Modification of sociometric status and social interaction with mainstreamed mild developmentally disabled children

Modification of sociometric status and social interaction with mainstreamed mild developmentally disabled children

Analysis and Intervention in Developmental Disabilities, Vol. 1, pp. 157-169, 1981 Printed in the USA. All rights reserved. 0270-4684/81/020157- 1350...

686KB Sizes 0 Downloads 34 Views

Analysis and Intervention in Developmental Disabilities, Vol. 1, pp. 157-169, 1981 Printed in the USA. All rights reserved.

0270-4684/81/020157- 13502.00/0 Copyright © 1981 Pergamon Press Ltd

Modification of Sociometric Status and Social Interaction with Mainstreamed Mild Developmentally Disabled Children Phillip S. Strain University of Pittsburgh

An intervention package based upon personal attraction theory was used to modify the sociometric standing and social interactions of three mildly developmentally disabled boys enrolled in a regular fourth-grade class. The study employed a withdrawal of treatment (ABAB) design with replication across subjects. Also, sociometric ratings and social interactions were assessed at two- and four-week follow-up points. Intervention consisted of three, interrelated components: (a) assignment of each target child as a team captain for a daily competition game; (b) manipulation of target children's scores such that the teams always won a prize; and (c) target children's distribution of prizes to teammates. Results indicate that: (a) the intervention procedure produced an immediate and durable increase in the sociometric standing of the boys; (b) an immediate and durable decrease in negative social interactions between the target children and regular class peers obtained; and (c) the positive interaction between the three boys and class peers was substantially increased when the intervention was in effect.

The considerable volume of research that has addressed the social standing of mentally retarded children in regular classroom settings has produced several consistent findings. First, mentally retarded youngsters are viewed by normal peers to be less academically and socially competent than their actual behavior would suggest (Cook & Wollersheim, 1976; Gottlieb, 1974; Jaffe, 1966). Second, mentally retarded children have been given consistently lower sociometric ratings than nonhandicapped children in the same classrooms (Bruininks, Rynders, & Gross, 1974; Goodman, Gottlieb, & Harrison, 1972; Gottlieb & Davis, 1973). Third, this sociometrically-measured rejection of mentally retarded youngsters is represented in limited, often negative interaction patterns between these children and normal class peers (Strain & Hill, 1979; Strain, Shores, & Reprintsmay be obtainedfromPhillipS. Strain,WesternPsychiatricInstituteand Clinic, University of PittsburghSchoolof Medicine, 3811 O'Hara St., Pittsburgh, PA 15261. 157

158

Phillip S. Strain

Timm, 1977). Viewed together, these findings suggest that the mere physical integration of normal and mentally retarded children will not necessarily lead to either social acceptance or positive social interaction. Moreover, this general pattern of social rejection is evident across preschool, elementary, and secondary class settings (Strain & Hill, 1979) as well as degree of mental retardation. A number of intervention strategies have been designed to alter the sociometric status and social interaction patterns of integrated mentally retarded children. These strategies may be divided into three basic groups. Initially, a number of early attempts to alter sociometric status employed what basically were highinterest, organized activities and games that would provide a positive context in which interaction between normal and mentally retarded children was likely to occur (Chennault, 1967; Lilly, 1971; McDaniel, 1970; Rucker & Vincenzo, 1970). These efforts produced equivocal results. Although some evidence indicated that sociometric choices were altered in a positive direction while the intervention was in effect, no maintenance of change following intervention or increase in positive social interaction was found. A second group of studies has emphasized the training of nonhandicapped peers who subsequently have encouraged social interaction between themselves and socially-withdrawn, mentally retarded classmates (Strain, 1977; Strain et al., 1977). Although these peer-mediated interventions have produced substantial changes in positive interaction between peer trainers and target children, the results leave a number of critical conceptual and applied issues unresolved. For example, these peer-mediated interventions have been applied exclusively in experimentally-created, free-play sessions with a limited number of children present. The generality of the techniques to more naturalistic classroom environments is undetermined. Second, the across-setting and across-time generalization of positive social behavior using these techniques has been extremely limited. Finally, no evidence of generalization of treatment effects to new social partners has been found. A final group of studies has employed global treatment "packages" in which mentally retarded children have been trained to imitate the social behaviors of normal class peers, who later prompt and reinforce their handicapped classmates for imitating socially appropriate behaviors (Apolloni & Cooke, 1978; Apolloni, Cooke, & Cooke, 1977; Guralnick, 1976). The basic intervention data on these techniques are encouraging; however, generalized behavior change has not been demonstrated. Moreover, Cooke and his colleagues have shown that direct treatment effects are severely limited when normal developing children have the option of playing with nonhandicapped peers. In summary, the currently available interventions for improving the social status and social interactions of integrated mentally retarded children have yet to produce other than transient, setting-specific sociometric and behavior changes. The present study was designed to test the efficacy of an intervention technique based upon the theoretical notions and empirical data of the personal attraction and friendship formation literature (Huston, 1974).

