Revised Knox Preschool Play Scale:
Interrater Agreement and Construct Validity
Milena Jankovich, Jacqueline Mullen, Erin Rinear,
Kari Tanta, Jean Deitz
KEY WORDS
• evaluation
• development
• pediatrics
• play
OBJECTIVE. Interrater agreement and construct validity of the Revised Knox Preschool Play Scale (RKPPS)
were examined.
METHOD. Two separately trained raters evaluated 38 typically developing children, ages 36 to 72 months.
For each child, the raters observed two 15-min free-play sessions.
RESULTS. For the overall play age, the scores of the two raters were within 8 months of each other 86.8% of
the time; for the 4 dimension scores, they were within 12 months of each other 91.7% to 100% of the time; and
for the 12 category scores, they were within one age level of each other 81.8% to 100% of the time. Construct
validity results showed a general match between the children’s chronological ages and their overall play age
scores.
CONCLUSIONS. Findings suggest that two raters can score the RKPPS with some consistency and that
scores on this measure progress developmentally, thus supporting its construct validity.
Jankovich, M., Mullen, J., Rinear, E., Tanta, K., & Deitz, J. (2008). Revised Knox Preschool Play Scale: Interrater agreement
and construct validity. American Journal of Occupational Therapy, 62, 221–227.
Milena Jankovich, MOT Student, Division of
Occupational Therapy, Department of Rehabilitation
Medicine, University of Washington, Seattle.
Jacqueline Mullen, MOT Student, Division of
Occupational Therapy, Department of Rehabilitation
Medicine, University of Washington, Seattle.
Erin Rinear, MOT Student, Division of Occupational
Therapy, Department of Rehabilitation Medicine, University
of Washington, Seattle.
Kari Tanta, PhD, OTR/L, FAOTA, is Clinical Assistant
Professor, Department of Rehabilitation Medicine,
University of Washington and Program Coordinator,
Children’s Therapy Department, Valley Medical Center,
Seattle.
Jean Deitz, PhD, OTR/L, FAOTA, is Professor,
Department of Rehabilitation Medicine, Box 356490,
University of Washington, Seattle, WA 98195;
deitz@u.washington.edu
P
lay is an important performance area of occupation to a child (American
Occupational Therapy Association [AOTA], 2002). Through play, children
learn to communicate, grow, and build the skills necessary to function in society
(Knox, 2005; Richmond, 1960). According to the Occupational Therapy Practice
Framework (AOTA, 2002), participating in play is a fundamental part of growth
and development throughout the life span. Motor, process, and communication
and interaction skills need to be integrated for a child to be a successful player. If a
child has sensorimotor, emotional, or social deficits, his or her ability to play may
be compromised. One of the roles of occupational therapists is to facilitate play to
ensure quality of life and optimal developmental outcomes (Takata, 1969).
Measures of play approximate cognitive and developmental level (Bergen,
1988), and observation of play gives occupational therapists the ability to evaluate
motor, process, and communication and interaction skills (AOTA, 2002). By evaluating play behaviors, a therapist can measure developmental competence and change
in a child (Schaaf & Mulrooney, 1989) and observe play deficits and the effectiveness of treatment on improving play skills (Bundy, 1991). Assessing the occupation
of play is also used to determine eligibility for services for children with special needs,
to facilitate intervention planning, and to understand a child’s preferred activities
(Lifter, 1996). To meet these aims, occupational therapists conducting play observations need “a systematic means of observing play behavior to determine a child’s
play assets, skills, and deficits” (Bledsoe & Shepherd, 1982, p. 783).
Although assessing play is important, it is not an easy construct to measure.
Knox (2005) contended that play is difficult and time-consuming to measure
because of the need to observe the child engaged in the occupation of play. Several
The American Journal of Occupational Therapy
Downloaded From: http://ajot.aota.org/ on 11/10/2015 Terms of Use: http://AOTA.org/terms
221
authors (Bundy, 1991; Knox, 2005; Smith, Takhvar, Gore,
& Vollstedt, 1985) stated that, to get a complete assessment
or observation of play, it must be conducted in multiple
environments. These challenges in measuring the construct
of play make it difficult for occupational therapists to find
an appropriate measurement tool to use.
