Assistive positioning as a control parameter of social-communicative interactions between students with profound multiple disabilities and classroom staff.

Physical therapists often recommend use of assistive positioning equipment for children who have cerebral palsy and other severe neuromotor impairments.[1-5] Wheelchairs, adapted chairs, standers, sidelyers, and prone positioners are some of the devices that are commonly used. Bergen et al[1] state that there are nine benefits of good positioning that utilizes such equipment: (1) normalization of neurological influences on the body, (2) prevention of contractures and deformities, (3) prevention of pressure ulcers, (4) improvement of functional activities, (5) comfort, (6) enhancement of autonomic nervous system function, (7) decreased fatigue, (8) facilitation of normalized movement in a development sequence, and (9) promotion of maximum function with minimal pathology.

Similar benefits of assistive positioning equipment have been described by others, but little research has been conducted to support the claims that have been made.[6] Most of the studies that have been reported have examined variations in seated positions, with dependent variables focusing primarily on physical functions, such as hand use,[7] respiratory function,[8] tonic muscle activity (electromyographic activity),[9] and spinal extension.[10] Although use of several positions during the day is recommended,[1,2] only a few studies have examined the effects of positions other than the seated position or have compared the effects of different positions.[11]

The limited research generally supports the observation that assistive positioning equipment can lead to various musculoskeletal and physiological responses and can influence motor behaviors. There are also indications that children's communication and their opportunities for social interaction can be influenced by assistive positioning devices.[11-13] When positioning leads to motor, communicative, or other behavior that was not exhibited until such positioning was provided, position can be viewed as a control parameter of the behavior in dynamic systems terms.

A dynamic systems perspective is being increasingly proposed as a foundation for theoretical bases to better understand and manage problems of motor development and motor control.[14-16] With older hierarchical models, the brain is seen as the controller of behavior. Using a dynamic systems perspective, behaviors are seen as the product of self-organizing, dynamic interactions of environmental and personal factors within task-specific contexts.(15) Neither personal nor environmental elements are seen as preeminent causes of behavior, and new behaviors can only be understood as the products of both kinds of elements.[15,17,18] Behavior and behavioral change result from a dynamic interaction of the system parts, the task, and the context and are not "prewired" in the genes or under the control of the central nervous system.[19] Within a dynamic system, a control parameter is the component that causes the system to shift to a new behavior.[15,18] The control parameter can be any endogenous or exogenous element, including those as diverse as myelinization of the nervous system or the type of surface on a floor. Even though elements necessary for manifestation of a behavior will not be exhibited until the controlling component reaches critical levels. For example, Bertenthal et al[20] identified infants' crawling as a control parameter for certain aspects of visual attention and forms of communication, the components of which were shown to be in place, but not expressed, until crawling commenced. Similarly, Thelen and Fisher[21] found that body fat (causing increased leg weight) served as a control parameter of cessation of infants' automatic stepping after the first few postnatal weeks.

The potential for external control parameters to evoke behavioral change has implications for physical therapy intervention with children who have severe disabilities. Although these children's own limitations may greatly restrict the pool of potential internal control parameters available to them, environmental controls need bear no such restrictions.

The purpose of this study was to determine whether one external control parameter--assistive positioning--could serve as a control parameter of social-communicative interactions between students with profound multiple disabilities and their classroom staff (ie, teachers and their assistants). The need to focus on such functional outcomes of physical therapy is receiving increasing emphasis, as is the need to examine the effects of physical therapy on other than strictly motor behaviors.[22,23] This is especially the case when working with children in public schools, where physical therapy services must support a child's educational program.[22,23]

One of the most important functional, educational goals for children with severe disabilities, a goal that is likely to be influenced by their positioning, is development of basic social-communication skills.[2,24] Accomplishment of this goal not only enables children to establish and maintain social closeness to others and to convey wants and needs, but can provide them with a means to influence their social and physical environments.[2,24] This ability to control what happens to themselves is often said to be the key to prevention or reduction of these students' pervasive passivity or "learned helplessness," a condition that can effectively block intervention efforts and that can be extremely resistant to change.[2]

