Effects of Testing Context on Ball Skill Performance in 5-Year-Old Children With and Without Developmental Delay.

Key Words: Ball-handling skills, Developmental delay, Peabody Developmental Motor Scales.

One of the most important changes in pediatric physical therapy has been a shift from use of neuromaturational and reflex hierarchical frameworks for evaluation of children to the measurement of disablement related to functional activities[1-3] This shift appears to have been driven by contemporary theories of motor development and motor control, which support motor learning and systems approaches to evaluation and intervention.[4] A related change is increased emphasis on evaluation and intervention in natural environments.[5] Therapists no longer assume that children's motor behaviors in an isolated therapy setting predict their behaviors in real-life environments or that performance in a therapeutic setting transfers to tasks that children need to accomplish in real-life situations.[6]

Physical therapy given in the public schools has changed over the past 20 years as Congress, the courts, an d public policy have modified and more fully defined the Individuals With Disabilities Education Act (IDEA).[7] The education of students with disabilities, for example, is moving from isolated special schools and classrooms to inclusion of children with disabilities in their neighborhood schools with their peers.[8] As children with disabilities have moved to more-inclusive settings, physical therapists have had to place greater emphasis on evaluation of the functional skills that children with disabilities need to participate in general education environments.[8,9]

The IDEA requires a multidisciplinary evaluation to determine the nature and extent of special education and related services needed by a child with a disability.[7] Because the law requires that physical therapy services be provided to assist children with disabilities to benefit from their educational programs,[7] physical therapists must evaluate skills important to children's functioning at school. Physical therapy evaluations should include observation of functional abilities in the natural environment (such as classrooms, cafeteria, bathrooms, hallways, physical education facilities, and the playground), evaluation of impairments as they affect function, and norm-referenced testing, when appropriate.[8]

One type of skill that is important for both social interaction on the playground and performance in physical education is ball-handling skills. These skills typically include throwing, catching, and kicking. Many physical educators consider ball-handling skills to be the most important motor skills taught in physical education because they are the cornerstones of many games.[10-12] Unlike motor skills that develop naturally with general movement experiences, even children without disabilities need to be taught and to practice ball-handling skills to become proficient.[13] For this reason, physical therapists may need to consider whether ball-handling skills are important for a child's educational program and, if so, to evaluate the child's proficiency.

A norm-referenced developmental test that measures ball-handling skills and is widely used by physical therapists is the Peabody Developmental Motor Scales (PDMS).[14] The PDMS is both criterion- and norm-referenced and measures gross and fine motor development from birth through 83 months of age. The Gross Motor Scale consists of 170 items divided into 17 age levels, with 10 items at each age level. Items are grouped into reflex, balance, nonlocomotor, locomotor, and receipt and propulsion skill categories. The receipt and propulsion category measures ball-handling skills.

The developers of the PDMS recommend its use with children with disabilities, and the PDMS has an adequate number of test items at each age level.[14] Through the years, the reliability and validity of measurements obtained with these scales have been under continued investigation.[14-17] Generally, reliability has been shown to be acceptable.[14]

Although the PDMS is widely used and can be used to identify children with developmental delays,[18] one criticism is that norm-referenced testing requires a standardized format of one-to-one testing by an examiner, so the results may not predict a child's performance when playing with other children.[19] Haley et al[20] demonstrated the importance of task context and environment from normative data for the Pediatric Evaluation of Disability Inventory.[21] They found a difference in the scores of children with and without disabilities when the environment or the demands of the task, such as chair size or walking surface, were varied. The authors concluded that context-specific functional assessments need to be developed. Other researchers have found similar effects of context on performance of adults.[22,23] People with hemiplegia[22] and patients with rheumatoid arthritis,[23] for example, performed more activities in the rehabilitation unit or hospital than they performed in their home environments.