According to Couch (1996, p. 77), “Assessment of play
skills in pediatrics is a critical aspect of developmental evaluation, yet therapists are challenged to find tools to reach this
aim.” The field of occupational therapy is currently addressing
the need for the development of measurements of play behaviors. Occupational therapists use several assessments of play,
each of which yields a different measure of the construct of
play. Although some of these assessments were developed years
ago, they are still being used or revised for current practice.
We highlight those frequently referenced in the literature.
The Parten Social Play Hierarchy (Parten, 1932),
Lunzer’s Scale of Organization of Play Behavior (Lunzer,
1959), Guide to Play Observation (Robinson, 1977),
Growth Gradient (Michelman, 1971), and Guide to Status
of Imitation (deRenne-Stephan, 1980) examine important
components of play (e.g., learning and organization of rules,
creativity, and imitation) but not play as an overall outcome. The Play History (Takata, 1969) is a qualitative
assessment of play through interview and observation that
provides a description of a child’s play behavior but does
not describe it according to a developmental scheme. The
Play Skills Inventory (Hurff, 1974) measures four play skills
in the areas of sensation, motor, perception, and intellectual
functioning and is conducted by observing predetermined
activities. The Preschool Play Scale (Bledsoe & Shepherd,
1982) examines play as an overall outcome according to a
developmental scheme.
In a study addressing the use of specific assessment tools
designed to evaluate play behaviors, Couch, Deitz, and
Kanny (1998) found that the Preschool Play Scale was used
most frequently by pediatric occupational therapists. These
investigators further identified the Preschool Play Scale as a
promising tool in need of additional study. Since then, the
Preschool Play Scale has undergone several changes and has
been used as a research and clinical tool (Bundy, 1989;
Restall & Magill-Evans, 1994). The current version of the
Preschool Play Scale, the Revised Knox Preschool Play Scale
(Knox, 1997), can be best understood by examining its history of development and revision.
Knox (1974) developed a play assessment, called A Play
Scale, to aid in determining children’s developmental play
ages. The tool has three main advantages. First, it was designed
to cover all areas of development and to reflect developmental
status (Knox, 1997, p. 46). Second, it is applicable for use
with children who are unable to perform a standardized
developmental assessment because the ­children are observed
participating in play as they normally would. Third, it assesses
children in their natural environments.
In 1982, Bledsoe and Shepherd revised and renamed the
scale the Preschool Play Scale to reflect current findings in
the play literature. The scale was further revised by Knox
(1997) after “review of its usefulness . . . and limitations” (p.
46) and renamed the Revised Knox Preschool Play Scale
(RKPPS). Changes were made related to scoring increments,
clarity of descriptors, and recent findings related to play
development (Bergen, 1988; Knox, 1997; Rubin, Fein, &
Vandenberg, 1983). Examples of changes, based on findings
related to stages of play development (Bergen, 1988), include
the addition of “pretend with replica toys” and “uses one toy
to represent another” as descriptors to the Dramatization
category and of “games with rules” as a descriptor to the
Cooperation category.
The RKPPS consists of the 4 dimensions and 12 categories of play behaviors (Knox, 1997) depicted in Figure 1. The
resulting play age and play profile provide therapists with
useful information to plan and implement intervention.
Although the RKPPS is available for use, occupational
therapists continue to use the Preschool Play Scale rather than
the RKPPS because reliability and validity data are available
for the Preschool Play Scale (K. Tanta, personal communication, September 15, 2005). To improve the RKPPS’s clinical
utility, its psychometrics merited examination. Specifically,
interrater agreement data were needed because of the subjectivity of scoring; also, to examine its construct validity, it was
important to know whether children who are typically developing score in their expected age categories. Therefore, the
research questions for this study were as follows:
• What is the interrater agreement for the RKPPS for the
overall play age score, for each of the 4 dimension scores,
and for each of the 12 category scores? and
• How do the RKPPS overall play age scores of typically developing children relate to the children’s chronological ages?
