program that relates them
together through cause and effect relationships. Reiber (1992; 1996) discussed microworlds, a
variant of simulations. He described microworlds as small representations of content areas or
domains that can be recognized by an expert, and simulations as designed to mimic real life
experiences, such as a flight simulator (Lee, 1999).
Thiagarajan (1998) argued that simulations do not reflect reality; they reflect someone’s
model of reality. According to Thiagarajan, a simulation is a representation of the features and
behaviors of one system through the use of another. Elements of a simulation correspond to
selected elements of the system being simulated. Some simulations focus on the physical features
of a real world object, such as a model airplane, while others focus on the processes and
interactions of real world events, such as mathematical equations that predict the number of
traffic fatalities during a holiday weekend (Thiagarajan, 1998). At the risk of introducing a bit
WAINESS PHD QUALIFYING EXAM
8
more ambiguity, Garris, Ahlers, and Driskell (2002) proposed that simulations can contain game
features, which leads to the final definition: sim-games.
Sim-Games
Thus, it is not too improper to consider games and simulations as similar in some
respects, keeping in mind the key distinction that simulations propose to represent reality and
games do not (Garris, Ahlers, & Driskell, 2002). Combining the features of the two media,
Rosenorn and Kofoed (1998) described simulation/gaming as a learning environment where
participants are actively involved in experiments, for example, in the form of role-plays, or
simulations of daily work situations, or developmental scenarios. Being away from the real
workplace, participants have the freedom to make wrong decisions and to learn from them.
This paper will use the definitions of games, simulations, and sim-games as defined by
Gredler (1996), which combine the most common features cited by the various researchers, and
yet provide clear distinctions between the three media. According to Gredler,
Games consist of rules that describe allowable player moves, game constraints
and privileges (such as ways of earning extra turns), and penalties for illegal
(nonpermissable) actions. Further, the rules may be imaginative in that they
need not relate to real-world events (p. 523).
This definition is in contrast to a simulation, which Gredler (1996) defines as “a dynamic
set of relationships among several variables that (1) change over time and (2) reflect authentic
causal processes” (p. 523). In addition, Gredler describes games as linear and simulations as nonlinear, and games as having a goal of winning while simulations have a goal of discovering
causal relationships. Gredler also defines a mixed metaphor referred to as simulation games or
WAINESS PHD QUALIFYING EXAM
9
gaming simulations, which is a blend of the features of the two interactive media: games and
simulations.
A major design weakness in game studies is that most studies compare simulations to
regular classroom instruction (lecture and/or classroom discussion). However, the instructional
goals for which each can be most effective often differ. The lecture method is likely to be
superior in transmitting items of information. In contrast, simulations have the potential to
develop the students’ mental models of complex situations as well as their problem-solving
strategies (Gredler, 1996).
MOTIVATIONAL ASPECTS OF GAMES
According to Garris, Ahlers, and Driskell (2002), motivated learners are easy to describe.
They are enthusiastic, focused and engaged, they are interested in and enjoy what they are doing,
they try hard, and they persist over time. Furthermore, they are self-determined and driven by
their own volition rather than external forces (Garris, Ahlers, & Driskell, 2002). Ricci, Salas, and
Cannon-Bowers (1996) defined motivation as “the direction, intensity, and persistence of
attentional effort invested by the trainee toward training.” Similarly, according to Malouf (19871988), continuing motivation is defined as returning to a task or a behavior without apparent
external pressure to do so when other appealing behaviors are available. And more simply, Story
and Sullivan (1986) commented that the most common measure of continuing motivation is
whether a student returns to the same task at a later time. In general, these descriptions of
motivation include the concept of continued motivation; persistence.
With regard to video games, and Asakawa and Gilbert (2003) argued that without sources
of motivation, players often lose interest and drop out of a game. However, there seems little
agreement among researchers on what those sources are—the specific set of elements or
characteristics that lead to motivation in any learning environment, and particularly with
educational games. According to Rieber (1996) and McGrener (1996), motivational researchers
WAINESS PHD QUALIFYING EXAM
10
have offered the following characteristics as common to all intrinsically motivating learning
environments: challenge, curiosity, fantasy, and control (Davis & Wiedenbeck, 2001; Lepper &
Malone, 1987; Malone, 1981; Malone & Lepper, 1987). Malone (1981) and others also included
fun as a criteria for motivation.
For interactive games, Stewart (1997) described some of the same elements as above, but
also included additional motivational elements; goals and outcomes. Locke and Latham (1990)
also commented that on of the most robust findings in the literature on motivation is that clear,
specific, and difficult goals lead to enhanced performance (Locke & Latham, 1990). They argued
that clear, specific goals allows the individual to perceive goal-feedback discrepancies, which are
seen as crucial in triggering greater attention and motivation. Clark (2001) argued that
motivation cannot exist without goals. The role of goals will be discussed in question 2. The
response to this question will focus on fantasy, control and manipulation, challenge and
complexity, curiosity, competition, feedback, and fun.
Fantasy
Research suggests that material may be learned more readily when presented in an
imagined context that interests the learner than when presented in a generic or decontextualized
form (Garris, Ahlers, & Driskell, 2002). Malone and Lepper (1987) defined fantasy as an
environment that evokes “mental images of physical or social situations that do not exist” (p.
250). According to Garris, Ahlers, and Driskell (2002), games involve imaginary worlds; activity
inside these worlds has no impact on the real world; and when involved in a game, nothing
outside the game is relevant. Rieber (1996) commented that fantasy is used to encourage learners
to imagine that they are completing the activity in a context in which they are really not present.
However, Rieber also described endognenous and exogenous fantasies. Endogenous fantasy
WAINESS PHD QUALIFYING EXAM
11
weaves relevant fantasy into a game, while exogenous simply sugar coast a learning environment
with fantasy. An example of an endogenous fantasy would be the use of a laboratory
environment to learn chemistry, since this environment is consistent with the domain. An
example of an exogenous environment would be a using a hangman game to learn spelling,
because hanging a person has nothing to do with spelling. Rieber (1996) describes endogenous
fantasy, but not exogenous fantasy, as important to intrinsic motivation, and further commented
that, unfortunately, exogenous fantasies are a common and popular element of many educational
games.
According to Malone and Lepper (1987), fantasies can offer analogies or metaphors for
real-world processes that allow the user to experience phenomena from varied perspectives. A
number of researchers (Anderson and Pickett, 1978; Ausubal, 1963; Malone and Lepper, 1978;
Malone and Lepper, 1987; Singer, 1973) argue that fantasies in the form of metaphors and
analogies provide learners with better understanding by allowing them to relate new information
to existing knowledge. According to Davis and Wiedenbeck (2001), metaphor also helps learners
to feel directly involved with objects in the domain so that the computer and interface become
invisible. The relationship of analogy and metaphor to learning is discussed in question 2.
Control and Manipulation
Hannifin and Sullivan (1996) define control as the exercise of authority or the ability to
regulate, direct, or command something. Control, or self-determination, promotes intrinsic
motivation because learners are given a sense of control over the choices of actions they may
take (deCharms, 1986; Deci, 1975; Lepper and Greene, 1978). Furthermore, control implies that
outcomes depend on learners’ choices and, therefore, learners should be able to produce
significant effects through their own actions (Davis, & Wiedenbeck, 2001). According to Garris,
WAINESS PHD QUALIFYING EXAM
12
Ahlers, & Driskell (2002), games evoke a sense of personal control when users are allowed to
select strategies, manage the direction of activities, and make decisions that directly affect
outcomes, even if those actions are not instructionally relevant (Garris, Ahlers, & Driskell,
2002).
However, Hannafin & Sullivan (1996) warned that research comparing the effects of
instructional programs that control all elements of the instruction (program control) and
instructional programs in which the learner has control over elements of the instructional
program (learner control) on learning achievement has yielded mixed results. Dillon and
Gabbard (1998) commented that novice and lower aptitude students have greater difficulty when
given control, compared to experts and higher aptitude students, and Niemiec, Sikorski, and
Walberg (DATE) argued that control does not appear to offer any special benefits for any type of
learning or under any type of condition.
Challenge and complexity
Challenge, also referred to as effectance, compentence, or mastery motivation (Bandura,
1977; Csikszentmihalyi, 1975; Deci, 1975; Harter, 1978; White, 1959), embodies the idea that
intrinsic motivation occurs when there is a match between a task and the learner’s skills. The
task should not be too easy nor too hard, because in either case, the learner will lose interest.
Similarly, Malone and Lepper (1987) have claimed that individuals desire an optimal level of
challenge; that is, tasks that are neither too easy nor too difficult to perform. Stewart (1997)
commented that games that are too easy will be dismissed quickly. According to Garris, Ahlers,
and Driskell (2002), there are several ways in which an optimal level of challenge can be
obtained. Goals should be clearly specified, yet the probability of obtaining that goal should be
uncertain, and goals must also be meaningful to the individual. They further argued that linking
WAINESS PHD QUALIFYING EXAM
13
activities to valued personal competencies, embedding activities within absorbing fantasy
scenarios, or engaging competitive or cooperative motivations could serve to make goals
meaningful (Garris, Ahlers, & Driskell, 2002).
Curiosity
According to Rieber (1996), challenge and curiosity are intertwined. Curiosity arises
from sitatuions in which there is complexity, incongruity, and discrepancy (Davis, &
Wiedenbeck, 2001). Sensory curiosity is the interest evoked by novel situations, cognitive
curiosity is the evoked by the desire for knowledge (Garris, Ahlers, & Driskell, 2002). Cognitive
curiosity motivates the learner to attempt to resolve the inconsistency through exploration
(Davis, & Wiedenbeck, 2001). Curiosity is identified in games by unusual visual or auditory
effects, and by paradoxes, incompleteness, and potential simplifications (Westbrook &
Braithwaite, 2002). Curiosity is the desire to acquire more information. This is a primary
component of the players’ motivation to learn how to operate the game (Westbrook &
Braithwaite, 2001).
Malone and Lepper (1987) noted that curiosity is one of the primary factors that drive
learning and is related to the concept of mystery. Garris, Ahlers, and Driskell (2002) commented
that make the distinction between curiosity and mystery to reflect the difference between
curiosity, which resides in the individual, and mystery, which is an external feature of the game
itself. Thus, mystery evokes curiosity in the individual, and this leads to the question of what
constitutes mystery (Garris, Ahlers, & Driskell, 2002). Research suggests that mystery is
enhanced by incongruity of information, complexity, novelty, surprise, and violation of
expectations (Berlyne, 1960), incompatibility between ideas and inability the predict the future
(Kagan, 1972), and information that is incomplete and inconsistent (Malone & Lepper, 1987).
Competition
Studies on competition with games and simulations have mixed results, due to
preferences and reward structures. In a study by Porter, Bird, and Wunder (1990-1991)
WAINESS PHD QUALIFYING EXAM
14
examining competition and reward structures found that the greatest effects of reward structure
were seen in the performance of those with the most pronounced attitudes toward either
competition or cooperation. The results suggested that performance was better when the reward
structure matched the individual’s preference. According to the authors, implications are that
emphasis on competition will enhance the performance of some learners but will inhibit the
performance of others (Porter, Bird, and Wunder, 1990-1991).
Yu (2001) investigated the relative effectiveness of cooperation with and without intergroup competition in promoting student performance, attitudes, and perceptions toward subject
matter studied, computers, and interpersonal context. With fifth-graders as participants, Yu
found that cooperation without inter-group competition resulted in better attitudes toward the
subject matter studies, and promoted more positive inter-personal relationships both within and
among the learning groups than cooperation/competition did (Yu, 2001). The exchange of ideas
and information both within and among the learning groups also tended to be more effective and
efficient when cooperation did not take place in the context of inter-group competition (Yu,
2001).
Feedback
Feedback within games can also easily be provided in order for learners to quickly
evaluate their progress against the established game goal. This feedback can take many forms,
such as textual, visual, and aural (Rieber, 1996). According to Ricci, Salas, and Cannon-Bowers
(1996), within the computer-based game environment, feedback is provided in various forms
including audio cues, score, and remediation immediately following performance. They argued
that these feedback attributes can produce significant differences in learner attitudes, resulting in
increased attention to the learning environment.
Fun
Learning that is fun appears to be more effective (Lepper & Cordova, 1992). Quinn
(1994, 1997) argued that for games to benefit educational practice and learning, they need to
combine fun elements with aspects of instructional design and system design that include
WAINESS PHD QUALIFYING EXAM
15
motivational, learning, and interactive components. According to Malone (1981a, b) three
elements (fantasy, curiosity, and challenge) contribute to the fun in games (as cited in Armory et
al, 1999). While fun has been cited as important for motivation and, ultimately, for learning,
there is no empirical evidence supporting the concept of fun. This might be due to the fact that
fun is not a construct but, rather, represents other concepts or constructs. Relevant alternative
concepts or constructs are play, engagement, and flow.
Play is entertainment without fear of present or future consequences; it is fun (Resnick &
Sherer, 1994). According to Rieber, Smith, and Noah (1998), play describes the intense learning
experience in which both adults and children voluntarily devote enormous amounts of time,
energy, and commitment and, at the same time, derive great enjoyment from the experience. This
is termed serious play to distinguish it from other interpretations which may have negative
connotations (Rieber, Smith, & Noah, 1998). Webster et al. (1993) found that labeling software
training as play showed improved motivation and performance. According to Rieber and Matzko
(2001) serious play is an example of an optimal life experience.
Csikszentmihalyi (1975; 1990) defines an optimal experience as one in which a person is
so involved in an activity that nothing else seems to matter; termed flow or a flow experience.
When completely absorbed in and activity, he or she is “carried by the flow,” hence the origin of
the theory’s name (Rieber and Matzko, 2001). Rieber and Matzko (2001) also content that a
person may be considered in flow during an activity when experiencing one or more of the
following characteristics: Hours pass with little notice; challenge is optimized; feelings of selfconsciousness disappear; the activity’s goals and feedback are clear; attention is completely
absorbed in the activity; one feels in control; and one feels freed from other worries (Rieber &
Matzko, 2001). And according to Davis and Wiedenbeck (2001), an activity that is highly
WAINESS PHD QUALIFYING EXAM
16
intrinsically motivating can become all-encompassing to the extent that the individual
experiences a sense of total involvement, losing track of time, space, and other events. Davis and
Wiedenbeck also argued that the interaction style of a software package is expected to have a
significant effect on intensity of flow. However, Rieber and Matzko contend that play and flow
differ in one respect; learning is an expressed outcome of serious play but not of flow (Rieber &
Matzko, 2001).
Engagement is defined as a feeling of directly working on the objects of interest in the
worlds rather than on surrogates. According to Davis and Wiedenbeck, this interaction or
engagement can be used along with the components of Malone and Lepper’s (DATE) intrinsic
motivation model to explain the effect of an interaction style on intrinsic motivation, or flow
(Davis, & Wiedenbeck, 2001). Garris, Ahlers, and Driskell (2002) commented that training
professional are interested in the intensity of involvement and engagement that computer games
can invoke, that the “holy grail” of training professionals is to harness the motivational
properties of computer games to enhance learning and accomplish instructional objectives.
Garris, Ahlers, and Driskell further argued that engagement in game play leads to the
achievement of training objectives and specific learning outcomes (Garris, Ahlers, & Driskell,
2002).
LEARNING/OUTCOMES
Druckman (1995) concluded that games seem to be effective in enhancing motivation and
increasing student interest in subject matter, yet the extent to which this translates into more
effective learning is less clear (Garris, Ahlers, & Driskell, 2002).
Anything that contributes to the increase of emotion (the quality of the design of video
games, for example) reinforces the attraction of the game but not necessarily its educational
interest (Brougere, 1999, p. 140).
WAINESS PHD QUALIFYING EXAM
17
Although students generally seem to prefer games over other, more traditional, classroom
training media, reviews have reported mixed results regarding the training effectiveness of
games (Garris, Ahlers, & Driskell, 2002).
In sum, liking the simulation does not translate to learning (Salas, Bowers, & Rhodenizer,
1998).
According to Ricci, Salas, and Cannon-Bowers (1996), results of their study provided
evidence that computer-based gaming can enhance learning and retention of knowledge. They
further commented that positive trainee reaction might increase the likelihood of student
involvement with training (i.e., devote extra time to training), but it is not a necessary factor for
enhanced learning.
Simulations and games have been cited as beneficial to a number of disciplines and for a
number of educational and training situations, including aviation training (Salas, Bowers, &
Rhodenizer, 1998), aviation crew resource management (Baker, 1993), military mission
preparation (Spiker & Nullmeyer, n.d.), laboratory simulation (Betz, 1995), chemistry and
physics education (Khoo & Koh, 1998), urban geography and planning (Adams, 1998; Betz,
1995), farm and ranch management (Cross, 1993), language training (Hubbard, 1991), disaster
management (Stolk, Alexandrian, Gros, & Paggio, 2001), and medicine and health care
(Westbrook & Braithwaite, 2001; Yair, Mintz, & Litvak, 2001). For business, games and
simulations have been cited as useful for teaching strategic planning (Washburn & Gosen, 2001;
Wolfe & Roge, 1997), finance (Santos, 2002), portfolio management (Brozik, & Zapalska,
2002), marketing (Washburn & Gosen), knowledge management (Leemkull, de Jong, de Hoog,
& Christoph, 2003), and media buying (King & Morrison, 1998).
Playing games is a way of learning laws of logic and methods of thinking. Older adults
can benefit from these experiences as much as younger populations (Weisman, 1994).
In addition to teaching domain-specific skills, games have been used to teach generalized
skills. Since the mid 1980s, a number of researchers have used the game Space Fortress, a 2-D,
simplistic arcade-style game, with a hexagonal “fortress” in the center of the screen surrounded
by two concentric hexagons and a space ship, to improve abilities that transferred far outside
gameplay, such as improving the results of fighter pilot training (Day, Arthur, and Gettman,
2001). According to Day, Arthur, and Gettman, 2001; Gopher, Weil, and Bareket, 1994;
Shebilske, Regian, Arthur, and Jordan,1992), Space Fortress includes “important informationprocessing and psychomotor demands” (p. 1024).
When the business game method was pitted against the case approach, and when casebased evaluation criteria were not employed, the game approach was superior to cases in
producing knowledge gains. Less can be stated, however, regarding the relationship between
gaming procedures and learning outcomes (Wolfe, 1997).
WAINESS PHD QUALIFYING EXAM
18
Whether verbal information, motor skills, or intellectual skills are the object of the
instruction, computer games can be designed to address specific learning outcomes (Dempsey,
Haynes, Lucassen, & Casey, 2002).
In a series of five experiments, Green and Bavelier (2003) showed the potential of video
games to alter visual selection attention, using a popular action video game, Medal of Honor (by
Electronic Arts). The control group played Tetris, a popular game requiring visual-motor control.
While both treatment group and control group improved visual selection attention, the amount of
improvement in visual selection attention was significantly higher in the treatment group (Green
& Bavelier, 2003).
By forcing players to simultaneously juggle a number of varied tasks (detect new
enemies, track existing enemies, and avoid getting hurt, among other tasks), action-video-game
playing pushes the limits of the three rather different aspects of visual attention. It leads to
detectable effects on new tasks and at untrained locations after only 10 days of training.
Therefore, although video-game playing may seem to be rather mindless, it is capable of
radically altering visual attention processing (Green & Bavelier, 2003).
In series of two experiment of college students (8 video game experts and 8 novices) to
test visual attention strategies, experienced video game players showed a marked ability to not
focus on low-probability areas for target objects as compared to novice video game player. And
compared to novices, video game experts were faster responders at both the low and high
probability locations (Greenfield, DeWinstanley, Kilpatrick, & Kaye, 1994).
Taken together, the two experiments showed that skilled or expert video game players
had better skills for monitoring two locations on a visual screen and that experimental video
game practice could alter the strategies of attentional deployment so that the response time for
the low-probability target was reduced (Greenfield, DeWinstanley, Kilpatrick, & Kaye, 1994).
According to Leemkull, de Jong, de Hoog, and Christoph (2003), much of the work on
the evaluation of games has been anecdotal, descriptive, or judgmental, but there are some
indications that they are effective and superior to case studies in producing knowledge gains,
especially in the area of strategic management (Wolfe, 1997).
According to Garris, Ahlers, and Driskell (2002), in an early meta-analysis of the
effectiveness of simulation games, Dekkers and Donatti (1981) found a negative relationship
between duration of training and training effectiveness. Simulation game became less effective
the longer the game was used (suggesting that perhaps trainees became bored over time).
The ultimate test of the knowledge and skill acquisition is usually not in the knowing but
in the ability to use knowledge appropriately—in the translation of knowledge into behavior
WAINESS PHD QUALIFYING EXAM
19
(Ruben, 1999). Moreover, coming to know, and especially being able to use knowledge and
skills generally, requires reinforcement, application, repetition, and often practice in a variety of
settings and contexts, in order for it to become fully understood, integrated, and accessible in
future situations (Ruben, 1999).
According to Ricci, Salas, and Cannon-Bowers (1996), the proper assessment of training
effects involves the examination of transfer or retention of the skills toward which training was
directed.
According to assimilation theory, there are two kinds of learning: rote learning and
meaningful learning. Rote learning occurs through repetition and memorization. It can lead to
successful performance in situations identical or very similar to those in which a skill was
initially learning. However, skills gained through rote learning are not easily extensible to other
situations, because they are not based on deep understanding of the material learning.
Meaningful learning, on the other hand, equips the learner for problem solving and extension of
learned concepts to situations different from the context in which the skill was initially learned
(Davis, & Wiedenbeck, 2001; Mayer, 1981).Meaningful learning takes place when the learner
draws connections between the new material to be learned and related knowledge already in
long-term memory, known as the “assimilative context” (Ausubel, 1963; Davis, & Wiedenbeck,
2001).
Meaningful learning results in an understanding of the basic concepts of the new material
through its integration with existing knowledge (Davis, & Wiedenbeck, 2001).
However, there is general consensus that learning with interactive environments such as
games, simulations, and adventures is not effective when no instructional measure or support are
added (Leemkull, de Jong, de Hoog, & Christoph, 2003).
According to Thiagarajan (1998), simulations can be used for instruction, awareness,
performance assessment, teambuilding, transfer, research, and therapy. However, if not
embedded with sound instructional design, games and simulations often end up truncated
exercises often mislabeled as simulations (Gredler, 1996). Gredler further commented that
poorly developed exercises are not effective in achieving the objectives for which simulations
are most appropriate—that of developing students’ problem-solving skills (Gredler, 1996).
In other words, outcomes are affected by the instructional strategies employed (Wolfe,
1997).
The generally accepted position is that games themselves are not sufficient for learning
but that there are elements of games that can be activated within an instructional context that
may enhance the learning process (Garris, Ahlers, & Driskell, 2002).
There are a number of empirical studies that have examined the effects of game-based
instructional programs on learning (Garris, Ahlers, & Driskell, 2002). For example, both
Whitehall and Mcdonald (1993) and Ricci et al. (1996) found that instruction incorporating game
features lead to improved learning (Garris, Ahlers, & Driskell, 2002).
de Jong and van Joolingen (1998), after reviewing a large number of studies on learning
from simulations, concluded, “there is no clear and univocal outcome in favor of simulations. An
explanation why simulation based learning does not improve learning results can be found in the
intrinsic problems that learners may have with discovering learning” (p. 181). These problems
are related to processes such as hypothesis generation, design of experiments, interpretation of
WAINESS PHD QUALIFYING EXAM
20
data, and regulation of learning. After analyzing a large number of studies, de Jong and van
Joolingen (1998) concluded that adding instructional support to simulations might help to
improve the situation (Leemkull, de Jong, de Hoog, & Christoph, 2003).
Berson (1996) cited Becker (1990) as stating that common problems encountered
throughout the literature on computer effectiveness in the social studies include: (a) design flaws
exacerbated by poor data collection procedures; (b) inadequate analysis of data and insufficient
presentation of results; and (c) poor description of the methodology, including the setting and
conditions under which the program was implemented (Berson, 1996).
With regards to research into the effectiveness of computers in social studies,
methodological problems persist in the areas of insufficient treatment definitions and
descriptions, inadequate sampling procedures, and incomplete reporting of statistical results.
Overall, there is paucity of empirical evidence, and most conclusions are impressionistic.
Consequently, there is not satisfactory evidence on which to base decisions to integrate
computers into social studies instruction (Berson, 1996).
While benefits were found with MIF, it was not compared to other forms of learned, and
therefore, it is unclear as to any sort of “benefit” that could be derived from using the software
(Henderson, Klemes, & Eshet, 2000).
When students were asked to assess what they learned from the learning unit, students
reported learning negotiating skills and the role that timing and deadlines played in the buying
and selling process. Learning outcomes were not assessed (King & Morrison, 1998).
In meta-analyzing a number of studies and meta-analyses on video games, Lee (1999)
commented that effect size never tells us under what conditions students learn more, less, or not
at all compared with the comparison group. For instructional prescription, we need information
dealing with instructional variable, such as instructional mode, instructional sequence,
knowledge domain, and learner characteristics. If we don’t know how these variables are
connected to learning outcomes, there is no way to prescribe appropriate conditions of
instruction for specific target learners. As a result, findings of these studies cannot contribute to
the quality of instruction in various educational settings (Lee, 1999).
REFLECTION/DEBRIEF
A game cannot be designed to directly provide learning. A moment of reflexivity is
required to make transfer and learning possible. Games require reflection, which enables the
passage from play to learning. Therefore, debriefing (after action review) appears to be an
essential contribution to research on play and gaming in education (Brougere, 1999).
Participants of a simulation are not in a position to learn anything worthwhile unless they
are required and encouraged to reflect on the experience through the process of debriefing.
WAINESS PHD QUALIFYING EXAM
21
Structured approaches to debriefing are more effective for learning than unstructured approaches
(Thiagarajan, 1998).
Debriefing is the review and analysis of events that occurred in the game itself.
Debriefing provides a link between what is represented in the simulation/gaming experience and
the real world. It allows the participant to draw parallels between game events and real-worled
events. The debriefing allows us to transform game events into learning experiences. Debriefing
may include a description of events that occurred in the game, analysis of why they occurred,
and the discussion of mistakes and corrective actions. Learning by doing must be coupled with
the opportunity reflect and abstract relevant information for effective learning to occur and for
learners to link knowledge gained to the real world. Debriefing and scaffolding techniques
provide the guidance and support to aid this process (Garris, Ahlers, & Driskell, 2002).
Student should be prepared and encouraged to study and critique the simulation model
event as it provides them with a degree of insight into urban processes that is not otherwise
possible (Adams, 1998).
Results of the study indicated that the four-phase approach (introduction, instruction,
engagement, and reflection/debriefing) was an effective approach to learning. The debriefing
process not only provided the opportunity to improve learning, it provided the opportunity for
feedback on the simulation, which resulted in many suggestions for improvement (Leemkull, de
Jong, de Hoog, & Christoph, 2003). For example, users indicated the game could be more
challenging and competitive (Leemkull, de Jong, de Hoog, & Christoph, 2003).
Using a simulation game, the “experimentarium,” Rosenorn and Kofoed (1998),
examined to nature of various phases of reflection in a learning process: reflection-before-action,
reflection-in-action, and reflection-on-action. Reflection-before-action occurs before learning,
and involves learners considering the types of problems they hope to solve more successfully.
Reflection-in-action occurs during a pause in learning and involves learners reviewing goals,
asking themselves whether they are moving in the right direction, and making necessary
adjustments. Reflection-on-action occurs after learning is completed, when learners consider
what was learning, why the outcome turned out to be as it did, and how the outcome may be
applied in the near future. According to the authors, these reflection periods contribute to the
depth and durability of the learning as well as to the changes in attitudes. To examine their
assumptions, participants worked in the experimentarium, virtual room for, according to the
researchers, experiments that are removed from the daily life of an organization, to give
employees, managers, and consultants the opportunity to develop and test new ideas.
Tools such as performance measurement, cognitive and task analysis, scenario design,
and feedback and debriefing mechanisms are necessary to ensure learning in simulation-based
training systems (Salas, Bowers, & Rhodenizer, 1998).
Debriefing is too important to be added on as an afterthought to an interactive simulation,
especially one used for training, increasing awareness, or team building. No simulation package
can be considered complete without an extensive debriefing guide (Thiagarajan, 1998).
CONCLUSION/DISCUSSION/SUMMARY
WAINESS PHD QUALIFYING EXAM
22
The modern computer technology has made possible a new and rich learning
environment, the simulation. In an instructional simulation, students learn by actually performing
activities to be learned in a context that is similar to the real world. Instructional simulation is
used in most cases as unguided discovery learning. Students can generate and test hypotheses in
a simulated environment by examining changes in the environment based on their input. Unlike
the traditional classroom instruction, in which students’ roles are passive in most cases, this
particular type of instruction requires students to be involved in their learning in an active way
(Lee, 1999).
Educators and trainers began to take notice of the power and potential of computer games
for education and training back in the 1970s and 1980s (Donchin, 1989; Malone, 1981; Malone
& Lepper, 1987; Ramsberger, Hopwood, Hargan, & Underfull, 1983; Ruben, 1999; Thomas &
Macredie, 1994). Computer games were hypothesized to be potentially useful for instructional
purposes and were also hypothesized to provide multiple benefits: (a) complex and diverse
approaches to learning processes and outcomes; (b) interactivity; (c) ability to address cognitive
as well as affective learning issues; and perhaps most importantly, (d) motivation for learning.
According to Ricci, Salas, and Cannon-Bowers (1996), motivation can be defined as “the
direction, intensity, and persistence of attentional effort invested by the trainee toward training.”
Currently, the increase power and flexibility of computers technology is contributing to
renewed interest in games and simulations. This development coincides with the current
perspective of effective instruction in which meaningful learning depends on the construction of
knowledge by the learner. Games and simulations, which can provide an environment for
learner’s construction of new knowledge, have the potential to become a major component of
this focus (Gredler, 1996).
If we are able to participate in games and simulations, it is because as children we learned
to master rules. We even ask ourselves if play does not prepare for a number of learning
situations characterized by a more or less explicit dimension of simulation, which supposes
master of the second degree and rules specific to certain situations. This is probably why the
Romans had the same name ludus for play and for school and why the teacher was called
magister ludi (Brougere, 1999).
Like other forms of instruction, simulations and games are likely to be more effective
with some students than with others (Gredler, 1996).
According to Ruben (1999), the theoretical foundations for simulations, games, and other
forms of interactive, experience-based learning had been in place at least since the writings of
Aristotle and the practices of Socrates (Ruben, 1999).
The public should not accept the rhetoric that technology makes learning easier and more
efficient, because ease and efficiency are not prerequisite conditions for deep and meaningful
learning (Schacter & Fagnano, 1999).
We than make the more important distinction that computer technologies, when designed
according to sound learning theory and pedagogy, have, and can substantially improve student
learning (Schacter & Fagnano, 1999).
Computer-Based Instruction (CBI) has been shown to moderately improve student
learning and achievement (Schacter & Fagnano, 1999).
WAINESS PHD QUALIFYING EXAM
23
Schacter and Fagnano (1999) conducted a meta-analysis of 12 meta-analyses on
computer-based instruction, comprised of a total of 546 individual studies, with subjects from
elementary, secondary, precollege, special, and college institutions (Schacter & Fagnano, 1999).
When computer technologies are designed around principles gleaned from learning
theories and implemented systematically, one can argue that the effect that these technologies
have on student learning and achievement are both powerful and transformative. Technologies
designed around educational and psychological theory compare favorably to other education
reform efforts because they have embedded proven teaching principles into the technology.
Thus, one gets the effects of both the teaching reform and the technology (Schacter & Fagnano,
1999).
Our findings from two experiments involving high school students suggest that the
effectiveness of CAI may go beyond basic cognitive processes, such as rote memory. While
previous work has found that CAI can be as effective as traditional teaching methods for rote
memory, it has not always been shown to be more effective (Taylor, Renshaw, & Jensen, 1997).
According to Cobb (1997), Clark’s work has prompted educators to be skeptical of
inflated media claims; to notice when expensive media are promoted where cheap would do; to
center instructional designs on the learner rather than the medium; to track learning effect to
instructional cause at the lowest level of analysis possible (medium attribute rather than medium
per se, method rather than medium, message rather than method; Cobb, 1997).
There have been a great number of experimental studies to examine the instructional
value of simulation. In most cases of these studies, researchers used expository instructional
methods, such as traditional classroom lectures or computer-based tutorials for comparison
groups. The research results from these studies were conflicting (Lee, 1999).
Instructional games offer the opportunity for the learner to learn by doing, to become
engaged in authentic learning experiences. However, people do not always learn by doing.
Sometimes we learn by observing; sometimes we learn by being told. “Learners are not passive
blotters at which we toss information; nor are they active sponges that absorb all they experience
unaided. We must temper our enthusiasm for the gaming approach with knowledge that
instructional games must be carefully constructed to provide both an engaging first-person
experience as well as appropriate learner support” (Garris, Ahlers, & Driskell, 2002, p. 461).
Games, simulations, and case studies have an important role in education and training in
putting learning into a context. Furthermore, they are constructivistic environments in which
students are invited to actively solve problems. Games and simulations provide students with a
framework of rules and roles through which they can learn interactively through a live
experience. They can tackle situations they might not be prepared to risk in reality, and they can
experiment with new ideas and strategies (Leemkull, de Jong, de Hoog, & Christoph, 2003).
They involve individual and group interpretations of given information, the capacity to suspend
disbelief, and a willingness to play with the components of a situation in making new patterns
WAINESS PHD QUALIFYING EXAM
24
and generating net problems (Jacues, 1995; as cited in Leemkull, de Jong, de Hoog, &
Christoph, 2003).
A type of learning environment, which is very close to games, is simulation. Simulations
resemble games in that both contain a model of some kind of system and learners can provide
input (changes to variable values or specific actions) and observe the consequences of their
actions (Leemkull, de Jong, de Hoog, & Christoph, 2003).
Play is traditionally viewed as applying only to young children (Rieber, 1996). There is
also a sense of risk attached to suggesting an adult is at play. Work is respectable, play is not.
Another misconception is that play is easy. Quite the contrary, even as adults we tend to engage
in unusually challenging and difficult activities when we play, such as sports, music, hobbies,
and games like chess (Rieber, 1996).
Play is a difficult concept to define. Play appears to be one of those constructs that is
obvious at the tacit level but extremely difficult to articulate in concrete terms—we all know it
when we see it or experience it. Its definition can also be culturally and politically constrained
(Rieber, 1996). Nevertheless, according to Rieber (1996), play is generally defined as having the
following attributes: (a) it is usually voluntary; (b) it is intrinsically motivating, that is, it is
pleasurable for its own sake and is not dependent on external rewards; (c) it involves some level
of active, often physical, engagement; and (d) it is distinct from other behavior by having a make
believe quality (Blanchard & Cheska, 1985; Csikszenmihalyi, 1990; Pellegrini, 1995; Pellegrini
& Smith, 1993; Yawkey & Pellegrini, 1984).
The commonsense tendency is for people to define play as the opposite of work, but this
is misleading (Rieber, 1996).
Computer games offer a new possibility for wedding motivation and self-regulated
learning within a constructivist framework, on which strives to combine both training nd
education, practice and reflection, into a seamless learning experience (Rieber, Smith, & Noah,
1998).
There are clearly many media for any instructional job, but this does not mean they all do
it at the same level of efficiency—whether economic, logistical, social, or cognitive (Cobb,
1997).
WAINESS PHD QUALIFYING EXAM
25
References for Question 1
Adams, P. C. (1998, March/April). Teaching and learning with SimCity 2000
Amory, A., Naicker, K., Vincent, J., & Adams, C. (1999). The use of computer games as an
educational tool: Identification of appropriate game types and game elements.
Asakawa, T., Gilbert, N. (2003). Synthesizing experiences: Lessons to be learned from Internetmediated simulation games.
Baker, D., Prince, C., Shrestha, L., Oser, R., & Salas, E. (1993). Aviation computer games for
crew resource management training.
Barnett, M. A., Vitaglione, G. D., Harper, K. K. G., Quackenbush, S. W., Steadman, L. A., &
Valdez, B. S. (1997). Late adolescents’ experiences with and attitudes toward
videogames.
Bauza, G. B., & Gelabert, M. E. (1995, June). The Hakayak’s last odyssey: A computer game
with a difference
Bell, H. H., & Crane, P. M. (1993) Training utility of multiship air combat simulation.
Benne, M. R., & Baxter, K. K. (1998). An assessment of two computerized vocabulary games
reveals that players improve as a result of review
Bernard, K. J. (1997, December). Strategic management games: A review [Electronic Version].
Simulation & Gaming Special Issue: Teaching Strategic Management, 28(4), 395-422.
Berson, M. J. (1996, Summer). Effectiveness of computer technology in the social studies: A
review of the literature. Journal of Research on Computing in Education, 28(4), 486499.
Betz, J. A. (1995/1996). Computer games: Increase learning in an interactive multidisciplinary
environment.
Brougere, G. (1999, June). Some elements relating to children’s play and adult
simulation/gaming.
Brozik, D., & Zapalska, A. (2002, June). The PORTFOLIO GAME: Decision making in a
dynamic environment.
Carr, P. D., & Groves, G. (1998). The Internet-based operations simulation game.
Carvalho, G. F. (1991). Evaluating computerized business simulators for objective learning
validity.
Choi, W. (1997). Designing effective scenarios for computer-based instructional simulations:
Classification of essential features.
Cobb, T. (1997). Cognitive efficiency: Toward a revised theory of media. Educational
Technology Research and Development, 45(4), 21-35.
Cross, T. L. (1993, Fall). AgVenture: A farming strategy computer game.
Davis, S., & Wiedenbeck, S. (2001). The mediating effects of intrinsic motivation, ease of use
and usefulness perceptions on performance in first-time and subsequent computer users.
Interacting with Computers, 13, 549-580.
Dempsey, J. V., Haynes, L. L., Lucassen, B. A., & Casey, M. S. (2002). Forty simple computer
games and what they could mean to educators.
Duke, R. D. (1995, December). Gaming: An emergent discipline. Simulation & Gaming, Silver
Anniversary Issue (Part 4), 426-438.
Fabiani, M., Buckley, J., Gratton, G., Coles, M. G. H., Donchin, E., & Logie, R. (1989). The
training of complex task performance.
Forbus, K. D. (2001). Articulate software for science and engineering Education.
WAINESS PHD QUALIFYING EXAM
26
Garris, R., Ahlers, R., & Driskell, J. E. (2002). Games, motivation, and learning: A research and
practice model.
Gopher, D., Weil, M., & Bareket, T. (1994). Transfer of skill from a computer game trainer to
flight.
Gopher, D., Weil, M., & Siegel, D. (1989). Practice under changing priorities: An approach to
the training of complex skills. Acta Psychologica, 71, 147-177.
Gredler, M.E. (1996). Educational games and simulations: a technology in search of a research
paradigm.
Green, C. S., & Bavelier, D. (2003, May 29). Action video game modifies visual selective
attention.
Greenfield, P. M., deWinstanley, P., Kilpatrick, H., & Kaye, D. (1996). Action video games and
informal education: Effects on strategies for dividing visual attention.
Henderson, L., Klemes, J., & Eshet, Y. (2000). Just playing a game? Educational simulation
software and cognitive outcomes.
Herselman, M. E. (2000). University students benefiting from the medium of computer games: A
case study.
Hindle, K. (2002, June). A grounded theory for teaching entrepreneurship using simulation
games.