Research and reported observations suggest not only that position could serve as a control parameter of children's own communicative behaviors, but that it could serve to control others' communicative behaviors directed toward a child.[5] Such opportunities for interaction are critical for the communication development of both normally developing infants and children with severe communication impairments.[2,24,25]

This study compared the effects of three positions in which children with severe neuromotor impairments are commonly placed on their communicative interactions with classroom staff. The positions were seated in a wheelchair, sidelying, and "freestyle" on a mat without assistive equipment. The following research questions were addressed:

1. During unstructured classroom observations: (1) What is the effect of students' position on the rate of communicative initiations by both students and adults? and (2) What is the effect of position on the frequency and average duration of communication-maintaining behaviors by both adults and students and on the proportional duration of the session during which such behaviors take place?

2. During a structured opportunity for interaction, what is the effect of position on the average duration of communicative interactions between students and staff, and what is the proportional duration of the session during which interactions take place?

Method

Subjects

Ten elementary school students, 6 to 12 years of age, with profound multiple disabilities participated in the study. All students attended self-contained special education classrooms for children with severe or profound disabilities, within regular elementary school buildings. Four of the students were in the same classroom. The other 6 students were in another school district, with 3 students in each of two classrooms in the same school. The three classrooms involved in the study were selected by administrators of the participating school districts. Informed consent was obtained from each student's parent or legal guardian.

Five criteria were used to select the students from the three classrooms: (1) classification as profoundly mentally retarded or as multiply handicapped with profound retardation, according to state guidelines for classifying disabled students; (2) less than a 6-month level of gross motor development (unable to sit independently); (3) an average score within the 12- to 21-month level, or below, on the Modified Interview Form[25] of the Early Social Communication Scales[26]; (4) use of assistive positioning equipment during the school day; and (5) an attendance record that suggested at least 6 days of school attendance during a 3-week period. Ten of the 21 students in the three classrooms met the selection criteria and participated in the study. Their characteristics are listed in Table 1.

Variables

The independent variable was the students' position (ie, seated in a wheelchair, positioned in a sidelyer, or positioned on a mat on the floor). I had intended to use a stander, rather than the mat, for the third position, but none of the students were using standers at the time of the study and all spent a considerable amount of time on mats in their classrooms.

The dependent variables were the adults' and students' communicative behaviors that initiated and maintained social-communicative interactions, according to definitions based on the Videotaped Interactions Analysis.[27] Initiation of communication was operationally defined as an attempt to start an interaction with a person with whom there had been none for at least 30 seconds. Maintenance of communication was defined as behaviors that served to continue an established interaction, with less than 2-second interruptions. Adult communicative behaviors were primarily spoken, whereas the children's communication was nonsymbolic, similar to that of typical infants.[2,24] Children communicated through such means as smiling, frowning, unintelligible vocalizations, and looking at a person. Complete information about coding categories and definitions can be obtained from the author.

Equipment

The wheelchairs were all custom-adapted and belonged to the students. Tumbleform sidelyers(*1) were used in two of the classrooms, and a Rifton sidelyer[unkeyable] was used in the third classroom. When on the mats, students were placed as they usually were when positioned without assistive equipment during their school day. Eight of the 10 students were placed on their backs with a pillow under their heads and a pillow or bolster under their knees. Pillows were not used with student 7 because she was able to roll from side to side when placed on the mat. Student 6 was placed prone, the position she used in the classroom because of excessive extension when she was positioned supine. When in the prone position, she was able to raise her head and support herself on her elbows.

A Hitachi VM-3100 VHS camcorder[unkeyable] with a Hitachi VM-CG20A character generator,[unkeyable] mounted on a tripod, was used to record all of the observations. The character generator superimposed a continuous digital time image on the videotape, which was used when coding the observations.

Procedure

Students were observed twice in each of the three positions, for a total of six sessions per child. The order of the positions was counterbalanced within and across students. The time of day each child was observed varied by no more than 30 minutes for each session.

Observations took place in the students' own classrooms, while they had an opportunity to play independently with a toy. During each session, students and the people interacting with them were videotaped by the investigator. A record was also kept of the students' interactions with people outside of camera range to assist with coding of the videotapes.