To compare children's ball-handling skills during one-to-one testing conditions and in the natural environment with peers, their ball-handling skills would need to be measured under both conditions using the same measurement procedures. We contend that one-to-one testing is relatively easy to do. Using the same measurement procedures in some natural environments, such as during spontaneous game playing, however, is more difficult. Brown and Snell[5] described 2 types of data gathered in natural environments: data obtained by direct testing and data obtained by observation under natural conditions. In direct testing, a child is presented with an opportunity to respond to a stimulus at a time that the child needs to use the skills being tested.[5] In observational data collection, behaviors are observed when they naturally occur, with no constraints being placed on the activity sequence or the location of the activities.[5] A disadvantage of this type of evaluation is that the desired behavior may not occur at the time of the observation.[5] An advantage is that even when testing materials and environments are simulated to be similar to the true tasks or environment, the child's true performance may be different in the natural environment.[5]

The purpose of our study was to compare children's ball-handling skills under 2 conditions: one-to-one standardized testing and direct testing during a game with peers. The research questions were: (1) Is the performance of 5-year-old children with developmental delay on the PDMS receipt and propulsion items different during one-to-one testing than in a group setting with peers? (2) Is the performance of 5-year-old children without developmental delay on the PDMS receipt and propulsion items different during one-to-one testing than in a group setting with peers? and (3) If the performance of children within each group is different in the 2 test conditions, is the magnitude of the difference in the 2 scores the same between groups?

Method

Subjects

Subjects were 5-year-old children in Lawton (Okla) Public Schools' special education program for children with developmental delay and the general kindergarten program. For the purposes of this study, "5 years" was defined as 60 through 71 months. Five-year-old children were selected for 3 reasons: (1) Children without disabilities do not enter school until they are 5 years of age, making it difficult to recruit younger children without disabilities; (2) other studies have found a difference in ball-handling performance of boys and girls at age 6 years[13,24]; and (3) the PDMS was designed for children up to 83 months of age, meaning that children near this age probably would be less influenced by task context than younger children because the tasks would be easier for them.

The inclusion criteria for subjects with developmental delay were the criteria used by the Oklahoma State Department of Education.[25] The subjects demonstrated delay in one or more of the following developmental domains: adaptive, cognitive, communication, physical (including gross and fine motor), and social or emotional. Delays were documented using a combination of 2 of the following tests: Battelle Developmental Inventory,[26] Birth to Three,[27] Early Learning Accomplishment Profile,[28] and PDMS.[14] "Delay" was defined by (1) scores that were at least 2 standard deviations below the mean or indicated at least 50% delay in one area of development or (2) scores that were at least 1.5 standard deviations below the mean or that indicated at least a 25% delay in 2 areas of development. Children who were not independently ambulatory or who performed more than 3 standard deviations below the mean in any area were not included.

The inclusion criteria for subjects without developmental delay were that they were 5 years of age, attended general kindergarten classes, and did not receive special education, physical therapy, or occupational therapy. They also could not have a history of receiving these services or early intervention services and could not be in the process of evaluation for special education services.

Twenty children with developmental delay (10 boys and 10 girls) and 20 children without developmental delay (10 boys and 10 girls) met the inclusion criteria and participated in the study. Table 1 shows the percentage of delay in each developmental domain for the children with developmental delay, as documented in their school records. Informed written consent was obtained from the parents of all children participating in the study.

Table 1.

Sex and Percentage of Delay of Children With Developmental Delay

                    Developmental Domain
Student No.   Sex   Cognitive   Communication   Adaptive

     1         F    44          46               0
     2         M    50          50              33
     3         F    45          72              47
     4         M    49          48              60
     5         M    50          25              35
     6         F    52          42              50
     7         F    33          31              50
     8         M     0          58               0
     9         F    44          45              48
    10         M    86          50              50
    11         F    43          43               0
    12         M    25          50              64
    13         M    43          42              53
    14         M    57          56              41
    15         M     0          50               0
    16         M    52          50              20
    17         F    25          33              12
    18         F     0           3              27
    19         F    29          60              26
    20         F    34           9              69

Student No.   Social/Emotional   Motor

     1         0                  7
     2        54                 28
     3        41                 21
     4        43                 60
     5        25                 50
     6        50                 45
     7        25                 41
     8         0                  0
     9        33                 31
    10        30                 30
    11        50                  0
    12        20                 25
    13        32                 38
    14        57                  0
    15         0                  0
    16         4                 36
    17         3                 39
    18        15                 38
    19        41                 50
    20        19                 51

Procedure

Each student participated in 2 sessions of ball skills testing using the receipt and propulsion items of the PDMS.[14] One session took place in a school gym during a game with peers, and the other session took place in the school physical therapy room with a therapist. The independent variable was the task in which skills were measured. The dependent variable was performance of ball skills as measured by PDMS scores.