Social Status and Interaction

159

Lott and Lott (1974) have described four basic processes by which interpersonal attraction is enhanced and friendships are formed. First, persons may possess such physical attributes (e.g., beauty, muscle tone) and personal characteristics (e.g., community leader, close relative) that, regardless of overt behavior, they are sought-out as social interactants. Second, a person may be inadvertently "paired" with some pleasant occasion or setting (e.g., Christmas season, vacation) such that this individual is associated with pleasurable consequences. Third, a person's behavior may be directly instrumental in another individual acquiring some positive consequence. For example, a star athlete's competence at a particular sport may insure that his teammates are successful in a competitive game. Finally, a person may provide specific, positive behaviors to social partners. Here, Lott and Lott (1974) maintain that positive verbal comments and expressions of approval and affection typically lead to an increased level of attraction. The intervention technique evaluated in this study represented a direct attempt to apply Lott and Lott's (1974) notions of interpersonal attraction to improve the sociometric status and social interactions of integrated mildly developmentally disabled children. Specifically, the intervention package was designed to: (a) create an ongoing, competitive game situation in which three sociallyrejected children were always responsible for their team's acquiring positive consequences; (b) pair the acquisition of positive consequences with the presence of the target youngsters; and (c) assign the target children leadership roles in the team competition context. The evaluation of treatment effectiveness included both sociometric and behavioral measures as well as immediate and follow-up assessment.

METHOD Subjects

The target subjects were three 10-year-old boys who had been integrated for a period of three months in a regular fourth grade class with 24 normal peers. The three boys were previously served in separate, segregated classes for mild mentally retarded children. Stanford-Binet Form L-M IQ scores for Tim, Joe, and Ralph were 79, 80, and 75, respectively. The classroom teacher reported that the target subjects' academic achievement was the lowest of any children in the class. Also, the teacher reported that these boys were seldom observed to play with each other or any other child in the class. Setting

This study was conducted in two separate settings. The intervention procedures were implemented in an area within the subjects' classroom that measured

160

Phillip S. Strain

2.5 x 6 m. This area included a "bull's eye" target that was drawn on the floor and a 2.2 x 2.8 m partition. On each day of the study, all children in the class participated in a team-based, bean bag toss game in this area. The behavioral effects of the intervention procedures were assessed during the subjects' daily play period that took place in the school gym. The play period was 30 minutes in length, with the first 15 minutes devoted to organized physical exercise. During the second 15-minute segment all 27 children were allowed to engage in self-selected activities. It was during this free-play period that social interaction data were collected. The gym period began 1.5 hours following the bean bag toss game. Teacher involvement in this time period was limited to stopping occasional fights and redirecting children who were engaging in physically dangerous behavior. Play equipment available to the children in the gym included: athletic balls, tumbling mats, hopscotch court, basketball goals, frisbees, plastic bowling set, and jump ropes.

Sociometric Assessment

The social standing of all children in this class was assessed using a sociometric device identical to that used by Goodman et al. (1972). Each child in the class was read a list of all classmates and asked to describe how they felt about each child. Three choices were available to the respondents. They could circle one of the three stick figure drawings that represented: (a) two children playing ball together (labeled "friend"); (b) two children at a blackboard (labeled "alright"); and (c) two children with their backs toward each other (labeled "wouldn't like"). The instrument was administered individually by one of the experimenters once during each experimental condition and at two- and fourweek follow-up points.

Behavioral Assessment

Social interactions between the three target subjects and their normal class peers were assessed via an observational system developed by Strain and Timm (1974). Here, four basic categories of social behavior were observed: positive motor-gestural, positive vocal-verbal, negative motor-gestural, and negative vocal-verbal. The operational definition for each category is detailed below: Positive Motor-Gestural: touch with hand or hands; hug; holding hands; kiss; wave; all cooperative responses involved with sharing a toy or materials. Positive Vocal-Verbal: all vocalizations directed to another child excluding screams, shouts, cries, whines, or other utterances which are accompanied by gestures which indicate negative, rejecting behavior.