Method
Participants
A convenience sample of children from the greater Seattle
area participated in this study. This sample consisted of 38
children, 17 boys and 21 girls. Seven boys and 7 girls were
36 to 48 months old; 6 boys and 11 girls were 48 to 60
months old; and 4 boys and 3 girls were 60 to 72 months
old. All children involved in the study were typically developing, had parental consent, and gave verbal assent to participate in the study. A child information form completed
by a parent was used to determine whether a child was
222
Downloaded From: http://ajot.aota.org/ on 11/10/2015 Terms of Use: http://AOTA.org/terms
March/April 2008, Volume 62, Number 2
communication, October 11, 2005), we decided to use two
15-min sessions. This is also congruent with the decision of
Harrison and Kielhofner (1986) to use two 15-min observations in their study using the Preschool Play Scale. Additionally,
although the age range for the RKPPS is from birth to 72
months, we focused on children ages 36 to 72 months because
our clinical experience suggested that this is the age group for
which this measure is used most frequently.
Procedure
Study Design and Scale Administrators
Figure 1. Dimensions and categories of the Revised Knox
Preschool Play Scale.
t­ ypically developing. “Typically developing” was defined as
not having diagnoses such as autism spectrum disorder, cerebral palsy, or developmental delay; not having current or
prior special education or therapy services; and not using
crutches, canes, walkers, wheelchairs, or other mobility aids
at the time of the study.
Setting
Following approval from the human subjects institutional
review board at the University of Washington, children were
recruited for this study from the Experimental Education
Unit at the University of Washington and two private preschools. Children in these programs tend to come from
families of middle to high socioeconomic status. All settings
had free-play periods, both indoors and outdoors, integrated
into children’s classroom routines. Each child had participated in his or her respective free-play periods for at least 1
month before observation. A variety of toys and play equipment were available, including items such as dramatic and
pretend play props, construction toys, arts and crafts supplies, and playground equipment. Two or more children
were in a play space (playground, gym, classroom) at any
given time, allowing opportunities for peer interaction.
Instrumentation
The RKPPS is an observational measure that allows therapists
to evaluate the play of children ages birth to 72 months in
their natural environments (Knox, 1997). Per the instructions, a child should be observed for two 30-min sessions in
two different settings, one indoors and one outdoors.
However, after consultation with the test developer, who
stated that it was acceptable to use two 15-min sessions (S.
Knox, personal communication, August 24, 2005), and with
an experienced pediatric therapist, who indicated that a full
hour of observation was not practical (K. Tanta, personal
If a test has good interrater reliability, it is assumed that two
individuals who prepare separately (i.e., therapists working
in different clinics) will score it similarly. Therefore, for the
purposes of this study, the primary scale administrators (two
master of occupational therapy students [MOT1 and
MOT2]) prepared separately and did not communicate with
each other with regard to concepts, administration, or scoring. Neither had experience with the RKPPS before the
study. Before data collection began, each student read and
studied information published about the RKPPS and practiced using it with several children who were typically developing. The latter continued until each of the scale administrators was scoring consistently with her partner. MOT1
trained with an experienced therapist who had used the
Preschool Play Scale extensively in her clinic and was very
familiar with the RKPPS. MOT2 trained with MOT3, who
also had previously been unfamiliar with the RKPPS. Before
and during data collection, MOT2 and MOT3 communicated with the creator of the RKPPS for clarification of concepts, definitions, and scoring. This allowed each primary
scale administrator access to a person with expert knowledge
of the RKPPS. Training continued until each of the scale
administrators was scoring consistently with her partner.
