<%BANNER%>

Extending Functional Analysis Methodology to Identify the Function of Peer-Related Withdrawn and Peer-Related Negative S...


PAGE 1

EXTENDING FUNCTIONAL ANALYSIS METHODOLOGY TO IDENTIFY THE FUNCTION OF PEER-RELATED WITHDRAWN AND PEER-RELATED NEGATIVE SOCIAL BEHAVIORS FOR YOUNG CHILDREN WITH AUTISM SPECTRUM DISORDERS By JENNIFER ANNE SELLERS A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2006

PAGE 2

Copyright 2006 by JENNIFER ANNE SELLERS

PAGE 3

iii ACKNOWLEDGMENTS I would like to thank my dissertation chairs, Drs. Tina Smith-Bonahue and Maureen Conroy, for their dedication and commitment in assisting me to complete this project. I would like to tha nk my dissertation committee, Drs. Jennifer Asmus, Terry Scott, and Gregory Valcante, for their expertis e and guidance on this project. I would like to thank the schools, teachers, students, and peers who partic ipated in this study. I would also like to thank Elizabeth McKenney a nd the staff of Project GATORSS, for their assistance in this study; without them this would not be possi ble. Finally, I would like to thank my parents, Arthur and Catherine Sellers, my grandmother, Helen Ellis, and friends, Lori and Johanna. Their love, prayer s, and support helped make this possible.

PAGE 4

iv TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iii LIST OF TABLES...........................................................................................................viii LIST OF FIGURES.............................................................................................................x ABSTRACT....................................................................................................................... xi CHAPTER 1 LITERATURE REVIEW.............................................................................................1 Introduction of Autism Spectrum Disorders (ASD).....................................................1 Diagnostic Criteria for ASD..................................................................................2 Prevalence of ASD................................................................................................4 Identification and Characterist ics of Children with ASD......................................4 Applied Behavior Analysis (ABA).......................................................................6 Summary of ASD..................................................................................................8 Assessment and Treatment of Problem Be havior in Young Children with ASD.........8 Introduction of Assessment Methodology.............................................................8 Functional Analysis...............................................................................................9 Structural Analysis..............................................................................................13 Summary of Assessment and Treat ment of Problem Behaviors.........................17 Review of the Literature.............................................................................................18 Review of Functional Analysis Literature...........................................................18 Review of Functional Analysis Studies...............................................................19 Participants included in review of literature................................................20 Therapists in functional analysis..................................................................21 Settings of functional analysis......................................................................21 Dependent measures.....................................................................................22 Experimental conditions...............................................................................22 Experimental condition modifications.........................................................23 Treatment of problem behavior....................................................................25 Design of functional analysis.......................................................................25 Findings of functional analysis....................................................................27 Summary of functional analys is review of literature...................................28 Review of Social Skills Literature.......................................................................29 Review of Social Skills Studies...........................................................................30

PAGE 5

v Participants included in review of literature................................................31 Therapists in studies.....................................................................................31 Settings of studies.........................................................................................33 Dependent measures.....................................................................................33 Design of studies..........................................................................................34 Assessment of social skills...........................................................................35 Treatment of social skills.............................................................................36 Findings of social skills assessment and treatment......................................37 Summary of social skills review of the literature.........................................37 Conclusions and Future Directions for Research.......................................................39 2 METHOD...................................................................................................................41 Participants.................................................................................................................41 Settings and Therapists...............................................................................................43 Materials.....................................................................................................................4 5 Response Definitions..................................................................................................46 Preference Assessment........................................................................................46 Functional Analysis.............................................................................................46 Participant peer-related withdrawn behavior...............................................47 Participant peer-relat ed negative behavior...................................................47 Participant peer-related appropriate behaviors.............................................48 Peer procedural integrity data......................................................................49 Observation System and Interobserver Agreement....................................................51 Observation System.............................................................................................51 Interobserver Agreement.....................................................................................54 Design and Data Analysis...........................................................................................56 Design of Procedures...........................................................................................56 Data Analysis.......................................................................................................57 Procedures...................................................................................................................58 Phase One: Descriptive Assessments..................................................................58 Phase Two: Preference Assessment....................................................................60 Phase Three: Functional Analysis.......................................................................61 Training to Peers..................................................................................................62 Free play condition.......................................................................................63 Ignore condition...........................................................................................63 Attention condition.......................................................................................63 Tangible condition........................................................................................64 Escape condition..........................................................................................64 Functional Analysis.............................................................................................65 Free play condition.......................................................................................65 Ignore condition...........................................................................................66 Attention condition.......................................................................................66 Tangible condition........................................................................................67 Escape condition..........................................................................................68 Second Functional Analysis................................................................................69

PAGE 6

vi 3 RESULTS...................................................................................................................72 Overview of Research Aims.......................................................................................72 Phase One: Descriptive Assessment Results..............................................................73 Shane Descriptive Assessment Results...............................................................73 Shannon Descriptive Assessment Results...........................................................74 Colin Descriptive Assessment Results................................................................75 Phase Two: Preference Assessment Results...............................................................77 Shane Preference Assessment Results.................................................................78 Shannon Preference Assessment Results............................................................79 First preference assessment results..............................................................79 Second preference assessment results..........................................................80 Colin Preference Assessment Results.................................................................81 Phase Three: Functional Analysis Results..................................................................82 Shane Functional Analysis Results.....................................................................84 Shane peer-related withdrawn and peer-related negative behaviors............84 Shane combined peer-related w ithdrawn and peer-related negative behaviors...................................................................................................86 Shane peer-integrity.....................................................................................89 Shane peer-related initiations.......................................................................90 Shane peer-related responses.......................................................................91 Summary of Shane’s functional analysis results..........................................92 Shannon Functional Analysis Results.................................................................93 Shannon peer-related withdrawn and peer-related negative behaviors........93 Shannon combined peer-related with drawn and peer-related negative behaviors...................................................................................................95 Shannon peer integrity.................................................................................96 Shannon peer-related initiations...................................................................97 Shannon peer-related responses.................................................................100 Summary of Shannon’s functi onal analysis results....................................101 Colin Functional Analysis Results....................................................................102 Colin peer-related withdrawn and peer-related negative behaviors..................102 Colin combined peer-related withdrawn and peer-related negative behaviors.................................................................................................103 Colin peer integrity.....................................................................................105 Colin peer-related initiations......................................................................107 Colin peer-related responses......................................................................108 Summary of Colin’s functional analysis results.........................................109 Summary of Functional Analysis Results.................................................................110 4 DISCUSSION...........................................................................................................115 Summary of Questions.............................................................................................115 Summary of Findings...............................................................................................116 Identified Functions Findings............................................................................117 Comparison of Functions Findings...................................................................123 Peer Integrity Findings......................................................................................125

PAGE 7

vii Participant Appropriate So cial Behavior Findings............................................126 Limitations of the Investigation................................................................................128 Future Research Directions.......................................................................................131 APPENDIX A REVIEW OF FUNCTIONAL ANALYSIS LITERATURE....................................134 B REVIEW OF SOCIAL SKILLS LITERATURE.....................................................171 C LIST OF PEER-RELATED POSITIVE INITIATIONS OBTAINED DURING DESCRIPTIVE ASSESSMENT (PHASE 1)...........................................................213 LIST OF REFERENCES.................................................................................................217 BIOGRAPHICAL SKETCH...........................................................................................224

PAGE 8

viii LIST OF TABLES Table page 3-1 Social Demands Used in Escape Condition of Functional Analysis........................76 3-2 Tangibles Identified Duri ng Descriptive Assessment..............................................76 3-3 Tangibles Included in Preference Assessment.........................................................77 3-4 Shane’s Rank Order of Stimuli Obtained from Preference Assessment..................79 3-5 Shannon’s Rank Order of Stimuli Obtain ed from First Preference Assessment.....80 3-6 Shannon’s Rank Order of Stimuli Obtain ed from Second Preference Assessment.81 3-7 Colin’s Rank Order of Stimuli Obtained from Preference Assessment...................82 3-8 Shane’s Mean Rate and Ranges pe r minute of Participant Peer-related Withdrawn, Negative, and Combined (Withdrawn and Negative) Behavior...........89 3-9 Shane’s Mean Rate and Ranges per mi nute of Peer Correct Presentation of Social Demands........................................................................................................90 3-10 Shane’s Average Percentage and Rang es of Peer Correct Delivery of Consequences...........................................................................................................90 3-11 Shane’s Mean Rate and Ranges pe r minute Peer Positive Initiations......................90 3-12 Shane’s Mean Rate and Ranges per mi nute of Participant Peer-related Positive and Negative Initiations...........................................................................................91 3-13 Shane’s Average Percentage and Ranges of Participant Peer-related Positive, No, and Negative Responses....................................................................................92 3-14 Shannon’s Mean Rate and Ranges per minute of Participant Peer-related Withdrawn and Negative Behavior..........................................................................95 3-15 Shannon’s Mean Rate and Ranges per mi nute of Peer Correct Presentation of Social Demands........................................................................................................97 3-16 Shannon’s Average Percentage and Ranges of Peer Correct Delivery of Consequences...........................................................................................................99

PAGE 9

ix 3-17 Shannon’s Mean Rate and Ranges pe r minute Peer Positive Initiations..................99 3-18 Shannon’s Mean Rate and Ranges per minute of Participant Peer-related Positive Initiations....................................................................................................99 3-19 Shannon’s Average Percentage and Range s of Participant Peer-related Positive and No Responses..................................................................................................101 3-20 Colin’s Mean Rate and Ranges pe r minute of Participant Peer-related Withdrawn, Negative, and Combined (Withdrawn and Negative) Behavior.........105 3-21 Colin’s Mean Rate and Ranges per minute of Peer Correct Presentation of Social Demands.................................................................................................................107 3-22 Colin’s Average Percentage and Ranges of Peer Correct Delivery of Consequences.........................................................................................................107 3-23 Colin’s Mean Rate and Ranges pe r minute Peer Positive Initiations.....................107 3-24 Colin’s Mean Rate and Ranges per minute of Participant Peer-related Positive Initiations................................................................................................................108 3-25 Colin’s Average Percentage and Ranges of Participant Peer-related Positive and No Responses.........................................................................................................109 3-26 Reinforcers Identified in Functional Analysis.......................................................113 C-1 Shane’s Verbal and Gestural Initiations.................................................................213 C-2 Shannon’s Verbal and Gestural Initiations.............................................................214 C-3 Colin’s Verbal and Gestural Initiations..................................................................216

PAGE 10

x LIST OF FIGURES Figure page 3-1 Rate per minute peer-related withdr awn (top panel), negative (middle panel), and combined social behavior (bottom panel) for Shane.........................................88 3-2 Rate per minute peer-related withdrawn (top panel) and negative social behavior (middle panel) in first functional analys is, and peer-related withdrawn social behavior (bottom panel) of sec ond functional analysis for Shannon.......................98 3-3 Rate per minute peer-related withdr awn (top panel), negative (middle panel), and combined social behavior (bottom panel) for Colin........................................106

PAGE 11

xi Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy EXTENDING FUNCTIONAL ANALYSIS METHODOLOGY TO IDENTIFY THE FUNCTION OF PEER-RELATED WITHDRAWN AND PEER-RELATED NEGATIVE SOCIAL BEHAVIORS FOR YOUNG CHILDREN WITH AUTISM SPECTRUM DISORDERS By Jennifer Anne Sellers May 2006 Chair: Tina Smith-Bonahue Cochair: Maureen Conroy Major Department: Educational Psychology Social skill deficits (e .g., withdrawn behavior) an d excesses (e.g., aggressive behavior) are common behaviors displayed by children with autism spectrum disorders (ASD), which can lead to decreased positiv e interactions with their peers. Several treatments have successfully increased appropriate peer-rela ted social behavior in young children with ASD (e.g., peer-networks); howev er, these treatments have not always produced generalizable results across particip ants, behaviors, or settings. Functional analysis methodology is one stra tegy that has been successful in identifying the functions maintaining a variety of problem behaviors linked to robust treatments, and may show promise for identifying functions of p eer-related problem social behavior. The primary purpose of this study was to evaluate the effectiveness of the functional analysis methodology to identify th e function of peer-rel ated withdrawn and peer-related negative social behaviors in three young children with ASD. An additional

PAGE 12

xii purpose of this investigation was to determine if peers could implement the functional analysis methodology with adequate integrity. Peers were included in the functional anal ysis conditions and provided participants with relevant reinforcement contingent on peer-related wit hdrawn and/or peer-related negative social behavior across all functional analysis conditions. The findings indicated the functional analysis methodol ogy was not successful in identifying a clear function for peer-related withdrawn or peer-related negative social behavior, singly, for any of the 3 participants However, when rate per minute of peerrelated withdrawn and peer-related negative behaviors were combin ed and evaluated post hoc, a negative reinforcement function for 1 of the 3 participants was identified. With the exception of one participant’s peer delivery of consequences in the escape and tangible conditions and another participan t’s peer delivery of positive initiations in the free play condition, peers we re able to implement the f unctional analysis procedures with adequate integrity.

PAGE 13

1 CHAPTER 1 LITERATURE REVIEW Introduction of Autism Spectrum Disorders (ASD) Social skill deficits (e.g. withdrawn behavior) and behavioral excesses (e.g., aggressive behavior) can result in decreas ed positive social interactions between individuals with autism spect rum disorders (ASD) and their peers. The incidence of ASD is on the rise (National Res earch Council, 2001; U.S. Depart ment of Education, Office of Special Education Programs, 2001), and an in creasing number of children with ASD are being educated in general education settings in which social skills deficits and behavioral excesses can impede academic and social succes s. This review addresses these issues by, first, examining ASD. This is followed by a review of the literature related to functional analysis of problem behavior with young child ren with ASD and social skills assessment and treatment with young children with ASD. Categories of ASD Autism spectrum disorder is a term coined by Allen in 1988, to include all types of disorders that are within the framework of pervasive developmental disorders (PDD) (Filipek et al ., 1999). Historically, disorders categorized under ASD have included infantile autism, PDD-residual type, ch ildhood schizophrenia, and autistic psychoses. Currently, the Diagnostic and Statistica l Manual of Mental Disorders Fourth Edition, Text Revision ( DSM-IV-TR ) (American Psychiatric Association [APA], 2000) PDD categories include autistic disorder, Rett's disorder, childhood disintegrative disorder Asperger's disorder, and PDD, not otherwise specified (NOS), including atypical autism. However, AS D is commonly used to refer to autistic

PAGE 14

2 disorder, Asperger's disorder, and PDD-NOS, and will be referred to as such for the purposes of this paper. In this chapter, an overview of ASD will be presented; next how applied behavior analysis has addressed problem behavior in young children with ASD will be discussed; an in depth literature review of the use of functional analysis methodology to assess problem behaviors in young children with ASD is provided, and finally an in depth literature review of pee r-related social skills assessment and treatment of peer-related social skills in young children with ASD is presented. Diagnostic Criteria for ASD Diagnostic criteria for autistic disorder, Asperger's disorder, and PDD-NOS include impairments in social interaction and repeti tive and stereotyped be havior. The level and degree of social impairments differ across different variati ons of ASD. In autistic and Asperger's disorder social impairments are described as deficits in nonverbal behaviors (e.g., eye contact, facial expr essions), developing and maintaining developmentally appropriate peer relations, and spontaneous in itiations (e.g., bringing ite ms of interest to show peers), responses (e.g., joining a play group when asked), or interactions (e.g., playing a game with pe ers) with peers (APA 2000). However, the criteria for social impairments for autistic disorder are identified as deficits in behavior s that display social and emotional indifference, whereas with As perger's disorder, social impairment is defined as eccentric and one-sided (e.g., talk ing to peers almost exclusively about a specific topic of interest only to the child with Asperger’s). Persons with autistic and Asperger's disorders also display restrictiv e, repetitive, and/or stereotyped behavior, interests, and activities. These behaviors ar e defined as motor mannerisms in autistic disorder (e.g., strictly following routines, stereotypic hand-flapping, rocking), and they

PAGE 15

3 are defined as a topic or intere st in gathering facts in Asperg er's disorder (e.g., learning as much as possible about frogs) (APA, 2000). Other differences in diagnostic criteria for autistic and As perger's disorder are age of onset, impairments in communication, and deficits in mental abilities. For the diagnosis of autistic disorder, the onset of symptoms must o ccur prior to age 3-years and there must be a delay in communication (APA, 2000). Both verbal and nonverbal communication abilities are impa ired and verbal communicati on may not develop at all. Communication attempts may be repetitive (e .g., repeating words of a video), language comprehension may be delayed (e.g., difficulty understanding questi ons or directions), and pragmatic aspects of language (e.g., irony, hu mor, social convention) are delayed or absent. In addition to social language dela ys, persons with autism may not develop imaginative or pretend play. There is not a cut off criterion for measures of mental abilities, though most people with autistic disorder also have an associated diagnosis of mental retardation, ranging from mild to pr ofound (APA, 2000). Therefore, the diagnosis of autistic disorder can be made with or without an associated diagnosis of mental retardation. Similarly, PDD-NOS diagnostic cr iterion requires a delay in communication skills; however, this diagnosis does not include criteria for mental ability or age of onset. In contrast to both autistic disorder and PDD-NOS, the diagno stic criterian of Asperger's disorder does not allow for either a delay in communication or mental retardation. In summary, children with ASD demonstrat e a number of diverse characteristics. They may have impairments in social inte ractions, verbal and nonverbal communication, and restrictive, repetitive, and/or stereotype d behavior, interests, and activities. Their impairments in social interactions and comm unication often lead to failure in developing

PAGE 16

4 age-appropriate peer relationships, which has many negative future implications for these children (Filipek et al., 1999). Prevalence of ASD Early identification of children with AS D has become increasingly critical as children who receive two or more years of intensive early education intervention are more likely to have greater developmental gains in speech and cognition than children who do not receive those services (Filipek et al., 1999). The overall prevalence of ASD is estimated at 10 to 20 per 10,000 people. There ar e estimates indicating that in the U.S. there are 60,000 to 115,000 children under 15-year s-old diagnosed with ASD (Filipek et al., 1999). More males than females are diagnos ed with ASD, though the ratio varies with the level of cognitive impairment. There is a ma le to female ratio of 2:1 for children with ASD who also have mental re tardation and 4:1 for children with ASD without mental retardation (Filipek et al., 1999). Therefore, th e early identification of the characteristics of ASD is critical to children receiving early education intervention to address behavioral excesses and deficits. Identification and Characteri stics of Children with ASD Children with ASD have a variety of behavi oral and social deficits. For example, young children with ASD often do not “cuddle” with their caregi vers, bring toys to show caregivers or peers, and may have temper ta ntrums when daily routines are changed (e.g., going to bed without reading a specific book) (Simpson & Zionts, 2000). In addition, they frequently engage in chronic problem behaviors that impa ct their learning and integration into society. Problem behavior Children with ASD may exhibit behavior problems such as a high level of motor activity, a ggression (e.g., hitting, kicking) self-injurious behavior

PAGE 17

5 (SIB) (e.g., head banging, hand biting), temp er tantrums, stereotypy (e.g., hand-flapping, spinning), and intense restrict ed patterns of interest/b ehavior (e.g., lining up objects, playing with only one toy) (APA, 2000; Simp son & Zionts, 2000). Problem behaviors are widely reported in the clinical and resear ch literature (Fisher, Ninness, Piazza, & OwenDeSchryver, 1996; Richman, Wacker, & Winborn, 2001). Social withdrawal and de ficits in communication Social withdrawal and a lack of communication, both verbal and nonverbal, are also charac teristics of children with ASD (APA, 2000). Children with ASD may not ma ke eye contact with adults or peers and may withdraw from peers in order to enga ge in solitary play. These children may not spontaneously initiate to peers (e.g., bring an item of interest to show a peer), may not respond to their name or other peer attempts to gain their attention (e.g., tapping shoulder, asking the child to play with them), and if the child does play with a peer, the play may be stereotypic (e.g., lining up t oys) rather than interactive (e.g., building a tower with Lego’s) or imaginative play (e.g., play ing pretend school) (APA, 2000; Simpson & Zionts, 2000). For example, McGee, Feldma n, and Morrier (1997) compared the social behavior of young children with ASD a nd young typically developing children, all attending an integrated preschool. The results suggest specific peer-r elated social skill deficits in young child ren with ASD in comparison to th eir typically developing peers, including deficits in initiations responses, and staying in prox imity (i.e., 3 ft) of peers. In addition, the results of this study suggest that these social skill de ficits do not improve with maturation over one year or with placement in an integrated preschool without treatments focused on their social behavior.

PAGE 18

6 In summary, children with ASD demonstr ate deficits in the areas of problem behaviors, social skills, a nd communication. Therefore, the early identification of ASD and development of treatments that can addre ss these children’s deficits is critical. Several types of treatments have been de veloped with the goa l of increasing the appropriate peer-related so cial behavior of young children with ASD (e.g., Laushey & Heflin, 2000; Pierce & Schreibman, 1997b). Howeve r, to date treatments to increase the social behavior of young children with ASD have primarily been based on a priori decisions (Brown, Odom, & Holcombe, 1996; Spence, 2003) rather than stemming from individual assessments. In the following sect ion, one approach to treatment, applied behavior analysis (ABA), that has been highly successful for use with children with ASD will be reviewed. Applied Behavior Analysis (ABA) The field of ABA includes several stra tegies to individually assess problem behaviors and to develop assessment-based tr eatments. Applied beha vior analysis has been demonstrated to be a successful tr eatment approach in assessing and treating problem behavior (Asmus, Ringdahl, Sellers, Call, Andelman, & Wacker, 2004; Iwata et al., 1994; Wacker, Berg, Harding, Derby, As mus, & Healy, 1998). Baer, Wolf, and Risley (1968), first, described ABA as an a pproach to treatment that focuses on socially important behaviors and incl udes reliable measurement and procedures that can differentiate variables that c ontrol behavior change, and can be replicated. Since that time, considerable research has been conduc ted on the effectiveness of various ABA strategies to decrease problem behavior (e.g., aggression, sel f-injury, destruction, stereotypy) and increase alte rnative appropriate behavior s (e.g., communication, social behaviors).

PAGE 19

7 Experimental analyses of consequences (functional analysis) and antecedents (structural analysis) are two common types of strategies used in the field of ABA to individually assess problem behaviors and to develop assessment-based treatments. Both functional and structural analyses lead to the identification of functi onal relationships that exist between environmental events (e.g., leve ls of attention provi ded, availability of preferred toys) and targ et behaviors (e.g., problem behavior peer-related problem social behavior) (Axelrod, 1987; Carr & Durand, 1985; Fisher et al., 1996; Peck, Sasso, & Jolivette, 1997). For example, Richman et al. (2001) conducted a functional analysis on the aggressive behavior of a 3-year-old ch ild diagnosed with ASD. The findings of the functional analysis indicated the participant engaged in aggressive behavior to gain access to tangible items. Based on this inform ation, a treatment was developed that taught the participant an appropriate alternative be havior to gain access to tangibles (i.e., handing a card or signing please to request the tangible item). This assessment-based treatment successfully reduced the aggressive behavior of the young child with ASD while simultaneously increasing his appropriate communication. Peck et al. (1997) conducted a hypothesis driven structural an alysis on the peerrelated problem social behavior of five child ren with disabilities (two with ASD) between the ages of 9 and 11-years. Based on the indi vidual findings of the structural analysis, two peer-mediated treatment protocols were developed and presented in an alternating treatment design. One treatment included the antecedents identified in the structural analysis that resulted in the hi ghest percentage of peer intera ctions or the highest rate of on-task behavior and lowest rate of problem behavior. In contrast, the second treatment protocol included antecedents identified in the structural analysis that resulted in low

PAGE 20

8 percentages of peer-related interactions or the lowest rate of on-task behavior and highest rate of problem behavior. The structural an alysis assessment in this investigation was successful in helping to develop an effective treatment that resulted in a higher rate of ontask behavior, lower rate of problem behavi or, and higher percenta ge of peer-related interactions in comparison to a treatment th at was not developed based on the findings of the hypothesis driven structural analysis. Summary of ASD In summary, children with ASD demonstrate deficits in social skills and excesses in problem behavior. Research has demonstrat ed the potential negative effects social behavior deficits can have on individuals with ASD from preschool through adulthood. However, even with and wit hout intensive treatments, gain s in peer-related social behaviors remains problematic for this popula tion. Applied behavior analysis appears to be a promising approach for developing indi vidual treatments to promote peer-related social behaviors in children with ASD. Th e remainder of this chapter will review the literature in two areas. First, assessment and treatment of problem behaviors will be discussed. Next, a review of the social skills treatment literature will be presented. Assessment and Treatment of Problem Be havior in Young Children with ASD Introduction of Assessment Methodology One of the primary approaches for am eliorating problem behaviors in young children with ASD is through experimental an alysis. Experimental analysis methods are based on Carr's (1977) theoretical model in which operant mechanisms are hypothesized to maintain problem behavior. Carr hypothesi zed that different i ndividuals engage in similar topographies of problem behavior for different environmental reasons. Some common topographies of problem behavior in clude aggression, destruction, SIB, and

PAGE 21

9 disruption. Carr proposed thre e potential operant mechanisms underlying the motivation for problem behavior: positive reinforcement, negative reinforcement, and sensory reinforcement. Carr's proposal that problem beha vior can have an operant basis has led to substantial research and the subsequent deve lopment of assessment techniques to identify the operant mechanisms underlying problem behaviors. Axelrod (1987) discussed tw o types of experimental an alyses: (1) manipulation of antecedent variables, and (2) manipulation of consequence variables. Axelrod defined a functional analysis as an assessm ent of factors that reinforce the occurrence of a behavior (e.g., escape from demands maintains SIB). Axel rod defined a structur al analysis as an assessment of antecedent variables that occasion the behavior (e.g., demands evoke SIB). Both functional and structural analyses ma y lead to the identif ication of functional relationships that exist between these envir onmental events and target behaviors. The influence of consequence events on problema tic behavior has been studied extensively via functional analyses, whereas the influence of antecedent events has been studied via structural analyses. Functional Analysis In ABA, the term “function” refers to th e purpose the behavior serves (i.e., the reinforcer maintaining the behavior) for the individual (Carr, 1977). Carr identified several reinforcers that maintain proble m behavior: positive (e.g., obtaining attention following problem behavior), negative (e.g., escaping demand following problem behavior), and self-stimulato ry (or automatic) reinforcemen t (i.e., problem behavior not maintained by social-mediation). In a seminal study based on Carr’s work, Iwata, Dorsey, Slifer, Bauman, and Richman (1982/1994) developed the functional analysis methodology to identify the

PAGE 22

10 function of SIB by experimentally evalua ting the consequences maintaining that behavior. Iwata et al. (1982/1994) de veloped the first operant methodology for identifying functional relationships between specific environmental variables and SIB. Nine adults with developmental disabilitie s who engaged in high levels of SIB participated in the study. Self-injurious behavior was measur ed across analogue conditions in which consequence events were manipulated. The analogue condition that resulted in the highest percent of SIB identified the class of consequences (positive or negative reinforcement) that served to maintain problem behavior. Results indicated that specific environmental variables were identified that maintained the occurren ce of SIB for 6 of 9 individuals. Iwata et al. (1982/1994) indi vidually assessed particip ant functions of SIB by experimentally evaluating four distinct environmental conditions : social disapproval, academic demand, free play, and alone. A free play condition was also evaluated as a control condition. In the social disapproval condition, the part icipant had access to tangibles and was told to play with the toys because the experimenter had to do some work. The participant's engagement in SIB resulted in statements of concern and disapproval along with non-punitive physical contact. An incr eased percent of SIB in this condition indicated the participant's SIB was maintained by positive reinforcement in the form of attention. In the academic demand condition, partic ipants sat at a table next to the experimenter and were presente d with difficult tasks (i.e., ta sks that were not completed

PAGE 23

11 spontaneously) using a three prompt sequence. The demand was first presented verbally. If the demand was not completed after 5 sec onds, the experimenter repeated the verbal demand and modeled how to complete the demand. If the demand was not completed after another 5 seconds, the experimenter repeated the verbal demand and physically guided the participant to complete the demand using the least amount of physical contact necessary. Praise was provided for completion of the demand, regard less of modeling or physical guidance. The next demand was then presented. If the participant engaged in SIB at any point during the demand, the experi menter ended the demand and turned away from the participant for 30 seconds before presenting the next demand. Repeated SIB resulted in an additional 30 second delay to presentation of the next demand. An increased percent of SIB in this condition indicated that the participant's SIB was maintained by negative reinforcement. In the free play condition the individual had access to tangibles and no demands were presented. Attention (soc ial praise, physical contact) was provided noncontingently for appropriate behavior (absence of SIB) at least every 30 seconds. Attention for SIB was not provided. If SIB were socially maintained, the lowest percent of SIB would be observed in the control condition. In the alone condition, the participant was in a room alone, and without any materials. An increased percent of SIB in this condition indicated the participant's SIB was maintained by self-stimulation. The length of assessment averaged 8 days across participants. Ei ght sessions were conducted per day and conditions lasted 15 minutes each. The order of conditions was predetermined by random drawings. Data we re collected using continuous 10 second

PAGE 24

12 interval recordings. A multielement design was used to analyze the data. Data collection continued until the data were stable, or the da ta were unstable in all conditions for 5 days, or until sessions had continued for 12 days. Results indicated that 4 of th e 9 participants displayed th e highest percent of SIB in the alone condition. Two partic ipants displayed the highest percent of SIB in the academic demand condition, one displayed the highest percent of SIB during the social disapproval condition, and two di splayed undifferentiated patterns of SIB. Six of the 9 participants consistently displayed a highe r percent of SIB in one specific condition, indicating the participants' f unction of SIB varied both betw een and within individuals, similar to Carr's (1977) proposal. That is, the topography of the beha vior (i.e., SIB) did not match the function of the behavior. Th is led to the conclusion that conducting individual assessments to determine the func tion of SIB would lead to more effective treatments that are based on the variables that are maintaining or reinforcing problem behavior rather than the t opography of the behavior. The majority of research studies published in the literature have used the functional analysis methodology developed by Iwata et al (1982/1994) to determine the maintaining contingencies of problem behavior; that is they have focused on the manipulation of consequence events during assessment. The result s of Iwata et al. have been consistently replicated by different clinical research teams and across different topographies of problem behavior (McComas, Hoch, Paone, & El-Roy, 2000; Steege, Wacker, Berg, Cigrand, & Cooper, 1989). In every case, th e functional analysis methodology was used to identify the effects of e nvironmental variables on the o ccurrence of problem behavior. The functional analysis methodology has been demonstrated to be successful in

PAGE 25

13 identifying the function for a variety of topogr aphies of behavior (e.g., SIB, aggression, destruction, stereotypy) and across a range of settings (e.g., home, school, clinic) (Hanley, Iwata, & McCord, 2003). Several epidemiological studies have also demonstrated the success of functional analysis in identifying reinfo rcers for a variety of problem behaviors of adults and children w ith and without devel opmental disabilities, and the successful reduction of problem behavi or utilizing assessment-based treatments (Asmus et al., 2004; Hagopian, Fisher, Sulliva n, Acquisto, & LeBlanc, 1998; Iwata et al., 1994; Wacker et al., 1998). A lthough extensive replication has occurred, one behavior that has not been investigat ed using functional analysis methodology is peer-related social behavior. Structural Analysis The results of a functional analysis ar e restricted to onl y one category of environmental variables: consequences. Other environmental variables that occur prior to or concurrently with behavior may also infl uence the occurrence of behavior due to their previous association with reinforcement. A variety of contextual factors have been assessed to evaluate their effect on problem behavior (e.g., off-task behavior, aggression) and peer-related social behavior (e.g., aggression toward peer s, negative vocalizations). To study these variables, a slightly differe nt methodology is used, referred to as a structural analysis. A structur al analysis is an assessment procedure that identifies the antecedent variables that set the occasion fo r problematic behavior; for example, demand situations may result in an increased likelihood of SIB (Axelrod, 1987; Carr & Durand, 1985). Thus, the purpose of a structural analysis is to more precisely identify antecedent events that reliably occasi on the target behavior.

PAGE 26

14 Carr and Durand (1985) first developed the structural analysis methodology, which was conducted with four children with AS D (ages 7 to 14-years). This assessment methodology differed from Iwata et al. (1982/1 994) in that behaviors were shown to correlate with changes in antecedent condi tions (high versus low demands and high versus low social attention) rather than w ith contingent access to consequences (escape and social attention). Despite this differe nce, the results were equally positive in identifying a specific operant basis for each participant's problem behavior. Peck et al. (1997) extended the stru ctural analysis methodology to include assessment of antecedents to examine the antecedent variables that occasion appropriate social behavior. Peck et al conducted hypothesis driven st ructural analyses with 5 participants in elementary school (ages 9 to 11-years) with developmental disabilities (two with ASD), mild to moderate mental retardation, and peer-rel ated negative social behavior. Hypothesis driven structural analys es were conducted with each participant to identify environmental factor s that increased the likelihood that appropriat e peer-related social behavior would occur. In this investigation, the special educ ation teachers and researchers developed hypotheses regarding the antecedents likel y to increase the likelihood of each participant’s appropriate and ne gative peer-related social be havior. There we re a total of seven antecedent events identified: noise le vel, peer group size, amount of interaction required for the task, task structure, gender of peer, level of so cial interaction, and transition warnings. The participants were in two self-contain ed classrooms for th e majority of the school day. Four typically de veloping peers that were identified by their general

PAGE 27

15 education teachers as having exemplary soci al skills and were the same age as the participants participated in the study. Each peer participated in the analyses of 2 participants. Individualized hypothesis driven structur al analyses were conducted in the participants' self-contained special educa tion classrooms during naturally occurring classroom activities. The hypothesized “most ef fective” and “least e ffective” antecedents identified were presented in 10 minute condi tions using an alternating treatment design (e.g., alternately presenting high and low noi se conditions). A same-aged typically developing peer was trained by the researcher s and participated in the conditions with each participant. Data were collected usi ng 6 second partial interval recording on the participants’ initiations (e.g., verbal or gestural behavior to peer to indicate communication when within at least 1 meter of a peer and physically oriented to peer), responses (verbal or gestural re sponse to peer within 3 seconds of that peer’s initiation), and continuations (third beha vior in initiation, response w ithin 3 seconds, continuation within 3 second sequence). The participants’ responses and continua tions were calculated as proportions of opportunities (i.e., the number of participant respons es or continuations divided by the number of peer initiations mu ltiplied by 100%). In addition, data were collected on the participants’ on-task and off-ta sk (not oriented toward or engaged with activity or peer) behavior, negative behavi or (e.g., aggression, swearing, SIB, escape [attempts to leave activity or area without permission]) and purposeful (rather than offtask) noncompliance. Peer initiations, respons es, and continuations we re scored using the same definitions as the participant. In additi on, peer behaviors were scored as cooperative

PAGE 28

16 (conversation or interaction that was not about playing the game) or instructional (conversation or interaction with or about the assigned task). The data were analyzed by combining the percentage of the participants' initiations and proportion of responses and continuations into one social inte raction graph and the percentage of participants' on-task beha vior on another graph. The “most effective” condition was identified using visual analysis, fi rst evaluating only social interactive data. If there was not a trend in the social intera ctive data, the “most effective” condition was identified based on the condition with the hi ghest percent of on-task behavior and the lowest percent of problem behavior. Individual treatments were develope d based on the assessment findings. The hypothesized “most effective” (Treatment A) condition was the combination of antecedents assessed in the structural analysis that resulted in highest level of on-task and effective communication. For example if high a nd low structure, interactive and social behavior were individually assessed and high structure, low interaction, and high socialization all resulted in the highest le vel of on-task and eff ective communication, all of those antecedents (i.e., hi gh structure, low interaction, and high socialization) were combined into one condition, i.e., Treatme nt A. Conversely, Treatment B was the combination of antecedents when the participan t displayed the most off-task behavior and problem behavior. Therefore, using the example above, Treatment B would be a combination of low structure, high interac tion, and low socialization. Antecedents that did not result in substantial or any differen ces in behavior were not included in the treatment evaluation (i.e., not in Treatment A nor Treatment B). The treatment conditions were conducted for 10 minutes during naturally occurring activities and presented in an

PAGE 29

17 alternating treatment design us ing a reversal sequence. Trea tment A was presented first and continued until a trend in on and offtask behavior and problem behavior was established. Treatment B was then conducted until a trend was established. Finally, Treatment A was repeated until a stable trend was again established. The results indicated that all 5 participan ts demonstrated differential appropriate on-task behavior in at leas t one antecedent condition. Th ree of the 5 participants demonstrated inconclusive patterns of initia tions, responses, and continuations across the antecedents and of the two who displayed differential behavior, only one displayed differential behavior for all three social inte ractive behaviors (i.e., initiations, responses, and continuations). Treatments that were based on the combination of antecedent variables that resulted in the most on-task behavior, least problem behavior, and for two of the participants, also the most social in itiation, response, and/or continuation behavior, resulted in higher levels of on-task and lo wer levels of problem behavior than the treatment based on the antecedent conditions with the most off-task and problem behavior. All Treatment A interventions succe ssfully increased on-task behavior, social interactive behavior, and decreased problem be havior (except for one participant who did not display problem behavior during treatmen t) in comparison to Treatment B. These findings suggest hypothesis driven structural analyses led to the development of successful peer-mediated treatments of peer-related social problems. Summary of Assessment and Treatment of Problem Behaviors Functional analysis is a strategy to use to assess consequences (i.e., reinforcers) maintaining target behavior, and structural analysis is a strategy to use to assess antecedent events preceding or co-occurri ng target behavior (Axelrod, 1987). Both experimental analyses of behavior have pr oven successful in identif ying the function of a

PAGE 30

18 variety of problem behaviors (e.g., aggression, destruction) and development of effective treatments (Carr & Durand, 1985; Iwata et al ., 1994). Treatments th at are based on the function of problem behavior are more effec tive than treatments that are not function based (Axelrod, 1987). To date, a functional an alysis of peer-related problem social behavior has not been conducted. Peck et al. (1997) conducted the first hypothesis driven structural analysis of peer-related problem social behavior. Given the success of functional analysis methodology to identify the functions of problem behavior and develop effective treatments, an extension of the functional anal ysis to include the assessment of peer-related problem social behavior appears to be a logical next step. Review of the Literature The purpose of this section is to provide a review of the literature on the use of experimental analyses to assess and treat problem behaviors in young children with ASD and to review the literature on soci al skills assessment and treatment. Two bodies of literature will be reviewed in order to demonstrate their contribution to the current study: (1) litera ture demonstrating th e effectiveness of functional analysis in identifying the function of problem behavior for young children with ASD, and (2) literature addressing assessment and treatment of social skill defi cits in young children with ASD. Review of Functional Analysis Literature Method to select reviewed studies. Appropriate articles for this literature review were identified by searching the following computerized databases: Academic Search Premier, Journal of Applied Behavior Analys is, and PsycINFO. Keywords used to search the databases included, but are not limited to functional analysis, autism, children, young, early, intervention, and behavior.