Social Status and Interaction

161

Negative Motor-Gestural: hit; pinch; kick; butt with head; "nonplaying" push or pull; grabbing object from child; destroying construction of another child. Negative Vocal-Verbal: screams, shouts, cries, whines, or other utterances which are accompanied by gestures which indicate negative, rejecting behavior. These social behavior categories were coded as to whether they occurred as initiated or responded events in an interaction sequence. Initiated behaviors included all motor-gestural and vocal-verbal acts that occurred either three seconds before or after another child's designated social behavior. Responded behaviors included all motor-gestural and vocal-verbal behaviors that occurred within three seconds of another child's social behavior. Prior observational research has shown that the three second criterion for discriminating initiated from responded behavior is a valid index (Tremblay, Strain, Hendrickson, & Shores, in press). Two categories of adult behavior during the free-play time were recorded also: (a) verbal and physical prompting to engage in positive interaction; and (b) social praise delivered contingent upon positive social interaction. Specific operational definitions of these categories are detailed in Strain, Shores, and Kerr (1976). Social behaviors of each target subject and class peers were recorded during free-play for five consecutive minutes, yielding 15 total minutes of observation daily. The order in which the subjects were observed was counterbalanced across all days. Interactions that did not involve one of the target children were not recorded. Entries of specific social behaviors were made by trained data collectors using coded symbols and a prepared record sheet (see Figure 1, Strain & Timm, 1974). The behaviors were recorded continuously within 10-second interval blocks, thus absolute frequency of occurrence data were obtained. Each social behavior was coded as having been emitted by either of the target subjects (Tim, Joe, or Ralph) or any of the peers; as being either motor-gestural or vocal-verbal in type; as being of positive or negative topography; and as having occurred as an initiated or responded event in an interaction episode. Adult praise, delivered contingent upon positive social interaction was entered by placing the symbol (t) just above the positive social behavior immediately preceding delivery of the event. Adult prompts to participate in positive interaction were recorded in a like manner, using the symbol (c). On each day of the study, two observers were present. Interobserver agreement was calculated by dividing the total number of target behaviors and corresponding topographic features recorded in agreement by that number plus those recorded in disagreement, then multiplying by 100. Agreement was reached only when both observers marked the same behaver, the same general category of behavior (motor-gestural or vocal-verbal), the same positive or negative and initiated or responded features, all within the same 10second interval of the recording sheet. A criterion of three consecutive days of agreement above 80% preceded formal data collection.

162

Phillip S. Strain

Experimental Design This study utilized a withdrawal of treatment (ABAB) design to assess the effects of the intervention package on the sociometric status and social interactions of the target subjects. The procedures in effect during each experimental condition are presented below:

Baseline I. During this condition, the children in the class were divided into three teams of nine children. One of the target children was assigned to each team. At the same time each day, the teams participated in a bean bag toss game in which the children tossed over the partition to a "bull's eye" target drawn on the floor. The partition was placed such that the children could not see the target. Each member of a team could receive a small piece of candy if the team's total score on the toss game was at least 100. The order in which the children on each team tossed was determined randomly by the teacher each day. Following each toss, the teacher called out the actual score obtained and the current total. Scores could range from 0 to 100, in intervals of 20. On two of the 10 days in this condition, each team surpassed 100 points, and the teacher handed out candy to the team members.

Intervention 1. At the beginning of this condition, the teacher explained to the children that the teams might be more successful if they had a captain. She then assigned each of the three boys as a captain of his respective team. It was further explained that the captain was to decide the daily order for the team members to toss, and that the captain would toss last each day. On each day of this condition the teacher called out a score for the captain of each team that was sufficient to bring the team's score above 100. Each team captain then distributed candy to his team members. This condition was in effect for 10 days for each team.

Baseline H. The identical procedures employed during Baseline I were reinstituted in this condition. Baseline II remained in effect for 10 days.

Intervention H. The treatment package implemented during Intervention I was reinstituted in an effort to replicate the sociometric and behavioral change exhibited during the first intervention condition. Intervention II continued for 10 days. Two and four weeks following the termination of Intervention II, follow-up sociometric and social interaction measures were taken.