During data collection, observer drift was monitored after
assessment of every 15 participants. This was accomplished
by each primary administrator scoring one child with her
original training partner.
Scoring
The RKPPS provides numerous scores, including an overall
play age, 4 dimension ages, and 12 category ages. Descrip­tors
(e.g., “uses words to communicate with peers,” “communicates with peers to organize activities”; Knox, 1997, p. 49)
illustrate the developmental play progression within the context of the 12 categories and 4 dimensions. The developmental progression is broken down in 6-month increments from
0 to 36 months (e.g., 12–18 months) and 12-month increments from 36 to 72 months (e.g., 36–48 months).
The American Journal of Occupational Therapy
Downloaded From: http://ajot.aota.org/ on 11/10/2015 Terms of Use: http://AOTA.org/terms
223
To determine each category score, the rater must review
all descriptors and decide at which age level they are representative of the child’s play behavior, as opposed to emerging
skills. Because some of the printed instructions for scoring
the RKPPS have been identified by therapists as being
unclear, Kari Tanta, with approval from the test developer,
wrote supplemental instructions (S. Knox & K. Tanta, personal communication, October 11, 2005). The Tanta
method of scoring involves giving a child credit within an
age category only for those behaviors that are truly representative of the child’s play behavior and not for emerging skills
or behaviors that occur fleetingly. This method was used in
combination with the instructions that accompany the
RKPPS. As raters observe play behaviors, a mark is placed
over the corresponding descriptors and then scored by looking at the respective age level (Knox, 1997). For example, if
a rater places a mark in the 36- to 48-month age level, the
child is scored as 48 months. A dimension score is the mean
of its category scores, and the overall play age score is the
mean of the four dimension scores (Knox, 1997).
In situations in which a category behavior is not exhibited during the time of observation, a rater has an option of
scoring a child with an NA (not applicable). Although not
described in the chapter regarding the RKPPS (Knox, 1997),
Knox and Kari Tanta (personal communication, October
11, 2005) indicated that a rater also has the option of asking
a child an open-ended question if necessary in an attempt to
more clearly identify the type of play in which the child is
engaging (e.g., “What are you playing?). For purposes of this
study, the requirement that both raters hear the question and
response was added.
Data Collection
Once parental consent and child assent were obtained and
the child information form was completed, the administrators began observing. They positioned themselves in the
child’s environment for a minimum of 5 min before beginning official observations. This was done to help control for
the novelty of having new people in the child’s environment.
Each child was observed for two 15-min free-play sessions
(Harrison & Kielhofner, 1986), one indoors and one outdoors. For each child, these generally took place on the same
day. When this was not possible, observations were completed within 1 week to control for the possibility of developmental changes.
The two administrators refrained from communicating
with each other with regard to the RKPPS throughout all
testing. Each administrator kept a log in which she described
the dimensions and categories she found to be particularly
challenging and the reasons why. These were later used to
determine whether there were specific sections of the RKPPS
that were consistently unclear to both administrators and thus
potentially in need of further review by the test developer.
Data Analysis
Because of the nature of the data generated by the RKPPS,
scorer consistency related to the first research question was
examined using percentage of agreement rather than reliability statistics (e.g., Pearson product–moment correlation
coefficients, intraclass correlation coefficients). According
to Tinsley and Weiss (1975), one of the benefits of this
approach is that it “represents the extent to which different
judges tend to make exactly the same judgments about the
rated subject” (p. 359). This approach is clinically relevant,
and using this approach also makes it possible to examine
the extent of agreement within increasing magnitudes of
difference (e.g., determining percentage of agreement for
ratings in the category of participation within a 6-month
window rather than looking exclusively at exact matches).
Relative to the second research question, number and percentage of children in each age category with each play
score were reported.
Results
Findings related to the first research question regarding the
interrater agreement for the RKPPS appear in Table 1 for
the overall play age score; in Table 2 for the four dimension
scores; and in Table 3 for the 12 category scores. Table 1
shows the magnitudes of difference in the RKPPS overall
play age scores assigned by the two raters and the cumulative
numbers and percentages according to the magnitudes of
difference in months. For the overall play age, the scores of
the two raters were within 8 months of each other almost
87% of the time.