Hubbard, P. (1991, June). Evaluating computer games for language learning.
Keys, J. B. (1997). Strategic management games: a review [Electronic Version].
Khoo, G.-s., & Koh, t.-s. (1998). Using visualization and simulation tools in tertiary science
education [Electronic Version].
King, K. W., & Morrison, M. (1998, Autumn). A media buying simulation game using the
Internet.
Kirriemuir, J. (2002, February). Video gaming, education, and digital learning technologies:
Relevance and opportunities.
Lee, J. (1999). Effectiveness of computer-based instructional simulation: A meta analysis.
Leemkuil, H., de Jong, T., de Hoog, R., & Christoph, N. (2003). KM Quest: A collaborative
Internet-based simulation game.
Malouf, D. (1987-1988). The effect of instructional computer games on continuing student
motivation.
Moreno, R., & Mayer, R. E. (2002). Learning science in virtual reality multimedia environments:
Role of methods and media.
Noyes, J. M., & Garland, K. J. (2003). Solving the Tower of Hanoi: Does mode of presentation
matter? Computers in Human Behavior, 19, 579-592.
Park, O.-C., & Gittelman, S. S. (1995). Dynamic characteristics of mental models and dynamic
visual displays. Instructional Science, 23, 303-320.
Porter, D. B. (1995). Computer games: Paradigms of opportunity.
Prislin, R., Jordan, J. A., Worchel, S., Semmer, F. T., & Shebilske, W. L. (1996, September).
Effects of group discussion on acquisition of complex skills. Human Factors, 38(3), 404416.
Resnick, H. (1994). Introduction: Electronic tools for education and training.
Resnick, H., & Sherer, M. (1994). Computerized games in the human services--An introduction.
Ricci, K. E. (1994, Summer). The use of computer-based videogames in knowledge acquisition
and retention.
WAINESS PHD QUALIFYING EXAM
27
Ricci, K. E., Salas, E., & Cannon-Bowers, J. A. (1996). Do computer-based games facilitate
knowledge acquisition and retention?
Rieber, L. P. (1996). Seriously considering play: Designing interactive learning environments
based on the blending of microworlds, simulations, and games.
Rieber, L. P., & Matzko, M. J. (Jan/Feb 2001). Serious design for serious play in physics.
Rieber, L. P., Smith, L., & Noah, D. (1998, November/December). The value of serious play.
Rosenorn, T., & Kofoed, L. B. (1998). Reflection in learning processes through
simulation/gaming. Simulation & Gaming, 29(4), 432-440.
Ruben, B. D. (1999, December). Simulations, Games, and experience-based learning: The quest
for a new paradigm for teaching and learning.
Salas, E., Bowers, C. A., & Rhodenizer, L. (1998). It is not how much you have but how you use
it: Toward a rational use of simulation to support aviation training. The International
Journal of Aviation Psychology, 8(3), 197-208.
Salies, T. G. (2002). Promoting strategic competence: What simulations can do for you.
Salomon, G. (1983). The differential investment of mental effort in learning from different
sources. Educational Psychology, 18(1), 42-50.
Santos, J. (2002, Winter). Developing and implementing an Internet-based financial system
simulation game.
Schacter, J., & Fagnano, C. (1999). Does computer technology improve student learning and
achievement? How, when and under what conditions? Journal of Educational
Computing Research, 20(4), 329-343.
Spiker, V. A., & Nullmeyer, R. T. (n.d.). Benefits and limitations of simulation-based mission
planning and rehearsal. Unpublished manuscript.
Standen, P. J., Brown, D. J., & Cromby, J. J. (2001). The effective use of virtual environments in
the education and rehabilitation of students with intellectual disabilities.
Stewart, K. M. (1997, Spring). Beyond entertainment: Using interactive games in web-based
instruction.
Stolk, D., Alexandrian, D., Gros, B., & Paggio, R. (2001). Gaming and multimedia applications
for environmental crisis management training.
Story, N., & Sullivan, H. J. (1986, November/December). Factors that influence continuing
motivation. Journal of Educational Research, 80(2), 86-92.
Taylor, G. L., & Disinger, J. F. (1997, Spring). The potential role of virtual reality in
environmental education [Electronic Version]. The Journal of Environmental Education,
28, 38-43.
Taylor, H. A., Renshaw, C. E., & Jensen, M. D. (1997). Effects of computer-based role-playing
on decision making skills. Journal of Educational Computing Research, 17(2), 147164.
Tennyson, r. D., & Breuer, K. (2002). Improving problem solving and creativity through use of
complex-dynamic simulations.
Thiagarajan, S. (1998, Sept/October). The myths and realities of simulations in performance
technology. Educational Technology, 38(4), 35-41.
Thiagarajan, S. (2001, May). Fun in the workplace.
Waller, D. (2000). Individual differences in spatial learning from computer-simulated
environments. Journal of Experimental Psychology, 6(4), 307-321.
Waller, D., Knapp, D., & Hunt, E. (2001, Spring). Spatial representations of virtual mazes: The
role of visual fidelity and individual differences. Human Factors, 43(1), 147-158.
WAINESS PHD QUALIFYING EXAM
28
Walters, B. A., Coalter, T. M., & Rasheed, A. M. A. (1997). Simulation games in business policy
courses: Is there value for students [Electronic Version]?
Washbush, J., & Gosen, J. (2001, September). An exploration of game-derived learning in total
enterprise simulations.
Weisman, S. (1994). Computer games for the frail elderly. In H. Resnick (Ed.), Electronic Tools
for Social Work Practice and Education (pp. 229-234). Bington, NY: The Haworth Press.
Westrom, M. & Shaban, A. (1992, Summer). Intrinsic motivation in microcomputer games.
Whitehill, B. V., & McDonald, B. A. (1993, September). Improving learning persistence of
military personnel by enhancing motivation in a technical training program. Simulation &
Gaming, 24(3), 294-313.
Winn, W., & Jackson, R. (1999, July/August). Fourteen propositions about educational uses of
virtual reality.
Wolfe, J. (1997, December). The effectiveness of business games in strategic management
course work [Electronic Version].
Wolfe, J., & Roge, J. N. (1997, December). Computerized general management games as
strategic management learning environments [Electronic Version].
Yair, Y., Mintz, R., & Litvak, S. (2001). 3D-virtual reality in science education: An implication
for astronomy teaching.
Yildiz, R., & Atkins, M., 1952- (1996, May). The cognitive impact of multimedia simulations on
14 year old students [Electronic Version].
Yu, F.-Y. (2001). Competition within computer-assisted cooperative learning environments:
Cognitive, affective, and social outcomes. Journal of Educational Computing Research,
24(2), 99-117.
WAINESS PHD QUALIFYING EXAM
2.
29
Review the theoretical and empirical literature on the relationship of cognitive load to
learning. Please, include a discussion of cognitive load in relationship to interactive
media (e.g., multimedia and games). Be sure to focus types of cognitive load (e.g.,
intrinsic, germane, and extraneous load).
Educational technology as a field now seems in a mood to move beyond the issue of
whether media contribute to learning, to acknowledge that media are here to stay in any case, and
drop the learning issues without resolving it (Cobb, 1997). However, Cobb (1997) contends that
the issue can be resolved in a more principles manner with one minor adjustment to Clark’s
position. He suggested that if a recurrent concept in Clark’s discourse, “efficiency,” is expanded
to include “cognitive efficiency,” then media choices become connected to learning, in some
circumstances (Cobb, 1997).
For first-time users, engagement appears to have resulted simply from the novelty of
learning a new computer application, regardless of the interaction style (Davis, & Wiedenbeck,
2001).
When the human operator has to master a very complex task, it may be advisable to
train different task components separately (Fabriani et al., 1989).
Briggs and Naylor (1962) and Naylor and Briggs (1963) proposed that two dimensions
are crucial in determining amenability of a task to part training: task complexity and task
organization. Complexity refers to the load imposed by each task component taken in isolation
while organization refers to the processing demands that originate from the interactions among
different task components. Briggs and Naylor claim that part-task training is most efficient when
task complexity is high while task organization is low. This is because it is under these
conditions that practice on individual task components makes it easier for the trainee to
determine the optimal means for dealing with each part, without the distraction imposed by other
task components. Thus, the trainee’s conception of the task is clarified and transfer of training
can occur (Fabriani et al., 1989).
However, one of the main advantages of part-task training—enabling the subjects to
perform parts in isolation—is also one of its main drawbacks. This is because skills practiced in
isolation may not integrate well with each other, and may not transfer well to the while task. In
addition, even if one agrees that it would be desirable to adopt some form of part-task training, it
is not always obvious how the task should be partitioned into components (Fabriani et al., 1989).
Gopher and his colleagues (Gopher et al., 1989) combined concepts derived from
schema models and attention theory to develop an approach to training that depends upon shifts
in attention and emphasis. They assumed that it is preferable to expose the subject to the entire
task throughout the training period. This assures integration by avoiding the partitioning of the
WAINESS PHD QUALIFYING EXAM
30
task. However, part-task training was achieved by emphasizing different task components during
different phases of training. This allowed the trainee to focus on the component that is
emphasized without losing sight of the whole task (Fabriani et al., 1989).
The hierarchical approach to training, developed by Frederiksen and White (1989),
drew from theories of the role of mental models in learning to devise a set of “problem
environments” which shaped the development of the trainee’s mental model. In addition,
Frederiksen and White determined the subject’s optimal strategy through an analysis of the task
based on a principled decomposition of its component skills. Then, a batter of training sub-tasks,
none of which need bear any similarity to the whole task, was developed. The training was
designed to emphasize the hierarchical nature of the sub-tasks. Sub-tasks administered later in
training required skills taught in previous sub-tasks, and the subject was let to incorporate the
elements of the optimal strategy in an integrated fashion (Fabriani et al., 1989).
Both regimes (emphasis on change, and hierarchical) were successful in improving the
subjects’ performance in a complex perceptual-motor task—the Space Fortress game. Those that
received both treatments achieved the highest performance improvement. Yet the hierarchical
group achieved the highest performance in absolute terms. A repeat of the studies at the
University of Illinois resulted in less extreme results. Differences were potentially attributed to
differences in subjects (Fabriani et al., 1989).
In addition, and perhaps more importantly, the exposure to the whole task—the
standard Space Fortress game—varied considerably for subjects in the two training regimes. The
participants in the Gopher et al. study played the whole game 200% to 400% more often than the
Frederiksen and White study. Therefore, differences may have been due to differences in
familiarity to with the whole task (Fabriani et al., 1989).
In this study comparing and integrated and a hierarchical approach to learning Space
Fortress, care was taken to eliminate the differences in training schedule and subject pool
(Fabriani et al., 1989).
It is equally important to assess the degree to which the acquired skills are robust to
interference. Therefore, in the present study we examined the degree to which subjects were
capable of performing the standard Space Fortress task concurrently with several other tasks.
These concurrent tasks formed a battery designed to assess the load placed on different
components of working memory (Fabriani et al., 1989). The study included 33 university
students (all right-handed males, 18-24 years old). There were three groups, a control group that
learning the play the game as a whole task, a treatment group (the integrated group) that received
emphasis training while playing the whole game, and the hierarchical group that trained on
subtasks and eventually the entire game (Fabriani et al., 1989).
The integrated group began with scores below the control group, but eventually
outperformed the control group. The hierarchical group performed more poorly than either of the
other groups in early stages, but eventually outperformed both the control and the integrated
group. During the dual task (the interference) stage of the training, the hierarchical group’s
performance gap increased over the other two groups increased. However, the hierarchical group
was the most affected least by the less disruptive secondary tasks but more with the more
disruptive secondary tasks. The integrated training group was more resistant to disruption by the
presence of concurrent secondary tasks than the other two groups (Fabriani et al., 1989).
In dividing performance into 28 variables (e.g. number of fortress hits, shooting
efficiency, perfect of foe mines killed), the hierarchical group outperformed the integrated group
on 20 of the variables and the control group on 22 of the variable.
WAINESS PHD QUALIFYING EXAM
31
In the initial stages of the study, a screening test of shooting ability was conducted. At
the end of training, the low scorers in the hierarchical group performed best, followed by the
control group, then the integrated group. It appears that integrated training is detrimental, or at
best of no value, for subjects with low screen scored. For the high scorers, the curve of the
integrated group is only slightly below that of the hierarchical group, and above the control
group. And for the medium scorers, the curve of the integrated group is intermediate with respect
to the curves of the other two groups (Fabriani et al., 1989).
In summary, on most performance variable, subjects trained with either of the part-task
training methods achieved higher scores than did subjects trained on the whole task. The
hierarchical task achieved superior performance when the game was performed alone. The
integrated group’s performance was more resistant to disruption by concurrently performed
secondary tasks. The training regime and the initial capability of the subject as measured by the
aiming screening task interacted in determining the effectiveness of training. Subjects who
scored high in the screening task taken before training began did well regardless of the part-task
training method to which they were subjected. On the other hand, the method of training did
make a difference for subjects with low and medium screening scores. The hierarchical method
was particularly beneficial, and the integrated methods particularly detrimental, for those
subjects who scored poorly on the screening task (Fabriani et al., 1989).
Improving science and engineering education is a critical problem for technological
societies, who, in addition to needing scientists, engineers, and technicians, need a scientifically
literate population in order to make wise decisions. We believe a new kind of educational
software, articulate software, can help solve this problem. Articulate software understands the
domain being learning in human-like ways, and can provide explanations and coaching to help
learners master it. Articulate software is made possible by advances in artificial intelligence,
particularly qualitative physics, combined with the ongoing revolution in computer technology
(Forbus, 2001).
By embedding human-like models of entities and processes in software, the software’s
understanding can be used to provide explanations that are directly coupled to how specific
results are derived. These explanations can delve into topics that traditional software cannot
handle, for example, why a process was considered to occur and hwy a specific approximation
makes sense (Forbus, 2001).
In additional to their commercial popularity, computer games have captured the
attention of training professionals and educators. There several reasons for this professional
interest. First, there has been a major shift in the field of learning from a traditional, didactic
model of instruction to a learner-centered model that emphasizes a more active learner role. This
represents a shift away from the “learning by listening” model of instruction to one in which
students learn by doing (Garris, Ahlers, & Driskell, 2002).
Space Fortress has a substantial history as a research instrument for complex problem
solving task, and is the instrument used in a number of the studies appearing in this article.
According to Day, Arthur, and Gettman (2001), Space Fortress includes “important informationprocessing and psychomotor demands” (p. 1024). Space Fortress is a visually simplistic, 2-D
video game, with a hexagonal “fortress” in the center of the screen surrounded by two concentric
hexagons and a space ship. The ship’s path and rotation are controlled by a joystick. Missile
firing is controlled by the mouse. Participants try to destroy the moving fortress by shooting
missiles, while trying to avoid collision with the fortress, and mine that periodically appear.
Participants benefit by shooting foe mines, but are penalized for shooting friendly mines.
WAINESS PHD QUALIFYING EXAM
32
Additionally, bonus events occur which require specific mouse actions. The ship works in
frictionless space, meaning that once it’s in motion, it will continue to move at a constant speed
unless altered by another joystick movement, to speed it, slow down, or stop. Speed and
movement can also affect score. The various events and conditions already described result in
points being added or deducted. To achieve a maximum score, subjects must destroy the fortress,
defend themselves, destroy all mines, manage their resources of missiles and point bonus, and
avoid being hit by either fortress or mines (see Arthur, Strong, Jordan, Williamson, Shelbilske,
and Regian, 1995, for a detailed description of the game).
Gopher, Weil, and Bareket (1994) stated that both flight training and Space Fortress
include continuous and discrete manual control, visual and spatial orientation, procedural
knowledge involving long- and short-term memory information, and high attention demands
under severe time constraints. Verbal communication was also introduced into the game to
simulate the demands in the flight situation. Gopher, Weil, and Bareket applied two approaches
to learning: emphasis change and hierarchical part-training. Under the emphasis change
approach, subjects practiced the whole game at all times, but they were led through instructions
and auxiliary feedback indicators to vary their focus of attention on different aspects of the game
during different game trials. Under this method, participants were exposed to the full load of the
task and taught alternative ways for coping with the task. In hierarchical part-training, the whole
task is decomposed, and before subjects are introduced to the full game, they are led through a
sequence of simplified part games, which gradually become more integrative and complex.
Subjects are given verbal tips on recommended behavior, based on subject matter experts.
The Gopher, Weil, and Bareket (1994) study involved 58 flight school cadets, with one
group learning Space Fortress and a control group that did not receive any video game training.
One group of cadets, the full training (FT) group, was given both emphasis change and
hierarchical part-task methods. The other group, the emphasis only training group (EOT), was
given emphasis change and attention-management procedures. The control group consisted of
cadets who were matched in ability to the experimental groups. Each group of flight cadets in the
experimental groups were trained for 10 one-hour sessions consisting of 10 to 14 trails of 2 or 3
minutes each. Transfer effects from the game training to actual flight were tested during eight
flights (45-60 minutes each) of the transition stage to the high-performance jet trainer.
The results from the study by Gopher, Weil, and Bareket (1994) provide strong support
for the emphasis change approach for teaching generalizable skills. Subjects in the FT group
obtained significantly higher final game scores on all measures of game performance, compared
with the EOT group. Despite the large differences in final game scores, the FT and EOT groups
did not differ in subsequent flight performance. The game group was significantly better in its
flight performance than the non-game group; About one-third of the subjects in the game group
were included in the highest score category, whereas only 3.4% were in the lower category.
None of the non-game subjects were included in the high scores category, whereas 28.6% were
in the lowest category. The game group increased its probability of graduation by 30%. The most
significant result of the games was that the percentage of graduates from the game group was
twice that of the non-game group. The authors contend that part-task training appears to focus
trainee attention on task specific elements, while emphasis-change training results in more
generalizable skills. Therefore, while the FT approach resulted in higher game scores (which
would benefit from the specific focus) than the EOT approach (which emphasized generalizable
skills), it did not transfer to higher scores in flight performance (because only the generalizable
skills which both groups acquired from the game experience were transferable to actual flight).
WAINESS PHD QUALIFYING EXAM
33
Information that is not held in working memory will need to be retained by the longterm memory system. Storing more knowledge in long-term memory reduces the load on
working memory. This results in a greater capacity being made available for active processing.
When problem solving, if the various rules have been learning and their application practiced,
this information can be held in long-term memory. Thus, once the individual is familiar with the
problem, s/he will be in a better position to plan how to solve the problem. There are two
important issues here. First, is the role of experience in terms of aiding the forming of mental
representations, reducing the memory loads and facilitating planning activities. Second, the
implications of having a display of the problem that acts like an “external memory” and provides
the user with information about the problem at all times. It is reasonable to conclude, therefore,
that an important characteristic of using a computer is that it reduces the load on working
memory (Noyes & Garland, 2003). The use of “memory” is of interest here, since it might be
argued that the display screen is merely providing an external representation of the problem
rather than a memory. However, whatever the terminology, there are many advantages to having
this situation when problem-solving (Noyes & Garland, 2003); It reduces the load on internal
working memory; storing less information in the internal working memory means that there is les
chance of forgetting information—this reduces the chance of the problem solver making errors;
Problem-solvers may consider the task to be less cognitively complex, because of the reduced
load on working memory—hence, they feel more confident about solving the problem; and it
allows the user to become more focused on solving the problem as opposed to remembering the
rules (Noyes & Garland, 2003).
In their simplest form, verbal protocols require individuals to report their thoughts as
they carry out the task. This is particularly appropriate for tasks that involve sequential
processing, because this mirrors the consecutive nature of the thought processes. it is then
relatively easy to talk through solving the problem. Verbal protocols, and in particular, the socalled “thinking aloud” techniques, have been shown to aid problem solving and this benefit has
been well documented (Ahlum-heath & DiVesta, 1986; Berry, 1983; Ericsson & Simon, 1993; as
cited in Noyes & Garland, 2003).
The Tower of Hanoi is a well-known problem-solving task that has been used many
times in an experimental setting (see Anderson & Douglass, 2001; as cited in Noyes & Garland,
2003). It involves a problem space about which the problem-solver has very little specific
domain knowledge, and solvers need to acquire additional knowledge to decompose a goal into
sub-goals. They need to learn how to evaluate the outcomes of their actions in order to sort the
actions that they carry out in terms of their contribution to solving a sub-goal (and ultimately, the
overriding goals of solving the Tower of Hanoi puzzle). It is a relatively straightforward taks
with a set of very simple instructions that can be easily represented (Noyes & Garland, 2003).
The Tower of Hanoi puzzle comprises a number of vertical pegs, and doughnut-shaped
disks of graduated sizes that fit onto these posts. At the start of the problem-solving exercise, all
the disks are arranged in pyramid form on one of the end pges with the largest disk on the
bottom. The ‘problem’ is to move all of the disks from this end peg to the other end peg, subject
to a number of constraints. These are: (1) only one disk can be moved at a time; (2) a disk cannot
be moved to be placed on a disk that is smaller than itself, and (3) no disk can be put aside. Any
number of disks can be used; the minimum number of moves is 2N – 1, where N equals the
number of disks. However, five disks and three pegs provide a problem of sufficient difficulty
that can be solved within a relatively short period, as only 31 moves need to be carried out
(Noyes & Garland, 2003).
WAINESS PHD QUALIFYING EXAM
34
In summary, it was hypothesized that for a simple, problem-solving task such as the
Tower of Hanoi, having access to a model of the problem well benefit performance in terms of
more successful problem-solving (i.e., completion of the puzzle), and more efficient problemsolving (i.e., fewer moves and faster times; Noyes & Garland, 2003).
In the first experiment, participants made fewer moves using the mental representation
than the physical and computer models, and more people gave up when trying to solve the puzzle
using the physical model. This suggests that problem-solving “in the head” is more efficient than
using a computer (Noyes & Garland, 2003). The physical model took the most moves.
The computer presentation of the Tower of Hanoi puzzle provides a means of
representing the problem pictorially. Thus, it provides an “intermediate representation” between
the physical and mental models. Compared to the physical model, manipulation of the computergenerated version of the puzzle was very easy and involved a “drag and drop” mouse oeration to
move the disks on the screen. Thus, individuals could very quickly elicit the desire moves;
perhaps, this ease of operation resulte din them not focusing on reaching the end-point by the
most efficient means, and as a result, a “trial and error” approach was being adopted (Noyes &
Garland, 2003). Problem-solvers did not have to be too careful about making sure the next move
was the right one (Noyes & Garland, 2003).
In all three experiments, participants were faster when using the computer version of
the puzzle in terms of moves per second (Noyes & Garland, 2003).
In the case of the Tower of Hanoi puzzle, making the moves relies on the internal
working memory, but the person also needs to apply the restriction rules to making the moves.
This information, although not shown on the screen in the same way as the puzzle, is “present”
in the computer; hence, the user has access to an “external (working) memory.” Further, there is
little cost of interruption of carrying out the task as information is not lost, for example, if
concentration momentarily lapses. This may help explain the greater number of moves when
using the computer version. Participants had so much information present on the display screen
that there is no need to be totally focused on solving the problem (Noyes & Garland, 2003).
In essence, it could be argued that display-based problem solving reduces the
complexity of the mental processes involved by reducing the loads on working memory (Noyes
& Garland, 2003).
When presented with a computer model for the Tower of Hanoi, there is no need to
make any effort to form your own mental representation, because there is an external
representation on the display screen. Consequently, the problem-solver faced with the computer
version of the problem is immediately at a disadvantage, because they are not having the benefit
of having to apply themselves to beginning to solve the problem (Noyes & Garland, 2003).
Further, the computer’s representation of the problem may not match their internal representation
of the problem. In effect, the computer model may be providing so-called “cognitive clutter” that
is interfering with the optimum route for problem solving. In contrast, solving the problem using
only a mental representation allows you to build a strong representation of the problem, and this
results in more efficient problem solving (Noyes & Garland, 2003).
A further explanation may lie in the use of verbal protocols. Individuals solving the
puzzle using the mental representation version of the Tower of Hanoi were required to talk
through the moves they were making, that is, to think aloud. This form of protocol analysis is
particularly appropriate to transformation problems such as the Tower of Hanoi (Noyes &
Garland, 2003).
WAINESS PHD QUALIFYING EXAM
35
One of the difficulties associated with the user of any computer-generated model is the
nature of the interface. This is particularly the case when considering problem solving as the
ergonomics of the display and the user’s interaction with it can influence the ease with which the
problem can be solved. The importance of the design of the interface must not be overlooked,
because as Zhang (1991) pointed out the external representations of the problem provide
memory aids. Hence, the design of these can change the nature of the task. The precise design of
the computer-generated Tower of Hanoi will, therefore, influence the solving of the puzzle. This
needs to be taken into account when generalizing from Tower of Hanoi studies (Noyes &
Garland, 2003).
Mental models explain human cognitive processes of understanding external reality,
translating reality into internal representation and utilizing it in problem solving (Park &
Gittelman, 1995).
Mental model formation depends heavily on the conceptualizations that individuals
bring to a task. When interacting with the environment, with others, and with the artifacts of
technology, people form mental models of themselves and the things with which they interact
(Norman, 1983; as cited in Park & Gittelman, 1995).
We can learn something from a source of information, given that it carries some
potentially useful information, if we perceive it to warrant the investment of effort needed for the
learning to take place (Salomon, 1983).
The argument of this article is that learning, in its generic sense, greatly depends on the
differential way in which sources of information are perceived, for these perceptions influence
the mental effort expended in the learning process. This argument is comprised of two ideas.
First, the amount of mental effort learners invest in extracting information from a source,
discriminating among its information units, remembering the information, or elaborating it in
their minds, is influenced by the way they perceive that source (Salomon, 1983).
Second, it is argues that learning is strongly influenced by the amount of mental effort
learners invest in processing the material—that is, the “depth” or “thoughtfulness” with which
they process it (Salomon, 1983).
It is often assumed that what determines effort investment is the difficulty of the
stimulus or task—that is, its novelty or complexity or the amount of “cognitive capacity” that is
uses as a function of its content density or structural complexity (Salomon, 1983).
Do the learners’ justified or unjustified perceptions of a medium’s quality—its typical
attributes and the task one usually performs with it—influence their learning as well (Salomon,
1983)?
There are at least two kinds of elaboration to be considered. Elaboration can be
automatic, carried out by well mastered mental processes over which a person exercises little
conscious control, and which are carried out with great ease in large chunks. Such elaborations
would usually be the result of much repeated practice and training (Salomon, 1983, p. 43).
Elaboration can, however, be controlled and nonautomatic, requiring attention and effort. Such
elaborations would generally be applied to relatively new, complex, or otherwise less practiced
material. Given a specific level of relevant skill mastery, it is the employment of controlled,
effortful elaborations that improves learning in the sense that better recall, more generated
inferences, and better integration of the material in memory (Salomon, 1983, p. 43).
Mental effort, relevant to the task and material, appears to be the feature that
distinguishes between mindless or shallow processing on the one hand, and mindful or deep
processing, on the other. Little effort is expended when processing is carried out automatically or
WAINESS PHD QUALIFYING EXAM
36
mindlessly (Salomon, 1983). Mindlessness refers to the ostensibly unattentive behavior of
otherwise intelligent people; as the absence of conscious processing (Salomon, 1983). According
to Salomon (1983), mindfulness refers to a cognitively active state characterized by the
conscious manipulation of the envisioned elements (Langer & Imber, 1980).
Shallow processing refers to automatic processing of well rehearsed features. Deep
processing refers to the effortful employment of non-automatic elaborations (Salomon, 1983).
Mental effort investment and motivation are not to be equated. Motivation is a driving
force, but for learning to actually take place, some specific relevant mental activity needs to be
activated. This activity is assumed to be the employment of nonautomatic effortful elaborations
(Salomon, 1983).
Mental effort invested in processing means the employment of nonautomatic
elaborations performed on the material (Salomon, 1983).
All things being equal, the amount of mental effort should be a combined function of
one’s mastery of the relevant mental skills, and the nature of the stimulus to be processed for a
particular task. One would expect that, given a particular stimulus task and a desired level of
performance, children with a better mastery of relevant skills with invest less effort in processing
a unit of material than children who have a poorer mastery of the requisite skills (Salomon,
1983).
Better skill mastery implies more automaticity of skill employment, and hence, by
definition, a smaller amount of mental effort is needed to teach the same pre-set level of message
comprehension by the more skillful child. Similarly, more demanding, difficult, or novel stimuli
are generally expected to evoke more effort investment than simple stimuli (Salomon, 1983).
The reason attributed to children’s shallow processing of television is the medium’s
shallowness, pictoriality, “crowdedness,” and rapid pace. On the other hand, the more serious,
deeper treatment of print is claimed to reflect the more demanding nature of that medium, its
relative abstractness, and imagery-generation requirements (Salomon, 1983).
The nature of the stimuli, their complexity, novelty, structuredness, pace, and the like,
in interaction with learners’ abilities, affect performance or learning outcomes only to some
extent. Perceptions, in the sense of dispositions, preconceptions, attitudes, and attributions, also
play an important role in the way one treats information. Furthermore, perceptions do not
always, nor necessarily, reflect the true nature of the given material (Salomon, 1983).
Langer and Benevento (1978) have shown that when people perceive a message as
highly familiar in structure, they forgo any detailed processing of its content and respond to it
mindlessly. Such mindlessness can take place based on its structural features (Salomon, 1983).
Strong preconceptions or perceptions of some material, source, or medium may affect
the actual investment of mental effort, and hence of learning (Salomon, 1983).
The material presented on TV is perceived to be shallower and less variable than the
material presented in print, even when the content areas (e.g., adventure stories, sport, science)
are held constant (Salomon, 1983).
Perceived self-efficacy refers to subjective judgments of how well one can execute a
course of action, handle a situation, learn a new skikll or unit of knowledge, and the like
(Salomon, 1983).
Perceived self-efficacy has much to do with how a class of stimuli is perceived. The
more demanding it is perceived to be the less efficacious would the perceivers be about it, and
the more familiar, easy, or shallow it is perceived, the more efficacious they would feel in
handling it (Salomon, 1983).
WAINESS PHD QUALIFYING EXAM
37
It follows from the above that perceived self efficacy should be related to the
perception of demand characteristics (the latter includes the perceived worthwhileness of
expending effort), and that both should affect effort investment jointly (Salomon, 1983).
Expectancy theory tells us that two factors are involved hers: the importance of a
particular yield, and the price to be paid for it. If one learns that information from one certain
sources is not very important, why should more effort be invested in it (Salomon, 1983)?
Learning is the process of acquiring knowledge (Tennyson & Breuer, 2002).
Thinking is the process of employing knowledge (Tennyson & Breuer, 2002).
The purpose of this article is to present an instructional method that has been shown to
significantly improve higher-order thinking strategies by enhancing the above descried
processes. The method employs computer-managed simulations that present contextually
meaningful problem situations that require learners to prepare solution proposals. The simulation
assesses the proposals and offers to learners the consequences of their decisions while also
iteratively updating the situational conditions. This type of simulation, unlike conventional
simulations that are used for the acquisition of knowledge, is complex-dynamic, requiring the
students to fully employ their knowledge base by generating soluations to domain-specific
problems (Tennyson & Breuer, 2002).
An important contribution of cognitive psychology in the past decade has been the
development of theories and models to explain the processes of learning and thinking. The value
of these theories is that they offer operational definitions of not only how learning and thinking
occur, but also why it occurs. The why explanation provides more direct means for
understanding how instructional methods may accomplish predictable improvements in both
learning and thinking (Tennyson & Breuer, 2002).
Tennyson and Breuer (2002) proposed an Interactive Cognitive Learning and Thinking
Model for cognitive learning based on a complexity theory perspective. The stages include:
external environment and behavior (action); sensory receptors (memory); executive control
(meta/automatic); cognitive strategies; affects; and knowledge base. The executive control
includes perceptions, attention, and resources (effort). The cognitive strategies include
construction of new knowledge and strategies, differentiation for selection of existing
knowledge, and integration, for restructuring and elaboration of knowledge. Affects include
motivation, feelings, attitudes, emotions, anxiety, and values. And knowledge base includes
declarative knowledge (knowing that), procedural knowledge (knowing how), and contextual
knowledge (knowing why, when, and where) (Tennyson & Breuer, 2002).
In Tennyson and Breuer’s (2002) model, there are bi-directional connections between
external environment and sensory receptions, sensory receptors and executive control, executive
control and cognitive strategies, executive controls and affects, and executive controls and
knowledge base (Tennyson & Breuer, 2002).
The cognitive processes of differentiation, integration, and construction of knowledge
are abilities that can be improved by effective instructional methods (Tennyson & Breuer, 2002).
In sequential information processing models, the labels short-term memory and
working memory are used synonymously to describe many of the functions of the executive
control component (Tennyson & Breuer, 2002).
There is agreement in the psychological field that the knowledge base has no capacity
limits and that knowledge is considered permanent, although it may become difficult to retrieve
in certain situations (Tennyson & Breuer, 2002).
WAINESS PHD QUALIFYING EXAM
38
The knowledge base consists of domains of knowledge that can be described as
complex networks (or schemas) of information (e.g., concepts or propositions). Within a domain,
knowledge is organized into meaningful modules called schemata. Schemata vary per individual
according to amount, organization, and accessibility (Tennyson & Breuer, 2002).
Motivation influences both attention and maintenance processes (Tennyson & Breuer,
2002).
Values and feeling would influence the criteria associated with acquisition of
contextual knowledge. Anxiety as an affect variable influences much of the internal processing
abilities. Along with emotions, anxiety can be a serious interfering variable in the cognitive
system (Tennyson & Breuer, 2002).
Differentiation is defined as the twofold ability to understand a given situation and to
apply appropriate contextual criteria by which to selectively retrieve specific knowledge from the
knowledge base (Tennyson & Breuer, 2002).
Integration is the ability to elaborate or restructure existing knowledge in the service of
previously unencountered problem situations (Tennyson & Breuer, 2002).
Construction is the ability to both discover and create new knowledge in novel or
unique situations (Tennyson & Breuer, 2002).
Higher-order thinking strategies involve three cognitive processes: differentiation,
integration, and construction of knowledge (Tennyson & Breuer, 2002).
The more fully developed the knowledge based in memory, the greater the
opportunities for differentiation and integration and possibly, creation of knowledge (Tennyson
& Breuer, 2002).
The cognitive processes of differentiation, integration, and construction of knowledge
are abilities that can be improved by effective instructional methods. Intelligence, on the other
hand, seems not to be directly influenced by instructional conditions (Tennyson & Breuer, 2002).
Thinking strategies represent a continuum of conditions ranging from a low-order of
automatic recall of existing knowledge to a high-order of constructing knowledge. From low to
high, the strategies are recall, problem solving, and creativity (Tennyson & Breuer, 2002).
Recall represents the retrieval of knowledge from memory. Recall strategies involve an
automatic differentiation of knowledge from the existing knowledge base. A higher-order recall
strategy is employed when more complex situations in which new conditions that have not been
previously encountered are part of the problem. With recall, the integration of all appropriate
schemata is required to succeed at a task (Tennyson & Breuer, 2002).
Problem solving is associated with situations dealing with previously unencountered
problems. That is, the term problem solving is most often defined for situations that require
employing knowledge in the service of problems not already in storage. In these types of
situations, the thinking strategies require the integration of knowledge to form new knowledge
(Tennyson & Breuer, 2002).
A first condition of problem solving involves the differentiation process of selecting
knowledge that is currently in storage using known criteria. Concurrently, this selected
knowledge is integrated to form a new knowledge. Cognitive complexity within this condition
focuses on elaborating the existing knowledge base (Tennyson & Breuer, 2002).
Problem solving may also involve situations requiring the construction of knowledge
by employing the entire cognitive system. Therefore, the sophistication of a proposed solution is
a factor of the person’s knowledge base, level of cognitive complexity, higher-order thinking
strategies, and intelligence (Tennyson & Breuer, 2002).
WAINESS PHD QUALIFYING EXAM
39
The highest order of human cognitive processing is the creating of a problem situation.
Rather than having the external environment dictate the situation, the individual, internally,
creates the need or problem. The highest cognitive condition exists when the individual creates
not only the situation but also constructs both the new knowledge and criteria necessary for
solution. Constructing knowledge involves the entire cognitive system. Creativity seems to
involve both the conscientious deliberations of differentiation and integration and the
spontaneous integrations that operate at a metacognitive level of awareness (Tennyson & Breuer,
2002).
Simulation in educational computing is a widely employed technique to teach certain
types of complex tasks. The purpose of using simulations is to teach a task as a complete whole
instead of in successive parts. For example, simulations are used in aviation training to replicate
the complex interaction of a number of variables needed to successfully pilot an airplane.
Learning the numerous variables simultaneously is necessary to fully understand the whole
concept of flying. We define these types of situations as task-oriented because the educational
objective is to learn the variables (declarative and procedural knowledge) and their context
(conceptual knowledge) (Tennyson & Breuer, 2002).
The assumption in complex-dynamic simulations is that that student has acquired
sufficient knowledge to proceed in the development of thinking strategies employing the
cognitive processes of differentiation, integration, and construction (Tennyson & Breuer, 2002).
The Minnesota Adaptive Instructional System (MAIS) is basically a computer research
tool in which we have investigated instructional variables associated with improving learning
according to individual differences and needs. As such, the instructional variables are
represented in adaptive instructional strategies that in turn are monitored for each student by an
expert tutor system using artificial intelligence techniques (Tennyson & Breuer, 2002).
The MAIS consists of two main components: (a) a curriculum component (or macro)
which maintains a student model (i.e., the cognitive, affective, and memory models of each
student) and a curricular level knowledge base; (b) an instructional component (or micro) that
adapts the instructional strategies according to moment-to-moment learning progress and need.
Both components are managed by expert tutor systems (Tennyson & Breuer, 2002).
Unlike task-oriented simulations, complex-dynamic simulations do not necessarily
employ the computer as an instruction delivery system. The main purpose of the computer in a
complex-dynamic simulation is to manage the simulation with the student doing most of the
learning activities with resources other than the computer. Depending on the learning situation,
the computer could certainly be used as a learning and instructional resource (Tennyson &
Breuer, 2002).
Research indicates that intra-group interactions in problem-solving situations
contribute to cognitive complexity development because learners are confronted with different
interpretations of the given simulation conditions by other group members (Tennyson & Breuer,
2002).
An important issue in cooperative learning is the procedure used to group students. Our
research shows that, for development of thinking strategies, group membership should be based
on similarity of ability in cognitive complexity (Tennyson & Breuer, 2002).
The expectations a user has about a computer’s behavior come from mental models,
while the “expectations” a computer has of a user come from user models. The two types of
WAINESS PHD QUALIFYING EXAM
40
models are similar in that they produce expectations that one “intelligent agent” has of another.
The fundamental distinction between them is that mental models inside the head while user
models occur inside a computer. Thus, mental models can be modified only indirectly by
training, while user models can be examined and manipulate directly (Allen, 1997).
Models are approximations to objects or processes which maintain some essential
aspects of the original (Allen, 1997, p. 49). In cognitive psychology, mental models are usually
considered to be the ways in which people model processes. The emphasis on process
distinguishes mental models from other types of cognitive organizers such as schemas. Models
of processes may be thought of as simple machines or transducers which combine or transform
inputs to produce outputs (Allen, 1997).
A mental model synthesizes several steps of a process and organizes them as a unit. A
mental model does not have to represent all of the steps which compose the actual process (e.g.,
the model of a computer program or a detailed account of the computer’s transistors) (Allen,
1997).