Each session consisted of an initial unstructured 5-minute period, during which the classroom staff and any others (such as therapists) were asked to go about their usual activities, interacting with all of the students as they usually would. Because pilot data revealed little interaction between adults and students during any unstructured classroom activities, a finding that was consistent with other investigations,[28] a 6-minute "interaction opportunity routine" was introduced, following the unstructured observation. This structured routine was designed to give students an attentive, communicative partner and opportunities to initiate and maintain interactions, with varying levels of prompts, as necessary. The routine is summarized in Table 2. The same adult (a teacher or classroom assistant) interacted with each student during the six experimental sessions. [TABULAR DATA OMITTED]

Adult and student behaviors that initiated or maintained interactions, during both the unstructured, and structured observations, were coded from the videotapes, according to the real time of the superimposed clock. Videotapes were coded by the investigator, a graduate student, and a third adult.

[TABULAR DATA OMITTED]

Reliability of Measures

Ten percent (6) of the 60 sessions were coded by two independent observers to assess interobserver reliability. Agreement was calculated using Cohen's Kappa, a measure of proportion of agreement that controls for agreement by chance.[29] Average agreement was .99 for the adults' communication (range=.98-.98) and .82 for the students' communication (range=.63-.99). Table 3. Adults' and Students' Initiations of Communications per Minute During the Unstructured Observations by Student Group(a) and Position

                   Position
                   Wheelchair     Sidelying      Freestyle

Adult initiations Group 1 (n=5)

  [unkeyable]      0.07           0.04           0.00
  SD               0.10           0.09           0.00

Group 2 (n=5)

  [unkeyable]      0.28           0.16           0.16
  SD               0.26           0.12           0.15

Combined (N=10)

  [unkeyable]      0.18           0.10           0.08
  SD               0.21           0.12           0.13

Student initiations

Group 1 (n=5)

  [unkeyable]      0.18           0.28           0.56
  SD               0.27           0.27           0.90

Group 2 (n=5)

  [unkeyable]       0.51          0.70           0.17
  SD               0.67           1.33           0.13

Combined (N=10)

  [unkeyable]      0.35           0.49           0.37
  SD               0.51           0.93           0.64

(a) Group 1 students functioned at lower levels of communication development than did group 2 students.

Data Analysis

The Micro-Analytic Data Analysis Package[30] was used to enter the coded data; merge the categories coded for each session; produce transcripts of the timed sequence of behaviors during each session; and provide basic calculations of frequency, rate, average duration, and proportional duration of the observation for each coded behavior. Frequencies indicated the number of times a behavior occurred during an observation. Rate was expressed as the average number of times a behavior occurred per minute during an observation. Average duration was calculated by adding the durations of each episode of communicative behaviors during an observation and dividing by the total number of episodes. Proportional durations were calculated by adding the durations of each episode of communication and dividing by the length of the observation. Proportional durations were used, rather than total durations of communication, because the length of the sessions varied by as much as 45 seconds. Means of the values obtained for the two observations of each position, during both the unstructured and structured interactions, were used for data analysis.

Prior to statistical analyses, raw data were plotted to examine any patterns revealed by the individual scores.[31] The plots indicated that five of the students exhibited considerably more communication and received more communication from adults than did the other five students. Identification of the students in each group suggested that one group could be functioning at a higher level of communiction development than the other group, an impression that was confirmed by the students' scores on the Social Interaction Scales of the Modified Interview Form[25] of the Early Social Communication Scales.[26] Five students (students 1, 2, 5, 8, and 10, comprising group 1) had average age-equivalent scores below 7 months, and five students (students 3, 4, 6, 7, and 9, comprising group 2) had average age-equivalent scores of 7 months or above. Observations of students' functioning were consistent with their groupings by the Social Interaction Scales scores. Because an effect of the students' levels of communicative functioning was possible, a group factor was included in the analyses to separately examine outcomes for the higher- and lower-functioning students.