Table 2 summarizes the cumulative number and percentage of children scored in each dimension according to
the magnitudes of difference in months. For example, in the
material management dimension, the two raters scored 36
children (94.7% agreement) within a magnitude of difference of 12 months or less.
Table 1. Magnitudes of Difference in Revised Knox Preschool
Play Scale Overall Play Age Scores Assigned by the Two Scale
Administrators
Magnitudes of Difference in Months
n
%
Cumulative n
Cumulative %
0
0.1–2.0
2.1–4.0
4.1–6.0
6.1–8.0
>8
1
2.6
1
2.6
11
28.9
12
31.5
9
23.7
21
55.2
7
18.4
28
73.6
5
13.2
33
86.8
5
13.2
38
100.0
Note. N = 38.
224
Downloaded From: http://ajot.aota.org/ on 11/10/2015 Terms of Use: http://AOTA.org/terms
March/April 2008, Volume 62, Number 2
Table 2. Cumulative Magnitudes of Difference in Age Scores for Each Dimension of the Revised Knox Preschool Play Scale
Magnitudes of Difference in Months
Dimension
Space management (n = 38)
Cumulative n
Cumulative %
Material management (n = 38)
Cumulative n
Cumulative %
Pretense–symbolic (n = 36a)
Cumulative n
Cumulative %
Participation (n = 38)
Cumulative n
Cumulative %
0
0.1–2.0
2.1–4.0
4.1–6.0
6.1–8.0
8.1–12.0
>12
16
42.1
16
42.1
16
42.1
31
81.6
31
81.6
38
100.0
38
100.0
6
15.8
9
23.7
16
42.1
27
71.1
29
76.3
36
94.7
38
100.0
15
41.7
15
41.7
15
41.7
26
72.2
26
72.2
33
91.7
36
100.0
8
21.1
8
21.1
21
55.3
24
63.2
33
86.8
37
97.4
38
100.0
One rater gave one child an NA rating, and two raters gave one child an NA rating.
a
Table 3. Magnitudes of Difference in Age Scores Assigned by the
Two Scale Administrators for Each Category Score of the Revised
Knox Preschool Play Scale
Magnitudes of Difference in Age Levela
Category
Space management
Gross motor (n = 38)
Cumulative n
Cumulative %
Interest (n = 22)
Cumulative n
Cumulative %
Material management
Manipulation (n = 35)
Cumulative n
Cumulative %
Construction (n = 28)
Cumulative n
Cumulative %
Purpose (n = 22)
Cumulative n
Cumulative %
Attention (n = 38)
Cumulative n
Cumulative %
Pretense/symbolic
imitation (n = 36)
Cumulative n
Cumulative %
Dramatization (n = 36)
Cumulative n
Cumulative %
Participation
Type (n = 38)
Cumulative n
Cumulative %
Cooperation (n = 38)
Cumulative n
Cumulative %
Humor (n = 11)
Cumulative n
Cumulative %
Language (n = 38)
Cumulative n
Cumulative %
0
1
2
3
NA
22
57.9
37
97.4
38
100.0
38
100.0
0
14
63.6
22
100.0
22
100.0
22
100.0
16
17
48.6
34
97.1
35
100.0
35
100.0
3
17
60.7
25
89.3
28
100.0
28
100.0
10
11
50.0
22
100.0
22
100.0
22
100.0
16
15
39.5
35
92.1
37
97.4
38
100.0
0
17
47.2
32
88.9
36
100.0
36
100.0
2
23
63.9
35
97.2
36
100.0
36
100.0
2
21
55.3
38
100.0
38
100.0
38
100.0
0
22
57.9
38
100.0
38
100.0
38
100.0
0
9
81.8
9
81.8
10
90.9
11
100.0
27
25
65.8
38
100.0
38
100.0
38
100.0
0
Table 4. Children in Each Category With Each Play Age Score
0 = same age level; 1 = within 1 age level difference; 2 = within 2 age levels
difference; 3 = within 3 age levels difference.
a
Table 3 relates to the category scores within each separate dimension. It shows the number of times the two raters
scored participants in the same age level, within one age level
difference, within two age levels difference, or within three
age levels difference. The two raters agreed within one age
level difference between 81.8% and 100% of the time on all
12 categories in the RKPPS.