Mental models may be incomplete and may even be internally inconsistent. The
representation of a mental model is, obviously, not the same as the real-world processes it is
modeling (Allen, 1997).
Because they are not directly observable, several different types of evidence have been
used to infer the characteristics of mental models: predictions, explanations and diagnosis,
training, and other methods. Users predict what will happen next in a sequential process and how
changes in one part of they system will be reflected in other parts of the system. Explanations
about the causes of an event and diagnoses of the reasons for malfunctions reflect mental models.
People who are trained to perform tasks with a coherent account of those tasks complete them
better than people who are not trained with the model. And evidence is also obtained from
reaction times for eye movements and answering questions about processes (Allen, 1997).
Models of mental models may be termed conceptual models. Conceptual models
include: metaphor; surrogates; mapping, task-action grammars, and plans; and prepositional
knowledg.e (Allen, 1997).
Metaphor uses the similarity of one process with which a person is familiar to teach
that person about a different process. Metaphors are rarely a perfect match to the actual process
and incorrect generalizations from the metaphor can results in poor performance on the task
(Allen, 1997).
Surrogates are descriptions of the mechanisms underlying the process. For a pocket
calculator, surrogate models would describe its functions in terms of registers and stacks.
Surrogate models are not well suited to describing user-level interaction (Allen, 1997).
Another class of conceptual model describes the links between the task the users must
complete and the actions required to complete those tasks. Mappings are suitable for describing
learnability and as a basis for design (Allen, 1997).
Grammars are of interest because of their ability to describe systematic variations of
complex sequences (Allen, 1997).
Planning models can also integrate tasks and actions (Allen, 1997).
According to Allen (1997), Laird (1983) describes propositional knowledge is the basis
for most logical thinking (Allen, 1997).
Although mental models have been studies in physics and mathematics, the vast
majority of research on them has been based on computer-human interaction. Many aspects of
WAINESS PHD QUALIFYING EXAM
41
human-interaction with computers involve complex processes, thus people who interact with
computer systems must have some type of mental model of these processes (Allen, 1997).
The most important practical application of understanding students’ mental models is
for training (Allen, 1997).
Selection of appropriate text and graphics can aid the development of mental models.
Training materials may highlight text, or include diagrams or other techniques for improving the
learners’ mental models (Allen, 1997).
Scaffolding is the process of training a student on core concepts and then gradually
expanding the training (Allen, 1997).
Animation of data or scenarios which evolve over time should be especially useful for
developing mental models because the causal relationships in a process can be clearly illustrated
(Allen, 1997).
According to Allen (1997), Gonzales examined many properties of aniamtions and
found that factors such as the smoothness of the transitions were important for performance on
tasks which had been presented with the animations (Allen, 1997).
In user models (the computer’s model), the task expert has information about what the
user is trying to accomplish and possible strategies for accomplishing those goals. The situation
expert contributes knowledge about the environment in which the user is trying to complete the
task (Allen, 1997).
User models are often said to adapt to users. However, there are different senses in
which a model may be adaptive. In the simplest sense, a model is adaptive if it gives different
responses to different categories of users. A more interesting sense is that a model adapts as it
gains experience with the individual user (Allen, 1997).
Feedback uses output from the model to refine it (Allen, 1997).
The main criterion for the effectiveness of a user model is in predicting important
behavior which facilitates the user’s activities. Among components contributing to this are
relevance, accuracy, and generality, adaptability, ease of development and maintenance, and
utility (Allen, 1997).
Relevance requires that models make predictions that apply to the target behavior or
user goals. Accuracy requires that the modesl make correct predictions. Generality of the model
requires robustness despite changes in tasks, situations, and users. The model should be scalable.
Adaptability requires that the model be able to respond to changes in user behavior. Ease of
development and maintenance is whether the effort in maintaining the user model is worthwhile.
And utility means that the model should improve the user’s behavior (Allen, 1997).
Instructional interaction between a computer and a human being may be viewed as a
specialized conversation (Allen, 1997). Personalization of tutoring may be modeled by observing
the conversations between tutors and students (Allen, 1997).
Although there is no precise definition of worked examples, they share certain family
resemblances. As instructional devices, they typically include a problem statement and a
procedure for solving the problem; together, these are meant to show how other similar problsm
might be solved. In a sense, they provide an expert’s problem-solving model for the learner to
study and emulate (Atkinson, Derry, Renkl, & Wortham, 2000).
We hold that the explicit understanding of learning processes obtained through
controlled experimentation, including laboratory experimentation, is an important part of the
scientific knowledge base about teaching and learning, which, in turn, has had a significant
positive impact on instructional research and practice in classrooms. Transfer from laboratory to
WAINESS PHD QUALIFYING EXAM
42
classroom is possible because, while there are may differences between laboratory and classroom
environments, there are also many constants across setting in terms of students’ basic neural and
cognitive processes, as well as the structure of the interventions and materials investigates
(Atkinson, Derry, Renkl, & Wortham, 2000, 185).
The worked examples literature is particularly relevant to programs of instruction that
seek to promote skills acquisition, a goals of many workplace training environments as well as
instructional programs in domains such as music, chess, athletics, programming, and basic
mathematics (Atkinson, Derry, Renkl, & Wortham, 2000, p. 185).
Although the early research demonstrated that worked examples were instructionally
effective, our review suggests specific factors that moderate their effectiveness. These include
(1) intra-example feature, in other words, how the example is designed, particularly the way the
example’s solution is presented, (2) inter-example features, principally certain relationships
among multiple examples and practice problems within a lesson, and (3) individual differences
in example processing on the part of students, especially the way in which students “selfexplain” the examples (Atkinson, Derry, Renkl, & Wortham, 2000, p. 186).
According to Atkinson, Derry, Renkl, and Wortham (2000), instructional materials
requiring a student to split attention among multiple sources of information might impose a
heavy cognitive load. The imposition of a heavy cognitive load was thought to negate the
benefits of studying worked examples. Tramizi and Sweller (1988) labeled this phenomenon the
split-attention effect and hypothesized that it interfered with the student’s acquisition of schemas
representing the basic domain concepts and principles that students should learn from examples
(Atkinson, Derry, Renkl, & Wortham, 2000).
Summary of intra-example features. Examples should be constructed to maximally
integrate all sources of information—including diagrams, text, and aural presentation—into one
unified presentation, since splitting students’ attention across multiple, non-integrated
information sources may cause cognitive overload and impair learning. However, when an
example display is complex, simultaneous aural explanation must be accompanied by a method
for explicitly directing students’ attention to pertinent parts of the example as it is being
described or discussed. Otherwise, students will expend too much effort trying to locate those
parts of the example that the aural presentation is referencing, which creates cognitive overload.
In addition, because subgoal tasks within complex problems typically represent important
conceptual ideas that students need to learn, instructional effectiveness is enhanced when
examples clearly demarcate a problem’s subgoal structure, either by labeling each step or by
visually isolating steps in an example display (Atkinson, Derry, Renkl, & Wortham, 2000).
Research on explanation effects suggests that self-explanations are an important
learning activity during the study of worked examples. Unfortunately, the present research
suggests that most learners self-explain in a passive or superficial way. Among the successful
learners, there seem to different subgroups employing different self-explanations styles
(anticipative reasoning and principle-based explanations). Both of these styles can be fostered by
instructional methods. Direct training appears to be effective, as are structural manipulations of
examples as adding subgoal labels, utilizing an integrated format, or using “incomplete”
examples. Less promising are the data on improving self-explaining (and problem solving)
through setting social incentive to explain, such as inducing students to prepare to tutor others. In
particular, students who have no prior tutoring experience and who are novices within the
domain being tutored appear to experience stress and overload when asked to provide
instructional explanations (Atkinson, Derry, Renkl, & Wortham, 2000).
WAINESS PHD QUALIFYING EXAM
43
We postulate that learning from worked examples causes learners to develop
knowledge structures representing important, early foundations for understanding and using the
comain ideas that are illustrated and emphasized by the instructional examples provide. These
representations guide problem solving and may be conceptualized as representing early stages in
domain schema development and in the acquisition of expertise in accordance with Anderson’s
model of skills acquisition (Atkinson, Derry, Renkl, & Wortham, 2000, p. 202).
Through use and practice, these representations are expected to evolve over time to
produce the more sophisticated forms of knowledge that experts us. Even after expertise is
achieved, learners can benefit from study of examples representing the performance of other
experts (Atkinson, Derry, Renkl, & Wortham, 2000).
Worked-examples lessons will promote transfer if they include variability. This means
that examples within lessons should differ from each other in terms of numerical values and
form (Atkinson, Derry, Renkl, & Wortham, 2000, p. 204).
There is evidence that the structure of worked examples enhances students’ selfexplanation behavior. Moreover, there is evidence that students’ self-explanation behavior during
study in turn mediates learning. However, it has not been determined that the effects of example
structure on learning outcomes are fully mediated by self explanation (Atkinson, Derry, Renkl, &
Wortham, 2000).
In additional to example structure, situational factors, such as training and social
incentives, can foster self-explanation (Atkinson, Derry, Renkl, & Wortham, 2000).
Problems are often presented to students as cases, such as medical cases, and students
are guided by a tutor as they analyze cases and seek solutions, for example, diagnoses and
treatments (Atkinson, Derry, Renkl, & Wortham, 2000).
Critics to worked examples may claim that students exposed to worked examples are
not able to solve problems with solutions that deviate from those illustrated in the examples, can
not clearly recognize appropriate instanced in which procedures can be applied, and have
difficulty solving problems for which they have no worked examples (Atkinson, Derry, Renkl, &
Wortham, 2000).
The current view suggests, however, that examples can in fact help educators achieve
the goal of fostering adaptive, flexible transfer among learners (Atkinson, Derry, Renkl, &
Wortham, 2000).
Worked-out examples typically consist of a problem formulation, solution steps, and
the final answer itself (Atkinson, Renkl, & Merrill, 2003).
Research indicates that exposure to worked-out examples is critical when learners are
in the initial stages of learning a new cognitive skill in well structured domains such as
mathematics, physics, and computer programming (Atkinson, Renkl, & Merrill, 2003).
Although worked-out examples have significant advantages, their use as a learning
methodology does not, of course, guarantee effective learning (Atkinson, Renkl, & Merrill,
2003).
According to Atkinson Renkl, and Merrill (2003), at first Chi and her colleagues (Chi
et al., 1989) postulated that the self-explanation effect principally involved inference generation
on the part of a learner. That is, by self-explaining, the learner is inferring information that is
missing from a text passage or an example’s solution. However, because of some inconsistencies
among this view and some of the findings in the self-explanation literuature, Chi (2000) revised
this initial view by suggesting that the self-explanation effect is actually a dual process, one that
involves generating inferences and repairing the learner’s own mental model. In the latter
WAINESS PHD QUALIFYING EXAM
44
process, it is assumed that the learner engages in the self-explanation process if he or she
perceives a divergence between his or her own mental representation and the mdoel conveyed by
the text passage or example’s solution. According to Chi, this new viewpoint extends the
inference generation by suggesting that “each student may hold a naïve model that may be
unique in some ways, so that each student is really customizing his or her self-explanation to his
or her own mental model” (p. 196; as cited in Atkinson, Renkl, & Merrill, 2003).
According to Atkinson, Renkl, and Merrill (2003), our findings on the usefulness of a
learning environment that combines fading worked-out steps with self-explanation prompts
support the basic tenets of one of the most predominant, contemporary instructional models,
namely the cognitive apprenticeships approach (Collins, Brown, & Newman, 1989). This
approach suggests that learners should work on problems with close scaffolding provided by a
mentor or instructor (Atkinson, Renkl, & Merrill, 2003).
According to Atkinson, Renkl, and Merrill (2003), this approach is characteristic of
Vygotsky’s (1978) “zone of proximal development” in which problems or tasks are provided to
learners that are slightly more challenging than they can handle on their own. Instead, solving the
problems or tasks independently, the learner must rely—at least initially—on the assistance of
their more capable peers and/or instructors to succeed. According to this approach, learners will
eventually make a smooth transition from relying on modeling to scaffolding problem solving to
independent problem solving (Atkinson, Renkl, & Merrill, 2003). In other words, this model
advocates the fading of instructional scaffolding during this transition. Correspondingly, our
partially worked-out examples provide a scaffold that permits learners to solve problems they
could not successfully solve on their own. The instructional scaffolding—in the shape of
worked-out solution steps—is gradually faded in our learning environment (Atkinson, Renkl, &
Merrill, 2003).
Learners are encouraged to reflect on their problem-solving process and to try to
identify ways of improving it. For instance, they are encouraged to reflect on the problems that
they have missed or to try to explain how to generate the correct solution, a process that can
increase the likelihood that the correct solution procedure will be internalized by the learner
(Atkinson, Renkl, & Merrill, 2003).
Overall, the use of prompts that encourage the learners to figure out the principle that
underlies a certain solution step can be recommended for several reasons, including the
following: (a) it produces medium to high effects on transfer performance, (b) these effects are
consistent across different age levels (university and high school), (c) it does not interfere with
fading, (d) it is very easy to implement (even without the help of computer technology), and (e)
it requires no additional instructional time. This prompting procedure is, however, not without its
drawbacks. Because this procedure is designed to elicit principle-based explanations, it is ideally
suited for well-structured domains such as mathematics and physics that contain clearly
identifiable domain principles “under” each solution step (Atkinson, Renkl, & Merrill, 2003).
As one can imagine, not all domains contain such clearly identifiable principles.
Hence, it is worth noting that our prompting procedure can nly be applied in an unmodified
manner when each solution step can be explained by a principle within a domain (Atkinson,
Renkl, & Merrill, 2003).
New technologies, such as the use of multimedia, can afford rich opportunities for
constructivist approaches in the field of education (Bailey, 1996).
Simply put, constructivism is learning by assembling meaning from pieces of reality
(D’Ignazio, 1992; as cited in Bailey, 1996).
WAINESS PHD QUALIFYING EXAM
45
Learning which is active becomes a reality as the learner is not a passive
nonparticipant who easily ignores or forgets the encounter. The initial “piece of reality” is
participation in the process. Constructivists then, advocate student-centered learning which is
self-directed, which has personal relevance to the learner, and which is manifested by a form of
active demonstration (Bailey, 1996).
Self-directed learning is more likely to have personal relevance, and as new technology
is assimilated with personalized associations, meaning and retention are increased (Bailey,
1996).
When using technology, initially, the learner must focus on the acquisition of skills and
knowledge related to learning the technology. However, once these are mastered and acclimated,
they may become—much like writing, typing, or keyboarding—tools for conveying information
(Bailey, 1996). The difference in this tool is that so many differing modes of communication are
possible in the context of the technology of multimedia, and deciding which is appropriate is in
itself a higher thinking decision (Bailey, 1996).
O’Neil’s Problem Solving Model
O’Neil’s Problem Solving model (O’Neil, 1999; see figure 2 below), is based on Mayer
and Wittrock’s (1996) conceptualization: “Problem solving is cognitive processing directed at
achieving a goal when no solution method is obvious to the problem solver” (p. 47). This
definition is further analyzed into components suggested by the expertise literature: content
understanding or domain knowledge, domain-specific problem-solving strategies, and selfregulation (see, e.g., O’Neil, 1999, in press). Self-regulation is composed of metacognition
(planning and self-checking) and motivation (effort and self-efficacy). Thus, in the specifications
for the construct of problem solving, to be a successful problem solver, one must know
something (content knowledge), possess intellectual tricks (problem-solving strategies), be able
to plan and monitor one’s progress towards solving the problem (metacognition), and be
motivated to perform (effort and self-efficacy; O’Neil, 1999, pp. 255-256). Each of these
problem-solving elements would have to be taught and assessed in the game context.
Fig 2. O’Neil’s Problem Solving Model
Problem Solving
Content
Understanding
Problem-Solving
Strategies
Self-Regulation
Metacognition
Planning
Domain
Specific
Domain
Independent
SelfMonitoring
Motivation
Effort
SelfEfficacy
WAINESS PHD QUALIFYING EXAM
46
Baker and Mayer’s CRESST Model of Learning
The CRESST model of learning (Baker & Mayer, 1999) links the components
required to assess problem solving in technology environments. The model is composed of six
families of cognitive demands: five families—content understanding, collaboration, problem
solving, communication and self-regulation—all radiating from learning, the sixth family of
cognitive demands. The shift from unidimensional measure of a construct to multidimensional
domains is rooted in the work of Glaser (1963), Hively, Patterson, and Paige (1968), and Baker
and Popham (1973). In the CRESST model, “each family consists of a task that can be used as a
skeleton for the design of instruction and testing” (Baker & Mayer, 1999, p. 275). For example,
understanding consists involve explanation, which in turn involves a variety of actions such as
having students read opposing views, invoking prior knowledge, and organizing and writing a
valid explanation. This task framework supports many different learning domains, such as
history or science. For problem solving, the task is instantiated in different domains so that a set
of structurally similar models for thinking about problem solving is applied in science,
mathematics, or social studies. In each domain, there is a need to identify the problem,
understand content, understand key principles, and fit solutions to constraints. The six families of
the CRESST model support all forms of learning.
Fig 3. Baker & Mayer’s CRESST model of learning: Families of cognitive demands
CRESST model of learning
Content
Understanding
Collaboration
Communication
Learning
Problem Solving
Self-Regulation
When the goal of instruction is meaningful learning (or student understanding),
assessments of problem-solving transfer are called for (Baker & Mayer, 1999).
Assessments that focus solely on the quantitative issue of how much was learned are
based on a view of learning as knowledge acquisition, i.e., that learning involves adding pieces
of information to one’s memory. In contrast, assessments that also focus on the qualitative issue
of how knowledge is structured and used by the learner are based on the view of learning as
knowledge construction, i.e., that learning involves making sense out of presented material by
building a mental model (Baker & Mayer, 1999).
Problem solving is cognitive processing directed at transforming a given situation into
a desired situation when no obvious method of solution is available to the problem solver
(Mayer, 1990 as cited in Baker & Mayer, 1999).
WAINESS PHD QUALIFYING EXAM
47
A problem exists when a problem solver has a goal but does not know how to reach it,
so problem solving is mental activity aimed at finding a solution to a problem (Baker & Mayer,
1999).
A promising direct approach to knowledge representation, more parsimonious than a
typical performance assessment, is knowledge or concept mapping, in which the learner
constructs a network consisting of nodes (e.g., key words or terms) and links (e.g., “is part of”,
“lead to,” “is an example of”) (Baker & Mayer, 1999, p. 274).
In problem solving, the skeletal structures are instantiated in content domains, so that a
set of structurally similar models for thinking about problem solving is applied to science,
mathematics, and social studies. These models may vary in the explicitness of problem
representations, the guidance about strategy (if any), the demands of prior knowledge, the focus
on correct procedure, the focus on convergent or divergent responses, and so on (Baker &
Mayer, 1999).
In each academic area, there is the need to identify the problem, the need to understand
content provided and omitted, the need to understand the key principle(s) at work, and the need
to fit solutions to constraints (Baker & Mayer, 1999).
Domain-specific aspects of problem solving (the part that is unique to geometry,
geology, or genealogy) involve the specific content knowledge, the specific procedural
knowledge in the domain, any domain-specific cognitive strategies (e.g., geometric proof, test
and fix), and domain specific discourse (O’Neil, 1998, as cited in Baker & Mayer, 1999).
Cognitive complexity is a concept at the heart of problem solving. It minimally requires
that students must process material beyond the recognition or recall level. Typically, cognitive
complex tasks have either implicit or explicit multiple steps by which the test taker must proceed
to develop an adequate solution (Baker & Mayer, 1999).
Two trends in technology are certain: the cost of computer technology will continue to
drop, and technology of all sorts will become easier to use (Baker & O’Neil, 2002).
Dreary intellectually, predictable pedagogically, despite cuter, more active graphics,
our learning systems will need massive rethinking to make them useful for the challenges facing
instruction for both children and adults (Baker & O’Neil, 2002, p. 611).
One key to their ultimate utility will be the degree to which technology can be used
simultaneously to teach and to measure better, more deeply and speedily, the complex tasks and
propensities needed for learners to achieve and to continue to learn in a rapidly changing society
(Baker & O’Neil, 2002, p. 611).
Problem solving is a family of cognitive demands that can be required in may subject
areas. The term problem solving goes far beyond the application of algorithms (e.g., subtraction
rules) to simple tasks (Baker & O’Neil, 2002).
Our definition of problem solving (Baker & Mayer, 1999; O’Neil, 1999) is already an
important component of educational reform efforts designed to raise the expertise of students
(Baker & O’Neil, 2002).
Problem-solving tasks can take a third form, dealing with simulations and problem for
which there is not a known solution, but which present, like the first case, a rapidly changing
scenario, for instance, with chance or the probability of existing faults occurring as “surprises”
during the examination sequence (Baker & O’Neil, 2002). Here, the intellectual task for the
learner varies and includes assimilation and incorporation of useful strategies and a running
internal record of the degree to which any combination of procedures or actions is likely to
optimize the outcome (Baker & O’Neil, 2002).
WAINESS PHD QUALIFYING EXAM
48
Not only can problems be obscured or embedded in distracting settings, or presented in
complex language, problems can also be provided sequentially to learners in a computerized
setting. Solving the first part of a task in a particular way may lead to a conditional
representation of the second part of the task. Contingent tasks may, on the other hand,
approximate reality, for there are consequences of correct and incorrect paths (Baker & O’Neil,
2002).
Both domain-independent and domain-dependent knowledge are usually essential for
problem solving. Domain-dependent analyses focus on the subject matter as the source of all
needed information (Baker & O’Neil, 2002).
Domain-independent analyses are those that attempt to capture the general strategies
that are in use across subject matters. These approaches should engender not general notions of
intelligence testing, but rather the attributes of performance that could be expected to transfer
over a wide task domain (Baker & O’Neil, 2002).
The use of visual materials to complement regular instruction has become a common
instructional technique at all levels of education. However, its integration into the instructional
environment has not realized its promise of increased effectiveness and efficiency in terms of
optimizing student learning. Analyses from existing, visual related research has failed to provide
generic guidelines for the integration of visualization to improve learning (Baker & Dwyer,
2000).
Learning will be more complete as the number of cures in the learning environment
increases. An increase in realism in the existing cues increased the probability that learning will
be facilitated (Baker & Dwyer, 2000).
In a meta-analysis of eight studies involving 2000 college and high schools students,
Baker and Dwyer (2000) found argued that an overall effect size of .71 demonstrated the general
positive effect that visualization can have in facilitating student achievement. However, (a) the
realism continuum is not an accurate predictor of instructional effectiveness, (b) not all types of
visuals are equally effective in facilitating achievement of different educational objectives, (c)
color can be an important instructional variable in facilitating achievement of specific types of
educational objectives, and (d) the type of visualization most effective for facilitating different
educational objectives may be dependent on the method of presentation (Baker & Dwyer, 2000).
Visuals which contain the essence of the message to be transmitted should be more
effective in facilitating achievement than the more realistic illustrations which have to be coded
by the central nervous system before being transmitted (Baker & Dwyer, 2000).
The effectiveness of the simple line presentations (color) may have resulted because
the use of color made the visuals more attractive and students attended to them more vigilantly
(Baker & Dwyer, 2000). This explanation is suggested since the only different between the b@w
and color treatments was that the color version consistedn fo blue lines on a pink background
rather than black on white and provided no additional information (Baker & Dwyer, 2000).
The effectiveness of the detailed shaded drawing presentation may have resulted
because the realistic detail in the visuals was accentuated by color enabling the students to
identify and interact with the relevant characteristics (Baker & Dwyer, 2000).
While attempting to focus on a mental activity, most of us, at one time or another, have
had our attention drawn by extraneous sound (Banbury, Macken, Tremblay, & Jones, 2001).
Sounds often seem to intrude on our awareness, without our invitation or, apparently,
control (Banbury, Macken, Tremblay, & Jones, 2001).
WAINESS PHD QUALIFYING EXAM
49
Evidently, these are instances in which our capacity to focus, to attend selectively to
thoughts or events, suffers some kind of breakdown (Banbury, Macken, Tremblay, & Jones,
2001).
There may be occasions when the system designer may wish to capture the attention of
the person, and knowledge of auditory distraction can be put to good use in the design of
auditory warnings and alarms (Banbury, Macken, Tremblay, & Jones, 2001).
In the present paper, we review a range of recent studies that focus on establishing the
conditions under which a person may be distracted while undertaking a relatively complex
mental task (Banbury, Macken, Tremblay, & Jones, 2001). Generically, these are know as
irrelevant sound studies (Banbury, Macken, Tremblay, & Jones, 2001).
Because hearing is omnidirectional and has the capacity to receive information at all
times, even in darkness or during sleep, it has been dubbed “the sentinel of the senses” (Banbury,
Macken, Tremblay, & Jones, 2001, p. 13).
Clearly, even when our attention is fastened to one activity, the brain is registering a
range of other events; otherwise, how do we manage to switch attention between sources of
information so purposefully and so adroitly (Banbury, Macken, Tremblay, & Jones, 2001)?
The general procedure for the irrelevant sound paradigm is straightforward. The
participant undertakes a short-term memory task involving recall of the order of a sequence of
verbal items (usually visually presented). While the task is being undertaken, irrelevant sound is
played, either narrative speech or isolated events at about one item per second. The participants
are told to ignore any sound they hear and are reassured they will never be required to report any
feature of it (Banbury, Macken, Tremblay, & Jones, 2001).
Because the memory task and irrelevant event are presented in different sensory
modalities, the effect cannot be attributed to some kind of interference (or masking) at the
sensory level. Instead, the disruption must be attributable to a confluence of processing from the
eye and the ear at some level beyond the sensory organs. This can be described as a breakdown
in attentional selectivity. Despite the intent of the person to concentrate on the memory task, the
irrelevant sound intrudes (Banbury, Macken, Tremblay, & Jones, 2001).
One explanation of why the sound is heard is the disruption is based on a conflict of
content between what is seen and what is heard. The other explanation is that interference arises
between concurrent common processes (specifically, the degree to which the two activities draw
on the ordering of material in the brain). This latter account, the changing state hypothesis, is
part of a more general model of working memory based on a blackboard analogy called the
object-oriented episodic record (Banbury, Macken, Tremblay, & Jones, 2001).
The irrelevant sound effect can be explained by supposing that interference results
form a conflict based on similarity of process between relevant and irrelevant sequences, not
similarity of content (Banbury, Macken, Tremblay, & Jones, 2001).
The disruptive effect of irrelevant sound on performance is independent of the level of
sound (the volume) (Banbury, Macken, Tremblay, & Jones, 2001).
Above three voices, the disruption is a decreasing function of the number of voices.
This effect is readily understood in terms of the masking of one sound by another. When the
sound contains a relatively large number of voices, words are no longer individually
distinguishable. In particular, there is evidence that changes in energy at the boundary of the
sounds are important in determining the degree of disruption (Banbury, Macken, Tremblay, &
Jones, 2001).
WAINESS PHD QUALIFYING EXAM
50
The paper page with orderly rows and alphanumeric symbols, and occasionally
images, is no longer the only nor, in many cases, even the dominant resource for contemporary
readers (Bangert-Drowns & Pyke, 2001, p. 214). Electronic media are increasing a preferred
means of information and entertainment (Bangert-Drowns & Pyke, 2001, p. 214).
In general terms, texts are any relatively permanent structures for the storage,
organization, and accessibility of a coherent body of information (Bangert-Drowns & Pyke,
2001). Electronic texts are information structures stored by and accessible through nonprint,
electronic media (Bangert-Drowns & Pyke, 2001).
Bangert-Drowns and Pyke (2001) developed a taxonomy of student engagement with
interactive computer media for text intrepretation. The taxonomy consisted of seven levels
ranging from a high of literate thinking to a low of disengagement. According to the authors,
disengagement occurred when “navigational and operational competence or interest is so
lacking, the student declines purposeful interaction” (Bangert-Drowns & Pyke, 2001, p. 226).
Also according the to authors, “the taxonomy’s ‘higher’ levels presuppose navigational and
operational competence” (Bangert-Drowns & Pyke, 2001, p. 233). Referring to Corno’s analysis
of volition in learning, Bangert-Drowns and Pyke (2001) argued that their higher taxonomic
levels reflect increasing capacity to employ metacognitive strategies to monitor progress toward
goals. Volitional capacities, strategic prioritization of goals, and perseverance in pursuit of
personal interests, appear clearly in self-regulated interest (Bangert-Drowns & Pyke, 2001).
In contrast to more traditional technologies that simply “deliver” information, current
computerized learning environments offer greater opportunities for interactivity and learner
control (Barab, Young, & Wang, 1999).
Nodes refer to the information units being displayed (e.g., paragraphs of text, pictures,
sets of questions), while links refer to the connections among nodes (Barab, Young, & Wang,
1999)..
Hypertext programs may simply offer sequencing and pace control, or they can allow
the learner to decide which, and in what order, information was be accessed (Barab, Young, &
Wang, 1999).
Learners are able to make navigational choices by activating clickable areas, allowing
them to jump from one location to another (Barab, Young, & Wang, 1999).
Increased affect does not necessarily lead to increased learning (Barab, Young, &
Wang, 1999).
In spite of the intuitive and theoretical appeal of hypertext environments, empirical
findings yield mixed results with respect to the learning benefits of learner control over program
control of instruction (Niemiec et al., 1996, Steinberg, 1989, as cited in Barab, Young, & Wang,
1999).
It appears that learner control wields a double-edged sword; for some users, it can
extend their intellectual performance, while for others, it may not facilitate performance—
possibly even leaving the user lost in a maze of information (Barab, Young, & Wang, 1999).
In generative learning (see Wittrock’s generative model in Whittrock, 1974, 1978),
learners are not passively receiving learning, but are actively engaged in the construction of
meaning as it relates to their beliefs, experiences, current goals, and the context in which learning
is occurring (Barab, Young, & Wang, 1999).
According to the results of a study by Barab, Young, and Wang (1999), increased
levels of learner control are beneficial when students are using a hypertext program to solve a
specific problem. In their study, university students were free to navigate directly to those nodes
WAINESS PHD QUALIFYING EXAM
51
of information they deemed appropriate. In the first of two experiments, which involved
problem-solving, those students did significantly better at the problem-solving task than those
who proceeded through the document in a linear manner (Barab, Young, & Wang, 1999). In the
second study, which involved reading comprehension, there were no differences between groups
(Barab, Young, & Wang, 1999).
The amount of learner control, from the perspective of the learning, afforded by a
hypertext system should not be assumed as high simply because the instructional designer creatd
the opportunity to visit links. Rather, it must be though of as a construct that is codetermined by
the learner’s perceptions of the affordances of the hypertext in relation to his/her particular goals
(Barab, Young, & Wang, 1999).
It appears that various goals and their inherent constaints (i.e., the goal paths they
establish) will affect both the process and product of learning (Barab, Young, & Wang, 1999).
In a four-factor MANOVA design, this exploratory study experimentally investigated
the influence of navigation mode (linear, nonlinear), distracting links (presence, absence),
sensation-seeking tendency (high, low) and spatial-synthetic ability (high low) on perceived
disorientation and incidental learning (accuracy of main point, example generation) in web
navigation (Baylor, 2001).
Incidental learning is conceptualized in two ways in this study: (a) from the macro
level of text processing, as one’s effectiveness at figuratively “getting the gist” of the website
content and developing a schematic mental representation to determine the main point; and (b)
from the micro level of text processing, as one’s effectiveness at generating and recalling
examples from the content. The distinctions between macro and micro levels of processing are
made for the purpose of describing this study (Baylor, 2001).
Disorientation is defined her as a user’s perception of his/her uncertainty of location
(Baylor, 2001).
While the implementation of a three-dimensional spatial environment is technically
feasible and would solve some disorientation problems for the learner, the use of such an
environment with its visualization facilitation may provide the learner with too much information
about locating information without letting the user discern the structure and meaning of the
information (Baylor, 2001).
Disorientation is a problem in terms of learning in open-ended learning environments
of both the navigational issue from the user’s perspective and also the external geography of the
website (Baylor, 2001).
In terms of navigation mode, two contrasting instantiations are linear or nonlinear. In a
linear navigation mode, the user moves through the website sequentially and is only allowed to
move forward or backward through the content; thus, the sequence of web pages is controlled by
the website. A nonlinear navigation mode is where the user has options to “jump” to any location
within the website at any time, providing more flexibility and control for the user (Baylor, 2001).
And advantage of a nonlinear navigation mode (typical of hypertext-based systems) is
that a learner could navigate in a personally meaningful way to access information (Baylor,
2001).
A disadvantage of a nonlinear navigation mode is it may not have the coherence that
would be provided when the learner is forced to process the information in a more systematic
way (from beginning to end). Specifically, in a nonlinear mode, the learner may not be able to
determine how the overall content is globally represented (Baylor, 2001).
WAINESS PHD QUALIFYING EXAM
52
In traditional forms of navigation, one must determine spatial position in relation to
landmarks or astral location to decide on the means of moving toward a goal. In a virtual world,
the feeling of being lost while navigating may result from a lack o connection among the
physical representations of the world. This suggests the need for some sort of mapping or
landmarking to serve as cues (Baylor, 2001).
The presence of distracters (extraneous and seductive details) had a negative effect on
both example generation and understanding the main point of the content (Baylor, 2001).
Disorientation was negatively correlated with the learners ability to generate examples
and to define the main point of the content (Baylor, 2001).
A moderately high effect size indicated that participants (average 30 years old,
predominantly white, with 56% males) reported more disorientation with the linear navigation
mode as compared to the nonlinear navigation mode. This indicates that users are more used to
and more comfortable with the nonlinear format of websites than when forced to navigate in a
linear configuration (Baylor, 2001).
Interestingly, the expected role of prior knowledge in facilitating orientation was not
supported. Participants in the linear mode had marginally more prior knowledge than those in the
nonlinear mode, yet the linear mode exhibited the higher level of disorientation. Therefore,
navigation mode may be a greater factor than prior knowledge in influencing orientation (Baylor,
2001).
The presence of distracters (extraneous and seductive details) negatively affected
example generation scores (Baylor, 2001).
That the pupil of the eye dilates during mental activity has long been known in
neurophysiology (Beatty, 1982). Only recently has this phenomenon been used as a tool in
investigating human cognitive processing (Beatty, 1982). Dilations occur at short latencies (100
to 100 msec) following the onset of processing and subside quickly once processing is
terminated. Perhaps, more important, the magnitude of papillary dilation appears to be a function
of processing load or “mental effort” required to perform the cognitive task (Beatty, 1982).
Pupillary dilations related to cognitive load occur both during the processing of new
information in working memory (e.g. hearing and repeating a series of numbers) and retrieval of
existing knowledge from long-term memory (e.g., recalling a series of number; Beatty, 1982).
Rehearsal strategies that improve performance on a working memory task act to reduce
the amplitude of the task-evoked papillary response (Beatty, 1982).
Selective attentional processing of sensory information occurs under conditions of high
information load when it is not possible to process adequately all incoming information (Beatty,
1982).
An iconic interface uses images to represent actions and objects that can be invoked or
manipulated by a user. There are a variety of icon types which convey meaning in different
ways. For example, representational icons are meant to represent actual physical objects and to
inherit the properties of those objects, while abstract icons are meant to convey abstract concepts
such as fragility (Benbasat & Todd, 1993). Representational icons are the most common type of
icon employed in computer interfaces (Benbasat & Todd, 1993).
When interfacing with a computer, a user is typically focused on some “primary
cognitive task” which may relate to problem-solving, analyzing, reading, or writing. Attention
devoted to the interface may interfere with the primary task. Since text-based processing is
associated with cognition, more interference will result between a cognitive task and a text-based
interface which demands the use of the same cognitive resources (Benbasat & Todd, 1993).
WAINESS PHD QUALIFYING EXAM
53
The less effort required to use the interface, the more likely it is that the primary task
will be successfully completed. If the iconic interface draws on a perceptual resource pool and
the primary task draws on a cognitive pool, than overall performance will improve (Benbasat &
Todd, 1993).
In evaluating the advantages of iconic interfaces, it is important to remember that there
is a difference between advantages attributable to some inherent property of icons and those that
are attributable to specific implementations (Benbasat & Todd, 1993). It is often argues that
iconic interfaces will be easier to use because they represent a collection of familiar objects; thus,
interference from the icons to system functions will be facilitated. While this may be true, the
advantage likely comes not from the icons per se, but from an implementation which permits
users to employ metaphors by which to map known attributes of familiar objects (Benbasat &
Todd, 1993).
According to the authors, the true advantage of icons may come from the fact that
visual cues can be processed more rapidly than text-based cues, and that an icon may carry more
information than a text-based cue (Benbasat & Todd, 1993).
Because the icons are present on screen, syntax error are eliminated (i.e., the syntax is
predefined; Benbasat & Todd, 1993).
It is argued that by facilitating the use of metaphor, iconic systems lead to significant
advantages. While it is true that most iconic interfaces rely on metaphors, such as the “desk top”
or “office” metaphor, this is a design and implementation issue, not an icon issue. Though it may
be less compelling, there is no reason why text-based cues could not be employed in lieu of icons
to represent such things as folders and documents (Benbasat & Todd, 1993).
Another advantage of icons is that common characteristics can be carried across
applications through consistent application of icons (such as a “quit” icon). However, this is a
feature or advantage that is not unique to iconic interfaces but rather is a property of good design
(Benbasat & Todd, 1993).
The ability to represent objects rather than abstract concepts through icons is another
claimed advantage. However, there is no real reason to believe that icons are the only way to
represent objects in system interfaces. Yet, it is possible that icons provide a superior way of
representing objects. This, however, is yet to be determined (Benbasat & Todd, 1993).
The disadvantages of iconic interfaces arise primarily from difficulties in
implementation rather than from any inherent properties of icons. For example, it is difficult to
design icons to convey the desired meaning without invoking other connotations. The
interpretation of a user and the intent of the designer may be different. When this happens,
problems arise and semantic errors occur. Such ambiguity in meaning arises because there is no
universal set of icons or principles to guide icon design (Benbasat & Todd, 1993).
Broadly defined, direct manipulation interfaces incorporate the concept of physical
manipulation of a system of interrelated objects which are analogous to objects found in the “real
world” (Schieiderman, 1983, as cited in Benbasat & Todd, 1993). Object representations may
take on a variety of forms. However, they are most commonly represented as icons although it is
possible to provide text-based implementations of the objects or a combined text-icon
presentations (Benbasat & Todd, 1993).
According to Benbasat and Todd (1993), Hutchins, Hollan, and Norman (1986),
developed a model to explain the effects of a direct manipulation interface. They claim that the
directness” of an interface results from the commitment of fewer cognitive resources in order to
complete a given task. Cognitive effort is minimized if the system interface maps directly into
WAINESS PHD QUALIFYING EXAM
54
the user’s view or mental model of a specific task. Directness is argued to be a function of two
factors: the first is distance, which must be minimized, and the second is engagement, which
must be maximized. Distance refers to the notion of the gap between the user’s thoughts and the
way they can be accomplished. Engagement relates to the degree of involvement the user
experiences with the system. Under conditions of high engagement or involvement, the sytem
interface becomes transparent and the user has the perception of working with the actual objects
of interest, rather than through an abstract computer system (Benbasat & Todd, 1993).