PAGE 31

19 The functional analysis body of literature me t the following criteria for inclusion in the review: (a) peer-reviewed journal, (b) functional analys is conducted (including brief functional analysis) using a single-subject design includ ing a variety of designs (including reversal, pair-wise, alternating treatm ent designs), (c) at least two analogue conditions conducted, (d) procedures for each analogue condition described, (e) individual results for particip ants presented in a line gr aph format, (f) at least one participant with ASD 8-years of age or younger and/or at le ast one participant with a diagnosis of developmental delays and char acteristics of ASD (e .g., social withdrawal, authors indicated the participant had autis tic like behaviors an d/or indicated the participant was suspected of having autism). Review of Functional Analysis Studies This section of the review demonstrates the utility of the functional analysis methodology in identifying the function of problem behavior for young children with ASD. A functional analysis is an experime ntal procedure designed to evaluate the environmental variables that are maintaining a target behavior(s). A total of 32 studies met the criteria mentioned above. The f unctional analysis methodology according to Iwata et al. (1982/1994) is rela tively new (i.e., since 1982); th erefore no studies prior to 1982 were included. Publication dates of the st udies included in the review range from 1982 to 2005. One noteworthy finding to mention is there has been an increase in studies that include a functional anal ysis with young children with ASD; only 9 studies were conducted between 1982 and 1999, whereas 20 studies were conducted between 2000 and 2004.

PAGE 32

20 Participants included in review of literature A total of 47 participants were includ ed across the 32 studies identified. These participants are described in Appendix A (Rev iew of Functional Analysis Literature). In addition, the number of partic ipants who met criteria in each study, age, gender, and disability are included. Age. All participants included in these studi es were between the ages of 3 and 8years-old. Forty-five percent (21) of the pa rticipants were between 5 and 6-years-old, 34% (16) were between 7 and 8-years-old, a nd 17% (8) were between 3 and 4-years-old. These numbers are not mutually exclusive, as two studies with more than one participant included one participant between the ages of 3 and 8-years-old, however, their exact age was not indicated. Number The majority (63%) of the 32 studies had only one participant that met the criteria for inclusion. Twenty-five percent (8) of the studies had two participants, 6% (2) had three participants, and one (3%) included four participants. Gender. Of the 47 participants included in this review, 85% (40) were male and 15% (7) were female. Disability. All 47 participants had either a di agnosis of PDD-NOS, autism, or if under the age of 6 years, char acteristics of ASD. Under th e category of ASD, 83% (39) had a diagnosis of autism, 13% (6) PDD-NOS, 2% (1) ASD, and 2% (1) was characterized as having autis tic-like behavior. Other di agnoses included mental retardation (level unspecified) (4%), mild me ntal retardation (6%), moderate to severe mental retardation (45%), and profound ment al retardation (6%). For 36% (17) of participants no level of mental retardati on was indicated. In addition, 6% (3) of the participants had a diagnosis of developmental delays, 6% (3) a diagnosis of attention

PAGE 33

21 deficit hyperactivity disorder (ADHD), 4% (2 ) oppositional defiant disorder (ODD), 4% (2) seizure disorder, and 2% (1) with a diagnosis of Fragile X syndrome. Therapists in functional analysis Of the 32 studies reviewed, 63% (20) of th e therapists were a staff member or the experimenter, 19% (6) were participant car egivers, and 13% (4) were participant teachers. Twenty-two percent (7) of the studies did not explicitly report who served as the therapist. The following are examples of how investigators have employed a variety of therapists when conducting functional analys es. Sasso et al. (1992) first conducted a functional analysis with experimenters serv ing as therapists a nd then, after providing training to the teacher, the teacher conducte d a functional analysis in the classroom, serving as the therapist. In another functi onal analysis study of food refusal, a parent served as the therapist throughout all functi onal analysis conditions (Najdowski, Wallace, Doney, & Ghezzi, 2003). English and Anderson (2004) manipulated who served as the therapists (experimenter or parent) while conducting a functional analysis. When the parent was the therapist, the functional analysis identified an attention and tangible function, however, when the experimenter served as the therapist, th e functional analysis only identified an attention func tion. Therefore, the results of the functional analysis were similar across therapists in identifying an attention function, though differed across therapists in identifying a tangible func tion. The findings of this study suggest the identity of the therapist influenced the findings of the functional analysis. Settings of functional analysis The settings in which the functional analys is procedures were implemented varied across the 32 studies, though the majority (56% ) were conducted in a session room while the participant was inpatient or attended an outpatient clinic. Twenty-two percent (7)

PAGE 34

22 were conducted in the partic ipant’s classroom, 16% (5) in an unused room at the participant’s school, 16% (5) in the home, and 3% (1) was conducted in the classroom on an inpatient unit in a hospital setting. In st udies with more than one participant, there were some instances of functional analyses be ing conducted in different settings (i.e., one participant in the home and the other at schoo l). When this occurred, both settings were recorded; therefore, the categories used to re port settings will exceed 100% when totaled. Dependent measures There were several behaviors measured in the functional analys es conducted across the 47 participants. The categories of behavi ors reported are not mutually exclusive and therefore, will not add up to 100% when tota led. In addition, some researchers measured one dependent variable (e.g., destructi on) but included seve ral behaviors (e.g., noncompliance, screaming, tantrums) within th e operational definition of destruction. As a result, the following should not be considered a comprehensives list of all behaviors the participants in these studies displayed. Thirty-six percent (17) participants engaged in aggression, 23% (11) in disruption, 19% (9 ) SIB, 11% (5) inappropriate mealtime behavior, 9% (4) destruction, 9% (4) stereot ypy, and 4% (2) of the participants engaged in screaming behavior. One participant enga ged in each of the following: tantrums, high frequency vocalizations, pica, saliva pla y, food refusal, spitting, elopement, hair manipulation, inappropriate language, and grabbing. Overall, 89% (42) of the participants’ dependent measures included one or more of the following behaviors: aggression, disruption, destruc tion, SIB, or tantrums. Experimental conditions The experimental conditions included in Iwata et al. (1982/19 94) included social disapproval (e.g., attention), demand, play, a nd alone. Out of the 32 studies reviewed,

PAGE 35

23 87% included each of those conditions (i.e., attention, demand, play, and alone/ignore). Ninety-seven percent included attention, demand and play conditions, 92% included these conditions with the addition of the ta ngible condition (i.e., at tention, demand, play, and tangible), and 85% with the addition of alone/ignore conditi ons (i.e., attention, demand, play, tangible, and alone/ignore). The demand condition was conducted in 100% of the studies reviewed. The attention condition was conducte d in 97% of the studies, th e play condition in 94%, tangible in 78%, and the alone/ignore condi tion was conducted in 56% of the studies evaluated. The inclusion of the tangible c ondition and the absence of an alone/ignore condition are the main differences of many of the functional analyses conducted in this review and the functional analysis procedur es originally described in Iwata et al. (1982/1994). Experimental condition modifications Variations of experimental conditions were occasionally conducted in attempts to clarify undifferentiated or inconclusi ve functional analysis findings. Behaviors reinforced. Asmus, Franzese, Conroy, and Dozier (2003) varied the behaviors reinforced (e.g., reinforcing stereotypy and/ or disruption) following a functional analysis with undiffere ntiated findings. These varia tions resulted in identifying a tangible function for disruption (when only di sruption was reinforced) and an automatic function for stereotypy (when both stereot ypy and disruption were reinforced). Thompson, Fisher, Piazza, and Kuhn (1998) va ried reinforcement of specific forms of the same behavior (i.e., aggression) follo wing an inconclusive functional analysis in which all forms of aggression were reinfo rced. Based on descriptive assessments, Thompson and colleagues hypothesized one form of aggression (i.e., chin grinding) was

PAGE 36

24 not maintained by social atte ntion, while all other forms of aggression were maintained by social attention. This hypot hesis was tested by conductin g two attention conditions, one in which all topographies of aggression we re reinforced with attention and one in which all topographies of aggression except ch in grinding were reinforced with attention. The findings confirmed the hypothesis generated from the descriptive assessments, other forms of aggression (all except chin grinding ) were maintained by social attention and chin grinding was not sensitive to atte ntion. This finding was significant as it demonstrated the potential for different fo rms of the same category of behavior (e.g., aggression) to be maintained by different re inforcers. In addition, it demonstrated the potential of aggression being maintained by a sensory function, a behavior not often associated with automatic reinforcement. Type of escape delivered. Following an inconclusive functional analysis, Hagopian, Wilson, and Wilder (2001) conduc ted a modified escape condition, escape from attention. Instead of providing the participant with a demand, attention was provided, and contingent on problem behavior attention (rather than the demand) was briefly removed. This modification resulted in conclusive functional analysis findings, indicating both an escape from attention and tangible function. In addition to modifying the escape condition to include escape from a ttention, rather than escape from demands, additional modifications incl uded escape from food (Levin & Carr, 2001; Piazza et al., 2003) and escape from noise (Tang, Kennedy, Koppekin, & Caruso, 2002). Type of attention delivered To identify specific variab les of social attention maintaining problem behavior, Fisher et al (1996) evaluated the type of attention delivered (verbal reprimand vers us unrelated verbal statement) contingent on destructive

PAGE 37

25 behavior. The findings indicated destructive be havior was maintained by attention in the form of verbal reprimands and not by attenti on in the form of unrelated statements. Along the same line, Hagopian, Contrucci Kuhn, Long, and Rush (2005) evaluated two types of attention (physical and verbal) in the maintenance of aggre ssive behavior and found that both forms of attention maintain ed the aggressive behavior. Type of tangible delivered Graff, Lineman, Libby, and Ahearn (1999) evaluated the type of tangible provided (i.e., toy ve rsus edible) contingent on screaming. In tangible-toy condition, screaming resulted in brief access to a toy and in the tangibleedible condition, screaming resu lted in a small edible. The fi ndings of this functional analysis were inconclusive. Treatment of problem behavior Twenty-five percent (8) of the studies reviewed conducted only a functional analysis and did not continue onto treatment. Thirteen percent (4 ) conducted a functional analysis and then continued with further assessment such as antecedent, preference, competing items, or choice assessments. The remaining 63% (20) conducted treatment within the study. Design of functional analysis The functional analysis procedures desc ribed by Iwata et al. (1982/1994) was an extended functional analysis conducted utili zing a multielement design. Of the 32 studies reviewed, 88% (28) utilized a multielement design, 13% (4) a multielement design with a brief functional analysis, 13% (4 ) reversal, and 3% (1) utilized a pair-wise (test-control) design. These categories are not mutually excl usive, therefore the percentages will not add up to 100% when totaled. The following di scussion illustrates how investigators have used several designs to identify functions of problem behavior. For example, Marcus,

PAGE 38

26 Vollmer, Swanson, Roane, and Ringdahl (2001) demonstrated a procedure to use the least amount of time necessary to conduct a functional analysis. Ma rcus and colleagues conducted brief functional analyses with 3 par ticipants. If a function was not identified in the brief functional analysis an extended functional anal ysis was conducted. The brief functional analysis was successful in identifyi ng functions for 2 of the 3 participants and the extended multielement functional analysis successfully identified the function of aggression for the third participant. Thes e findings suggest an extended functional analysis may not always be necessary to id entify the maintaining variables of aggression. Along the same line, Brown et al. (2000) conducted a brief multielement functional analysis with rapidly changing conditions with one participant and an extended multielement functional analysis was a s econd participant. For both participants, functions of their problem behaviors were identified. This was further confirmed when Finkel, Derby, Weber, and McLaughlin (2003) conducted a brief functional analysis of elopement and disruption in the classroom with the teacher serving as the therapist. The brief functional analysis met the time limita tions required in the natural setting and successfully identified an escape function. One exception is a brief functional analysis of saliva play conducted by Luisel li, Ricciardi, Schmidt, and Ta rr (2004) in the participant’s classroom. The findings of the brief functi onal analysis were undifferentiated, further supporting Marcus et al. (2001) fi ndings that brief functional analysis may not always be sufficient for identifying maintaining variables of target behavior for all participants. In addition to extended multielement and brief f unctional analysis designs, reversal designs were also employed to identify f unctions of problem behavior.

PAGE 39

27 Richman, Wacker, Asmus, Casey, and Ande lman (1999) varied topographies of behavior reinforced to evaluate a potential hierarchy of behaviors. A brief functional analysis was conducted using a reversal desi gn to evaluate the topography of behavior reinforced. A reversal design was also employe d when evaluating the effect of therapists (experimenter and parent) on the findings of an extended multielement functional analysis (English & Anderson, 2004). Piazza et al (2003) conducted a functional analysis of inappropriate mealtime behavior util izing a pair-wise comp arison (test condition compared to control condition) in a reversal design. Additional variations include following an extended multielement functional analysis with a series of consecutive ignor e conditions (Sidener, Carr, & Firth, 2005) and only including two experimental conditions (escape preferred food and escape nonpreferred food) (Levin & Carr, 2001) in a reversal design. Findings of functional analysis The functional analysis methodology succe ssfully identified the maintaining variable of the target behavior(s) for 92% ( 43) of the 47 participants. The majority (74%) of the participant’s target be haviors were maintained by a single function. When only one maintaining variable was identified, 28% ( 13) were maintained by an escape function, 13% (6) attention, 13% (6) tangible, and 13% (6) an automatic function. Thirty-two percent (15) of participants’ target behaviors were maintained by multiple functions. When multiple maintaining variables were identified, 17% (8) were maintained by escape and tangible functions, 9% (4) by attention and tangible, 2% (1) atte ntion and escape, 2% (1) tangible and escape from attention, a nd 2% (1) escape, attention, and tangible functions. The functional analysis methodology did not successfully identify maintaining variables for 8% (4) participants. Four percen t (2) of the participan ts’ functional analysis

PAGE 40

28 findings were undifferentiated and 4% (2) we re inconclusive. The categories are not mutually exclusive and therefore will not add up to 100%. For example, one participant had two target behaviors, each behavior with a separate, single function. Each function was categorized separately, as a single functi on, not as multiply maintained. A participant was indicated as having multiply maintained f unctions if the same target behavior(s) had multiple functions. In addition, some studies conducted multiple functional analyses in which more than one finding was indicated from each analysis. Rather than choosing between which findings to report, both were included. Summary of functional analysis review of literature The review of these 32 studies has demonstr ated the effectiveness of the functional analysis methodology in identifying the vari ables maintaining a variety of problem behaviors displayed by young children with ASD. To date, a functional analysis of peerrelated social behavior, partic ularly peer-related withdraw n and peer-related negative social behavior has not been conduc ted with young children with ASD. Strengths The functional analysis literatur e has demonstrated good success in identifying the functions of a variety of problem behaviors (e.g., SIB, aggression, destruction, disruption, noncom pliance, food refusal), across a variety of settings (schools, homes, inpatient units, outpatient clinics), and therapists (teachers, parents, experimenters) with young children with AS D. The increasing trend in the number of studies conducted in the past four years is promising to th e continuing dem onstration of the effectiveness of functional analysis methodology in the assessment of problem behavior in young children with ASD. Despite the increase in the number of studies conducted in the past four years, the vast majo rity (88%) only included 1 or 2 participants

PAGE 41

29 per study, which does not provide a large partic ipant pool from which to base the studies findings. The functional analysis methodology has b een implemented across a variety of therapists, though the majority (63%) were sta ff or experimenters and the majority (56%) of studies were conducted in a session room located in an inpatient or outpatient clinic. Additional studies utilizing pare nts, teachers, and peers in na tural settings such as home and school would provide addi tional support of the utility of this methodology in young children with ASD. Limitations Out of the 32 studies reviewed, no functional analysis included an assessment of any peer-related social behavi ors, including peer-r elated withdrawn and peer-related negative social be havior. In addition, peers were not included in any of the studies as therapist or change agents in the assessment of target behaviors with young children with ASD. However, some research ers did find that the functional analysis identified different functions of problem be havior for the same child with ASD when different therapists conducted the functiona l analysis conditions (English & Anderson, 2004). In addition, Thompson et al. (1998) found that different forms of the same behavior (i.e., aggression) were maintained by different environmental variables. Future directions Given the success of the f unctional analysis methodology to identify environmental variables of a variety of behaviors across a variety of therapists, utilizing the functional analys is methodology to identify envir onmental variables of peerrelated withdrawn and peer-related negative so cial behaviors is a logical next step. Review of Social Skills Literature The purpose of this section is to review the literature on assessment and treatment of social behavior in young children with ASD. Similar to the literature on problem

PAGE 42

30 behavior with young children with ASD, the literature on soci al skills with young children with ASD suggests target behaviors (e.g., appropriate social behavior) can be manipulated (i.e., increased or decreased) based on modificat ions made to environmental variables within the child’s se tting (e.g., peer-training). Method to select reviewed studies. Appropriate articles for this literature review were identified by searching the following computerized databases: Academic Search Premier and PsycINFO. Keywords used to se arch the databases included, but are not limited to social skills interv ention, social skills assessment, social competence, social skills, and autism. The social skills body of literature met th e following criteria for inclusion in the review: (a) peer-reviewed jour nal, (b) published on or afte r 1985, (c) assessment and/or treatment of peer-related social behavior (at least one of the peer-related social behaviors had to be observable) conducted using a si ngle-subject design incl uding a variety of designs (e.g., multiple baseline across participants, multiple baseline across settings, reversal), (d) at l east one participant with ASD 8-years of age or younger. Review of Social Skills Studies This section reviews the l iterature on social skills training and treatment of young children with ASD. A total of 16 studies met the criteria mentioned above. The studies in the review ranged from 1992 to 2005. One notew orthy finding to mention is there has been a decrease in studies that include asse ssment and/or treatment of peer-related social behaviors using a single s ubject design with young children with ASD; 10 studies were conducted between 1992 and 1997, whereas 6 studies were conducted between 1998 and 2005. The most studies conducted in one year were five, in the year 1997. Since that year, the most studies conducted in one year were two.

PAGE 43

31 Participants included in review of literature There were a total of 50 participants acro ss the 16 studies identified. The number of participants that met the criteria for incl usion are described in Appendix B (Review of Social Skills Literature). In addition, the numbe r of participants who met criteria in each study, age, gender, and disability are reported in Appendix B. Age. All participants included in these studi es were between the ages of 3 and 8years old. Sixty percent (30) of the particip ants were between 3 a nd 5-years old and 40% (20) were between 6 and 8-years old. Number. Of the 16 studies included in this re view, 44% (7) had th ree participants that met the criteria for inclusion. Twenty -five percent (4) of the studies had two participants, 25% (4) had four participants, and 6% (1) had five participants that met criteria for inclusion in this review. Gender Of the 50 participants included in this review, 96% (46) were male and 8% (4) were female. Disability. All 50 participants had either a diagnosis of PDD-NOS, autism, or Asperger’s disorder. Under the category of AS D, 90% (45) had a diagnosis of autism, 8% (4) a diagnosis of Asperger’s, a nd 2% (1) a diagnosis of PDD-NOS. Therapists in studies Of the 16 studies reviewed, peers were involved in each study. In some studies peers were just present during the assessment and/or treatment and in other studies the peers served as therapists to the participants Specifically, peers were therapists in 69% (11) of the studies reviewed. Th irty-eight percent (6) of the therapists were participants’ teachers, 31% (5) were experimenters, and 6% (1) were participants’ caregivers. There was more than one therapist (e.g., a peer a nd teacher) in 50% of the studies reviewed.

PAGE 44

32 Specifically, 25% (4) of the studies include d peer and teacher therapists, 19% (3) included peer and experimenter therapists and 6% (1) included the participants’ caregiver and experimenter as therapists. In st udies that included more than one therapist, the data from both therapists were reporte d individually (e.g., p eer) and together (e.g., peer and teacher); therefore, the categorie s used to report ther apists will exceed 100% when totaled. The following are examples of how investigators employed a variety of therapists in the assessment and/or treatment of social skills. Ga rfinkle and Schwartz (2002) investigated the effect s of peer imitation training on the social interaction of young participants with ASD. Teacher therap ists led small group peer imitation training with the participant and 4 to 5 peers, all attending an integrated preschool. Training consisted of the participant and peers imitati ng the play of the peer leader. The leader switched throughout each session, so that during each session, each participant and peer was a leader at least once. Therefore, peers se rved as models for the participant to imitate and the teacher served as the trainer, providi ng least to most prompts to the participant and peers to ensure imitation of the p eer leader. In another study, Thiemann and Goldstein (2004) utilized peers more directly as therapis ts. In this study, the investigators trained peers to implement written text treat ment with participants and examined the effects of this intervention on the social be havior of young participants with ASD. This investigation employed both the experimenter and peers as therapists. The peers were trained by experimenters to use written text cues to facilitate social behavior between the peers and the participan t. After training, treatment began w ith the peer therapists and the participant. The peer therapis ts were given scripts and in structions to follow throughout the treatment sessions. The experimenter’s involvement was limited to pre-session

PAGE 45

33 directions, nonverbal feedback (i.e., circ ling a happy face for a correct prompt) during sessions, and post-session feedback to peer therapists. Therefore, in this investigation, the peers and experimenters served as therapists to th e participant, though the peers were the primary change agents, with the experimenter mainly serving as a trainer to the peers and the participant. Alternatively, in an inves tigation conducted by Shabani, Katz, Wilder, Beuchamp, Taylor, and Fischer (2002), peer s were present during treatment; however, peers were not provided with any training or instruction during treatment. Instead, the experimenter provided initiati on training to the participant using a tactile prompt (i.e., vibrator) and then evaluated the effects of the tactile prompt on the frequency of participant initiations during free play with p eers. Therefore, in this investigation, the experimenter was the therapist and the peers were simply present during the baseline and treatment assessment. Settings of studies All 16 studies were, in part, conducted in a school setting. Fifty percent (8) were conducted in an integrated classroom, 31% (5) were conducted in a classroom when the majority of the classmates were not present, 25% (4) in a general education classroom, 19% (3) in a special education classroom, and 13% (2) in the participants’ home. In some studies assessment and/or treatment was c onducted across more than one setting (i.e., participant’s special education classroom a nd general education classroom). When this occurred, both settings were recorded; therefore, the categories used to report settings will exceed 100% when totaled. Dependent measures There were several social behaviors meas ured across the 16 studies. The categories of behaviors reported are not mutually excl usive and therefore, will exceed 100% when

PAGE 46

34 totaled. For example, one study measured init iations and responses. Both initiations and responses were included as categories of beha vior reported. Therefore, when totaled, the number of categories of behavi or recorded (i.e., 2) divided by the number of studies (i.e., 1) multiplied by 100%, would exceed 100%. In addition, some researchers reported measuring the same dependent variable (e.g., initiation) but include d different behaviors (e.g., request for attention, re quest for action, helping) w ithin their own operational definitions of initiation. As a result, th e following should not be considered a comprehensive list of all peer-related soci al behaviors measured in these studies. A noteworthy point is that direct measures of observable beha viors, rather than scores from standardized achievement, intelligence, behavi oral, or other measures were included in this section. Sixty-nine per cent (11) studies measured pa rticipant initiations, 50% (8) participant responses, and 44% (7) measured in teractions. Thirteen percent (2) of the 16 studies measured unprompted/spontaneous in itiations, 6% (1) of studies measured initiations and responses in the form of a compliment, verbal initiations, use of augmentative communication to initiate, respond, and interact, social behavior (including, initiations, responses, and negative behavior), and appropriate social skills. Six percent (1) of studies measured use of language dur ing interactions (i.e., number of words per sentence), peer imitation, proximity to peers, engagement with materials while interacting, and disr uptive behavior. Design of studies The majority (75%) of the 16 studies u tilized a multiple baseline design in the measurement and treatment of social skills. Fifty-six percent (9) utilized a multiple baseline design across participants, 25% (4) a reversal, 13% (2) a multiple baseline across settings, 13% (2) a multiple baseline across peer trainers, 6% (1) a multiple baseline

PAGE 47

35 across activities, and 6% (1) a multiple base line across behaviors. The categories are not mutually exclusive as some studies included multiple designs (e.g., multiple baseline across participants and sett ings), therefore the percen tages will exceed 100% when totaled. For example, Garrison-Harrell, Kamp s, and Kravits (1997) evaluated the effects of peer networks on the duration of social interactions using a multiple baseline design across settings within a multiple baseline acros s 3 participants. Therefore, peer networks were systematically implemented across three settings for each of the 3 participants. Assessment of social skills Unlike the functional analysis studies reviewed in the pr evious section, a separate assessment of social behaviors prior to treat ment was not conducted. However, all studies included a baseline phase that provided assessment information on the levels of social behaviors of the participants prior to treatm ent. For example, Wert and Neisworth (2003) measured the frequency of spontaneous reque sts made by four young ch ildren with ASD. These measurements were obtained duri ng 30 minute play sessions prior to the implementation of treatment. The frequency of spontaneous requests made during these baseline observations were then compared to the frequency of spont aneous requests made during treatment (video self-modeling). Simila rly, Laushey and Heflin (2000) obtained a percent of appropriate social skills for two young children with ASD over a four week period, prior to the implementation of trea tment (i.e., active pe er tutoring). The percentage of appropriate so cial skills from baseline (i .e., pre-treatment) was then compared to the percentage of appropriate social skills during treatment. One noteworthy difference between the functional analysis stud ies reviewed in the previous section and the social intervention studies reviewed in th is section, is that a lthough baseline measures

PAGE 48

36 provided an assessment of the participants’ so cial behaviors, the treatments were not driven by the results of these assessments. Treatment of social skills All 16 studies reviewed incl uded a treatment phase that focused on increasing the appropriate social behavior of the participants. A variety of treatments were implemented across these studies, though all treatments were determined a priori (i.e., treatments were not assessment-based). In the majority, 63% ( 10), of treatments peers played a major role in the delivery and implementation of the treat ment (i.e., peer pivot al response training, peer networks, peer traini ng, peer incidental teaching, peer activity tutoring, peer imitation, peer video modeling, written text trea tment, peer self-evaluation). In 33% (5) of the studies teachers were a part of the treat ment (i.e., teacher led social skills training with and without feedback and contingent re wards, teacher incidental teaching, teacher facilitation of participant peer interactions). Finally in 33% (5), tr eatments that included only the participants’ were implemented (i .e., tactile prompt, priming, self-video modeling, self-management, positive reinfor cement). The purpose of these treatments was to increase the occurrence of appropriate social behavi or from baseline levels. For example, Pierce and Schreibman (1997a) investig ated the effects of peer pivotal response training (PRT) on the use of language duri ng interactions by young children with ASD. The participants increased thei r frequency of language durin g interactions from baseline to intervention. Similarly, in another example, the duration of social interactions for three young children with ASD increased from base line following the implementation of peer networks in an investigation by Garrison-Harrell et al. (1997).

PAGE 49

37 Findings of social skills assessment and treatment All 16 studies reviewed reported increases in the occurrence of social behavior following treatment, in comparison to baseline measures. For 88% (14) of the studies, all participants in the studies displayed incr eases across all target appropriate social behaviors. For 12% (2) of the studies, all partic ipants displayed an increase in at least one target social behavior. Five of the 16 studies reviewed included a generalization component in the treatment analysis (e.g., gene ralization across settings, peers, time). For 4 (80%) of those 5 studies, al l participants successfully generalized the target social behavior. For example, Zanolli, Daggett, a nd Adams (1996) evaluated the effects of priming on the spontaneous initiations of two young children with ASD. Spontaneous initiations increased for both participan ts, however priming only increased the spontaneous initiations for both settings fo r one participant. Th e second participant increased spontaneous initiations in one, but not both settings. Summary of social skills review of the literature The review of these 16 studies has illustra ted the success of a va riety of treatments to increase the appropriate social behavi ors displayed by young children with ASD. A variety of topographies of so cial behavior have been targeted for assessment and treatment. To date, a treatment of social be havior driven by the fi ndings of a functional analysis of peer-related social behavior has not been conducted w ith young children with ASD. Strengths The social skills literature has de monstrated good success in developing treatments that increase a va riety of appropriate social behaviors (e.g., initiations, responses, interactions, peer imitation, la nguage used during interactions), across a variety of settings (e.g., genera l education classrooms, inte grated classrooms, special

PAGE 50

38 education classrooms, home) and therapists (e.g., peers, teacher, experimenter) with young children with ASD. All 16 studies reviewed were conducted, at least in part, in a school setting and with peers pr esent. This demonstrates th e feasibility of implementing treatments focused on increasing participants’ ap propriate peer-related social behavior in the participant’s natural environment with their peers. Treatments were evaluated using sing le subject design methodology, with the majority (75%) of studies employing multiple baseline design and the remaining 25% of studies employing a reversal design. The use of these single subject designs provided experimental control to demons trate that the implementation of the treatment components rather than an unknown variable resulted in the increases in social behavior (Kazdin, 1982). Limitation One of the main limitations in this literature is the lack of assessmentbased treatments. Despite the success the pre-determined treatments have had in increasing social behavior and demonstrati ng experimental control of these successes through the use of multiple base line and reversal designs, th e reasons young children with ASD engage in social interactions with thei r peers is unknown (i.e., the outcomes of their social interactions). Without this informa tion it is unknown whether treatments linked to the functions or outcomes of peer-related so cial behaviors would result in more robust and durable changes in behavior s, (e.g., increases in social behavior across participants, behaviors, and settings, and improvement s in generalization and maintenance). One interesting findings is that even t hough all 16 studies reviewed were conducted in applied settings, teachers serv ed as therapists in less than half the studies (38%). In order for these treatments to continue over time it may be necessary to increase the use of

PAGE 51

39 teachers as change agents. It is unknown if the low percent of teachers serving as therapists is due to the willingness of the te achers to participate, feasibility of teacher implementation of treatment procedures or othe r reasons. However, if the treatment is not feasible for a teacher to implem ent, even if it is successful in increasing social behavior during the course of the inve stigation, the treatment may not be continued, resulting in diminished treatment gains. Future directions Given the demonstrated effectiv eness of the functional analysis methodology to identify variables maintaining a variety of beha viors, across a variety of settings and therapists, a logica l next step in the social skil ls literature is to examine the use of functional analysis methodology to develo p assessment based treatments. Conclusions and Future Directions for Research In this chapter, two bodies of literature were review ed to (a) demonstrate the effectiveness of functional analysis met hodology in identifying the function of problem behavior for young children with ASD, and (b ) illustrate assessment and treatment of social behaviors in young children with ASD. The review of the functional analysis literature demonstrated the success of this methodology to identify maintaining variables of a variety of problem behaviors of young ch ildren with ASD. Despite the variety of problem behaviors assessed, the functional analysis literature does not include an assessment of peer-related social behavior for young children with ASD. The review of the social skills literature illustrated a vari ety of treatments successfully implemented to increase the appropriate soci al behavior of young children wi th ASD. However, not all treatments were demonstrated to be successful across all participants settings, or time. The social skills literature does not include treatments based on functions or outcomes that maintain participants’ social behaviors. There is a paucity of research utilizing

PAGE 52

40 functional analysis of peer-related social behavior with young children with ASD. Given the success of the functional an alysis to identify variables maintaining problem behavior and develop robust treatments, the extensi on of the functional analysis methodology to assess peer-related social behavior appears to be the next step. This study seeks to expand the litera ture by applying functional analysis methodology as a means of assessing peer-r elated social beha viors of young children with ASD. To that end, the study has th e following purposes: (a) to evaluate the effectiveness of the use of the functional analysis methodology to determine the function of peer-related withdrawn a nd peer-related negative social behavior for three young participants with ASD, (b) to compare the re sults of a functional analysis of peer-related withdrawn and peer-related ne gative social behavior acro ss participants with ASD to determine if the functions differed across participants, (c) to determine if peers were able to implement the functional analysis pro cedures with acceptable integrity, and (d) to examine rate per minute of participant peer-re lated positive initiations and percentage of participant peer-related positive and peer-related no response social behavior across functional analysis conditions.