Social Status and Interaction

163

RESULTS Observer Agreement

Mean percent of agreement across the 42 observation days ranged from 82 to 100%, with a mean of 88% for all behaviors reported. Teacher Behaviors

The various teachers who at different times were located in the gym area carefully followed the experimenter's request not to promote peer interaction with the target children. Across all experimental days only three prompts and four praise events were observed. Sociometric Changes

Table 1 depicts the number of "Wouldn't like," "Alright," and "Friend" ratings for Tim, Joe, and Ralph across each experimental condition. During Baseline I these boys were by far the lowest rated children in the class. In fact, only one other child received a single "Wouldn't like" rating. On the ninth day of Intervention I the sociometric was conducted again. At this point, Tim, Joe, and Ralph's ratings had changed considerably. In fact, there were only seven children who received more "Friend" ratings than the target subjects. When the intervention procedures were terminated during Baseline II, the three boys' ratings were less favorable, but still well above their initial ratings. During the final intervention condition, ratings for each boy closely replicated the level of favorability exhibited during Intervention I. When the sociometric instrument was given again at two- and four-week follow-up points, no appreciable changes from the Intervention II assessment were noted.

Social Interaction Changes

Figure 1 presents the total number of positive and negative social behaviors exhibited by the three target subjects and their class peers across all experimental conditions. Each target subject data point depicts the total number of initiated and responded, vocal-verbal and motor-gestural behaviors of a positive or negative type exhibited on that day. Each peer data point depicts the total number of initiated and responded, vocal-verbal and motor-gestural behaviors of a positive or negative type exhibited during interaction with the specified target subject on that day. No social behaviors exhibited during interaction between the target subjects are included in Figure 1. Figure 1 shows that Tim, Joe, and Ralph rarely engaged in any positive

164

Phillip S. Strain

e~ I

°~

,d

z

Social Status and Interaction • O • ~,

Target Bubjecl's Peers' Positive Target aubjoct's Peors' Negative

BASELINE I

165

Positive k h l v i o r Behavior Negative Behavior Behavior

INTERVENTION

BABE LINE II

INTERVENTION II

i g Tim =

o

~_-_..__-__~-__ 25 m

o >

20

<

Z

Joe

w Z IM G

m I

5

i 0

A

25

20

Ralph

o

5

I0

15

20

25

30

35

40"

DAYS

FIGURE 1. Total number of positive and negative social behaviors exhibited by each target subject and class peers across all experimental conditions and follow-up points.

166

Phillip S. Strain

interaction with peers during Baseline I. In fact, negative contacts between these boys and classmates were three to four times as frequent as positive contacts during this condition. When the treatment procedure was implemented during Intervention I, both positive and negative interaction patterns were altered immediately. Negative interactions declined to zero levels and positive social behaviors increased twentyfold. When the social attraction intervention was terminated in Baseline II, the frequency of positive social behaviors reduced gradually, while negative interactions remained at a very low level. During the reinstatement of treatment procedures in Intervention II, the level of behavior change closely replicated that observed in the initial treatment condition. At the first follow-up point, no negative contacts between Tim, Joe, Ralph and their peers were observed. Some maintenance of positive interaction was evident. Tim, Joe, and Ralph engaged in 8, 7, and 8 positive behaviors, respectively. Peers of Tim, Joe, and Ralph engaged in 9, 8, and 10 positive behaviors. The final follow-up session resulted in a similar behavior pattern with no negative behaviors observed, and a stable level of positive contact between the target children and their peers. Total positive and total negative behaviors were analyzed also with respect to who initiated what kind of contact with whom. Of particular interest was the kind of initiations typically received by the target subjects. During the initial baseline condition, over 90% of all negative social behaviors exhibited by peers were ones initiated toward the target children. The relatively small number of positive behaviors observed during Baseline I were largely composed (85%) of responses to positive initiations by the three boys. During Intervention I, 55% of the peers' positive behaviors were initiations to target children. This percentage did not vary more than three percentage points across the remaining experimental conditions and follow-up points. DISCUSSION The results of this study indicate that: (a) the personal attraction intervention produced very clear and positive changes in the sociometric standing of the three target subjects; (b) the intervention also resulted in an immediate increase in the number of positive social contacts, and an immediate reduction in the number of negative social contacts between each boy and normal peers; (c) analyses of the initiated and responded dimensions of interaction showed that the intervention procedure resulted in a reciprocal exchange of positive social behavior; (d) the intervention also resulted in a partial maintenance of sociometric and behavior change during the return to baseline and both follow-up points. Those results regarding the changes in the sociometric status of the target children indicate that the initial intervention condition produced an improved social standing for these youngsters that did not change appreciably across the remaining experimental conditions and follow-up points. Although some re-