Data were analyzed using Statistical Analysis System (SAS)[32] general linear models procedures for repeated-measures analysis of variance (ANOVA), with Tukey post hoc analyses of main effects. Nonparametric procedures were used when data did not meet assumptions of the ANOVA models. Effect sizes (f) were calculated according to Cohen,[33] using [[unkeyable].sub.1] as the proportion of total variance accounted for by the effect.[34]

Prior to data collection, it was decided that effects would be said to be significant if the probability of a Type I error was .10 or less. This level was chosen because few strategies have been shown to be effective with children who have profound multiple disabilities[35] and the risk of a Type II error, which could cause potentially useful findings to be discounted, was greater than the risk of a Type I error. It has been argued that the nature of the problem should dictate the optimal alpha level,[36,37] and the nature of this study meets many of the criteria that Labovitz[38] suggests for selecting alpha levels greater than .05. These criteria include: (1) the gravity of consequences of an error (eg, decisions about administering a potentially fatal drug versus decisions about interventions with less serious consequences), (2) lack of existing theory and empirical evidence to suggest the plausibility of results, and (3) the need to explore interrelations for the purpose of developing hypotheses to be tested in other investigations.

Results

Unstructured Observations

During the 5-minute unstructured observations, when adults and students were free to interact as they chose, little interaction took place, as had been anticipated by pilot data and reports of other investigations.

Adult initiations. Adults initiated interactions an average of 0.12 times per minute, or less than once during each 5-minute session (Tab. 3). Even though the rates of initiation were low, effects were found for position (F=3.11; df=2,16; P=.07) and for student group (F=4.24; df=1,8; P=.07), with adults initiating more interactions with the higher-functioning group 2 students. Post hoc comparisons of the main effect for position showed that adults initiated interactions more often when students were in their wheelchairs than when they were in the freestyle (P<.08) or the sidelying (P<.10) position (Fig. 1). Differences were not shown between the sidelying and freestyle positions. The effect size of student position on adult initiations was .24, which Cohen[33] considers to be a small to medium effect size in the behavioral sciences.

Students' initiations. Students initiated communication somewhat more often than adults, averaging 0.40 initiations per minute (compared with an average of 0.12 initiations per minute for the adults) (Tab. 3). The Wilcoxon Matched-Pairs, Signed-Ranks Test was used for pair-by-pair comparisons of the students' average rates of initiations when they were in each of the three positions. No effects of position were shown. Comparisons of wheelchair and sidelyer initiations, wheelchair and freestyle initiations, and sidelyer and freestyle initiations yielded Z values of -.204 (P=.84), -.255 (P=.80), and -.612 (P=.54), respectively. The students' group was also not shown to be related to their average rate of initiations, with a chi-square (1 df) value of 0.61 (P=.43), as determined by the Kruskal-Wallis test.

Maintenance of communication.

Maintenance of communication was coded when a communicative behavior occurred within 30 seconds of the communicative partner's communication. Because the rates of initiation were so low, there were few opportunities for maintaining behaviors, by either children or adults. The frequency of students' communication that maintained interactions was not affected by their position (F=0.18; df=2,16; P=.83), nor were the average duration of maintaining behaviors (F=1.73; df=2,16; P=.21) and the proportion of the session in which maintaining behaviors occurred (F=1.03; df=2,16; P=.37). Because there was only one instance of a communication maintaining behavior by a group 1 student, an effect of student group was found for frequency (F=8.71; df=1,8; P=.02), average duration (Kruskal-Wallis test, [X.sup.2][1]=9.29, P=.002), and proportional duration ([X.sup.2][1]=9.30, P=.002) of communication maintenance during the unstructured observations. Adults' communication maintaining behaviors were too few to include in the statistical analyses.

[TABULAR DATA OMITTED]

Observations During the Interaction Opportunity Routine

The average durations of each episode of communication and the proportional duration of the session during which students and adults interacted during the interaction opportunity routine were compared by group and position. As during the unstructured component, there were differences in the communication of group 1 and group 2 students, this time in the proportional duration of the session during which they interacted (F=21.32; df=1,8; P=.002). The average durations of their interaction episodes were less disparate (F=1.79; df=1,8; P=.22).