PAGE 53

41 CHAPTER 2 METHOD Overview This study is an exte nsion of an OSERS f unded project (Asmus & Conroy, 2003). The project is a 3-year study that examined the use of ABA techniques in the assessment and treatment of peer-related withdrawn and peer-related negative social behavior displayed by young children (between the ages of 18 months and 5-years) with ASD in their natural settings. Participants Participants were the first three children enrolled in an OSERS funded project that reached the functional analysis phase of assessment (Asmus & Conroy, 2003). The project involved the assessment and treatment of peer-related withdr awn and peer-related negative social behavior for participants diagnosed with ASD. The participants were referred to the project by their parents or th e University of Florida Center for Autism and Related Disabilities (CARD). Enrollment criteria for inclusion in the project included: (a) the participant was between the ages of 18 months and 5-years of age, (b) the participant had a diagnosis that fell under the category of ASD, (c) the participant was identified and directly observed to engage in peer-related w ithdrawn and/or peer-related negative social behavior in the school setting, and (d) the pare nts, teacher, and principal/director agreed to participate in the project. The participants included in this study were Shane, Shannon, and Colin. Pseudonyms were used to protect participants’ anonymity. Shane. Shane was a 5-year-old boy diagnosed with Pervasive Developmental Disorder, Not Otherwise Specified (PDDNOS). He attended a pre-school varying

PAGE 54

42 exceptionalities classroom at a local public school where he received speech/language and occupational therapy (OT) services. Afte r school, Shane attended a private general education pre-school with 18 students and one teacher, where all assessments were conducted. Shane had mild cognitive delays and significant expressi ve language delays with echolalia. Shane could follow simple requests. He rarely used speech to communicate spontaneously, though he would repeat phrases such as, “Help please” when repeatedly asked to do so by an adult. He mainly used gestures such as handing items or leading adults to communicate his wants and needs. Shane had limited play skills (i.e., played with Lego's by banging th em together or on the floor instead of building with them) and spent the majority of his time engaging in stereotypic behavior. Shane's peer-related withdrawn social behavior was leaving the area in which peers were present. Shane's peer-related negative soci al behavior was aggression (i.e., pinching, biting, hitting, kicking, pushing, throwing toys at peers). His stereo typic behavior was pacing back and forth in the classroom, rep eatedly shaking a small toy with his hand, and making nonfunctional vocalizations. Shannon. Shannon was a 4-year-old girl diagnos ed with autism. She attended a private general education pre-school classroom with 15 students and one teacher, where all assessments were conducted. Previousl y, Shannon had participated in a separate, unrelated study in her home environment that evaluated treatment for problem behavior (e.g., noncompliance, tantrums). Concurrently with this study, Shannon participated in a separate, unrelated study that focused on in creasing her verbalization skills. Shannon could follow simple requests and spoke in full sentences, though he r speech was often difficult to understand as she often spoke very softly. She displayed independent play

PAGE 55

43 skills (e.g., pushed buttons to open and close th e drawer of a play cash register) and independent pretend play (e.g., pretended sh e was making pizza with play food/oven and pretended Barbie's were a family with a mother and father). Shannon's peer-related withdrawn social behavior was leaving the ar ea in which peers were present. Her peerrelated negative social behavi or included yelling and/or wh ining, “No” and “I want the ___ (e.g., toy).” Shannon did not en gage in stereotypic behavior. Colin. Colin was a 3-year-old boy diagnosed with PDD-NOS. He attended a public special education pre-school classroom w ith five students, one teacher, and two paraprofessionals (one was assigned just to Colin). Colin received 30 minutes of speech/language services twice a week and 30 minutes of OT services once a week. Colin had extremely limited expressive communica tion and displayed immediate and delayed echolalia. He inconsistently us ed gestures such as handing an d leading adults to obtain his wants and needs. Colin displayed inde pendent play skills (e.g., putting together puzzles) and some parallel play skills (e.g., put ting together a puzzle wh ile sitting next to a peer). Colin's peer-related withdrawn social behavior was leaving the area where peers were present. His peer-related negative so cial behavior includ ed aggression (i.e., pinching, scratching, pulling hair, biting). Co lin’s stereotypic be havior was hand/arm flapping. Settings and Therapists This study included three phases: (1) de scriptive assessmen t, (2) preference assessment, and (3) functional analysis. Each phase was conducted in the participant's school setting. Phase 1 was conducted in the participant's classroom with all other students present. Phase 2 and 3 were conducte d in the classroom (except Colin) either when all of the students were present (Sha ne) or when other students were at recess

PAGE 56

44 (Shannon). For Shane, assessments were conduc ted during classroom activities that were student directed (e.g., centers) to avoid inte rrupting the class schedule. For Colin, Phase 2 and 3 assessments were conducted in the atrium directly next to his classroom. During Phase 3, one part of the classroom (designa ted by the teacher) was used to conduct the sessions, with the exception of Colin, whose assessment was conducted in a designated area of the atrium. Masking tape was placed on the floor to separa te the assessment area from the general classroom or atrium. The a ssessment area was then further split into equal halves, by placing tape on the floor, to identify the in and withdrawn areas. The in area was the area where peers were present and the withdrawn area was the area where only the participant could go to be alone. The dimensions for both Shane's in and withdrawn areas were 5 ft by 3 ft, Colin's we re both 3 ft by 3 ft, and Shannon's were each 4 ft by 3 ft. For each participan t, the dimensions for the in area were the same as the dimensions of the withdrawn area. Additi onal physical borders (e.g., bookshelf) were used for Shane and Shannon, as needed, to block off the assessment area (e.g., to decrease opportunities for th e participant to leave the entire assessment area). Phase 1 (descriptive assessment) involved naturalistic classroom observations and interviews with teachers; therefore there was no therapist. A graduate research assistant with training in ABA served as the primary therapist for Phase 2 (preference assessment). During Phase 3 (functional analys is), a graduate research as sistant with training in ABA trained the peers, provided coaching and assistance to the peers as needed in each condition, and oversaw all sessions. In a ddition, during Phases 2 and 3, a second therapist, either a graduate re search assistant or undergradu ate assistant, was present at

PAGE 57

45 each session to operate the video camera, a nd in Phase 2 to serve as a second data collector. In Phase 3, a pool of five peers were sele cted from the participant's class (except for Colin) based on their parent/guardian consent, consistent attendance, willingness to participate, compliance and skills to follow one step verbal instructions, and intelligibility of speech (volume and clearness). There were two peers present who participated in each functional analysis condition. The two peers were chosen prior to each condition based on availability (e.g., presence in the classroom) and willingness to participate. Attempts were made to utilize the same two peers as often as possible. If two peers were not available or were not willing to participate, sessions were cancelled for that day. Colin's classmates did not participate due to lack of compliance and skills to follow one step verbal instructions, and/or intelligibility of speech (volume and clearness). His peers were chosen, using the same guidelines stated above, from a general e ducation kindergarten classroom at his school, with peers he had oppor tunities to interact with on previous occasions. Materials The materials used in Phase 1 were materi als from the participants' school setting. Materials for use in the functional analysis (P hase 3) were identifie d for each participant via a preference assessment (Phase 2). Seve n stimuli were selected for the preference assessment based on experimenter observations of materials the participants' played with in the classroom during Phase 1; and materials that were cond ucive to socialization (e.g., peer-related social behavior could occur when playing with Lego's though would be more difficult when listening to music on headphones). The experimenter reviewed all sessions from Phase 1 and selected from a running tally of items the participan t played with. Only

PAGE 58

46 materials that could be played with by more than one child were selected for use during the preference assessment. In addition, because there needed to be more than one item available for three children to play with, onl y materials that had at least one identical match (e.g., two Thomas the Engine electr onic books, multiple wooden blocks) in the classroom or that could be easily purchased we re selected for inclusion in the preference assessment. Preference assessment. Prior to the functional analysis, a multiple stimulus without replacement (MSWO) preference assessment was conducted with each participant based on the procedures describe d in DeLeon and Iwata (1996). From this assessment, high preferred, neutral, and low pr eferred tangibles were identified for each participant, for use in the functional analys is (Phase 3). The materials differed for each participant and varied acro ss functional analysis conditi ons. Neutral tangibles were available in all functional an alysis conditions, except the ignore conditions (which had no materials). The participant's high preferre d tangible was available only during the tangible condition of the functional analysis. Response Definitions Preference Assessment During Phase 2, the participant's choice of a stimulus was scored for each trial. A choice was defined as physically touching a stimulus. If two or more stimuli were touched, the stimulus touched first was scored as the stimulus chosen. If two or more stimuli were touched simultaneously, no stimul us was scored and the trial was repeated. Functional Analysis Several participant and peer behaviors were scored during Phase 3. The primary dependent variables were the participants' peer-related withdrawn and peer-related

PAGE 59

47 negative social behaviors. In addition, other pa rticipant peer-related social behaviors were scored as appropriate (initi ations, responses, no response). The primary peer behaviors scored in Ph ase 3 were (1) presentation of social demands to the participant in the escape condition, and (2) delivery of relevant consequences following participant peer-relat ed withdrawn and/or peer-related negative social behavior in the escape, attention, a nd tangible conditions Correct delivery of social demands and consequences were record ed as indicators of procedural integrity. Additional peer behaviors recorded were posi tive initiations in the free play condition. Participant peer-relate d withdrawn behavior Peer-related withdrawn soci al behavior was defined as more than 50% of the participant's body crossing the vertical plane of the tape into the withdrawn area (the tape separated the withdrawn and in areas). This included entering the withdrawn area and leaving the entire assessment area (i.e., more than 50% of the participant's body crossing the vertical plane of the tape that led out of the in area, though not into the withdrawn area). Participant peer-relate d negative behavior Peer-related negative social behavior was individually de fined for each participant. Peer-related negative social behavior incl uded the following behavi ors: aggression (i.e., pinching, scratching, biting, hitting, kicking, pushing, pulling hair, th rowing materials at peers) and yelling/whining (i.e., “No,” “I want the ___[toy]”). For Shane, negative initiations and responses were also recorde d. A negative initiation was defined as Shane initiating any non-accidental verbal or gestural social behavior with the potential to cause harm to the peer(s) in attempt to elicit a social response. It in cluded destruction (e.g., destroying peers’ objects/st ructures) and disruption (e.g., yelling). If Shane’s negative

PAGE 60

48 initiation also met the definition for peer-rel ated negative social behavior, peer-related negative social behavior was scored instead of negative in itiation. A negative response was defined as any non-accidental verbal or gestural social be havior with the potential to cause harm to peer(s) when it overtly acknowledged an initia tion within 3 seconds of the end of the initiation (e.g., if the initiation was sayi ng “Hi,” a negative response was scored if it occurred after the peer said, “Hi” and before 3 seconds had elapsed following the termination of the peer in itiation). This included physic al destruction (e.g., destroying peers’ objects/structures) a nd disruption (e.g., yelling). If Shane’s negative response also met the definition for peer-related negative so cial behavior, peer-related negative social behavior was scored instead of negative response. Participant peer-related appropriate behaviors Participant peer-relate d positive initiation A positive initiation was defined as the participant initiating any verbal or gestural social behavior direct ed toward a peer in an attempt to elicit a social response. Positive in itiations included verbal or gestural social behavior that (a) provided assistance, comfor t, or affection (e.g., gently putting hand on peers back); (b) elicited peer attention or access to objects/acti vities (e.g., inviting peer to join play activity, calling out peer name, offering materials, seeking information, commenting, and complimenting, and/or seek ing help/assistance); and (c) elicited attention or access to objects including role assignment, directing an activity, and coordinating pretend play and social contact. Participant peer-relate d positive response A positive response was defined as any verbal or gestural soci al behavior that the participant engaged in to overtly acknowledged an initiati on within 3 seconds of the end of the peer’s initiation (e.g., if a peer initiated to the particip ant by saying their name, a positiv e response was scored if the

PAGE 61

49 response occurred after the pa rticipant’s name was stated and before 3 seconds had elapsed following the name being stated). This included the participant helping a peer to complete a task or other ac tivity, following a peer’s l ead, helping behaviors, and reciprocal social behaviors. Participant peer-related no response A no response was defined as the participant either knowingly or unknowingly ignoring a peer(s) initiation or continuing to engage in the same play behavior. No res ponse was scored if th e participant did not engage in peer-related negative social behavior and if the participan t did not engage in a peer-related verbal or gest ural positive response (or nega tive response, for Shane only) within 3 seconds of the termination of the p eer’s initiation (e.g., If a peer said, “Colin, Colin, Colin” with no time gap, one no respons e rather than three no responses was scored). During the escape condition participant peer -related no response behavior was also scored when the peer presented a social de mand to the participant and the participant either did not complete the social dema nd or did not complete the social demand correctly (e.g., the par ticipant was told to turn the pa ge of a book and instead pressed a button on the book). Peer procedural integrity data Peer social demands Peer social demands were defi ned as any verbal or gestural social demand presented to the participan t during the escape condition. Correct peer presentation of social demands was define d as the peer presenting a social demand (independently or following a prompt from the experimenter) using a two (verbal, gestural) or three step (verbal, gestural physical) prompt sequence every 15 seconds (excluding the 15 seconds the participant receiv ed contingent negative reinforcement). A

PAGE 62

50 criterion rate of 4.0/minute that was consiste nt across escape conditions was identified as a target to demonstrate adequate procedural integrity. Presenting social demands a mean rate of at least 3.2/minute was consider ed to meet the integrity guideline. Peer delivery of consequences Peer delivery of consequences was defined as a peer providing the participant with 15 sec onds of the relevant consequence (attention, escape, or access to the high preferred tangibl e) contingent on the participant's peerrelated withdrawn and/or peer -related negative social behavior. Consequence delivery was defined for each condition as follows: 15 se conds of attention in the form of verbal reprimands/requests to stay and play near peers in the attentio n condition; 15 seconds access to high preferred tangible in the tangi ble condition, and a 15 second break (escape) from social demands in the escape conditi on contingent on par ticipant peer-related withdrawn and/or peer-related negative soci al behavior. Correct peer consequence delivery was defined as the peer providing th e relevant consequen ce (independently or following a prompt from the experimenter) w ithin 5 seconds of the participant's peerrelated withdrawn and/or peer -related negative social behavi or and providing the relevant consequence for at least 15 seconds (the experi menter kept track of the 15 seconds using a timer and told the peers when the 15 sec onds had elapsed). A mi nimum criterion of 80% that was consistent acro ss escape, tangible, and atten tion conditions was identified as a target to demonstrate adequate inte grity. Correct delivery of consequences an average of 80% was considered to meet the in tegrity guideline. Peer consequences were not provided for participant peer-related appropriate initiation s and/or responses. Peer positive initiations A peer positive initiation was defined as the peer initiating any verbal or gestur al social behavior directed toward the participant in an

PAGE 63

51 attempt to elicit a social response. This was recorded only during the free play condition. Positive initiations included verbal or gest ural social behavior that (a) provided assistance, comfort, or aff ection (e.g., gently putting hand on participant’s back); (b) elicited participant attention or access to objects/ activities (e.g., invi ting participant to join play activity, calling out participant’s name, offering ma terials, seeking information, commenting, and complimenting, and/or seek ing help/assistance); and (c) elicited attention. Correct presentation of a peer pos itive initiation was de fined as the peer positively initiating to the participant every 15 seconds (the experime nter kept track of the 15 seconds using a timer and told the peers when the 15 seconds had elapsed). A criterion rate of 4.0/minute that was consiste nt across free play conditions was identified as a target to demonstrate adequate procedur al integrity. Presenting positive initiations a mean rate of at least 3.2/minute was cons idered to meet the integrity guideline. Observation System and Interobserver Agreement Observation System During all three phases, all sessions a nd observations were videotaped using a standard mini DV video camcorder. Audio was obtained from a wireless microphone connected to the pa rticipant by hooking the microphone p ack on the participant's pant belt or waste band and clipping the microphone to the participant's shirt. An audio receiver unit was connected to the camcorder to transmit audio from the wireless microphone to the mini DV. This assisted in enhancing the participant and surrounding peers’ verbalizations and limiting background noise from the classroom and atrium. For all phases, the mini DV tape was played back on the camcorder, in VCR mode, and downloaded to Movie Star on an HP or Dell computer via DAZZLE or PINNACLE. In Phase 1, all sessions were then viewed for da ta collection on a computer via Movie Star,

PAGE 64

52 with video and audio occurring simultaneously. In Phase 2, data were taken both live and via Movie Star. In Phase 3, all sessions were vi ewed for data collection via Movie Star or via the camcorder (the DV was played on the camcorder on VCR m ode and viewed on a television screen by connecting the camcorde r video and audio devi ces to a television monitor). During Phase 1, five (Shane and Colin) to six (Shannon) hours of naturalistic observations across pre-specified contexts we re collected. Data were collected using paper and pencil to tally each ma terial the participant engaged with for use in Phase 2. In addition, a list of every participant and peer initiation was obtained to identify social demands for use in Phase 3. During the preference assessment (Phase 2) data were collected via paper and pencil. A data sheet with fi ve sessions and seven trials per session was utilized. The stimulus chosen for each trial was written down in the corresponding session and trial. The percentage of trials each stimulus wa s selected was calculated by dividing the number of times a stimulus was chosen by the number of trials the stimulus was presented. For example, if a stimulus was select ed on the first trial of all five sessions the formula was 5 / 5 x 100 = 100%. If a stimulus wa s selected on the second trial of all five sessions the formula was 5 / 10 x 100 = 50%. Th e stimuli were then ranked, based on the percentage of trials each stimulus was sele cted, as high preferred (stimulus chosen the highest and/or second highest pe rcentage of trials), neutral (1 to 3 stimuli chosen second through fifth percentage of tria ls), and low preferred (2 to 3 stimuli chosen fifth through seventh percentage of trials).

PAGE 65

53 During Phase 3 (functional analysis), data we re collected via Dell palm pilots using the Tap It program (Tapp, 2003). Subsequent to scoring each day's sessions, the data were downloaded from the palm pilots to a co mputer. The data were then analyzed using the Multiple Option Observation System for Experimental Studies (MOOSES, Tapp, 2002) from which frequency, duration, and perc entage reports of each behavior were obtained. From these reports responses per mi nute were calculated by dividing the total frequency of the behavior by the duration (i.e ., minute) of the condition (i.e., 5 minutes). Phase 3 behaviors During Phase 3 participant peer-related positive initiations, responses, no responses, withdrawn, and nega tive social behavior were scored as frequency data. In addition, peer social de mands, peer consequence delivery, and peer positive initiations were scored as frequency data. Frequency behaviors were scored at the start of the behavior. For example, each time an initiation started (e.g., participant stated peer's name ) the initiation (e.g., participant peer-related positive initiation) was scored. For each peer initiation a participant response was recorded. Similarly, for each peer social demand, a participant response was recorded. The duration behavior scored in Phase 3 was peer delivery of consequences. For example, when a peer started to provide contingent attenti on (in the attention condition) following the participant’s peer-related wit hdrawn and/or peer-related negative social behavior, the peer delivery of consequences button on the palm pilot was turned on. The delivery of consequences duration code was tu rned off 3 seconds after the peer stopped providing the participant with the consequence (e.g., attention). Peer consequence delivery was analyzed using occurrence, latency to onset, and duration measures.

PAGE 66

54 Interobserver Agreement Interobserver agreement was not collected during Phase 1 as it involved compiling information from the descriptive observation to determine which materials to include in the preference assessment (Phase 2) and what social demands peers would present participants during the escape condition of the functional analysis (Phase 3). During Phase 2, the primary and reliability observer i ndependently collected data either live, via the camcorder (the DV was played on the camcorder on VCR mode and viewed on a television screen by connecting the camcorde r video and audio devi ces to a television monitor), or from a computer via Movie Star. The primary data collector, reliability data collector, or an additional graduate research assist ant conducted the preference assessment with each participant. Data collec tors consisted of the first author and/or trained graduate research assistants. Interobserver agreement was calculated for each session by dividing the number of agreemen ts by the number of agreements plus disagreements and multiplying by 100% (K azdin, 1982). Interobserver agreement was collected for an average of 63% of all preference assessment sessions. Interobserver agreement was collected for 100% of Shan e’s sessions, 30% of Shannon's sessions, and 60% of Colin’s preference assessment se ssions. Interobserver agreement was 100% across all sessions a nd participants. During the functional analysis (Phase 3) the primary and reliability data collectors consisted of the first author a nd/or trained graduate research assistants. Graduate research assistants were trained to criteria prior to taking reliability data. A graduate research assistant was considered trained after co llecting data on three functional analysis conditions with interobserver agreement of at least 80% when compared to the first author. Both the primary and reliability obs ervers collected data independently via a

PAGE 67

55 camcorder (by playing the tape on VCR m ode and connecting the camcorder to a television screen) or Movie Star on a computer. Interobserver agreement on frequency and duration behavior was calculated fo r each session using MOOSES (Tapp, 2002). Interobserver agreement on frequency beha viors was calculated by MOOSES by dividing the number of agreements by the number of agreements plus disagreements multiplied by 100% (Kazdin, 1982). An agreement for fre quency behaviors was defined as both observers recording the same behavior within 5 seconds before or after each other. A disagreement for frequency be haviors was defined as the tw o observers not recording the same behavior within 5 seconds before or after each other. Interobserver agreement on duration be haviors was calculated by MOOSES by dividing the smaller duration by the larger duration multiplied by 100% (Miltenberger, 1997). An agreement for duration behaviors was defined as both observers recording the beginning of the same behavior within 5 sec onds before or after each other and recording the end of that same behavior within 5 sec onds before or after each other. Interobserver agreement was obtained on an average of 58% of sessions across participants. The interobserver agreement for each behavior was averaged across conditions for each participant to obtain an overall ag reement for the functional analysis. Interobserver agreement was obtained fo r 33% of Shane's functional analysis sessions. Agreement averaged 85% (range 33% to 100%) for Shane's peer-related withdrawn social behavior, 95% (range 80% to 100%) for peer-rel ated negative social behavior, 100% for peer-related negative init iations, 96% (range 80% to 100%) for peerrelated positive response, 100% for peer-related negative response, and 91% (range 78% to 100%) for peer-related no response. Although positive initiations occurred,

PAGE 68

56 interobserver agreement was not collected duri ng those conditions and therefore was not reported. Agreement averaged 76% (range 69% to 83%) for Shane's peer social demands and 95% (range 89% to 99%) for Shane’ s peer delivery of consequences. Interobserver agreement was obtained fo r 78% of Shannon's functional analysis sessions. Agreement averaged 100% for Sh annon's peer-related withdrawn social behavior and 100% for peer-rel ated negative social behavior Agreement averaged 100% for Shannon’s peer-related positive initiations 66% (range 0% to 100%) for peer-related positive responses, and 87% (range 50% to 100%) for Shannon’s peer-related no response behavior. Agreement averaged 84% (range 65% to 95%) for Shannon’s peer social demands and 96% (range 95% to 97% ) for peer delivery of consequences. Interobserver agreement was obtained fo r 52% of Colin's functional analysis sessions. Agreement averaged 100% for Colin’s peer-related withdraw n social behavior and 100% for his peer-related negative social behavior. Agreement was 100% for peerrelated positive initiations and for peer-related positive responses and 89% (range 68% to 100%) for peer-related no responses. Agreem ent averaged 95% (ra nge 79% to 100%) for peer social demands and 99% (range 99% to 100%) for peer delivery of consequences. Design and Data Analysis Design of Procedures No design was used in Phase 1 (descriptive assessment) to select items for use in the preference assessment (Phase 2) or to se lect social demands to be presented to the participant in the escape condition (Phase 3) In Phase 2, (preference assessment) stimuli were counterbalanced (placement in the line of items presented) across sessions and after each trial each stimulus was moved one pl ace to the left, to avoid possible sequence effects (DeLeon & Iwata, 1996). Phase 3 (functional analysis) was conducted by

PAGE 69

57 alternating between four to five conditions using a mu ltielement design (Barlow & Hayes, 1979; Iwata et al., 1982/1994). The func tional analysis conditi ons were presented in a random order and/or based on continuous analysis of data, to avoid possible sequence effects. The order was determined by placing the name of each condition in a hat and pulling out one at a time or based on ongoing analysis of conditions. For Shannon, a second functional analysis was conduc ted with minor procedural variations, though conditions and design remained the sa me as the first functional analysis. Data Analysis During Phase 3, data were analyzed via vi sual inspection of single-subject design line graphs. Although data were collected on a variety of part icipant and peer behaviors, decisions regarding what conditions to conduc t and when to stop data collection were based only on rate per minute of participant peer-related withdrawn and/or peer-related negative social behaviors. Rate per mi nute of these behaviors was calculated by MOOSES (Tapp, 2002) by dividing the total freq uency of each behavior by five (i.e., total session time in minutes). To evaluate the function of both the participant's peerrelated withdrawn social be havior and peer-related ne gative social behavior, the behaviors were graphed separately. Therefore, one graph depicted rate per minute of the participant's peer-related withdrawn social be havior and a second graph depicted rate per minute of the participant's peer-related negative social behavior. Decisions on when to end data collection were made separately ba sed on each graph. Each graph was visually inspected for stability base d on procedures described by Kazdin (1982). A three point trend rule was used as a guideline to make decisions on when to stop data collection. The condition(s), in which each be havior occurred most often, in a stable trend, were considered the function of that behavior.

PAGE 70

58 Visual inspection techniques were used to determine which condition(s) the participant engaged in more peer-related withdrawn social behavior and/or negative social behavior. The functional analysis was continued until a clear function (at least three increasing, decreasing, or st able data points) was determined for the rate per minute of each of these behaviors. The line graphs we re compared to determ ine if the function of peer-related withdrawn and/or peer-related negative social behaviors differed for each participant. Following the analyses of rate per minute of peer-related withdrawn and peer-related negative social behavior, averag es of the peer integrity measures were compared to determine the extent with which the peers implemented the functional analysis procedures with acceptable procedural integrity. Mean rate per minute of participant peer-r elated positive initiations, responses, and no responses were compared post-hoc to determine if there were differences across functional analysis conditions. Procedures Phase One: Descriptive Assessments During the descriptive assessment seven stimuli were selected for use in the preference assessment (Phase 2). Tangibles we re identified through direct observation. Social demands were identified for use in th e escape condition of th e functional analysis (Phase 3). Social demands were se lected through direct observation. Direct observations were conducted in each participant's classroom. Five hours of direct observation were conducted for Shan e and Colin, and 6 hours were conducted for Shannon. Observations were videotaped fo r subsequent viewing and scoring. These observations were conducted across several days in order to obtain 30 minutes of data on participant and peer social behavior in each of the following contextual factors: activity

PAGE 71

59 (manipulatives/blocks, cognitive/books/pre-academic, art/sensory, pretend/sociodramatic, dance/music, snack/meal, computer, and games with rules), play format (adult directed, child directed), a dult engagement (adult active, passive, and disengaged), and peer group size (one-on-one, small, and larg e). From these observations, a list of tangibles and social demands were developed for use in the preference assessment (Phase 2) and functional analysis (Phase 3). The pr imary investigator or a research assistant viewed the videotapes and made a list of all the tangibles the participant engaged with during the direct observations. From this list and indirect assessments, tangibles were selected for use in the preference assessment. The primary investigator or a research assistant viewed the videotapes a second tim e and made a list of all the peer-related initiations made by the participant and pee r(s) during the direct observations. The exact topography of each participant and peer in itiation observed during the descriptive observations was listed (e.g., peer initiate d to participant by asking, “Will you hand me the block?” or the participant initiated to the peer by holding up her doll in the direction of the peer). From these peer-r elated initiations, social dema nds were selected for use in the escape condition of the functional analysis. Identification of tangibles From information obtained fr om the direct descriptive assessments, seven stimuli were selected for inclusion in the preference assessment. In addition, stimuli were selected ba sed on more than one person being able to play with the stimulus (e.g., social behavior could occu r when playing with Lego’s though would be more difficult when listening to music on headphones) and the availability of multiple exact stimuli in the classroom setting. For Shannon, one additional stimulus (play cash register) was included in the preference assess ment that was not selected from the direct

PAGE 72

60 descriptive assessment. The play cash register was included, because Shannon had been observed to play with the cash register at other times and her teacher reported Shannon would often play with this toy. Identification of social demands Social demands were determined individually for each participant and based on the participan t and peer initiations that were recorded from the descriptive observations. Initiations that facilitated peer-related positive social behavior (e.g., “Hand me the book”) were sel ected as social demands in the escape condition. Initiations that did not facilitate peer-related positive social behavior (e.g., “No!” “Stop it”) were not selected to be used as social demands. Phase Two: Preference Assessment A preference assessment was conducted at each participant's school in either their classroom or atrium (Colin). The seven stim uli that were selected for each participant were evaluated for preference using proce dures similar to DeLeon and Iwata (1996). There were five sessions conducted, with seve n trials per session. Additional preference assessments were conducted as needed if ther e was concern that pref erence for items had changed. The primary investigator or graduate research assistant served as the therapist across all sessions. On the first trial of each session, all seven stimuli were presented in a straight line across a table or on the floor, facing the participan t. The participant sat at a table (or on the floor) with the seven stimuli spread evenly across the table (or floor). After being shown how to manipulate each stim ulus the participant was asked to choose what item s/he wanted to play with. The first stimulus the participant touched was recorded as the chosen stimulus for that trial. The participant was then given the opportunity to engage with the chosen stimul us for 30 seconds. Before beginning the next trial the chosen stimulus was removed from the participant's view and the remaining

PAGE 73

61 stimuli were moved one place to the left and even ly spread out in front of the participant. The participant had 30 seconds on each trial to choose a st imulus. If no stimulus was chosen, the session was terminated and the remain ing trials were scored as “not chosen”. Attempts to choose two stimuli were bl ocked and resulted in the choice being represented. Each stimulus was ranked by divi ding the number of trials the stimulus was chosen by the number of trials the stimu lus was presented, then multiplying by 100%. The stimulus (or stimuli) chosen the highest and/or second highest pe rcentage of trials was identified as the high preferred tangi ble item and was presented only during the tangible condition. The stimuli chosen second through fifth percentages of trials were ranked as the neutral tangibles and were pr esent in all conditi ons except ignore. In addition, the same ratio of neut ral tangibles were present in the withdrawn and in areas. For example, if Thomas the Train book wa s identified as a neutral stimulus, three Thomas the Train books would be present in the in area and one Thomas the Train book would be present in the wit hdrawn area in the attention, ta ngible, and escape conditions of the functional analysis. This procedure wa s adapted for Shannon, as only one type of stimulus (i.e., Barbie’s) was ranked as neutral. Therefore, three Barbies were present in the in area and one Barbie in the withdrawn area. Phase Three: Functional Analysis A functional analysis was conducted at each participant's school in the participant's classroom (except Colin, whose functional anal ysis was conducted in the atrium outside of his classroom). The procedures were based on Iwata et al. (1982/1994), though differed in the following ways: (a) relevant re inforcement was delivered in test conditions (i.e., attention, tangible, escape) contingent on peer-related wi thdrawn and/or peer-related negative social behavior, rath er than problem behavior di rected toward adults, (b)

PAGE 74

62 relevant reinforcement (i.e., access to attenti on, access to high preferred tangible, escape from social demands) was delivered by peers rath er than an adult, and (c) in the escape condition, social demands were presented by pe ers rather than adults, and the demands were to engage in peer-related positive res ponses rather than academic/vocational tasks. In addition, the session area was split in ha lf by placing masking tape on the floor and entering one half (i.e., the w ithdrawn side) was designated as peer-related withdrawn social behavior. Approximately three to fi ve 5 minute conditions were conducted during each 30 to 60 minute session. Sessions were conducted one to three times per week. Reinforcement was provided for 15 seconds c ontingent on peer-relate d withdrawn and/or peer-related negative social behavior in the attention, ta ngible, and escape conditions. Training to Peers Functional analysis conditions were pres ented to peers as games, each game (condition) having different rules for how to play with the participant. Prior to each condition, peers were told the rules for that condition (game) and the length (i.e., 5 minutes) of the condition. After the therapist ve rbally instructed the peers on the rules for that condition, the peers were required to demonstrate unde rstanding of the rules by roleplaying with each other and the participant. Each peer role-p layed as both the participant and the peer. The condition only began after both peers demonstrated knowledge of the rules by successfully role-playi ng the relevant responses to participant behavior with a peer or the participant. Therefore, a peer was required to demonstrate 100% accuracy in delivering the relevant consequence prior to participating in each condition. The instructions provided to the peers are descri bed below. The therapist provided reminders and instructions to the peers as needed to ensure procedural integrity of the condition.

PAGE 75

63 Free play condition Peers were instructed to fre quently talk, look at, and play with the participant. The peers were also instructed to withhold c onsequences (e.g., attention, tangible, escape) contingent on participant peer-related withdr awn or peer-related nega tive social behavior. As needed, the therapist provide d peers with reminders to tal k, look at, and play with the participant, and provided examples of things to talk to the participant about, and/or ways to play with participant. Ignore condition Peers were instructed to tal k, look at, and play only with each other and to not talk, look at, or play with the part icipant. No tangible items (i.e ., toys) were present in this condition; therefore, peers were directed to pi ck a game they could play together without toys (e.g., patty-cake) prior to the beginning of the condition. The therapist assisted, as needed to identify an appropriate activity/g ame for the peers to play. As needed, the therapist provided peers with reminders to ta lk, look at, and play only with each other, examples of things to talk about with each ot her, and/or games to play with each other. Attention condition Peers were instructed to ta lk, look at, and play only w ith each other, unless the participant engaged in peer-related withdrawn or peer-r elated negative social behavior. Contingent on the participant engaging in peer -related withdrawn or peer-related negative social behavior, peers provided the participant with 15 seconds of atte ntion in the form of verbal reprimands and/or reque sts to stay and play near th em (e.g., “No!,” “Stop leaving all the time,” “We want you to play by us”). The therapist kept track of the 15 second contingent attention using a tim er and told the peers when the 15 seconds had elapsed. As needed, the therapist provided peers with re minders to talk, look at, and play only with

PAGE 76

64 each other, to provide contingent attention to the participant, and when to stop providing contingent attention. Tangible condition Peers were instructed to frequently talk about the participan t's preferred tangible item (selected in Phase 2 duri ng the preference assessment), l ook at the participant, and play with the preferred tangibl e, unless the participant engage d in peer-related withdrawn or peer-related negative social behavior. Contingent on the participant engaging in peerrelated withdrawn or peer-relat ed negative social behavior, p eers provided the participant with 15 seconds access to the preferred tangi ble. The therapist kept track of the 15 seconds contingent access to the preferred item using a timer and told the peers when the 15 seconds had elapsed. As needed, the th erapist provided peers with reminders to frequently talk, look at, and play with part icipant, to provide the participant with contingent access to his/her preferred ta ngible, to provide noncontingenet access to neutral tangibles, and when to stop providi ng the participant access to the preferred tangible. Escape condition Peers were instructed to present the part icipant with social demands (selected by the therapist based on information obtained in Phase 1) using a two or three step prompt sequence every 15 seconds, unless the particip ant engaged in peer-r elated withdrawn or peer-related negative social be havior. Verbal social demands (e.g., “Say hi to me”) were presented using a two step prompt sequence (v erbal, model) and gestural social demands (e.g., “Wave to me”) were presented using a th ree step prompt sequence (verbal, model, guide). Contingent on the part icipant correctly completing the social demand within 3 seconds of either a verbal or model prompt, the peers provided brief verbal praise (e.g.,

PAGE 77

65 “Good”) to the participant. Contingent on the participant enga ging in peer-related withdrawn or peer-related negative social be havior, peers told the participant “OK, you don't have to” and provided the participant wi th a 15 second break from social demands. The therapist kept track of th e 15 seconds contingent escape from social demands using a timer and told the peers when the 15 seconds had elapsed. The therapist provided peers with the social demands to present to the pa rticipant and the sequenc e of prompts to use, reminders to provide the part icipant with contingent esca pe from social demands, and when to present the next social demand. Functional Analysis Prior to each condition the participant was directed/guided to stand over the taped line (i.e., one foot in the in area and one foot in the wit hdrawn area) where the therapist told the participant the rules for the relevant condition. The th erapist then directed/guided the participant to the in area prio r to the start of the condition. Free play condition Prior to the free play condition the therapist reviewed the rules with the participant, “You can play here (pointing to withdrawn area) or here ( pointing to in area). Your friends will talk and play with you no matte r where you go or what you do.” The therapist then directed/guided the participant to sit in the in area next to the two peers. The peers were seated next to each othe r so that when the participant sat, a semicircle was created. Therefore, the condition began with the peer s and the participant f acing each other. The participant had continuous access to neutral tang ibles and access to peer attention at least every 15 seconds. No peer consequences we re provided for participant peer-related withdrawn or peer-related ne gative social behavior.

PAGE 78

66 Ignore condition Prior to the ignore condition the therapist re viewed the rules with the participant, “You can play here (pointing to withdrawn area) or here (pointing to in area), your friends are only going to talk and play w ith each other no matter where you go or what you do.” The therapist then direct ed/guided the participant to sit in the in area, at the point closest to the withdrawn area, sitting sideways to the peers. The two peers sat across from each other in the in area, at the point furthest away from the withdrawn area. No tangibles were available for the particip ant or peers and no p eer consequences or attention was provided for part icipant peer-related withdrawn, negative, or appropriate social behavior. Attention condition Prior to the attention conditi on, the therapist reviewed the rules with the participant, “You can play here (pointing to withdrawn area) or here (poi nting to in area). Right now it is your friends turn to talk only with each other, but if you go over here (pointing to withdrawn area) or if you (peer-related negative social behavior) your friends will play with you.” Participant peer-rel ated negative social behavior was individually defined for each participant and stated specifically, such as, “You need to whine, play with me.” After the rules were given, the therapist then directed/guided the par ticipant to sit in the in area, at the point closest to the withdrawn area, sitting si deways to the peers. The two peers sat across from each othe r in the in area, at the poi nt furthest away from the withdrawn area. Neutral tangibles were ava ilable to the participant and peers throughout the condition. The neutral tangibles in the in ar ea were split in half prior to the condition, so that the participant and p eers would have toys near th em at the beginning of the condition (i.e., toys were located by the peers and where the participant would be seated).

PAGE 79

67 Peers provided 15 seconds of attention in the form of verbal reprimands and requests to stay and play near them contingent on part icipant peer-related w ithdrawn or peer-related negative social behavior. If the participant was in the withdrawn area after the 15 seconds of contingent attention, the th erapist neutrally dir ected/guided the par ticipant to the in area. No attention was provided for participan t peer-related appropria te social behavior. Tangible condition Prior to the tangible condition the part icipant had access to the high preferred tangible for 2 minutes. The participant was then directed/guided to the line while still maintaining access to the high preferred tangibl e (i.e., holding on to it) as the therapist reviewed the rules with the pa rticipant, “You can play here (pointing to withdrawn area) or here (pointing to in area). Right now it is your friends turn to play with (high preferred tangible), but if you go over here (pointing to withdrawn area) or (peer-related negative social behavior) you can play with (high pref erred tangible).” Part icipant high preference tangible was defined individua lly based on preference assess ment results and stated specifically, such as “play-doh.” Participant peer-related ne gative social behavior was defined individually and stated sp ecifically, such as, “Yell, I want to play with it.” After the rules were given, the therapist directed/guide d the participant to the in area next to the two peers. The condition began immediatel y following the peer removing the high preferred tangible from the par ticipant, while stating, “It is our turn.” and then gave the participant a neutral tangible The participant had continuous access to neutral tangibles and access to peer attention at least ever y 15 seconds. Peers played with the high preferred tangible unless the participant e ngaged in peer-related withdrawn or peerrelated negative social behavi or. Peers provided the participant with 15 seconds access to the high preferred tangible contingent on peer -related withdrawn or peer-related negative

PAGE 80

68 social behavior. After the 15 seconds access to the high preferred tangible, the peers removed the high preferred tangible from the pa rticipant, while stati ng, “It is our turn,” and then gave the participant a neutral tang ible. If the participant was in the withdrawn area after the 15 seconds of c ontingent access to the high pref erred tangible, the therapist neutrally directed/guided the participant to the in area af ter the 15 seconds. The peers then removed the high preferred tangible and replaced it with a neutral tangible. No access to the high preferred tangi ble was provided for participan t peer-related appropriate social behavior Escape condition Prior to the escape condition, th e therapist reviewed the ru les with the participant, “You can play here (pointing to withdrawn area) or here (poi nting to in area). If you play here (pointing to in area) your friends are going to tell you how to play. If you don't want to do what your friends tell you to do, you can (peer-related negative social behavior) or go over here (pointing to wit hdrawn area) and play how you wa nt to.” Participant peerrelated negative social behavior was defined individually and st ated specifically, such as, “Yell, no!” After the rules were given, the ther apist directed/guided the participant to the in area between the two peers. The particip ant had continuous access to neutral tangibles and access to peer attention at least every 15 seconds. Peers presented the participant with social demands every 15 seconds using a two prompt sequence (verbal, model) for social demands that required a verbal response a nd a three prompt sequence (verbal, model, guide) for social demands that required a gestural response (e.g., “Turn the page”). Participant social demands were define d individually and based on descriptive observations conducted in Phase 1. Contingent on participant peer-r elated withdrawn or peer-related negative social behavior, peers stated, “OK, you don't have to” and provided

PAGE 81

69 15 seconds of escape from social demands. If the participant did not complete the social demand or did not complete the social demand correctly, the social demand was repeated again (for a total of two pr esentations) prior to presen ting the next social demand. Participant compliance with social demands resulted in peers providing brief neutral praise (e.g., “Thanks for talking to me,” “OK,” “Good”) and continued presentation of social demands every 15 seconds. If the part icipant was in the wit hdrawn area after the 15 seconds of contingent escape from social dema nds, the therapist neut rally directed/guided the participant to the in area after the 15 seconds and the pe ers presented a social demand. Escape was not provided for participant pee r-related appropriate social behavior. Second Functional Analysis A second functional analysis was conducted with Shannon only. In the first functional analysis, Shannon did not engage in peer-related withdr awn social behavior, thus this behavior could not be evaluated. In addition, Shannon engaged in low to zero rates per minute of peer-related negative so cial behavior, resulti ng in inconclusive findings. As a result, a second functional anal ysis was conducted. The procedures in the second analysis were similar to those in th e first functional analys is, though altered in attempts to reduce response effort to engage in peer-related wit hdrawn social behavior. Shannon’s second functional analysis was c onducted in the same setting and with the same peers as the first functional analysis. Approximately three 5 minute conditions were conducted during each 30 minute session. Condi tions included free play, escape, tangible, and attention. The ignore condition was not conducted, as Shannon did not engage in peer-related withdrawn or peer-re lated negative social behavior in the free play or ignore conditions of the first functional analysis. Sess ions were conducted one to three times per

PAGE 82

70 week. Reinforcement was provided for 15 sec onds contingent on peer-related withdrawn or peer-related negative social behavior in the attention, ta ngible, and escape conditions. The free play, attention, and escape cond itions were conducted exactly the same as described in the first functional analysis with one exception. The only difference was every 15 seconds the therapist directed/guide d Shannon to stand over the taped line (i.e., one foot in the in area and one foot in the withdrawn area) and asked her, “Which side do you want to go to?” Only Shannon’s physical choice (i.e., which side she entered) was recorded. When standing over the taped line, the response effort to enter the withdrawn and in area was seemingly equal, as both required one step to e ither her right or left side. This differed from the first functional anal ysis when the conditions began with Shannon sitting in the in area, which required Shannon to stand up and walk at least two steps to enter the withdrawn area. Pr actically, Shannon could have moved her body from the in area to the withdrawn area w ithout standing (e.g., crawling) though the number of motor movements required for her to enter the wit hdrawn area would have remained higher than the one step required in the second functi onal analysis. Shannon was only observed to enter the withdrawn or in area by walking. The tangible condition differed from the first functional analysis in two ways. First, every 15 seconds the therapist directed/guide d Shannon to stand over the taped line (i.e., one foot in the in area and one foot in the withdrawn area) with the high preferred tangible in her hand, and asked her, “Which side do you want to go to?” Only Shannon’s physical choice (i.e., which side she entered) was the reco rded. When standing over the taped line, the response effort to either en ter the withdrawn or in area was seemingly equal, as both required one step to either her right or left side Secondly, peers did not

PAGE 83

71 deliver the consequence (i.e., access to high preferred item) for peer-related withdrawn social behavior. If Shannon entered the in ar ea, the therapist directed/guided her to give the high preferred tangible to her peers. P eers then played with the high preferred tangible. Contingent on peer-r elated negative social beha vior, peers provided Shannon access to the high preferred tangible for 15 s econds. However, if Shannon entered the withdrawn area, she simply maintained access to the high preferred tangible. Therefore, peers did not deliver the consequence (i.e., a ccess to high preferred tangible) contingent on peer-related withdrawn social behavior. This procedure was different from the first functional analysis when Shannon was require d to engage in peer-related withdrawn social behavior prior to receiving the conse quence (i.e., high preferred tangible) and the peers delivered the consequence rather than Shannon maintaining access to the high preferred tangible from st anding at the line.