The group testing in the gym consisted of 5 short games that included all of the PDMS receipt and propulsion items: (1) throw/catch a playground ball (items 68, 69, 78, 104, 116, 135, and 145), (2) throw/catch a tennis ball (items 83, 95, 122, 124, and 153), (3) bouncing game (items 112, 125, and 164), (4) throwing to a target game (items 123, 126, and 150), and (5) kicking the ball (items 77, 85, 103, 158, 162, and 163). Table 2 lists the specific skills that correspond to the item numbers. The isolated testing with the therapist consisted of the same items, administered in the same order. To aid in testing and scoring, colored tape was used to mark distances and 15-degree deviations on the floors prior to all sessions. Two score sheets for the tested items were developed. One score sheet was used to record all of the children's scores during the testing of each group, and the other score sheet was used to record each child's scores during the isolated testing.

Table 2.

Mean Peabody Developmental Motor Scales(14) Raw Scores by Group(a)

                                        Isolated Testing Environment

                                         FWODD   MWODD   FWDD   MWDD

12-14 mo
  68) Rolling ball                       20      20      20     20 H
  69) Flinging ball                      20      20      20     20 H
15-17 mo
  77) Kick ball, step on it              20      20      18     20
  78) Throw ball 3 ft,(b)
      <45 [degrees] deviation            20      20      20     20 H
18-23 mo
  83) Throw tennis ball 3 ft,
      <45 [degrees] deviation            20      20      20     20 H
  85) Kick ball 3 ft, <20 [degrees]
      deviation                          19      20      16     20
24-29 mo
  94) Throw ball 5 ft, <20 [degrees]
      deviation                          19      20      14     19
  95) Throw tennis ball 7 ft,
      <20 [degrees] deviation            18      20      15     17
30-35 mo
  103) Kick ball 6 ft, <15 [degrees]
       deviation                         17      18      12     19
  104) Catch ball from 5 ft, with
       arms straight, 2/3 times          20      20      15     20
36-41 mo
  112) Bounce throw tennis ball to
       the wall, from 5 ft                9      18      10     12
  116) Catch ball from 5 ft, with
       elbows bent, 1/2 times            20      20      16     19
42-47 mo
  122) Throw tennis ball 10 ft, with
       shoulder rotation, <20
       [degrees]                         10      12       6     12
  123) Throw ball to wall target,
       from 5 ft, 2/3 times              18      16      12     16
  124) Catch tennis ball from 5 ft,
       with elbows bent, 2/3 times       12      18       3     12
  125) Bounce throw ball to wall from
       5 ft                               8      12      11     12
  126) Overhand throw ball to wall
       target, from 5 ft, 2/3 times      17      16       8     13
48-53 mo
  135) Throw ball 10 ft, <20 [degrees]
       deviation                         12      15       3     11
54-59 mo
  145) Catch ball, from 6 ft, with
       elbows at side, 2/3 times         13      16       4      9
  150) Throw tennis ball, overhand,
       to wall target, from 12 ft,
       2/3 times                          2       8       1      7 L
60-71 mo
  153) Bounce tennis ball on floor 2
       times and make 2-handed catch     12      17       4      7 L
  158) Kick ball from stationary
       point into air 12 ft               7      14       4      7
72-83 mo
  162) Kick rolling ball 8 ft after
       running 4 ft                       8      18       6      6
  163) Drop-kick ball 5 ft, 1/2 times     5      17       1      9 L
  164) Throw ball to wall, from 5 ft,
       and catch after it bounces 1
       time                              11      20      10     11