Social Status and Interaction

167

duction in the number of "Friend" ratings was evident during Baseline II and both follow-up evaluations, the level of acceptance during these periods was far different from the singular social rejection that characterized the Baseline I judgments by normal peers. These data are in general agreement with Gottman, Gonso, and Rasmussen (1975) and Gottman, Gonso, and Schuler (1976) who suggest that sociometric status among normal children represents a relatively stable metric that can be used to predict later social adjustment and acceptance. However, previous attempts to improve the sociometric status of mildly handicapped children have not resulted in acceptance levels that maintain in the absence of treatment (e.g., McDaniel, 1970; Rucker & Vincenzo, 1970). The maintenance of treatment effects in this study likely is related to the specific, programmed, positive contact between the boys and their peers. In those earlier studies, actual interaction between children, or opportunities for target children to be observed as competent members of a group were left to chance. The observed change in negative contacts between the boys and their nonhandicapped peers closely parallels the sociometric data at two levels. First, the vast majority of negative interaction between the three boys and their peers during Baseline I was initiated by peers. Anecdotal reports by the observers indicated that many of these negative social initiations were composed of taunting, name-calling incidents related specifically to the boys' previous diagnosis as retarded. This behavior pattern is consistent with the verbal reports of these normal youngsters when they provided initial sociometric ratings. At a second level, both sociometric and negative behavior changes maintained following Intervention I. In effect, the initial presentation of the social attraction treatment essentially eliminated further negative interactions between the boys and their class peers. Similar effects on the negative interactions of isolate preschool children have been obtained with teacher prompting and reinforcement of positive social behavior (Strain & Timm, 1974) and peer-mediated social initiations (Strain et al., 1977). It would appear, therefore, that relatively long-term changes in the negative contacts between previously isolate children and normal peers can be affected by intervention aimed specifically at promoting positive social contacts. Whereas the initial sociometric and negative behavior changes evident during Intervention I maintained across the subsequent experimental conditions, very clear experimental control over positive interactions between the target children and their peers was demonstrated. As important as the increased number of positive social behaviors by the boys and their peers may be, the reciprocity of interaction that characterized these positive exchanges is equally significant. Specifically, there was a similar number of positive contacts that were initiated by target children and peers. Such an equal exchange of positive social initiations has been recognized as a principal characteristic of normal preschool and schoolage children's interactions (Tremblay et al., in press; Strain & Hill, 1979). Moreover, the reciprocal exchange of positive social behaviors has been observed

168

Phillip S. Strain

in children who maintain a high social status and many friendships (Hartup, 1978). Not only was the pattern of interaction between target children and peers comparable to normative data, but the overall frequency of positive behavior during intervention conditions is similar to that reported for children identified as socially competent (e.g., Strain & Fox, in press). Although the present intervention procedures produced a most favorable sociometric and behavioral change with the three boys, at least three critical conceptual and applied issues remain unresolved. First, without further replications, it is not possible to determine the efficacy of the personal attraction intervention for mainstreamed children with behavioral handicaps more severe than those evidenced by these three boys. Each of the boys was skilled at the various games and group activities that occurred during the recess periods, and this fact may have well contributed to the observed magnitude, generalization, and maintenance of behavior change. Significantly, a number of authors have proposed that perceived incompetence is a major contributor to the frequent social rejection directed toward mainstreamed children (e.g., Goodman et al., 1972; Strichart & Gottlieb, 1975). It may be that developmentally disabled children with specific play and social skills deficits will require more extensive intervention in order to alter patterns of social rejection and isolation. A second issue centers on the specification of those elements of the treatment package responsible for the sociometric and behavioral change. In future research on the personal attraction intervention, a logical component analysis might include the sequential application of: (a) assignment of target child as a team captain; (b) manipulation of target child's score to insure group reinforcement; and (c) target child's distribution of prizes to teammates. Finally, the fact that deception was used as part of the treatment package may reasonably limit the wide-spread adoption of the procedure. It is possible, of course, that a component analysis might show that assignment as team captain and distributing prizes are conditions sufficient to produce socially significant behavior change. Also, individuallytailored modifications to the general intervention package, such as selecting game-related skills at which target children were particularly competent, might obviate the need to falsify child performance.

Acknowledgement--This research was supported, in part, by a grant from the National Institute of

Child Health and Human Development (HD00973).