PAGE 84

72 CHAPTER 3 RESULTS Overview of Research Aims The purpose of this investigation was to (a) evaluate the effectiveness of the use of the functional analysis methodology to determin e the function of pee r-related withdrawn and peer-related negative social behavior for three young participan ts with ASD, (b) compare the results of a functional analysis of peer-related withdrawn and peer-related negative social behavior acro ss participants with ASD to determine if the functions differed across participants, (c) to determine if peers were able to implement the functional analysis procedures with adequate integrity, and (d) to examine rate per minute of participant peer-related positive initiations and percentage of participant peer-related positive and no response social behavior across conditions. These research aims were addressed via three phases of assessment. During Phase 1, a list of participant and peer initiations a nd a list of tangibles (e.g., toys) was generated for each participant based on descriptive assessment. Data from Phase 1 were obtained by observing videotapes from 5 (Shane a nd Colin) to 6 (Shannon) hours of direct observation and listing the topograp hy of participant and peer in itiations and the tangibles the participants engaged with. The initial li st of peer-related initiations was then narrowed down, according to several criteria, to 4 to 11 initiations, to be used as social demands for the escape condition of the functi onal analysis (Phase 3). The initial list of tangibles was narrowed down to seven, according to the ability to be manipulated by more than one child and the feasibility of obtaining at least two id entical tangibles, for

PAGE 85

73 use in the preference assessment (Phase 2). During Phase 2, seven tangibles were assessed for preference and then rank ordered (high, neutral, and low) for use in the functional analysis (Phase 3). The tangible(s) chosen on the hi ghest percentage(s) of trials was ranked as high preferred and used dur ing the tangible condition in Phase 3. The tangible(s) ranked as neutral were used dur ing the free play, atte ntion, tangible, and escape conditions in Phase 3. During Phase 3, a functional analysis based on Iwata et al. (1982/1994) of peer-related withdrawn and p eer-related negative social behavior was conducted. The functional analysis included fr ee play, attention, ta ngible, escape, and ignore conditions. Data from Phase 3 were analyzed using visual analysis. Phase One: Descriptive Assessment Results The descriptive assessment consisted of identifying 4 to 11 social demands and seven tangibles for each participant for use in Phases 2 and 3. The results of Phase 1 are presented separately for each participant. Shane Descriptive Assessment Results Social demands identified During direct observation, Sh ane and his peers were observed to engage in a total of 14 peer-rel ated verbal initiations and 17 peer-related nonverbal (gestural) initiations (see Appe ndix C-1, Shane’s Verbal and Gestural Initiations). Of those initiations six were selected as social demands and were used in the escape conditions during Phase 3 (as described in Chapter 2) The seven social demands selected for use in the escape condition of Phase 3 were: 1) “T urn the page,” 2) “Press the button,” 3) “Look at me,” 4) “Touch my shoul der,” 5) “Hold my hand,” and 6) “Give me a high five” (see Table 3-1, Social Demands Used in Escape Condition of Functional Analysis).

PAGE 86

74 Tangibles identified The tangibles identified for Sh ane via direct observation are listed in Table 3-2. One point that should be noted is Shane typically engaged with these tangibles in a stereotypical rather than a functional manner. It was hypothesized that when Shane engaged in stereotypy he was le ss likely to engage in social behavior regardless of environmental consequences (i .e., stereotypy may have been automatically reinforced and competed with social behavi or). Therefore, the procedures to identify tangibles for use in the preference assessm ent were supplemented with descriptive assessments in attempts to identify tangibl es Shane would engage with in a functional manner. The additional assessment included Sh ane’s teacher identifying tangibles Shane had been observed to engage with in a f unctional manner. Shane’s teacher identified the ABC Phonics, sound-effect books, and a plastic ball. Of these tangibles, seven were selected, using direct observation procedures (as described in Chapter 2) and information obtained from supplementary descriptive asse ssments. The seven tangibles selected for use in the preference assessment (Phase 2) were: 1) Scooby-do book, 2) foam fish, 3) Thomas the Engine book, 4) plastic ball, 5) ABC Phonics, 6) Bob the Builder book, and 7) Lego’s (see Table 3-3, Ta ngibles Included in Preference Assessment). Of note, the books identified via direct observation did not include Scooby-do, Thomas the Engine, or Bob the Builder, as those books were not pres ent in the classroom. The classroom had a limited number of books with sound effects and Shane had a history of bending and breaking books. Therefore, based on the teach er’s recommendation, these specific books were purchased for use in the preference assessment. Shannon Descriptive Assessment Results Social demands identified During direct observation, Shannon and her peers were observed to engage in a total of 63 peer-rel ated verbal initiations and 41 peer-related

PAGE 87

75 nonverbal (gestural) initiations (see Appe ndix C-2, Shannon’s Verb al and Gestural Initiations). Of those initiations, 11 were select ed as social demands and were used in the escape conditions during Phase 3. The 11 social demands selected for use in the escape condition of Phase 3 were: 1) “Pick a dinosau r,” 2) “Show me what you are making,” 3) “Hand me the ___ (e.g., block),” 4) “Look at ___ (e.g., me),” 5) “ Put the ____ (e.g., block) here,” (where the peer points) 6) “Help me make a snake with ___ (e.g., the playdoh), you roll this side,” (where the peer points) 7) “Put your hands on your lap,” 8) “Let's play patty cake, come sit by me,” 9) “Say hi,” 10) “Let's sing a song, sing with me, Shannon,” and 11) “Tell me ____ (e.g., what you are making)” (see Table 3-1, Social Demands Used in Escape Conditi on of Functional Analysis). Tangibles identified The tangibles identified fo r Shannon via direct observation are listed in Table 3-2. Of these tangibles, seven were selecte d. The seven tangibles selected for use in the preference assessmen t (Phase 2) were: 1) T-Rex dinosaur, 2) Barbie dolls, 3) play cash re gister, 4) small dinosaurs, 5) wooden blocks, 6) play-doh, and 7) Dr. Seuss book (see Table 3-3, Tangibles Included in Preference Assessment). Colin Descriptive Assessment Results Social demands identified During direct observation, Colin and his peers were observed to engage in a total of 12 peer-relat ed verbal initiations and three peer-related nonverbal (gestural) initiations (see Appe ndix C-3, Colin’s Verbal and Gestural Initiations). Of those initiations four were selected as soci al demands and were used in the escape condition during Phase 3. The four social demands selected for use in the escape conditions of Phase 3, were: 1) “Look at me,” 2) “Hand me the ___ (e.g., toy),” 3) “Take this ___ (e.g., toy),” and 4) “Give me a high five” (see Table 3-1, Social Demands Used in Escape Condition of Functional Analysis).

PAGE 88

76 Tangibles identified The tangibles identified for Colin via direct observation are listed in Table 3-2. Of these tangibles, seven were selected. The seven tangibles selected for use in the preference assessment (Phase 2) were: 1) drawing, 2) Magnadoodle, 3) Mr. Potato Head, 4) puzzle, 5) pl ay-doh, 6) wooden blocks, and 7) blocks on a strings (see Table 3-3, Tangibles Included in Preference Assessment). Table 3-1. Social Demands Used in Es cape Condition of Functional Analysis Shane Shannon Colin Turn the page Pick a dinosaur Look at me Press the button Show me what you are making Hand me the ___ (e.g., toy) Look at me Hand me the ___ (e.g., block) Take this ___ (e.g., toy) Touch my shoulder Look at ___ (e.g., me) Give me a high five Hold my hand Put the ___ (e.g., block) here (where the peer points) Give me a high five Help me make a snake with ___(e.g., play-doh), you roll this side (where the peer points) Put your hands on your lap Let’s play patty cake, come sit by me Say hi Let’s sing a song, sing with me, Shannon Tell me ___ (e.g., what you are making) Table 3-2. Tangibles Identified During Descriptive Assessment Shane Shannon Colin Foam alphabet letters Pr etend kitchenware (e.g., food, plates, spoons) Computer game Bean bag Pretend physician tools (e.g., blood pressure monitor) Listening to music Plastic tools (e.g., pliers) Art activities (e.g., gluing, painting, drawing on paper with pens, crayons, and colored and pencils) Dancing to music Play-doh Cutting with scissors Hair rollers

PAGE 89

77 Table 3-2. Continued Shane Shannon Colin Play stethoscope Matching cards by color, insects, and shape Blocks on strings Yellow spoon Sorting by colors Puzzles Wooden blocks Puzzles Drawing with markers and crayons Small green dinosaur Play-doh Large and small plastic blocks Mr. Potato Head arm Plastic necklaces Wooden blocks Small yellow glasses Plastic and w ooden blocks Books (looking at and listening) Clip Plastic animals (e.g., turtle, giraffe, lion, horses, small dinosaurs, T-rex) Magnadoodle Small plastic dinosaurs Musical instruments (e.g., maracas, tambourine) Play-doh Plastic snake Looking at a book Blue toy car Plastic dragon fly Barbie’s Mr. Potato Head Beads Plastic bottle Books Water Yellow foam fish Barbie doll Lego’s Plastic cookware and food Big Lego’s Plastic tools Play phone Table 3-3. Tangibles Include d in Preference Assessment Shane Shannon Colin Scooby-do book T-rex dinosaur Drawing Foam fish Barbie dolls Magnadoodle Thomas the Engine book Play cash register Mr. Potato Head Plastic ball Small dinosaurs Puzzle ABC Phonic Wooden blocks Play-doh Bob the Builder book Play-doh Wooden blocks Lego’s Dr. Seuss book Blocks on strings Phase Two: Preference Assessment Results During Phase 2, preference for the seven ta ngibles selected for each participant during the descriptive assessment was evalua ted. Criteria for identifying a high preferred tangible(s) included having the hi ghest ranking (i.e., highest per centage of trials chosen). If there was a tie between two tangibles for the highest ranking or if the two highest

PAGE 90

78 rankings were within 10% of each other, bot h tangibles were cons idered high preferred. Criteria for identifying neutral preferred ta ngible(s) included having a ranking between second and fifth. Criteria for identifying lo w preferred tangible( s) included having a ranking between fourth and seventh. Tangibles were not ranked if they did not meet the above criteria (e.g., if there was more than a 10% difference between the tangible chosen the first and second highest per centage of trials). The resu lts of Phase 2 are presented separately for each participant. Shane Preference Assessment Results Five preference assessment sessions were conducted with Shane. Shane’s high preferred tangibles were the Scooby-do book and the foam fish, chosen 42% and 33% of the trials, respectively (see Table 3-4, Shan e’s Rank Order of Stimuli Obtained from Preference Assessment). His neutral preferre d tangibles were the Thomas the Engine book, the plastic ball, and the ABC Phonics, chosen 24%, 20%, and 18% of trials, respectively. Shane’s low pref erred tangibles were the Bob the Builder book and Lego’s, chosen 14% and 11% of the trials, respectively. The high preference stimuli used in the tangible conditions of Shane’s functional analysis were the Scooby-do book and the fo am fish. The Scooby-do book and foam fish were rotated across tangible conditions so that only one high pr eferred tangible was available in each tangible condition. The neutra l tangibles available in Shane’s free play, attention, tangible, and escape conditions were the Thomas the Engine book and the ABC Phonics. The plastic ball was identified as a neutral preferred tangible; however, each time Shane chose the plastic ball during the preference assessment, he threw it out of the assessment area. As a result of the distra ction this caused to the ongoing classroom activities, the plastic ball was not included in any of th e functional analysis conditions.

PAGE 91

79 Table 3-4. Shane’s Rank Order of Stimuli Obtained from Preference Assessment Stimuli Percentage of Trials Chosen Rank Functional Analysis Conditions Stimuli Available Scooby-do book 42 High preferred Tangible (rotated) Foam fish 33 High prefer red Tangible (rotated) Thomas the Engine book 24 Neutral preferred Free play, attention, tangible, escape Plastic ball 20 Neutral preferred None ABC Phonics 18 Neutral preferred Free play, attention, tangible, escape Bob the Builder book 14 Low preferred None Lego’s 11 Low preferred None Shannon Preference Assessment Results Ten preference assessment sessions were conducted with Shannon; five sessions were conducted in the first preference a ssessment and five sessions in the second preference assessment. First preference assessment results In the first preference assessment, Sha nnon’s high preferred tangible was the T-rex dinosaur, chosen 83% of the trials (see Table 3-5, Shannon’s Rank Order of Stimuli Obtained from First Preference Assessment). Her neutral preferred tangibles were the play cash register, small dinosaurs, and w ooden blocks, chosen 28%, 23%, and 20% of the trials, respectively. Shannon’s low prefe rred tangibles were play-doh and the Dr. Seuss book, chosen 18% and 16% of the trials respectively. The tangible not ranked was the Barbie dolls, chosen 56% of the trials. In the first preference assessment, the high preference stimulus used in the tangible conditions of Shannon’s functional analysis wa s the T-rex dinosaur. The neutral tangibles available in Shannon’s free pla y, attention, tangible, and esca pe conditions were the small dinosaurs and the wooden blocks. The play ca sh register, ranked high preferred, was not

PAGE 92

80 included in the tangible conditions due to one of the two play cash registers breaking prior to the beginning of the functional analysis (i.e., two play cash re gisters were needed to conduct a tangible condition a nd only one was available). The results of the first preference assessm ent were utilized in functional analysis conditions 1 through 15. Toward the end of the first functional analysis there was concern that the high preferred tangible was no longer high preferred. Therefore, a second preference assessment was conducte d prior to condition number 16. Table 3-5. Shannon’s Rank Order of Stimuli Obtained from First Preference Assessment Stimuli Percentage of Trials Chosen Rank Functional Analysis Conditions Stimuli Available T-rex dinosaur 83 High preferred Tangible Barbie dolls 56 Not ranked None Play cash register 28 Neutral preferred None Small dinosaurs 23 Neutral preferred Free play, attention, tangible, escape Wooden blocks 20 Neutral preferred Free play, attention, tangible, escape Play-doh 18 Low preferred None Dr. Seuss book 16 Low preferred None Second preference assessment results In the second preference assessment, Sh annon’s high preferred tangibles were the play cash register and play-doh, both chosen 45% of the trials (see Table 3-6, Shannon’s Rank Order of Stimuli Obtained from Second Preference Assessment). Her neutral preferred tangible was the Ba rbie dolls, chosen 29% of the trials. Shannon’s low preferred tangibles were the wooden blocks the small dinosaurs, and the Dr. Seuss book, all chosen 17% of the trials. The tangible no t ranked was the T-rex dinosaur, chosen 42% of the trials.

PAGE 93

81 In the second preference assessment, the high preference stimuli used in the tangible conditions of Shannon’s functional anal ysis were the play cash register (the second play cash register had been fixed and was available for use) and the play-doh. The tangibles were rotated across conditions so that only one high preferred tangible was available in each tangible condition. The ne utral tangible available in Shannon’s free play, attention, tangible, and escape conditions was the Barbie dolls. Table 3-6. Shannon’s Rank Or der of Stimuli Obtained from Second Preference Assessment Stimuli Percentage of Trials Chosen Rank Functional Analysis Conditions Stimuli Available Play cash register 45 High preferred Tangible (rotated) Play-doh 45 High preferred Tangible (rotated) T-rex dinosaur 42 Not ranked None Barbie dolls 29 Neutral preferred Free play, attention, tangible, escape Wooden blocks 17 Low preferred None Small dinosaurs 17 Low preferred None Dr. Seuss book 17 Low preferred None Colin Preference Assessment Results Five preference assessment sessions were conducted with Colin. Colin’s high preferred tangible was drawing, chosen 71% of the trials (see Table 3-7, Colin’s Rank Order of Stimuli Obtained from Preference A ssessment). His neutral preferred tangibles were the Magnadoodle and Mr. Potato Head, chosen 29% and 28% of the trials, respectively. Colin’s low pref erred tangibles were the puz zle, play-doh, wooden blocks, and blocks on strings, chosen 17%, 14%, 12%, and 9% of the trials, respectively. The high preference stimulus used in the tangible conditions of Colin’s functional analysis was drawing. The neutral tangibles available in Colin’s free play, attention, tangible, and escape conditions were th e Magnadoodle and Mr. Potato Head.

PAGE 94

82 Table 3-7. Colin’s Rank Order of Stim uli Obtained from Preference Assessment Stimuli Percentage of Trials Chosen Rank Functional Analysis Conditions Stimuli Available Drawing 71 High preferred Tangible Magnadoodle 29 Neutral preferred Free play, attention, tangible, escape Mr. Potato Head 28 Neutral pr eferred Free play, attention, tangible, escape Puzzle 17 Low preferred None Play-doh 14 Low preferred None Wooden blocks 12 Low preferred None Blocks on strings 9 Low preferred None Phase Three: Functional Analysis Results Functional analyses of peer-related wit hdrawn and peer-related negative social behavior were conducted with each participant incorporating social demands in the escape condition (see Phase 1 descriptive assess ment results). The tangible(s) identified as high preferred, via the preference assessmen t, were incorporated into the tangible condition (see Phase 2 preference assessment results). The tangibl es identified as neutrally preferred via the preference assessmen t (with the exception of the plastic ball in Shane’s preference assessment and play cas h register in Shannon’s first preference assessment), were incorporated into the free play, attention, escape, and tangible conditions. Data were graphically displayed and analyzed for each participant. Peerrelated withdrawn and peer-related negative so cial behaviors were graphically displayed using two different methods for each participan t. First, graphs were created for each behavior displaying the rate of that beha vior per minute in each functional analysis condition. Second, the rate per minute of peer-related withdr awn and peer-related negative social behaviors were combined post hoc and graphed for each functional analysis condition. Throughout each 5 minute f unctional analysis condition, participants

PAGE 95

83 had the opportunity to engage in peer-relate d withdrawn and peer-related negative social behaviors, with each behavior resulting in the same consequence in the given condition (e.g., in the attention condition, attention wa s provided contingent on both peer-related withdrawn and peer-related nega tive social behavior). At some point in the functional analysis conditions, 2 of the 3 participant’s (i.e., Shane and Colin) engaged in both social behaviors (peer-related withdr awn and peer-related negati ve) during the same 5 minute condition. Therefore, the behaviors were combin ed post-hoc to displa y the total rate per minute of combined peer-relate d (withdrawn and nega tive) social behaviors displayed for each functional analysis condition. These data addressed the first and second research questions “Can the functional analysis me thodology of evaluating pr oblem behavior be used to identify the function of peer-relate d withdrawn and peer-rel ated negative social behavior?” and “Are the functi ons of peer-related withdraw n and peer-related negative social behaviors different across young childre n with ASD?” Tables were generated to present the mean rate per minute and ranges of each peer-related social behavior across functional analysis conditions and to pr ovide a comparison between participants. Next, peer integrity behavior s were analyzed. Tables were generated to present the mean rate per minute and ranges of peer corre ct presentation of social demands in the escape condition, average pe rcentage and ranges of correct pe er delivery of consequences in the tangible, attention, and escape conditi ons, and mean rate per minute and ranges of peer correct delivery of positive initiations in the free play condition. These data addressed the third research question “C an peers implement functional analysis procedures with accep table integrity?”

PAGE 96

84 Finally, tables were generated to present the mean rate per minute and ranges of participant peer-related positive initiations, average percentage and ranges of participant peer-related positive responses, and average percentage and ranges of participant peerrelated no response social behaviors. These data addressed the final research aim “To examine rate per minute of participant peer-re lated positive initiations and percentage of participant peer-related positive and no response social behavior across conditions.” The results from Phase 3 are presented separately for each participant. Shane Functional Analysis Results Shane peer-related withdrawn and peer-related negative behaviors The rate per minute of peer-related wit hdrawn social behavior (top panel), peerrelated negative social behavior (middle pa nel), and the combination of peer-related withdrawn and negative social behaviors (bottom panel) from Shane's functional analysis are displayed in Figure 3-1. The mean rate per minute and ranges of peer-related withdrawn, peer-related negative, and the co mbination of peer-related withdrawn and negative social behaviors are displayed in Table 3-8. Shane exhibited a three point increasing trend of peer-relat ed withdrawn social behavior in the ignore and tangible conditions. The ignore and ta ngible conditions ended on a three point upward trend, though peer-related withdrawn social beha vior was variable throughout the tangible condition, with 4 of 10 conditions with zer o rates per minute of withdrawn social behavior. The attention cond ition also began on an increa sing trend, though ended on a three point downward trend. Peer-related withdr awn social behavior was exhibited in 3 of the 4 escape conditions, though the rates per mi nute were low (0.2/minu te to 0.4/minute) and a stable three point tre nd (i.e., upward/downward/stable ) was not established. With the exception of one condition, peer-related w ithdrawn social behavi or occurred at zero

PAGE 97

85 rates per minute in the free play condition. Shane’s peer-related withdrawn social behavior was variable and ove rlapped between all functional analysis conditions. Visual analysis suggests that Shane’s peer-related withdrawn social beha vior (Figure 3-1, top panel) maybe multiply maintained by automatic reinforcement and positive reinforcement in the form of access to prefer red tangibles, though due to variable rates of peer-related withdrawn social behavior across all conditions the results may best be described as inconclusive. Shane exhibited a three point upward/stable trend of peer-related negative social behavior in the ignore and tangible cond itions, though the ignore condition ended on a downward point (i.e., zero rates per minute of peer-related negative social behavior were displayed in the final ignore condition). Th e tangible condition ended on a three point upward/stable trend, though peer-r elated negative social behavior occurred at low to zero rates per minute in the first seven tangible conditions, with 6 of 10 conditions with zero rates per minute of negative social behavior. Peer-related negative social behavior was displayed in 3 of 4 escape conditions, though the first three data points established a downward trend. In the final escape conditi on, peer-related negative social behavior increased to 0.4/minute, though a three point upward trend was not established and there was overlap between ignore, attention, and tangible conditions. Low to zero rates per minute of peer-related negative social behavi or were observed in the attention and free play conditions. Visual analysis suggests that Shane’s peer-related negative social behavior (Figure 3-1, middle panel) maybe ma intained by positive reinforcement in the form of access to preferred tangibles. However, due to variable rates of peer-related negative social behavior in th e ignore and escape conditions a nd low and variable rates of

PAGE 98

86 peer-related negative social be havior in the tangible conditi on, the results may best be described as inconclusive. Shane combined peer-related withdraw n and peer-related negative behaviors As noted previously, the opportunity to e ngage in both peer-re lated withdrawn and peer-related negative social behaviors wa s available across all functional analysis conditions and during the escape, tangible, and attention conditions, reinforcement was delivered contingent on the occurrence of eith er peer-related wit hdrawn or peer-related negative social behavior. For example, in tangible condition number 23, Shane engaged in one instance of peer-related withdrawn soci al behavior and one instance of peer-related negative social behavior. The occurrence of peer-related withdrawn social behavior was reinforced with 15 seconds access to his hi gh preferred tangible and the occurrence of peer-related negative social be havior was reinforced with 15 seconds access to his high preferred tangible, for a total of two reinforcement deliveries in that tangible condition. Visual analysis of Shane’s peer-related w ithdrawn (Figure 3-1, top panel) and peerrelated negative social behavi or (Figure 3-1, middle panel) indicated both peer-related withdrawn and peer-related ne gative social behaviors occu rred in the same functional analysis conditions. Specifically, when peer -related withdrawn or peer-related negative social behavior occurred, both occurred in 100% (3 of 3) escape (i.e., peer-related withdrawn social behavior and peer-related negative social behavior occurred in escape condition numbers 9, 16, and 24; no peer-related withdrawn or negative social behavior occurred in escape condition number 22), 67% (4 of 6) tangible, and 50% (2 of 4) of ignore conditions. Therefore, when peer-rel ated withdrawn (or peer-related negative) social behavior occurred in the escape condi tion, peer-related negative (or peer-related withdrawn) social behavior also occurred in 100% of escap e conditions. Similarly, when

PAGE 99

87 peer-related withdrawn (or peer-related negati ve) social behavior occurred peer-related negative (or peer-related withdrawn) social behavior also occurred in 67% of tangible conditions and 50% of ignore conditions. Peer-related withdraw n and peer-related negative social behaviors did not occur in th e same condition for attention and free play conditions. Shane exhibited a three poi nt upward trend of combin ed peer-related (withdrawn and negative) social behavior in the i gnore and tangible condi tions. The ignore and tangible conditions ended on a three point upw ard trend, though combined peer-related (withdrawn and negative) social behavior was variable thr oughout the tangible condition, with 4 of 10 conditions with zero rates per minute of combined peer-related (withdrawn and negative) social behavior. Combined p eer-related (withdrawn and negative) social behavior was initially elevated in the a ttention condition, though ended on a decreasing and variable trend. Combined peer-related (w ithdrawn and negative) social behavior was exhibited in 3 of the 4 escape conditions, though the rates per minute were variable (2.4/minute to 0.6/minute) and a stable three point trend was not established. With the exception of one condition, combined peer-r elated (withdrawn and negative) social behavior occurred at zero rates per minute in the free play condition. Visual analysis suggests that Shane’s combined peer-related (withdrawn and negative) social behavior (Figure 3-1, bottom panel) maybe multiply maintained by automatic reinforcement and positive reinforcement in the form of access to tangibles. However, due to variable rates of combined peer-related (withdrawn and ne gative) social behavi or in the tangible, attention, and escape conditions, the results may best be described as inconclusive.

PAGE 100

88 0 0.4 0.8 1.2 1.6 2 2.4 13579111315171921232527Rate per min Negative 0 0.5 1 1.5 2 2.5 3 13579111315171921232527 Attention Escape Tangible Free Play I g noreRate per min Withdrawn 0 0.5 1 1.5 2 2.5 313579111315171921232527Rate per min Withdrawn an d NegativeSessions Figure 3-1. Rate per minute peer-related w ithdrawn (top panel), negative (middle panel), and combined social behavior (bottom panel) for Shane.

PAGE 101

89 Table 3-8. Shane’s Mean Rate and Range s per minute of Participant Peer-related Withdrawn, Negative, and Combined (Withdrawn and Negative) Behavior Condition Withdrawn Negative Combined Free play 0.55 (0 to 2.22) 0 0.55 (0 to 2.22) Tangible 0.34 (0 to 1.6) 0.12 (0 to 0.4) 0.46 (0 to 2.0) Attention 0.48 (0 to 1.2) 0.08 (0 to 0.4) 0.56 (0.2 to 1.2) Escape 0.2 (0 to 0.4) 0.75 (0 to 2.2) 0.95 (0 to 2.4) Ignore 0.85 (0.2 to 2.6) 0.4 (0 to 0.8) 1.25 (0.2 to 2.6) Shane peer-integrity Shane’s peer correct presentation of social demands occurred at a mean rate of 3.4/minute in the escape condition, very close to the goal mean rate of 4.0/minute and within the 80% criteria range (see Table 39). Peer correct delivery of consequences averaged the highest in free play (M = 100%) and ignore conditions (M = 100%). Lower average percentages of peer correct delivery of consequences occurred in the attention (M = 88%), tangible (M = 71%), and escape c onditions (M = 65%) (see Table 3-10). Peer positive initiations occurred at a mean rate of 3.4/minute in the free play condition, very close to the goal mean rate of 4.0/minute and within the 80% criter ia range (see Table 311). In general (with the exception of the tangible and escape conditions) these data indicate that Shane’s peers’ were able to im plement the procedures (i.e., correctly present social demands in the escape condition, deliver consequences contingent on peer-related withdrawn and peer-related ne gative social behavior, and present positive initiations) with acceptable integrity (i.e., met the 80% criteria range).

PAGE 102

90 Table 3-9. Shane’s Mean Rate and Ranges pe r minute of Peer Correct Presentation of Social Demands Condition Shane peer correct pr esentation of social demands Escape 3.4 (2.2 to 4.6) Table 3-10. Shane’s Average Percentage a nd Ranges of Peer Correct Delivery of Consequences Condition Shane peer correct delivery of consequences Free play 100 Tangible 71 (0 to 100) Attention 88 (38 to 100) Escape 65 (43 to 100) Ignore 100 Table 3-11. Shane’s Mean Rate and Range s per minute Peer Positive Initiations Condition Shane peer correct presentation of positive initiations Free play 3.4 (3.0 to 4.6) Shane peer-related initiations Shane’s peer-related positive initiations occu rred at a mean rate 0.15/minute in the free play condition (see Table 3-12 ). Shane engaged in peer-rel ated positive initiations in only 1 of the 27 functional analysis conditions (i.e., free play condi tion). He engaged in zero rates per minute of peer-related positive in itiations in the tangible, attention, escape, and ignore conditions. Shane’s peer-related nega tive initiations occurr ed at a mean rate 0.08/minute in the tangible condition and mean rate 0.15/minute in the escape conditions (see Table 3-12). He did not di splay peer-related negative in itiations in the free play, attention, or ignore conditions.

PAGE 103

91 Table 3-12. Shane’s Mean Rate and Range s per minute of Part icipant Peer-related Positive and Negative Initiations Condition Positive initiations Negative initiations Free play 0.15 (0 to 0.6) 0 Tangible 0 0.08 (0 to 0.4) Attention 0 0 Escape 0 0.15 (0 to 0.4) Ignore 0 0 Shane peer-related responses Shane’s highest average percentage of peer-related positive response occurred in the escape condition (M = 33%). Lower aver age percentages of peer-related positive responses occurred in the tangible (M = 16%), attention (M = 10%), free play (M = 5%), and ignore (M = 0%) conditions (see Table 3-13 ). Shane’s highest average percentage of peer-related negative response occurred in the attention condition (M = 8%). Lower average percentages occurred in the escape (M = 4%), tangible (M = 1%), and free play (M = 0%) conditions (see Table 3-13). Shan e’s highest average pe rcentage of peerrelated no response occurred in the free pl ay condition (M = 95%). Lower average percentages of peer-related no response occurred in the attention (M = 82%), tangible (M =80%), and escape (64%) conditions (see Tabl e 3-13). In comparison to peer-related positive and peer-related negative responses, Shane engaged in peer-related no response behavior most often across f unctional analysis conditions (m ajority of all sessions Shane engaged in peer-related no response behavior at least 50% or more). It is important to note that peer-related positive, negative and no response behavior were based on Shane’s opportunities to respond to peers initiations. Pe ers did not initiate to Shane during the ignore condition, as they were in structed to only talk and play with each other. Therefore,

PAGE 104

92 Shane did not have an opportunity to engage in peer-related res ponse behavior in the ignore condition of the functional analysis. Table 3-13. Shane’s Average Percentage and Ra nges of Participant Peer-related Positive, No, and Negative Responses Condition Positive response No response Negative response Free play 5 (0 to 13) 95 (87 to 100) 0 Tangible 16 (0 to 58) 80 (47 to 100) 1 (0 to 6) Attention 10 (0 to 50) 82 (50 to 100) 8 (0 to 40) Escape 33 (6 to 48) 64 (52 to 81) 4 (0 to 9) Ignore N/A N/A N/A Summary of Shane’s functional analysis results In summary, analysis of Shane’s pee r-related withdrawn social behavior was inconclusive due to variability and overla p between conditions, though due to elevated rates per minute of peer-related withdrawn soci al behavior in the i gnore condition and the final three tangible conditions, multiple functions of automatic reinforcement and positive reinforcement in the form of access to preferred tangibles cannot be ruled out. The function of Shane’s peer-related nega tive social behavior was unclear due to variability and overlap between conditions, t hough due to elevated rates of peer-related negative social behavior per minute in the final three tangible c onditions, a function of positive reinforcement in the form of access to high-preferred tangibles cannot be ruled out. The function of Shane’s combined peer -related (withdrawn and negative) social behavior was also unclear due to variab ility and overlap between conditions, though elevated rates per minute of combined peer -related (withdrawn and negative) social behavior in the ignore condition, the final thr ee tangible conditions, and in the attention condition, multiple maintaining functions of automatic reinforcement and positive

PAGE 105

93 reinforcement in the form of both access to high preferred tangibles and access to peer attention cannot be ruled out. In addition, with the exception of the escape and tangible condition, peers generally had acceptable inte grity with implementing the functional analysis procedures (i.e., correctly presen ting social demands, con tingent consequences, and positive initiations). Finally, the additi onal data evaluated indicated Shane rarely initiated to his peers, when Shane’s peers initiated to him, Shane most often did not respond to their positive init iations or social demands. Shannon Functional Analysis Results Shannon peer-related withdrawn and peer-related negative behaviors The rate per minute of peer-related wit hdrawn social behavior (top and bottom panel) from Shannon’s first and second functi onal analysis, respectiv ely, and peer-related negative social behavior (mi ddle panel) from Shannon's first functional analysis are displayed in Figure 3-2. Initial functional analysis Shannon exhibited a three poi nt stable trend of zero rates per minute of peer-related withdrawn so cial behavior in each condition of the first functional analysis (Figure 3-2, top panel) Visual analysis suggests Shannon did not engage in peer-related withdrawn social be havior during the firs t functional analysis (Figure 3-2, top panel), therefore th e function could not be evaluated. Shannon exhibited a three point stable trend of zero rates per minute of peer-related negative social behavior in the free play, attention, escape, and ignore conditions of the initial functional analysis (Figure 3-2, mi ddle panel). Shannon’s peer-related negative social behavior occurred at variable and lo w to zero rates per minute in the tangible condition of the first functiona l analysis (Figure 3-2, middl e panel). Visual analysis suggests (Figure 3-2, middle pane l) Shannon only engaged in p eer-related negative social

PAGE 106

94 behavior in the tangible condition of the firs t functional analysis. However, due to low and variable rates per minute of peer-related negative social beha vior in the tangible condition, in addition to endi ng on a three point downward trend, the results are best described as inconclusive. Shannon did not engage in peer-related w ithdrawn social behavior in the first functional analysis (see Tabl e 3-14), therefore combined peer-related (withdrawn and negative) social behaviors were not presented graphically or in table format. As a result of not being able to evaluate a function of peer-related withdrawn social behavior and inconclusive findings of Shannon’s peer-rel ated negative social behaviors, a second functional analysis was conducted. Since Sh annon did not engage in peer-related withdrawn or peer-related negative social behavi or in the free play or ignore conditions during the first functional an alysis (see Table 3-14), the ignore condition was not conducted in the second functional analysis. Second functional analysis Shannon exhibited elevated rates per minute of peerrelated withdrawn social behavi or in the tangible condition (Figure 3-2, bottom panel) of the second functional analysis. However, the last data point in the tangible condition ended on a slight downward tre nd, preventing a three point st able trend (i.e., 0.8/minute, 0.8/minute and 0.4/minute) of peer-related wi thdrawn social behavi or in the tangible condition. Low to zero rates of peer-related w ithdrawn social behavior was displayed in the attention condition. Zero rates per minute of peer-related withdr awn social behavior were displayed in 2 of the 4 attention conditi ons. In the final atten tion condition, the rate of peer-related withdrawn social behavior increased to 0.2/minute, though a three point upward trend was not establis hed. With the exception of one escape and one free play

PAGE 107

95 condition, zero rates per minute of peer-related withdrawn social behavior occurred in the escape and free play conditions. Visual anal ysis suggests that Sh annon’s peer-related withdrawn social behavior (Figure 3-2, bottom panel) maybe maintained by positive reinforcement in the form of access to prefer red tangibles, though due to variable rates of peer-related withdrawn social behavior in the tangible and attention conditions (i.e., three point trend was not established), the result s are best described as inconclusive. Shannon combined peer-related withdraw n and peer-related negative behaviors Shannon did not engage in peer-related nega tive social behavior in any condition of the second functional anal ysis. Therefore, combined peer-related (withdrawn and negative) social behaviors were not presented graphically or in table format. The mean rate and ranges of Shannon's peer-related withdrawn and peer-rel ated negative social behavior for both functional analyses a nd each functional analysis condition, are displayed in Table 3-14. Table 3-14. Shannon’s Mean Rate and Ranges per minute of Participant Peer-related Withdrawn and Negative Behavior Condition Withdrawn (initial functional analysis [FA #1]) Negative (initial functional analysis [(FA #1]) Withdrawn (second functional analysis [FA #2]) Negative (second functional analysis [FA #2]) Free play 0 0 0.07 (0 to 0.2) 0 Tangible 0 0.15 (0 to 0.6) 1.2 (0.4 to 2.2) 0 Attention 0 0 0.25 (0 to 0.8) 0 Escape 0 0 0.07 (0 to 0.2) 0 Ignore 0 0 N/A N/A

PAGE 108

96 Shannon peer integrity Shannon’s peer correct presentation of soci al demands occurred at a mean rate of 3.53/minute in the escape condition of the firs t functional analysis, slightly below the goal mean rate of 4.0/minute (though within the 80% criteria range) and occurred at a mean rate of 2.73/minute in the escape conditi on of the second functional analysis, below the goal mean rate of 4.0/minute and below the 80% criteria range (s ee Table 3-15). Peer correct delivery of consequen ces averaged the highest in the free play (M =100%), attention (M =100%), escape (M =100%), a nd ignore (M=100%) conditions of the first functional analysis (see Table 316). Lower average percentage of peer correct delivery of consequences occurred in the tangible condition (M = 97%) of the first functional analysis. Peer correct delivery of consequences averaged the highest in the free play (M = 100%), tangible, and attention (M = 97%) c onditions of the second functional analysis. Lower average percentage of peer correct delivery of consequences occurred in the escape condition (M = 83%) of the second f unctional analysis (see Table 3-16). Peer consequences were not delivered contingent on Shannon’s peer-related withdrawn social behavior in the tangible condition of the s econd functional analysis, as Shannon, rather than peers, maintained access to the high pref erred item. All data for peer delivery of consequences were within the 80% established criteria. Peer positive initiations occurred at a mean rate of 1.85/minute in the free play condition of the first functional analysis, below the mean goal rate of 4.0/minute and be low the 80% criteria (see Table 3-17). In the second functional analysis, peer positive in itiations occurred at a mean rate of 2.47/minute in the free play condition, below the mean goal rate of 4.0/minute and below the 80% criteria. In general (with the excep tion of presenting social demands in the escape condition of the second functional analysis) these da ta indicate that Shannon’s

PAGE 109

97 peers’ were able to implement the procedures (i.e., correctly presen ting social demands in the escape conditions and correctly delivering consequences continge nt on peer-related withdrawn and peer-related nega tive social behavior) with acceptable integrity (i.e., met the 80% criteria range). An exception was peer presentation of positive social initiations in the free play conditions of the first a nd second functional anal yses, occurring below the established 80% criteria range. Table 3-15. Shannon’s Mean Rate and Ranges per minute of Peer Correct Presentation of Social Demands Condition Shannon’s peer correct presentation of social demands (FA #1) Shannon’s peer correct presentation of social demands (FA #2) Escape 3.53 (3.0 to 4.2) 2.73 (2.6 to 2.88) Shannon peer-related initiations Shannon’s mean rate per minute and ranges of peer-related positive initiations for the first and second functional an alysis are presented in Tabl e 3-18. In the first functional analysis, the highest mean rate per minute of Shannon’s positive initiations occurred in

PAGE 110

98 0 0.2 0.4 0.6 0.8 1 13579111315171921 Free Play Attention Tangible Escape IgnoreRate per min Withdrawn 0 0.2 0.4 0.6 0.8 1 13579111315171921Rate per min Negative 0 0.4 0.8 1.2 1.6 2 2.4 13579111315Rate per min Withdrawn Sessions Figure 3-2. Rate per minute peer-related withdrawn (top panel) and negative social behavior (middle panel) in first functional anal ysis, and peer-related withdrawn social behavior (bottom pane l) of second functi onal analysis for Shannon.