Position was not found to have an effect on either the average duration (F=2.45; df=2,16; P=.12) or the proportional duration (F=1.42; df=2,16; P=.27) of communication of the students as a total group. A group x position effect was clear for both the average duration of the lower-functioning students' episodes of communication (F=9.20; df=2,16; P=.002) and the proportional duration of the session during which they interacted with the attentive adult (F=5.16; df=2,16; P=.02). Post hoc comparison of main effects for position showed that the average duration of the lower-functioning students' interactions was higher (P<.05) when the students were positioned freestyle than when they were positioned in either their wheelchairs or the sidelyer. Differences were not found between wheelchair and sidelyer positioning. The effect size of lower-functioning students' position on the average duration of their communication was large[33] (f=1.89).

The proportional durations of the students' interactions were also higher in the freestyle position than in the sidelying and wheelchair positions (P<.08), with no differences between wheelchair and sidelyer positioning. The greater amount of communication in the freestyle position was reflected by a large effect size of 1.26. The average durations and proportional durations of the students' interactions by position and group are shown in Figures 2 and 3. Descriptive statistics are presented in Table 4.

Discussion

The interactions between these students with profound multiple disabilities and their classroom staff support the proposition that the effects of positioning are complex and that there is no one best position for all students or for all purposes.[2,6,11] Even among this group of students who were considered to have the most severe disabilities and who communicated at a nonsymbolic level, there were individual effects of position and effects that were related to the students' functional levels within the classification of profound disability. In spite of the variability, position did appear to serve as a control parameter of two distinct elements of interactions between the students and classroom staff: (1) the spontaneous, unstructured communicative initiations of the adults and (2) the interactions of the lower-functioning students when they were provided a responsive, communicative partner.

The finding that adults initiated communication at higher rates when students were in their wheelchairs lends support to observations that wheelchair positioning provides greater opportunities for social interaction than other positions.[39] This may be the case, at least in part, because wheelchairs place students nearer the usual interaction height of ambulatory individuals than do positioning devices placed on the floor.[5] Adults may also expect students to be more communicative when they are upright than when they are in recumbent positions. For example, classroom personnel made several comments, such as, "Uh, oh, she's lying down, she'll go to sleep." Such adult expectations of student receptivity are likely to influence their own communicative behaviors directed toward the student, as demonstrated by a number of studies examining effects of teacher expectation on their own classroom behaviors.[40]

Because there were few adult initiations overall and because factors other than position may have influenced the initiations that did take place, caution must be taken not to overextend the results. The findings, however, are consistent with reported observations and suggest that more extensive research concerning the effect of students' positioning on adult behaviors could have important implications for intervention, both in providing students with positioning that leads to maximal opportunities for communication and by increasing the awareness of therapists and teachers of the need to ensure that students have communication opportunities in all of the positions in which they may be placed.

Effects of position on students' communication were not found for either group of students during the unstructured observations, or for the higher-functioning students during the structured observations. Variability between and within students, and the low volume of communication during unstructured observations, however, could have obscured differences that may have been revealed had there been more students, or more homogeneous students, and a greater amount of interaction.

The finding that the lowest-functioning students communicated more in the freestyle position during the structured interactions was not expected, because the positioning literature generally cites only detrimental or nonbeneficial effects of the supine position, or positioning without assistive equipment.[1,3,4] The result, however, is consistent with the results of a study[12] in which communication of 3- to 6-month-old infants, as indicated by duration of their gaze at their mothers, was greater when the infants were positioned supine, as compared with semireclined or upright. Although caution is needed when speculating about possible application of infant research findings to students such as those who participated in the study, the similar effect of position on their communication suggests that the infant research literature may provide fruitful directions for future research with students who have profound disabilities.