Table 4 summarizes the findings for the second research
question regarding how children who are developing typically score on the RKPPS and the relationship of these scores
to their chronological ages. This question examines one
aspect of construct validity, defined by Anastasi and Urbina
(1997, p. 126) as “the extent to which the test may be said
to measure a theoretical construct or trait” such as play.
According to these authors, “Since abilities are expected to
increase with age during childhood, it is argued that the test
scores should likewise show such an increase, if the test is
valid” (Anastasi & Urbina, 1997, p. 127). The findings
reported in Table 4 support the construct validity of the
RKPPS because the vast majority of the overall play ages
assigned by the two raters matched the children’s chronological ages. However, the match was stronger for the older age
groups than for those in the 36- to 47-month age range. For
that group, 6 of the 14 children earned play age scores that
were higher than their chronological ages.
Average Play Age Score (Months)
Actual Age
(Months)
36–47 (n = 14)
48–59 (n = 17)
60–72 (n = 7)
The American Journal of Occupational Therapy
Downloaded From: http://ajot.aota.org/ on 11/10/2015 Terms of Use: http://AOTA.org/terms
36–47
48–59
60–72
n
%
n
%
n
%
8
57.1
5
16
35.7
94.1
1
1
7
7.1
5.9
100.0
225
Discussion
This study focusing on the RKPPS resulted in two major
findings. First, two independently trained raters can generally score within 8 months of each other on the overall play
age, within 12 months on the 4 dimension scores, and within
one age level on the 12 category scores. Second, the construct
validity of the RKPPS was supported because there was a
general match between the children’s chronological ages and
their overall play ages. Because play is a developmental construct, a valid measure of it should produce scores that reflect
a developmental progression (Dunn, 1989).
Experiences associated with the data collection, combined with the research findings, resulted in three recommendations for improving the RKPPS. The first is to provide
more detail regarding interpreting play behaviors and scoring. Although the interrater agreement was acceptable, each
rater reported challenges in scoring whereby she often
debated between two scores for a single item. Therefore, it
is likely that interrater agreement could be improved by
providing more information with the measure. Specifically,
it would be helpful to have thorough descriptions of specific
behaviors and examples of play behaviors with their appropriate scores. These descriptions could provide a framework
for interpreting specific play behaviors and measuring these
behaviors consistently. Also, a scoring module could be developed that therapists could complete when learning to use the
RKPPS. It could include a practice video of a child playing
and a scoring key that the therapist could use to check his or
her scores against after completing the RKPPS.
A second recommendation is that the guidelines for the
use of an occasional open-ended question to the child to
clarify a play scenario (S. Knox, personal communication,
October 11, 2005) be adopted and included in future versions of the RKPPS. An example of an open-ended question
is “What are you doing?” The therapist might ask this question when he or she sees a boy sweeping with a broom. The
child’s response might reveal that the boy is not just sweeping but pretending to clean his house for expected guests.
Knowing such information facilitates interpretation of the
behavior and accuracy in scoring.
Third, the possibility of allowing a rater to use an occasional prompt to encourage a child to engage in a type of
play the rater needs to observe should be explored. The raters
in the current study noted that some children engaged in the
same play activity for the duration of the observation.
Therefore, to encourage engagement in other types of play
so the child’s play can be assessed accurately, prompts may
be useful. An example of a prompt might be to place blocks
in front of a child and ask, “What can you do with these?”