PAGE 111

99 Table 3-16. Shannon’s Average Percentage and Ranges of Peer Correct Delivery of Consequences Condition Shannon’s peer correct delivery of consequences (FA #1) Shannon’s peer correct delivery of consequences (FA #2) Free play 100 100 Tangible 97 (75 to 100) 100 Attention 100 97 (86 to 100) Escape 100 83 (50 to 100) Ignore 100 N/A Table 3-17. Shannon’s Mean Rate and Range s per minute Peer Positive Initiations Condition Shannon’s peer correct presentation of positive initiations (FA #1) Shannon’s peer correct presentation of positive initiations (FA #2) Free play 1.85 (1.4 to 3.0) 2.47 (2.0 to 3.4) the free play condition (M = 0.75/minute). Lower mean rates per minute of peer-related positive initiations occurred in the escape (M = 0.33/minute), tangible (M = 0.28/minute), attention (M = 0.07/minute), and ignore (M = 0/minute) conditions. Similar to the first functional analysis, Shannon’s highest mean rate per minute of peer-related positive initiations also occurred in the free play condition (M = 0.47/minute). Lower mean rates per minute of peer-related positive initiations occurred in the escape (M = 0.13/minute), tangible (M = 0.08/minute), and atten tion (M = 0/minute) conditions. Table 3-18. Shannon’s Mean Rate and Ranges per minute of Participant Peer-related Positive Initiations Condition Positive initiation (FA #1) Positive initiation (FA #2) Free play 0.75 (0.6 to 1.0) 0.47 (0 to 1.0) Tangible 0.28 (0 to 0.6) 0.08 (0 to 0.2) Attention 0.07 (0 to 0.2) 0

PAGE 112

100 Table 3-18. Continued Condition Positive initiation (FA #1) Positive initiation (FA #2) Escape 0.33 (0 to 0.8) 0.13 (0 to 0.2) Ignore 0 N/A Shannon peer-related responses Shannon’s highest average percentage of peer-related positive responses occurred in the escape (M = 57%) and free play (M = 56%) conditions of the first functional analysis (see Table 3-19). Lower average perc entage of peer-related positive responses occurred in the tangible condition (M = 26%). Similar to the first functional analysis, Shannon’s highest average percentage of pee r-related positive responses occurred in the free play (M = 73%) and escape (M = 70%) conditions. Lower to zero average percentages of peer-related positive respons es occurred in the tangible (M = 7%) and attention (M = 0%) conditions. Shannon’s highest average percentage of peer-related no response behavior occurred in the tangible condition (M = 73 %) of the first functional analysis. Lower average percentages of peer-related no res ponse behavior occurred in the free play (M = 44%) and escape (M = 27%) conditions (see Table 3-19). In the second functional analysis, Shannon’s highest average percenta ge of peer-related no response behavior occurred in the attention (M = 100%) and tangible (M = 87%) conditions. Lower average percentages of peer-related no response behavi or occurred in the escape (M = 30%) and free play (M = 27%) conditions.

PAGE 113

101 Table 3-19. Shannon’s Average Percentage and Ranges of Partic ipant Peer-related Positive and No Responses Condition Positive response (FA #1) No response (FA #1) Positive response (FA #2) No response (FA #2) Free play 56 (43 to 71) 44 (29 to 57) 73 (60 to 88) 27 (12 to 40) Tangible 26 (11 to 50) 73 (50 to 89) 7 (0 to 14) 87 (57 to 100) Attention N/A N/A 0 100 Escape 57 (50 to 61) 27 (27 to 28) 70 (43 to 87) 30 (13 to 57) Ignore N/A N/A N/A N/A Summary of Shannon’s functi onal analysis results In summary, Shannon did not engage in peer-related withdrawn social behavior during the initial functional analysis, ther efore the function coul d not be evaluated. Shannon engaged in variable and low to zero rates per minute of peer-related negative social behavior in the first functional analysis, resulting in inconclusive findings. As a result, a second functional analysis was c onducted, with minor procedural variations. Shannon’s peer-related withdraw n social behavior was elevat ed in the tangible condition in comparison to all other conditions in the second functional analys is. However, a three point upward/stable trend was not established in the tang ible condition, therefore the results are best described as inconclusive. Shannon did not engage in peer-related negative social behavior in the second functi onal analysis, therefore the function of peerrelated negative social behavior could not be evaluated. In addition, with th e exception of the escape condition in the second functional analysis condition, and free play conditions in the first and second functional analysis conditions, peers generally had acceptable integrity with implementing the functional anal ysis procedures (i.e ., correctly presenting social demands and delivering consequences). Finally, the additional data evaluated indicated Shannon initiated to her peers most of ten in the free play conditions and less

PAGE 114

102 often in the ignore and attention conditions. When peers initiated to Shannon, she was most likely to respond during free play and escape (social demand) conditions, and least likely to respond in the attention and tangible conditions. Colin Functional Analysis Results Colin peer-related withdrawn and peer-related negative behaviors The rate per minute of peer-related wit hdrawn social behavior (top panel), peerrelated negative social behavi or (center panel), and the co mbination of peer-related withdrawn and negative social behavior (bottom panel) from Colin's functional analysis are displayed in Figure 3-3. The mean rate per minute of peer-related withdrawn, peerrelated negative, and the combination of peer-related withdrawn and negative social behaviors are displayed in Table 3-20. Colin exhibited a three point upward trend of peerrelated withdrawn social beha vior in the escape condition. However, despite an initial three point upward trend, peer -related withdrawn social beha vior was variable throughout the escape condition and did not end on a thre e point upward trend. Colin’s peer-related withdrawn social behavior was elevated in the escape condition in comparison to all other functional analysis conditions. With the ex ception of one free play and one attention condition, peer-related withdrawn social behavior occurred at zero rates per minute in the free play, tangible, attention, and ignore c onditions. Visual analys is suggests Colin’s peer-related withdrawn social behavior (F igure 3-3, top panel) maybe maintained by negative reinforcement in the form of escap e from social demands. However, due to variable rates per minute of peer-related withdrawn social behavior in the escape condition, the results are best described as inconclusive. Colin exhibited a three point upward trend of peer-related negative social behavior in the escape condition. However, peer-related negative social behavior was variable

PAGE 115

103 throughout the escape condition, beginning with a three point downward trend of peerrelated negative social behavi or that was immediately foll owed by a three point upward trend. Peer-related negative social behavior occurred at zero rates per minute in the free play, attention, tangible, and ignore conditions. Peer-related negative social behavior occurred at zero rates per minute in 2/7 es cape conditions. With the exception of the two escape conditions that occurred at zero rate s per minute, Colin’s peer-related negative social behavior was elevated in the es cape condition in comp arison to all other conditions. Visual analysis s uggests that Colin’s peer-relat ed negative social behavior (Figure 3-3, middle panel) maybe maintained by negative reinforcement in the form of escape from social demands. However, due to variable rates per minute of peer-related negative social behavior in the escape condition and the esca pe condition not ending with a three point upward/stable trend of peer-relat ed negative social beha vior, the results are best described as inconclusive. Colin combined peer-related withdraw n and peer-related negative behaviors As noted previously, the opportunity to e ngage in both peer-re lated withdrawn and peer-related negative social behavior wa s available across all functional analysis conditions and during the escape, tangible, and attention conditions, reinforcement was delivered contingent on the occurrence of eith er peer-related wit hdrawn or peer-related negative social behavior. For example, in escape condition number 3, Colin engaged in two instances of peer-related withdrawn soci al behavior and three instances of peer related negative social behavior. Each occu rrence of peer-related withdrawn social behavior was reinforced with 15 seconds escape from social demands and each occurrence of peer-related negative social behavior was reinforced with 15 seconds

PAGE 116

104 escape from social demands, for a total of fi ve deliveries of negative reinforcement in that escape condition. Visual analysis of Colin’s peer-related w ithdrawn (Figure 3-3, top panel) and peerrelated negative social behavi or (Figure 3-3, middle panel) indicated both peer-related withdrawn and peer-related ne gative social behaviors occu rred in the same functional analysis (i.e., escape) conditions. Peer-relat ed withdrawn and/or peer-related negative social behavior occurred in 100% (7/7) of the escape conditions. Both peer-related withdrawn and peer-related nega tive social behavior occurre d in 43% (3/7) of the escape conditions. Peer-related negative social behaviors did not occur in the free play, attention, tangible, and ignore conditions, therefore p eer-related withdraw n and peer-related negative social behaviors did not occur in those same conditions. Colin exhibited a three point stable trend of combined peer-rel ated (withdrawn and negative) social behavior in the escape c ondition. Combined peer-r elated (withdrawn and negative) social behavior in the escape condi tion was elevated in comparison to all other conditions. In addition, there was little ove rlap between escape conditions and other conditions, as combined peer-related (wit hdrawn and negative) so cial behavior only occurred in one free play and one attenti on condition. With the exception of combined peer-related (withdrawn and nega tive) social behavior occurri ng at a rate of 0.2/minute in 1/4 free play conditions and a rate of 0.2/mi nute in 1/4 attention conditions, combined peer-related (withdrawn and ne gative) social behavior occu rred at zero rates per minute across free play, attention, and ignore conditi on. In the escape condition, combined peerrelated (withdrawn and negative) social beha vior occurred at a rate that ranged from 0.2/minute to 1.0/minute. In the last three escape conditions, combined peer-related

PAGE 117

105 (withdrawn and negative) social behavior occurred at a rate of 0.4/minute. Visual analysis indicated that Colin’s combined peer-relate d (withdrawn and negativ e) social behavior (Figure 3-3, bottom panel) was maintained by negative reinforcement in the form of escape from social demands. Table 3-20. Colin’s Mean Rate and Range s per minute of Part icipant Peer-related Withdrawn, Negative, and Combined (Withdrawn and Negative) Behavior Condition Withdrawn Negative Combined Free play 0.5 (0 to 0.2) 0 0.05 (0 to 0.2) Tangible 0 0 0 Attention 0.04 (0 to 0.2) 0 0.04 (0 to 0.2) Escape 0.26 (0 to 0.8) 0.26 (0 to 0.6) 0.51 (0.2 to 1.0) Ignore 0 0 0 Colin peer integrity Colin’s peer correct presentation of social demands occurred at a mean rate of 3.54/minute in the escape condition, very close to the goal mean of 4.0/minute and within the 80% criteria range (see Table 3-21). Peer correct delivery of c onsequences averaged the highest in the free play, ignore, atten tion, and tangible conditions (M = 100% for all conditions). A slightly lower av erage percentage of peer corr ect delivery of consequences occurred in the escape condition (M = 94%) (s ee Table 3-22). Peer positive initiations occurred at a mean rate of 3.5/minute in th e free play condition, very close to the goal mean rate of 4.0/minute and within the 80% criteria range (see Table 3-23). All peer integrity data were within the 80% established criteria. In general th ese data indicate that Colin’s peers were able to implement the procedures (i.e., correctly present social demands in the escape condition, correctly de liver consequences contingent on peerrelated withdrawn and peer-related negativ e social behavior, and present positive

PAGE 118

106 0 0.2 0.4 0.6 0.8 1 13579111315171921 Free Play Attention Tangible Escape IgnoreRate per min Withdrawn 0 0.2 0.4 0.6 0.8 1 13579111315171921Rate per min Negative 0 0.2 0.4 0.6 0.8 1 13579111315171921Rate per min Withdrawn and NegativeSessions Figure 3-3. Rate per minute peer-related w ithdrawn (top panel), negative (middle panel), and combined social behavior (bottom panel) for Colin.

PAGE 119

107 initiations in the free play condition) with acceptable integrity (i.e., met the 80% criteria range). Table 3-21. Colin’s Mean Rate and Ranges pe r minute of Peer Correct Presentation of Social Demands Condition Colin’s peer correct pr esentation of social demands Escape 3.54 (2.0 to 5.6) Table 3-22. Colin’s Average Percentage a nd Ranges of Peer Correct Delivery of Consequences Condition Colin’s peer correct delivery of consequences Free play 100 Tangible 100 Attention 100 Escape 94 (67 to 100) Ignore 100 Table 3-23. Colin’s Mean Rate and Range s per minute Peer Positive Initiations Condition Colin’s peer correct presentation of positive initiations Free play 3.5 (3.2 to 3.8) Colin peer-related initiations Colin displayed very low mean rates per minute of peer-related positive initiations in the tangible (M = 0.07/minute), escap e (M = 0.06/minute), free play (M = 0.05/minute), and attention (M = 0.04/minut e) conditions (see Table 3-24). Colin engaged in peer-related positive initiations in only 5 of the 21 functional analysis conditions (i.e., 2 escape, 1 free play, 1 tangi ble, and 1 attention condition). He engaged in zero rates per minute of peer-related posit ive initiations in the ignore condition.

PAGE 120

108 Table 3-24. Colin’s Mean Rate and Range s per minute of Part icipant Peer-related Positive Initiations Condition Positive initiation Free play 0.05 (0 to 0.2) Tangible 0.07 (0 to 0.2) Attention 0.04 (0 to 0.2) Escape 0.06 (0 to 0.2) Ignore 0 Colin peer-related responses Colin’s highest average percentage of peer-related positive responses occurred in the escape condition (M = 32%) (see Table 325). Lower average pe rcentages of peerrelated positive response s occurred in the free play (M = 7%), tangible (M = 4%), and attention (M = 0%) conditions Colin’s highest average pe rcentages of peer-related no responses occurred in the attention (M = 100 %), tangible (M = 96%), and free play (M = 92%) conditions (see Table 3-25). Lower average percentage of peer-related no responses occurred in the escape condition (M = 60%). In comparison to peer-related positive responses, Colin engaged in peer-related no response behavior most often across functional analysis conditions. Specifically, with the exceptio n of the escape condition, Colin engaged in peer-related no response behavi or at least 80% or more of all functional analysis conditions. Colin engaged in peer-rel ated no response behavi or in less than 80% of all escape conditions. In addi tion, his peer-related responses were more variable in the escape condition (percenta ge of peer-related no response behavior ranged from 30% to 79%) in comparison to all other functional anal ysis conditions (percentage of peer-related no response behavior ranged from 82% to 100%). It is important to note that peer-related positive and no response behavior were base d on Colin’s opportunities to respond to

PAGE 121

109 peers initiations. Peers did not initiate to Colin during the ignore condition, as they were instructed to only talk and play with each other. Therefore, Colin did not have an opportunity to engage in peer-related respons e behavior in the ignore conditions of the functional analysis. Table 3-25. Colin’s Average Percentage and Ranges of Participant Peer-related Positive and No Responses Condition Positive response No response Free play 7 (0 to 18) 92 (82 to 100) Tangible 4 (0 to 12) 96 (88 to 100) Attention 0 100 Escape 32 (12 to 60) 60 (30 to 79) Ignore N/A N/A Summary of Colin’s functional analysis results In summary, Colin’s peer-related withdr awn and peer-related negative social behaviors appeared to be maintained by ne gative reinforcement in the form of escape from social demands. Colin’s peer-related withdrawn and peer-rel ated negative social behaviors were elevated in the escape condi tion in comparison to all other functional analysis conditions. However, due to variab ility and not ending on a three point stable trend, the findings of peer-related withdrawn and peer-related nega tive social behaviors are not definitive. Colin’s combined peer -related (withdrawn and negative) social behavior was maintained by negative reinfor cement in the form of escape from social demands. In addition, peers had acceptable integrity with implem enting the functional analysis procedures (i.e., correctly presenting social demands, delivering contingent consequences, and presenting positive initiati ons). Finally, the addi tional data evaluated indicated Colin rarely initiated to his peers, and when Colin’s peers initiated to him,

PAGE 122

110 Colin most often did not respond to their init iations. However, in comparison to Colin’s responses to his peers’ initiations, he re sponded positively more often to their social demands. Summary of Functional Analysis Results In summary, the results indicated that the functional analysis identified a function (i.e., negative reinforcement) for the combin ed peer-related (wit hdrawn and negative) social behavior for 1 of the 3 participants (i.e., Colin) when using visual analysis and a three point trend, stab ility, overlap, and magnitude of differences to guide decision making. The results for Colin’s peer-related wi thdrawn and peer-related negative social behaviors were not clear. Colin’s peer-relate d withdrawn and peer-related negative social behaviors were elevated in the escape condi tion in comparison to all other functional analysis conditions, however rates per minute we re variable and the analysis did not end on a three point stable or upward trend. Th e results of Shane and Shannon’s functional analyses of peer-related w ithdrawn and peer-related nega tive social behavior were inconclusive. Shane’s functional analysis re sults were inconclusive due in part to variability and overlap between conditions. His peer-related withdrawn social beha vior was variable across attention, escape, tangible, and ignore conditions. A three point upward trend was established in the ignore c ondition, however peer-related withdrawn social behavior was variable increasing from ra tes of 0.2/minute, 0.2/minute, 0.4/minute to 2.2/minute in the last ignore condition. Shane’s pe er-related negative social be havior was also variable, though zero rates per minute of peer-related ne gative social behavior occurred in the free play condition and the ignore condition ended on a downward (i.e., 0/ minute) trend. Both peer-related withdrawn and peer -related negative social beha viors ended on a three point

PAGE 123

111 upward trend in the tangible condition. Shane’s combined peer-related (withdrawn and negative) social behavior remained variab le in the tangible co ndition, though appeared somewhat less variable in the ignore c ondition. Despite that, Shane’s combine peerrelated (withdrawn and negative) social behaviors were variable across conditions, resulting in inconclusive findings. Shannon did not engage in peer-related w ithdrawn social behavior in the first functional analysis, therefore it could not be evaluated. She engaged in low to zero rates per minute of peer-related negative social behavior in the first functional analysis, resulting in inconclusive fi ndings. A second functional analysis was then conducted, with minor procedural variations. Shannon did not engage in peer-related negative social behavior in the second functional analysis, therefore it could not be evaluated. She engaged in elevated rates per minute of peer-related withdrawn social behavior in the tangible condition in comparison to all othe r functional analysis conditions, though the tangible condition ended on a slightly downwar d trend, resulting in inconclusive findings. Shannon did not engage in peer-related wit hdrawn and peer-related negative social behavior in the same functional analysis condition, therefore her combination of peerrelated (withdrawn and negative) soci al behavior could not be evaluated. The functions of peer-related withdrawn a nd peer-related negative social behaviors could not be compared across participants, as a function of combined peer-related (withdrawn and negative) social behaviors was only identif ied for one participant. However, some similarities across the particip ant’s functional analyses did occur. All 3 participants did not engage in peer-related negative social behavior in the free play condition and 2 of the 3 (i.e., Shannon and Co lin) did not engage in peer-related negative

PAGE 124

112 social behavior in the free play or ignore conditions of the functional analysis. In addition, Shannon and Colin only engaged in pee r-related negative social behavior in one test condition (i.e., tangible a nd escape condition, respectively) of the functional analysis. Finally, all 3 participants enga ged in peer-related withdrawn social behavior in only one free play condition of the func tional analysis and Shannon a nd Colin did not engage in any peer-related withdrawn social behavior in the ignore conditions of the functional analysis. However, it is important to note th at an ignore conditi on was not conducted in Shannon’s second functional analysis since sh e had engaged in zero rates per minute of peer-related withdrawn social behavior in the ignore c ondition of the first functional analysis. Peer integrity data indicated that in genera l, all 3 participants’ peers were able to implement the functional analysis procedures with acceptable integrity (i.e., 80% criteria range). Colin’s peers implemented all functio nal analysis procedur es (i.e., correctly presenting social demands in the escape condition, correctly deliv ering consequences contingent on peer-related w ithdrawn and/or peer-related negative social behavior, and presenting positive initiations in the free pl ay condition) within the 80% criteria range. Shane’s peers also implemented procedures (with exception of correct delivery of consequences contingent on peer-related w ithdrawn and/or peer-related negative social behavior in the escape and tangible conditions ) within the 80% criteria range. Shannon’s peers implemented the procedures of correctly presenting social demands in the escape condition (with the exception of the escape c ondition in the second functional analysis) and correct delivery of consequences continge nt on peer-related w ithdrawn and/or peerrelated negative social behavi or within the 80% criteria range. However, Shannon’s peers

PAGE 125

113 did not present positiv e initiations in the free play c ondition within the 80% criteria range. Participant peer-related positive initiation data indicated that Shane and Colin rarely positively initiated to their peers throughout all f unctional analysis conditions. When Shannon initiated to her peers, it was most often in the free play condition. Shane and Colin rarely responded to their peers’ initiations, though when they did respond, it was most often to social demands in th e escape condition. Simila rly, Shannon rarely responded to peer initiations in the tangi ble condition and never responded to peer initiations in the attention condition. However, when she di d respond, it was most often in the free play and escape (social demand) c onditions. All 3 participants engaged in no response behavior an average of 80% or more in the atte ntion and tangible conditions. Shane and Colin also engaged in no response be havior an average of 92% or higher in the free play condition. Table 3-26. Reinforcers Identi fied in Functional Analysis Peer-related Behavior Shane Shannon (FA #1) Shannon (FA #2) Colin Withdrawn Inconclusive (possible automatic and positive reinforcement in the form of access to preferred tangibles) Not evaluated Inconclusive (possible positive reinforcement in the form of access to preferred tangibles) Inconclusive (possible negative reinforcement in the form of escape from social demands) Negative Inconclusive (possible positive reinforcement in the form of access to preferred tangibles) Inconclusive Not evaluated Inconclusive (possible negative reinforcement in the form of escape from social demands)

PAGE 126

114 Table 3-26. Continued Peer-related Behavior Shane Shannon (FA #1) Shannon (FA #2) Colin Combined (Withdrawn and Negative) Inconclusive (possible positive reinforcement in the form of access to preferred tangibles) N/A N/A Negative reinforcement in the form of escape from social demands

PAGE 127

115 CHAPTER 4 DISCUSSION Summary of Questions Children with ASD demonstrate problems w ith peer-related social behaviors (e.g., withdrawn and/or negative) (APA, 2000). Seve ral treatments have been implemented and successfully increased the appropr iate peer-related social beha vior of participants with ASD (Garrison-Harrell et al., 1997; Pierce & Schreibman, 1997b); however, these treatments were not always successful across participants, behaviors, or settings (Apple Billingsley, & Schwartz, 2005; Kamps, Potucek, Lopez, Kravits, & Kemmerer, 1997; Thiemann & Goldstein, 2004). Functional an alysis methodology has demonstrated success in treating a variety of problem behaviors (e.g., a ggression, SIB, destruction) with young children with ASD (Piazza et al., 199 8; Richman, et al., 2001; Wacker et al., 1990); however, this methodology has not been utilized to treat peer-related social behaviors. Given the success of the functiona l analysis to identify variables maintaining problem behavior and develop robust treatments the extension of the functional analysis methodology to assess peer-related social be havior appears to be the next step. The purpose of this study was fourfold: (a) to evaluate the effectiveness of the use of the functional analysis methodology to determine the function of peer-related withdrawn and peer-related ne gative social behavior for three young participants with ASD, (b) to compare and determine differe nces of the functions of peer-related withdrawn and peer-related nega tive social behavior across pa rticipants with ASD, (c) to determine if peers were able to implemen t the functional analysis procedures with

PAGE 128

116 acceptable integrity, and (d) to examine rate per minute of participant peer-related positive initiations and percentage of partic ipant peer-related positive and participant peer-related no response social behavior across conditions. Summary of Findings The primary purpose of this study was to evaluate the effectiveness of the functional analysis methodology to identify th e function of peer-rel ated withdrawn and peer-related negative social behaviors. Unfortunately, the functional analysis methodology was not successful in identifyi ng a definite function for peer-related withdrawn or peer-related negative social beha vior, singly, for any of the 3 participants. However, when rate per minute of combin ed peer-related (wit hdrawn and negative) social behaviors was evaluated post hoc, a ne gative reinforcement function for 1 of the 3 participants (i.e., Co lin) was identified. Since a clear function of combined peer-r elated (withdrawn a nd negative) social behavior was only identified for one partic ipant, the second purpose, to compare the functions of peer-related w ithdrawn and peer-related nega tive social behavior across participants with ASD to determine if the f unctions differed across participants, could not be evaluated. The third purpose was to determine if peers were able to implement the functional analysis procedures with acceptabl e integrity. With the exception of Shane’s peer delivery of consequences in the escap e and tangible conditions and Shannon’s peer presentation of positive initiations in th e free play condition, peers were able to implement the functional analysis procedures (i.e., provide social demands in the escape condition, provide relevant consequences in the escape, tangible, and attention conditions, and provide positive initiations in the free play condition) within the 80% criteria. Finally, the fourth purpose was to ex amine rate per minute of participant peer-

PAGE 129

117 related positive initiations and percentage of participant peer-related positive and peerrelated no response social behavior across condi tions. All 3 participants rarely engaged in peer-related positive initiations. When they di d respond to peer-posi tive initiations, they did so most often in response to social demands (with the exception of Shannon, who responded most often to peer-positive initia tions and social demands). However, all 3 participants most often engaged in pee r-related no response behavior (e.g., not responding to or acknowledging a peer’s initia tion within 3 seconds of the initiation). Colin and Shannon engaged in peer-related no response behavior in 100% of attention conditions, whereas Shane engaged in peer-rel ated no response behavior in 95% of free play conditions. Each of these findings w ill be examined further in the following sections. Identified Functions Findings Shane’s functional analysis findings were th e most variable of the 3 participants. Shane’s peer-related withdrawn social behavi or ended on a three point upward trend in the tangible and ignore conditions. However, the rates per minute of peer-related withdrawn social behavior ranged between 0/minute and 0.4/minute in 9/10 tangible conditions and ranged from 0.2/minute to 0.4/m inute in 3/4 ignore conditions. The final ignore and the final tangible c onditions are th e only conditions when the rate per minute of peer-related withdrawn social behavior exceeded 0.4/minute. Similarly, rate per minute peer-related withdrawn social beha vior ranged between 0/minute and 0.4/minute in the escape condition. In the free play c ondition, rates per minute of peer-related withdrawn social behavior were 0/minute with the exception of one condition with a rate of 2.2/minute. Initially rate per minute of p eer-related withdrawn so cial behavior in the attention condition was stable elevated, and did not overl ap with other conditions.

PAGE 130

118 However, rates per minute of peer-related wit hdrawn social behavior decreased with 3 of the 5 attention conditions ranging between 0/ minute and 0.4/minute. Of the 27 functional analysis conditions, the rate of peer-related withdrawn so cial behavior only exceeded 0.4/minute in five conditions (twice in the at tention condition, and once in the free play, ignore, and tangible conditions). The tangibl e and ignore conditions ended on variable upward trend (i.e., the only condition above 0.4/minute with an increased rate of 2.2/minute in the ignore condition and incr eased rate of 1.2/minute in the tangible condition). Further analysis until Shane’s rate per minute of peer-related withdrawn social behavior was stable may have resulted in more conclusive findings. When Shane’s peer-related withdrawn and p eer-related negative social behaviors were combined, the functional analysis findings remained variable Similar to his peer-r elated withdrawn and peer-related negative social be haviors, visual analysis of Shane’s peer-related combined (withdrawn and negative) social behaviors in dicated overlap between functional analysis conditions. Shane’s functional analysis fi ndings were inconclusive, though further analysis of peer-related withdrawn, peer-rel ated negative, and combined peer-related (withdrawn and negative) social behavior s may have resulted in clearer findings. Shannon’s functional analysis resu lts were less variable than Shane’s. In fact, in the initial functional analysis, Shannon engaged in zero rates per minute of peer-related withdrawn social behavior acr oss all functional analysis c onditions and only engaged in peer-related negative social be havior in one (i.e ., tangible) functional analysis condition. In the tangible condition of the initial functional analysis, Shannon’s variable peer-related negative social behavior was hypot hesized to be related to sa tiation of the high preferred tangible (dinosaur) (i.e., the ta ngible was no longer high preferre d). Change in preference

PAGE 131

119 in the course of an assessment has been noted elsewhere in the literature (Roane, Vollmer, Ringdahl, & Marcus, 1998). In additi on, Mueller, Wilczynski, Moore, Fusilier, and Trahant (2001) found that a lthough a participant would e ngage in problem behavior to gain access to low and neut ral preferred tangibles, the participant would engage in higher levels of problem behavi or to gain access to high pref erred tangibles than to gain access to neutral or low prefe rred tangibles. As a result of the hypothesis that Shannon’s preference of tangibles had changed, a sec ond preference assessment was conducted prior to tangible condition number 17 (DeLeon & Iwata, 1996). The results of the second preference assessment identified two different tangibles as hi gh preferred, verifying the hypothesis that the tangible (i.e., dinosaur) was no longer high preferred. The newly identified high preferred tangibles were ava ilable (rotating the tangi bles so only one was available each tangible condition) starting in tangible condition number 17. An initial relative increase (i.e., increas e from 0/minute to 0.6/minute) was observed in tangible condition number 17, though it was followed by th ree tangible conditions with zero rates per minute of peer-related negative social behavior. Therefore, access to the high preferred tangible did not appear to main tain Shannon’s peer-related negative social behavior. This was consistent with teacher report and direct ob servation, indicating Shannon engaged in zero to low rates per minut e of peer-related nega tive social behavior at school. In contrast, direct observations indicated Shannon did enga ge in peer-related withdrawn social behavior at school. Despite this, and verification that a high preferred tangible was present in the tangible conditi on, Shannon’s peer-related withdrawn social behavior remained at zero rates per minute in all conditions of the initial functional

PAGE 132

120 analysis. As a result, a second functional anal ysis was conducted in attempts to capture the variables maintaining Shannon’s peer-r elated withdrawn social behavior. It was hypothesized that response effort may have decreased the likelihood Shannon would engage in peer-r elated withdrawn social beha vior, as the response effort to engage in peer-related w ithdrawn social behavior was hi gher than the response effort to remain in the in area. Richman et al. (2001) found differentiated occurrence of topographies of behavior, despite reinforcem ent for all topographies of behavior, based on the efficiency of responding. Based on the hypothesis that response effort decreased Shannon’s likelihood to engage in peer-related withdrawn social be havior, the response effort required for Shannon to engage in p eer-related withdrawn social behavior was equalized with the response effort required to enter the in area (i.e., Shannon had to make one step to enter the in or withdrawn area) across all condit ions of the second functional analysis. Shannon engaged in elevated rates pe r minute of peer-related withdrawn social behavior in the tangible c ondition in comparison to all other functional analysis conditions. In addition, she engaged in peer-rel ated withdrawn social behavior at least one time in each of the functional analysis conditions, therefore she experienced the contingencies (i.e., access to high preferred tangible, access to attention, escape from social demands, and no response) of peer-rel ated withdrawn social behavior in each functional analysis condition. Since Shannon di d not engage in peer-related withdrawn social behavior in any condition of the ini tial functional analysis and because it was not experimentally evaluated, it remains unclear whether reduced response effort or simply experiencing the contingencies of peer-related withdrawn social beha vior resulted in an

PAGE 133

121 increase in the behavior in the second func tional analysis, particularly in the tangible condition of the second functional analysis. In the second functional analysis, Shannon engaged in zero to 0.2 rates per minute of peer-related withdrawn social behavior across free play, escape, and attention conditions, with an exception in one atten tion condition when peer-related withdrawn social behavior occurred at a rate of 0.8/minute. Shannon’s pe er-related withdrawn social behavior was elevated in the tangible c ondition in comparison to all other functional analysis conditions. In additi on, with the exception of one attention condition, there was no overlap in Shannon’s peer-relat ed withdrawn social behavior in the tangible condition and all other functional analysis conditions. However, Shannon’s p eer-related withdrawn social behavior was on a general downward trend (i.e., rates started at 1.8/minute, 2.2/minute, 0.8/minute, 0.8/minute, and ended at 0.4/minute) in the tangible condition, despite the lowest rate of 0.4/minute in th e tangible condition rema ining higher than all but one (i.e., attention) of all other f unctional analysis conditions. One possible explanation for the decline in peer-related withdrawn social beha vior in the tangible condition is satiation. As noted previous ly, preference for tangibles has been demonstrated to change throughout assessments (Roane et al., 1998). If preference for a tangible changes and is no longer high preferre d, the occurrence of the behavior that results in access to that tangible, has been demonstrated to reduce as the preference decreases (Mueller et al., 2001). The same high preferred tangibles from the second preference assessment were utili zed in the last four tangible conditions of the initial functional analysis and the five tangible condi tions of the second functional analysis, for a total of nine tangible conditions with th e same two high preferred tangibles being

PAGE 134

122 rotated across tangible conditions. Shannon’s preference was demonstrated to change after only four tangible conditi ons had been conducted in th e initial functional analysis. Therefore, it is possible Sha nnon’s peer-related withdrawn so cial behavior decreased in the tangible condition of the second functiona l analysis due to th e tangible no longer being high preferred. However, a third pref erence assessment was not conducted to evaluate this possibili ty. Had this been evaluated and preference had changed, the results of Shannon’s second functional analysis may ha ve been more stable and thus provided clearer results (Mueller et al.). Similar to Shannon, Colin’s functional anal ysis was less variable than Shane’s. However, Colin had slightly more overlap of peer-related withdrawn social behavior across functional analysis conditions than Shannon. In addition, Co lin’s peer-related withdrawn social behavior was elevated in the escape condition in comparison to all other functional analysis conditions, whereas Sh annon’s peer-related withdrawn social behavior was elevated in the tangible condi tion in comparison to all other conditions in the second functional analysis. The magnitude of difference between Colin’s rate per minute of peer-related withdrawn social be havior in the escape condition and all other conditions was low. Only 2 of the 7 escap e conditions had a rate of peer-related withdrawn social behavior above 0.2/mi nute and Colin engaged in peer-related withdrawn social behavior at a rate of 0.2/minute in one attention and one free play condition. Therefore, only 2 of the 7 escape conditions did not overlap with another functional analysis condition. Co lin only engaged in peer-relate d negative social behavior in one functional analysis condition, another similarity with Shannon. However, Colin’s peer-related negative social be havior was more stable, with 4 of the 7 escape conditions

PAGE 135

123 with a rate between 0.2/minute and 0.4/minute and 5 of the 7 conditions with a rate between 0.2/minute and 0.6/minute of peer-re lated negative social behavior. An additional escape condition with a rate of 0.2/ minute or higher of peer-related negative social behavior may have made the results more definitive. When all peer-related (withdrawn and negative) social behaviors were combined for each functional analysis condition, the escape condition was elevated above all other functional analysis conditions, with the exception of one free play and attenti on condition, and a three point stable trend was established to verify a function of negative reinforcement. Comparison of Functions Findings An exact comparison of functions of p eer-related withdraw n and peer-related negative social behavior cannot be examined as a definitive function was only identified for Colin, and that function (i .e., escape from social demands) was only definitive when peer-related withdrawn and peer -related negative social be haviors were combined across conditions. However, similarities and differences of functional analysis results were still observed across participants. One similarity was the social behavior displayed in the free play condition. Peer-related nega tive social behavior did not occur for any participant in the free play condition and peer-r elated withdrawn social beha vior only occurred in one free play condition for each pa rticipant. Similarly, Shannon and Colin never engaged in peer-related withdrawn or peer -related negative social behavi or in the ignore conditions. The free play and ignore conditions are consider ed control conditions, that is, conditions in which to compare the test conditions (e.g., escape, tangible, attention conditions) (Iwata et al., 1982/1994). When the target beha vior occurs at lower rates per minute in the free play and ignore conditi on in comparison to one or more test conditions in the functional analysis (i.e., Shannon and Colin), the target behavior is typically concluded to

PAGE 136

124 be maintained or manipulated by environmenta l variables, rather than automatic (i.e., non-environmental) reinforcement. However, wh en peer-related social behavior occurs at lower rates per minute in the free play condi tion in comparison to all other conditions, including the ignore condition (i.e., Shane), the peer-related social behavior is not as clearly identified as maintained by environm ental variables, though is identified as being controlled by environmental variables, as the peer-related social behavior can be decreased when access to tangibles and atte ntion are available and demands are not presented. For example, Sidener et al. (2005) conducted functional analysis of problem behavior with two participants. The findings of the functional anal ysis indicated both participants engaged in problem behavior during the most intervals in the ignore condition and least in the free play condition. Based on these findings, environmental enrichment was implemented, where attenti on and access to materials were available. This treatment was successful in reducing th e problem behavior of both participants. Differences between the participants incl uded the functional analysis conditions in which peer-related social behaviors were c onsistently elevated. P eer-related withdrawn social behavior was consistently elevated in the ignore condition in Shane’s functional analysis, the tangible condition in Shannon’s second functional analysis, and the escape condition in Colin’s functional analysis. Peer-related negative social behavior was not displayed consistently in any functional anal ysis condition for all 3 participants, though was elevated in the ignore and escape condi tions in Shane’s functional analysis, the tangible condition in Shannon’s functional an alysis, and escape condition in Colin’s functional analysis.