REFERENCES Apolloni, T., & Cooke, T.P. Integrated programming at the infant, toddler, and preschool levels. In M.J. Guralnick (Ed.), Early intervention and the integration of handicapped and nonhandicapped children. Baltimore: University Park Press, 1978. Apolloni, T., Cooke, S.A., & Cooke, T.P. Establishing a normal peer as a behavioral model for delayed toddlers. Perceptual and Motor Skills, 1977, 44, 231-241. Bruininks, R.H., Rynders, J.E., & Gross, J.C. Social acceptance of mildly retarded pupils in resource rooms and regular classes. American Journal of Mental Deficiency, 1974, 78, 377-383.

Social Status and Interaction

169

Chennault, M. Improving the social acceptance of unpopular educable mentally retarded pupils in special classes. American Journal of Mental Deficiency, 1967, 72, 455-458. Cook, J.W., & Wollersheim, J.P. The effect of labeling of special education students on the perceptions of contact versus non-oont0ct normal peers. Journal of Special Education, 1976, 10, 187-198. Goodman, H., Gottlieb, J., & Harrison, R.H. Social acceptance of EMRs integrated into a nongraded elementary school. American Journal of Mental Deficiency, 1972, 76, 412-417. Gottlieb, J. Attitudes toward retarded children: Effects of labeling and academic performance. American Journal of Mental Deficiency, 1974, 79, 268-273. Gottlieb, J., & Davis, J.E. Social acceptance of EMRs during overt behavioral interaction. American Journal of Mental Deficiency, 1973, 78, 141-143. Gottman, J., Gonso, J., & Rasmussen, B. Social interaction, social competence, and friendship in children. Child Development, 1975, 46, 709-718. Gottman, J., Gonso, J., & Schuler, P. Teaching social skills to isolated children. Journal of Abnormal Child Psychology, 1976, 4, 179-197. Guralnick, M.J. The value of integrating handicapped and nonhandicapped preschool children. American Journal of Orthopsychiatry, 1976, 46, 236-245, Hartup, W.W. Peer interaction and the processes of socialization, In M.J. Guralnick (Ed.), Early intervention and the integration of handicapped and nonhandicapped children. Baltimore: University Park Press, 1978. Huston, T.L. Foundations of interpersonal attraction. New York; Academic Press, 1974. Jaffe, J. Attitudes of adolescents toward the mentally retarded. Amgrican Journal of Mental Deficiency, 1966, 70, 907-912. Lilly, M.S. Improving social acceptance of low sociometric status, low achieving students. Exceptional Children, 1971, 37,431--448. Lott, A.J., & Lott, B.E. The role of reward in the formation of positive interpersonal attitudes. In T.L. Huston (Ed.), Foundations of interpersonal attraction. New York: Academic Press, 1974. McDaniel, C.O. Participation in extracurricular activities, social acceptance, and social rejection among educable mentally retarded students. Education and Training of the Mentally Retarded, 1970, 5, 4-14. Rucker, C.N., & Vincenzo, F.M. Maintaining social acceptance gains made by mentally retarded children. Exceptional Children, t970, 36, 679-680. Strain, P.S. Effects of peer social initiations on withdrawn preschool children: Some training and generalization effects. Journal of Abnormal Child Psychology, 1977, 5, 445-455. Strain, P.S., & Fox, J.E. Peers as ~havior change agents for withdrawn classmates. In A.E. Kazdin & B. Lahey (Eds.), Advances in clinical child psychology. New York: Plenum, in press. Strain, P.S., & Hill, A.D. Social interaction training with the severely handicapped. In P. Wehman (Ed.), Leisure time skiUsfor the severely handicapped. University Park Press, 1979. Strain, P.S., Shores, R.E., & Kerr, M.M. An experimental analysis of "spillover" effects on social interaction among behaviorally handicapped prescho0~ children. Journal of Applied Behavior Analysis, 1976, 9, 31-40. Strain, P.S., Shores, R.E., & Timm, M.A. Effects of peer social initiations on the behavior of withdrawn preschool children. Journal of Applied Behavior Analysis, 1977, 19, 289-298. Strain, P.S., & Timm, M.A. An experimental analysis of social interaction between a behaviorally disordered preschool child and her classroom peers. Journal of Applied Behavior Analysis, 1974, 7, 583-590. Strichart, S.S., & Gottlieb, J. Imitation of retarded children by their nonretarded peers. American Journal of Mental Deficiency, 1975, 79, 506-513. Tremblay, A., Strain, P.S., Hendrickson, J.M., & Shores, R.E. Social interactions of normally developing preschool children: Using normative data for subject and target behaviors selection. Behavior Modification, in press.