Questions about why the lowest-functioning students interacted more during freestyle positioning than during wheelchair or sidelying positioning cannot be answered by this study. One possible influence on their interactions that should be investigated further is head control, as it affects the position of the head and orientation of the head and eyes toward the communicative partner. Although all of the group 1 students maintained their heads in an upright position when they were in their wheelchairs, they did appear to have more difficulty orienting toward the adult than did group 2 students. Occasionally, these students even appeared to be "stuck" to one side, perhaps as an influence of the asymmetrical tonic neck reflex (ATNR). Considering that primitive reflexes (including the ATNR and tonic labyrinthine reflex) are reportedly stronger in the supine position than when in upright adaptive positioning devices, the freestyle position would be expected to lead to greater difficulty controlling the position of the head than when seated. This difficulty would, in turn, create a more negative influence on communication, which was not the case in this study.

It is likely that there was no single cause for the finding that the lowest-functioning students engaged in more interactions in the freestyle position than in the other two positions. Rather, from a dynamic systems point of view, the behavioral outcomes were self-organized from an interaction of the students' characteristics, the task at hand, and the various environmental elements. In this case, the qualities of the lower-functioning students merged with the position, the structured social-communicative task, and an attentive communication partner to produce a greater amount of communication when the students were in the freestyle position. Under these conditions, the freestyle position served as a control parameter of the students' communication.

The blend of exogenous and endogenous factors of the higher-functioning students, including their position, did not influence the measured communicative behaviors. This finding is also consistent with a dynamic systems perspective, because control parameters change with developmental changes and varying system conditions.[18] As such, it is likely that position will differentially affect the interactions of students who are functioning at different developmental levels and who are involved in various educational tasks. As an example of an effect of position on a higher-level communication skill, it was found that access to communication board locations was less effective when students were positioned in a sidelyer than when they were positioned using a chair, a prone-wedge, or a stander.[11]

Conclusions and Implications

The potential of position to serve as a control parameter of communication by both students and adults was demonstrated by this study. Implications for physical therapists are suggested by the findings that position affected adults' communicative behaviors and that there may be a mismatch between the position in which adults offer the greatest opportunity for interaction and the position in which some students are the most communicative. These findings could be clinically important if positioning leads to even small improvements in the educational experiences of children with the most severe disabilities, for whom few effective intervention strategies have been found.[35]

Therapists and others involved with students who have profound multiple disabilities should consider the possibility that students' opportunities for communication may be limited when they are placed in freestyle or sidelying positions. When such positions are used, it would be well to be alert to any social isolation and ensure that appropriate opportunities for interaction are provided.

The finding that the most profoundly disabled students interacted more during the structured observations when in the freestyle position was unexpected. Further investigation is needed, but the supine position (with appropriate pillows or other supports) could be the most effective position in which to place students with the most severe disabilities when attempting to teach fundamental interaction skills. Because prolonged placement in the supine position may be both physically detrimental and socially isolating, it is suggested only for active intervention, when engaged in social interaction.

Several avenues for future research have been suggested throughout this discussion, and, in general, research is needed to identify any consistent effects of position that are associated with various student characteristics, contexts, and instructional goals. One of the potential effects of positioning that was not clarified by this study and that deserves further investigation is the effect of position on the students' level of arousal. Low levels of arousal and inattentiveness are frequently cited as primary problems of educational intervention with the lowest-functioning students,[35] and improved level of arousal may be a useful educational goal for some students.[41]

There may also be effects of the standing position, which could not be examined in this study. Standing puts students near the normal interaction level of adults and ambulatory classmates. Thus, standing, like wheelchair positioning, could offer more opportunities for interaction than positioning on the floor. Standers may also provide postural supports that could enhance the students' own communicative behaviors.

One other potentially important effect of a positioning device that was considered, but not explored, in this study is the mobility offered by a wheelchair. During the observations, the students' wheelchairs remained in one location, but wheelchairs can allow mobility to locations at which greater opportunities for interaction exist and may help to reduce the amount of time students are not engaged in instruction, reported to be common in programs for the most severely disabled students.[42]

Acknowledgments

I am grateful to the school personnel and students who participated so enthusiastically in the study; to the Purdue University augmentative and alternative communication research group for their continuing support and helpful suggestions; to Alan Fogel, PhD, for introducing me to dynamic systems and his infant research; and to Barbara Neas, PhD, for her help with data analysis. I particularly thank George R Karlan, PhD, and Lyle L Lloyd, PhD, for sharing their considerable expertise, energy, and enthusiasm.