PAGE 137

125 Peer Integrity Findings Research has shown that training can increase the frequency in which young children provide initiations, reinforcement, a nd instructions to peers (McGee, Almeida, Sulzer-Azaroff, & Feldman, 1992; Pierce & Schreibman, 1997a; Zanolli, et al., 1996). Consistent with this previous work, in this study, participants’ peer s were generally able to implement functional anal ysis procedures of deliver ing relevant consequences contingent on peer-related withdrawn and p eer-related negative social behaviors within the mean 80% criteria range. Exceptions incl uded Shannon’s peers’ correct presentation of social demands and positive initiations in the free play condition and Shane’s peers’ correct delivery of consequen ces in the tangible and escape condition, all of which were below the mean 80% criteria range. Corr ect delivery of consequences was not systematically evaluated in this investigati on, therefore it is unclea r what, if any, role it had on participant peer-related social behavior. One possibility is that inadequate (i.e., below the 80% criteria) delivery of consequences could result in varied and thus unstable patterns of social behavior. In terestingly, Colin’s functional an alysis appeared to have the least amount of variability across conditions and his peers’ correct delivery of consequences exceeded an average of 87% acr oss all functional analysis conditions. In contrast, Shane’s peers’ averaged correct delivery of consequences within the 80% criteria for only 3 of the 5 functional analys is conditions, and Shane’s functional analysis appeared to have the most variability acr oss conditions. However, without systematic manipulation of peer correct delivery of reinfo rcement, the effect of peer implementation of procedures on participant social beha vior is unknown. Additional peer integrity behaviors measured included correct presenta tion of social demands in the escape condition and positive initiations in the free pl ay condition. In contrast to peer correct

PAGE 138

126 delivery of consequences, both Shane and Colin’s peers’ averaged correct presentation of social demands and positive in itiations within the 80% criteria. Despite this, Shane’s peer-related (withdrawn and ne gative) social behaviors we re variable in the escape condition of the functional analysis. In contra st, Shannon’s peer’s did not average correct presentation of social demands or positive initiations in the second functional analysis within the 80% criteria. However, despite that, her peer-relate d withdrawn social behavior was stable across both the escape and free play conditions. Again, these peer integrity measures were not systematically evaluated, therefore the effect, if any, on participant social behavior is unknown. Participant Appropriate So cial Behavior Findings McGee et al. (1997) compared the peer-r elated social behavior of young children with ASD to the peer-related social beha vior of young children who were typically developing. McGee et al. found that the ch ildren with ASD initi ated, responded, and stayed within proximity (i.e., 3 ft) of peers less often than typically developing children. McGee et al. also found that time spent in an integrated preschool (i.e., 1 year) did not increase the appropriate peer-r elated social behavior of young children with ASD. In this investigation, all 3 participants (Shane, Sha nnon, Colin) were in preschool settings for at least one year, though none had a specific trea tment plan that focused on increasing peerrelated social behaviors (i.e., initiations a nd/or responses). Similar to McGee et al. findings, all 3 participants e ngaged in zero to low rates pe r minute of initiations across functional analysis conditions and zero to low (with the exception of Shannon’s free play and escape conditions) percentage of peer-r elated positive responses. When peers were prompted to frequently (i.e., every 15 seconds ) initiate to participants, Shane and Colin averaged positive responses to these initiations less than 35% of the time. Interestingly,

PAGE 139

127 Shane and Colin’s average percentage of positive responses doubled in the escape condition in comparison to all other func tional analysis conditions. This was not systematically evaluated, though one possible explanation is that when given specific instructions (by peers) on how to positively respond, Shane and Colin were more successful (e.g., “Press the button”) than res ponding to the vague statements (e.g., “You have the book”) that were provided frequent ly (i.e., every 15 seconds) in the free play condition. Therefore, being provided examples of positive responses may be the first step in the chain of behaviors l eading to spontaneous social responses. Similarly Shannon averaged a higher percentage of positive responses in the escape condition, though also averaged a higher percentage of positiv e responses in the free play condition. Anecdotally, Shannon displayed more sophistic ated language and play behavior than either Shane or Colin. Therefore, alt hough Shannon did respond to specific social demands, she may have acquired sufficient langu age and play behavior that she was also able to positively respond to vague or non-di rective statements provided by peers in the free play condition. This may suggest that young children with ASD who have more developed language and play skills may requi re less instruction to positively respond to peers initiations. Another po ssibility is that Shannon did not require additional reinforcement to engage in positive respons es in the free play and escape conditions, whereas Shane and Colin may have engaged in higher percentages of positive responses had an additional reinforcer been provid ed. For example, Gonzalez-Lopez and Kamps (1997) increased the appropriate peer-related social behavior of young participants with ASD with the combination of social skills tr aining and positive reinforcement contingent on appropriate peer-related social behavior. Since the focus of the present investigation

PAGE 140

128 was to determine the functions of problem peer-related social behaviors, no reinforcement for positive initiations or responses was provided in the functional analysis; therefore, it is unclear what eff ect that might have had on participant social behavior. Interesting, Colin engaged in the highest average peer-related positive responses in the escape condition when he also engaged in the highest rate per minute of peer-related (withdrawn and negative) soci al behavior in the escape condition. Limitations of the Investigation Although this investigation presented many interesting findings, there are several limitations that need to be considere d. The main limitations of the study were methodological. First, decisions used to termin ate data collection were not always based on a three point trend. The results of the func tional analyses were not conclusive enough to adequately answer the question of th e usefulness of the functional analysis methodology in identifying functions for peer-rel ated social behavior This was at least partially due to not always using a three point trend to determine when to end the analysis. Using a three point trend to end analyses may have led to more definitive findings across participants. Second, it is unclear if the measurement of peer-related withdrawn social behavior was representative of the withdrawn behavior participants’ engaged in with peers in their natural environment. According to the de finition of peer-related withdrawn social behavior over half of the pa rticipant’s body had to cross th e plane of the line (indicated by masking tape on the floor) in to the withdrawn area to be measured as the participant engaging in peer-related withdrawn social behavior. The definition of peer-related withdrawn social behavior was the act of ente ring an area where peers were not present. However, withdrawn social behavior could also have been measured as proximity (e.g., 3

PAGE 141

129 ft or more between the participant and peer), rate per minute of positive initiations (e.g., 0/minute to 0.2/minute), percentage of positive responses (e.g., 0% to 20%), percentage of no responses (e.g., 80% to 100% ), facing away from peers, or not playing with the same materials as the peers, which is mo re consistent with how withdrawn social behavior is defined within th e diagnosis of ASD (APA, 2000) and how peer-related social behavior deficits have been described in pr evious literature (e.g., see McGee et al., 1997; Strain & Danko, 1995). For example, Strain & Danko (1995) individu ally described the social behavior deficits di splayed by five young children with ASD (e.g., infrequent peerrelated interaction, ignoring p eer initiations). McGee et al. (1997) found children with ASD displayed peer-related social behavior deficits, including d ecreased proximity and decreased facial orientation to peers, when compared to same-age typically developing children. In this study, the assessment areas (i.e., withdrawn and in ) were at least 3 ft in length for each participant, which is the dist ance of proximity evaluated by McGee et al. However, this assessment area was only a sm all portion of the natural environment (i.e., classroom or atrium). It is unc lear what affect the size of the assessment area and/or the unnatural set-up (i.e., tape on floor and not having access to the entire classroom or atrium) had on peer-related social behaviors. Further analysis is needed to determine if the definition of peer-related withdrawn social behavior utilized in this investigation captured the peer-related withdrawn social beha vior displayed in the natural environment. Third, modification of the methodology to adju st for response effort to engage in peer-related withdrawn social behavior was only conducted with one participant (i.e., Shannon). It is unknown what, if any, effect response effort had on Shane and Colin’s peer-related withdrawn social behavior. Further evaluation of response effort may have

PAGE 142

130 led to clearer functiona l analysis findings of peer-relat ed withdrawn social behavior for Shane and Colin. Fourth, measurement of peer correct im plementation of the functional analysis procedures included both prompted and unpr ompted peer correct implementation. For example, if the peer correctly delivered the relevant consequence in the tangible condition (i.e., peer delivered the high prefe rred tangible to the participant within 5 seconds of the participant displaying peer-rel ated withdrawn and/or peer-related negative social behavior, and allowed the participant access to the high preferred tangible for 15 seconds) it met the criteria to be scored as correct regardless if th e peer delivered the consequence independently or after being prompted by the ex perimenter. Therefore, it is unknown what percentage of peer correct im plementation of the functional analysis procedures were implemented independently by the peers or were implemented only after the experimenter prompted the peer. Anecdot ally, the level of experimenter prompts varied across functional analysis conditions. For example, peers often required prompts to present social demands to participants in the escape c ondition, though rare ly required experimenter prompts to only talk to each other (i.e., the second peer) in the ignore condition. Systematic evaluation of inde pendent and prompted correct peer implementation of functional analysis procedur es is needed to determine the affect of experimenter prompts on peer-related soci al behavior across functional analysis conditions. Additional limitations of this study include d the small sample size. The sample size was low (N = 3) and there was little variati on in age (3-years to 5years-old). A larger sample size is needed to answer the questi on of the utility of the functional analysis

PAGE 143

131 methodology to identify the function of peer-re lated withdrawn and peer-related negative social behaviors for young children with ASD. Also, implementing a treatment analysis based on the findings of the functional analysis may have been helpful in confirming or disconfirming functional analysis findings. Specifically, a tr eatment analysis may have served to further evaluate a tangible function for Shannon’s pe er-related withdrawn social behavior and escape function for Colin’s peer-related withdrawn and peer-related negative social behavior. Future Research Directions The effectiveness of the functional anal ysis methodology to identify environmental variables maintaining peer-rel ated withdrawn and peer-related negative social behaviors of children with ASD should continue to be evaluated. First, methodology issues surrounding the measurement of peer-related with drawn social behavior should be further evaluated to ensure the topogr aphy of peer-related withdraw n behavior assessed in the functional analysis represents the topography of peer-related withdrawn social behavior displayed in the natural environment. This may broaden the measurement of peer-related withdrawn social behavior to include topogr aphies such as proximity, initiations, and responses. For example, peer-related withdraw n behavior could be defined individually based on observations in th e natural environment. Second, the utility of the assessment and ev aluation of peer-related withdrawn and peer-related negative social behaviors as separate and combined topographies when analyzing functional analysis findings s hould be further evaluated. For example, evaluating combined peer-related (withdrawn and negative) social behaviors throughout the functional analysis may result in cleare r functional analysis findings. Assessing the peer-related social behaviors separately, so that only one behavior results in

PAGE 144

132 reinforcement in the functional analysis te st conditions, may also result in clearer functional analysis findings. This may be mo re time consuming as it would require two functional analysis be conducted, one for each peer-related social behavior; however, clearer functional analysis findings may outweigh additional time requirements. Third, further analysis is n eeded to determine what effect reinforcing peer-related appropriate social behaviors would have on participant peer -related withdrawn and peerrelated negative social behaviors. If both peer-related appropriate and peer-related withdrawn and negative social behaviors were reinforced, one possible outcome is an increase in appropriate social behavior and decrease in withdrawn and negative social behaviors. Additionally, potential negative side effects should be evaluated in relation to history of reinforcement for appropriate peer -related social beha vior in the natural environment and not reinforcing (i.e., placi ng on extinction) appropr iate peer-related social behavior in the functional analysis. Fourth, treatment based on the functions iden tified in the functional analysis should be conducted to further verify reinforcers id entified in the functional analysis and to determine if the treatments are successful in decreasing peer-related withdrawn and peerrelated social behaviors. For example, trea tments based on functions identified could be compared to treatments based on reinforcers that were not identified to maintain peerrelated withdrawn and/or peer -related negative social beha viors to determine if the treatment based on the function identified resulted in a better treatment outcome than the treatment that was not based on the function identified. These findings could assist in determining the utility of functional analys is methodology in assessing the environmental variables maintaining peer-relat ed withdrawn and peer-related negative social behaviors.

PAGE 145

133 Fifth, peers’ correct implementation of func tional analysis procedures (i.e., correct presentation of social demands in the escape condition, positive initiations in the free play condition, and correct delivery of consequen ces in the attention, tangible, and escape condition) should be systematical ly evaluated to determine wh at, if any, effect it has on participant peer-related withdr awn and peer-related negative social behavior, including stability across conditions. Fu rther evaluation of procedural integrity would assist in identifying criteria for determining the valid ity of functional analysis conditions when assessing peer correct implem entation of functional analys is procedures, as well as assessing the independence of peer correct implementation of functional analysis procedures (i.e., peers implemented procedures independent of expe rimenter prompt or peers implemented procedures after receiv ing a prompt from the experimenter). Finally, future research should focus on ev aluating the role of the participant’s communication, play, and cognitive skills in re lation to the findings of the functional analysis. Systematic evaluation of these f actors (i.e., communication, play, and cognitive skills) could assist in determining if these f actors influence the func tions identified in the functional analysis.

PAGE 146

APPENDIX A REVIEW OF FUNCTIONAL ANALYSIS LITERATURE

PAGE 147

135Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings Iwata, Dorsey, Slifer, Bauman, & Richman (1994) Journal of Applied Behavior Analysis reprint from Analysis and Intervention in Developmental Disabilities (1982) Manual Identify functional properties of self-injurious behavior (SIB) N: 1, A: 6-7, G: male; D: profound mental retardation (MR), autistic-like behavior T: staff, S: therapy room, N: 10 SIB C: social disapproval, academic demand, unstructured play, alone P: Functional analysis (FA) D: multielement FA: automatic function

PAGE 148

136 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings Wacker, Steege, Northrup, Sasso, Berg, Reimers, et al. (1990) Journal of Applied Behavior Analysis Manual Systematically evaluate functional communication training (FCT) via component analysis and generality of FCT across behaviors N: 1, A: 7 years, G: male, D: autism, severe to profound MR, seizure disorder T: inpatient unit staff (graduate students, therapists, and teacher) S: inpatient unit, N: information not provided SIB: hand biting Sign C: FA: escape, tangible, alone, social attention, tangible P: FA, treatment (TX): FCT D: multielement, then alternating treatment FA: Tangible function Required FCT and time-out to eliminate SIB and maintain signing

PAGE 149

137 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings Sasso, Reimers, Cooper, Wacker, Berg, Steege, Kelly, et al. (1992) Journal of Applied Behavior Analysis Journal of Applied Behavior Analysis Find: Autism Compare results of experimental and descriptive analyses; could the analyses and treatments be conducted with procedural integrity by teachers in the classroom? N: 1, A: 7-5, G: female, D: autism T: experimenter and then teacher S: session room and classroom, N: 5 Aggression and inappropriate language C: ignore, attention, escape, tangible, play P: experimenter conducted FA, teacher training and conducting antecedentbehaviorconsequence (ABC) analysis and then teacher training and conducting FA, both in classroom under natural conditions; TX D: FA: multielement, study: alternating treatment design Experiment FA: escape, tangible function Teacher ABC: escape, tangible Teacher FA: escape, tangible function

PAGE 150

138 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings Fisher, Ninness, Piazza, & OwenDeSchryver (1996) Journal of Applied Behavior Analysis Academic Search Premier Find: Functional analysis and autism What was the specific form of attention maintaining the destructive behavior? N: 1, A: 4 years, G: male, D: autism, moderate MR, & oppositional defiant disorder (ODD) T: staff, when attention conditions began, one therapist associated with related, other with nonrelated verbal content, then therapists reversed. S: session room (inpatient), N: information not provided Destruction C: tangible, demand, attention, toy play, verbal reprimand, unrelated verbal statement P: FA (Iwata et al.), then two attention conditions varying content of attention. Controlled for therapist effect. TX: noncontingent reinforcement (NCR) D: multielement Destructive behavior maintained by attention in the form of verbal reprimands.

PAGE 151

139 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings Mace, Shapiro, & Mace (1998) Journal of Applied Behavior Analysis Journal of Applied Behavior Analysis Find: Autism Effect treatment with extinction and NCR would have with and without warning stimuli N: 1, A: 7 years, G: female, D: autism and moderate MR T: experimenter, then teacher S: special education classroom N: 5 SIB C: play, escape, tangible, attention P: FA (reinforcement on variable ration [VR] 2); NCR + extinction + warning; NCR + extinction + No warning D: multielement FA: escape and tangible function NCR + extinction + Warning resulted in largest decrease in SIB

PAGE 152

140 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings Piazza, Fisher, Hanley, LeBlanc, Worsdell, Lindauer, et al. (1998) Journal of Applied Behavior Analysis PsycINFO Find: Functional analysis and autism Would functional analysis identify function for pica? Would function based treatments reduce socially maintained pica? N: 1, A: 5 years, G: male, D: autism, ADHD, moderate MR, & severe esophagitis T: staff S: inpatient session room N: information not provided Pica C: FA: attention, demand, alone (in baited room), toy play (high preferred toys) TX: baseline, noncontingent attention (NCA), then extended alone; preference assessment, treatment analysis (NCA + matched v unmatched stimuli) P: FA, TX (continuous NCA) D: FA: multielement; TX: ABAB FA: pica maintained by access to social attention TX: NCA reduced pica, but not to clinically acceptable levels; Pica continued in extended alone sessions TX: NCA + matched or unmatched stimuli effective in reducing pica

PAGE 153

141 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability(D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings Thompson, Fisher, Piazza, & Kuhn (1998) Journal of Applied Behavior Analysis (1 of 2) Journal of Applied Behavior Analysis Find: Autism Purpose: evaluate several topographies of aggression that appeared to have different functions (social and nonsocial) and treat social and nonsocial topographies N: 1, A: 7 years, G: male, D: PDD, NOS, severe MR, & severe hemophilia T: staff, S: session room, then generalized to unit living room and classroom after discharge, N: information not provided Chin grinding aggression, other aggression (hitting, kicking, pinching, scratching), appropriate communication (card exchange), alternative chin grinding (on device to stimulate chin) C: demand (staff performed self-care with him), social attention, tangible, and play; C: second FA; Attention (all aggression), attention (other aggression), NCA P: FA, descriptive assessment (DA) (hypothesized generated from therapist –child interactions in FA, and First FA: inconclusive; DA: hypothesized chin grinding automatic (unstructured times) and other aggression attention and unstructured Second FA: Other aggression maintained by social attention, chin grinding not sensitive to adult attention Aggression can occur in absence of

PAGE 154

142 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability(D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings Thompson, Fisher, Piazza, & Kuhn (1998) Journal of Applied Behavior Analysis (2 of 2) parent – child and staff child observations on unit; second FA, treatment of attention maintained aggression, treatment of chin grinding D: multielement social consequences Categorize behavior by function rather than topography Contingency of one response may not effect that behavior, but may change probability of another behavior occurring

PAGE 155

143 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability(D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings Graff, Lineman, Libby, & Ahearn (1999) Behavioral Interventions Academic Search Premier Find: Functional analysis and autism Compared findings from descriptive assessment and functional analysis N: 1, A: 6 years, G: female, D: autism & severe MR T: teacher S: empty room at residential program, N: 4 to 5 Screaming C: attention, demand, alone, tangible-toy, tangible-edible, & play and shared attention condition P: DA (ABC), preference assessment, FA, DA (setting events), FA (added shared attention condition phase); TX D: multielement DA results did not identify antecedent, suggesting automatically reinforced; FA multielement was at first undifferentiated then inconclusive (decreasing screaming across all conditions), in shared attention, again low and decreasing screaming. TX of time-out for screaming and edible for compliance decreased screaming.

PAGE 156

144 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings Richman, Wacker, Asmus, Casey, & Andelman (1999) Journal of Applied Behavior Analysis (1 of 2) PsycINFO Find: Functional analysis and autism What methods can assist in clarifying undifferentiated functional analyses? Will relative latencies of onset of behavior provide support for response class hierarchy hypothesis? N: 2, A: 8 and 4-year-old, G: male, D: PDD, autism, ADHD, moderate MR, developmentally delayed (DD) T: parent or staff S: inpatient session room N: information not provided Participant (P) P1: screams, grabbing, aggression, appropriate vocalizations, P2: disruption, aggression, appropriate mands, compliance C: FA: free play, escape, attention, alone, & tangible C: extinction P1 (tangible) screaming only, grabbing only, aggression only P2 (escape) disruption only, aggression only P: FA (all topographies of aberrant behavior reinforced); extinction analysis: FA: P1: tangible & attention functions P2: escape, tangible, attention Extinction analysis: P1: screaming only, screaming occurred, grabbing only, grabbing increased, screaming decreased, aggression only, aggression

PAGE 157

145 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings Richman, Wacker, Asmus, Casey, & Andelman (1999) Journal of Applied Behavior Analysis (2 of 2) P1: (tangible), P2 (escape) D: FA: multielement; extinction analysis: reversal and brief multielement increased, others stable, bias toward screaming (hierarchy of screaming, grabbing, disruption) P2: when aggression reinforced, disruption rarely occurred, when aggression only reinforced, displayed disruption first, then aggression, bias to disruption (hierarchy disruption then aggression)

PAGE 158

146 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings Brown, Wacker, Derby, Peck, Richman, Sasso, et al. (2000) Journal of Applied Behavior Analysis Journal of Applied Behavior Analysis Find: Autism and children Use FCT to determine if a mand matched to function and reinforced only when establishing operation (EO) present would decrease aberrant behavior and increase manding? N: 2; A: 7 and 5 years; G: male; D: autism, severe MR, possible seizure disorder, PDD, NOS, moderate MR T: graduate student, mother, and father; S: inpatient unit classroom and home (playroom); N: information not provided Aberrant behavior (SIB, aggression, noncompliance, destruction, stereotypy); mands C: free play, attention, tangible, attention/ tangible, escape P: brief and extended FA, mand assessment, FCT (EO present, EO absent, control) D: brief (multielement with rapidly changing conditions) for one, extended: multielement for second FA: escape; attention/ tangible function Mainly manded relevant mands in EO present

PAGE 159

147 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings Cunningham & O’Neill (2000) Education and Training in Mental Retardation and Developmental Disabilities PsycINFO Find Functional analysis and autism What is the agreement on function of behavior when comparing a rating scale, structured interview, descriptive observation, and functional analysis when assessing problem behavior of young children with autism? N: 3, A: 3 and 5 years, G: male, D: autism T: FA: experimenter S: DA, descriptive observation (DO): classroom, FA: other rooms in school, N: information not provided SIB, aggression, & tantrums C: demand, attention, tangible, play, & alone (2) P: general teacher question; Functional Analysis Interview (FAI), Motivation Assessment Scale (MAS), DO: Functional Assessment Observation form (FAO); FA D: FA: multielement FA: tangible and demand functions for all three participants Agreement between teacher question and FAI Two of three same primary and secondary functions identified in FA as in indirect and direct measures. For third, tangible and demand had equal ratings, but in FA tangible was primary

PAGE 160

148 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability(D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings McComas, Hoch, Paone, & El-Roy (2000) Journal of Applied Behavior Analysis Journal of Applied Behavior Analysis Find: autism Identify function of destruction, evaluate variables of demands on occurrence of escape maintained destruction. N: 2, A: 8 years, G: male, D: autism and developmental delays T: instructor affiliated with the school, familiar in Applied Behavior Analysis (ABA), S: participant’s classroom with 6 students and 4 teachers (1), college campus laboratory (1), N: information not provided Destruction (disruption and SIB) and compliance with task demands C: free play, attention, escape, tangible (only 1 participant) P: FA, DA & hypothesis generation, EO analysis D: multielement FA: escape maintained DA Hypothesis: instructional strategy, choice of task sequence Variables of tasks isolated from effects of consequence responding

PAGE 161

149 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability(D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings Rapp, Dozier, Carr, Patel, & Enloe (2000) Behavioral Interventions Academic Search Premier Find: Functional analysis and autism Determine variables maintaining hair manipulation and identify objects that might compete with hair manipulation N: 1, Age: 4 years, G: male, D: autism T: tutors, S: unfamiliar therapy room then child’s therapy room with all materials available, N: 4 conditions in four days Hair manipulation C: no interaction (materials available), attention (within academic task), demand (nonpreferred task), control P: FA conducted within typical academic schedule; sensory assessment; preference assessment D: multielement FA: hair manipulation maintained by automatic reinforcement, gloves reduced hair manipulation along with manipulating hairy/furry toys.

PAGE 162

150 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings Hagopian, Wilson, & Wilder (2001) Journal of Applied Behavior Analysis Academic Search Premier Find: Functional analysis and autism Following an inconclusive functional analysis, will a modified functional analysis lead to clearer results? N: 1, A: 6 years, G: male, D: autism & mild MR T: information not provided S: inpatient N: information not provided Aggression, disruption, SIB, & spitting C: social attention, demand [1st & 2nd FA), play (1st FA); escape from attention, tangible (2nd FA) P: 1st FA (Iwata, et al.); 2nd FA modified escape from attention and tangible; (TX): FCT and NCR D: Multielement FA; TX multiple baseline (MBL) 1st FA according to Iwata et al. was inconclusive. Modified FA identified tangible and escape from attention functions. TX based on modified FA decreased problem behavior

PAGE 163

151 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings Levin & Carr (2001) Behavior Modification (1 of 2) PsycINFO Find: Functional analysis, autism, children What variables controlled problem behavior accompanied by food selectivity? What role does access to preferred foods have prior to intervention with and without positive reinforcement? N: 4, A: 5, 6, and 7 years, G: male (3), female (1), D: autism & moderate to severe MR T: experimenter S: preference assessment (outside classroom); FA & TX: classroom N: information not provided Problem behavior: aggression, throwing food, SIB, & banging Grams of food consumed C: escape (preferred food) & escape (nonpreferred food) P: 3 food log (one at home, one at school), food preference assessment, FA, preference assessment, treatment analysis (varying assess to preferred foods prior to sessions with and without positive FA: all participants displayed more problem behavior and less consumption of food in escape nonpreferred condition TX: when food not available prior to session, consumption of less preferred food increased to access preferred food

PAGE 164

152 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings Levin & Carr (2001) Behavior Modification (2 of 2) reinforcement for consumption) D: FA: reversal, TX 1 & 3 (MBL across participants), changing criterion (condition D)

PAGE 165

153 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings Marcus, Vollmer, Swanson, Roane, & Ringdahl (2001) Behavior Modification PsycINFO Find: Functional analysis and autism Will functional analysis methodology according to Iwata et al. 1982/1994 replicate to assessment of aggression? What are the functions of aggression? N: 3, A: 4, 5, and 6 years, G: male, D: autism, ODD, mild and moderate MR T: at least 2 staff S: home and clinic N: 2 to 5 Aggression C: escape, attention, tangible, play, & ignore (1) P: DA, FA: brief FA (all 3) D: extended as needed; brief multielement (within session data analysis), extended multielement, pairwise (test – control) multielement FA: 2 of 3 participants functions of aggression were identified in brief FA, both escape maintained Third participant’s function of aggression, tangible, was identified in extended multielement functional analysis

PAGE 166

154 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability(D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings Mueller, Wilczynski, Moore, Fusilier, & Trahant (2001) Journal of Applied Behavior Analysis Journal of Applied Behavior Analysis Find: autism What effect does high, medium, or low preference tangibles have on aggression maintained by tangible function? N: 1, A: 8 years, G: male, D: autism, moderate MR T: information not provided, S: unused classroom at school, N: information not provided Aggression & engagement with restricted and noncontingently available alternative stimuli C: attention, toy play, demand, & tangible; antecedent analysis (restricting high preferred, low, and medium and replacing with high, medium, or low preferred) P: DA, FA, preference assessment, antecedent analysis D: multielement FA: aggression maintained by tangible Antecedent analysis: higher rates of aggression when high preferred tangibles restricted, but still occurred with medium and low, regardless of what alternative tangible was (high preferred, medium or low preferred)

PAGE 167

155 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings Richman, Wacker, & Winborn (2001) Journal of Applied Behavior Analysis Journal of Applied Behavior Analysis Find: Autism and children Purpose: replicate effects of response efficiency described by Horner and Day (1991), extend with response efficiency on positive reinforcement N: 1, A: 3 years, G: male, D: PDD, NOS T: mother, S: session room N: 4 to 22 Aggression, handing mother communication card, signing “please” C: free play, attention, tangible, escape C: mand analysis: card to aggression; card to signing please P: FA, mand analysis D: FA: multielement; mand analysis (fixed ratio concurrent schedule assessment) FA: aggression maintained by tangible Mand: response more to card than aggression Sign please over card and aggression

PAGE 168

156 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings Anderson & Long (2002) Journal of Applied Behavior Analysis PsycINFO Find: Functional analysis, autism, children How do results of structured descriptive assessments (SDA) compare to analogue functional analyses? Do structured descriptive assessments lead to effective treatments? N: 2, A: 8 & 6 years, G: male, D: autism & moderate to severe MR T: FA: experimenters, SDA: parent or teacher; TX: teachers aid or speech therapist S: FA: therapy room; SDA: home or classroom N: information not provided SIB, aggression, & disruption C: FA: attention, task, play, alone (1 participant), & tangible (1 participant) C: SDA: attention, task, tangible, & play P: DA, FA, SDA D: FA: multielement FA: escape (1) and tangible function (1) FA and SDA matched for one of two participants. The one that didn’t match, the FA identified tangible and the SDA identified escape TX: matched to SDA reduced problem behavior

PAGE 169

157 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings Harding, Wacker, Berg, Barretto, & Rankin (2002) Education and Treatment of Children Academic Search Premier Find: Functional analysis and autism When problem behavior is maintained by both negative and positive reinforcement, how does preferred toys and parent attention effect choice making of two children with PDD, NOS? N: 2, A: 4 and 6 years, G: male, D: PDD, NOS, communication disorder, MR level unspecified T: mother, S: home and session room, N: FA: information not provided, treatment probes: 5 FA: Problem behavior: aggression, destruction, & noncomplia nce Concurrent operant (CO): time allocation, task completion C: FA: play, attention, escape, & tangible C: CO: attention/ high preferred (HP) toy v alone/ less preferred (LP) toy and attention/LP v alone/HP P: FA; preference assessment; choice assessment, TX (multi-step choice procedure) D: FA: multielement, Choice assessment: concurrent – schedules with reversal FA: escape and tangible functions for both participants Choice assessment: Both participants chose preferred toy over parental attention TX: work then play with high preferred, for one a mand was inserted, reduced problem behavior, increased compliance

PAGE 170

158 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability(D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings LaBelle & CharlopChristy (2002) Journal of Positive Behavior Interventions Academic Search Premier Find: Functional analysis and autism Was aberrant behavior hypothesized to be multiply maintained from naturalistic observations confirmed in analogue functional analysis and functions changed within sessions N: 2, A: 8-6 years, and 8-8 years, G: male, D: autism T: graduate and undergraduate students S: therapy room at treatment center, N: 2 High frequency vocalization s and disruption C: attention, tangible, escape, play, alone, changing P: naturalistic observation; FA with changing contingency condition D: mutlielement Both participants had multiply maintained functions in the DA and confirmed in the FA. DA accurate but incomplete. Bobby: DA: attention, escape; FA: automatic, escape, changing. Richard DA: escape, attention; FA: attention, escape, changing. Both also demonstrated aberrant behavior in tangible.