There are several advantages to direct manipulation. First, novices can learn basic
functionality quickly, because the system incorporates a model of the task as held by the user
(Benbasat & Todd, 1993). Second, experts with both the system and the task and/or task domain
can work extremely rapidly (Benbasat & Todd, 1993). Third, intermittent or casual users can
retain operational concepts. Casual users may have to go through a learning period each time
they use an application. With a direct manipulation interface, such relearning will be reduced
since the interface maps into the user’s model of the task. Fourth, users can immediately see if
their actions are furthering their goals (Benbasat & Todd, 1993). It should be noted, however,
that none of these advantages are unique to direct manipulation (Benbasat & Todd, 1993).
There are also a number of advantages of direct manipulation. Some are inherent
disadvantages and some are disadvantages in implementation (Benbasat & Todd, 1993).
In terms of inherent disadvantages, first is model specificity. Direct manipulation
interfaces gain much of their power from the development of specific models which the user can
understand and apply. Such specific models may sacrifice flexibility. Users may be required to
learn many specialized systems rather than fewer generalized ones. Second is constraint of the
solution space. The success of direct manipulation systems depends on their ability to capture a
user’s model faithfully. As a result, it is unlikely that they will lead to new ways to think about
problems. Rather, the interface will reinforce current thinking; thus, discouraging innovative
solutions. And third is repetitive operations. Repetition, or looping, of functions can be tedious to
perform in a direct manipulation environment (Benbasat & Todd, 1993).
In terms of implementation disadvantages, first is the question of whose model of
interaction is to be built into the system. To design a general interface, one assumes that that is a
prototypical user’s model to draw upon. This may not be the case. The second disadvantage is
precision in manipulation. In a direct manipulation system, invoking a command requires precise
manipation by the user. Virtually every user of a mouse-driven interface has experience the
frustration of attempting detailed work on a screen and having incorrect objects activated
(Benbasat & Todd, 1993).
In a study involving 48 university students (27 males and 21 females), little or no
advantages were found for icons. However, the author did state that these results may not
generalize to other applications such as games (Benbasat & Todd, 1993).
The general study of human-machine interaction began in WWII with a focus on
understanding the psychology of soldiers interacting with weapon and information systems.
After the war, human-machine interaction began to examine more broadly the relationship of
work and computer product environments. Human-computer interaction (HCI) developed from
this work and is a multi-disciplinary field involving computer science, psychology, engineering,
ergonomics, sociology, anthropology, and design. HCI is concerned with the design, evaluation,
and implementation of interactive computing systems for human use (Berg, 2000).
HCI is generally used to mean human-computer interaction, but sometimes is
described as human-computer interface or man-machine interface (MMI; Berg, 2000).
WAINESS PHD QUALIFYING EXAM
55
The literature on HCI focuses in part on cognitive processes (mental processes),
especially in terms of the capacities of users and how these affect users’ abilities to carry out
specific tasks with computer systems (Berg, 2000). The cognitive aspects of HCI include motor,
perceptual, and cognitive systems, as well as two types of memory: working and long-term
(Berg, 2000).
Usability refers to the degree to which a computer is effectively used by its users in the
performance of tasks (Berg, 2000). According to Berg (2000), designing for experienced users is
difficult, but designing for a broad audience of unskilled users presents a far greater challenge
(Berg, 2000).
Interface design is a subset of HCI and focuses specifically on the computer input and
output devices such as the screen, keyboard, and mouse, and has its roots in the ergonomic study
of instrument panels during WWII (Berg, 2000).
In addition to visual interface issues, the HCI literature also touches on topic related to
visual perception and how the specifics of human visual perception may impact human-computer
interaction (Berg, 2000).
Interface metaphors are often discussed in HCI literature as they pertain to interface
design. Interface metaphors work by exploiting previous user knowledge of a mental model
(Berg, 2000).
Research suggests that metaphors stand in the way of making new connections and
associations and that, while similar representations creatd by metaphors can be useful, they can
also be detrimental to user behavior under specific conditions, particularly if the metaphor does
not fie appropritately (Berg, 2000).
The term animation is used to describe movements of either text or graphics on the
computer screen (Berg, 2000).
Agents are active and ever-present software components that perceive, appear to
reason, act, and communicate. They are also referred to as guides and personal assistants (Berg,
2000).
For the past 50 years, computers have had a profound effect on humans and have
advanced our lives in immeasurable ways (Chalmers, 2000). According to Hokanson and
Hooper (2000), computers were predicted to improve both teaching and student achievement.
Students would learn more through computers: test scores would rise, students would remember
more, and they would learn at a faster rate. Computer-assisted education would help students
prepare to compete in a modern, global workforce. Despite continued optimism, we now find
criticisms and concerns being raised regularly. Principal among the complaints is the failure to
find an improvement in learner’s performance (Hokanson & Hooper, 2000).
Two of the most significant developments during the 1980s in the domain of humancomputer interaction (HCI)—direct manipulation (DM) and graphical user interfaces (GUI)—
combine to form direct manipulation interfaces (DMIs). These two innovations were introduced
and proposed, hand in hand, as vehicles to user-friendly design promoting ease of use and
improved task performance (Kaber, Riley, & Tan, 2002). Hypertext and hypermedia make wide
use of the graphical user interface (GUI), which operates on the metaphorical premise of direct
manipulation and engagement by the users (Brown & Schneider, 1992). Hypermedia relies
heavily on the use of windows, icons, menus, and pointer systems (Brown & Schneider, 1992).
In addition, interface metaphors such as the Microsoft Windows “desktop” metaphor, are
widely used. Some of the newest metaphor developments can be found in interfaces created for
presenting information structures, multimedia, group work, and virtual reality (Neale & Carroll,
WAINESS PHD QUALIFYING EXAM
56
1997, p. 442). Along with the proposed benefits of metaphors, the way in which the user
interacts with the computer environment has also been suggested to influence performance.
Allen (1997) has suggested that interaction between a computer and a human being may be
viewed as a specialized conversation. According to de Jong, de Hoog, and de Vries (1993), one
way of interacting with a computer, where objects can be manipulated directly, is found in the
so-called direct manipulation metaphor. In the early 1980s, Shneiderman coined the term direct
manipulation along with its key concepts. With direct manipulation, objects on the screen are
representations of real world objects, and interactions with that world, in the simplest form, are
manipulated through clicking and dragging with a mouse. Because there is a minimum of
syntactic knowledge required by the user, he can concentrate fully on the semantics of the
objects and the actions of the task (de Jong et al., 1993). While this is expected to lead to
improved performance (Benbasat & Todd, 1993), it has not always been the case.
Review of the Literature
This review is divided into three parts. First is a discussion of mental models and their
relationship to computer interfaces. In particular, the second part of the discussion centers on
metaphors, a specific type of mental model which is used in modern computer interfaces. The
last section is a discussion of the three interface types, which are defined by their mode of
interaction: command, menu, and direct manipulation. In this third section, research results are
discussed along with conflicting findings and possible causes for those conflicts.
Mental Models
Models are approximations of objects or processes which maintain some essential
aspects of their original form (Allen, 1997), and mental models are usually considered the way in
which people model processes. This emphasis on process distinguishes mental models from
other types of cognitive organizers such as schemas (Allen, 1997). The majority of the research
on mental models has been with studies of computer-human interaction. Many aspects of humaninteraction with computers involve complex processes; therefore people who interact with
computer systems must enlist a mental model for those processes (Allen, 1997). According to
Eberts and Brittianda (1993), the user forms a mental model of how the computer system or
program works, which then guides the user’s actions and behaviors. The mental model can be
thought of as the user’s understanding of the relationships between the input and output of the
computer so the user can predict the output that would be produced by possible inputs (Eberts &
Brittianda, 1997).
In addition to mental models, other models relevant to human-computer interaction
include user models and data models. “The expectations a user has about a computer’s behavior
come from mental models, while the ‘expectations’ a computer has of a user comes from user
models (Allen, 1997, p. 49). A Graphical User Interface (GUI) is a type of representation of a
data model from the perspective of user interaction; It determines how the data are displayed to
the user. Therefore, the GUI can only be effectively designed after the data model has been
developed (Stary, 1999). Computer-related design tasks, including software design of
educational applications and video games, may involve the interaction of several mental models.
They may include models of the capabilities of the tools, models of the partially completed work,
and models of the user’s interests and capabilities (Allen, 1997). A number of visual and
auditory components can aid in the development of mental models, including text, graphics, and
animation (Allen, 1997).
Metaphors
WAINESS PHD QUALIFYING EXAM
57
One mental model, the metaphor, uses “the similarity of one process with which a
person is familiar to teach that person about a different process” (Allen, 1997, p. 50). Metaphors
also help learners feel directly involved with objects in the domain so the computer and interface
become invisible (Wiedenbeck & Davis, 2001). There are several types of metaphors, including
activity metaphors, mode of interaction metaphors, and task domain metaphors. Activity
metaphors are determined by the user’s highest level goal; for example, controlling a process,
communicating, or playing a game (Neale & Carroll, 1997). Mode of interaction metaphors
organize the principal nature in which users think about interacting with the computer; these
metaphors are task independent. The third type of metaphor, the task domain metaphor, provides
an understanding for how tasks are structured. Most of the user interface literature discusses
metaphors at the task domain level (Neale & Carroll, 1997).
According to Neale and Carroll (1997), the mode of interaction metaphor can be
divided into three interaction categories: conversation, declaration, and model world. Two of
these categories, the conversation and model world metaphors, will be discussed here, due to
their relevance to command and direct manipulation interfaces. The conversation metaphor
creates a conversational interface (e.g. command line) which functions as an implied
intermediary between the computer and user, and is modeled after human to human
conversations (Neale & Carroll, 1997). The model world metaphor is what most in the user
interface community thinks of when working with metaphors. The model world is usually based
on the metaphor of the physical world, and the user interacts directly with the modeled world. A
combined metaphor, the collaborative manipulation metaphor, is a combination of the
conversational and model world metaphors (Neale & Carroll, 1997).
The ways in which a metaphor is incorporated into a mental model are difficult to
examine and probably vary greatly from user to user. In addition, a metaphor can be
counterproductive because the metaphor is rarely a perfect match to the actual process and
incorrect generalizations from the metaphor can result in poor performance on the task (Allen,
1997). Metaphor mismatches can occur for several reasons. Small dissimilarities between the
source and target domains cause mismatches. Combining several metaphor source domains will
typically result in mismatches among the metaphor mapping of the composite; the metaphors in
the composite can be inherently different, often directly contradicting each other. Mismatches
can also occur when the user’s task characteristics and goals change (Neale & Carroll, 1997).
Interface Style
Interface design is an effective way to accommodate user differences (Sein, Olfman,
Bostron, & Davis, 1993). According to Kaber et al. (2002), Graphical User Interfaces (GUIs)
were, in part, an outgrowth of direct manipulation, implying that the term interface includes both
the screen design and the style of interaction. Wiedenback and Davis (1997) contend that
interaction style may have a strong impact on perceptions of software and ultimately on its use,
particularly for users who are not computer professionals and who are characterized by an
irregular or less-intense pattern of use (Wiedenback & Davis, 1997). Three types of interfaces
are defined by the literature, based on their interaction style: conversational (or command),
direct manipulation, and menu.
Command interface. The conversational interface requires the user to read and
interpret either words or symbols which tell the computer to perform arithmetic operations and
processes (Brown & Schneider, 1992). In conversational interfaces, the user typically uses a
keyboard to type commands telling the computer what he or she wants to have happen. Often
these commands are similar to, but still unlike, natural languages (de Jong et al, 1993). A more
WAINESS PHD QUALIFYING EXAM
58
common term for the conversational interface is command interface (or command line interface).
Command interfaces are operated by the user typing a command string in a language and syntax
recognized by the system. The user must remember an array of commands, as well as their
syntax. And since several command lines, or a single complex line, may be required to achieve
the desired outcome, the user must also structure a sequence of actions correctly to achieve the
intended result (Wiedenback & Davis, 1997). Because interactions are carried out using a
keyboard, rather than by pointing, clicking, and dragging with a mouse, the results of the actions
are often not as visible to the user as when using the other two interface types; direct
manipulation and menu (Wiedenback & Davis, 1997).
Direct manipulation interface. Researchers credit Schneiderman with coining the phrase
direct manipulation in the 1980s (Brown & Schneider, 1992; Eberts & Brittianda, 1993; Kaber et
al., 2002; Phillips, 1995). Direct manipulation (DM) is a collective term that refers to a style of
HCI for user interfaces showing the properties of continuous representation of objects and
actions of interest, object manipulation with physical interaction with icons and buttons rather
than the use of complex syntax, and rapid incremental reversible operations with rapid, visible
feedback (Eberts & Brittianda, 1993; Kaber et al., 2002).
The direct manipulation interface (DMI) is defined as one in which the user has direct
interaction with the concept world; the domain. The user is able perceive a direct connection
between the interface and what it represents (Brown & Schneider, 1992). Broadly defined, direct
manipulation interfaces represent the physical manipulation of a system of interrelated objects
analogous to objects found in the real world. While the object representations may take on a
variety of forms, they are most commonly represented as icons; although it is possible to provide
text-based implementation of the objects or combined text-icon presentations (Benbasat & Todd,
1993). DMIs allow users to carry out operations as if they were working on the actual objects in
the real world. The gap between the user’s intentions and the actions necessary to carry them out
is small. These two characteristics of direct manipulation are referred to as engagement and
distance. High engagement with small distance lead to a feeling of directness in a system
(Wiedenbeck & Davis, 1997).
Engagement is defined as a feeling of working directly with the objects of interest in the
world rather than with surrogates (Wiedenbeck & Davis, 2001). Engagement refers to the
perceived locus of control of action within the system (Frohlich, 1997). A critical determination
for level of engagement is whether users feel they are the principle actors within the system. In
systems based on a conversational style of interaction, the locus of control appears to reside with
a “hidden intermediary” (Frohlich, 1997, p. 465). This interaction is considered indirect because
the user is not directly engaged with the objects of interest. In systems based on a graphical style
of interaction, with use of a pointing, clicking, and dragging device, the locus of control appears
to reside with users who manipulate the objects of interest (Frohlich, 1997). This creates a sense
of engagement.
The cognitive effects of direct manipulation can be described in terms of distance
(Frohlich, 1997; Wiedenbeck & Davis, 2001). Distance is a cognitive gap between the user’s
intentions and the actions needed to carry them out. This distance is in part a syntactic distance,
consisting of the translation of user intentions to a language and syntax understood by the
computer. It is also partly a semantic distance consisting of the translation of a user’s “real
world” understanding of the task to its computer implemented form (Wiedenbeck & Davis,
2001). With direct manipulation the syntactic distance is reduced by presenting the user with a
predefined list of visible options. The semantic distance is reduced by the use of an interface
WAINESS PHD QUALIFYING EXAM
59
metaphor that allows the user to click and drag familiar objects in a well-understood context
(e.g., the Windows or Macintosh desktop metaphor). The metaphor is most often complemented
with icons meant to evoke the metaphor in a concrete, visual way (Wiedenbeck & Davis, 2001).
According to Frohlich (1997), distance refers to “the mental effort required to translate goals into
actions at the interface and then evaluate their effects” (p. 466). Each intended action must span a
cycle of goal, action, and result. Interfaces which make it easier for users to perform these cycles
are said to be more direct (Frohlich, 1997).
Menu interface. In a menu style of interaction, objects and possible actions are
represented by a list of choices, usually as text. Menus are similar to direct manipulation in that
they help guide the user, which, like with direct manipulation, reduces mental burden. The menu
may help to structure the task and eliminate syntactic errors (Wiedenbeck & Davis, 1997).
However, menu-based systems are generally less direct than DMIs because the hierarchical
structure of the menus provide a kind of syntax that the user must learn. As a result, users do not
feel as directly connected to the objects they are manipulating through their actions (Wiedenbeck
& Davis, 1997).
Comparing interfaces. A number of studies have been conducted comparing command,
direct manipulation, and menu interfaces; some with consistent results and some without. The
findings of studies comparing menu to command line interfaces have been relatively consistent.
Overall, recognition and categorization may be faster for pictures than text (Benbasat & Todd,
1993). Menu interfaces lead to better task performance than the command interfaces, which is
attributed to a smaller gap between user intentions and actions with menu interfaces.
(Wiedenbeck & Davis, 2001). The results of studies comparing DMI to menu or DMI to
command line have been less consistent.
Widenbeck and Davis (1997) suggested that direct manipulation interfaces lead to more
positive perceptions of ease of use than does a command interface. With elementary school
students, Brown and Schneider (1992) found DMI more comfortable and enhanced the speed of
problem solving. DMI was also less frustrating compared to the conversational interface. Sein et
al. (1993), contended that because a direct manipulation interface provides an “explicit,
comprehensible, analogical conceptual model of the computer system, it can reduce the demands
placed upon subjects to internalize system states, which in turn leads to improved performance”
(p. 615). In support of this view, de Jong et al. (1993) found direct manipulation interfaces
enhanced the efficiency of task performance for both simple and complex tasks, with the
improved performance more pronounced for complicated tasks.
Other findings for direct manipulation interfaces are mixed or unclear. In an analysis
of empirical studies into the benefits of icons, and therefore direct manipulation, Benbasat and
Dodd (1993) found no clear advantage for the use of icons. According to Kaber et al. (2002),
although direct manipulation may minimize cognitive distance and maximize engagement, the
interface is less effective from the perspective of repetitive or complex tasks, particularly those
where one action is to affect multiple objects. They argue that, to achieve semantic directness
(the distance between the user’s intentions and the objects and operations provided by the
system), the user should be able to communicate those intentions to the system in a simple and
concise manner at all times. The need for repetitive actions in order to affect multiple objects is
not supported by DM and, therefore, increases mental effort and the amount of time needed to
complete a task (Kaber et al., 2002).
In a comparison of the DMI to the command interface, Westerman (1997) found that
the performance strategies of novices were relatively insensitive to command complexity while
WAINESS PHD QUALIFYING EXAM
60
experts were aware of this factor and used the command line less frequently as complexity
increased. And with regards to experts, Frohlich (1997) found that performance slows, rather
than speeds up, with direct manipulation interfaces, for two reasons. First, as was also suggested
by Kaber et al. (2002) and Westerman (1997), the language of DM limits complex actions.
Second, use of familiar real-world metaphors may limit users to existing ways of doing things;
while this may make learning and remembering easier for novices, it is more constraining for
experts. In communications, Frohlich (1997) found that direct manipulation interfaces increase
the cognitive load on conversational partners, even though it decreased the interactional work
between them.
A number of causes have been suggested to account for the discrepancies in the findings
for direct manipulation interfaces. Eberts and Brittianda (1993) questioned the validity of
interface comparison studies. They suggested that comparing performance differences across
interface design is difficult because the predicted execution times are intrinsically different for
each interface and, therefore, difficult to compare (Eberts & Brittianda, 1993). In contrast,
Benbasat and Todd (1993) argued that direct manipulation interfaces are often compared to both
command or menu type interfaces in studies. The menu interface eliminates the confounding
effect of time on performance found with command line. Also, since the menu interface is
usually made up of menu panels containing a list of options which may be words or icons, the
selection of menus facilitates an experimental design to test the main and interaction effects of
direct manipulation versus menus, and text versus icon-based interfaces (Benbasat & Todd,
1993). With this in mind, in a study of adult learners, Benbasat and Todd (1993) found no
performance advantages for icon-based systems, when compared to other interfaces. However,
both Benbasat and Todd (1993) and Frohlich (1997) have suggested that the icons themselves
may be influencing the findings. A number of factors have been found to affect the value and
usability of icons: specifically, complexity, meaningfulness, and concreteness. These factors
combine to define an icon’s distinctiveness. Distinctiveness refers to whether one icon is
confused with other icons (McDougall, de Bruin, & Curry, 2000). According to McDougall et al.
(2000), icon complexity is concerned with the level of detail used in constructing the icon’s
imagery. It is particularly important when simple icons are presented against a complex array, or
when complex icons are presented against a simple array. Icon meaningfulness refers to how
well an icon presents the user with its intended function; how much it portrays the action it
generates. And icon concreteness is the degree to which an icon depicts real world objects users
are familiar with (McDougall et al., 2000).
The effects of these various characteristics are influenced by the way in which icons are
grouped, concreteness of one icon compared to the other icons, and the complexity of an icon
compared to the complexity of other icons. According to the researchers, meaningfulness, rather
than complexity, appeared to be the primary determinant of icon distinctiveness when the
concreteness of the icon arrays was varied. Concrete icons (i.e., pictorially representing realworld objects), were seen as more meaningful against an abstract array. Conversely, abstract
icons were seen as more meaningfully against a concrete array. When simple and complex icons
were presented, which consisted of a mixture of both abstract and concert icons, both of these
effects were observed (McDougall et al., 2000). In their study, the effects of icon concreteness
were found to be short lived and limited to user’s early experience with an icon set. By contrast,
the effects of icon complexity were most apparent in tasks involving a search component and did
not diminish as a result of experience (McDougall et al., 2000).
WAINESS PHD QUALIFYING EXAM
61
While Benbasat and Todd (1993) found little or no performance advantages for icons,
and while Frohlich (1997) found no general advantages to using icons rather than textual menus,
Frohlich did suggest that in particular cases icons may be better because of the additional
information they carry. According to Benbasat and Todd (1993), icons may lead to improved
performance for novices and casual learners because of the superiority of visual memory over
verbal memory. Frohlich (1997) argued that the quality of the icon can affect both user
performance and study results. Frohlich has contended that poorly designed icons can actually be
worse than labels because those icons carry less information (Frohlich, 1997). And even with
well designed icons, it is difficult to design icons to convey the desired meanings without
invoking other connotations. When this happens, problems arise and semantic errors occur.
Ambiguous meanings arise because there is no universal set of icons or principles to guide icon
design (Benbasat & Todd, 1993).
A final possible confound in the findings with regards to direct manipulation interfaces
may be due to how specific interface implementations are defined. Many so called direct
manipulation interfaces include elements from several interface styles, and are more accurately
referred to as mixed mode interfaces (Frohlich, 1997). They include menus and windows, as well
as conversational interaction such dialog boxes, fill-in forms, and command languages (de Jong
et al., 1993; Phillips, 1995). The Macintosh operating system is one such example. While it is
typically referred to as a direct manipulation interface, it covers a range of interactions involving
a pointing device and keyboard for menu selection, dragging, and drawing, along with dialogue
boxes and text entry (Phillips, 1991). Pure direct manipulation interfaces according to the
framework would be “model-world interfaces based on Action in/Action out modality involving
only the media of sound, graphics, and motion. Dialog boxes, forms, and short-cut commands
are not part of this definition” (Frohlich, 1997, p. 478). Using this framework, many interfaces
which have traditionally been thought of as direct manipulation interfaces are in actuality mixed
mode interfaces (Frohlich, 1997).
Conclusion
Computers assist people in performing tasks of an increasingly difficult, complex, and
comprehensive nature, and using an application can be made easier by introducing “transparent”
interfaces (de Jong et al., 1993). A fundamental motivation of graphical user interfaces (GUI) is
to improve the medium and content of human-computer communication. The implementation of
such an interface can be achieved by providing visual representations of the concepts or items of
interest through the use of objects which can be directly manipulated by a pointing or selecting
device such as a graphics tablet, mouse, or track-ball (Edmonds, O’Brien, Bayley, & McDaid,
1993).
According to Kaber et al. (2002), at the heart of the direct manipulation concept is the
promotion of manual interaction with objects, rather than the use of a communications language
and syntax, to reduce the mental load placed on the learner’s cognitive system. The DM
paradigm has been found to support ease of learning for novice users, reduced error rates, and
decreased computer related anxieties (Kaber et al, 2002). These characteristics, along with a
greater perception of the relationship between input and output, can be found in a variety of
systems ranging from video games, interactive graphics packages, and spreadsheet programs to
computer-aided design systems, virtual-control systems, and many office systems (Kaber et al.,
2002).
The premise behind the expectation that direct manipulation interfaces would improve
performance is that, according to Benbasat and Todd (1993), when interacting with a computer a
WAINESS PHD QUALIFYING EXAM
62
user is typically focused on some cognitive task Attention devoted to the interface may interfere
with that task. Since text-based processing is associated with cognition, more interference should
be expected. The belief was that, with direct manipulation interfaces, the less effort required for
interaction would result in more mental resources being available for the learning task (Benbasat
& Todd, 1993). However, findings do not always support this expectation. Instead of an interface
that would provide benefit in all situations, these interfaces seem to improve selected aspects of
usability on a restricted set of tasks (Frohlich, 1997). Possibly, the solution lies in mixed
interfaces. According to Svendsen (1991), some tasks should have a command interface, others a
direct manipulation interface, and yet others some kind of hybrid interface. As Frohlich (1997)
suggested, manual interfaces are not always better than conversational ones, and combined
interfaces can leave the choice very effectively in the hands of users. Ultimately, the challenge is
to fine-tune computer interfaces to make computers easier to use and accessible to all learners
(Chalmers, 2000).
The purpose of this review is to explore the various task related constructs and conditions
that affect motivated behavior and, ultimately, mental effort. “Motivation generates the mental
effort that drives us to apply our knowledge and skills. “Without motivation, even the most
capable person will not work hard” (Clark, 2003, p. 21). Motivated behavior involves attempting
and persisting at academic achievement tasks (Corno & Mandinah, 1983), and learning is
strongly influenced by the amount of mental effort, the depth or thoughtfulness, learners invest in
processing material (Salomon, 1983). Mental effort is defined as “working ‘smarter’ at either a
new or old performance goal” (Condly, Clark, & Stolovitch, in press, p. 1). A number of items
affect motivation and mental effort. In an extensive review of motivation theories, Eccles and
Wigfield (2002) discuss Brokowski and colleagues’ motivation model that highlights the
interaction of the following cognitive, motivational, and self-processes: knowledge of oneself
(including goals and self perceptions), domain-specific knowledge, strategy knowledge, and
personal-motivational states (including attributional beliefs, self-efficacy, and intrinsic
motivation). In a study of college freshmen, Livengood (1992) found that psychological
variables (i.e., effort/ability reasoning, goal choice, and confidence) are strongly associated with
academic participation and satisfaction. And Corno and Mandinah (1983) commented that
students in classrooms actively engage in a variety of cognitive interpretations of their
environments and themselves which, in turn, influence the amount and kind of effort they will
expend on classroom tasks (Corno & Mandinah, 1983).
Task characteristics can be divided into three broad categories: A) the nature and content
of the task, B) the learner’s perceptions and interpretations of the task, and C) the context in
which the task is occurs, all of which can affect task perceptions, motivations, and mental effort.
The nature and content of the task includes elements such as task difficulty and whether the task
is collaborative or individualistic, as well as the task’s domain, the information to be learned, and
the instructional elements applied to the task. Individual perceptions and interpretations of the
task are based on a number of personal factors such as goal orientation, self-efficacy,
expectancies for success, and the value placed on the task. The context in which the task occurs
includes a variety of elements such as the classroom structure (e.g., whether a classroom is
collaborative or competitive), the instructional design, the presence or absence of rewards or
other incentives, the nature of the evaluative processes, the amount and type of instructional
support offered, and the goal orientation of the classroom. Each of the components within and
across the three categories interacts to create a complex network of influences and
interdependencies, which ultimately affect motivation and mental effort. The various
WAINESS PHD QUALIFYING EXAM
63
components can be referred to as task characteristics, since each explicitly defines the task or
applies interpretations to the task that alter perceptions and the personal definition of the task.
Each component serves to either support or undermine the investment of mental effort.
This review is based on the relationships and constructs defined in the CANE
(Commitment And Necessary Effort) model of motivation, (Clark, 1999). The model is a
compilation of a number of smaller, disparate motivational models (see Pintrich & Schunk,
2002). The one area of divergence from the CANE model in this review is with respect to
persistence and mental effort. In the CANE model, persistence and mental effort are seen as
distinct indicators of motivation that do not directly interact. Other researchers (e.g., Miller,
Greene, Montalvo, Ravindran, Nichols, 1996; Thompson, Meriac, & Cope, 2002) suggest that
mental effort can be an indicator of persistence, creating a relationship where persistence is an
independent variable and mental effort is a dependent variable. This review adopts a blend of
these two perspectives, where persistence is seen as an indicator of mental effort for those
findings that either explicitly connect the two and for findings that refer to persistence in a way
that suggests the application of mental effort. I have divided this review into four sections: goal
setting and goal orientation, expectancy-value theory and self-efficacy, instructional design, and
cognitive engagement and self-regulation. Each of these four sections are further divided into
subsections. The goal setting and goal orientation section includes an introduction of goal
orientation, followed by discussions of task orientation and performance orientation. The
expectancy-value theory and self-efficacy section is subdivided into an introduction of the
expectancy-value theory, followed by task value, and self-efficacy. The Instructional design
section is subdivided into an introduction, task difficulty, support and feedback, collaboration,
and incentives. The final section, cognitive engagement and self-regulation, is subdivided into an
introduction, followed by cognitive engagement, and effective strategy use.
Review of the Literature
Eccles and Wigfield (2002) discuss Pintrich and colleagues’ model of the relations
between motivation and cognition. The model incorporates a variety of components including
student characteristics (such as prior achievement levels), the social aspects of the learning
setting (e.g., the social characteristics of the task and classroom interactions between students
and teachers), several motivational constructs derived from expectancy-value and goal theories
(expectancies, values, and affect), and various cognitive constructs (e.g., background knowledge,
learning strategies, and self-regulatory and metacognitive strategies. Both the cognitive and
motivational constructs are assumed to influence students’ involvement with their learning and,
consequently, achievement outcomes (Eccles & Wigfield, 2002). Students achievement values
determined initial engagement and their self-efficacy facilitated both engagement and
performance in conjunction with cognitive and self-regulation strategies. In sum, the social
cognitive view of self-regulation emphasizes the importance of self-efficacy beliefs, casual
attributions, and goal setting in regulation of behavior directed at accomplishing a task or
activity. (Eccles & Wigfield, 2002).
Goals Setting and Goal Orientation
Individuals without specific goals (such as “do your best”), do not work as long as those
with specific goals, such as “list 70 contemporary authors” (Thompson et al., 2002; Locke &
WAINESS PHD QUALIFYING EXAM
64
Latham, 2003). Goal setting theory, according to Thompson et al. (2002), is based on the simple
premise that people exert effort toward accomplishing goals. Goals may increase performance as
long as a few factors are taken into account, such as acceptance of the goal, feedback on progress
toward the goal, a goal that is appropriately challenging, and a goal that is specific (Thompson et
al., 2002). Goal orientation theory is concerned with the prediction that those with high
performance goals and a perception of high ability will exert great effort, and those with low
ability perceptions will avoid effort. (Miller et al., 1996). Whether or not a person adopts a goal
is not only influenced by his view of his ability, it is influenced by other, salient evaluation
criteria (Bong, 2001).
In a study by Ames and Archer (1988) on the relationship of goal orientation to task
choice, and selection and use of effective learning strategies, the researchers found that the use of
learning strategies may be related to whether students adopted a mastery or performance goal
orientation in the classroom. A mastery goal orientation is when students undertake challenging
tasks for the sake of learning and improving abilities. Those who adopt a performance goal
orientation are concerned with how their abilities are perceived or evaluated by others.
Those with a performance orientation try to validate their superior ability or receive an
extrinsic incentive (Jagacinski & Nicholls, 1984; Bong, 2001). Depending on the situation, those
with a performance orientation may either try do demonstrate their ability (performanceapproach) or hide a perceived lack of ability (performance-avoidance). Those with a mastery
orientation also might try do demonstrate ability (mastery-approach) or avoid a situation where
they are not entirely sure of their ability to succeed (mastery-avoidance; Archer & Scavek,
1998). For mastery oriented learners, effort is seen as a way to increase ability and to succeed.
For performance oriented students, effort is seen as a sign of inability and, therefore, the
appearance of effort is to be avoided (Archer & Scavek, 1998).
There are a number of alternative terms for mastery orientation, including intrinsic
orientation, task orientation, task-involved orientation, and learning orientation. Alternatives to
performance orientation include extrinsic orientation, ability-focused orientation, ego-orientation
and ego-involved orientation (Jagacinski & Nicholls, 1984; Ames & Archer, 1988; Archer &
Scevak, 1998; Coffin & MacIntyre, 1999; Bong, 2001). In this review, the terms mastery and
performance will be used for these two constructs, respectively.
Mastery orientation. With mastery orientation, the belief is that more effort will lead to
greater mastery. If we try hard and increase mastery, that success leads to a greater feeling of
competence. Mastery is an end in itself—for challenge, curiosity, and mastery (Jagacinski &
Nicholls, 1984). A mastery orientation can be fostered by the way a task is structured, by the
nature of the evaluative system in which instruction is embedded, by the level of autonomy
afforded students, and by the opportunity to work collaboratively with other students (Archer &
Scevak, 1998). For example, providing students an opportunity to resubmit assignments as a way
to improve skills and grades has been found to promote a mastery orientation. Informational
feedback versus ranking feedback has also been found to promote mastery orientation;
Informational feedback gives students an indication of strengths and weaknesses and where to
focus future effort. (Archer & Scevak, 1998). According to Covington and Omelich (1984),
mastery oriented learning structures promote a number of factors thought to initiate and sustain
task involvement, persistence, and improved performance (Covington & Omelich, 1984). When
students perceive their class as emphasizing a mastery goal, they were more likely to use
WAINESS PHD QUALIFYING EXAM
65
effective learning strategies, prefer challenging tasks, enjoy their class more, and believe that
effort and success covary (Ames & Archer, 1988).
Performance orientation. In contrast to mastery orientation, individuals who are
performance oriented hold a differentiated conception of ability (i.e., effort and ability covary),
because their assessment of ability is based on normative information (comparison to others).
Perceived success occurs for when they demonstrate superior ability by outperforming peers
rather than displaying high effort or personal improvement (Fry & Duda, 1997). Activation of
the differentiated conception of ability will be likely when learners are directly concerned with
evaluating their own or another’s ability, such as with academic test performance and grading
systems, where competition with others is emphasized (Jagacinski & Nicholls, 1984).
When a performance orientation was salient to students, there was a tendency to see the
work as too difficult, reflecting a maladaptive motivational pattern that was unlikely to support
continued effort (Ames & Archer, 1988); Evaluative conditions can have this effect. Testing
situations commonly involve norm-referenced evaluations on performance, increasing the
likelihood that a differentiated conception will be activated. The differentiated conception is
necessary for adequate or objective evaluations of ability; If we don’t compare our effort and
performance with that of others, we can’t tell whether our performance is due to task difficulty or
effort, as opposed to ability (Jagacinski & Nicholls, 1984). High effort in mastery involving
situations can lead to feelings of competence, accomplishment, and pride. High effort in
performance involving situations generally results in lower feelings of competence (Jagacinski &
Nicholls, 1984).
Expectancy-Value Theory
Expectancy-value theories propose that the probability of behavior depends on the value
of a goal and expectancy of obtaining that goal (Coffin & MacIntyre, 1999). Expectancies refer
to beliefs about how we will do on different tasks or activities, and values have to do with
incentives or reasons for doing the activity (Eccles & Wigfield, 2002). From the perspective of
expectancy-value theory, goal hierarchies (the importance and the order of goals) also could be
organized around aspects of task value. Different goals may be perceived as more or less useful,
or more or less interesting. Eccles and Wigfield (2002) suggest that the relative value attached to
the goal should influence its placement in a goal hierarchy, as well as the likelihood a person will
try to attain the goal and therefore exert mental effort.
Task value. Task value refers to an individual’s perceptions of how interesting,
important, and useful a task is (Coffin & MacIntyre, 1999). Interest in, and perceived importance
and usefulness of, a task comprise important dimensions of task value (Bong, 2001). Citing
Eccles’ expectancy-value model, Townsend and Hicks (1997) stated that the perception of task
value is affected by a number of factors, including the intrinsic value of a task, its perceived
utility value, and its attainment value. Thus, engagement in an academic task may occur because
of interest in the task, or because the task is required for advancement in some other area
(Townsend & Hicks, 1997). According to Corno and Mandinah (1983), a task linked to one’s
aspirations (a “self-relevant” task) is a key condition for task value (Corno & Mandinah, 1983).
However, task value can be affected by other perceptions. For example, if a person has a
performance orientation, it is predicted that motivated behavior should decrease on self-relevant
WAINESS PHD QUALIFYING EXAM
66
tasks if the performance of significant others are interpreted as relatively more successful (Corno
& Mandinah, 1983).
Low task relevance can have a similar effect. If, for example, the results of a test are
nonconsequential (they have no utility value), or if the student perceives a test as
nonconsequential, he may not invest sufficient effort on complex (and therefore more mentally
taxing) test items (Wolf, Smith, & Birnbaum, 1995). In addition, participation in any task may
also carry negative aspects or costs which can affect the individual’s perception of the task.
These costs may include the amount of effort necessary for success or loss of valued alternative
activities. According to Townsend and Hicks (1997), because of the limitations of time and
energy, a student’s decision to participate in a valued academic task might result in an inability
to participate in another highly valued activity, such as a social activity. Thus, an activity in one
life domain may have high intrinsic, utility, and attainment values, yet may act as an obstacle to
success in an activity in some other life domain. The cost of involvement in the first activity
would decrease the overall value of that activity (Townsend & Hicks, 1997). Goal satisfaction or
dissatisfaction in any domain may be related to how activities in other domains are perceived,
through task value, suggesting that Eccles expectancy-value model of motivated behaviors can
be used to consider not only academic achievement behaviors but also achievement in the wider
sense of social goals (Townsend & Hicks, 1997). In addition, social satisfaction also influences
the value of social tasks, and their position in the goal hierarchy. The more socially satisfied a
person is, the greater the perceived value and the lower the cost. For those low in social
satisfaction, a classroom structure that supports the social domain, such as a classroom that
promotes collaboration and cooperative learning, can have a positive effect on students’ task
values. For example, students in a math or language classroom with a cooperative goal structure
reported higher task values for those classes (Townsend & Hicks, 1997).
There are other ways a classroom can be structured to increase perceived value. Miller et
al. (1996) suggested that an emphasis on the coordination of proximal goals with distal valued
outcomes (future consequences) is one such solution. The distal goals are expected to help
sustain effort in academic areas that are of low interest to students. The proximal goals are to
help promote the utilitarian component of task value. Archer and Scevak (1998), suggest that
choice in a task or topic can promote interest; another component of task value (Archer &
Scevak, 1998).
Self Efficacy. Academic self-efficacy is a student’s beliefs about his or her capabilities to
perform academic tasks at specific levels (Bong, 2001). People’s beliefs about their ability to
successfully perform a task influence their willingness to attempt the task, the level of effort they
will expend on the task, as well as their persistence in the face of challenge (Miller et al., 1996).
Self-efficacy can also determine the goal orientation of a student. According to Livengood
(1992), students low in confidence in their intelligence tend to be performance oriented, to
validate their ability and perform in order to look good, even at the risk of not learning. Those
high in confidence tend to be mastery oriented and participate in activities to develop their
abilities and increase mastery.
The goal orientation of the task, whether it is performance or mastery oriented, can affect
students differently, depending on the students’ levels of self-efficacy. Jagacinski and Nicholls
(1984) suggested that people who perceive themselves as able will perform equivalently in
performance and mastery situations. Those will low self-efficacy will perform worse in
performance oriented situations than in task oriented situations. Performance in task oriented
WAINESS PHD QUALIFYING EXAM
67
situations appears to be equivalent regardless of whether a person has low or high self-efficacy.