                                         Group Testing Environment

                                         FWODD   MWODD   FWDD   MWDD

12-14 mo
  68) Rolling ball                       20      20      20     20 H
  69) Flinging ball                      20      20      20     20 H
15-17 mo
  77) Kick ball, step on it              20      20      18     20 H
  78) Throw ball 3 ft,(b)
      <45 [degrees] deviation            16      20      19     16
18-23 mo
  83) Throw tennis ball 3 ft,
      <45 [degrees] deviation            20      20      19     19
  85) Kick ball 3 ft, <20 [degrees]
      deviation                          19      20      16     19
24-29 mo
  94) Throw ball 5 ft, <20 [degrees]
      deviation                          14      15      15     11 D
  95) Throw tennis ball 7 ft,
      <20 [degrees] deviation            19      20       9     14
30-35 mo
  103) Kick ball 6 ft, <15 [degrees]
       deviation                         17      19      16     17
  104) Catch ball from 5 ft, with
       arms straight, 2/3 times          18      19       5     16
36-41 mo
  112) Bounce throw tennis ball to
       the wall, from 5 ft                8      12      10     11
  116) Catch ball from 5 ft, with
       elbows bent, 1/2 times            16      20       4     17 D
42-47 mo
  122) Throw tennis ball 10 ft,
       with shoulder rotation,
       <20 [degrees]                     12       8       2      8
  123) Throw ball to wall target,
       from 5 ft, 2/3 times              16      19      11     17
  124) Catch tennis ball from 5 ft,
       with elbows bent, 2/3 times       10      16       0      6
  125) Bounce throw ball to wall
       from 5 ft                         12      17       7     13
  126) Overhand throw ball to wall
       target, from 5 ft, 2/3 times      17      20       7     17
48-53 mo
  135) Throw ball 10 ft,
       <20 [degrees] deviation            6      10       3      6 L
54-59 mo
  145) Catch ball, from 6 ft, with
       elbows at side, 2/3 times          6      11       0      7
                                                                L, D
  150) Throw tennis ball, overhand,
       to wall target, from 12 ft,
       2/3 times                          4      10       3      4 L
60-71 mo
  153) Bounce tennis ball on floor
       2 times and make 2-handed
       catch                             10      18       3     12
  158) Kick ball from stationary
       point into air 12 ft               8      14       3      7
72-83 mo
  162) Kick rolling ball 8 ft after
       running 4 ft                      10      19       2      7
  163) Drop-kick ball 5 ft, 1/2
       times                              8      14       1      6
  164) Throw ball to wall, from 5
       ft, and catch after it
       bounces 1 time                    11      18       6     12

(a) Minimum score=0, maximum score=20, 10 subjects per group. FWODD=females without developmental delay, MWODD=males without developmental delay, FWDD=females with developmental delay, MWDD=males with developmental delay. H=highest score test item under each condition, L=lowest score test item under each condition, D=largest difference in score for a test item under each condition.

(b) 1 ft=0.3048 m.

The tester was a physical therapist with 3 years of pediatric experience who did not know the research questions of the study. The researcher (AKD) trained the tester during 3 sessions, each approximately 30 minutes long. During the first training session, the score sheet and setup for the group games were discussed. During the second training session, scoring criteria, standard verbal instructions, and standard verbal praise were covered. Plans for interrater and interrater reliability testing with students in a pilot study also were discussed. During the third session, the tester and the researcher practiced setting up and administering the group session.

A pilot test was conducted to examine the testing procedures and to identify any needed changes. Three students with developmental delay and 3 students without developmental delay participated in the pilot test. Based on the pilot test, the following procedural changes were made prior to data collection: (1) The group size was changed from 6 subjects to 4 subjects due to difficulty scoring 6 children at one time, and (2) the standard verbal instruction for kicking the ball up was changed from "kick the ball" to "kick the ball up" because most of the children in both groups did not attempt to kick the ball up.

During data collection, all children were tested with their classes. Children in 4 morning and 4 afternoon classes for children with developmental delay and children in 2 morning and 1 afternoon kindergarten classes participated in the study. To control for possible order effects of testing, half of children with developmental delay and half of the children in kindergarten classes were chosen at random to be tested in the isolated setting first. The children in the other classes were tested in the group setting first. A group of children in kindergarten was tested first to control for the possibility of inflated scoring of the children with developmental delay if they were tested first, as discussed by Field.[29]

Group testing took place in the school gym and lasted approximately 40 minutes. During the group testing, the children wore front and back name tags and the researcher assisted the tester in keeping the children in their places and dealing with any problem behaviors that occurred. When the groups of children with developmental delay were tested, the group included one child from the kindergarten group who had been tested previously and was known to have age-appropriate throwing skills. This student threw the ball while all of the other children's catching skills were scored to ensure that the children's ability to catch was not influenced by a peer's poor throwing ability. Isolated testing took place in the school physical therapy room, using the same order of items and standardized instructions and praise as testing in the group setting. The researcher videotaped all sessions.

Reliability

To determine interrater reliability, the researcher and the tester simultaneously scored children's performance during group games and isolated testing. Although they stood near each other so they could observe children from the same angles, they could not see each other's score sheets. Intrarater reliability was determined by the researcher and the tester rescoring children's performance from the videotapes. All of the trials during the pilot test were examined for reliability to identify the need for changes in procedures and scoring that were made prior to the actual data collection. During data collection, one group session and one isolated session per week were scored by both the researcher and the tester and were rescored from the videotapes to determine reliability and identify any problems with observer drift.

Each set of observations (scores) was compared for interrater or intrarater agreement or disagreement, and the kappa statistic was calculated by group and isolated test setting. As shown in Table 3, the kappa statistics for interrater agreement were between .89 and .94 and the kappa statistics for intrarater agreement were between .84 and .91. Better interrater agreement than intrarater agreement is unusual[30] and probably resulted from some difficulty scoring students from the videotapes; camera angles did not always give a clear view of students, and lights sometimes caused glare from the floor.

Table 3.

Interrater and Intrarater Agreement by Children's Developmental Status and Test Setting

                                  No. of    No. in
                                  Scores   Agreement   Kappa

Interrater reliability
  Isolated testing environment
    With developmental delay        75        72        .94
    Without developmental delay     99        93        .91
  Group testing Environment
    With developmental delay       275        255       .89
    Without developmental delay    372        354       .92
Intrarater reliability
  Isolated testing environment
    With developmental delay        75        67        .84
    Without developmental delay    100        93        .90
  Group testing environment
    With developmental delay       269        242       .84
    Without developmental delay    372        349       .91

Data Analysis

The 25 receipt and propulsion items were scored according to instructions in the PDMS test manual.[14] Items are scored 0, 1, or 2, so the minimum score possible was 0 and the maximum score possible was 50.

A preliminary analysis was performed using a 3-way analysis of variance to determine whether differences existed in PDMS scores due to order of testing, sex, and developmental status. Then, a repeated-measures analysis of variance using a 3-factor design with the variables of sex, developmental status, and test setting was performed using SAS software.[31](*) An alpha level of .05 was used to determine statistical significance.

Results

The preliminary analyses showed that whether students were tested first under group or isolated conditions did not affect their scores; therefore, data from the 2 orders of testing were combined for further analyses. As shown in Table 4, preliminary analyses showed a difference in PDMS scores by developmental status and sex. Children without developmental delay scored higher than children with developmental delay, and boys scored higher than girls. Regardless of sex, the mean score of children without developmental delay ([bar]X=38.75, SD=5.93) was higher than the mean score of children with delay ([bar]X=29.25, SD=8.44; P=.0001). Regardless of developmental status, boys had a higher mean score ([bar]X=38.45, SD=6.96) than girls ([bar]X=29.55, SD=7.97; P=.0001).

Table 4.

Scores of Boys and Girls With Developmental Delay (WDD) and Without Developmental Deity (WODD) on the Peabody Developmental Motor Scales(14) Receipt and Propulsion Items in the Isolated Test Setting and in a Group Test Setting With Peers

                                          Score

                        Developmental   X       SD
                        Status

Isolated test setting
  Girls                 WDD             26.90   7.29
  Boys                  WDD             35.80   4.81
  Girls                 WODD            35.70   4.81
  Boys                  WODD            43.80   2.70
Group test setting
  Girls                 WDD             21.90   6.52
  Boys                  WDD             32.40   7.26
  Girls                 WODD            33.70   4.76
  Boys                  WODD            41.80   4.69

                              Score

                        Maximum   Minimum

Isolated test setting
  Girls                 37        16
  Boys                  44        23
  Girls                 42        26
  Boys                  48        39
Group test setting
  Girls                 38        15
  Boys                  42        20
  Girls                 39        25
  Boys                  48        34

Preliminary results did not show an interaction between sex, developmental status, and test setting, so scores were averaged over the variables to answer the following research questions:

1. Is the performance of 5-year-old children with developmental delay on the PDMS receipt and propulsion items different in the isolated test setting than in a group setting with peers? The mean score of children with developmental delay was higher in the isolated setting ([bar]X=31.35, SD=7.92) than in the group setting with peers (X=27.15, SD=8.60), and this difference was statistically significant (F=8.97; df=1,19; P=.007).

2. Is the performance of 5-year-old children without developmental delay on the PDMS receipt and propulsion items different in an isolated test setting than in a group setting with peers? The mean score of children without developmental delay in the isolated setting was 39.75 (SD=5.63). The mean score of children without developmental delay in the group setting with peers was 37.75 (SD=6.20). There was no difference in performance between test settings (F=3.38; df=1,19; P=.08).

The effect size of a 3.1-point difference between the mean scores of children in the isolated test setting and the mean scores of children in the group test setting across developmental status was large, with a 76% probability of demonstrating a difference with 20 students in each group. Because the performance of the children without developmental delay did not differ in the 2 test conditions, we could not answer the third question about whether the difference in performance in the 2 conditions was greater for one group than for the other group.

Table 2 shows that no patterns appeared to exist in the types of ball skill items on which the children scored lowest or on which they performed differently under the 2 test conditions. They scored consistently higher on the test items at the developmentally younger end of the scale.

Discussion

The finding that the children with developmental delay performed better in the isolated test setting than they did in a structured game with peers supports the observations of Haley and colleagues[19,20] that children with disabilities perform differently in different environments. In his explanation of motor learning, Schmidt[32] emphasized that if a skill is to be transferred to various environments, the environments in which the skill is practiced must be similar, requiring the same processes. It is likely that performance in the isolated testing environment with a therapist did not require the same processes as performing what appeared to be the same skills in the gym with peers.

One difference between the isolated testing situation and the group game with peers was that, as in most standardized testing situations, the room in which the testing took place was relatively quiet and free from distractions. This was not the case when children were tested in the gym with their peers. One observed difference between the groups was that the children without developmental delay made comments suggesting that they were competing with each other during group games. Children with developmental delay did not make such comments, suggesting that if competition causes children to try harder to do well, then the children with developmental delay may not have benefited from the competition to the same extent as children without developmental delay.

Children in both groups appeared to imitate the behavior of other children. If the first student to throw the ball, for example, threw it poorly so that it bounced before reaching the catcher, the other children tended to throw the ball with the same pattern. Children also might have been influenced by the good skills of their peers during the group games. Research has shown that observation of peers can promote improvement in an imitator's skills.[33] The children without developmental delay had better peer models than did the students with developmental delay. A limitation of this study was that for the group testing, the children with developmental delay were with other children with developmental delay. If they had been included with the children without developmental delay, they might have had better models and have performed better.

In the isolated test setting, the children did not have peer models, either good or poor. All children had the advantage of receiving the ball from the tester, who ensured that the child was ready and carefully threw the ball to optimize each child's chances of catching or kicking it successfully. This was not always the case in the gym environment with peers.

The finding that boys performed better than girls was not expected. Previous reports[13,24] suggest that a sex difference does not occur before 6 years of age. In our study, boys without developmental delay performed best, followed, in order, by girls without developmental delay, boys with developmental delay, and girls with developmental delay. The boys with developmental delay had the same scores as girls without developmental delay. This may be a useful finding when planning group intervention. To avoid frustration and promote inclusion of 5-year-old boys with developmental delay in group games, it may be helpful to place them in a group that is not predominantly boys without developmental delay.

Because norm-referenced tests do not measure motor skills in natural environments, many authors advocate observation in natural environments to obtain information about children's disabilities and how they may be affected by societal limitations.[5,19,20] The findings of our study may support the argument that observation in natural environments is important because, as Haley and colleagues stated,[19] motor behaviors are more than just competencies unfolding over time; motor behaviors are responses to meet desired goals within specific physical and social contexts. For this reason, future research that compares direct testing with observation of children in naturally occurring situations for performing ball skills may reveal even greater differences in performance between the isolated test setting and observation in the natural environment than were found between the isolated test setting and a structured group test setting.

Two limitations of our study were the inclusion of only 5-year-old children, which prevents generalization to children of other ages, and the use of a relatively small sample. Another limitation was occurrence of some distracting behaviors during the group testing situations, but these behaviors are likely to occur in most natural environments. Finally, the effects of teaching ball-handling skill,; in isolated and group environments on task generalization should be investigated.