PAGE 171

159 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings Tang, Kennedy, Koppekin, & Caruso (2002) Journal of Applied Behavior Analysis Academic Search Premier Find: Functional analysis and autism Could social or nonsocial functions of ear covering be identified? N: 1, A: 5 years, G: male, D: autism T: information not provided, S: classroom DA, empty classroom (FA), N: 5 Stereotypy: ear covering C: alone, attention, control, demand, alone + noise, demand + noise, attention + noise, P: forty three 30 min DO, audiological exam, FA D: multielement DO: antecedent was peer screaming, consequence no change in interaction. FA highest stereotypy in alone + noise, indicating maintained by negative sensory reinforcement (noise attenuation) Importance of linking DA to FA

PAGE 172

160 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings Asmus, Franzese, Conroy, & Dozier (2003) School Psychology Review Academic Search Premier Find: Functional analysis and autism Would consequating only one of two problem behaviors in functional analysis clarify the function of the consequated behavior? N: 1, A: 7 years, G: male, D: autism and language impairment T: doctoral level student, S: unused room at school, N: information not provided Stereotypic vocalizations & disruptive behavior C: free play, attention, tangible, escape, & alone P: FA consequating disruption and/or stereotypy; functional analysis consequating just disruption D: multielement When stereotypy and disruption were reinforced, stereotypy automatically maintained, no function identified for disruption. When only disruption reinforced, disruption maintained by tangibles, when view graphs with separate topographies. For both FA’s aggregate graph looks automatic

PAGE 173

161 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings Crosland, Zarcone, Lindauer, Valdovinos, Zarcone, Hellings, et al. (2003) Journal of Autism and Developmental Disorders Academic Search Premier Find: Functional analysis and autism Whether destructive behavior was reduced with risperidone and if it occurs under similar environmental conditions? N: 1, A: 6 years, G: male, D: autism, profound MR, & fragile X syndrome. T: information not provided S: empty classroom at school N: 1 Destruction, appropriate (complianc e), & number of demands C: demand, attention, free play, & tangible P: initial placebo, medication, placebo. Phases 1 to 2 weeks. D: multielement (FA), doubleblind, placebocontrolled trial (medication) In BL and placebo consistent destructive behavior in demand and tangible. With medication destructive behavior decreased in demand, but remained consistent in tangible across phases. When back to BL, destruction in demand increased. Destruction at zero in all play conditions across phases

PAGE 174

162 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability(D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings Finkel, Derby, Weber, & McLaughlin (2003) Journal of Positive Behavioral Interventions (1 of 2) Academic Search Premier Find: functional analysis, young, children How choice can be used to identify function when functional analysis inconclusive? N: 1, A: 5 years, G: female, D: autism, developmental delays T: special education teacher S: classroom N: information not provided Elopement Disruption (noncompliance) Appropriate behavior; choice responding C: FA: free play, escape, attention, tangible C: choice P: parent teacher interview; brief FA, preference assessment, reinforcer assessment, choice analysis D: Brief FA: multielement rapid reversal Preference assessment: multipleschedule Choice: concurrent FA: escape function Choice: chose less demanding task

PAGE 175

163Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability(D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings Finkel, Derby, Weber, & McLaughlin (2003) Journal of Positive Behavioral Interventions (2 of 2) schedules combined with reversal design

PAGE 176

164 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings Najdowski, Wallace, Doney, & Ghezzi (2003) Journal of Applied Behavior Analysis PsycINFO Find: Functional analysis, autism, young Purpose: parent implement functional analysis and serve as primary data collector during treatment. N: 1, A: 5 years, G: male, D: ASD T: parent, S: session room, then home, and restaurant N: information not provided FA: Food refusal TX: bites accepted and swallowed C: no interaction, attention, play, & escape P: FA; differential reinforcement of alternative behavior (DRA), DRA + escape extinction (EE) + demand fading D: FA: multielement, TX: MBL across settings FA: escape function DRA + EE + demand fading effective over time. Parent implemented all phases with little supervision

PAGE 177

165 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings Piazza, Fisher, Brown, Shore, Patel, Katz, et al. (2003) Journal of Applied Behavior Analysis Manual Will functional analyses methodology assess inappropriate mealtime behavior of children with pediatric feeding disorders? N: 2, A: 7 and 4 years, G: male, D: autism, ADHD, severe MR, cerebral palsy, & poor vision T: natural observations: parent, FA: staff S: session room N: 15 In natural observations: parent behavior: escape, attention, tangible; child: acceptance, inappropriate mealtime behaviors FA: Inappropriate mealtime behaviors: batting, aggression, throwing food, head turning, & negative vocalizations C: natural observations, FA: play, escape, attention, & tangible P: natural observation (1 participant) and FA (both participants) D: FA: reversal (test condition compared to control) Natural observations: parent provide attention and escape from meal contingent on inappropriate mealtime behavior. FA: undifferentiated and attention function

PAGE 178

166 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings English & Anderson (2004) Research in Developmental Disabilities PsycINFO Find: Functional analysis Effect caregiver versus experimenter conducting FA had on findings? Are variables effecting behavior in natural environment present in FA? N: 1, A: 5 years, G: male, D: autism & mild MR T: mother and experimenter S: therapy room N: information not provided Aggression & disruption C: play, tangible, attention, & demand P: DA (scatter plot and FAI), caregiver training of FA, FA D: FA: multielement within a reversal FAI: hypothesized escape function and attention; Scatter plot hypothesized attention function FA caregiver: attention and tangible functions FA experimenter: attention function (lower responses per min than with parent)

PAGE 179

167 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings Luiselli, Ricciardi, Schmidt, & Tarr (2004) Child & Family Behavior Therapy PsycINFO Find: Functional analysis Can brief FA identify function of saliva play? Can function based treatment of saliva play be evaluated quickly? N: 1, A: 6 years, G: male, D: autism T: information not provided S: classroom N: information not provided, total of 8 days for FA, 3 for treatment Saliva play C: FA: attention, demand, play, & ignore, TX: gum & chew object P: Brief FA, TX D: FA: multielement, TX: alternating treatment FA: undifferentiated function TX: chew object decreased saliva play

PAGE 180

168 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings Hagopian, Contrucci Kuhn, Long, & Rush (2005) Journal of Applied Behavior Analysis PscyINFO Find: Functional analysis Would providing competing stimuli during FCT allow schedule thinning to occur more quickly? N: 1, A: 7 years, G: male, D: autism, ADHD, & moderate MR T: staff S: inpatient session room N: information not provided FA: aggression Competing items assessment: engagement TX: handing a picture and saying, “I want to play” C: FA: social attention, physical attention, tangible, demand, alone, & toy play P: FA; (RAISD, competing stimulus assessment; FCT + extinction, FCT + extinction + competing stimuli; schedule thinning D: FA multielement TX: reversal within multielement (ABAB [BC]) FA: attention in the form of verbal and physical attention, and tangible functions Competing stimuli identified Schedule thinning proceeded more quickly with FCT when competing stimuli were present

PAGE 181

169 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings Hanley, Piazza, Fisher, & Maglieri (2005) Journal of Applied Behavior Analysis PsycINFO Find: Autism Purpose: Use of preference assessment methodology to assess child’s preference for function based intervention with and without punishment N: 1, A: 5 years, G: male, D: moderate MR, autism, & seizure disorder T: staff S: Inpatient session rooms, N: 5 FA: SIB, aggression, & disruption FCT: hand card Preference: pressing 1 of 3 microswitches prior to session to determine which condition in effect C: FA: attention, escape, tangible, alone, & play P: FA; FCT training trials, FCT + extinction, NCR, FCT + Punishment (hands down for 30 s); TX: preference evaluation (choices: FCT, FCT + punishment, punishment only) D: FA: multielement TX: multielement and reversal FA: attention function FCT + extinction and NCR reduced problem behavior, but not consistently, FCT + punishment decreased problem behavior consistently. Preference: FCT + punishment

PAGE 182

170 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Setting (S) Number session per week (N) Dependent Measures Experimental Conditions (c), Procedures (P) Design (D) Findings Sidener, Carr, & Firth (2005) Journal of Applied Behavior Analysis Academic Search Premier Find: Functional analysis and autism Effectiveness of superimposition and withholding of edible consequences as treatment for automatically maintained stereotypy (replicating and extending through by pretreatment functional analysis)? N: 2, Age: 6 years, G: female, D: autism, S: home T: information not provided S: home, N: 4 to 5 Stereotypy: scratching C: attention, demand, no interaction, & control. 1 participant had tangible condition P: FA, TX evaluation D: multielement design, followed by several consecutive no interaction sessions FA: Automatic and undifferentiated function Superimposition and withholding tangibles didn’t decrease stereotypy, but environmental enrichment did

PAGE 183

APPENDIX B REVIEW OF SOCIAL SKILLS LITERATURE

PAGE 184

172 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings McGee, Almeida, SulzerAzaroff, & Feldman (1992) Journal of Applied Behavior Analysis (1 of 3) Academic Search Premier Find: social competence and autism Will peer incidental teaching increase the peers social interaction to children with autism? Will it increase the initiations and responses in children with autism? N: 3, A: 3, 4, & 5 years, G: all males, D: all autism Peer tutors: N: 3, A: 4, 4, & 4 years, G: all female, D: typically developing Chosen based on age (oldest), attendance, teacher report of compliance and age appropriate social skills, high status (peer sociometrics and teacher rating) Comparison peers N: 2, A: 3 & 3 T: experimenter and peer S: integrated preschool N: 5 Social peerinteractions, initiations, and responses D: MBL across participants BL: Peer training of incidental teaching: training occurred in free play with peer and target child with high preferred toys. Training: wait for target child to initiate, ask to label toy, give toy after labels it, praise for labeling, prompting to take turns. Experimenter trained peer by For all 3 participants, interactions increased with peer training (and only increased when peer training implemented) In fading, interactions remained above baseline, though, even though above baseline in fading 2, it was lower than fading 1

PAGE 185

173 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings McGee, Almeida, SulzerAzaroff, & Feldman (1992) Journal of Applied Behavior Analysis (2 of 3) years, G: one male, one female D: typically developing modeling and then faded to use of picture prompt checklist. After all pairs had 8 sessions of training started rotation of peer trainers to target children to assist with generalization Training ended after 9 consecutive session of 45% (of opportunities) successful prompted responses from target child

PAGE 186

174 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings McGee, Almeida, SulzerAzaroff, & Feldman (1992) Journal of Applied Behavior Analysis (3 of 3) TX: Peer incidental teaching: Fading 1: experimenter decreased proximity and prompts Continued until at least 3 sessions, and when interactions continued with no prompts Fading 2: experimenter out of room

PAGE 187

175 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings Sainato, Goldstein, & Strain (1992) Journal of Applied Behavior Analysis (1 of 3) PsycINFO Find: social skills interventions Will peers use of selfevaluation increase their independent use of facilitative strategies with a child with autism? Will that result in higher rates of social behavior with the child with autism? Will it increase use of these strategies N: 3, A: 3, 4, & 4 years, G: all males, D: all autism Peer confederates: N: 3, A: 3, 4, & 4 years, G: one male, two females, D: typically developing Untrained Peers: N: 2, A: 3 & 4 years, G: information not provided, D: typically developing T: teacher S: integrated preschool During sociodramatic area during play N: 5 Social behavior Peer selfevaluation ratings D: MBL across participants Pre training and post training BL: Triads: 1 child with autism, 1 peer confederate, 1 untrained peer Poster with the four facilitative strategies present. Peer confederate strategy training (as a group): attention getting, play organizers, For all 3 peers, strategy use increased during play, though it did not increase until selfevaluation was implemented. Decrease in the number of teacher prompts for all 3 participants when selfevaluation implemented For all 3 target children with autism, an increase in

PAGE 188

176 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings Sainato, Goldstein, & Strain (1992) Journal of Applied Behavior Analysis (2 of 3) with other children with autism? Will untrained peers who observe these peers increase their facilitative strategies with children with autism? sharing, and responding Peer selfevaluation training (individually): forecasting and practicing with adult, role reversal with adult, and training with target child with autism. Peer had selfevaluation book to mark if did correctly or incorrectly. Peer had to have 3 out of 5 correct without prompting to move on. an increase in social behavior occurred when peers began selfevaluation Two of three peers had to be systematically taught selfevaluation with generalization target children, did not generalize use of strategies on their own, but did with further training.

PAGE 189

177 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings Sainato, Goldstein, & Strain (1992) Journal of Applied Behavior Analysis (3 of 3) Peer selfevaluation condition: experimenter rated during session, peer self-evaluated after session. If peer met criteria and matched experimenter ratings, earned small toy. Generalization: with different target children with ASD

PAGE 190

178 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings Strain & Kohler (1994) Journal of Emotional and Behavioral Disorders (1 of 4) Academic Search Premier Find: N/A Purpose: extend selfmonitoring to social interactions between young children with autism and their typically developing peers. Compare these results across home and school. Systematically fade adult prompts and reinforcement. N: 3, A: 3, 4, & 5 years, G: all males, D: autism Peers: N: 10, A: 3 to 5 years, G: information not provided, D: typically developing T: teacher and peer S: quiet room in integrated preschool N: 5 Duration of positive social interaction (positive initiations, concurrents, and yes responses) D: MBL across participants BL: Groups: dyads (target child and one peer), peers rotated each day Teacher introduced activity, no prompts TX: social skills package by Kohler et al. (1990) implemented by teacher. The peer group of 2 to 3. Three skills taught: Social skills training with use of poster did not increase any participants’ social interactions. Teacher facilitation increased social interaction for the one participant received it. Selfmonitoring package increased social interactions for all 3 participants

PAGE 191

179 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings Strain & Kohler (1994) Journal of Emotional and Behavioral Disorders (2 of 4) suggesting play organizers, share offers and requests, and assistance offers and requests Then taught three strategies for using each skills, 1) initiate/extend interactions, 2) respond positively to initiations, & 3) persist Teacher taught skills by introducing, modeling, having children practice with

PAGE 192

180 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings Strain & Kohler (1994) Journal of Emotional and Behavioral Disorders (3 of 4) adult and each other and then practicing independently Poster illustration: with illustration of skills, no prompts provided Teacher facilitation: implemented for only one participant, verbal prompts and reinforcement poster Selfmonitoring:

PAGE 193

181 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings Strain & Kohler (1994) Journal of Emotional and Behavioral Disorders (4 of 4) poster available, teacher provided prompts and post session reinforcers. If engaged in positive initiation, response or concurrent, put foam picture in container. If reached goal, target child and peer received reinforcer

PAGE 194

182 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings Strain & Danko (1995) Journal of Emotional and Behavioral Disorders (1 of 2) Academic Search Premier Find: N/A What is the development of engagement and social competence skills in young children with autism and does increasing one skill increase the other? N: 5, A: 4, 4, 5, 5, & 6 years, G: all males, D: autism Peers N: 17, A: 3 to 5 years G: 7 male, 10 female, D: typically developing T: teachers S: Classroom in half day integrated preschool for children with autism and typically developing children N: 5 Classroom structure (classroom ecology, teacher participation, child participation) Social interaction (initiations, response, & concurrents) D: reversal BL: 25 min free play Engagement TX: teacher instructed to use incidental teaching to facilitate engagement (following day, teacher told if child’s engagement increased or decreased) Social intervention: teachers facilitated target child’s Engagement intervention conducted for 4 of 5 participants fifth engaged in high levels of engagement in BL. Of these 1 participant displayed a marked increase in engagement, the remaining 2 displayed a slight increases in engagement. Zero displayed an increase in

PAGE 195

183 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings Strain & Danko (1995) Journal of Emotional and Behavioral Disorders (2 of 2) interaction with peers (following day, teacher told if child’s social interaction increased or decreased) social interaction. Social intervention increased peer-related social interaction for all participants and did not increase engagement

PAGE 196

184 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings Zanolli, Daggett, & Adams (1996) Journal of Autism and Developmental Disorders (1 of 3) Manual Can priming increase the spontaneous initiations of children with autism to their typically developing peers? Can priming address these issues (in implementing interventions to increase initiations in classrooms): initiations occurring spontaneously, result in high probability of a response, N: 2, A: 4 & 4 years, G: both males, D: both autism Peers: N: 10, A: 4 to 6 years, G: 4 males, 6 females, D: typically developing Peers chosen if approached target child in prebaseline, verbally expressed interest in them, or observed to frequently play with same T: teacher and peer S: university integrated preschool, attended 3 hours per day, 4 days per week N: 3 Target child initiation, peer response, peer initiation, peer delivery of tangible, teacher praise, & teacher prompt D: MBL across activities BL: Activity: target child brought to randomly selected preferred activity area, peer asked to join, teacher instructs them to greet each other and then to play (peer responds and gives reinforcer during play time Unprompted initiations increased for both participants following priming. For 1 of the 2, unprompted initiations increased only following the implementation of priming across 3 activities. The other participant did not increase unprompted initiations in the second activity until it was

PAGE 197

185 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings Zanolli, Daggett, & Adams (1996) Journal of Autism and Developmental Disorders (2 of 3) occur same rate as typically developing peers, use a variety of initiations, not burden to teacher? preferred tangibles, and nominated by teacher as having good social skills. contingent on initiations). TX: Target child priming: conduct immediately before activity session. Teacher prompted target child to complete a variety of initiations to peer (initiations resulted in reinforcer) Whole class: priming and activity sessions generalized to the whole class and priming was reduced. For both participants the number of topographies of initiations increased only with the when priming was introduced to each activity. The number remained higher than baseline when generalized to whole class and reduced consequence to priming, but

PAGE 198

186 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings Zanolli, Daggett, & Adams (1996) Journal of Autism and Developmental Disorders (3 of 3) conducted during free play when all students present Reduced priming: from 14 to 7 and 5 trials did decrease for both participants. Peers: Unprompted delivery of participant increased from near zero to 64% and 61% in priming. In final phase, the teacher did not have to provide more than one prompt to peers per session to deliver the consequence (12 and 6 weeks for participants, respectively).

PAGE 199

187 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings GarrisonHarrell, Kamps, & Kravits (1997) Focus on Autism and Other Developme ntal Disabilities (1 of 2) Manual What are the effects of peer network (utilizing augmentative communicatio n) on target child social interaction; generalization across settings and activities; language and disruptive behavior; and peer acceptance in general education classrooms? N: 3, A: 7, 7, and 6 years, G: 2 males & 1 female, D: all three had diagnosis of autism, 1 with developmental apraxia Additional: criteria of communication difficulties and limited social skills T: 15 typically developing first graders (5 from each class). Selection criteria: high-status from sociometric assessment, compliance with adults, consistent school attendance, ageappropriate social skills and expressive/recepti ve language skills S: Self contained classroom, general education classroom, lunchroom, library, computer room, & playground. Duration of social interaction (initiationresponse sequence) Duration using augmentative communicat ion to initiate, respond, and/or interact D: MBL across settings (within MBL across participants) BL: groups of 4 students (rotated peers) across settings; TX: Augmentative communication training for target students Peer training (in groups of 5) with augmentative communication and social skills (initiation, response, complimenting, Peer network with augmentative communicat ion, increased frequency and duration of social interactions for all 3 participants from baseline, and generalized across settings and activities

PAGE 200

188 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings GarrisonHarrell, Kamps, & Kravits (1997) Focus on Autism and Other Developme ntal Disabilities (2 of 2) N: 3 to 4 sharing, providing instructions, sharing ideas, and maintaining conversation); Peer network across settings: target student and 5 peers engaged in 20 min cooperative activity

PAGE 201

189 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings GonzalezLopez & Kamps (1997) Focus on Autism and Other Developmental Disabilities (1 of 2) Academic Search Premier Find: N/A Purpose: Effect of social skills training alone and in small group teaching, and with reinforcement on social behavior of young children with autism and behavior problems with typically developing peers. Can peers be trained in basic behavior management? N: 4, A: 5, 5, 7 & 7 years, G: 2 males and 2 females, D: all with autism Peers N: 12, A: 5 to 8 years, G: 7 males, 5 females, D: typically developing Peers matched to target child by age T: 2 special education teachers and peers S: special education classroom N: 3 to 4 Social interaction (frequency and duration) Use of specific social skills Disruption D: reversal with two interventions BL: 20 min play groups with teacher, no prompts Groups were 4 children and 1 teacher at table with materials. 1st TX: 10 min of teacher led social skills training, 10 – 15 min play (no feedback) 2nd TX: same as first TX, though with Duration and frequency of social interaction increased for 3 of 4 target children with social skills training and for all participants with social skills training plus reinforcement All peers had some decrease in duration of interaction in final TX condition, variable for both duration

PAGE 202

190 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings GonzalezLopez & Kamps (1997) Focus on Autism and Other Developmental Disabilities (2 of 2) verbal feedback and received star for appropriate interactions. Three or more stars received prize Return to BL 2nd TX implemented and frequency of interaction Inappropriate behaviors decreased in treatment for 3 of the 4 participants. 2 peers engaged in disruptive behavior frequently

PAGE 203

191 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings Kamps, Potucek, Lopez, Kravits, & Kemmerer (1997) Journal of Behavioral Education (1 of 2) Manual Will peermediation (with support networks, training, scripts, and reinforcement): increase social interaction? Feasible for natural settings? Implemented by teachers, paraprofessionals? Generalize to new settings? N: 3, A: 6, 8, & 8 years, G: three males, D: three with autism Peers: N: 31, A: not given, but 22 in second grade and 9 in Kindergarten, G: specifics not given, but close to half male and half female, D: typically developing Chosen by teacher based on good social skills, parent consent, and no T: teacher and peers S: Kindergarten and second grade public schools Activities: academic, center/game, lunch, & recess N: 3 to 4 Initiations, responses, & interactions D: MBL across settings BL across all activities: no set prompts or opportunities for social interaction TX: Peer network intervention, individualized to each participant and activity. Criteria for intervention: 3 to 4 session per week, at least 10 min session with at least 1 peer (prefer 2For all 3 participants, a consistent increase in social interactions only occurred across all four settings with the implementation of the intervention. For two of the three participants, increases in the two nontrained setting began to occur over time without specific training (i.e.,

PAGE 204

192 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings Kamps, Potucek, Lopez, Kravits, & Kemmerer (1997) Journal of Behavioral Education (2 of 2) negative history with target child 5), adult present for training, reinforcement, feedback, social activity, peer training (modeling, adult-student practice, and peer-student practice), scripts, visual reinforcement for peer interaction, and adult feedback at end of session generalization occurred

PAGE 205

193 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings Pierce & Schreibman (1997b) Focus on Autism and Other Developmental Disabilities (1 of 2) Academic Search Premier Find: Social skills and autism Purpose: to replicate effects of peer pivotal response training (PRT) to increase language in children with autism. Determine effects on toy play and generalization after peer PRT. N: 2, A: 7 & 8 years, G: both males, D: both autism Peers: Typically developing peers N: 8, A: 7 to 8, G: 4 males, 4 females, D: typically developing Chosen based on: cooperative, friendly, good attendance T: peers S: classroom and recreation room Generalization to novel classroom N: information not given Language during interaction (appropriate words and number of words per sentence) D: MBL across participants and peers BL: dyad told to play with toys, no instructions or prompts given TX: PRT: Conducted with experimenter over a 2 week period. Taught: paying attention, giving choices, vary toys, model appropriate social behavior, Both participants increased their average use of language (frequency and quality) with each peer following PRT training

PAGE 206

194 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings Pierce & Schreibman (1997b) Focus on Autism and Other Developmental Disabilities (2 of 2) reinforce attempts, encourage & extend conversation, take turns, narrate play, teach responding to several cues.

PAGE 207

195 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings Pierce & Schreibman (1997a) Journal of Applied Behavior Analysis Academic Search Premier Find: Social skills intervention and autism Replicate Pierce & Schreibman (1995) and assess generalization with untrained peers N: 2, A: 7 & 8 years, G: male, D: autism T: peers, trained by experimenters S: training: in classroom during recess (1) and recreation room (1), generalization setting: third grade classroom during recess Peers: 8 typically developing (3 for each participant, from different classrooms and one for each from different classrooms in generalization). Peers were 8 years old, one generalization peer was 9 years old. N: 1 Maintaining interactions, initiates conversation, initiates play D: MBL across peer trainer BL: toys available, dyad told to play together TX: peers taught pivotal response training strategies Post treatment probes Initiations and maintaining interactions increased from baseline across all trained peers for both participants Participants generalized across participants, settings, and stimuli.

PAGE 208

196 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings Laushey & Heflin (2000) Journal of Autism and Developmental Disorders Academic Search Premier Find: Social skills and autism Will peerinitiation training increase social skills of children with autism more effectively if all peers in class are taught (including other children with autism) or if the child with autism is simply put in proximity to untrained peers N: 2, A: 5 & 5 years, G: both males, D: one with autism, one with PDD, NOS T: Peers ages 5 to 6 years, half male and half female S: two separate, integrated kindergarten classrooms N: data taken once every 10 days Appropriate social skills: (Asking for an object and responding according to the answer given; Appropriately getting the attention of a peer; Waiting for a turn; Looking at or in direction of peer speaking to target child) D: reversal BL: passive proximity peer tutoring TX: active peer tutoring (“buddy system”) trained to all, including child with autism. All children had a buddy that rotated daily during free play. Buddy pairs who did well got special treat (treasure chest), discontinued after 4 weeks, as not required In comparison to baseline, buddy pairs intervention increased percent of appropriate social skills for both participants

PAGE 209

197 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings Garfinkle & Schwartz (2002) Topics in Early Childhood Special Education (1 of 3) Academic Search Premier Find: Social skills and autism Purpose: evaluate effects of peer imitation package on social initiation and social interaction, and the generalization of those social behaviors N: 3, A: 3, 4, & 5 year, G: three males, D: three with autism, three with cognitive delays Peers: same 4 to 5 peers (some with and some without disabilities) in small group intervention and all of class in free play, N: 15 and 16, A: 3 to 6 years, G: unknown, D: approximately T: assistant teacher (bachelor’s degrees and current graduate students) S: integrated university affiliated preschool (approximately half of class with special needs and other half typically developing) N: small group: 1, free play 4 Small group: target child and peer initiations, positive responses, negative responses, no responses, independent peer imitations, prompted peer imitations Free play: same as small group, with the addition of: nonsocial engagement, proximity, prompt (adult directive or support) D: MBL across 2 participants (2 were in same class, so they started TX at same time) BL: Small group (same 4 to 5 children, some with and without disabilities), supervised by assistant teacher for 15 min Free play: 40 to 60 min after small group, and lasted 30 No peer imitation observed in BL. A small increase in peer imitation in treatment. Two of three participants independently imitated peers, one required more prompting. Social interactions were variable and low in baseline and did not increase in treatment. Generalization: low and variable peer imitation in free play. Slight increase in social interaction, but still low and variable

PAGE 210

198 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings Garfinkle & Schwartz (2002) Topics in Early Childhood Special Education (2 of 3) half with disability and half typically developing min. Same classroom used, but access to larger space in room. Continued in BL throughout small group intervention TX: small group peer imitation training for 10 min in small group. 1) teacher instruction, 2) selecting leader, 3) prompts for peer imitation, & 4) praise for imitation. Each days TX continued until all children were leaders twice. On follow-up only 1 of 3 participants displayed peer imitation, and that was for only one day All 3 participants had increase in nonsocial engagement in treatment and it maintained in follow-up. All 3 participants had increase in proximity

PAGE 211

199 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings Garfinkle & Schwartz (2002) Topics in Early Childhood Special Education (3 of 3) Leader’s switched after all children were simultaneously imitating the leaders play (teacher prompts and praise used to achieve this). Generalization: continued free play baseline after small group Follow-up: returned to BL in small group for 2 of 3 participants

PAGE 212

200 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings Shabani, Katz, Wilder, Beuchamp, Taylor, & Fischer (2002) Journal of Applied Behavior Analysis Academic Search Premier Find: Social skills and autism Replicate Taylor and Levin, extend by evaluating participant responses, and fading prompts N: 3, A: 6, 7, & 7 years, G: male, D: autism T: experimenter Peers: Peers not provided with training or instruction. Variety of peers from participants’ general education class in schools, same peer at home (no mention of development). S: school (2) and home (1) N: information not provided Participant verbal initiations, participant verbal responses to peer initiations D: ABAB BL: during free play situation TX: training (to initiate to adult following prompt every 1 min). Trained on three initiations. Tactile prompts: pager vibrator when activated by remote at least once every 25 s repeated if no initiation was made) during free play with peer. Prompt fading Tactile prompt increased verbal initiations and verbal responses for all participants Prompt fading partially successful for one of two participants

PAGE 213

201 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings Wert & Neisworth (2003) Journal of Positive Behavior Interventions PsycINFO Find: social skills and autism Will video selfmodeling (VSM) increase spontaneous requests of young children with autism? N: 4, A: 4, 4, 5, & 5 years, G: all males, D: all autism T: experimenters and trained ABA therapists employed by families S: observations of spontaneous requests at school during play sessions, Home (making and watching video tape) N: 5 Spontaneous requesting (SR): independently asking for objection or action D: MBL across participants BL: at school during 30 min play sessions Adult prompters (ABA therapist) trained 5 min video tape of each child’s SR behaviors made at home TX: watch 5 min videotape (within 60 min of school) each day for 5 consecutive days, then observation at school For all 4 participants, spontaneous requesting increased following BL with VSM intervention Maintenance data for 3 of 4 and they maintained SR’s over 2-6 week period

PAGE 214

202 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings Thiemann & Goldstein (2004) Journal of Speech, Language, and Hearing Research (1 of 6) Academic Search Premier Find: social skills and autism Will an intervention with peer training and written text treatment (WTT) increase target children’s social behavior? What are collateral effects on peers (trained and untrained) response & acceptance of target children? Are changes socially N: 4, A: 6, 7, 7, 7 years old, G: four males, D: two autism, two Asperger’s Peers: N: 10, A: 7 to 9 years, G: information not given, D: typically developing, 2 received resource for learning disabilities Selected based on: teacher recommendation of high status peers (sociometrics scored in top 30%), age-appropriate social skills, willingness to T: experimenter and peers S: 3 of 4 in general ed classroom during activities social academic activities, board games, etc, 1 of 4 in a resource room N: 3 to 4 for 10 to 24 min each session Appropriate social behaviors: securing attention, initiating (comments, compliments, requests for information, requests for actions/objects) and responses Inappropriate behaviors: other and no response Peer behavior measured: appropriate initiations & responses (same as above), and D: MBL across participants and behaviors Preand post: SSRS teacher report social skills subtest BL: triad (1 target child, 2 peers) had 10 min social activity. Prior to session peers received session agenda, job list, and materials. No prompts given to peers (unless to stay in area) Peer training: conducted with After peer training, some increase in social behavior for 3 of 4 participants, the fourth continued to have variable levels. Three of four continued to increase social behavior from peer training to WTT, though the fourth remained variable. All participants

PAGE 215

203 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings Thiemann & Goldstein (2004) Journal of Speech, Language, and Hearing Research (2 of 6) valid? participate appropriate initiations (from above) only each peer dyad before or after school by researcher. Taught 5 skills, each skill was taught in 30 min for a total of 150 min over 5 days. Skills: 1) look, wait and listen, 2) answer questions, 3) keep talking, 4) say something nice, & 5) start talking Training: 1) discussing 4 behavioral steps per skill, 2) writing peergenerated increased use of the skills target for intervention, only when intervention for that skill, and only after intervention for that skill was implemented. Maintenance was achieved for 3 of 4 participants. All peers increased responsiveness to target child post WTT, though only 3

PAGE 216

204 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings Thiemann & Goldstein (2004) Journal of Speech, Language, and Hearing Research (3 of 6) examples of skill by pictures of children talking, 3) role play with adultpeer and peerpeer, 4) adult feedback, 5) review of behavioral steps Post-peer training BL: prior to 10 min activity peer dyad shown list of 5 skills and told if have 5 happy faces (use skill two or more times) receive prize from treasure chest). If no skill used in 1 of 4 increase responsiveness following peer training One of four teacher reported positive change in social behavior on SSRS.

PAGE 217

205 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings Thiemann & Goldstein (2004) Journal of Speech, Language, and Hearing Research (4 of 6) min, adult prompted peer by showing index card with skill described TX: WTT: 3 target communication skills individually chosen for each target child based on BL info (no upward trend observed in skill) Target child and peers instructed on each skill for 25 min (10 min instruction, 10 min activity, 5 min feedback & reinforcement),

PAGE 218

206 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings Thiemann & Goldstein (2004) Journal of Speech, Language, and Hearing Research (5 of 6) 3 to 4 times per week using written text and picture cues. Prior to session, adult reviewed target skill, wrote words in bubble (visual cue for session), students rehearsed scripts in role play, target child read scripts, students set goal for number of times use skill (based on previous session). Adult provided feedback to target child

PAGE 219

207 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings Thiemann & Goldstein (2004) Journal of Speech, Language, and Hearing Research (6 of 6) during session by marking happy face for use of skill or prompting use of skill if not used in 1 min. If met pre-set goal each child received prize. After increase in skill for four sessions, second skill treatment began. After third skill, maintenance data collected on first 2 skills

PAGE 220

208 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings Apple, Billingsley, & Schwartz (2005) Journal of Positive Behavior Interventions (1 of 5) PsycINFO Find: Social skills and autism Experiment1 Will teaching high functioning children with ASD compliment giving initiations and response through video modeling increase the behaviors? Experiment2 Will selfmanagement increase complement Experiment 1 N: 2, A: 5 & 5 years, G: both males, D: Asperger’s & autism, no MR Experiment 2 N: 3 (1 from experiment 1), A: 4, 5, & 5 years, G: 2 males, 1 female, D: 2 Asperger’s, 1 autism, no MR Peers present Experiment 1 T: experimenter and peers S: integrated preschool N: 3 Experiment 2 T: experimenter and peers S: integrated preschool and kindergarten N: 3 Experiment 1 Compliment giving responses Experiment 2 Compliment giving initiations and responses D: MBL across participants BL: 15 min free play in classroom TX: Video modeling: 4 video segments of classmates for each participant. 3 of 4 videos modeled complimentgiving responses (1 segment for each sentence structure) and the fourth Both participants increased complimentgiving with video modeling though did not initiate. When reinforcer added, initiations increased for both. When video withdrawn and kept reinforcer, initiations and compliment giving remained stable. When

PAGE 221

209 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings Apple, Billingsley, & Schwartz (2005) Journal of Positive Behavior Interventions (2 of 5) giving? modeled complimentgiving initiations One response and one initiation video segment shown daily, prior to free play, then data taken in free play Video modeling plus reinforcement: tangible for 4 compliments and 2 peers were prompted to approach target child to provide opportunities reinforcement withdrawn, responses remained stable, initiations decreased to zero. Experiment 2 In training phase, all met criteria for compliment giving initiations and continued to classroom – they all passed that to move to selfmanagement.

PAGE 222

210 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings Apple, Billingsley, & Schwartz (2005) Journal of Positive Behavior Interventions (3 of 5) Reinforcement only: video taken out Withdrawal of reinforcer: BL though with 2 peer approaches Experiment 2: D: MBL across participants to replicate Experiment 1 and assess selfmanagement The two participants who were not in Experiment 1, participated in the same BL Increased compliment giving initiations for all three with selfmanagement. 2 of 3 needed only one prompt during this phase to initiate a compliment, otherwise they independently met criteria using selfmanagement. Both maintained response compliments

PAGE 223

211 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings Apple, Billingsley, & Schwartz (2005) Journal of Positive Behavior Interventions (4 of 5) and video modeling phases as Experiment 1. Screening: verified could differentiate compliments. Selfmanagement teaching phase trained until could respond twice with compliments, self record then, and pick prize with no prompts. Then conducted in classroom during free in generalization to outside and small group

PAGE 224

212 Author/Year Search Research Questions Participants -Number (N) -Ages (A) -Gender (G) -Disability (D) Therapist (T) Settings (S) Number of sessions per week (N) Dependent Measures Design (D) Assessment & Treatment Components Findings Apple, Billingsley, & Schwartz (2005) Journal of Positive Behavior Interventions (5 of 5) play, had to complete (as above) for 2 sessions with no prompts to leave training phase Self management: during free play in integrated classroom, reduced prompts, if 2 compliment responses, earned prize. Then generalized for 2 of 3 participants to outside and small group

PAGE 225

213 APPENDIX C LIST OF PEER-RELATED POSITIVE INITIATIONS OBTAINED DURING DESCRIPTIVE ASSESSMENT (PHASE 1) Table C-1. Shane’s Verbal and Gestural Initiations. Verbal Initiations Gestural Initiations Give me five Shane laying on top of peer Do you want a sticker? Shane sitting on peer’s lap Lets go get a sticker Gently touching another’s head Squish it (play-doh) Gently touching another’s arm Do the play-do like this Gently rubbing another’s arm Shane, can I have that? (toy) Gently touching another’s hand Calling out peers name (i.e., “Shane!”) Looking at another (i.e., eye contact) Can I have that? Holding another’s hand Pass it! (card) Taking Shane's hand to dance Hand it! (card) Physically guiding Shane to make the letter “B” with play-doh Go, Shane Shane entering area where other peers are located Shane, stop Shane playing with Lego’s next to a peer Don’t sit there Holding a toy for a peer to see it No Aggression (e.g., pinching, slapping, punching) Taking a toy from the hands of a peer Playing with Lego’s next to a peer

PAGE 226

214 Table C-2. Shannon’s Verbal and Gestural Initiations. Verbal Initiations Gestural Initiations Shannon, come play with me Shannon handing play plate to peer Shannon, help us clean up Shannon showing bent box to a peer Shannon, you got a (unintelligible)! Peer putting pretend blood pressure wrap on Shannon’s arm Hey, a lunch box! Peer handed Shannon materials Shannon laughing, while looking at peer Peer gave a toy to Shannon Hot. It’s very hot. Shannon reached over and took a toy from a peers pile Give me the green Peer moved Shannon’s hand If you want, I got some (unintelligible) Peer approached Shannon’s play materials Here you go, have some orange juice Approached peer Ice cream! Eye contact with peer Hey, look! Sharing tangibles with peer Pizza's here Showing toys to peers Bent (Shannon showing bent box to peer) Peer helped Shannon clean up Singing song to peer (song uninte lligible) Peer patted Shannon's arm Peer reaching over to Shannon’s toys Sh annon held a worm up to a peers face What’s that? Shannon tickled a peer Hello, Shannon Peer tickled Shannon Shannon, come play with me Peer held up a toy for Shannon to see Shannon, help us clean up Shannon reached to peer and held her hand out for a toy Shannon, you got a (unintelligible)! Peer put a toy on Shannon’s head Hey, a lunch box! Peer approached Shannon with a toy Shannon laughing, while looking at peer Peer handed Shannon a toy That’s, (peer name)! Shannon gui ded a peers hand to organize the peers play area Shannon, look! Peer gave Shannon materials Come on, Shannon Peer showed Shannon what she should be doing during the bingo activity Here, Shannon Peer points to screen (for Shannon to look at) Shannon, we can make some things with the blocks Peer tugs on Shannon’s arm to get her to sit down You guys play over there and me and Shannon will play over here Peer touches Shannon's hand to get her attention Shannon, will you help me with this? Peer takes Shannon’s hand, puts it on the computer mouse, a nd guides her through the activity Shannon, want to make a cage now? (for farm animals) Peer points to screen where Shannon is suppose to be looking

PAGE 227

215 Table C-2. Continued Verbal Initiations Gestural Initiations That’s my lion! Peer arranges Shannon’s materials Lets put some animals over here Peer points to Shannon’s materials Shannon, will you move back a little ? Peer touches Shannon’s dress We’re gonna go make it a cage Peer makes eye-contact with Shannon Shannon, come here. Help me Shannon picks up peers materials at clean up time No, we’re making an animal cage P eer cleans up Shannon’s materials at clean up time Shannon, can we just keep all of the animals spread out? Shannon showed a toy to a peer Shannon laughing, directed toward peer Peer touches Shannon’s materials Don’t do that! I think Shannon’s making that kind of necklace Shannon, take these off if you want Look! Shannon, your name’s on my list (of friends) A turtle! Ooh! Look what I got, Shannon Look at the wiggle worm Mom! Shannon? Shannon! Grass, right here Here, Shannon Shannon needs some more Red or yellow one? Come sit down Shannon, can we put these here? Wanna see something? See? Will you please move, Shannon? Hey, Shannon I got the lion over there Hey, look, a giraffe! Oh!, Mommy! Baby! Here Wanna pour these in there? Come on Shannon, quick, quick No, Shannon, not yet Shannon, put your hands in your lap!