In performance situations, low self-efficacy students will perform poorly due to fears of negative
appraisals; they see performance tasks as a test of their abilities. Those with high self-efficacy
perform well because they do not have that overemphasis on being evaluated. They are more
concerned with the task and the learning process; they tend to approach the performance
situation as if it were a mastery situation (Jagacinski & Nicholls, 1984).
In contrast to self-efficacy, which is somewhat global, task value and goal orientation are
more domain specific. How much value students attach to particular subject matter and their
preferences toward task mastery and challenge in that subject varies across domains (Bong,
2001). Furthermore, task value (importance, usefulness, and intrinsic interest) may play a more
meaningful role than self-efficacy in guiding students to a mastery orientation. In a study of high
school students, Bong (2001) found that task-value perceptions were clearly differentiated across
diverse subjects. In addition, mastery orientation followed the same pattern as task value,
suggesting a correlation of cross-domain associations. In contrast, self-efficacy perceptions were
only moderately correlated across subjects (Bong, 2001).
Instructional Design.
Self-efficacy interacts with a number of theories, including attribution theory, social
cognitive theory, and achievement theory. In classrooms of students at cognitive levels where
attributional explanations for behavior make sense, situations most likely to induce attributions
vary. According to Corno and Mandinah (1983), these situations can include grading, testing,
and other evaluation procedures; skill training or drilling exercises; and problem solving or
competitive games, as well as assignments involving the simultaneous application of various
academic skills (e.g., leading a discussion, writing) and performance of sex-typed tasks by
students of the opposite sex (Corno & Mandinah, 1983). Social Cognitive Theory posits that
those who perceive a more positive outcome will work harder to increase learning, and will
therefore perform better (Wiedenbeck & Davies, 2001). According to Miller et al. (1996),
theories of achievement motivation built around competence-related goals have suggested that
students’ desires to increase their knowledge, understanding, or skills (i.e., mastery orientation)
are major factors in guiding their level of engagement in academic tasks. However, the extent to
which students hold valued long-term goals and the extent to which they perceive their current
school experiences as related to the attainment of those goals must also be considered. This
suggests that educators must enlist a variety of tools in their efforts to foster cognitive
engagement and learning (Miller et al., 1996). Achievement motivation is also enhanced to the
extent that the learner perceives the positive relationship between the amount of study time
expended and the rewards (e.g., proximal rewards such as grades) attained. This covariation
strengthens the saliency of effort as a primary cause of one’s successes and failures (Covington
& Omelich, 1984).
There are a number of ways a classroom environment can be structured to encourage
learning and enhance student motivation. As an example of an effective teaching and
motivational tool, providing multiple-retesting, in order to attain a goal with a predetermined
level of achievement, provides feedback regarding what is yet to be learned (Covington &
Omelich, 1984). According to Hughes, Sullivan, and Mosely (1985), the few studies dealing
explicitly with the effects of teacher evaluation on continuing motivation (i.e., return to task)
have only partially supported the contention that motivation is reduced by teacher evaluation.
WAINESS PHD QUALIFYING EXAM
68
One explanation is the nature of the environment in which the evaluation occurred. One effective
format for evaluation is to first allow students to gain mastery of a task and only evaluate
performance after the task becomes relatively easy for the majority of students. Teacher
comments and judgments, while students are still learning to perform the task well, should take
the form of constructive feedback designed to help students improve their performance, rather
than as evaluation for some other purpose—a similar instructional and evaluative process
recommended for mastery learning (Hughes et al., 1985). A second example of how to improve
learning and student motivation is to provide a classroom environment that supports social
interactions. According to Townsend and Hicks (1997), social satisfaction should be higher in
classrooms where teachers utilize methods of instruction that provide greater opportunities for
involvement and affiliation with other students. One such form is cooperative learning, where
small groups of students work together to accomplish shared goals.
Goal orientation. Goal orientation also plays a significant role in how students utilize
mental effort, as well as their attitudes. Ames and Archer (1988) commented that when we ask
why students fail to use effective learning strategies, we may not be giving enough attention to
the conditions of learning that may affect the use of learning strategies. We may need to consider
how the student perceives the goal orientation of the learning environment. Situational demands
can affect the salience of specific goals, which in turn results in differential patterns of cognition,
affect, and performance (Ames & Archer, 1988). For example, when social comparison is made
salient, students focus on their ability, and these self-perceptions mediate performance and
affective reactions to success and failure. By contrast, when absolute standards, selfimprovement, or participation are emphasized, students focus more on mental effort and task
strategies (Ames & Archer, 1988). In many classrooms, the informational cues that serve to
emphasize one goal or another are often mixed and tend to be inconsistent over time. Further,
students in the same classroom may differ in the degree to which they focus on certain cues, as
well as how they interpret them (Ames & Archer, 1988). The degree to which a classroom
climate emphasizes mastery orientation, rather than performance orientation, is predictive of how
students choose to approach tasks and engage in learning (Ames & Archer, 1988). However, it is
the students’ perception of the classroom orientation that matters more than the teachers intended
orientation. Archer and Scevak (1998) found that the way lecturers approach their teaching—the
attitudes and behavior they display—is related to students’ motivation to learn. Students teachers
who perceived the lecturer to be encouraging a mastery orientation made use of the types of
study strategies that are expected to enhance understanding, they enjoyed their tutorials, they saw
the subject as relevant to their future (teaching) careers, and they were willing to tackle difficult
rather than easy tasks. This adaptive approach was displayed not only by the highly competent
students but by students who saw themselves as only average or below average, as well (Archer
& Scevak, 1998).
Another instructional practice that can foster mental effort is related to absolute grading
standards (criterion-based assessment). However, while absolute grading standards contribute to
performance improvements, it is the level of standards expected, rather than whether they were
defined in relative or absolute terms, that primarily affected the increased performance. This
raises the question of the optimal motivational level of task difficulty (Covington & Omelich,
1984).
WAINESS PHD QUALIFYING EXAM
69
Task difficulty. According to Davis & Wiedenbeck (2001) cognitive curiosity, which
arises from situations in which there is complexity, incongruity, and discrepancy, motivates the
learner to attempt to resolve the inconsistencies through exploration. Salomon (1983) suggests
that learning, and the amount of mental effort expended, greatly depends on the differentiated
way in which sources of information are perceived, and that those perceptions influence the
mental effort expended in the learning process. The amount of mental effort learners invest in
extracting information from a source, discriminating among information units, remembering, or
elaborating is influenced by the way they perceive that source. Perceptions of a source refer to
the mental effort requirements of the message, its attributions (e.g., depth, complexity,
importance), the tasks to be performed, as well as the context in which the learner is exposed to
the source (Salomon, 1983).
According to Archer and Scevak (1998), task difficulty is an elusive thing to define. One
influential component in that definition is the probability of error or the time or effort required to
avoid error, pointing to the importance and interrelatedness of the subtasks that constitute smaller
and simpler cognitive skills (Archer & Scevak, 1998). Crawford (1978) comments that the
instructional difficulty level that best facilitates learning has been examined in a number of
different contexts. The findings have indicated that no single best difficulty level exists for
optimally promoting knowledge acquisition for all types of learners in all situations (Crawford,
1978). A 50% difficulty level is suggested for individuals with a high need for achievement,
since this is neither too easy and boring nor overly difficult and frustrating. For individuals with
a strong fear of failure and a low need for achievement, instruction that is neither very low nor
very high in difficulty is predicted as being optimal—either because there is a very low
probability of failure at the low difficulty level, or an excuse for failure (an opportunity for
external attribution) at the high (above 50%) degree of difficulty. Therefore, these learners would
prefer difficulty levels tending toward the extremes (0% or 100%; Crawford, 1978). According
to Clark (2003) an impossible task is one where the perceived probability of success is less than
15%.
In a study by Archer and Scevak (1998), participants performed better when trials were
more difficult to initiate. These results are consistent with theories in which attention (mental
effort) is allocated in response to the high level of task difficulty. By creating more difficult
initial tasks, allocation mechanisms are “tricked” into investing more attention and effort than is
necessary (or longer than is necessary) so that on subsequent tasks cognitive performance
benefits from that initial boost initiated by the trial-initiation demands. This would suggest that
conditions of difficult trial initiation result in relatively increased cognitive arousal, which in turn
yields corresponding increases in the capacity of available attention (Archer & Scevak, 1998).
These findings have implications for instructional practices, particularly in the form of computerbased instruction or drills, which are frequently designed with the goals of making procedures as
easy as possible and introducing material slowly. The findings of Archer and Scevak suggest that
such practices may be counterproductive, because students are unlikely to sustain mental effort if
the initial tasks are too easy and do not produce high mental effort demands. Therefore, each
type or dimension of task difficulty should be carefully considered in the design and analysis of
tasks, to determine the optimal initial levels of task difficulty for eliciting and sustaining
attention, accelerating learning, and improving performance (Archer & Scevak, 1998).
In contrast, Hughes et al. (1985) suggested that students have been shown to return to a
task at a greater rate as they feel more competent on the initial task. Return to task was
significantly higher when subjects initially were given an easy task rather than a hard one. As
WAINESS PHD QUALIFYING EXAM
70
students’ performance improved, they returned to task more often. Return rates were also
significantly higher for students who reported they thought they did not perform well on initial
tasks. And students returned to the easier task at a higher rate than they did to the harder task
(Story & Sullivan, 1986).
The effects of task difficulty on performance may be moderated by other variables such
as goal orientation. Jagacinski and Nicholls (1984) commented that presentation of a moderately
difficult or challenging task (i.e., at the 50% difficulty level) in mastery oriented conditions
should generate an expectation that higher effort would lead to more mastery, thereby
demonstrating higher ability. As long as the task is not perceived as too difficult to support a gain
in mastery, all individuals should apply high effort and perform effectively (Jagacinski &
Nicholls, 1984). However, if the same moderately difficult task were presented in a performance
oriented context, individuals might face the dilemma that although high effort could increase
performance, it could also become a demonstration of low ability. Individuals who believe their
ability is low (as compared to others) would expect to perform poorly (relative to others), even if
they tried hard, and therefore demonstrate low ability. For these individuals, low effort might be
seen as a way of reducing the degree to which failure would imply low ability (Jagacinski &
Nicholls, 1984).
Crawford (1978) suggested that learners with strong cognitive structures learn optimally
under less redundant (i.e., more difficult) conditions. However, for less able students, the
instruction is probably best if it proceeds in smaller steps and presents the information in a more
redundant format. For these less able students (or those who perceive themselves as less able),
success on a task appears to improve performance on subsequent attempts at the same task, and
success on one task effects the speed of learning on the second task. However, success on one
task does not always facilitate success on a subsequent task (Crawford, 1978). Small steps, and
prior success may not be beneficial to all students, depending on their goal orientation.
According to Latta (1978), for students who are not mastery oriented but must master a difficult
task, prior success can be detrimental to the learning process. In contrast, prior success helps
students with a master orientation when attempting to master a difficult task (Latta, 1978).
Support and feedback. A positive, personalized, and encouraging comment may not be
powerful enough to motivate students to return to task and exert mental effort (Story & Sullivan,
1986). The context of the comment is an important mediator. For example, challenging tasks
may be less threatening and possibly even more attractive to students who view the situation as
emphasizing the process of learning, encouraging effortful activity, and deemphasizing the
negative consequences for making errors—a mastery orientation (Ames & Archer, 1988).
According to Hughes et al. (1985), students returned to task more often after a hard activity
under self-evaluation or after an easy activity under teacher evaluation. It was suggested that the
reason for low return to task on difficult tasks with teacher evaluation was do to the threat of
exposure. Students felt threatened that their poor performance would be observed and evaluated
by the teacher. This threat reduces motivation and therefore reduces return to task (continued
mental effort). By providing self evaluation for difficult tasks, that threat was removed. As a
result, students commonly perceived performance on a hard task as more of a challenge under
self-evaluation and as more of a threat under teacher evaluation (Hughes et al., 1985).
One type of feedback that seems particularly helpful in motivating students is success
feedback. Success feedback may function as a reinforcer, a cue for eliminating errors, or an
incentive (Latta, 1978). The immediate effects of success feedback will lead to better
WAINESS PHD QUALIFYING EXAM
71
performance by individuals low in achievement orientation compared to those high in
achievement orientation. An individual high in achievement orientation is predominantly
motivated to approach success, while a person low in achievement orientation is predominantly
motivated to avoid failure. Therefore, the probability of success on a task is an important,
moderating factor. The differences between those with high and low achievement orientation
occur because those high in achievement orientation prefer to work on tasks with a probability of
success of about .5, while those low in achievement orientation prefer to work on tasks with a
probability of success closer to either 1.0 or 0.0. Thus, any facilitation of performance by success
feedback observed on the first task should be moderated by initial achievement orientation, with
success exerting a more positive impact on individuals initially low in achievement orientation
(Latta, 1978).
Collaboration. Cooperative task structures are situations in which two ore more
individuals are allowed, encouraged, or required to work together on a task. The task structures
used in cooperative (collaborative) learning situations can be divided into two categories: task
specialization and group study. With task specialization, each group member is responsible for a
specific part of the group activity. With group study methods, all group members study together
and do not have separate tasks (Slavin, 1984).
According to Slavin (1984), there are several reasons collaborative tasks might be
expected to improve student achievement. Collaborative tasks can promote peer tutoring, group
discussions, and controversy—all which appear to increase comprehension. However, the effects
of cooperative learning tasks on achievement depend on the behaviors of the group and the
characteristics of participants, as well as other factors. Cooperative environments have been
found to be beneficial in some circumstances and harmful in other (Slavin, 1984). Cooperative
learning (i.e., collaboration) can either support or deter mental effort, depending on student
attitude and on classroom structure. Students may use collaboration as a way of doing less. For
example, Archer & Scevak (1998) found that for students who worked with a partner, some
stated they chose to partner in order to halve the workload. Others, though, chose collaboration
for positive reasons. Some students chose to work with others to increase the number of ideas
generated (Archer & Scevak, 1998).
In a study that used television as the medium for content delivery, Klein, Erchul, and
Pridemore (1994) found that students who worked alone performed better than those who
worked cooperatively. The structure in which the students worked on the first tasks, influenced
their preferences for the way subsequent tasks were structured. Students working alone
expressed more interest in individual activities, while those who worked cooperatively expressed
a desire for activities that required cooperative learning (Klein et al., 1994). However, the results
of the study may have been skewed, due to the nature of the incentive structure (the rewards).
Klein et al. (1994) suggested that the results of their experiment indicated that the positive effects
of these methods on student achievement resulted from the use of cooperative incentives, not
from the use of cooperative tasks. Slavin (1984) contended it is not just the administration of
rewards, but the nature of the rewards that may affect outcomes (Slavin, 1984). Slavin stated that
the most successful cooperative learning methods do little to alter the content or deliver of
instruction. While the methods do change the way students study, the group study aspect of the
cooperative learning methods has not been found to contribute to achievement effects. However,
the evidence indicated that a simple change in a classroom incentive system produces relatively
consistent changes in student achievement (Slavin, 1984).
WAINESS PHD QUALIFYING EXAM
72
Incentives. Students who are unmotivated to learn do now learn (Slavin, 1984). Student
motivation refers to students’ interest in doing academic work and learning academic material.
Continuing motivation (persistence) is defined as returning to a task or behavior without
apparent external pressure to do so when other behavior alternatives are available (Malouf, 19971998). Classroom incentives refers to methods teachers use to motivate students to do academic
work and learn materials (Slavin, 1984). Student motivation is influenced in part by classroom
incentives, but also by such factors as interest in the task, parents’ interest in the students’
achievement, and students’ perceptions of their abilities and chances of success (Slavin, 1984).
According to Malouf (1997-1998), several factors have been found to influence the effects of
inducements upon subsequent effort, including the power of the inducement, the initial level of
motivation, the effects on self-perceived competence and task enjoyment, and the relationship
between inducement and behavior. For example, incentives based on mental effort have been
shown to produce a performance gain of 20% (Condly et al., in press).
A clear distinction should be made between the terms reward and reinforcer. A reinforcer
acts to strengthen a behavior, (e.g., by increasing rate, intensity, duration, or quality). If a reward
is delivered but no strengthening of behavior is observed, it cannot be said that reinforcement has
occurred. The majority of studies on the reduced continuing motivation have not reported strong
reinforcement effects on behavior (Malouf, 1983). Continued effort is only one of several
possible ways in which rewards may influence behavior. Rewards may also convey information
about the probability of future reinforcement, promote the development of skills which may
allow a student to enjoy previously unenjoyable activities, or convey information about a
students ability or competence. The net effect of a reward on subsequent behavior may be from a
combination of these and other messages conveyed by the reward (Malouf, 1983).
A number of researchers (e.g., Hughes et al., 1985; Coffin & MacIntyre, 1999) have
commented that, in many cases, extrinsic motivation (rewards) decreases initial intrinsic
motivation (interest) and may even interfere with the process of learning. This effect, know as
the over-justification effect, commonly occurs when both intrinsic and extrinsic reasons for
participating in the task are present. Because there is an overabundance of justification, the
attribution of intrinsic interest is discounted by the presence of the external incentive. In general,
this occurs because extrinsic rewards may distract attention away from a student’s interest and
enjoyment of a task, as well as the actual process of learning (Coffin & MacIntyre, 1999). The
over-justification hypothesis has been used to explain the results of the apparent negative effects
of rewards on intrinsic interest. The plausible explanation is that offering external motivators for
an inherently interesting activity will result in a reduction of interest in the activity. This
prediction is an outgrowth of the self-attribution theory and of the study of personal causation
(Hughes et al., 1985). According to the self-attribution theory, the reasons for engaging in
activities are perceived and inferred from the environment. When there are no external
motivators, the reason for pursuing an activity is attributed to personal interest and desire.
However, when an external motivator is introduced, the reason for engaging in the activity is
attributed to that external force. Personal causation hypothesizes that for a person to be
motivated to pursue an activity, she must feel she is the cause of that action. Rewards change this
perception of personal causation and thus undermine intrinsic interest (Hughes et al., 1985).
There are a number of researchers who dispute the over-justification hypothesis. For
example, Hughes et al. (1985) believe there are inconsistencies in the over-justification
hypothesis. They suggest that the lack of consistent relationship across studies between teacher
WAINESS PHD QUALIFYING EXAM
73
evaluation and continuing motivation may indicate that the over-justification hypothesis doesn’t
adequately explain the relationship between grades and motivation. For example, it appears the
hypothesis only applies with an activity initially of high interest for an external reward to reduce
that interest (Hughes et al., 1985). Eisenstein (1985), too, commented that rewards that
undermine interest for initially high interest subjects appear to raise interest for initially low
interest subjects. Other researchers have found a number of factors that also might affect overjustification. According to Miller et al. (1996), it may be the nature of the reward and not just
any reward that affects intrinsic interest. Immediate extrinsic rewards are typically presented in a
manner that reduces a person’s sense of self-determination. However, the pursuit of distant
outcomes (distant rewards), rather than proximal rewards, is likely to be viewed as selfdetermined rather than imposed; The result would be continued intrinsic interest (Miller et al.,
1996). Malouf (1983) suggested that exogenous (rewards unrelated to a task) may support the
over-justification effect, while endogenous rewards (rewards related to the task) do not.
Eisenstein (1985) also found that endogenous rewards enhance an activity so that the activity
itself is the end, and when the rewards are exogenous, the activity simply becomes a means to an
end (Eisenstein, 1985).
In addition to offering rewards for individual work, rewards can also be offered to those
working in a group. According to Slavin (1984), there are two primary components of
cooperative learning methods: a cooperative task structure and a cooperative incentive structure.
Cooperative learning methods always involve cooperative tasks, but not all of them involve
cooperative incentives. Cooperative task structures are situations in which two ore more
individuals are allowed, encouraged, or required to work together on some task, coordinating
their efforts to complete a task. The critical feature of a cooperative incentive structure is that
group members are interdependent for a reward they will share if they are successful as a group.
Cooperative incentive structures usually involve cooperative tasks, but the two are conceptually
distinct (Slavin, 1984). There are three types of incentive structures used in cooperative learning
methods: rewarding a group, rewarding the individual, or offering no rewards at all. A group
reward structure provides all group members the same reward, based on performance of the
group as a whole. An individual reward structure provides each individual in the group with a
reward, based on that individual’s performance (Klein et al., 1994). Through a metanalysis of 46
field experiments on cooperative learning, Slavin (1984) suggested that the optimum reward
structure for group tasks is group rewards, because rewards based on group performance creates
group member norms supporting performance; group members try to make the group successful
by encouraging each other to excel. In support of Slavin’s comments, in a meta-analysis by
Condly et al. (in press), findings indicated a 48% increase in performance for team-based
incentives. Slavin hypothesizes that groups create an internal, very sensitive, and very effective
socially based reward system for each other, in which they pay a great deal of attention to each
other’s efforts and socially reinforce efforts to help the group achieve its goal. The group is also
likely to apply social disapproval to group mates who are underperforming or playing around
instead of learning (Slavin, 1984).
The individual reward system for a group can also take the form of a competitive reward
system, and promote a competitive learning environment. In a competitive learning mode,
rewards are restricted to top performers (or the top group) so the likelihood of a student or group
receiving a reward is reduced by the presence of other able students or groups. In contrast, under
an individualistic reward structure, the likelihood of attaining a reward does not depend on the
performance of others. Such noncompetitive conditions lead to a classroom mastery orientation,
WAINESS PHD QUALIFYING EXAM
74
where improvement in performance over time becomes the basis for evaluation and selfimprovement becomes a dominant goal (Covington & Omelich, 1984). An alternative approach
to the competitive reward structure is to reward the group based on the highest scores. In an
analysis of a number of studies, Slavin (1984) found that when the group was rewarded based on
the highest scores, high achievers learned the most, while low achievers learned the most only
when the group depended on their scores. Student achievement is best enhanced by cooperative
learning methods that use group rewards for individual learning, and by learning methods that
maintain high individual accountability for students.Cooperative learning where groups are
rewarded on the basis of the sum of all members provides the greatest learning benefit, and
therefore the greatest expenditure of mental effort, to all group members (Slavin, 1984).
Cognitive Engagement and Self-Regulation.
Students in classrooms actively engage in an array cognitive interpretations of their
environments and themselves. This in turn, influences motivation in the form of the amount and
type of effort exerted (Corno & Mandinah, 1983). Goals initiate and direct behavior, and the
content of the goals help to determine the strategy used for achieving them (Rosswork, 1977).
According to Corno and Mandinah (1983), evidence suggests that students use varied processing
strategies to carry out common academic tasks. These strategies are variations of self-regulated
learning, and students differ in their spontaneous use of these variations. Students apply different
cognitive engagement strategies because tasks vary in novelty, difficulty, and competitive
features, because teachers provide different types of instruction and guidance, and because
students have different goals, past experiences with the task or the domain, and general ability
levels and mental sets (Corno & Mandinah, 1983). Jones, Yokoi, Johnson, Lum, Cafaro, and Kee
(1996) also supported the effect of the availability and accessibility of relevant knowledge on
strategy processing.
Cognitive engagement. According to Corno and Mandinah (1983), there are four forms of
cognitive engagement: self-regulation, task focus, resource management, and recipience. Each
form is defined by the amount of acquisition (alerting, monitoring, and high-level planning) and
transformation (selectivity, connecting, and low-level planning) processes used. Transformative
processes are cognitive processes that directly help in generating knowledge (Corno and
Mandinah, 1983). Examples of transformative processes include hypothesis generation and data
interpretation (de Jong, de Hoog, & de Vries, 1993). Transformation processes (i.e., selecting,
connecting, and planning) have both metacognitive and cognitive features; They can activate
other cognitive schemata that may be relevant for the task (Corno & Mandinah, 1983).
According to de Jong et al. (1993), alertness, monitoring, and high-level planning are
predominantly information acquisition processes; the information is gathered primarily from the
environment. Acquisition processes bound and control the transformation processes. The
acquisition processes are viewed as metacognitive because they regulate the transformation
processes. The transformation processes have both metacognitive and cognitive aspects (Corno
& Mandinah, 1983). de Jong et al. (1993) defined similar processes, using the term regulative
processes, which combines some aspects of both acquisition and transformation. de Jong et al.
(1993) stated that regulative processes help manage learning through processes such as
monitoring, planning, and verifying, and that monitoring and planning together can be called
navigation. For this discussion, Corno and Mandinah’s terms and definitions will be used.
WAINESS PHD QUALIFYING EXAM
75
Considered the highest form of cognitive engagement, self-regulation is one of Corno and
Mandinah’s four forms of cognitive engagement and consists of specific cognitive activities,
such as deliberate planning and monitoring, which learners carry out as they encounter academic
tasks (Corno & Mandinah, 1983). Self-regulation processes include elaboration, problem
solving, decision making, integration, and planning (Corno & Mandinah, 1983). According to
Eccles and Wigfield (2002), self-regulated learners have three important characteristics: They
use an assortment of self-regulated strategies; they are self-efficacious; and they have numerous
and varied self-determined goals. Self-regulated learners engage in three important processes:
self-observation (monitoring personal actions); self-judgment (evaluation and comparison to a
performance goal or other standard, such as the performance of others); and self reactions
(reactions to performance outcomes). When these reactions are favorable, students are more
likely to persist and apply mental effort. The reactions to failure are of particular importance.
The favorableness of a learner’s reaction to failure is determined by how the learner interprets
his difficulties and failures (Eccles & Wigfield, 2002). Corno and Mandinah (1983) suggested
that self-regulated learners is are forever increasing, deepening, and manipulating specific
content networks or associative memory networks, including the strength of the bonds between
propositions. Therefore, self-regulated learning is an effort to deepen and manipulate the
associative network in a particular area (including non-academic domains) and to monitor and
improve that deepening process.
In the second form of cognitive engagement, task focus, students activate relatively more
information transformation processes than acquisition processes; selectivity, connecting new
information to existing knowledge, and task-specific planning are the key cognitive activities.
Task focus is appropriate when tasks require quick analytic responses and little self-checking or
use of external cues. Task focus can be promoted by instruction that systematically eliminates
the irrelevant features of an object, idea, argument, or event; for example, demonstrating the
steps a learner would take in determining information relevant to completing a task, and sorting
and chunking of that information into meaningful categories. Task focused instruction would
emphasize the separation of the relevant from the irrelevant information and further emphasize
that only the relevant information is important to achieve the desired performance. Task focused
instruction should also emphasize the importance of using what the student already knows to
help categorize and anchor new information in memory, and to visualize changes in design and
visual fields. This type of instruction can help students prepare for some types of achievement
and ability tests (Corno & Mandinah, 1983).
The third form of cognitive engagement is resource management. According to Corno
and Mandinah (1983), although self-regulated learning is the highest form of cognitive
engagement, self-regulated learning is somewhat taxing. When tasks create cognitive demands,
students may engage in self-regulated learning; or they may shift the mental burden by calling on
available external resources, such as a knowledgeable peer; this process of acquiring external
cognitive resources is termed resource management. With resource management, learners
intentionally avoid the mental effort of carrying out information transformation on their own,
instead enlisting the help of others for some or all task components (Corno & Mandinah, 1983).
The social character of the classroom setting can encourage resource management. Cooperative
learning environments, where group work or peer support is encouraged, is an example of a
classroom situation that can encourage resource management (Corno & Mandinah, 1983).
Recipience, the fourth form of cognitive engagement, is a form of passive response or
learning, where the environment provides much of the transformation and low-level monitoring
WAINESS PHD QUALIFYING EXAM
76
processes; termed short-circuiting. In these environments, most of the mental burden is removed
from the learner and provided by an external source, similar to resource management. The
difference is that with resource management, the learner must enlist external support. With
recipience, the external support is automatically provided to the learner through the instructional
process. For example, advanced organizers that provide short-circuiting promote the use of
recipience (Corno & Mandinah, 1983). Short-circuiting organizers include charts and diagrams,
summaries and reviews, outlines and marginal notes, markers of important points, and advance
organizers. Whether or not these organizers provide short-circuiting depends the type and extent
of the contain they contain. They only short-circuit if they provide most or all of the to-belearned information. In these instances, all the student has to do is memorize the information
provided by the organizers. No transformational mental processes are required, just acquisition
processes; and only some of the acquisition processes are required, such as rehearsal. In addition
to short-circuiting immediate learning, an implicit message is being sent that learning is rote or
associational, rather than requiring problem solving and mental elaboration. The result of shortcircuiting is reduced development of cognitive skills, compared to the amount of development
promoted by either self-regulation or resource management (Corno & Mandinah, 1983). In
contrast, if the roll of the organizer is simply to guide learning (if the organizer provides only key
terms or information, or examples that support and assist learning), short-circuiting does not
occur (Corno & Mandinah, 1983).
Effective strategy use. When confronted with tasks, learners automatically use the
knowledge and skills they have already acquired and are perceived to be relevant to goal
attainment (Locke & Lathan, 2002). There are a number of ways the various cognitive strategies
can be utilized by students and by instructional design to promote learning. In addition,
classroom instruction can be designed to assist learners in gaining and developing cognitive
strategies—to help learners learn to learn. For example, while short-circuiting is generally
viewed negatively from an educational perspective, short-circuiting can serve as a learning tool.
For low or even average ability students, short-circuiting can be beneficial. For these learners,
short-circuiting can provide achievement of immediate, lower-level objectives, thereby
increasing task-specific efficacy expectations for those students. Those students are then more
apt to apply higher order cognitive strategies (Corno & Mandinah, 1983).
For high achieving students, even though they should be more knowledgeable and aware
of effective learning strategies, their use of those strategies is dependent on their perception of
the goal emphasis of the class (Ames & Archer, 1988). Task perceptions also affect high
performer’s strategy selections. According to Corno and Mandinah (1983), in some instances,
high achieving students may prefer low level processes, such as short-circuiting, as a way to
shortcut certain learning requirements. In contrast, these more able students use active mental
approaches for complex tasks. In classroom environments, able learners shift between active and
less active learning processes as interest or task perceptions dictate (Corno & Mandinah, 1983).
According to Corno and Mindinah (1983), able learners have cognitive strategies for
accomplishing tasks that may not be present in the repertoires of less able learners. Less able
learners may approach tasks passively (recipience) or by seeking external assistance (resource
management), because they are unfamiliar with higher order processes. Students must be taught
alternative cognitive engagement strategies, alternatives that are more effective for some tasks
(Corno & Mandinah, 1983). The context of the learning environment as well as the instructional
design can affect development and use of various cognitive strategies. According to Eccles and
WAINESS PHD QUALIFYING EXAM
77
Wigfield (2002), some environments do not allow much latitude in choice of activities or
approaches, making self-regulation more difficult. Corno and Mandinah (1983) added that
learning is less self-regulated when some of the processes are overtaken by classroom teachers,
other students, or features of written instructions (short-circuiting).
Instructional design methods can be utilized to foster not only the use of cognitive
strategies but the development of those strategies and an awareness of when to use them as well.
One example is classroom recitation. According to Corno and Mandinah (1983), classroom
recitation is when a teacher conducts a lesson dialog involving repetition of goals and content,
asking students questions to cognitively engage students and to elicit responses, and responding
to student questions and comments. An advantage of classroom recitation is it can encourage
cognitive engagement on several levels, without enveloping any one instructional strategy long
enough to harm the motivation or performance of students with differing abilities. It may also
restrict the likelihood of any one cognitive engagement strategy becoming automatic or habitual
(Corno & Mandinah, 1983).
Another possible determinate of strategy use is future consequences. Future
consequences are “anticipated and valued distant consequences thought to be at least partially
contingent on task performance but not inherent in the performance itself” (Miller et al., 1996, p.
390). The researchers commented that future consequences contribute to the explanation of
variance in both academic engagement (e.g., effort, strategy use, self-regulation) and
achievement, even when controlling for other goals and perceived abilities. A need to please the
teacher was also found to increase self-regulation, and it covaries with reported use of selfregulatory behaviors such as setting goals, monitoring progress, and making adjustments in study
behavior (Miller et al., 1996). Self-setting goals has also been shown to lead to more task
commitment and better task strategies for learners with high self-efficacy (Locke & Latham,
2002). In general, people with high efficacy are more likely than those with low efficacy to
develop effective task strategies (Lock & Latham, 2002).
Knowledge can be stored in memory in a variety of forms. One way is in isolated and
disconnected pieces of information, often the result of learning by rote. Much of this knowledge
that students acquire in school seems to be in this form. In contract, knowledge can be organized
into large, interconnected bodies, where pieces of knowledge are conceptually linked to other
pieces. This network of interconnections can extend and link to other information to broaden the
range of cognitive activities, such as answering a variety of domain-specific questions, drawing
analogies, making inferences, and generalizing to other domains (Blanton, 1998).
A second aspect of the information-processing approach which is also an integra. part
of an instructional design concerns the activating of relevant background knowledge (Blanton,
1998).
The media selected for the design must be consistent with the operational objectives.
Media can be books, pamphlets, brochures, handouts, slides, film strips, television, computer,
etc. (Blanton, 1998).
Hypertext and hypermedia production tools are making wide use of the graphical user
interface (GUI). This interface operates on the metaphorical premise of direct manipulation and
engagement by the user. Authors of hypermedia constructs are relying heavily on the avaialblity
of windows, icons, menus, and pointer systems in producing and implementing software
presentations (Brown & Schneider, 1992).
WAINESS PHD QUALIFYING EXAM
78
The direct manipulation interface (DMI) is defined as one in which the subject has
direct interaction with their concept world. The subject hs the ability to perceive a direct
connection between the interface and what it represents (Brown & Schneider, 1992).
In a study with eighty-seven elementary school students, grades three through six,
students had little trouble assimilating the direct manipulation interface, and had more difficulty
with the conversational computer interface. While the study examined attitudinal differences
and found a DMI preferable to a conversational interface, learning outcomes were not examined
(Brown & Schneider, 1992).
According to Brunken, Plass, and Leutner (2003), Sweller (1999) distinguished three
types of load: one type that is attributed to the inherent structure and complexity of the
instructional materials and cannot be influenced by the instructional designer, and two types that
are imposed by the requirements of the instruction and can, therefore, be manipulated by the
instructional designer (Brunken, Plass, & Leutner, 2003).
The cognitive load caused by the structure and complexity of the material is called
intrinsic cognitive load. The complexity of any given content depends on the level of item or
component interactivity of the material, that is, the amount of information units a learner needs
to hold in working memory to comprehend the information (Pollock, Chandler, & Sweller, 2002,
as cited in Brunken, Plass, & Leutner, 2003). Cognitive load imposed by the format and manner
in shich information is presented and by the working memory requirements of the instructional
activities is referred to as extraneous cognitive load, a term that highlights the fact that this load
is a form of overhead that does not contribute to an understanding of the materials (Brunken,
Plass, & Leutner, 2003). Finally, the load induced by learners’ efforts to process and comprehend
the material is called germane cognitive load (Gerjets & Scheiter, 2003; Renkl & Atkinson,
2003; as cited in Brunken, Plass, & Leutner, 2003).
According to Brunken, Plass, and Leutner (2003), both extraneous and germane load
can be manipulated by the instructional design fo the learning material (Brunken, Plass, &
Leutner, 2003). Among the instructional strategies that have been found to reduce extraneous
cognitive load and optimize germane cognitive load are worked examples (Kalyuga, Chandler,
Tuovinen, & Sweller, 2001); goal-free activities (Sweller, 1999); and activities that are based on
the completion effect (van Merrienboer, Schuurman, de Croock, & Paas,2 002), modality effect
(Brunker & Leutner, 2001; Mayer & Moreno, 2003; Sweller, 1999), and redundancy effect
(Sweller, 1999) as cited in Brunker, Plass, and Leutner (2003).
Cognitive load theory is based on several assumptions regarding human cognitive
architecture: the assumption of a virtually unlimited capacity of long-term memory, schema
theory of mental representations of knowledge, and limited-processing capacity assumptions of
working memory (Brunken, Plass, & Leutner, 2003).
According to Brunken, Plass, and Leutner (2003), the Baddeley (1986) model of
working memory assumes the existence of a central executive that coordinates two slave
systems, a visuospatial sketchpad for visuospatial information such as written text or pictures,
and a phonological loop for phonological information such as spoken text or music (Baddeley,
1986, Baddeley & Logie, 1999). It is also assumed that both slave systems are limited in capacity
and independent from one another in that the processing capacities of one system cannot
compensate for lack of capacity in the other (Brunken, Plass, & Leutner, 2003).
For ech of the two working memory subsystems, the total amount of cognitive load for
a particular individual under particular conditions is defined as the sum of intrinsic, extraneous,
and germane load induced by the instructional materials. Therefore, a high cognitive load can be
WAINESS PHD QUALIFYING EXAM
79
a result of a high intrinsic cognitive load (i.e., a result of the nature of the instructional content
itself). It can, however, also be a result of a high extraneous or germane cognitive load (i.e., a
result of activities performed on the materials that result in a high memory load). In other wors,
the same learning material can induce different amounts of memory load when different
instructional strategies and designs are used for its presentation, because the different cognitive
tasks required by these strategies and designs are likely to result in varying amounts of
extraneous and germane load (Brunken, Plass, & Leutner, 2003).
If the difference between total cognitive load and the processing capacity of the visual
or auditory working memory approaches zero, then the learner experiences a high cognitive load
or overload (Brunken, Plass, & Leutner, 2003).
The foundation and implications of CLT can be especially well investigated in the
context of multimedia learning, because the use of this technology as instructional medium
involves perceiving and processing information in different presentation modes and sensory
modalities. A process theory that supplements CLT in the description of the cognitive processes
in multimedia learning was introduced by Mayer (2001) as the generative theory of multimedia
learning (Brunken, Plass, & Leutner, 2003).
Two of the principle foundations of the generative theory of multimedia learning are
the dual-coding assumption and the dual-channel assumption (Brunken, Plass, & Leutner, 2003).
The dual-coding assumption refers to the presentation mode of the information and posits that
verbal material (e.g., written and spoken text) and pictorial material (e.g., pictures, graphics, and
maps) are processed and mentally represented in separate but interconnected systems, an
assumption taken from dual-coding theory (Paivio, 1986). The dual-channel assumption refers to
the sensory modality of information perception and proposes that visual information (e.g.,
written text) and auditory information (e.g., spoken text) are processed in different systems that
correspond to the visuospatial and phonological subsystems in Baddeley’s (1986) working
memory model (Brunken, Plass, & Leutner, 2003).
According to Brunken, Plass, and Leutner (2003), the generative theory of multimedia
learning combines these two assumptions with a gerative approach to learning (Wittrock, 1974,
1990) by stating that learners actively select relevant visual and verbal information from the
learning material and organize them in visual and verbal working memory, respectively, by
building associative connections between them (Brunken, Plass, & Leutner, 2003). Learners then
integrate the mental representations as well as prior knowledge by building referential
connections (Mayer, 2001).
The modality effect may best illustrate how these principles allow for design of
multimedia instruction that enhances learning outcomes. Focusing on the sensory modality of
information, this principle states that knowledge acquisition is better facilitated by materials
presented in a format that simultaneously uses the auditory and the visual sensory modalities,
than by a format that uses only the visual modality (Mayer, 2001). Using CLT, the modality
effect can be explained by describing the memory load condition for each of the treatments. The
picture-and-text variant induces a higher load in visual working memory, because both types of
information are processed in this system. The picture-and-narration variant induces a lower
amount of load in visual working memory, because auditory and visual information are being
processed in their respective systems. Thus, the load total load induced by this variant of the
instructional materials is distributed among the visual and the auditory system (Brunken, Plass,
& Leutner, 2003).
WAINESS PHD QUALIFYING EXAM
80
Cognitive load can be treated as a theoretical construct, describing the internal
processes of information processing that cannot be observed directly (Brunken, Plass, & Leutner,
2003). The various methods of assessing cognitive load that are currently available can be
classified along two dimensions, objectivity (subjective of objective) and causal relation (direct
or indirect). The objectivity dimension describes whether the method uses subjective, selfreported data or objective observations of behavior, physiological conditions, or performance.
The causal relation dimension classifies methods based on the type of relation of the
phenomenon observed by the measure and the actual attribute of interest (Brunken, Plass, &
Leutner, 2003). Self-report questionnaires on the amount of mental effort individuals feel they
exerted are an example of subjective-indirect measurements. Self-reports of the difficultly level
of materials are an example of subjective-direct measurements. Analyzing performance
outcomes are an example of indirect-objective measurements. And neuroimaging techniques
(e.g, MRIs), physiological techniques (e.g., papillary response), and dual-task analysis are
examples of objective-direct measurements (Brunken, Plass, & Leutner, 2003).
Human-computer interaction (HCI) as a multidisciplinary and multifaceted area is
strongly influenced by technological, organizational, and socioeconomic factors (Bullinger,
Ziegler, & Bauer, 2002).
Preferential selection refers to choices that we make about what is, and about what is
not, attention-worthy (Calvert, Watson, Brinkley, & Bordeaux, 1989).
The purpose of thise study was to examine the effects of presentational features on
children’s preferential select and memory for information presented in oral story format and
depicted in a computer microworld. As expected, preschoolers preferentially selected and
recalled words that had been presented with moderate levels of actions better than words that had
been presented with no action (Calvert, Watson, Brinkley, & Bordeaux, 1989). Action was both
inherently interesting to children, as demonstrated by their preferential selection scores, and
memorable to children, as demonstrated by their free recall scores (Calvert, Watson, Brinkley, &
Bordeaux, 1989).
Objects presented without sounds were better recalled than objects presented with
sounds (Calvert, Watson, Brinkley, & Bordeaux, 1989).
Sex differences in children’s preferential selection scores suggested that action is more
inherently interesting to boys than to girls (Calvert, Watson, Brinkley, & Bordeaux, 1989).
How easily users, or learners in the case of educational technology, become disoriented
in a computerized text may be a function of the user interface (Chalmers, 2003).
One area where disorientation can be a problem is in the use of links. Links enable
users to expand their knowledge to include thousands of related topics. Although links create the
advantage of exploration, there is always the chance that the explorer may get lost, not knowing
where they were, where they are going, or where they are (Chalmers, 2003).
Learning theories have traditionally been applied to venues of instruction such as
textbook instruction, classroom instruction, and one-on-one tutoring. However, it cannot be
assumed that learning theories applied to these venues can automatically be applied to learning
with computers (Chalmers, 2003).
Schemas are generally though of as ways of viewing the world and in a more specific
sense, ways of incorporating instruction into our cognition. According to Chalmers (2003),
Satzinger (1998) described schema theory to include knowledge structures that concepts in
human memory, including procedural knowledge of how to use the concepts (Chalmers, 2003).
WAINESS PHD QUALIFYING EXAM
81
Piaget proposed that learning is the result of forming new schemas and building upon
previous schema (Chalmers, 2003).
Scaffolding is a term used to describe the process of forming and building upon a
schema. Interface scaffolding refers to a schema support for computer-assisted learning. A key
component of one kind of interface scaffolding is that it can be made fadeable. That is, interface
scaffolding can be faded in or out as needed. This fading can be a function of the learner or the
computer. In learner induced fading, learners describe whether or not to show the scaffold. The
trouble with this idea is that learners may not make good decisions about which scaffolding to
show and which scaffolding to hide. In computer induced fading, the computer decides whether
or not to fade the scaffolding, based on a model of the learner’s understanding. The main
problem with this approach is that an extensive model of the learner’s knowledge may be hard to
specify or evaluate in more open ended domains (Chalmers, 2003).
In addition to schemas, another closely related cognitive learning theory is that of
cognitive load. Cognitive load is a term used to describe the amount of information processing
expected of the learner. Intuitively, it makes sense that the less cognitive load a learner has to
carry,t he easier learning should be. In fact, researchers have proposed that working memory
limitations can have an adverse effect on learning (Sweller, 1993; Sweller and Chandler, 1994,
Yeung, 1999; as cited in Chalmers, 2003).
In addition to learning, students also need to retain information, if they are to use their
knowledge beyond the learning situation. Retention refers to the amount of knowledge which can
be remembered after a given amount of time (Chalmers, 2003). Retention can be subdivided into
two types, depending on the amount of time which as elapsed between the point of learning and
the point of recall. These subdivisions are called short-term retention (i.e., in working memory)
and long-term retention, (i.e., in long-term memory; Chalmers, 2003). Short-term retention is
assessed during or immediately after the material has been presented. Long-term retention is
assessed at least one week after material has been presented (Chalmers, 2003).
To enhance retention, a number of techniques have been suggested. One of these
techniques is chunking; that is, to group the multiple pieces of information into chunks
(Chalmers, 2003).
Good screen design leads to completing lessons in less time and with a higher
completion rate (Chalmers, 2003).
Outline organizers may be presented in the form of an agenda before a tutorial or
lecture (Chalmers, 2003).
Post organizers are used to help learners summarize information. They can appear in
the form of a summary at the end of a chapter or lecture (Chalmers, 2003).
Graphic organizers are organizers of information in a graphic format. Graphics
organizers can be described as spatial displays of text information that can be provided to
students as study aids that accompany text (Chalmers, 2003).
A continuous organizer is an organizer that is continuously updated and context
sensitive (Chalmers, 2003).
Concepts maps contain nodes and linkages to identify interrelationships between
pieces of knowledge. The learner generates concept maps (Chalmers, 2003).
Through hypertext, using associative links and taking advantage of the structure of the
information, learners are encouraged to explore and find the information they need, then progress
to other learning activities (Chou & Lin, 1998).
WAINESS PHD QUALIFYING EXAM
82
After initial information needs have been satisfied, the next stage is knowledge
acquisition—integrating new knowledge with existing knowledge (Chou & Lin, 1998).
One hundred twenty one students (98 males and 23 females) from two mid-sized
universities in northern Taiwan in a required freshman Introduction to Information Technology
course participated in the experiment using a hypertext learning course requiring a search task
and three map types: a global map, a local map, and a local tracking map. The global map
showed the entire hierarchical structure, listing the names of the 94 nodes contained in the
course. The names represented the concept taught within the node, and a tree-like overview map
provided the conceptual structure of the information. Using the map, learners could find where
they were, where they had visited, and where they had not visited (Chou & Lin, 1998). The local
maps could be described as parts of the global map showing particular knowledge areas in the
course. They focused only on neighborhoods of activated nodes, that is, one level above and two
levels below the current node (or concept). Users were always in the current map, but did not
know exactly where they were in the overall courseware. Local maps were updated once a user
moved to a node outside the current local map (Chou & Lin, 1998). The local tracking map was
similar to the local map, but always showed the activated node in the center of the may in a
“You-are-her” fashion. The local tracking maps were updated whenever the users moved to other
nodes (Chou & Lin, 1998). The experiment included five treatments, one treatment for each of
the three map types, one no-map group, and one group that received all three maps (Chou & Lin,
1998).
Map type caused significant effects on the subjects search steps, search efficiency, and
development of cognitive maps. Subjects in the global map and all-map groups took fewer steps
(jumping from one node to another using either the global map or hot keys) than subjects in the
other three groups. The all-maps group used the global map 84% of the time they used a map,
therefore it was concluded that the global map helped learners find particular information in
fewer steps (Chou & Lin, 1998).
The search steps for the local map and tracking local map groups was similar to the no
map group, suggesting that the limited scope of the local maps gave them little advantage of
having no map (Chou & Lin, 1998).
The search efficiency of the global and all-maps groups was better than the other three
groups (Chou & Lin, 1998).
Given no time limit, the map type, including having no map, did not affect search-task
completion (Chou & Lin, 1998), suggesting that map usage was not beneficial to task completion
but was to task efficiency and speed (Chou & Lin, 1998).
The problems of cognitive overhead and disorientation are inter-related. Cognitive
overhead is the additional mental effort learners must make in order to choose which links to
follow and which to abandon from a large number of options. No knowing one’s location and
continual decision making can be distracting and complicate the learning journey in a hypertext
environment. These two problems can become even more serious if the hypertext system as a
large number of nodes and links (Chou, Lin, & Sun, 2000).
Sixty four students (51 males and 13 females) from a mid-sized university in northern
Taiwan participated in a study, using a hypertext course for a search task, with next and previous
buttons for moving forward or backward a screen at a time, and hot keys for jumping to specific
screens. Two types of maps were provided: global maps and local maps. There were three
randomly assigned groups: global map, local map, and no map. Participants were assessed on
WAINESS PHD QUALIFYING EXAM
83
completion of the search task and on their ability to create concept maps of the relationships
among the course nodes, that is, concepts (Chou, Lin, & Sun, 2000).
Map type significantly affected subjects’ search steps, revisitation ration, hyper
jumping, and cognitive map development. Subjects in the global map group took fewer steps
(jumping from one node to another using either the global map or the hot keys), then subjects in
the other two groups (Chou, Lin, & Sun, 2000).
The search steps for the local map group was not significantly different from those of
the no map group (Chou, Lin, & Sun, 2000).
Subjects in the global map group has a lower re-visitation ration and lower hyperjumping scores than those in the other two groups (Chou, Lin, & Sun, 2000).
The no map group has the highest mean score for cognitive map development,
followed by the global map group, then the local map group (Chou, Lin, & Sun, 2000). The
difference between the no map and the global map groups was not significant, but the difference
between both these groups and the local map group was significant (Chou, Lin, & Sun, 2000).
Once we are committed to a goal, we must make a plan to achieve a goal. A key
element of all goal-directed planning is our personal assessment of the necessary skills and
knowledge required to achieve a goal. A key aspect of self efficacy assessment is our perception
of how novel and difficult the goal is to achieve. The ongoing results of this analysis is
hypothesized to determine how much effort we invest in the goal (Clark, 1999).
It is presumed that if we perceive a task as very difficult, that perception reflects an
analysis of our own task-relevant skills. The usual solution to such perceptions is to attempt to
increase self efficacy or self regulation perceptions and therefore reduce the perceptions that
tasks are difficult. The problem with this strategy is that people when people lack knowledge or
skills, no increase in their self efficacy alone, apart from a concurrent increase in knowledge and
skills, will increase performance (Clark, 1999).
Effort is primarily influences by specific and detailed self efficacy assessments of the
knowledge required to achieve tasks (Clark, 1999).
Automated expertise, developed over many hundreds of hours of practice, requires no
cognitive effort to experess (Clark, 1999).
Effort diminishes at either exceptionally low or high self efficacy levels and the
relationship between self efficacy and effort follows the shape of an inverted “U” (Clark, 1999).
According to Clark (1999), Sweller (1988;1994) has considerable evidence that when
the “cognitive load” of a task exceeds the capacity of working memory, effort ceases. Clark
further commented that Paas and Van Merrienboer (1993) have provided evidence that excessive
cognitive load reduces both mental effort and performance (Clark, 1999).
The more novel the goal is perceived to be, the more effort will be invested until we
believe that we might fail. At the point where failure expectations begin, effort is reduces as we
“unchoose” the goal to avoid a loss of control. This inverted U relationship suggests that effort
problems take two broad forms: over confidence and under confidence (Clark, 1999).
The level of mental effort necessary to achieve work goals can be influenced by
adjusting perceptions of goal novelty and the effectiveness of the strategies people use to achieve
goals (Clark, 1999).
Motivation generates the mental effort that drives us to apply our knowledge and skills.
Without motivation, even the most capable person will not work hard (Clark, 2003).
Motivation is the result of our beliefs about what makes us successful and effective
(Clark, 2003).
WAINESS PHD QUALIFYING EXAM
84
Easy goals are not motivating (Clark, 2003).
Peoples’ belief about whether they have the skills required to succeed at a task is
perhaps the most important factor in the quality and quantity of mental effort people invest in
their work (Clark, 2003).
People will more easily and quickly choose to do what interests them (Clark, 2003).
The term metacognition continues to be used in two distinct ways: the conscious and
purposeful reflection on various aspects of knowing and learning, and the unconscious regulation
of knowledge structures and learning that some information-processing theorists posit to be
under the control of executive processes (Clements & Nastasi, 1999).
The basic contention of achievement goal theory is that, depending on their subjective
purposes, achievement goals differentially influence school achievement via variations in the
quality of cognitive self-regulation processes (Covington, 2000).
Cognitive self-regulation refers to students being actively engaged in their own
learning, including analyzing the demands of school assignments, planning for and mobilizing
their resources to meet these demands, and monitoring their progress toward completion of
assignments (Covington, 2000).
There are several cognitive and other factors that may be important in using VEs.
These include individual differences that may affect efficient use of VEs , the effectiveness of
passive exploration of a VE as opposed to active exploration, the kinds of features or cues within
a VE that facilitate tracking one’s position during movement through the VE (i.e., navigation),
and adverse sensory factors associate with immersion (Cutmore, Hine, Maberly, Langford, &
Hawgood, 2000).
The term navigation refers to a process of tracking one’s position in a physical
environment to arrive at a desire destination. A route through the environment consists of either a
series of locations or a continuous movement along a path (Cutmore, Hine, Maberly, Langford,
& Hawgood, 2000).
Navigation becomes problematic when the whole path cannot be viewed at once but is
largely occluded by objects in the environment. These can include walls or large environmental
objects such as trees, hills, or buildings. Under these conditions, once cannot simply plot a direct
visual course from the start to finish locations. Rather, knowledge of the layout of the space is
required. Maps or other descriptive information may provide this knowledge (Cutmore, Hine,
Maberly, Langford, & Hawgood, 2000).
Effective navigation of a familiar environment depends upon a number of cognitive
factors. These include working memory for recent information, attention to important cues for
location, bearing and motion, and finally, a cognitive representation of the environment which
becomes part of a long-term memory, a cognitive map (Cutmore, Hine, Maberly, Langford, &
Hawgood, 2000). This representation is what permits the local perceptual cures to be of use in
tracking or maintaining a sense of “knowing where one is.” The representation also permits
generation of expectancies for encountering future landmark (Cutmore, Hine, Maberly,
Langford, & Hawgood, 2000).
A cognitive map makes some aspects or attributes of the world explicit, while
permitting other aspects to be computed or approximated as needed (i.e., they are implicit;
Cutmore, Hine, Maberly, Langford, & Hawgood, 2000).
The interaction of cognitive style and environment type indicates that visual-spatial
ability is an important predictor of navigation performance in the absence of flow-field
WAINESS PHD QUALIFYING EXAM
85
information, but when this information is provided, the cognitive style groups perform similarly
(Cutmore, Hine, Maberly, Langford, & Hawgood, 2000).
In a series of five maze experiments, Cutmore, Hine, Maberly, Langford, and
Hawgood (2000) found the following. A simple VE that presented the human with nothing more
than a series of bare frames, each providing a view of a VE maze room, supported the acquisition
of spatial knowledge about the VE (Cutmore, Hine, Maberly, Langford, & Hawgood, 2000).
Compass headings did not seem to help exit finding tasks. Also, while landmarks provided useful
cues, males utilized them significantly more often than females (Cutmore, Hine, Maberly,
Langford, & Hawgood, 2000).
In the field of psychology, there has been a demise of the behaviorist view in favor of
the cognitive view of learning. A behaviorist view of learning emphasizes teaching strategies that
involve repetitive conditioning of learner response. A cognitive view places importance on the
learners’ cognitive activity and the mental models they form (Dalgarno, 2001).
In the field of psychology, there has been a gradual rejection of the assumption, help
by many cognitivists, that there is some objectively correct knowledge representation. The
alternative view, termed constructivist, is that, within a domain of knowledge, there may be a
number of individually constructed knowledge representations that are equally valid. The focus
of teaching then becomes one of guiding learners are they build on and modify their existing
mental models, that is, a focus on knowledge construction rather than knowledge transmission
(Dalgarno, 2001).
Typically, a simulation is defined as a model of a real workd environment (Dalgarno,
2001), while a microworld is defined as a model of a concept space, which may be a very
simplified version of a real world environment, or it may be a completely abstract environment
(Dalgarno, 2001).
The existence of bookmarks is important not exactly to avoid disorientation problems
but mainly to enable recovering from an eventual possibility of disorientation. The bookmark
mechanisms allow the user to mark a node in the hypermedia document so that he can reach that
node at any time during the navigation process and from any point of the hypermedia document
(Dias, Gomes, & Correia, 1999). History lists can also be used (Dias, Gomes, & Correia, 1999).
With one eye on the future, many educators and literary scholars are predicting nothing
less than a paradigm shift in the manner in which we understand the learning experience and the
education process as a result of hypermedia technologies in general and the World Wide Web in
particular (Dillon & Gabbard, 1998).
Previous researchers have buried navigational elements within their studies of the
affects of learner control and interactivity on achievement and attitude. Consequently, the result
of the many studies are conflicting. With the definition of learner control there also exist
elements of interactivity and navigation. These terms are not synonymous, although the research
sometimes treats them as such. Interactivity implies a relationship between the learner and the
instructional module with varying degrees of engagement. Navigation is a function of
interactivity along with feedback, pacing, inquiry, and elaboration. The presence of interactivity
creates an opportunity for navigation (Farrell & Moore, 2000/2001).
Cognitive load “the total amount of mental activity imposed on working memory at an
instance in time” (Cooper, 1998)
The major factor that contributes to cognitive load is the number of elements that
need to be attended to at any one time during learning (Cooper, 1998).
WAINESS PHD QUALIFYING EXAM
86
Message complexity, stimulus features, and additional cognitive demands inherent in
hypermedia may combine to exceed the cognitive resources of some learners (Daniels & Moore,
2000).
Giving learners control and autonomy over an environment can either facilitate
learning or lead to disorientation and confusion (Dias, Gomes, & Correia, 1999).
Learner control: Generically: “Student choice of practice items, reviewing and
feedback.(Niemiec, Skorski, & Walberg, 1996, p. 158). Specific to computers: “…giving learners
control over elements of a computer-assisted instructional program. (Hannafin & Sullivan, 1995,
p. 19).
Navigation: the paths learners chose to view information to accomplish various
cognitive and learning tasks; a relational property among parts of a systems (individual, task,
hypermedia, learning context) (Barab, Bowdish, & Lawless, 1997).
[Navigation] determines the amount and quality of information retrieved from a
hypermedia source (Farrell & Moore, 2000, p. 170).
Cognition: the intellectual processes through which information is obtained,
represented mentally, transformed, stored, retrieved, and used.
Learning: The product of the interaction among what learners already know, the
information they encounter, and what they do as they learn(Bruning, Shraw, & Ronning, 1999, p.
6).
Hypermedia: …environments in which the information representation and
management system is developed around a network of multimedia nodes connected by various
links. (Barab, Bowdish, & Lawless, 1997, p. 23). …a generic term covering hypertext,
multimedia, and related applications involving chunking of information into nodes that could be
selected dynamically (McKnight, Dillon, & Richardson, 1991).
The evidence suggests that the instructional methods and information content, not
media, improve learning (Clark, 2001) Instructional methods are external representations of
internal cognitive processes that are necessary for learning but which learners cannot or will not
provide for themselves (Clark, 2001) Examples of instructional methods: Provide learning goals.
Use examples (e.g., demonstrations, simulations and analogies). Monitor (in the form of practice
exercises, and tests). Feedback (synchronous or asynchronous). Selection (highlighting important
information)
Novice and lower aptitude students have the greatest difficulty with hypermedia
(Dillon & Gabbard,1998).
Control does not appear to offer special benefits for particular learners or under
specific conditions (Niemiec, Sikorski, & Walberg).
More positive attitude can indicate less learning (Salomon, 1984).
Recent evidence that some collaborative group work results in learning losses for some
participants (Stipek, 2003).
Media, including CBI and Distance Learning, does not increase learning (Bernard et al,
AECT 2003).
Six extensive meta analyses of distance and media learning studies in the past decade
have found the same negative or weak results (see Bernard, et al, 2003)
Use navigation maps (Chou & Lin, 1998; Ruddle et al, 1999, Chou, Lin, & Sun, 2000)
Use menus (Benbasat & Todd, 1993; Farrell & Moore, 2000-2001)
Grouping, location, color, concreteness, complexity, distinctiveness of icons
(McDougall et al., 2000; Niemela & Saarinen, 2000; Zammit, 2000).
WAINESS PHD QUALIFYING EXAM
87
Use simple control in support of theory.
Interactivity improved learner understanding only when it was used in a way that
minimized cognitive load and allowed for two-stage construction of a mental model (Mayer &
Chandler, 2001).
Part-Whole is better than Whole-Part for transfer.
Part-Part and Part-Whole are better than Whole-Whole for transfer.
Intelligence is not a function of how hard the brain works, but rather how efficiently it
works. This efficiency may derive from a more focused use of brain areas relevant for good task
performance (Gerlic & Jausovec, 1999).
Gerlic and Jausovec (1999) conducted EEG studies of brain activity during multimedia
performance. The content that results of the study showed a clear difference between multimedia
presentations and text presentations. The video and picture presentations increased activity of the
occipital and temporal lobes. The text presentation increased activity of the frontal lobes. Finding
from this study support prior medical findings that one of the basic functions of the temporal
cortex is the processing of auditory input, while the exclusive function of the occipital lobes is
vision. It is also believed that the occipital cortex is involved in imagery. The prefrontal cortex
appears to be involved in controlling and monitoring our thoughts and actions, and the frontal
lobes control working memory (Gerlic & Jausovec, 1999). These findings suggest that it is
reasonable to believe that multimedia presentations trigger visualization strategies such as mental
imagery, which is critical to many kinds of problem solving and discovery (Rieber, 1995, as
cited in Gerlic & Jausovec, 1999).
Another explanation for the reported differences could be that video and picture
presentations increased occipital activity because they included visual material, whereas the text
presentation had no such material. However, the authors cite a number of reasons that this is a
less plausible explanation of the differences (Gerlic & Jausovec, 1999).
The EEG study also showed that gifted individuals exhibited lower mental activity
when involved in learning the material. These differences were more pronounced for the video
presentation than the text presentation, and could indicate a tendency that multimedia
presentations are less effective for gifted students. However, one must bear in mind that the tasks
used in the present study were rather simple involving only knowledge about facts. It is
questionable if a similar trend would have been obtained for more complex information to be
learned (Gerlic & Jausovec, 1999).
Working memory refers to the limited capacity for holding information in mind for
severl seconds in the context of cognitive activity (Gevins, Smith, Leong, McEvoys, Whitfield,
Du, & Rush, 1998).
Overload of working memory has long been recognized as an important source of
performance errors during human-computer interaction and is particularly acute in unskilled
users for whom unfamiliar procedures are likely to require greater commitment of cognitive
resources. Furthermore, overload of working memory capacity had been found to be a limiting
factor in the early stages of procedural skill acquisition. As a result, the need to minimize
working memory load has been cited as a primary guiding principle for the design of intelligent
tutoring systems (Gevins, Smith, Leong, McEvoys, Whitfield, Du, & Rush, 1998).
“As shared symbol systems, media are potent cultural tools for the selective sculpting
of profiles of cognitive processes” (Greenfield, Brannon, & Lohr, 1994, p. 87).
WAINESS PHD QUALIFYING EXAM
88
A medium is not simply an information channel; as a particular mode of
representation, it is also a potential influence on information processing (Greenfield, Brannon, &
Lohr, 1994, p. 88).
Each medium has its particular design features such that it presents certain kinds of
information easily and well and other kinds poorly and with difficulty. Each medium, therefore,
presents certain opportunities to construct particular kinds of representations. As a consequence,
each medium stimulates different kinds of representational processes; it provides a particular
kind of cognitive socialization (Greenfield, Brannon, & Lohr, 1994).
From the point of view of development and socialization, video games are particularly
important because they affect children during the formative years of childhood, when
socialization is taking place (Greenfield, Brannon, & Lohr, 1994).
Video games go beyond print and photography in their presentation of twodimensional representations of three-dimensional space. The consumer must be able to interpret
not only static two-dimensional images into three-dimensional space, but dynamic images as
well. Additionally, the user must not only interprets, but also mentally transform, manipulate,
and relate dynamic and changing images (Greenfield, Brannon, & Lohr, 1994). It is the transfer
of this skill to spatial contexts outside the game that is the focus of the present research. The
question is: Can video game practice develop transferable skills in manipulating threedimensional spatial representations (Greenfield, Brannon, & Lohr, 1994)?
Although we were not able to demonstrate the predicted experimental effect of shortterm practice of a game on mental paper folding, we were able to show a causal relationship of
expertise gained over long-term and mental paper folding (Greenfield, Brannon, & Lohr, 1994).
Effective use of mnemonic strategies has been characterized as developing through
three stages. During the first stage, children are not capable of utilizing the strategy effectively.
This difficulty is referred to as a mediational deficiency. During the second stage, children still
do not use the strategy spontaneously; however, they are now capable of using the strategy
effectively if specifically instructed to do so. This failure of children to spontaneously utilize a
strategy which they are actually capable of using is referred to as a production deficiency. The
final stage involves mature use of the strategy, by which time children produce the strategy
spontaneously while performing strategy-appropriate tasks (Guttentag, 1984).
In a study of second graders, it was found that the mental effort requirement of
instructed cumulative rehearsal was significantly greater for production deficient children than
for children who normally utilized a cumulative rehearsal strategy spontaneously. One possible
explanation for this finding is that the decrease with age in the mental effort required of strategy
use resulted form an increase with age in spontaneous use of the strategy. That is, because
practice generally decreases the mental effort requirement of task performance, the children who
use a cumulative strategy spontaneously may simply have been more highly practiced at using
the strategy than were the children classified as production deficient (Guttentag, 1984).
Alternatively, the mental effort requirement of strategy use may be one factor affecting
children’s strategy selection. That is, there may be a tendency for children to avoid using
strategies which require a very large expenditure of mental effort on their part (Guttentag, 1984).
Haggas and Hantula (2002) conducted a study with university students to determine
the effects of covert and overt computer responses to performance. An example of a covert
format would be “THINK of the correct answer. When you have though of the correct answer,
click the READY button to see if you were right.” Clicking the READY button brought up
feedback: “The correct answer is [ ].” In the overt format, the display was “CLICK on the correct
WAINESS PHD QUALIFYING EXAM
89
answer.” Clicking on the answer brought up differential feedback. A correct answer called up
feedback in black text with an orange box, such as, “Answer choice [ ] is CORRECT!!”. An
incorrect answer called up feedback in black text with a light blue box, for example, “Incorrect.
The correct answer is [ ]” (Haggas & Hantula, 2002).
The majority of participants showed preference for the overt format, and the difference
between time taken to complete covert and overt questions was not significant. However, a
negative relationship was found between the time taken to complete the program and the number
of overt questions answered quickly (Haggas & Hantula, 2002).
The seductive detail effect is the reduction of retention caused by the inclusion of
extraneous details (Harp & Mayer, 1998). Seductive details are details that not part of the to-belearned material but tend to enhance the presentation of the material.
Many models of learning (e.g. the CANE model: Clark, 1999) include the executive
processes of selecting, organizing, and integrating.
Selecting involves paying attention to the relevant pieces of information in the text
(Harp & Mayer, 1998).
Organizing involves building internal connections among the selected pieces of
information, such as causal chains (Harp & Mayer, 1998).
Integrating involves building external connections between the incoming information
and prior knowledge existing in the learner’s long-term memory (Harp & Mayer, 1998).
It is believe that seductive details interfere with some or all of these three
metacognitive processes (Harp & Mayer, 1998).
According to the distraction hypothesis, seductive details do their damage by
“seducing” the reader’s selective attention away from the important information. A possible
solution is to leave the details, but guide the learner away from them and to the relevant
information (Harp & Mayer, 1998).
According to the disruption hypothesis, seductive details are damaging because the
interrupt the transition from one main idea to the next. In order for the reader to be able to
construct a coherent mental model of the chain of events leading to the formation of lightning,
links between the steps in the causal chain must be constructed in working memory. Because
seductive details are presented between the steps of the causal chain, the reader is not able to see
how to link the steps. As a result, the learner interprets each step as a separate, independent
event, rather than as part of a causal chain. A way to solve the problem and keep the seductive
details is to provide support that helps the reader to more effectively organize the important main
ideas. For example, rewriting a passage by using organizational signals such as preview
sentences and number signals in a passage about a process should help the reader to realize the
steps explained in the passage are related to one another (Harp & Mayer, 1998).
According to the diversion hypothesis, the learner builds a representation of the text
organized around the seductive details, rather than around the important main ideas contained in
the lesson. In this case, seductive details prime the activation of inappropriate prior knowledge as
the organizing schema for the lesson. If the diversion hypothesis is correct, then revising a lesson
by presenting all of the irrelevant information at the beginning of the lesson would exacerbate
the seductive details effect. Conversely, revising the passage by moving the seductive details to
the end of the lesson would result in reducing the seductive details effect, by coming after the
important information, and therefore, too late to become the central component of the schema
(Harp & Mayer, 1998).
WAINESS PHD QUALIFYING EXAM
90
In their experiment on seductive details, Harp and Mayer (1998) found that the
diversion hypothesis is the most likely explanation for the effect. Therefore, one way to
discourage inappropriate schema activation is to delay the introduction of seductive information
until after the reader has processed the information material. Another way, of course, is to simply
not introduce seductive details at all (Harp & Mayer, 1998).
Many educators believe that young children do not have the cognitive capacity to
interact and make sense of the symbolic representations of computer environments. Early
childhood educators believe that young children learn best by investigating with their senses, by
examining that which is tactile and tangible (Howland, Laffey, & Espinosa, 1997).
Simply because use of computers may be categorized as a concrete activity, we cannot
assume this means that children’s involvement with computers necessarily results in high quality
learning (Howland, Laffey, & Espinosa, 1997).
The experiential mode is reactive and automatic, resulting in a response without
conscious thought. Because the relevant information needed for decision-making already exists
in our memory, our actions can be driven by the events as they occur. Computer games and drill
and practice computer lessons result in this type of cognition (Howland, Laffey, & Espinosa,
1997).
The reflective mode, on the other hand, is reasoned and conceptual, allowing the
thinker to consider various alternatives. This type of explorative and discover orientation is at the
heart of the developmentally appropriate practices we hope will take place in primary education
(Howland, Laffey, & Espinosa, 1997).
During “event-driven” computer games, one engages in the experiential mode by
immersion in the recurring challenges and events. Although experiential learning can be a good
motivator, the act of experiencing can easily become the sole outcome, with little or no actual
thinking, connecting to other concepts, or generating new ideas (Howland, Laffey, & Espinosa,
1997).
The challenge of computer-using primary educators is to find and use computational
environments that meet the requirements of presenting “meaningful and manipulable”
developmentally appropriate activities which do not simply rely on experiential cognition which
may defeat the educational purpose of the activity (Howland, Laffey, & Espinosa, 1997).
Intensity and direction are two variables of motivation which can be influenced by
internal characteristics of the task or by extrinsic outcomes, such as rewards or praise (Howland,
Laffey, & Espinosa, 1997). While computer games can provide fantasy, and while fantasy has
been defined as a motivating factor (Malone, 1980), different children are drawn to different
fantasies. While one child may be intrigued by the pretend notion of being a deep sea diver,
another child might find that particular fantasy irrelevant to his own fantasy preferences
(Howland, Laffey, & Espinosa, 1997).
Curiosity also plays a role in motivation. An environment which is too simple to a
child will fail to spark curiosity just as surely as one which proves too difficult (Howland,
Laffey, & Espinosa, 1997). This point is similar to Vygotsky’s (1978) zone of proximal
development (Howland, Laffey, & Espinosa, 1997).
An optimal computer environment for learning might be one that matches the
motivating factors of fantasy and curiosity with the child’s motivation toward mastery and
competence (Howland, Laffey, & Espinosa, 1997).
It is incorrect to consider language as correlative of thought; language is a crrelative of
unconsciousness. The mode of language correlative to consciousness is meanings. The work of
WAINESS PHD QUALIFYING EXAM
91
consciousness with meanings leads to the generation of sense, and in the process consciousness
acquires a sensible (meaningful) structure (Hudson, 1998).
The interface of a contemporary CBI program frequently can be likened to a control
panel from which users access information in an oftentimes sophisticated and complicated piece
of software (Jones, Farquhar, & Surry, 1995).
One theoretical construct in the field of cognitive psychology is the notion of cognitive
strategies. Cognitive strategies include rehearsal strategies, elaboration strategies, organization
strategies, affective strategies, and comprehension monitoring strategies. These strategies are
cognitive events that describe the way in which one is processing information (Jones, Farquhar,
& Surry, 1995).
Metacognition is a type of cognitive strategy that has executive control over other
cognitive strategies. In the context of learning through a computer-based learning environment,
metacognition refers to the activities of a user when monitoring, regulating, and orchestrating
learning processes (Jones, Farquhar, & Surry, 1995). Strategy selection, attention, goal setting,
and goal checking are four individual strategies within metacognition. These categories can be
grouped into two major categories: (1) control processes and (2) monitoring processes (Jones,
Farquhar, & Surry, 1995).
Control processes: As the executive controller of cognitive processes, metacognition
select the appropriate strategy for the task at hand. The selection of a cognitive strategy depends
upon the individual’s understanding of the current problem or cognitive situation. Personal
experiences in solving similar tasks and using various strategies will affect the selection of a
cognitive strategy (Jones, Farquhar, & Surry, 1995).
Control processes: To aid in the learner’s attention to the content, an individual cn also
choose to attend to particular cognitive strategies. This strategy, attention, is important in followthrough, completing, and correctly performing the steps of subordinate cognitive strategies
(Jones, Farquhar, & Surry, 1995).
Monitoring processes: Cognitive processes such as learning and problem solving begin
with the identification of a goal. In learning, this might be an understanding of a particular topic.
In problem solving, the goal would be to find a solution (or the best solution) to the problem
(Jones, Farquhar, & Surry, 1995).
Goal setting guides the cognitive strategies in a certain direction. Goal checking are
those monitoring processes that check to see if the goal has been accomplished, or if the selected
strategy is working as expected. The monitoring process is active throughout an activity and
constantly evaluates the success of other processes. If a cognitive strategy appears not bo be
working, an alternative may then be selected (Jones, Farquhar, & Surry, 1995).
Metacognitive guidance includes many familiar methods, such as advance organizers,
graphic representations of problems, and hierarchical knowledge structures. These instructional
methods should be used to aid the novice in developing an expert’s awareness of the problem
space. Teaching the student problem space representational skills may be the most effective way
to turn a “poor” novice problem solve into a “good” novice problem solver (Jones, Farquhar, &
Surry, 1995).
Metacognition is the “management” of though processes as one learns and solves
problems. Learners using CBI are presented with large amounts of information and asked to
manage that information to solve a particular problem or learn about a particular topic. In order
to assist the management of information, the interface should provide users with relevant data
WAINESS PHD QUALIFYING EXAM
92
about the program, how to use the program, where they are in the program, and how well they
are doing (Jones, Farquhar, & Surry, 1995).
Not every program needs a metaphor. Not all programs can support a metaphor. Study
the content carefully and decide what the program is intended to do. Providing users with a
theme can be more helpful than a forced or inappropriate metaphor (Jones, Farquhar, & Surry,
1995).
If a metaphor can be used, use a metaphor that reflects the program’s content. Users
should not have to learn the meaning of a metaphor along with the content of the program (Jones,
Farquhar, & Surry, 1995).
Provide maps so that users can find where they are, and allow provisions to jump to
other information of interest from the map (Jones, Farquhar, & Surry, 1995).
Provide visual effects to give users visual feedback that their choices have been made
and registered by the program (Jones, Farquhar, & Surry, 1995).
Designers should provide users with visual or verbal cues to help them navigate
through unfamiliar territory. Overviews, menus, icons, or other interface design elements within
the program should serve as advance organizers for information contained in the program (Jones,
Farquhar, & Surry, 1995).
Provide cues such as maps and menus as advance organizers to help users
conceptualize the organization of the information in the program (Jones, Farquhar, & Surry,
1995).
Metacognition, or the management of cognitive processes, involves goal-setting,
strategy selection, attention, and goal checking (Jones, Farquhar, & Surry, 1995).
The user interface can indicate the content covered in the program through the user of
advance organizers such as menus. Placing checkmarks after visiting a section will indicate to
the user sections that have been visited. However, it should be noted that visiting a section
doesn’t necessarily mean that the learner viewed or engaged in the content (Jones, Farquhar, &
Surry, 1995).
In this article, we survey evidence that a large number of cognitive load theory (CLT)
effects that can be used to recommendation instructional design are, in fact, only applicable to
learners with very limited experience (Kalyuga, Ayers, Chandler, & Sweller, 2003). With
additional experience, specific experimental effects can first disappear and then reverse. As a
consequence, the instructional design recommendations that flow from the experimental effects
also reverse (Kalyuga, Ayers, Chandler, & Sweller, 2003). We call the reversal of cognitive load
effects with expertise the expertise reversal effect. Like all cognitive load effects, it originates
from some of the structures that constitute human cognitive architecture (Kalyuga, Ayers,
Chandler, & Sweller, 2003).
Working memory limits profoundly influence the character of human information
processing (Kalyuga, Ayers, Chandler, & Sweller, 2003). Only a few elements (or chunks) of
information can be processes at any time without overloading capacity and decreasing the
effectiveness of processing. Conversely, long-term memory contains huge amounts of domainspecific knowledge structures that can be described as hierarchically organized schemas that
allow us to categorize different problem states and decide the most appropriate solution moves
(Kalyuga, Ayers, Chandler, & Sweller, 2003).
Controlled use of schemas requires conscious effort, and therefore, working memory
resources. However, after having being sufficiently practiced, schemas can operate under
automatic, rather than controlled, processing. Automatic processing of schemas requires minimal
WAINESS PHD QUALIFYING EXAM
93
working memory resources and allows for problem solving to proceed with minimal effort
(Kalyuga, Ayers, Chandler, & Sweller, 2003).
CLT (see Sweller, 1999, and Sweller, van Merrienboer, & Paas, 1998, for recent
reviews) is based on the assumptions that schema construction and automation are the major
goals of instruction, but these goals can be thwarted by the limited capacity of working memory.
Because of the limited capacity working memory, the proper allocation of available cognitive
resources is essential to learning (Kalyuga, Ayers, Chandler, & Sweller, 2003).
Experts possess a larger (and potentially unlimited) number of domain-specific
schemas. Hierarchically organized schemas represent experts’ knowledge in the domain and
allow experts to categorize multiple elements of related information into a single, higher level
element. When confronted with a specific configuration of elements, experts are able to
recognize the pattern as a familiar schema and treat (and act on) the whole configuration as a
single unit. When brought into working memory, a single, high-level element requires
considerably less working memory capacity for processing then the many low-level elements it
incorporates, thus reducing the burden on working memory. As a consequence, acquired
schemas, held in long-term memory, allow experts to avoid processing overwhelming amounts
of information and effectively reduce the burden on limited capacity meoory. In addition, as
already mentioned, experts are able to bypass working memory capacity limits by having many
of their schemas highly automated due to extensive practice (Kalyuga, Ayers, Chandler, &
Sweller, 2003).
The level of learner experience in a domain primarily influences the extent to which
schemas can be brought into working memory to organize current information. Novices lack
sophisticated schemas associated with a task or situation at hand. For these inexperienced
learners, no guidance for handling a given situation or task is provided by relevant schemas in
long-term memory. Instructional guidance can act as a substitute for missing schemas and, if
effective, acts as a means of constructing schemas (Kalyuga, Ayers, Chandler, & Sweller, 2003).
If the instructional presentation fails to provide necessary guidance, learners will have
to resort to problem-solving search strategies that are cognitively inefficient, because they
impose a heavy working memory load (Kalyuga, Ayers, Chandler, & Sweller, 2003).
Expertise reversal effect: In contrast, experts bring their activate schemas to the
process of constructing mental representations of a situation or task. They may not need any
additional instructional guiadance because their schemas provide full guidance. If, nevertheless,
instruction provides information designed to assist learners in constructing appropriate mental
representations, and experts are unable to avoid attending to this information, there will be an
overlap between the schema-based and the redundant instruction-based components of guidance
(Kalyuga, Ayers, Chandler, & Sweller, 2003). Cross-referencing and integration of redundant
components will require additional working memory resources and might cause a cognitive
overload. This additional cognitive load may be imposed even if a learner recognizes the
instructional materials to be redundant and so decides to ignore that information as best has he or
she can (Kalyuga, Ayers, Chandler, & Sweller, 2003). For more experienced learners, rather than
risking conflict between schemas and instruction-based guidance, it may be preferable to
eliminate the instruction-based guidance (Kalyuga, Ayers, Chandler, & Sweller, 2003).
Split attention effect: When dealing with two or more related sources of information
(e.g., text and diagrams), it is often necessary to integrate mentally corresponding representations
(verbal and pictorial) to construct a relevant schema and achieve understanding. When different
sources of information are separated in space or time, this process of information integration may
WAINESS PHD QUALIFYING EXAM
94
place an unnecessary strain on limited working memory resources. Intensive search-and-match
processes may be involved in cross-referencing the representations. These search-and-match
processes may severely interfere with constructing integrated schemas, thus increasing the
burden on working memory and hindering learning (Kalyuga, Ayers, Chandler, & Sweller,
2003). Superiority of physically integrated materials that do not require split attention over
unintegrated materials that do require split attention and mental integration before they can be
understood provides and example of the split-attention effect (Kalyuga, Ayers, Chandler, &
Sweller, 2003).
Redundancy Effect: Physical integration of two or more sources of information to
reduce split attention and cognitive load is important if they sources of information are essential
in the sense that they are not intelligible in isolation for a particular learner. Alternatively, if they
sources are intelligible in isolation with one source unnecessary, elimination rather than physical
integration of the redundant source is preferable (Kalyuga, Ayers, Chandler, & Sweller, 2003).
Whether two sources of information are unintelligible in isolation and so require
integration or whether one source is redundant and so should be eliminated does not depend just
on the nature of the information, it also depends on the expertise of the learner. A source of
information that is essential for a novice may be redundant for an expert (Kalyuga, Ayers,
Chandler, & Sweller, 2003).
Text coherence depends on the learner’s expertise. Text that is minimally coherent for
novices may well be fully coherent for experts. Providing additional text is redundant for experts
and will have negative rather than positive effects, thus demonstrating the expertise reversal
effect (Kalyuga, Ayers, Chandler, & Sweller, 2003).
Modality effect: Using a combination of both auditory and visual sources of
information is an alternative way of dealing with split attention. According to dual-processing
models of memory and information processing, the capacity to process information is distributed
over several partly independent subsystems. As a consequence, effective working memory
capacity can be increased by presenting some information in an auditory and some in an visual
modality (Kalyuga, Ayers, Chandler, & Sweller, 2003).
Many studies (Mayer, 1997; Mayer & Moreno, 1998; Mousavi, Low, & Sweller, 1995)
have demonstrated that learners can integrate words and diagrams more easily when the worls
are presented in auditory form rather than visually, providing an example of the modality effect
(Kalyuga, Ayers, Chandler, & Sweller, 2003).
However, auditory explanations may also become redundant when presented to more
experienced learners. Kalyuga, Chandler, and Sweller (2000) demonstrated that if experienced
learners attend to auditory explanations, learning might be inhibited (Kalyuga, Ayers, Chandler,
& Sweller, 2003).
Worked example effect: Worked examples consisting of a problem statement followed
by explanations of all solution details represent a case of fully guided instruction. Exploratory
learning environments, discovery learning, or problem solving, however, represent a form of less
or even relatively unguided instruction. A considerable number of studies, such as Quilici and
Mayer (1996) demonstrated that when adults learned laws of mechanics from unstructured
simulations (designed as free exploration), the results were significantly worse than those for an
example-based, tutorial condition (Kalyuga, Ayers, Chandler, & Sweller, 2003).
When solving unfamiliar problems, learners normally use a means-end search strategy
directed toward reducing differences between current and goal problems states by using suitable
operators. These activities are unrelated to schema construction and automation and are
WAINESS PHD QUALIFYING EXAM
95
cognitively costly because they impose heaving working memory load (Sweller, 1988, as cited in
Kalyuga, Ayers, Chandler, & Sweller, 2003).
Providing worked examples instead of problems eliminates the means-ends search and
directs a learner’s attention toward a problem state and its associated moves (Kalyuga, Ayers,
Chandler, & Sweller, 2003). Of course, worked examples should be appropriately structured to
eliminate an unnecessary cognitive load do to, for example, split-attention effects (Kalyuga,
Ayers, Chandler, & Sweller, 2003).
As learners experience in a domain increased, solving a problem may not require a
means-end search and its associated working memory load due to a partially, or even fully,
constructed schemas. When a problem can be solved relatively effortlessly, analyzing a
redundant worked example and integrating it with previously acquired schemas in working
memory may impose a greater cognitive load than problem solving. Under these circumstances,
practice in problem solving may result in more effective learning than studying worked
examples, because solving problems may adequately facilitate further schema construction and
automation (Kalyuga, Ayers, Chandler, & Sweller, 2003).
Worked examples are most appropriate when presented to novices, but they should be
gradually faded out with increased levels of learner knowledge and replace by problems
(Kalyuga, Ayers, Chandler, & Sweller, 2003; Renkl & Atkinson, 2003).
Some material imposes an intrinsically high cognitive load because the elements that
must be learned interact and so cannot be processed in isolation without compromising
understanding. Learners must process many interacting elements of information simultaneously
in working memory where understanding is defined as the ability to process all necessary
interacting elements in working memory simultaneously. However, the assessment of element
interactivity is always relative to the level of expertise of an intended learner. If the learner holds
an appropriate set of previously acquired domain-specific schemas, the whole set of interacting
elements may be incorporated into a schema and regarded as a single element. Conversely, a
novice learner may need to attend to each of the elements and learn all interaction between
elements individually. If element interactivity is sufficiently high for the learner, this mental
activity will overload the limited capacity of working memory and cause a learning failure
(Kalyuga, Ayers, Chandler, & Sweller, 2003).
How can novices acquire the schemas necessary to allow the processing of very highelement interactivity material if they cannot process all of the element in working memory
simultaneously and if those interacting elements cannot be processed in isolation because they
interact? See the Space Fortress dyadic protocol studies (Kalyuga, Ayers, Chandler, & Sweller,
2003). Another method is to initially present the information as isolated elements of information
(Kalyuga, Ayers, Chandler, & Sweller, 2003). However, this method may not be beneficial to
expert learners (Kalyuga, Ayers, Chandler, & Sweller, 2003).
The Imagination Effect: The imagination effect occurs when learners are asked to
imagine the content of instruction learn more than learners simply asked to study the material.
More knowledgeable students who held appropriate prerequisite schemas found imagining
procedures and relations more beneficial for learning compared with studying working examples,
where less knowledgeable students found imagining procedures and relations had a negative
effect compared with studying worked examples (Kalyuga, Ayers, Chandler, & Sweller, 2003).
The process of mental imagining is closely associated with constructing and running
mental representations in working memory. Because inexperienced learners have no appropriate
schemas to support this process, attempts to engage in imagining are likely to fail (Kalyuga,
WAINESS PHD QUALIFYING EXAM
96
Ayers, Chandler, & Sweller, 2003). When asked to study worked examples rather than imagine
procedures, novices can construct schemas of interacting elements, an essential first step to
learning (Kalyuga, Ayers, Chandler, & Sweller, 2003).
Schema is defined as a cognitive construct that permits people to treat multiple
subelements of information as a single element, categorized according to the manner in which it
will be used (Kalyuga, Chandler, & Sweller, 1998).
Schema have a dual function: storing learned information in long-term memory and
reducing the burden on working memory by allowing multiple elements of information to be
treated as a single element (Kalyuga, Chandler, & Sweller, 1998).
Automation allows information to be processed with less working memory resources
than if not automated. Schemas are stored in long-term memory with varying degrees of
automaticity. A schema can be stored and retrieved from long-term memory either in fully
automated form or in a form that requires conscious consideration of each of the elements and
their relations (Kalyuga, Chandler, & Sweller, 1998).
If a schema can be brought into working memory in automated form, it will make
limited demands on working memory resources, leaving more resources available to search for a
possible solution problem (Kalyuga, Chandler, & Sweller, 1998). Cognitive load theory, which
incorporates this architecture, has been used to design a variety of instructional procedures,
based on the assumption that working memory is limited and that skilled performance is driven
by automated schemas held in long-term memory (Kalyuga, Chandler, & Sweller, 1998).
Split attention effect (Kalyuga, Chandler, & Sweller, 1998).
Redundancy effect (Kalyuga, Chandler, & Sweller, 1998).
Limits of working memory (Kalyuga, Chandler, & Sweller, 2000).
Split attention effect (Kalyuga, Chandler, & Sweller, 2000).
Many ways to deal with split attention (Kalyuga, Chandler, & Sweller, 2000).
Modality effect (Kalyuga, Chandler, & Sweller, 2000).
Redundancy effect (Kalyuga, Chandler, & Sweller, 2000).
Problem solving may inhibit schema construction and automation because the strategy
normally used to solve problems, means-end analysis, imposes a heavy working memory laod
that interferes with learning. A means-end strategy is directed towards reducing difference
between current and goal problem states. To use the strategy, the solver must simultaneously
consider the current problem state, the goal state, the difference between the current and goals
states, the relevant operators and their relations to the differences between the current and goal
states, and lastly, any subgoals that have been established (Kalyuga, Chandler, Tuovinen, &
Sweller, 2001).
Providing solution examples instead of problems should reduce cognitive load because
it obviates the need for means-end search and instead requires learners to study each example
state and its associated move or moves (Kalyuga, Chandler, Tuovinen, & Sweller, 2001).
There are established conditions under which the worked example effect does not
occur. If a worked example is structured in a manner that imposes a heavy cognitive load, there
is no reason to predict that worked examples will be superior to solving the equivalent problems
and the effect should disappear (Kalyuga, Chandler, Tuovinen, & Sweller, 2001).
The redundancy effect can also affect the value of worked examples. For more
experienced learners, some of the worked example information may be unnecessary, because the
information is already know to the learner and, therefore, redundant. Trying to incorporate that
redundant information with the schema already in working memory can create more cognitive
WAINESS PHD QUALIFYING EXAM
97
load than necessary and even overload working memory (Kalyuga, Chandler, Tuovinen, &
Sweller, 2001). For these learners, problem solving would be superior to worked examples,
because problem solving allows them to using existing schema to solve a goal condition, and
doesn’t require the inclusion of redundant schema information (Kalyuga, Chandler, Tuovinen, &
Sweller, 2001).
In the adaptive training (AT) method, the training system monitors and evaluates the
performance of a student. Based on this evaluation, a new level of task difficulty is set for
practice in an effort to maintain optimal learning conditions for the individual trainee (Mane,
Adams, & Donchin, 1989).
Two hypotheses underlie the concept of AT: (1) the learning of a complex perceptualmotor skill is better accomplished if the learner starts with a less difficult version of the task and
then makes the transition to an increasingly more difficult version of the task; (2) learning of a
task is better when the transition from one level of difficulty to another is based on the
individuals’ level of proficiency rather than a fixed order transition (Mane, Adams, & Donchin,
1989).
Part training (PT) is a method in which parts of the task are presented in isolation. A
number of part training methods such as pure part, progressive part, repetitive part, retrogressive,
and isolated parts have been developed (Mane, Adams, & Donchin, 1989). PT can be an
effective method of training because it provides the student with an opportunity to study in
isolation the relationship among a subset of the elements in the task. When training a task as a
whole, it is often difficult to determine which of several factors in a situation determines the
outcome of any given action, or to isolate the relationship of two variables from the influence of
other variables. Breaking the task into parts is a good way to overcome that difficulty (Mane,
Adams, & Donchin, 1989).
In an experiment using a flight simulator computer game, when comparing PT to the
whole task control group, there is a clear advantage which persists throughout the entire period
of training. Subjects in the PT group performed better in game aspects which were directly
related to the PT manipulation, with shorter performance times. PT also resulted in higher
transfer performance (Mane, Adams, & Donchin, 1989).
Successful problem solving depends on three components—skill, metaskill, and will—
and each of these components can be influenced by instruction. When the goal of instruction is
the promotion of nonroutine problem solving, students need to possess the relevant skill,
metaskill, and well. Metacognitiion—in the form of metaskill—is central in problem solving
because it manages and coordinated the other components (Mayer, 1998).
Perhaps the most obvious way to improve problem solving performance is to teach the
basic skills. The general procedure is to analyze each problem into the cognitive skills needed for
solution and then systematically teach each skill to mastery (Mayer, 1998). One approach is to
break apart a task into its component skills and then systematically teach each skill to mastery
(part task component training). In this approach, any large task can be broken down into a
collection of “instructional objectives” (Mayer, 1998). Another method is to break the task into a
hierarchy of components (Mayer, 1998).
Metaskills (or metacognitive knowledge) involves knowledge of when to use, how to
coordinate, and how to monitor various skills in problem solving (Mayer, 1998).
An important instructional implication of the focus on metacognition is the problem
solving skills should be learned in the context of realistic problem-solving situations. Instead of
using drill and practice on component skills in isolation—as suggested by a skill-based
WAINESS PHD QUALIFYING EXAM
98
approach—a metaskill-based approach suggest modeling of how and when to use strategies in
realistic academic tasks (Mayer, 1998).
Rather than practicing of basic component skills in isolation, successful
comprehension strategy instruction requires learning within the context of real tasks. By
embedding strategy instruction in academic tasks, students also acquire the metacognitive skills
of when and how to use the new strategies (Mayer, 1998).
Effort theory and interest theory yield strikingly different educational implications.
The effort theory is most consistent with the practice of teaching skills in isolation, and with
using instructional methods such as drill-and-practice. The interest theory is most consistent with
the practice of teaching skills in context, and with using instructional methods such as cognitive
apprenticeship (Mayer, 1998).
Individual interest refers to a person’s dispositions or preferred activities, and therefore
is a characteristic of the person. Situational interest refers to a task’s interestingness, and
therefore is a characteristic of the environment. Interest theory predicts that students think harder
and process the material more deeply when they are interested, rather than uninterested (Mayer,
1998). Interest theory also predicts that an otherwise boring task cannot be made interesting by
adding a few interesting details, such as seductive details (Mayer, 1998).
Self-efficacy theory predicts that students work harder on a learning task when they
judge themselves as capable than when they lack confidence in their ability to learn. Selfefficacy theory also predicts that students understand the material better when they have high
self-efficacy than when they have low self-efficacy (Mayer, 1998).
Simple user interaction in a multimedia explanation refers to user control over the
words and pictures that are presented in the multimedia explanation—namely, the pace of the
presentation. Simple user interaction may affect both cognitive processing during learning and
the cognitive outcome of learning (Mayer & Chandler, 2001).
The parts-first hypothesis assets that learners are more likely to experience cognitive
overload when the whole presentation is given first (whole-part) and it therefore cannot serve as
an effective context for organizing the subsequent parts presentation. Instead, when the parts
presentation comes first (part-whole), learners can build separate component models for each of
the key parts of the system. These component models will serve as chunks that can be more
easily organized into a mental model when the whole presentation is given (Mayer & Chandler,
2001).
In a whole-whole presentation, learners receive the entire multimedia explanation and
then receive it again. In a part-part presentation, learners receive the parts presentation and then
receive it again.
In their experiment, the PW group performed better on the transfer test than the WP
did, and the PP group performed better on the transfer text than the WW group did. For measures
of deep understanding, there was a clear advantage of PW presentation over WW presentation
and PP presentation over WW presentation. PW and PP seem the best methods for deep learning
as measured by transfer (Mayer & Chandler, 2001).
The locus of the redundancy effect seems to be at the point of visual attentional
scanninc, as posited by the split-attention hypothesis. The onscreen text competes with the
animation for visual attention, thus reducing the chances that the learner will be able to attend to
relevant aspects of the animation and text (Mayer, Heiser, & Lonn, 2001).
WAINESS PHD QUALIFYING EXAM
99
We interpret the redundancy as a new piece of support for the cognitive theory of
multimedia learning and, in particular, the idea that humans possess separate visual and auditory
processing channels that are each limited in capacity (Mayer, Heiser, & Lonn, 2001).
According to a cognitive theory of multimedia learning, not all techniques for
removing redundancy are equally effective. For example, in the case of multimedia explanations
consisting of animation, narration, and on-screen text, one effective solution is to remove the onscreen text, but it does not follow that the same benefits would occur by instead removing the
narration (Mayer, Heiser, & Lonn, 2001).
Coherence effect refers to situations in which adding words or pictures to a multimedia
presentation results in poorer performance on tests of retention or transfer (Mayer, Heiser, &
Lonn, 2001).
Seductive details (Mayer, Heiser, & Lonn, 2001).
Mayer defines as the presentation of information in two or more formats, such as in
words and pictures (Mayer, 1997; Mayer & Moreno, 1998).
According to the dual-processing theory, visually presented information is processed—
at least partially—in visual working memory whereas auditorily presented information is
processed—at least partially—in auditory working memory (Mayer & Moreno, 1998).
The major result of these studies is a split-attention effect in which students learned
better when pictorial information was accompanied by verbal information presented in an
auditory rather than a visual modality (Mayer & Moreno, 1998). The results also extend previous
research on contiguity effects in which students learned better when an animation depicting the
workings of a scientific system and the corresponding narration were presented concurrently
rather than successively (Mayer & Moreno, 1998).
According to the dual-processing theory of working memory, students learn better in
multimedia environments when words and pictures are presented in separate modalities than
when they are presented in the same modality (Mayer & Moreno, 1998).
In split attention situations, the learner’s attentional resources (or central executive
resources) are used to hold worlds and pictures in visual working memory sort her is not enough
left over to build connections between words and pictures. In contrast, when learners can
concurrently hold worlds in auditory working memory and picture in visual woking memory,
they are better able to devote attentional resources to building connection between them (Mayer
& Moreno, 1998).
In split-attention situations, an overload in visual working memory reduces the
learner’s ability to build coherent mental models that can be used to answer transfer questions. In
contrast, when words are presented in an auditory working memory and pictures are presented in
visual working memory, the learner is better able to organize representations in each store and
integrate across stores (Mayer & Moreno, 1998).
We define multimedia learning as learning from words and pictures, and we define
multimedia instruction as presenting words and pictures that are intended to foster learning. The
words can be printed or spoken. The pictures can be static or dynamic (Mayer & Moreno, 2003).
We define meaningful learning as deep understanding of the material, which includes
attending to important aspects of the presented material, mentally organizing it into a coherent
cognitive structure, and integrating it with relevant existing knowledge. Meaningful learning is
reflected in the ability to apply what was taught to new situations; problem solving transfer. In
our research, meaningful learning involves the construction of a mental model of how a causal
system works (Mayer & Moreno, 2003).
WAINESS PHD QUALIFYING EXAM
100
A central challenge facing designers of multimedia instruction is the potential for
cognitive overload—in which the learner’s intended cognitive processing exceeds the learner’s
available cognitive capacity (Mayer & Moreno, 2003).
Dual channel assumption (auditory and visual channels): (Mayer & Moreno, 2003).
Limited processing capacity (working memory) (Mayer & Moreno, 2003). This is the
central assumption of Chandler and Sweller’s (1991; Sweller 1999) cognitive load theory.
Cognitive overload defined (Mayer & Moreno, 2003).
Three kinds of load: essentially processing, incidental processing, and representational
holding. Essential processing refers to cognitive processes that are required for making sense of
the presented material, such as the five core processes in the cognitive theory of multimedia
learning—selecting words, selecting images, organizing words, organizing images, and
integrating. Incidental processing refers to cognitive processes that are not required for making
sense of the presented material but are primed by the design of the learning task.
Representational holding refers to cognitive processes aimed at holding a mental representation
in working memory over a period of time (Mayer & Moreno, 2003).
Reducing cognitive load can involve redistributing essential processing, reducing
incidental processing, or reducing representational holding (Mayer & Moreno, 2003).
Split attention effect: (Sweller, 1999) (Mayer & Moreno, 2003).
Modality effect (Mayer & Moreno, 2003).
Segmenting (part task): (Mayer & Moreno, 2003).
Pretraining (part task): (Mayer & Moreno, 2003).
Weeding involves making the narrated animation as concise and coherent as possible,
so the learner will not be primed to engage in incidental processing (Mayer & Moreno, 2003).
Signaling provides cues to the learner about how to select and organize the material
(Mayer & Moreno, 2003).
Aligning words and pictures (spatial contiguity) (Mayer & Moreno, 2003).
Eliminating redundancy (Mayer & Moreno, 2003).
Temporal contiguity effect (Mayer & Moreno, 2003).
Constructivist learning occurs when learners construct meaningful mental
representations from presented information (Mayer, Moreno, Boire, & Vagge, 1999).
A design principle is a technique for constructing multimedia environments that foster
constructivist learning. Although learners are not physically active in the multimedia
environment, it may possible to promote some degree of cognitive activity that results in
constructivist learning (Mayer, Moreno, Boire, & Vagge, 1999).
Internal connections = selecting relevant information of the modal and organizing them
into causal chains (Mayer, Moreno, Boire, & Vagge, 1999).
External connections (aka referential connections). = integrating the internal
connections to one another and with relevant prior knowledge (Mayer, Moreno, Boire, & Vagge,
1999).
Constructivist learning occurs when learners are able to build referential connections
between corresponding aspects of the visual and verbal representations of a multimedia
presentation (Mayer, Moreno, Boire, & Vagge, 1999). Constructivist learning is fostered when
the learner is able to hold a visual representation in visual working memory and a corresponding
verbal representation in verbal working memory at the same time. The model implicates working
memory (or cognitive load) as a major impediment to constructivist learning (Mayer, Moreno,
Boire, & Vagge, 1999).
WAINESS PHD QUALIFYING EXAM
101
The contiguity effect: learners perform better on retention and transfer when they view
animated materials concurrently with corresponding narration than when the animation is viewed
either before or after the narration (Mayer, 1997; Mayer, Moreno, Boire, & Vagge, 1999).
If modalities must be presented successively, rather than concurrently, reducing the
material to smaller bites reduces the detrimental learning effects of the contiguity effect (Mayer,
Moreno, Boire, & Vagge, 1999).
An explanation is a description of a causal system containing parts that interact in a
coherent way. A change in one part causes a change in another part (Mayer & Sims, 1994).
Multimedia learning occurs when students use information presented in two or more
formats to construct knowledge. This definition also applies to the term multimodal, since
learners are exposed to more than one sense modality, rather that multimedia, which refers to the
idea that the instructor uses more than one presentation medium (Mayer & Sims, 1994).
The dual coding theory involves three processes: A verbal explanation is presented
along with a visual explanation, then in working memory the learner constructs mental
representations of the two explanations and accesses relevant prior knowledge from long term
memory, and lastly the two representations are combined or linked with referential connections
(Mayer & Sims, 1994).
Contiguity effect: From the dual coding theory, it is expected that meaningful learning
occurs in working memory when multiple modes of information are process and linked with
referential connection. This in turn leads to better transfer effects. Therefore, if the material is not
presented concurrently, this process is ill-supported. (Mayer & Sims, 1994).
Students performed better on problem-solving transfer when the voice in the
multimedia message was from a human speaking with a standard accent rather than a human
speaking with a foreign accent or a machine voice. These results are consistent with the social
agency theory and cognitive load theories. The social agency theory suggests that social cues in a
multimedia message can prime the social conversation schema in learners. Once the social
conversation schema is activated, learners are more likely to acts as if they are in a conversation
with another person. Thus, at least to some extent, the social rules of human-to-human
communication come into play. Therefore, the learner tries harder to make sense of what the
computer is asying by engaging in deep cognitive processing. The deep processes include
selecting relevant information for further processing, organizing the pieces of information into
coherent representations, and integrating verbal and visual representations with each other and
with prior knowledge. Deep cognitive processing results in meaningful learning outcomes, which
enable learners to apply (or transfer) what they have learned to new situations (Mayer, Sobko, &
Mautone, 2003).
Concrete icons enable users to use their everyday knowledge about the objects they
depict to understand the likely function of the icon. The effects of icon concreteness are short
lived and limited to users’ early experience of an icon set when users are unsure of the meaning
of icons. The effect of icon complexity, in contrast, are more most apparent in tasks involving a
search component and do not diminish as a result of experience (McDougall, de Bruijn, & Curry,
2000).
Distinctiveness of an icon, and the features underpinning distinctiveness, vary
depending on the nature of the array in which an icon is presented. Concrete icons in an array of
abstract icons become distinct. Abstract icons in an array of concrete icons become distinct.
Contrast is the distinguishing factor (McDougall, de Bruijn, & Curry, 2000).
WAINESS PHD QUALIFYING EXAM
102
Virtual Reality (VR) is a multi-sensory highly interactive computer based environment
where the user becomes an active participant in a virtually real world. First person’s point of
view, freedome in navigation, and interaction are esessential for a computer environment to be
characterized as a VR environment, or VE (virtual environment; Mikropoulos, 2001).
A virtual environment designed to educate the ser is called a virtual learning
environment. It should have and educational objective and provide users with experiences they
would otherwise not be able to experience in the physical world (Mikropoulos, 2001).
VR proposes the adaptation of technology to people and not the opposite
(Mikropoulos, 2001).
The physical structure of the human brain is affected by the way it is used. Different
kinds of experiences configure the brain, especially children’s brains. The reorganization of
children’s brains is an important factor in the educational process, specifically in the case of the
involvement of mediand and educational technology (Mikropoulos, 2001).
The goals of this article is to compare the electrical brain activity taking place in
virtual versus real environments. A further goals is to measure and analyze the cognitive changes
that users of educational VR systems experience and to evaluate the consequences of such a kind
of educational software (Mikropoulos, 2001).
Electroencephalography (EEG) showes the electrical activity of a number of neurons
that can be recorded from the scalp. Techniques have been developed to extract information from
the signals recorded in order to obtain an understanding of the brain processes underlying
psychophysical and cognitive functions (Mikropoulos, 2001).
College students were exposed to an educational VE with landscapes for geography
and astronomy teaching, buildings and rooms for environmental and physics education, and the
incised of cells for biology teaching. Movements in these environments were compared to
movements in real world counterparts, with the real world versions occurring first (Mikropoulos,
2001).
Subjects were more attentive when navigating in the virtual world. Less mental effort
was used in the real world version of tasks than in the virtual version. All findings can be
attributed to experience in the real world versus inexperience in the virtual world. Overall,
though, the findings reported similar brain activity for the same task in both the real and virtual
environment. This activity is connected with visual perception, attentional demands, and mental
effort. The results thus indicate that users behave similarly in virtual and real environments. They
also indicate that virtual reality provides educational environments for students to concentrate,
perceive, and judge as a result of less eye-movement and alpha signal dimunition. Additionally,
there is need for users to be trained in and comfortable with VR (Mikropoulos, 2001).
Contiguity effect: Temporal-contiguity effect and spatial-contiguity effect. Spatial =
modalities integrated or physically separated. Temporal = order of presentation. contiguity =
split-attention effect (Moreno & Mayer, 1999).
Modality principle = dual-channel effects (Moreno & Mayer, 1999).
Mixed modalities are better (Moreno & Mayer, 1999).
Suggests replacing split-attention effect with multiple terms: spatial-contiguity effect,
temporal-contiguity effect, and modality effect, according to the situation because they results in
different effects on working memory (Moreno & Mayer, 1999).
Arousal theory suggests that adding entertaining auditory adjuncts will make a learning
task more interesting, because it creates a greater level of attention so that more material is
processed by the learner (Moreno & Mayer, 2000a).
WAINESS PHD QUALIFYING EXAM
103
Adding extraneous sentences or illustrations, referred to as seductive details, results in
poorer retention and transfer performance, even when the material was meant to entertain
(Moreno & Mayer, 2000a).
The coherence principle or theory holds that auditory adjuncts can overload the
auditory channel (or auditory working memory). Any additional material (including sound
effects and music) that is not necessary to make the lesson intelligible or that is not integrated
with the rest of the materials will reduce effective working memory capacity and thereby
interfere with the learning of the core material, and therefore, resulting in poorer performance on
transfer tests (Moreno & Mayer, 2000a).
The more relevant and integrated sounds are, the more they will help students’
understanding of the materials (Moreno & Mayer, 2000a).
On the surface, seductive details and auditory adjuncts seem similar. However, the
underlying cognitive mechanisms are quire different. Whereas seductive details seem to prime
inappropriate schemas into which incoming information is assimilated, auditory adjuncts seem to
overload auditory working memory (Moreno & Mayer, 2000a).
The self-reference effect on memory is based on a very efficient mechanism to process
material that is very familiar to oneself. Personalizing the context improves learning by helping
learners interpret and interrelate important information in the familiar versus abstract problem
statement (Moreno & Mayer, 2000b).
Providing personalized message in media communication seems more likely to ease
the processing of the message by being more consistent with the social rules and schemas of
normal conversations (Moreno & Mayer, 2000b).
A self-reference effect for problems-solving transfer in multimedia messages was
observed across five experiments: Student who learned by means of a personalized explanation
(either as speech or as on-screen text) were better able to use what they learned to solve new
problems than students who received a neutral monologue (Moreno & Mayer, 2000b).
The beneficial effects of introducing self-referencing into a multimedia science lesson
occur independently of the behavioral interaction required during a computer lesson. When the
presentation is linear, so that students are required only to watch an animation while listening to
or reading an explanation, and when students are required to make choices by clicking on the
computer screen, self-referencing seems to promote the mental interactions needed to actively
involve the learner in the process of understanding (Moreno & Mayer, 2000b).
Learning environments can vary in immersion from no immersion (such as illustrated
text) to low immersion (such as an educational game presented using a computer display and
speakers) to high immersion (such as a computer display presented using a head-mounted
display [HMD] and earphones; Moreno & Mayer, 2002).
The Design-A-Plant game puts learners on an alien planet where they must make a
plant flourish. The games uses a static 3D environment with the plant centered horizontally on
the screen. It uses a pedagogic agent who offers individualized advice concerned the relation
between plant features and climate conditions (Moreno & Mayer, 2002).
The questions for the study included: do the same instructional design principles there
were discovered with a non-immersive medium also apply to low-immersion media (e.g.,
desktop games) and more immersive media (e.g., HMD games)? The researchers focused on
retention, transfer, and program ratings (Moreno & Mayer, 2002).
Modality effect (Moreno & Mayer, 2002).
WAINESS PHD QUALIFYING EXAM
104
The study provided evidence that students felt a stronger sense of presence in more
immersive VREs. Also, students who learn in a more immersive VRE do not necessarily learn a
computer-based lesson more deeply as compared with students in a less immersive VRE
(Moreno & Mayer, 2002).
The researchers argued the lack of media effects might have been do to the low quality
of the graphics and a less compelling environment (Moreno & Mayer, 2002).
Modality effects appear to be consistent across non-, low-, and high-immersive
environments (Moreno & Mayer, 2002).
Because some media may enable instructional methods that are not possible with other
media, it might be useful to explore instructional methods that are possible in immersive
environments but not in others (Moreno & Mayer, 2002).
People have a limited working memory that is able to hold and process only a few
items of information at a time (Mousavi, Low, & Sweller, 1995).
People have a huge long-term memory that is effectively unlimited in size (Mousavi,
Low, & Sweller, 1995).
Schema acquisition is a primary learning mechanism. Schemata have the functions of
storing information in long-term memory and of reducing working memory load by permitting
people to treat multiple elements of information as a single element (Mousavi, Low, & Sweller,
1995).
Automation of cognitive processes, including automatic use of schemata, is a learning
mechanism that also reduces working memory load by effectively bypassing working memory.
Automated information can be processed without conscious effort (Mousavi, Low, & Sweller,
1995).
A limited working memory is central to this architecture and central to cognitive load
theory (Mousavi, Low, & Sweller, 1995).
Split-attention effect (Mousavi, Low, & Sweller, 1995). Can be reduced through dualmodality presentations (Mousavi, Low, & Sweller, 1995).
In general, the data (using Space Fortress) show facilitation in skill acquisition through
the employment of various part-task procedures and specific instructional strategies over the
baseline control conditions. However, there are a number of caveats (Newell, Carlton, Fisher, &
Rutter, 1989).
The part-task effect is strongly influenced by the nature of the part task selected for
prior practice. It appears that only part tasks that reflect “natural” units of coordinated activity
facilitate skill acquisition (Newell, Carlton, Fisher, & Rutter, 1989).
Control of difficulty levels did not affect performance (Newell, Carlton, Fisher, &
Rutter, 1989).
Results showed that both the presence of icons (versus textual indicators) and the
spatial grouping of icons speeded the search for a target file (Niemela & Saarinen, 2000).
Our results support the notion that icons, by their pictorial nature, may have other
inherent properties that lead to improved user performance at the interface (Niemela & Saarinen,
2000).
The grouping of items reduces the number of items to be searched. Spatially close
items tend to be grouped, but in more stimulus condition, humans are able to attend selectively to
spatially scattered subsets of elements (Niemela & Saarinen, 2000).
WAINESS PHD QUALIFYING EXAM
105
In this study, the grouping based on both the spatial closeness and similar appearance
if icons seemed to enable more efficient search than did the visual grouping based on the
similarity of icons (Niemela & Saarinen, 2000).
Cognitive load theory (CLT) originated in the 1980s and underwent substantial
development and expansion in the 1990s (Paas, Renkl, & Sweller, 2003).
Low-element interactivity refers to environments where each element can be learned
independently of the other elements, and there is little direct interaction between the elements.
High-element interactivity refers to environments where there is so much interaction between
elements that they cannot be understood until all the elements and their interactions are
processed simultaneously. As a consequence, high-element interactivity material is difficult to
understand (Paas, Renkl, & Sweller, 2003).
Element interactivity is the driver of our first category of cognitive load: intrinsic
cognitive load, because the demands on working memory capacity imposed by element
interactivity are intrinsic to the material being learned. Different materials differ in their levels of
element interactivity and thus intrinsic cognitive load, and they cannot be altered by instructional
manipulations; only a simpler learning task that omits some interacting elements can be chosen
to reduce this type of load (Paas, Renkl, & Sweller, 2003).
Working memory, in which all conscious cognitive processing occurs, can handle only
a very limited number of novel interacting elements; possibly no more than two or three. Longterm memory can contain vast numbers of schemas—cognitive constructs that incorporate
multiple elements of information into a single element with a specific function (Paas, Renkl, &
Sweller, 2003).
With schema use, a single element in working memory might consist of a large number
of lower level, interacting elements, which if processed individually might have exceeded the
capacity of working memory (Paas, Renkl, & Sweller, 2003).
The automation of schema so that they can be processed unconsciously further reduces
the load on working memory (Paas, Renkl, & Sweller, 2003).
As well as element interactivity, the manner in which information is presented to
learners and the learning activities required of learners can also impose a cognitive load. When
that load is unnecessary and so interferes with schema acquisition and automation, it is referred
to as extraneous or ineffective cognitive load. Cognitive theorists spend much of their time
devising alternative instructional design and procedures that reduce extraneous cognitive load
compared to conventionally used procedures (Paas, Renkl, & Sweller, 2003).
Extraneous cognitive load is primarily important when intrinsic cognitive load is high
because the two forms of cognitive load are additive (Paas, Renkl, & Sweller, 2003).
The last form of cognitive load is germane or effective cognitive load. Germane
cognitive load is influenced by the instructional design. The manner in which information is
presented to learners and the learning activities required of learners are factors relevant to levels
of germane cognitive load. Whereas extraneous cognitive load interferes with learning, germane
cognitive load enhances learning.
Cognitive load theory (CLT) is concerned with the development of instructional
methods that efficiently use people’s limited cognitive processing capacity to stimulate their
ability to apply acquired knowledge and skills to new situations (i.e., transfer). CLT is based on a
cognitive architecture that consists of a limited working memory, with partly independent
processing units for visual/spatial and auditory/verbal information, which interactis with a
comparatively unlimited long-term memory. Central to CLT is the notion that working memory
WAINESS PHD QUALIFYING EXAM
106
architecture and its limitations should be a major consideration when designing instruction (Paas,
Tuovinen, Tabbers, & Van Gerven, 2003).
The most important learning processes for developing the ability to transfer acquired
knowledge and skills are schema construction and automation. According to CLT, multiple
elements of information can be chunked as single elements in cognitive schema, which can be
automated to a large extent. Then, they can bypass working memory during mental processing
thereby circumventing the limitations of working memory. Consequently, the prime goals of
instruction are the construction and automation of schemas (Paas, Tuovinen, Tabbers, & Van
Gerven, 2003).
Cognitive load is not simply considered as a by-product of the learning process but as a
major factor that determines the success of an instruction intervention (Paas, Tuovinen, Tabbers,
& Van Gerven, 2003).
Cognitive load can be defined as a multidimensional construct representing the load
that performing a particular task imposes on the learner’s cognitive system. The construct has a
causal dimension reflecting the interaction between task and learner characteristics and an
assessment dimension reflecting the measurable concepts of mental load, mental effort, and
performance (Paas, Tuovinen, Tabbers, & Van Gerven, 2003).
Task characteristics that have bee identified in CLT research are task format, task complexity,
use of multimedia, time pressure, and pacing of instruction. Relevant learner
characteristics comprise expertise level, age, and spatial ability (Paas, Tuovinen, Tabbers,
& Van Gerven, 2003).
Mental load is the aspect of cognitive load that originates from the interaction between
task and subject characteristics. Mental load provides an indication of the expected cognitive
capacity demands and can be consider an a priori estimate of the cognitive load (Paas, Tuovinen,
Tabbers, & Van Gerven, 2003).
Mental effort is the aspect of cognitive load that refers to the cognitive capacity that is
actually allocated to accommodate the demands imposed by the task; thus, it can be considered
to reflect the actual cognitive load. Mental effort is measured while participants are working on a
task (Paas, Tuovinen, Tabbers, & Van Gerven, 2003).
Performance, also an aspect of cognitive load, can be defined in terms of learner’s
achievements. It can be determined while people are working on a task or thereafter (Paas,
Tuovinen, Tabbers, & Van Gerven, 2003).
Mental models explain human cognitive processes of understanding external reality,
translating reality into internal representations and utilizing it in problem solving (Park &
Gittelman, 1995).
Mental model formation depends heavily on the conceptualizations that individuals
bring to a task. When interacting with the environment, with other, and with the artifacts of
technology, people form mental models of themselves and the things with which they interact
(Park & Gittelman, 1995).