(*1)Tumble Forms Inc, 60 Page Rd, Clifton, NJ 07012.

[unkeyable]Rifton, Rte 213, Rifton, NY 12471.

[unkeyable]Hitachi, 1400 Morse Ave, Elk Grove Village, IL 60007.

References

1 Bergen AF, Presperin J, Tallman T. Positioning for Function: Wheelchairs and Other Assistive Technologies. Valhalla, NY: Valhalla Rehabilitation Publications Ltd; 1990.

2 Campbell PH. Dysfunction in posture and movement with individuals with profound disabilities. In: Brown F, Lehr DH, eds. Persons with Profound Disabilities: Issues and Practices. Baltimore, Md: Paul H Brookes Publishing Co; 1989:163-189.

3 Connolly B, Schultz J, Greer BB. Motor development. In: Greer JG, Anderson RM, Olde SJ, eds. Strategies for Helping Severely and Multiply Handicapped Citizens. Baltimore, Md: University Park Press; 1982:157-213.

4 Fraser BA, Hensinger RN, Phelps JA. Physical Management of Multiple Handicaps: A Professional's Guide. 2nd ed. Baltimore, Md: Paul H Brookes Publishing Co; 1990.

5 McEwen IR, Lloyd LL. Positioning students with cerebral palsy to use augmentative and alternative communication. Language, Speech, and Hearing Services in Schools. 1990;21:15-21.

6 Mac Neela JC. An overview of therapeutic positioning for multiply-handicapped persons, including augmentative communication users. In: Attermeier SM, ed. Augmentative Communication: Clinical Issues. New York, NY: The Haworth Press Inc; 1987:39-60.

7 Nwaobi OM. Seating orientations and upper extremity function in children with cerebral palsy. Phys Ther. 1987;67:1209-1212.

8 Nwaobi OM, Smith PD. Effect of adaptive seating on pulmonary function of children with cerebral palsy. Dev Med Child Neurol. 1986;28:351-354.

9 Nwaobi OM. Effects of body orientation in space on tonic muscle activity of patients with cerebral palsy. Dev Med Child Neurol. 1986;28:41-44.

10 Miedaner JA. The effects of sitting positions on trunk extension for children with motor impairment. Pediatric Physical Therapy. 1990;2:11-14.

11 McEwen IR, Karlan GR. Assessment of the effects of position on communication board access by individuals with cerebral palsy. Augmentative and Alternative Communication. 1989;5:235-242.

12 Fogel A, Dedo JY, McEwen IR. Effect of postural position and reaching on gaze during mother-infant face-to-face interaction. Infant Behavior and Development. In press.

13 Hulme JB, Gallacher K, Walsh J, et al. Behavioral and postural changes observed with use of adaptive seating by clients with multiple handicaps. Phys Ther. 1987;67:1060-1067.

14 Giuliani CA. Theories of motor control: new concepts for physical therapy. In: Contemporary Management of Motor Control Problems: Proceedings of the II STEP Conference. Alexandria, Va: Foundation for Physical Therapy Inc; 1991:29-35.

15 Heriza CB. Implications of a dynamical systems approach to understanding infant kicking behavior. In: Movement Science. Alexandria, Va: American Physical Therapy Association; 1991:214-227.

16 Horak FB. Assumptions underlying motor control for neurologic rehabilitation. In: Contemporary Management of Motor Control Problems: Proceedings of the II STEP Conference. Alexandria, Va: Foundation for Physical Therapy Inc; 1991:11-27.

17 Thelen E, Kelso JAS, Fogel A. Self-organizing systems and infant motor development. Development Review. 1987;7:39-65.

18 Fogel A, Thelen E. The development of early expressive and communicative action: reinterpreting the evidence from a dynamic systems perspective. Child Dev. 1987;23:747-761.

19 Kelso JAS. Epilogue: two strategies for investigating action. In: Kelso JAS, ed. Human Motor Behavior: An Introduction. Hillsdale, NJ: Lawrence Erlbaum Associates Inc; 1982:283-287.