Conclusion

Physical therapists who evaluate students with developmental delay should carefully consider the methods and the environment in which they test children. To determine whether a child has a developmental delay, standardized administration of norm-referenced tests is appropriate. To determine how students function in their everyday environments, observation and talking with teachers and parents can help to obtain a clear picture of students' capabilities and needs.

Acknowledgments

We thank the teachers and administrators in Lawton Public School,; for their support during this project. We also thank the children and their parents for their willingness to participate.

(*) SAS Institute Inc, PO Box 8000, Cary, NC 27511.

References

[1.] McEwen IR, Shelden ML. Pediatric therapy in the 1990s: the demise of the educational versus medical dichotomy. Physical and Occupational Therapy in Pediatrics. 1995;15:33-45.

[2] Heriza CB. Pediatric physical therapy: reflections of the past and visions of the future. Pediatric Physical Therapy. 1994;6:105-106.

[3] Heriza CB, Sweeney JK. Pediatric physical therapy, part I: practice scope, scientific basis, and theoretical foundation. Infants and Young Children. 1994;7:23-32.

[4] Stuberg W, Harbourne R. Theoretical practice in pediatric physical therapy: past, present, and future considerations. Pediatric Physical Therapy. 1994;6:119-125.

[5] Brown F, Snell ME. Meaningful assessment. In: Snell ME, ed. Instruction of Students With Severe Disabilities. New York, NY: Macmillan Publishing Co; 1993:61-98.

[6] Harris SR, McEwen IR. Assessing motor skills. In: McLean ML, Bailey D, Wolery M, eds. Assessing Infants and Toddlers With Special Needs. Columbus, Ohio: Merrill; 1996:305-333.

[7] Individuals With Disabilities Education Act Amendments of 1991, 105 Stat 587-608.

[8] Effgen SK. The educational environment. In: Campbell SK, ed. Physical Therapy for Children. Philadelphia, Pa: WB Saunders Co; 1994: 847-872.

[9] Giangreco MF, Edelman S, Dennis R. Common professional practices that interfere with integrated delivery, of related services. Remedial and Special Education. 1991;12:16-24.

[10] Anderson G, Elliot B. The overarm throw: a movement back to the basics. Australian Council for Health, Physical Education and Recreation National Journal. 1991;132:27-30.

[11] Butterfield SA, Loovis EM. Influence of age, sex, balance, and sport participation on development of throwing by children in grades K-8. Percept Mot Skills. 1993;76:459-464.

[12] Williams N. Throwing and catching: a steady diet. Journal of Physical Education Recreation and Dance. 1992;64:14.

[13] Halverson LE, Roberton MA, Langendorfer S. Development of the overarm throw: movement and ball velocity changes by seventh grade. Res Q. 1982;48:311-318.

[14] Folio R, Fewell R. Peabody Developmental Motor Scales and Activity Cards. Chicago, Ill: The Riverside Publishing Co; 1983.

[15] Palisano RJ, Lydic JS. The Peabody Developmental Motor Scales: an analysis. Physical and Occupational Therapy in Pediatrics. 1984;4:69-75.

[16] Hinderer KA, Richardson PK, Atwater SW. Clinical implications of the Peabody Developmental Motor Scales: a constructive review. Physical and Occupational Therapy in Pediatrics. 1989;9:81-106.

[17] Schmidt LS, Westcott SL, Crowe TK. Interrater reliability of the gross motor scale of the Peabody Developmental Motor Scales with 4- and 5-year-old children. Pediatric Physical Therapy. 1993;5:169-175.

[18] Campbell SK. Measurement in developmental therapy: past, present, and future. Physical and Occupational Therapy in Pediatrics. 1989;9:1-13.

[19] Haley SM, Baryza MJ, Blanchard Y. Functional and naturalistic frameworks in assessing physical and motor disablement. In: Wilhelm IJ, ed. Clinics in Physical Therapy: Physical Therapy Assessment in Early Infancy. New York, NY: Churchill Livingstone Inc; 1994:225-256.