People sometimes develop the dynamic characteristics of mental models showing
direction of processes, motion, and changes over time. These indicate that the dynamic
characteristics of mental models seem to be determined primarily by subjects’ understanding of
the system features and functions more than by the visual content of the externally presented
training contents or the system (Park & Gittelman, 1995).
WAINESS PHD QUALIFYING EXAM
107
Motional or motion cures simulating system functions (visible or invisible) in visual
displays seem to facilitate the formation process of dynamic characteristics of mental models
(Park & Gittelman, 1995).
Simulations that include the actual movements in a task are more directive and
effective for teaching the dynamic nature of a given task and aiding the formation of dynamic
characteristics of mental models for the task (Park & Gittelman, 1995).
Worked out examples usually consist of a problem formulation, solution steps, and the
final solution itself (Renkl, & Atkinson, 2003).
In later stages of skill acquisition, emphasis is on increasing speed and accuracy of
performance, and skills, or at least subcomponents of them should become automated. During
these stages, it is important that the learners actually solve problems as opposed to studying
examples (Renkl, & Atkinson, 2003).
Cognitive skills refer to the learners’ capabilities to solve problems from intellectual
domains such as mathematics, medical diagnosis, or electronic troubleshooting. Cognitive skill
acquisition is, therefore, a narrow term as compared to learning. For example, it does not include
acquisition of declarative knowledge for its own sake, general thinking or learning skills, general
metacognitive knowledge, and so on (Renkl, & Atkinson, 2003).
Intrinsic load (Renkl, & Atkinson, 2003).
Germane load (Renkl, & Atkinson, 2003).
Extraneous load (Renkl, & Atkinson, 2003).
Fading worked out solution steps (Renkl, & Atkinson, 2003).
Worked out examples defined (Renkl, Atkinson, Maier, & Staley, 2002).
Animation can avoid being distracters to learning if it is clear to the learner that the
animation (e.g. a moving spaceship) is not part of the to-be learned material (Rieber, 1996).
The term seductive details is used to describe highly interesting but unimportant test
segments. These segments usually contain information that is tangential to the main themes of a
story, but are memorable because they deal with controversial or sensational topics (Schraw,
1998).
Dyadic protocol (Shebilske, Wesley, Regian, Arthur, & Jordan, 1992).
Learner-controlled instruction was superior to the program controlled instruction with
regard to student performance in a novel procedural task (Shyu & Brown, 1995).
Prior knowledge did not significantly contribute to performance based on control type
(Shyu & Brown, 1995).
Direct manipulation with visual feedback was the most user-friendly interface in the
study (Svendsen, 1991).
Results indicate that direct manipulation may hinder effective problem solving,
because the interface is so supportive of thoughtless action that the user neglects to look for rules
where these are called for (Svendsen, 1991).
When a program is characterized as user friendly, it entails that users are able to learn
the functionality of a program fairly quickly, are able to use this functionality, and like using the
program (Svendsen, 1991).
The focus on authentic learning tasks—whole tasks that are based on real-life tasks—
can be found in practical educational approaches, such as subject-based education, the case
method, problem-based learning, and competency-base learning (van Merrienboer, Kirshner, &
Kester, 2003).
WAINESS PHD QUALIFYING EXAM
108
A severe problem with authentic whole tasks is that learners may have difficulty
learning because they are overwhelmed by the task complexity (van Merrienboer, Kirshner, &
Kester, 2003).
Scoffolds, according to their original meaning with educational psychology, include all
devices or strategies that support students’ learning (van Merrienboer, Kirshner, & Kester, 2003).
In both cognitive apprenticeship learning and our framework, scaffolding explicitly
pertains to a combination of performance support and fading. Initially, the support enables a
learner to achieve a goal or action not achievable without that support. When the learner achieves
the desired goal, support gradually diminishes until it is no longer needed (van Merrienboer,
Kirshner, & Kester, 2003).
Because excessive or insufficient support can hamper the learning process, it is critical
to determine the right type and amount of support and to fade at the appropriate time and rate
(van Merrienboer, Kirshner, & Kester, 2003).
CLT emphasizes the need to integrate support for novice learners with the task
environment fully; otherwise, split-attention effects increase extraneous cognitive load because
learners have to integrate information mentally from the task environment with the given support
(van Merrienboer, Kirshner, & Kester, 2003).
It is clearly impossible to use highly complex learning tasks from the start of a course
or graining because this would yield excessive cognitive load for the learners. The common
solution is to let learners start their work on relatively simple learning tasks and progress toward
more complex tasks (van Merrienboer, Kirshner, & Kester, 2003).
Complex performances are broken down into simpler parts that are trained separately
or, in a part-whole approach, are gradually combined into whole-task performance. It is not until
the end of the training program that learners have the opportunity to practice the whole task (van
Merrienboer, Kirshner, & Kester, 2003).
Part-task approaches to sequencing are highly effective to prevent cognitive overload
because the load associated with a part of the task is lower than the load associated with the
whole task (van Merrienboer, Kirshner, & Kester, 2003).
However, part-task approaches to sequencing and instructional models driven by
separate objectives do not work well for complex performances that require the integration of
skills, knowledge, and attitudes and extensive coordination of constituent skills in new problem
situations (van Merrienboer, Kirshner, & Kester, 2003).
Whole-task approaches attend to the coordination and integration of constituent skills
from the very beginning, and they stress that learners quickly develop a holistic vision of the
whole task that is gradually embellished during the training (van Merrienboer, Kirshner, &
Kester, 2003).
Learning tasks are often equated with conventional problems. Such tasks confront the
learner with a given state and a set of criteria for an acceptable goal state. There is overwhelming
evidence that such conventional task are exceptionally expensive in terms of working memory
capacity. E.g., means-end analysis (van Merrienboer, Kirshner, & Kester, 2003).
Learning tasks that take the form of worked examples confront learners not only with a
given state and a desired goal state but also with an example solution. Studying those examples
as a substitute for performing conventional problem solving tasks may be beneficial, because it
focuses attention on problem states and associated solution states and so enables learners to
induce generalized solutions or schemas (van Merrienboer, Kirshner, & Kester, 2003). A
WAINESS PHD QUALIFYING EXAM
109
disadvantage of worked-out examples is that they do not foce learners to study them carefully.
(van Merrienboer, Kirshner, & Kester, 2003).
An alternative to worked examples is completion tasks that present a given state, a goal
state, and a partial solution to the learners who must complete the solution. Completion tasks
combine the strong points of worked out examples and conventional learning tasks. Like
conventional learning tasks, they directly encourage learners to be active because learners have
to complete the solution, which is only possible by the careful study of the partial example
provided by the completion task (van Merrienboer, Kirshner, & Kester, 2003).
One way is to present necessary information before the learners start working on a
learning task or series of tasks. The other way is to present the necessary information precisely
when the learners need it during task performance (just-in-time information). CLT does not yield
an unequivocal answer to the question of which of the two ways is best (van Merrienboer,
Kirshner, & Kester, 2003).
In contrast to supportive information, procedural information pertains to consistent
task components or recurrent task aspects that are performed as routines by experts. These tasks
can become automated by experts. CLT not only indicates that procedural information is best
presented when learners need it, but it also raises two related design issues. First, presenting
procedural information precisely when it is needed to perform particular actions prevents
temporal split-attention effects. Second, presenting procedural information so that is it fully
integrated with the task environment prevents spatial split-attention effects (van Merrienboer,
Kirshner, & Kester, 2003).
Supportive information may be helpful in performing the nonrecurrent aspects of
learning. It is best presented before a class of equivalent learning tasks, and it is critical that the
learners elaborate on it so that it can be easily retrieved from long-term memory when necessary
for the learning task. Elaboration are used to develop schemas whereby nonarbitrary relations
are established between new information elements and the learner’s prior knowledge (van
Merrienboer, Kirshner, & Kester, 2003).
Split-attention effect: (Yeung, Jin, & Sweller, 1997).
Redundancy effect: (Yeung, Jin, & Sweller, 1997).
Trail and error in computer gaming is defined as the absence of a systematic strategy in
playing a game. This particular strategy involves actions and reactions to circumstances,
consequences, and feedback within the game framework. Knowledge of how to play the game is
accumulated through observation and active participation in the gaming process, not be reading
rules and instructions (Dempsey, Haynes, Lucassen, & Casey, 2002). In this study, strategies in
playing computer games included trial and error, reading instructions, relying on prior
knowledge or experiences, and developing a personal game-playing strategy. Trial and error was
by far the predominant strategy across all game types (126 of the 160 games played). It was often
employed even in cases where participants reported that they know a more efficient strategy
(Dempsey, Haynes, Lucassen, & Casey, 2002).
Self-efficacy is defined as one’s belief about one’s ability to successfully carry out
particular behaviors (Davis, & Wiedenbeck, 2001).
WAINESS PHD QUALIFYING EXAM
110
We define problem solving as the intellectual skill to propose solutions to previously
unencountered problem situations (Tennyson & Breuer, 2002).
Creativity is defined as the cognitive skill of creating a problem situation and proposing a
solution(s) (Tennyson & Breuer, 2002).
It goes without saying that the most efficient medium would not necessarily be ideal for
every stage of learning. The goal is to have a principled and empirical way to calculate optimal
information distributions at various points in different types of learning processes, including of
course terminal distributions (Cobb, 1997). Airline pilots are destined always to share major
parts of their cognitive work with their instruments, trapeze artists to get most of the work
packed into their heads. The way forward in media design is to model learner and medium as
distributed information systems, with principled, empirically determined distributions of
information storage and processing over the course of learning (Cobb, 1997).
A distribution-of-information analysis suggests that schematized information is to a large
extent preprocessed in a consumer culture, and so imposes a low memory demand when called
up for problem solving. But unfamiliar relations between decontextualized letters and numbers
are fully processed in working memory with predictably poor results (Cobb, 1997).
WAINESS PHD QUALIFYING EXAM
111
References for Question 2
Allen, R. B. (1997). Mental models and user models. In M. Helander, T. K. Landauer & P.
Prabhu (eds.), Handbook of Human Computer Interaction: Second, Completely Revised
Edition (pp. 49-63). Amsterdam: Elsevier
Atkinson, R. K., Derry, S. J., Renkl, A., & Wortham, D. (2000). Learning from examples:
Instructional principles from the worked examples research. Review of Educational
Research, 70(2), 181-214.
Atkinson, R. K., Renkl, A., Merrill, M. M. (2003). Transitioning from studying examples to
solving problems: Effects of self-explanation prompts and fading worked-out steps.
Journal of Educational Psychology, 95(4), 774-783.
Avouris, N., Dimitracopoulou, A., & Komis, V. (2003). On analysis of collaborative problem
solving: An object-oriented approach. Computers in Human Behavior, 19, 147-167.
Bailey, D. H. (1996). Constructivism and multimedia: Theory and application; innovation and
transformation. International Journal of Instructional Media, 23(2), 161-165.
Baker, E. L., & Mayer, R. E. (1999). Computer-based assessment of problem solving. Computers
in Human Behavior, 15, 269-282.
Baker, R., & Dwyer, F. (2000). A meta-analytic assessment of the effect of visualized
instruction. International Journal of Instructional Media, 27(4), 417-426.
Banbury, S. P., Macken, W. J., Tremblay, S., & Jones, D. M. (2001, Spring). Auditory
distraction and short-term memory: Phenomena and practical implications. Human
Factors, 43(1), 12-29.
Bangert-Drowns, R. L., & Pyke, C. (2001). A taxonomy of student engagement with educational
software: An exploration of literate thinking with electronic text. Journal of Educational
Computing Research, 24(3), 213-234.
Barab, S. A., Young, M. F., & Wang, J. (1999). The effects of navigational and generative
activities in hypertext learning on problem solving and comprehension. International
Journal of Instructional Media, 26(3), 283-309.
Bargh, J. A. (2002). Beyond simple truths: The human-Internet interaction. Journal of Social
Issues, 58(1), 1-8.
Baylor, A. L. (2001). Perceived disorientation and incidental learning in a web-based
environment: Internal and external factors. Journal of Educational Multimedia and
Hypermedia, 10(3), 227-251.
Beatty, J. (1982). Task-evoked papillary responses, processing load, and the structure of
processing resources. Psychological Bulletin, 91(2), 276-292.
Benbasat, I., & Todd, P. (1993). An experimental investigation of interface design alternatives:
Icon vs. text and direct manipulation vs. menus. International Journal of Man-Machine
Studies, 38, 369-402.
Berg, G. A. (2000). Human-computer interaction (HCI) in educational environments:
Implications of understanding computers as media [Electronic Version]. Journal of
Educational Multimedia and Hypermedia, 9(4), 349-370.
Blanton, B. B. (1998). The application of the cognitive learning theory to instructional design.
International Journal of Instructional Media, 25(2), 171-177.
Brown, D. W., & Schneider, S. D. (1992), Young learners’ reactions to problem solving
contrasted by distinctly divergent computer interfaces. Journal of Computing in
Childhood Education, 3(3/4), 335-347.
WAINESS PHD QUALIFYING EXAM
112
Brunken, R., Plass, J. L., & Leutner, D. (2003). Direct measurement of cognitive load in
multimedia learning. Educational Psychologist 38(1), 53-61.
Calvert, S. L., Watson, J. A., Brinkley, V. M., & Bordeaux, B. B. (1989). Computer
presentational features for young children’s preferential selection and recall of
information. Journal of Educational Computing Research, 5(1), 35-49.
Castelli, C., Colazzo, L., & Molinari, A. (1998). Cognitive variables and patterns of hypertext
performances: Lessons learned for educational hypermedia construction [Electronic
Version]. Journal of Educational Multimedia and Hypermedia, 7(2-3), 177-206.
Chadwick, J. (1992). The development of a museum multimedia program and the effect of audio
on user completion rate. Journal of Educational Multimedia and Hypermedia, 3(1), 331340.
Chou, C., & Lin, H. (1998). The effect of navigation map types and cognitive styles on learners’
performance in a computer-networked hypertext learning system [Electronic Version].
Journal of Educational Multimedia and Hypermedia, 7(2/3), 151-176.
Chou, C., Lin, H, & Sun, C.-t. (2000). Navigation maps in hierarchical-structured hypertext
courseware [Electronic Version]. International Journal of Instructional Media, 27(2),
165-182.
Clark, R. E. (1999). The CANE model of motivation to learn and to work: A two-stage process
of goal commitment and effort [Electronic Version]. In J. Lowyck (Ed.), Trends in
Corporate Training. Leuven, Belgium: University of Leuven Press.
Clark, R. E., & Estes, F. (1999, November/December). Authentic educational technology: The
lynchpin between theory and practice. Educational Technology, 39(6), 5-13.
Clements, D. H., & Nastasi, B. K. (1999). Metacognition, learning, and educational computer
environments [Electronic Version]. Information Technology in Childhood Education, 10,
5-38.
Corno, L., & Mandinach, E. B. (1983). The role of cognitive engagement in classroom learning
and motivation. Educational Psychologist, 18(2), 88-108.
Covington, M. V. (2000). Goal theory, motivation, and school achievement: An integrative
review. Annual Review of Psychology, 51, 171-200.
Cutmore, T. R. H., Hine, T. J., Maberly, K. J., Langford, N. M., & Hawgood, G. (2000).
Cognitive and gender factors influencing navigation in a virtual environment.
International Journal of Human-Computer Studies, 53, 223-249.
Dalgarno, B. (2001). Interpretations of constructivism and consequences for computer assisted
learning. British Journal of Educational Technology, 32(2), 183-194.
Daniels, H. L., & Moore, D. M. (2000). Interaction of cognitive style and learner control in a
hypermedia environment. International Journal of Instructional Media, 27(4), 369-383.
Davidson-Shivers, G. V., Shorter, L., & Jordan, K. (1999). Learning strategies and navigation
decisions of children using a hypermedia lesson [Electronic Version]. Journal of
Educational Multimedia and Hypermedia, 8(2), 175-188.
Dias, P., Gomes, M. J., & Correia, A. P. (1999). Disorientation in hypermedia environments:
Mechanisms to support navigation. Journal of Educational Computing Research, 20(2),
93-117.
Dienes, Z., & Fahey, R. (1998). The role of implicit memory in controlling a dynamic system.
The Quarterly Journal of Experimental Psychology, 51A(3), 593-614.
WAINESS PHD QUALIFYING EXAM
113
Dillon, A., & Gabbard, R. (1998, Fall). Hypermedia as an educational technology: A review of
the quantitative research literature on learner comprehension, control, and style. Review
of Educational Research, 63(3), 322-349.
Eberts, R. E., & Bittianda, K. P. (1993). Preferred mental models for direct-manipulation and
command-based interfaces. International Journal of Man-Machine Studies, 38, 769-785.
Farrell, I. H., & Moore, D. M. (2000). The effect of navigation tools on learners’ achievement
and attitude in a hypermedia environment. Journal of Educational Technology Systems,
29(2), 169-181.
Feldman, S. (2001). The link, and how we think: Using hypertext as a teaching & learning tool.
International Journal of Instructional Media, 28(2), 153-158.
Fletcher-Flinn, C. M., & Gravatt, B. (1995). The efficacy of computer assisted instruction (CAI):
A meta-analysis. Journal of Educational Computing Research, 12(3), 219-242.
Flottemesch, K. (2000, May/June). Building effective interaction in distance education: A review
of the literature. Educational Technology, 40(3), 46-51.
Friedrichsen, P. M., Dana, T. M., & Zembal-Saul, C. (2001). Learning to teach with technology
model: Implementation in secondary science teacher education [Electronic Version]. The
Journal of Computers in Mathematics and Science Teaching, 20(4), 377-394.
Frohlich, D. M. (1997). Direct manipulation and other lessons. In M. Helander, T. K. Landauer
& P. Prabhu (eds.), Handbook of Human Computer Interaction: Second, Completely
Revised Edition (pp. 463-488). Amsterdam: Elsevier
Gerlic, I., & Jausovec, N. (1999). Multimedia: Differences in cognitive processes observed with
EEG [Electronic Version]. Educational Technology Research and Development, 47(3),
5-14.
Gevins, A., Smith, M. E., Leong, H., McEvoy, L., Whitfield, S., Du, R., & Rush, G. (1998).
Monitoring working memory load during computer-based tasks with EEG pattern
recognition methods. Human Factors, 40(1), 79-91.
Guttentag, R. E. (1984). The mental effort requirement of cumulative rehearsal: A developmental
study. Journal of Experimental Child Psychology, 37, 92-106.
Haggas, A. M., & Hantula, D. A. (2002). Think or click? Student preference for overt vs. covert
responding in web-based instruction. Computers in Human Behavior, 18, 165-172.
Harp, S. F., & Mayer, R. E. (1998). How seductive details do their damage: A theory of
cognitive interest in science learning. Journal of Educational Psychology, 90(3), 414434.
Hokanson, B., & Hooper, S. (2000). Computers as cognitive media: Examining the potential of
computers in education. Computers in Human Behavior, 16, 537-552.
Howland, J., Laffey, J., & Espinosa, L. M. (1997). A computing experience to motivate children
to complex performances [Electronic Version]. Journal of Computing in Childhood
Education, 8(4), 291-311.
Hudson, B. (1998). Group work with multimedia: The role of the computer in mediating
mathematical meaning-making [Electronic Version]. The Journal of Computers in
Mathematics and Science Teaching, 17(2/3), 181-201.
Jones, M. G., Farquhar, J. D., & Surry, D. W. (1995, July/August). Using metacognitive theories
to design user interfaces for computer-based learning. Educational Technology, 35(4),
12-22.
Kalyuga, S., Ayres, P., Chandler, P., & Sweller, J. (2003). The expertise reversal effect.
Educational Psychologist, 38(1), 23-31.
WAINESS PHD QUALIFYING EXAM
114
Kalyuga, S., Chandler, P., & Sweller, J. (1998). Levels of expertise and instructional design.
Human Factors, 40(1), 1-17.
Kalyuga, S., Chandler, P., & Sweller, J. (2000). Incorporating learner experience into the design
of multimedia instruction. Journal of Educational Psychology, 92(1), 126-136.
Kashihara, A., Kinshuk, Oppermann, R., Rashev, R., & Simm, H. (2000). A cognitive load
reduction approach to exploratory learning and its application to an interactive
simulation-based learning system. Journal of Educational Multimedia and Hypermedia,
9(3), 253-276.
Kozma, R. (2000). The relationship between technology and design in educational technology
research and development: A reply to Richey [Electronic Version]. Educational
Technology Research and Development, 48(1), 19-21.
Mane, A. M., Adams, J. A., & Donchin, E. (1989). Adaptive and part-whole-training in the
acquisition of a complex perceptual-motor skill. Acta Psychologica, 71, 179-196.
Mayer, R. E. (1998). Cognitive, metacognitive, and motivational aspects of problem solving.
Instructional Science, 26, 49-63.
Mayer, R. E., & Chandler, P. (2001). When learning is just a click away: Does simple user
interaction foster deeper understanding of multimedia messages? Journal of Educational
Psychology, 93(2), 390-397.
Mayer, R. E., Heiser, J., & Lonn, S. (2001). Cognitive constraints on multimedia learning: When
presenting more material results in less understanding. Journal of Educational
Psychology, 93(1), 187-198.
Mayer, R. E., & Moreno, R. (1998). A split-attention effect in multimedia learning: Evidence of
dual processing systems in working memory. Journal of Educational Psychology, 90(2),
312-320.
Mayer, R. E., & Moreno, R. (2003). Nine ways to reduce cognitive load in multimedia learning.
Educational Psychologist, 38(1), 43-52.
Mayer, R. E., Moreno, R., Boire, M., & Vagge, S. (1999). Maximizing constructivist learning
from multimedia communications by minimizing cognitive load. Journal of Educational
Psychology, 91(4), 638-643.
Mayer, R. E., & Sims, V. K. (1994). For whom is a picture worth a thousand words? Extensions
of a dual-coding theory of multimedia learning. Journal of Educational Psychology,
86(3), 389-401.
Mayer, R. E., Sobko, K., & Mautone, P. D. (2003). Social cues in multimedia learning: Role of
speaker’s voice. Journal of Educational Psychology, 95(2), 419-425.
McDougall, S. J. P., de Bruijn, O., & Curry, M. B. (2000). Exploring the effects of icon
characteristics on user performance: The role of icon concreteness, complexity, and
distinctiveness. Journal of Experimental Psychology: Applied, 6(4), 291-306.
Mikropoulos, T. A. (2001). Brain activity on navigation in virtual environments. Journal of
Educational Computing Research, 24(1), 1-12.
Miller, G. A. (1956). The magical number, seven, plus or minus two: Some limits on our
capacity for processing information. Psychological Review, 63, 81-97.
Moreno, R., & Mayer, R. E. (1999). Cognitive principles of multimedia learning: The role of
modality and contiguity. Journal of Educational Psychology, 91(2), 358-368.
Moreno, R., & Mayer, R. E. (2000a). A coherence effect in multimedia learning: The case of
minimizing irrelevant sounds in the design of multimedia instructional messages. Journal
of Educational Psychology, 92(1), 117-125.
WAINESS PHD QUALIFYING EXAM
115
Moreno, R., & Mayer, R. E. (2000b). Engaging students in active learning: The case for
personalized multimedia messages. Journal of Educational Psychology, 92(4), 724-733.
Moreno, R., & Mayer, R. E. (2002). Learning science in virtual reality multimedia environments:
Role of methods and media. Journal of Educational Psychology, 94(3), 598-610.
Mousavi, S. Y., Low, R., & Sweller, J. (1995). Reducing cognitive load by mixing auditory and
visual presentation modes. Journal of Educational Psychology, 87(2), 319-334.
Newell, K. M., Carlton, M. J., Fisher, A. T., & Rutter, B. G. (1989). Whole-part training
strategies for learning the response dynamics of microprocessor driven simulations.
Acta Psychologica, 71, 197-216.
Niemela, M., & Saarinen, J. (2000, Winter). Visual search for grouped versus ungrouped icons in
a computer interface. Human Factors, 42(4), 630-635.
Paas, F., Renkl, A., & Sweller, J. (2003). Cognitive load theory and instructional design: Recent
developments. Educational Psychologist, 38(1), 1-4.
Paas, F., Tuovinen, J. E., Tabbers, H., & Van Gerven, P. W. M. (2003). Cognitive load
measurement as a means to advance cognitive load theory. Educational Psychologist,
38(1), 63-71.
Park, O.-C., & Gittelman, S. S. (1995). Dynamic characteristics of mental models and dynamic
visual displays. Instructional Science, 23, 303-320.
Renkl, A., & Atkinson, R. K. (2003). Structuring the transition from example study to problem
solving in cognitive skill acquisition: A cognitive load perspective. Educational
Psychologist, 38(1), 13-22.
Renkl, A., Atkinson, R. K., Maier, U. H., & Staley, R. (2002). From example study to problem
solving: Smooth transitions help learning. The Journal of Experimental Education, 70(4),
293-315.
Rieber, L. P. (1996). Animation as a distractor to learning. International Journal of Instructional
Media, 23(1), 53-57.
Salomon, G. (1983). The differential investment of mental effort in learning from different
sources. Educational Psychology, 18(1), 42-50.
Schacter, J., & Fagnano, C. (1999). Does computer technology improve student learning and
achievement? How, when and under what conditions? Journal of Educational
Computing Research, 20(4), 329-343.
Schraw, G. (1998). Processing and recall differences among seductive details. Journal of
Educational Psychology, 90(1), 3-12.
Shebilske, W. L., Regian, W., Arthur, W., Jr., & Jordan, J. A. (1992). A dyadic protocol for
training complex skills. Human Factors, 34(3), 369-374.
Shyu, H.-y., & Brown, S. W. (1995). Learner-control: The effects of learning a procedural task
during computer-based videodisc instruction. International Journal of Instructional
Media, 22(3), 217-230.
Svendsen, G. B. (1991). The influence of interface style on problem solving. International
Journal of Man-Machine Studies, 35, 379-397.
van Merrienboer, J. J. G., Kirschner, P. A., & Kester, L. (2003). Taking a load off a learner’s
mind: Instructional design for complex learning. Educational Psychologist, 38(1), 5-13.
Yeung, A. S., Jin, P., & Sweller, J. (1997). Cognitive load and learner expertise: Split-attention
and redundancy effects in reading with explanatory notes. Contemporary Educational
Psychology, 23, 1-21.
WAINESS PHD QUALIFYING EXAM
116
3. Review the theoretical and empirical literature on the impact of scaffolding on
learning. Include a discussion of types (e.g., graphical scaffolding) and contexts (e.g.,
K-12).
Mayer, Mautone, and Prothero (2002) commented that a major instructional issue in
learning by doing within simulated environments concerns the proper type of guidance, which
they refer to as cognitive apprenticeship. Using a discovery-based geological game, the
researchers argued that results of the study indicate that adding pictorial scaffolding to the
learning materials lead to improved performance on a transfer for both high- and low-spatial
students in the Profile Game.
Moreno and Mayer (2000) have shown how personalization can improve learning (based
on performance outcomes in both retention and transfer), based on theories that “self-referential
language promotes the elaboration of the instructional materials” (p. 725), and “personalized
messages are more consistent with our schemas for communicating in normal conversations and
therefore require less cognitive effort to process” Their study focused on the use of an active
pedagogical agent, a form of scaffolding where an animate object (either visual or auditory or
just auditory) provides support during learning. As a result of their findings, the researchers
argued that “multimedia science programs can result in broader learning if the communication
model is centered around shared environments in which the student is addressed as a participant
rather than as an observer” (p. 731).
In a series of five random assignment experiments using University of California, Santa
Barbara Psychology students as subjects, Moreno and Mayer (2000) examined the impact of
personalization of multimedia messages on learning outcomes. The experiments were based on
the assumptions that “self-referential language promotes the elaboration of the instructional
materials” (p. 725) and “personalized messages are more consistent with our schemas for
communicating in normal conversations and therefore require less cognitive effort to process” (p.
725; see Moreno and Mayer, 2000, for evidence supporting these assumptions). In each
experiment, a computer program was used for teaching how lightning works (students were pretested for prior knowledge). One group was given neutral messages, while the other received
personalized messages.
Results in the Moreno and Mayer (2000) experiments supported use of personalized
messages to increase performance. In all five experiments, those receiving personalized
messages (whether textual or auditory) scored significantly higher on transfer tests, while results
for retention varied. Retention increased for the game group when a pedagogical agent was
added to the game. An interesting result of the experiment was that even though the addition of
the pedagogical agent increased retention, the favorableness rating for using a pedagogical agent
with or without personalization was not significant. The researchers suggested the lack of
significance for might have been due to the nature of the questions and that a more sensitive set
of survey questions might produce different results. As a result of the study, Moreno and Mayer
(2000) argued that “multimedia science programs can result in broader learning if the
communication model is centered around shared environments in which the student is addressed
as a participant rather than as an observer” (p. 731). It should be noted that, while Moreno and
WAINESS PHD QUALIFYING EXAM
117
Mayer, referred to the instrument as a game, it appears to fit the Gredler’s definition of a
simulation, not a game or simulation game (Gredler, 1996).
An overview of player position was considered an important feature in adventure games.
Players reported that help functions, hints, and examples were necessary in adventure,
miscellaneous, and word games. Mystery, intrigue, and suspense were pleasing to some players.
Many liked the idea of games with familiar scenarios or stories (Dempsey, Haynes, Lucassen, &
Casey, 2002).
Instructional supports include the following elements that are listed by Alessi (2000):
explaining or demonstrating the phenomenon or procedure; giving hints and prompts before
student actions; giving feedback following student actions; providing a coach, advice, or help
system; providing dictionaries and glossaries; providing user controls not needed in noninstructional simulation; and giving summary feedback or debriefing (Leemkull, de Jong, de
Hoog, & Christoph, 2003).
The success of a VR highly depends on the friendliness of the user interface. Upon
entering the dynamic 3-D virtual representation of the solar system, the user has to project
himself into this “reality” and to adopt new looking points, which is by no means an easy
cognitive task, especially at young ages (Yair, Mintz, & Litvak, 2001).
The lose of orientation and “vertigo” feeling which often accompanies learning in a
virtual-environment is minimized by the display of a traditional, two-dimensional dynamic map
of the solar system. The map helps to navigate and to orient the user, and facilitates an easier
learning experience (Yair, Mintz, & Litvak, 2001).
The Touch the Sky, Touch the Universe program lets students interact directly with
various forms of multimedia that simulate resources used by practicing scientists. Journeys
through the virtual simulations of the solar system and the Milky Way help students bridge the
gap between the concrete world of nature and the abstract realm of concepts and models. As
students examine images, manipulate three-dimensional models, and participate in these virtual
simulations, they enhance their understanding of scientific concepts and processes. Students are
not simply passive recipients of prepackaged multimedia content, and cause use a variety of
computerized tools to view, navigate, and analyze a realistic three-dimensional representation of
space (Yair, Mintz, & Litvak, 2001).
Three simulation packages were selected, DRAX, FLOWERS, and LAB, because each
contained a different type of simulation: physical, procedural, and process simulations,
respectively. This categorization is related directly to different types of mental processing and is
particularly useful in students of conceptual learning (Yildiz & Atkins, 1996).
A physical simulation requires the learner to construct a mental model of how a system
functions based on causal relationships between entities that form part of that system. Free
discovery or guided discovery methods may be embedded in this type of simulation (Yildiz &
Atkins, 1996).
A procedural simulation is one designed to train the user to perform certain tasks in a
systematic way, correcting anomalies, mistakes, or disturbances which may arise. Feedback on
errors made, and the opportunity to repeat procedures many times, are characteristic features of
this type of simulation (Yildiz & Atkins, 1996).
WAINESS PHD QUALIFYING EXAM
118
A process simulation tries to help the student to understand the progression of a
dynamic system. Normally it is run several times with different initial values for the parameters
(Yildiz & Atkins, 1996).
DRAX, which fits the general characteristics of a physical simulation, was designed to
improve students’ understanding of how electricity is made in power stations by enabling them
to obtain a surrogate experience of what a coal fired power station is like, what happens in each
of the main buildings, and the process by which electricity is made (Yildiz & Atkins, 1996).
FLOWERS, which contains the characteristics of a procedural simulation, was
designed to illustrate the probabilistic nature of experimental results and to teach students
scientific investigation methods. It included a wide range of statistical tools which could be
called up as required by the user and were intended to improve students’ skills in constructing
and interpreting graphs. Users were placed in the situation of conducting an experiment in
growing flowers in which they had contro over four key interrelated variables that affect growth:
nitrogen, temperature, potash, and length of daylight (Yildiz & Atkins, 1996).
LAB, which fit the characteristics of a process simulation, enabled students to
understand the relationships between gravity, speed, height, time, etc. The LAB was a room of
on-screen experiments relating to energy. On-screen tools allowed the users to measure distance,
time, and velocity in several different ways (Yildiz & Atkins, 1996).
A study using the three simulations was conducted using 2296 students aged 11 to 18
years, randomly selected from two schools in North East England. A test was designed to cover
the specific learning objectives of each simulation (Yildiz & Atkins, 1996).
Results indicate that Interactive Video (IV) simulations can interact in complex ways
with both gender and prior achievement characteristics. Nevertheless, DRAX, the physical
simulation based on the power station, produced the greatest cognitive gain. The reason for this
may well lie in the design of this simulation, which applied several important principles derived
from learning theories. For example, at every point, it enabled students to obtain advance
information (scaffolding) about what they could do and could expect; it helped students to relate
new information to what they already knew from school physics, and it made use of animations,
computer graphics, and games to reinforce nascent understanding. It also allowed students to
decide their own learning route through the material, and it gave students immediate feedback on
how they were doing with the on-screen experiments (Yildiz & Atkins, 1996).
By comparison, although FLOWERS, the procedural simulation, provided some
conceptual scaffolding and had real life relevance, game students little freedom of choice about
how to solve the problem they were presented with. It also required a sophisticated approach
(e.g., holding one variable constant while altering another). Due to curriculum time constraints,
the task seemed to be beyond the capability of the middle and low achieving students (Yildiz &
Atkins, 1996).
The process simulation, LAB, lacked an explicit explanatory framework. There were
no links to real life referents or examples of the application of the principles of physics being
demonstrated. The feedback had to be worked out by the students themselves by interpreting the
on-screen read outs of distance, speed, etc., making it more difficult to develop explicit
hypothesis-test-interpret-hypothesis-test chains. For the middle achieving students, the facility to
repeat the same experiment many times seemed to have been helpful, perhaps building
confidence in their understanding. For high achieving students, the lack of challenge and variety
may have become obstacles to developing understanding and may have been the factors which
led to a lower score on the post-test than the pre-test. For pupils with low prior achievement, the
WAINESS PHD QUALIFYING EXAM
119
lack of clear learning goals and advice may have prevented learning from occuring (Yildiz &
Atkins, 1996).
Carroll, W. M. (1994). Using worked examples as an instructional support in the algebra
classroom. Journal of Educational Psychology, 86(3), 360-367.
Cary, M., & Carlson, R. A. (1999). External support and the development of problem-solving
routines. Journal of Experimental Psychology: Learning, Memory, and Cognition, 24(4),
1053-1070.
Chi, M. T. H., Bassok, M., Lewis, M. W., Reimann, P., & Glaser, R. (1989). Self-explanations:
How students study and use examples in learning to solve problems. Cognitive
Science, 13, 145-182.
de Jong, T., de Hoog, R., & de Vries, F. (1993). Coping with complex environments: The effects
of providing overviews and a transparent interface on learning with a computer
simulation. International Journal of Man-Machine Studies, 39, 621-639.
Gertjets, P., & Scheiter, K. (2003). Goal configurations and processing strategies as moderators
between instructional design and cognitive load: Evidence from hypertext-based
instruction. Educational Psychologist, 38(1), 33-41.
Jimenez, L., & Mendez, C. (2001). Implicit sequence learning with competing explicit cues. The
Quarterly Journal of Experimental Psychology, 54A(2), 345-369.
Kalyuga, S., Chandler, P., Tuovinen, J., & Sweller, J. (2001). When problem solving is superior
to studying worked examples. Journal of Educational Psychology, 93(3), 578-589.
Kee, D. W., & Davies, L. (1988). Mental effort and elaboration: A developmental analysis.
Contemporary Educational Psychology, 13, 221-228.
WAINESS PHD QUALIFYING EXAM
120
Kee, D. W., & Davies, L. (1990). Mental effort and elaboration: Effects of Accessibility and
instruction. Journal of Experimental Child Psychology, 49, 264-274.
Kee, D. W., & Davies, L. (1991). Mental effort and elaboration: A developmental analysis of
accessibility effects. Journal of Experimental Child Psychology, 52, 1-10.
Khine, M. S. (1996). The interaction of cognitive style with varying levels of feedback in
multimedia presentation. International Journal of Instructional Media, 23(3), 229-237.
Mayer, R. E., Mautone, P., & Prothero, W. (2002). Pictorial aids for learning by doing in a
multimedia geology simulation game. Journal of Educational Psychology, 94(1), 171185.
Mautone, R. D., & Mayer, R. E. (2001). Signaling as a cognitive guide in multimedia learning.
Journal of Educational Psychology, 93(2), 377-389.
Murphy, N. & Messer, D. (2000). Differential benefits from scaffolding and children working
alone. Educational Psychologist, 20(1), 17-31.
Mwangi, W., & Sweller, J. (1998). Learning to solve compare word problems: The effect of
example format and generating self-explanations. Cognition and Instruction, 16(2), 173199.
Neale, D. C., & Carroll, J. M. (1997). The role of metaphors in user interface design. In M.
Helander, T. K. Landauer & P. Prabhu (eds.), Handbook of Human Computer
Interaction: Second, Completely Revised Edition (pp. 441-462). Amsterdam: Elsevier
Quilici, J. L., & Mayer, R. E. (1996). Role of examples in how students learn to categorize
statistics word problems. Journal of Educational Psychology, 88(1), 144-161.
WAINESS PHD QUALIFYING EXAM
121
Renkl, A., & Atkinson, R. K. (2003). Structuring the transition from example study to problem
solving in cognitive skill acquisition: A cognitive load perspective. Educational
Psychologist, 38(1), 13-22.
Renkl, A., Atkinson, R. K., Maier, U. H., & Staley, R. (2002). From example study to problem
solving: Smooth transitions help learning. The Journal of Experimental Education, 70(4),
293-315.
Rosswork, S. G. (1977). Goal setting: The effects on an academic task with varying magnitudes
of incentive. Journal of Educational Psychology, 69(6), 710-715.
Seaward, M. R. (1998). Interactive assistants provide ease of use for novices: The development
of prototypes and descendants. Computers in Human Behavior, 14(2), 221-237.
Tarmizi, R. A., & Sweller, J. (1988). Guidance during mathematical problem solving. Journal of
Educational Psychology, 80(4), 424-436.
Tuovinen, J. E., & Sweller, J. (1999). A comparison of cognitive load associated with discovery
learning and worked examples. Journal of Educational Psychology, 91(2), 334-341.
van Merrienboer, J. J. G., Clark, R. E., & de Croock, M. B. M. (2002). Blueprints for complex
learning: The 4C/ID-model. Educational Technology Research & Development, 50(2),
39-64.
Weidenbeck, S. (1989). Learning iteration and recursion from examples. International Journal of
Man-Machine Studies, 30, 1-22.
Wu, A. K. W., & Lee, M. C. (1998). Intelligent tutoring systems as design. Computers in Human
Behavior, 14(2), 209-220.