PAGE 228

216 Table C-3. Colin’s Verbal and Gestural Initiations. Verbal Initiations Gestural Initiations Colin Colin bumped peer while dancing Yummy Colin looked at peer Ahhh Colin pinched peer Peer approached Colin Peer sat down next to Colin, and played with a puzzle next to him Peer grabbed Colin’s hand that contained a puzzle piece in it Peer offers Colin a puzzle piece Peer shows Colin where to put a puzzle piece Peer pokes Colin in the chest Peer walking nudges Colin forward Peer looked at Colin Colin sat down next to a peer and watched him play a computer game

PAGE 229

217 LIST OF REFERENCES American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author. Anderson, C. M., & Long, E. S. (2002). Use of a structured descriptive assessment methodology to identify variables affecting problem behavior. Journal of Applied Behavior Analysis 35 137-154. Apple, A. L., Billingsley, F., & Schwartz, I. S. (2005). Effects of video modeling alone and with self-management on complimentgiving behaviors of children with highfunctioning ASD. Journal of Positive Behavior Interventions 7(1), 33-46. Asmus, J. M. & Conroy, M. A. (2003). Evidenced based practices to address social and behavioral problems in young childre n with Autism Spectrum Disorders #H324D020023. United States Department of Education, Office of Special Education and Rehabilitative Services (OSERS). Asmus, J. M., Franzese, J. C., Conroy, M. A., & Dozier, C. L. (2003). Clarifying functional analysis outcomes for disrup tive behaviors by controlling consequence delivery for stereotypy. School Psychology Review 32(4), 624-630. Asmus, J. M., Ringdahl, J. E., Sellers, J. A ., Call, N. A., Andelman, M. S., & Wacker, D. P. (2004). Use of a short-term inpatient model to evaluate aberrant behavior: Outcome data summaries from 1996 to 2001. Journal of Applied Behavior Analysis, 37, 283-304. Axelrod, S. (1987). Functional a nd structural analyses of behavior: Approaches leading to reduced use of punishment procedures? Research in Developmental Disabilities, 8, 165-178. Baer, D. M., Wolf, M. M., & Risley, T. R. (1968). Some current dimensions of applied behavior analysis. Journal of Applied Behavior Analysis, 1 91-97. Barlow, D. H., & Hayes, S. C. (1979). Alte rnating treatments de sign: One strategy for comparing the effects of two tr eatments in a single subject. Journal of Applied Behavior Analysis, 12, 199-210.

PAGE 230

218 Brown, K. A., Wacker, D. P., Derby, K. M., Peck, S. M., Richman, D. M., Sasso, G. M., et al. (2000). Evaluating th e effects of functional co mmunication training in the presence and absence of establishing operations. Journal of Applied Behavior Analysis, 33, 53-71. Brown, W. H., Odom, S. L ., & Holcombe, A. (1996). Observational assessment of young children's social behavior with peers. Early Childhood Research Quarterly, 11, 19-40. Carr, E. G. (1977). The motivation of self -injurious behavior: A review of some hypotheses. Psychological Bulletin, 84(4), 800-816. Carr, E. G., & Durand, V. M. (1985). Re ducing behavior problems through functional communication training. Journal of Applied Behavior Analysis, 18, 111-126. Crosland, K. A., Zarcone, J. R., Lindauer, S. E., Valdovinos, M. G., Zarcone, T. J., Hellings, J. A., et al. (2003). Use of functional analysis methodology in the evaluation of medication effects. Journal of Autism and Developmental Disorders, 33(3), 271-279. Cunningham, E., & O’Neill, R. E. (2000). Compar ison of results of f unctional assessment and analysis methods with young children with autism. Education and Training in Mental Retardation and Developm ental Disabilities, 35(4), 406-414. DeLeon, I. G., & Iwata, B. A. (1996). Eval uation of a multiple-stimulus presentation format for assessing reinforcer preferences. Journal of Applied Behavior Analysis, 29, 519-533. English, C. L., & Anderson, C. M. (2004). Eff ects of familiar versus unfamiliar therapists on responding in the analog functional analysis. Research in Developmental Disabilities, 25, 39-55. Filipek, P. A., Accardo, P. J., Baranek, G. T., Cook, E. H. Jr., Dawson, G., Gordon, B., et al. (1999). The screening and diagnos is of autistic sp ectrum disorders. Journal of Autism and Developmental Disorders, 29(6), 439-484. Finkel, A. S., Derby, K. M., Weber, K. P., & McLaughlin, T. F. (2003). Use of choice to identify behavioral function following an inconclusive brief functional analysis. Journal of Positive Behavi oral Interventions, 5(2), 112-121. Fisher, W. W., Ninness, H. A. C., Piazza, C. C., & Owen-DeSchryver, J. S. (1996). On the reinforcing effects of the content of verbal attention. Journal of Applied Behavior Analysis, 29, 235-238.

PAGE 231

219 Garfinkle, A. N., & Schwartz, I. S. (2002). P eer imitation: Increasing social interactions in children with autism and other devel opmental disabilities in inclusive preschool classrooms. Topics in Early Childhood Special Education, 22(1), 26-38. Garrison-Harrell, L., Kamps, D., & Kravits, T. (1997). The effects of peer networks on social-communicative behavior s for students with autism. Focus on Autism and Other Developmental Disabilities, 12(4), 241-254. Gonzalez-Lopez, A., & Kamps, D. M. (1997). Social skills traini ng to increase social interactions between children with autism and their typical peers. Focus on Autism and Other Developmental Disabilities, 12(1), 2-14. Graff, R. B., Lineman, G. T., Libby, M. E., & Ahearn, W. H. (1999). Functional analysis and treatment of screaming in a you ng girl with severe disabilities. Behavioral Interventions, 14, 233-239. Hagopian, L. P., Contrucci Kuhn, S. A., L ong, E. S., & Rush, K. S. (2005). Schedule thinning following communicati on training: Using competing stimuli to enhance tolerance to decrements in reinforcer density. Journal of Applied Behavior Analysis, 38, 177-193. Hagopian, L. P., Fisher, W. W., Sullivan, M. T., Acquisto, J., & LeBlanc, L. A. (1998). Effectiveness of functional communicati on training with and without extinction and punishment: A summary of 21 inpatient cases. Journal of Applied Behavior Analysis, 31, 211-235. Hagopian, L. P., Wilson, D. M., & Wilder, D. A. (2001). Assessment and treatment of problem behavior maintained by escape from attention and access to tangible items. Journal of Applied Behavior Analysis, 34, 229-232. Hanley, G. P., Iwata, B. A., & McCord, B. E. (2003). Functional analysis of problem behavior: A review. Journal of Applied Behavior Analysis, 36, 147-185. Hanley, G. P., Piazza, C. C., Fisher, W. W., & Maglieri, K. A. (2005). On the effectiveness of and preference for p unishment and extinction components of function-based interventions. Journal of Applied Behavior Analysis, 38, 51-65. Harding, J. W., Wacker, D. P., Berg, W. K., Barretto, A., & Rankin, B. (2002). Assessment and treatment of severe be havior problems using choice-making procedures Education and Treatment of Children 25(1), 26-46. Iwata, B. A., Dorsey, M. F., Slifer, K. J., Bauman, K. E., & Richman, G. S. (1994). Toward a functional analysis of self-injury. Journal of Applied Behavior Analysis 27, 197-209. (Reprinted from Analysis and Intervention in Developmental Disabilities, 2, 3-20, 1982)

PAGE 232

220 Iwata, B. A., Pace, G. M., Dorsey, M. F., Za rcone, J. R., Vollmer, T. R., Smith, R. G., et al. (1994). The functions of self-inj urious behavior: An experimentalepidemiological analysis. Journal of Applied Behavior Analysis, 27, 215-240. Kamps, D. M., Potucek, J., Lopez, A. G., Kravits, T., & Kemmerer, K. (1997). The use of peer networks across multiple settings to improve social interaction for students with autism. Journal of Behavioral Education, 7(3), 335-357. Kazdin, A. E. (1982). Single-case research designs New York, NY: Oxford University Press. LaBelle, C. A., & Charlop-Christy, M. H. ( 2002). Individualizing func tional analysis to assess multiple and changing functions of severe behavior problems in children with autism. Journal of Positive Behavi or Interventions, 4(4), 231-241. Laushey, K. M., & Heflin, L. J. (2000). Enhanc ing social skills of kindergarten children with autism through the training of multiple peers as tutors. Journal of Autism and Developmental Disorders, 30(3), 183-193. Levin, L., & Carr, E. G. (2001). Food selectiv ity and problem behavi or in children with developmental disabilities: Analysis and intervention. Behavior Modification, 25(3), 443-470. Luiselli, J. K., Ricciardi, J. N., Schmidt, S ., & Tarr, M. (2004). Brief functional analysis and intervention evaluation for treatment of saliva-play [Electronic version]. Child & Family Behavior Therapy, 26(3), 53-61. Mace, A. B., Shapiro, E. S., & Mace, F. C. (1998). Effects of warning stimuli for reinforcer withdrawal and task onset on self-injury. Journal of Applied Behavior Analysis, 31, 679-682. Marcus, B. A., Vollmer, T. R., Swanson, V., Roane, H. R., & Ringdahl, J. E. (2001). An experimental analysis of aggression. Behavior Modification 25(2), 189-213. McComas, J., Hoch, H., Paone, D., & El -Roy, D. (2000). Escape behavior during academic tasks: A preliminary analysis of idiosyncratic establishing operations. Journal of Applied Behavior Analysis, 33, 479-493. McGee, G. G., Almeida, M. C., Sulzer-Azaroff, B., & Feldman, R. S. (1992). Promoting reciprocal interactions vi a peer incidental teaching. Journal of Applied Behavior Analysis, 25, 117-126. McGee, G. G., Feldman, R. S., & Morrier, M. J. (1997). Benchmarks of social treatment for children with autism. Journal of Autism and Developmental Disorders, 27(4), 353-364.

PAGE 233

221 Miltenberger, R. (1997). Behavior modification: Principles and procedures Pacific Grove, CA: Brooks/Cole. Mueller, M. M., Wilczynski, S. M., Moore, J. W., Fusilier, I., & Trahant, D. (2001). Antecedent manipulations in a tangible condi tion: Effects of stimulus preference on aggression. Journal of Applied Behavior Analysis, 34, 237-240. Najdowski, A. C., Wallace, M. D., Doney, J. K., & Ghezzi, P. M. (2003). Parental assessment and treatment of food selectivity in natural setting. Journal of Applied Behavior Analysis, 36, 383-386. National Research Council. (2001). Educating children with autism (Committee on Educational Interventions for Children w ith Autism. Division of Behavioral and Social Sciences and Education). Wa shington, DC: National Academy Press. Peck, J., Sasso, G. M., & Jolivette, K. (1997) Use of the structural analysis hypothesis testing model to improve social interactions via peer-mediated intervention. Focus on Autism and Other Developmental Disabilities, 12(4), 219-230. Piazza, C. C., Fisher, W. W., Brown, K. A., Shor e, B. A., Patel, M. R., Katz, R. M., et al. (2003). Functional analysis of in appropriate mealtime behaviors. Journal of Applied Behavior Analysis, 36, 187-204. Piazza, C. C., Fisher, W. W., Hanley, G. P., Le Blanc, L. A., Worsdell, A. S., Lindauer, S. E., et al. (1998). Treatment of pica thr ough multiple analyses of its reinforcing functions. Journal of Applied Behavior Analysis, 31, 165-189. Pierce, K., & Schreibman, L. (1997a). Multiple peer use of pivotal response training to increase social behaviors of classmates with autism: Results from trained and untrained peers. Journal of Applied Behavior Analysis 30, 157-160. Pierce, K., & Schreibman, L. (1997b). Using peer trainers to promote social behavior in autism: Are they effective at enhancing multiple social modalities? Focus on Autism and Other Developmental Disabilities, 12(4), 207-218. Rapp, J. T., Dozier, C. L., Carr, J. E., Patel, M. R., & Enloe, K. A. (2000). Functional analysis of hair manipulation: A replication and extension. Behavioral Interventions, 15, 121-133. Richman, D. M., Wacker, D. P., Asmus, J. M., Casey, S. D., & Andelman, M. (1999). Further analysis of problem behavi or in response class hierarchies. Journal of Applied Behavior Analysis, 32, 269-283. Richman, D. M., Wacker, D. P., & Winbor n, L. (2001). Response efficiency during functional communication trai ning: Effects of effort on response allocation. Journal of Applied Behavior Analysis, 34, 73-76.

PAGE 234

222 Roane, H. S., Vollmer, T. R., Ringdahl, J. E., & Marcus, B. A. (1998). Evaluation of a brief stimulus preference assessment. Journal of Applied Behavior Analysis, 31, 605-620. Sainato, D. M., Goldstein, H., & Strain, P. S. (1992). Effects of self-evaluation on preschool children’s use of so cial interaction strategies with their classmates with autism. Journal of Applied Behavior Analysis, 25, 127-141. Sasso, G. M., Reimers, T. M., Cooper, L. J ., Wacker, D., Berg, W., Steege, M., et al. (1992). Use of descriptive and experimental analyses to identify the functional properties of aberrant beha vior in school settings. Journal of Applied Behavior Analysis, 25, 809-821. Shabani, D. B., Katz, R. C., Wilder, D. A., Be uchamp, K., Taylor, C. R., & Fischer, K. J. (2002). Increasing social initiations in ch ildren with autism: Effects of a tactile prompt. Journal of Applied Behavior Analysis, 35, 79-83. Sidener, T. M., Carr, J. E., & Firth, A. M. (2005). Superimposition and withholding of edible consequences as treatment for automatically reinforced stereotypy. Journal of Applied Behavior Analysis, 38, 121-124. Simpson, R. L., & Zionts, P. (2000). Autism: Information and re sources for professionals and parents (2nd ed.). Austin, TX: Pro-ed. Spence, S. H. (2003). Social skills trai ning with children and young people: Theory, evidence and practice. Child and Adolescent Me ntal Health, 8(2), 84-96. Steege, M. W., Wacker, D. P., Berg, W. K., Cigrand, K. K., & Cooper, L. J. (1989). The use of behavioral assessment to prescrib e and evaluate treatments for severely handicapped children. Journal of Applied Behavior Analysis, 22, 23-33. Strain, P. S., & Danko, C. D. (1995). Activ ity engagement and social interaction development in young children with autism: An examination of “free” intervention effects. Journal of Emotional and Behavioral Disorders, 3(2), 108123. Retrieved April 16, 2005, from http://weblinks3epnet.com Strain, P. S., & Kohler, F. W. (1994). Teachi ng preschoolers with autism to self-monitor their social interactions: An analysis of results in home and school settings. Journal of Emotional and Be havioral Disorders, 2(2), 78-88. Retrieved April 16, 2005, from http://weblinks3.epnet.com Tang, J-C., Kennedy, C. H., Koppekin, A., & Caru so, M. (2002). Functional analysis of stereotypical ear covering in a child with autism. Journal of Applied Behavior Analysis, 35, 95-98.

PAGE 235

223 Tapp, J. (2002). The MOOSES Lodge: Introduction Retrieved June 19, 2002, from Vanderbilt University Web site: http://kc.vanderbilt.edu/~jont/mooses.html Tapp, J. (2003). Minimoose [Compu ter software]. Retrieved from http://kc.vanderbilt.edu/tapp/minimooseppc.zip Thiemann, K. S., & Goldstein, H. (2004). Effect s of peer training a nd written text cueing on social communication of school-age ch ildren with pervasive developmental disorder. Journal of Speech, Language, and Hearing Research, 47, 126-144. Thompson, R. H., Fisher, W. W., Piazza, C. C., & Kuhn, D. E. (1998). The evaluation and treatment of aggression maintained by attention and automatic reinforcement. Journal of Applied Behavior Analysis, 31, 103-116. U.S. Department of Education, Office of Special Education Programs (1994-2001). Annual report to Congress on the impl ementation of the Individuals with Disabilities Education Act (16th through 23rd). Washington, DC: Author. Wacker, D. P., Berg, W. K., Harding, J. W ., Derby, K. M., Asmus, J. M., & Healy, A. (1998). Evaluation and long-term treatmen t of aberrant behavior displayed by young children with disabilities. Journal of Developm ental and Behavioral Pediatrics, 19(4), 260-266. Wacker, D. P., Steege, M. W., Northrup, J., Sasso, G., Berg, W., Reimers, T., et al. (1990). A component analysis of functi onal communication tr aining across three topographies of severe behavior problems. Journal of Applied Behavior Analysis, 23, 417-429. Wert, B. Y., & Neisworth, J. T. (2003). Eff ects of video self-modeling on spontaneous requesting in children with autism. Journal of Positive Behavior Interventions 5(1), 30-34. Zanolli, K., Daggett, J., & Adams, T. (1996). Te aching preschool age autistic children to make spontaneous initiations to peers using priming. Journal of Autism and Developmental Disorders, 26(4), 407-422.

PAGE 236

224 BIOGRAPHICAL SKETCH Jennifer Anne Sellers completed her underg raduate degree at the University of Iowa in Iowa City, Iowa. She graduated in 1997 with a Bachelor of Arts degree in psychology. Jennifer completed her master’s degree at the University of Northern Iowa in Cedar Falls, Iowa. She graduate d in 2000 with a Master of Arts in Education. During her time at the University of Florida, Jennifer worked as a research assistant on a grant designed to facilitate the in clusion of children with autis m in the general education classroom and on a grant designed to assess and treat peer-related withdrawn and peerrelated negative social behaviors for young ch ildren with autism spectrum disorders in their natural settings. She completed her pr e-doctoral internship requirements on the Behavior Center at the Marc us Institute in Atlanta, Ge orgia. Upon completion of her Doctor of Philosophy degree in school psyc hology, Jennifer plans to pursue employment that will allow her to utilize the principles of applied behavior analysis while making a positive difference in schools, families, and children.


Permanent Link: http://ufdc.ufl.edu/UFE0013440/00001

Material Information

Title: Extending Functional Analysis Methodology to Identify the Function of Peer-Related Withdrawn and Peer-Related Negative Social Behaviors for Young Children with Autism Spectrum Disorders
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0013440:00001

Permanent Link: http://ufdc.ufl.edu/UFE0013440/00001

Material Information

Title: Extending Functional Analysis Methodology to Identify the Function of Peer-Related Withdrawn and Peer-Related Negative Social Behaviors for Young Children with Autism Spectrum Disorders
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0013440:00001


This item has the following downloads:


Full Text












EXTENDING FUNCTIONAL ANALYSIS METHODOLOGY TO IDENTIFY THE
FUNCTION OF PEER-RELATED WITHDRAWN AND PEER-RELATED
NEGATIVE SOCIAL BEHAVIORS FOR YOUNG CHILDREN WITH AUTISM
SPECTRUM DISORDERS















By

JENNIFER ANNE SELLERS


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2006

































Copyright 2006

by

JENNIFER ANNE SELLERS















ACKNOWLEDGMENTS

I would like to thank my dissertation chairs, Drs. Tina Smith-Bonahue and

Maureen Conroy, for their dedication and commitment in assisting me to complete this

project. I would like to thank my dissertation committee, Drs. Jennifer Asmus, Terry

Scott, and Gregory Valcante, for their expertise and guidance on this project. I would like

to thank the schools, teachers, students, and peers who participated in this study. I would

also like to thank Elizabeth McKenney and the staff of Project GATORSS, for their

assistance in this study; without them this would not be possible. Finally, I would like to

thank my parents, Arthur and Catherine Sellers, my grandmother, Helen Ellis, and

friends, Lori and Johanna. Their love, prayers, and support helped make this possible.
















TABLE OF CONTENTS



A C K N O W L E D G M E N T S ......... .................................................................................... iii

L IST O F TA B L E S ......... ............................. .......... ... .............. viii

LIST OF FIGURES ............................... .... ...... ... ................. .x

ABSTRACT .............. .......................................... xi

CHAPTER

1 L IT E R A TU R E R E V IE W ......... ................. .............................................................1

Introduction of Autism Spectrum Disorders (ASD)................................ ..................
D iagnostic Criteria for A SD ........................................................ ............... 2
P revalence of A SD .............. .............................. .... ......... ... ...... ..... .. .... ..4.
Identification and Characteristics of Children with ASD...................................4
A applied B behavior A analysis (A B A ) ............................................ .....................6
Sum m ary of A SD .................... .. ..... ... ................................... ............. ... .. ...8
Assessment and Treatment of Problem Behavior in Young Children with ASD......... 8
Introduction of Assessment Methodology..........................................................8
Functional A analysis ..... .... .................................................... ................ .9
Structural A analysis ................................................ ................. ............. 13
Summary of Assessment and Treatment of Problem Behaviors......................17
R review of the Literature .......... .............. ....................... .......... ............... 18
Review of Functional Analysis Literature................................................18
Review of Functional Analysis Studies.................................... ............... 19
Participants included in review of literature ................... .............. 20
Therapists in functional analysis ............ ............................................. 21
Settings of functional analysis ................................................................ 21
D dependent m measures ...................................................................... 22
Experim ental conditions......................................... .......................... 22
Experim ental condition m modifications ............................... ............... 23
Treatment of problem behavior............................... .............. 25
Design of functional analysis .............. .............................................25
Findings of functional analysis ................... .............................................27
Summary of functional analysis review of literature .............................. 28
Review of Social Skills Literature..................... .... ......................... 29
R eview of Social Skills Studies...................................... ......................... 30









Participants included in review of literature .............................................31
T herapists in studies ......................................... ............... ......... ...... .... 31
Settings of studies ........................................... ......... ................33
D dependent m easures......................................................... ............. 33
D esign of studies ...................... ...... .............. .. ....... .... ...........34
A ssessm ent of social skills....................................... ......... ............... 35
Treatm ent of social skills ...................... ................ ............... .... 36
Findings of social skills assessment and treatment.............. ...................37
Summary of social skills review of the literature........................................37
Conclusions and Future Directions for Research ................................................. 39

2 M E T H O D .......................................................... ................ 4 1

P artic ip an ts ......................................................................... 4 1
Settings and Therapists ...................................................... .. ............ 43
M a te ria ls ........................................................................... 4 5
R response D definitions .................. .................................... ................. 46
Preference A ssessm ent ......................................................... ............... 46
Functional A nalysis............... ..................................... ......... ... .............. 46
Participant peer-related withdrawn behavior ............................................47
Participant peer-related negative behavior...............................................47
Participant peer-related appropriate behaviors............... .. ............. 48
Peer procedural integrity data ............................ ...... ....... ............... 49
Observation System and Interobserver Agreement............................... ...............51
Observation System ................. ................................... .. .. .. ............ 51
Interobserver Agreement ........................... .................... 54
D esign and D ata A nalysis................................................. .............................. 56
D design of Procedures .......... .......... .............................. .. ........ .... 56
D ata A n a ly sis ................................................................................................. 5 7
Procedures ......... ...... ... ....... .................. 58
Phase One: D escriptive A ssessm ents ....................................... ............... 58
Phase Two: Preference Assessm ent ....................................... ............... 60
Phase Three: Functional A analysis ............................................ ............... 61
T raining to P eers.....................................................62
Free play condition .......... .... .... .................. ........................ 63
Ign ore con edition ................................................. ... ........ .................63
A attention condition ........... ............................................ ........ .............. 63
T angible condition .......... .................................... ...... ........ .... .......... 64
E scape condition .................................. ................. .... ..... .. 64
F unctional A naly sis ........... ...................................................... .. .... ...... 65
Free play condition .......... .... .... .................. ........................ 65
Ign ore con edition ................................................. ... ........ ....... ............66
A attention condition ........... ............................................ ........ .............. 66
T angible condition .......... .................................... ...... ........ .... .......... 67
E scape condition .................................. ................. .... ..... .. 68
Second Functional A analysis ........................................ .......................... 69



v









3 R E SU L T S ....................................................... 72

Overview of Research Aim s .......................................................... .............. 72
Phase One: Descriptive Assessment Results ......... ................... .... .............73
Shane Descriptive Assessment Results .................................... ............... 73
Shannon Descriptive Assessment Results....................................................74
Colin D escriptive A ssessm ent Results ..................................... ............... ..75
Phase Two: Preference Assessment Results............................................................77
Shane Preference Assessment Results ............... ......................................78
Shannon Preference Assessm ent Results ................................. ................ 79
First preference assessm ent results ................................... ............... ..79
Second preference assessment results .....................................................80
Colin Preference Assessm ent Results ...................................... ............... 81
Phase Three: Functional Analysis Results....................................... ............... 82
Shane Functional Analysis Results .......................... .......... ............... 84
Shane peer-related withdrawn and peer-related negative behaviors............84
Shane combined peer-related withdrawn and peer-related negative
behaviors ...................................................... .............................. 86
Shane peer-integrity ........................... ............. .................. ......89
Shane peer-related initiations .................................................. ............... 90
Shane peer-related responses ................................................. ............... 91
Summary of Shane's functional analysis results........................................92
Shannon Functional Analysis Results ......................... .. ....................93
Shannon peer-related withdrawn and peer-related negative behaviors........93
Shannon combined peer-related withdrawn and peer-related negative
b e h av io rs ............................................................................... 9 5
Shannon peer integrity ........................................ ........................... 96
Shannon peer-related initiations .............. ............................................ 97
Shannon peer-related responses ............. .............................................100
Summary of Shannon's functional analysis results.................................. 101
Colin Functional A analysis R results ....................... .............. ............... .... 102
Colin peer-related withdrawn and peer-related negative behaviors................02
Colin combined peer-related withdrawn and peer-related negative
behaviors ................................. ........................... ... ....... 103
Colin peer integrity ....................................... .. ............... .... .. ........ .... 105
Colin peer-related initiations .............. ............................................. 107
Colin peer-related responses ............................................................... 108
Summary of Colin's functional analysis results............... ... ...............109
Summary of Functional Analysis Results ............... .....................................110

4 D ISC U S SIO N ...................................................... 115

Sum m ary of Q questions ............................................................ ............ 115
Sum m ary of Findings ....................................................... .................. 116
Identified Functions Findings................................... ............................... ....... 117
Comparison of Functions Findings ............... .......................................123
Peer Integrity Findings ......................................................... .............. 125









Participant Appropriate Social Behavior Findings ................. ... .................126
Limitations of the Investigation.................... .. ......... ........ .. ............... 128
Future Research Directions .......... .... ...... ......... ........ .... ............... 131

APPENDIX

A REVIEW OF FUNCTIONAL ANALYSIS LITERATURE.........................134

B REVIEW OF SOCIAL SKILLS LITERATURE .................................................171

C LIST OF PEER-RELATED POSITIVE INITIATIONS OBTAINED DURING
DESCRIPTIVE ASSESSMENT (PHASE 1)................................ ............... 213

LIST OF REFEREN CES ........................................................... .. ............... 217

B IO G R A PH IC A L SK E TCH ........................................ ............................................224
















LIST OF TABLES


Table page

3-1 Social Demands Used in Escape Condition of Functional Analysis.....................76

3-2 Tangibles Identified During Descriptive Assessment............................................76

3-3 Tangibles Included in Preference Assessment.................................... ............... 77

3-4 Shane's Rank Order of Stimuli Obtained from Preference Assessment..................79

3-5 Shannon's Rank Order of Stimuli Obtained from First Preference Assessment .....80

3-6 Shannon's Rank Order of Stimuli Obtained from Second Preference Assessment.81

3-7 Colin's Rank Order of Stimuli Obtained from Preference Assessment ................. 82

3-8 Shane's Mean Rate and Ranges per minute of Participant Peer-related
Withdrawn, Negative, and Combined (Withdrawn and Negative) Behavior...........89

3-9 Shane's Mean Rate and Ranges per minute of Peer Correct Presentation of
S o cial D em an d s ................................................................... ............... 9 0

3-10 Shane's Average Percentage and Ranges of Peer Correct Delivery of
Consequences .................................... ................................ .........90

3-11 Shane's Mean Rate and Ranges per minute Peer Positive Initiations....................90

3-12 Shane's Mean Rate and Ranges per minute of Participant Peer-related Positive
an d N eg ativ e Initiation s ........................................ .............................................9 1

3-13 Shane's Average Percentage and Ranges of Participant Peer-related Positive,
N o, and N negative Responses ........................................................ ............. 92

3-14 Shannon's Mean Rate and Ranges per minute of Participant Peer-related
W withdraw n and N negative B ehavior.................................. ..................................... 95

3-15 Shannon's Mean Rate and Ranges per minute of Peer Correct Presentation of
S o cial D em an d s ................................................................... ............... 9 7

3-16 Shannon's Average Percentage and Ranges of Peer Correct Delivery of
Consequences .................................... ................................ .........99









3-17 Shannon's Mean Rate and Ranges per minute Peer Positive Initiations.................. 99

3-18 Shannon's Mean Rate and Ranges per minute of Participant Peer-related
Positive Initiations .................. ................................. ..... .. ........ .... 99

3-19 Shannon's Average Percentage and Ranges of Participant Peer-related Positive
and N o R responses ................................................................... .. .... .. 101

3-20 Colin's Mean Rate and Ranges per minute of Participant Peer-related
Withdrawn, Negative, and Combined (Withdrawn and Negative) Behavior......... 105

3-21 Colin's Mean Rate and Ranges per minute of Peer Correct Presentation of Social
D em hands ........... ......................................... .................... .......... 107

3-22 Colin's Average Percentage and Ranges of Peer Correct Delivery of
Consequences ................................ ................................ ........ 107

3-23 Colin's Mean Rate and Ranges per minute Peer Positive Initiations.....................107

3-24 Colin's Mean Rate and Ranges per minute of Participant Peer-related Positive
Initiation s ................................................ ............. .. ..... .............. .. 108

3-25 Colin's Average Percentage and Ranges of Participant Peer-related Positive and
No Responses ............................................................. ............ 109

3-26 Reinforcers Identified in Functional Analysis ......................................................113

C-l Shane's Verbal and Gestural Initiations.........................................................213

C-2 Shannon's Verbal and Gestural Initiations ............... ..................... ............... 214

C-3 Colin's Verbal and Gestural Initiations............... .................. ............... 216















LIST OF FIGURES


Figure page

3-1 Rate per minute peer-related withdrawn (top panel), negative (middle panel),
and combined social behavior (bottom panel) for Shane ........................................88

3-2 Rate per minute peer-related withdrawn (top panel) and negative social behavior
(middle panel) in first functional analysis, and peer-related withdrawn social
behavior (bottom panel) of second functional analysis for Shannon.....................98

3-3 Rate per minute peer-related withdrawn (top panel), negative (middle panel),
and combined social behavior (bottom panel) for Colin..............................106















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

EXTENDING FUNCTIONAL ANALYSIS METHODOLOGY TO IDENTIFY THE
FUNCTION OF PEER-RELATED WITHDRAWN AND PEER-RELATED
NEGATIVE SOCIAL BEHAVIORS FOR YOUNG CHILDREN WITH AUTISM
SPECTRUM DISORDERS

By

Jennifer Anne Sellers

May 2006

Chair: Tina Smith-Bonahue
Cochair: Maureen Conroy
Major Department: Educational Psychology

Social skill deficits (e.g., withdrawn behavior) and excesses (e.g., aggressive

behavior) are common behaviors displayed by children with autism spectrum disorders

(ASD), which can lead to decreased positive interactions with their peers. Several

treatments have successfully increased appropriate peer-related social behavior in young

children with ASD (e.g., peer-networks); however, these treatments have not always

produced generalizable results across participants, behaviors, or settings. Functional

analysis methodology is one strategy that has been successful in identifying the functions

maintaining a variety of problem behaviors linked to robust treatments, and may show

promise for identifying functions of peer-related problem social behavior.

The primary purpose of this study was to evaluate the effectiveness of the

functional analysis methodology to identify the function of peer-related withdrawn and

peer-related negative social behaviors in three young children with ASD. An additional









purpose of this investigation was to determine if peers could implement the functional

analysis methodology with adequate integrity.

Peers were included in the functional analysis conditions and provided participants

with relevant reinforcement contingent on peer-related withdrawn and/or peer-related

negative social behavior across all functional analysis conditions.

The findings indicated the functional analysis methodology was not successful in

identifying a clear function for peer-related withdrawn or peer-related negative social

behavior, singly, for any of the 3 participants. However, when rate per minute of peer-

related withdrawn and peer-related negative behaviors were combined and evaluated post

hoc, a negative reinforcement function for 1 of the 3 participants was identified.

With the exception of one participant's peer delivery of consequences in the escape

and tangible conditions and another participant's peer delivery of positive initiations in

the free play condition, peers were able to implement the functional analysis procedures

with adequate integrity.














CHAPTER 1
LITERATURE REVIEW

Introduction of Autism Spectrum Disorders (ASD)

Social skill deficits (e.g., withdrawn behavior) and behavioral excesses (e.g.,

aggressive behavior) can result in decreased positive social interactions between

individuals with autism spectrum disorders (ASD) and their peers. The incidence of ASD

is on the rise (National Research Council, 2001; U.S. Department of Education, Office of

Special Education Programs, 2001), and an increasing number of children with ASD are

being educated in general education settings in which social skills deficits and behavioral

excesses can impede academic and social success. This review addresses these issues by,

first, examining ASD. This is followed by a review of the literature related to functional

analysis of problem behavior with young children with ASD and social skills assessment

and treatment with young children with ASD.

Categories of ASD. Autism spectrum disorder is a term coined by Allen in 1988,

to include all types of disorders that are within the framework of pervasive

developmental disorders (PDD) (Filipek et al., 1999). Historically, disorders categorized

under ASD have included infantile autism, PDD-residual type, childhood schizophrenia,

and autistic psychoses. Currently, the Diagnostic and Statistical Manual of Mental

Disorders, Fourth Edition, Text Revision (DSM-IV-TR) (American Psychiatric

Association [APA], 2000) PDD categories include autistic disorder, Rett's disorder,

childhood disintegrative disorder, Asperger's disorder, and PDD, not otherwise specified

(NOS), including atypical autism. However, ASD is commonly used to refer to autistic









disorder, Asperger's disorder, and PDD-NOS, and will be referred to as such for the

purposes of this paper. In this chapter, an overview of ASD will be presented; next how

applied behavior analysis has addressed problem behavior in young children with ASD

will be discussed; an in depth literature review of the use of functional analysis

methodology to assess problem behaviors in young children with ASD is provided, and

finally an in depth literature review of peer-related social skills assessment and treatment

of peer-related social skills in young children with ASD is presented.

Diagnostic Criteria for ASD

Diagnostic criteria for autistic disorder, Asperger's disorder, and PDD-NOS include

impairments in social interaction and repetitive and stereotyped behavior. The level and

degree of social impairments differ across different variations of ASD. In autistic and

Asperger's disorder social impairments are described as deficits in nonverbal behaviors

(e.g., eye contact, facial expressions), developing and maintaining developmentally

appropriate peer relations, and spontaneous initiations (e.g., bringing items of interest to

show peers), responses (e.g., joining a play group when asked), or interactions (e.g.,

playing a game with peers) with peers (APA, 2000). However, the criteria for social

impairments for autistic disorder are identified as deficits in behaviors that display social

and emotional indifference, whereas with Asperger's disorder, social impairment is

defined as eccentric and one-sided (e.g., talking to peers almost exclusively about a

specific topic of interest only to the child with Asperger's). Persons with autistic and

Asperger's disorders also display restrictive, repetitive, and/or stereotyped behavior,

interests, and activities. These behaviors are defined as motor mannerisms in autistic

disorder (e.g., strictly following routines, stereotypic hand-flapping, rocking), and they









are defined as a topic or interest in gathering facts in Asperger's disorder (e.g., learning as

much as possible about frogs) (APA, 2000).

Other differences in diagnostic criteria for autistic and Asperger's disorder are age

of onset, impairments in communication, and deficits in mental abilities. For the

diagnosis of autistic disorder, the onset of symptoms must occur prior to age 3-years and

there must be a delay in communication (APA, 2000). Both verbal and nonverbal

communication abilities are impaired and verbal communication may not develop at all.

Communication attempts may be repetitive (e.g., repeating words of a video), language

comprehension may be delayed (e.g., difficulty understanding questions or directions),

and pragmatic aspects of language (e.g., irony, humor, social convention) are delayed or

absent. In addition to social language delays, persons with autism may not develop

imaginative or pretend play. There is not a cut off criterion for measures of mental

abilities, though most people with autistic disorder also have an associated diagnosis of

mental retardation, ranging from mild to profound (APA, 2000). Therefore, the diagnosis

of autistic disorder can be made with or without an associated diagnosis of mental

retardation. Similarly, PDD-NOS diagnostic criterion requires a delay in communication

skills; however, this diagnosis does not include criteria for mental ability or age of onset.

In contrast to both autistic disorder and PDD-NOS, the diagnostic criterian of Asperger's

disorder does not allow for either a delay in communication or mental retardation.

In summary, children with ASD demonstrate a number of diverse characteristics.

They may have impairments in social interactions, verbal and nonverbal communication,

and restrictive, repetitive, and/or stereotyped behavior, interests, and activities. Their

impairments in social interactions and communication often lead to failure in developing









age-appropriate peer relationships, which has many negative future implications for these

children (Filipek et al., 1999).

Prevalence of ASD

Early identification of children with ASD has become increasingly critical as

children who receive two or more years of intensive early education intervention are

more likely to have greater developmental gains in speech and cognition than children

who do not receive those services (Filipek et al., 1999). The overall prevalence of ASD is

estimated at 10 to 20 per 10,000 people. There are estimates indicating that in the U.S.

there are 60,000 to 115,000 children under 15-years-old diagnosed with ASD (Filipek et

al., 1999). More males than females are diagnosed with ASD, though the ratio varies with

the level of cognitive impairment. There is a male to female ratio of 2:1 for children with

ASD who also have mental retardation and 4:1 for children with ASD without mental

retardation (Filipek et al., 1999). Therefore, the early identification of the characteristics

of ASD is critical to children receiving early education intervention to address behavioral

excesses and deficits.

Identification and Characteristics of Children with ASD

Children with ASD have a variety of behavioral and social deficits. For example,

young children with ASD often do not "cuddle" with their caregivers, bring toys to show

caregivers or peers, and may have temper tantrums when daily routines are changed (e.g.,

going to bed without reading a specific book) (Simpson & Zionts, 2000). In addition,

they frequently engage in chronic problem behaviors that impact their learning and

integration into society.

Problem behavior. