research - Istituto Marco Belli

Overview of Research Outline
Ways of Knowing
The Scientific Approach
What is Research?
Scientific Method: The Research Loop
Goals of Science
Types of Research
4 Types of Research Designs
Types of Variables
Hypotheses and Theory
Which Design to Use?
Evaluating Research
Intuitive vs. Empirical Sources of
 Tenacity (e.g., habit, repeated exposure,
 Authority (e.g., moral, psychological,
intellectual leaders)
 Reason (i.e., logical syllogisms)
 Serendipity (e.g., Columbus’ & India)
Empiricism (relies on formal methods of
Limitations of Intuition
Intuition relies unquestioningly on personal
Involves cognitive and motivational biases
Erroneous conclusions about cause and effect
(e.g., Fundamental Attribution Error)
Skepticism, Science, and the Empirical
Peer review
Integrating Intuition, Skepticism, and
Elements of Goodstein’s Evolved
Theory of Science
Observations accurately reported to others
Search for discovery and verification of ideas
Open exchange and competition among ideas
Peer review of research
Characteristics of pseudoscience
Hypotheses generated are not typically testable
If scientific tests are reported, methodology is not
scientific and validity of data is questionable
Supportive evidence is anecdotal and does not cite
scientific references
Claims ignore conflicting evidence
Claims tend to be vague, and appeal to pre-conceived
Claims are never revised
What is Research?
A systematic process of asking and attempting
to answer questions about the world.
Scientific Method: 5 Steps
Identify a Research Problem or Question
Develop Hypotheses and Research Design
Conduct the Study
Test the Hypotheses
Communicate Results**
-File Drawer Problem
Research findings often create new research
questions which result in a “Research Loop”
Scientific Method Steps
Goals of Science
 Description:
-What is the relationship found between alcohol
consumption and aggression?
 Prediction:
-How many aggressive acts will a person exhibit
if he/she were to consume 3 beers?
 Causation & Explanation:
-Does alcohol intoxication and/or social
expectations about drinking behavior increase
Determining Causality
Temporal Precedence: The cause must
precede the effect
2. Covariation Principle: The effect occurs
when the cause is present, but does not
occur when the cause is not present
3. Elimination of Alternative Explanations:
Plausible explanations for the causal
relationship are ruled out.
Types of Research Approaches
Basic Science vs. Applied Approaches
 Qualitative vs. Quantitative Approaches
Basic Science
Applied Approaches
Basic Science:
Further an understanding of psychological
functioning (e.g., What is the nature or basis of
people’s self-esteem?)
Applied Research:
-Provide practical solutions to optimize psychological
functioning (e.g., How do we increase people’s selfesteem?)
Quantitative Approaches
Qualitative Approach:
Collects and examines responses from people without
translating the responses into numbers; Emphasis on
Broad Themes of phenomena of interest.
Quantitative Approach:
Similar to qualitative but does rely on numbers to
represent phenomena. Emphasis on Specific Tests of
phenomena of interest.
3 Basic Types of Research
Non Experimental Designs
-Program Evaluation
True Experimental Designs
Manipulation –Randomization -Control Group
Quasi-Experimental Designs
Similar to true experimental method but lacks 1 or more
of its core features
Psychological Variables &
The “Soft Science”
A Variable = any observable characteristic that is not constant
Psychological variables = characteristics of psychological
functioning (i.e., thoughts, feelings, and behaviors, and the
processes that regulate them) that vary in people.
-”Hypothetical Constructs” that by nature are abstract ideas/concepts
-Often are ambiguous,
-Difficult to define and measure
Psychology is considered a “soft science” because of the
inherent qualities of the variables that are examined.
-To scientifically examine research questions involving psychological
variables they must be precisely defined and measured.
Conceptual and Operational
Definitions of Variables
Empirical examination of psychological variables,
requires that they be (1) conceptually defined, and
(2) translated into concrete, observable forms.
Conceptual Definitions = broadly based dictionary
like descriptions of a variable.
(e.g., Self-Esteem = The general extent to which a
person has positive or negative feelings of self-worth).
2. Operational Definition of variables = the specific
way that a variable is measured, manipulated, or
“operating” within the context of a study.
(e.g., Self-Esteem = the score a person gets on a selfesteem questionnaire that he/she had completed).
The Basic Structure of a Study
Studies examine the pattern of
relationships among variables:
Independent Variables (IVs)
In general, an IV is theorized to exert meaningful influence on a
particular outcome (i.e., DV).
Within true-experiments an IV is systematically manipulated to
cause changes in the DV.
Dependent Variables (DVs)
In general, a DV reflects the characteristic that researchers are
seeking to describe, predict, or explain.
Reflects the outcome that emerges in response to the IVs.
Differences in the DV are attributed to (or are caused by) the
varying amounts or levels of the IV (i.e., the values of the DV
depend upon the levels of the IV)
Types of Variables
Participant/Subject Variables
 Situational Variables
 Response Variables
 Mediating Variables
 Control Variables
 Confounding Variables
Assessing Variables:
4 Types of Measures
 Behavioral
 Behavioroid
 Physiological
Issues of Interpreting IV and DV
3rd Variable
Limits often exist when interpreting the relationship
between IVs and DVs, particularly “causal” relations.
Directionality problem
X  Y; Y  X; or bi-directionality?
The Third variable problem
When the IV & DV relationship is attributable to another variable
Systematic vs. Random Influences Upon
the IV and DV Relationships
Confounding Variable
 Reasoned to exert
meaningful influence on the
IV & DV relationship
IS NOT measured
Because its value is
unknown, its influence is also
Control Variable
 Reasoned to exert meaningful
influence on the IV & DV
reflects random error, or
(uncontrolled) influence.
Blurs, distorts, or obscures
the true IV & DV
relationships, complicates
theoretical understanding of
IS intentionally measured
Because its value is known,
its influence is detectable and
systematically observed.
i.e., Determining the IV & DV
relationship after having
controlled for (kept constant)
the influence of the 3rd
Helps to more clearly
illustrate the IV & DV
relationships and a theoretical
understanding of them.
Hypotheses and Theories
Hypothesis = a formally stated testable prediction
about the pattern of relationships to be found
among variables
Theory = A set of interrelated statements
proposed to explain phenomena
Interplay Between Theory and
Research tests theory
-suggests changes
Theory guides research
Approaches to Research
Top-Down vs. Bottom-Up Approaches
Hypotheses may be used to create or broaden
theories by inductive reasoning
Theories may be used to test hypotheses by
deductive reasoning (Top-Down)
Bottom-Up/Inductive Reasoning
Reasoning that proceeds from specific cases to
general conclusions or theories.
Used in early stages of development of studying a
Relies on observations or data to help create or
extend prior ideas or theories about relationships
that exist in the world
Top-Down: Deductive Reasoning
Reasoning that proceeds from general theories to
specific cases.
Used when there is a well-established theory that
predicts certain observable outcomes.
Once a comprehensive theory is established specific
predictions may be made from the theory and tested
through experimentation
Which Design to Use?
It depends on your particular goals, but…
Begin with non experimental methods that capitalize on identifying and
describing relationships between variables.
-For example, first
observe behaviors in natural settings (i.e.,
field studies), and then utilize interviews or survey research (i.e.,
correlational studies).
Then, use Quasi-Experimental and True-Experimental Designs to help
explain the relationships between variables.
For example, determine the causal relationships between variables or how
much change occurs in a DV in response to manipulated levels of an IV
Ultimately, Convergence (similarity) in the pattern of findings across
multiple methods is best!
Evaluating Research
A variety of standards exist for assessing the quality of a
theory and research findings.
1. Reliability = consistency in measurement
2. Validity = accuracy of constructs and measures
-construct validity
3. Replication = consistency in research findings when
comparing equivalent or similarly conducted studies
Distinguishing Reliability and Validity
Not Reliable,
Not Valid
Not Valid