It would be necessary to outline a protocol for the use of such
prompts, ensuring that they are compatible with a natural
play context, as defined by the measure.
Clinical Implications
This study provides suggestions regarding clinical use of the
RKPPS. First, because the magnitudes of difference between
the two raters’ scores for the overall play age were within 8
months of each other the vast majority of the time (86.8%
agreement; Table 2), the findings suggest that if a child
receives a score up to 8 months below his or her chronological
age, there is potentially no cause for concern. This discrepancy between the chronological age and the overall play age
score could be caused by individual differences in scoring and
is not necessarily an indication that the child has play deficits.
However, if a child’s score falls more than 8 months below
his or her chronological age and this finding is congruent
with other assessment results, further evaluation may be warranted. Second, if an occupational therapist were to use the
RKPPS, he or she can have some confidence in the developmental progression of the scale in light of the general match
between the overall play ages and the chronological ages of
the children. However, therapists should note that 43% of
the 36- to 47-month-olds had average play age scores above
their play age category, suggesting that this scale may provide
an overestimate of these children’s play ages. Last, therapists
should be cautious about generalizing these results to children
from other cultures and from other socioeconomic groups.
Strengths and Limitations
An important strength of this study’s design is that the two
raters trained separately and did not communicate with
regard to the scale during data collection. This approach
supports the clinical utility of the RKPPS because two raters
from different settings should be able to obtain similar results
when evaluating the same child if they each study the instrument thoroughly, practice giving the scale, and check interrater agreement with a colleague.
There were two primary limitations of this study. The
first was that the sample was homogeneous, with the children
being typically developing and tending to come from middle
to upper socioeconomic status urban and suburban areas. The
second was that the raters for the construct validation portion
of the study were not unaware of the children’s approximate
ages while they were collecting data. This could have created
a possible bias in scoring. Controlling for this factor is difficult, if not impossible, because children are assigned to specific classes on the basis of age, and their physical size and
other behaviors provide cues regarding their ages. As a partial
control, at the time of data collection the two raters were
unaware of the fact that their scores would be reported in
relationship to the children’s chronological ages.
226
Downloaded From: http://ajot.aota.org/ on 11/10/2015 Terms of Use: http://AOTA.org/terms
March/April 2008, Volume 62, Number 2
Directions for Future Research
Continued examination of the psychometric properties of
the RKPPS is recommended. As changes are made in the
information provided for those administering the RKPPS,
new studies should be conducted. Specifically, interrater and
test–retest agreement should be examined for children who
are typically developing and for those with developmental
delays. Children in both groups should represent varying
geographic regions, cultures, and socioeconomic groups. In
addition, further validation studies should be completed, and
normative data should be collected.
Conclusion
Play is a challenging and important construct for occupational therapists to measure. The creation of an accurate
picture of a child’s developmental play level and participation in play activities is integral to occupational therapy
evaluation and intervention planning. The RKPPS shows
promise for meeting this important need in that the current
research suggests that two raters can score this measure with
some consistency, and the scores on this measure progress
developmentally. It is only through continued refinement
and development of play assessments that the need for accurate and valid measurement of this key area of occupation
will be met. s
Acknowledgments
The authors thank the teachers, parents, and students who
made this project possible.
References
American Occupational Therapy Association. (2002). Occupational
therapy practice framework: Domain and process. American
Journal of Occupational Therapy, 56, 609–639.
Anastasi, A., & Urbina, S. (1997). Psychological testing (7th ed.).
Upper Saddle River, NJ: Prentice Hall.
Bergen, D. (1988). Stages of play development. In D. Bergen
(Ed.), Play as a medium for learning and development: A handbook of theory and practice (pp. 49–66). Portsmouth, NH:
Heinemann Educational.
Bledsoe, N., & Shepherd, J. (1982). A study of reliability and validity of a Preschool Play Scale. American Journal of Occupational
Therapy, 36, 783–788.
Bundy, A. (1989). A comparison of the play skills of normal boys
with sensory integrative dysfunction. Occupational Therapy
Journal of Research, 9(2), 84–100.
Bundy, A. C. (1991). Play theory and sensory integration. In
A. G. Fisher, E. A. Murray, & A. C. Bundy (Eds.), Sensory­integration theory and practice. Philadelphia: F. A. Davis.
Couch, K. J. (1996). The use of the Preschool Play Scale in
published research. Physical and Occupational Therapy in
Pediatrics, 16, 77–84.
Couch, K., Deitz, J., & Kanny, E. (1998). The role of play in pediatric occupational therapy. American Journal of Occupational
Therapy, 52, 111–117.
de Renne-Stephan, C. (1980). Imitation: A mechanism of play
behavior. American Journal of Occupational Therapy, 34,
95–102.
Dunn, W. (1989). Validity. In L. J. Miller (Ed.), Developing normreferenced standardized tests (pp. 149–168). Binghamton, NY:
Haworth Press.
Harrison, H., & Kielhofner, G. (1986). Examining reliability
and validity of the Preschool Play Scale with handicapped
children. American Journal of Occupational Therapy, 40,
167–173.
Hurff, J. (1974). A play skills inventory. In M. Reily (Ed.), Play as
exploratory learning (pp. 267–284). Beverly Hills, CA: Sage.
Knox, S. (1974). A Play Scale. In M. Reilly (Ed.), Play as exploratory learning (pp. 247–266). Beverly Hills, CA: Sage.
Knox, S. (1997). Development and current use of the Knox
Preschool Play Scale. In L. D. Parham & L. S. Fazio (Eds.),
Play in occupational therapy for children (pp. 35–51). St. Louis,
MO: Mosby/Year Book.
Knox, S. (2005). Play. In J. Case-Smith (Ed.), Occupational therapy
for children (5th ed., pp. 571–586). St. Louis, MO: Mosby.
Lifter, K. (1996). Assessing play skills. In M. McLean, D. B.
Bailey, & M. Wolery (Eds.), Assessing infants and preschoolers with special needs (pp. 435–461). Englewood Cliffs, NJ:
Prentice Hall.
Lunzer, E. (1959). Intellectual development in the play of young
children. Educational Review, 11, 205–217.
Michelman, S. (1971). The importance of creative play. American
Journal of Occupational Therapy, 25, 285–290.
Parten, M. B. (1932). Social participation among pre-school children.
Journal of Abnormal and Social Psychology, 28, 1243–1269.
Restall, G., & Magill-Evans, J. (1994). Play and preschool children
with autism. American Journal of Occupational Therapy, 48,
113–120.
Richmond, J. (1960). Behavior, occupation, and treatment of children. American Journal of Occupational Therapy, 14, 183–186.
Robinson, A. (1977). Play: The arena for acquisition of rules
for component behavior. American Journal of Occupational
Therapy, 31, 248–253.
Rubin, K., Fein, G., & Vandenberg, B. (1983). Play. In P. Mussin
(Ed.), Handbook of child psychology (pp. 694–774). New York:
Wiley.
Schaaf, R. C., & Mulrooney, L. L. (1989). Occupational therapy
in early intervention: A family-centered approach. American
Journal of Occupational Therapy, 43, 745–754.
Smith, P. K., Takhvar, M., Gore, N., & Vollstedt, R. (1985).
Play in young children: Problems of definition, categorization and measurement. Early Child Development and Care,
19, 24–41.
Takata, N. (1969). The play history. American Journal of Occupa­
tional Therapy, 23, 314–318.
Tinsley, H., & Weiss, D. (1975). Interrater reliability and agree­
ment of subjective judgments. Journal of Counseling Psychol­
ogy, 22, 358–376.
The American Journal of Occupational Therapy
Downloaded From: http://ajot.aota.org/ on 11/10/2015 Terms of Use: http://AOTA.org/terms
227