[20] Haley SM, Coster WJ, Binda-Sundberg K. Measuring physical disablement: the contextual challenge. Phys Ther. 1994;74:443-451.

[21] Haley SM, Coster WJ, Ludlow LH, et al. Pediatric Evaluation of Disability Inventory: Development, Standardization, and Administration Manual. Boston, Mass: New England Medical Center Hospitals Inc and PEDI Research Group; 1992.

[22] Strub N, Levine RE. Self-care: a comparison of patients' institutional and home performance. Occupational Therapy Journal of Research. 1981; 7:53-56.

[23] Haworth RJ, Hollings EM. Are hospital assessments of daily living activities valid? Int Rehabil Med. 1979;1:59-62.

[24] Wickstrom RL. Fundamental Motor Skills. Philadelphia, Pa: Lea & Febiger; 1983:101-163.

[25] Policies and Procedures for Special Education in Oklahoma. Oklahoma City, Okla: Oklahoma State Department of Education; 1993.

[26] Newborg J, Stock J, Wneck L. Battelle Developmental Inventory. Allen, Tex: DLM; 1988.

[27] Bangs T. Birth to Three. Allen, Tex: DLM; 1986.

[28] Glover M, Preminger J, Sanford A. Early Learning Accomplishment Profile for Developmentally Young Children. Chapel Hill, NC: Chapel Hill Training Outreach Project; 1988.

[29] Field T. Ecological variables and examiner biases in assessing handicapped preschool children. J Pediatr Psychol. 1981;6:155-163.

[30] Portney LG, Watkins MP. Foundations of Clinical Research: Applications to Practice. East Norwalk, Conn: Appleton & Lange; 1993.

[31] SAS/STAT User's Guide: Version 6. Vol 2. 4th ed. Cary, NC: SAS Institute Inc; 1989.

[32] Schmidt RA. Motor learning principles for physical therapy. In: Lister MJ, ed. Contemporary Management of Motor Control Problems: Proceedings of the II STEP Conference. Alexandria, Va: Foundation for Physical Therapy Inc; 1991:49-64.

[33] Larin HM. Understanding motor performance in children. In: Campbell SK, ed. Physical Therapy for Children. Philadelphia, Pa: WB Saunders Co; 1994:157-181.

AK Doty, PT, PCS, is employed by Lawton (Okla) Public Schools. This work was completed in partial fulfillment of the requirements for her Master of Science degree front the University of Oklahoma Health Sciences Center. Address all correspondence to Ms Doty at 4813 SE Trenton, Lawton, OK 73501 (USA) ([e-mail]).

IR McEwen, PhD, PT, PCS, is Presbyterian Health Foundation Presidential Professor, Department of Physical Therapy, University of Oklahoma Health Sciences Center, Oklahoma City, Okla.

D Parker, PhD, is Professor, Department of Biostatistics and Epidemiology, University of Oklahoma Health Sciences Center.

J Laskin, PT, is Assistant Professor, Department of Physical Therapy, University of Montana, Missoula, Mont. At the time the study was completed, he was Adjunct Assistant Professor, Department of Physical Therapy, University of Oklahoma Health Sciences Center.

Concept and research design were provided by Doty, McEwen, Parker, and Laskin; writing, by Doty and McEwen; data collection, by Doty, with the assistance of Jean Goodman, PT; data analysis, by Doty, McEwen, and Parker. Doty provided subjects, with the assistance of the Special Education Director of Lawton Public Schools; she provided facilities/equipment, with the assistance of the Principal of Robert E. Lee Elementary School.

This research was approved by the Institutional Review Board of the University of Oklahoma Health Sciences Center.

Support for the research was provided by a preparation of related services personnel grant (#H029F30020) from the US Department of Education, Office of Special Education and Rehabilitative Services; an Interdisciplinary Leadership in Education for Health Professionals Caring for Children With Neurodevelopmental and Related Disabilities grant (MC409411) from the Maternal and Child Health Bureau; and a research grant from the Section on Pediatrics of the American Physical Therapy Association. The article, however, does not necessarily reflect the policies of these organizations, and official endorsement should not be inferred.

This research was presented in abstract format at the American Physical Therapy Association Combined Sections Meeting; Dallas, Tex; February 12-16, 1997.

This article was submitted May 1, 1998, and was accepted May 13, 1999.3