20 Bertenthal BI, Campos JJ, Barrett KC. Self-proudced locomotion: an organizer of emotional, congnitive, and social development in infancy. In: Emde RN, Harmon RJ, eds. Continuities and Discontinuities in Development. New York, NY: John Wiley & Sons Inc; 1984:49-76.

21 Thelen E, Fisher DM. Newborn stepping: an explanation for a "disappearing" reflex. Dev Psychol. 1982;18:760-775.

22 Haley SM. Commentary. Phys Ther. 1991;71:577-578.

23 Harris SR. Functional abilities in context. In: Contemporary Management of Motor Control Problems: Proceedings of the II STEP Conference. Alexandria, Va: Foundation for Physical Therapy Inc; 1991:253-259.

24 Siegel-Causey E, Downing J. Nonsymbolic communication development: theoretical concepts and educational strategies. In: Goetz L, Guess D, Stremel-Campbell K, eds. Innovative Program Design for Individuals with Dual Sensory Impairments. Baltimore, Md: Paul H Brookes Publishing Co; 1987:15-48.

25 Karlan GR, Pennington G, Granlund M, Ward MF. Modified Interview Version of the Early Social Communication Scales. West Lafayette, Ind; Special Education, Purdue University; 1986.

26 Seibert JM, Hogan AE. Early Social Communication Scales. Miami, Fla: Mailman Center for Child Development, University of Miami; 1982.

27 Karlan GR, Ward MF. Videotaped Interactions Analysis: West Lafayette, Ind: Special Education, Purdue University; 1986.

28 Houghton J, Bronicki GJB, Guess D. Opportunities to express preferences and make choices among students with severe disabilities in classroom settings. The Journal of the Association for Persons with Severe Handicaps. 1987;12:18-27.

29 Hollenbeck AR. Problems of reliability in observational research. In: Sackett GP, ed. Observing Behavior, Volume 2: Data Collection and Analysis Methods. Baltimore, Md: University Park Press; 1978;79-98.

30 Kienapple K. Micro-Analytic Data Analysis Package. Halifax, Nova Scotia, Canada: Mount Saint Vincent University; 1986.

31 Borg WR, Gall MD. Educational Research. 5th ed. New York, NY: Longman Inc; 1989.

32 SAS User's Guide: Statistics, Version 5 Edition. Cary, NC: SAS Institute Inc; 1985.

33 Cohen J. Statistical Power Analysis for the Behavioral Sciences. 2nd ed. Hillsdale, NJ: Lawrence Erlbaum Associates Inc; 1988.

34 Winer BJ. Statistical Principles in Experimental Design. New York, NY: McGraw-Hill; 1971.

35 Thompson B, Guess D. Students who experience the most profound disabilities: teacher perspectives. In: Brown F, Lehr DH, eds. Persons with Profound Disabilities: Issues and Practices. Baltimore, Md: Paul H Brookes Publishing Co; 1989:3-41.

36 Rothstein JM. Editor's note: much ado about probability. Phys Ther. 1990;70:535-536.

37 Skipper JK Jr, Guenther AL, Nass G. The sacredness of .05: a note concerning the uses of statistical levels of significance in social science. The American Sociologist. 1967;2:16-18.

38 Labovitz S. Criteria for selecting a significance level: a note on the sacredness of .05. The American Sociologist. 1986;4:200-222.

39 Fulford GE, Cairns TP, Sloan Y. Sitting problems of children with cerebral palsy. Dev Med Child Neurol. 1982;24:48-53.

40 Good TL. Two decades of research on teacher expectations: findings and future directions. Journal of Teacher Education. 1987;38(4):32-47.

41 Rainforth B. Biobehavioral state and orienting: implications for educating profoundly retarded students. Journal of the Association for the Severely Handicapped. 1982;6:33-37.

42 Lehr DH, Noonan MJ. Issues in the education of students with complex health care needs. In: Brown F, Lehr DH, eds. Persons with Profound Disabilities: Issues and Practices. Baltimore, Md: Paul H Brookes Publishing Co; 1989:139-160.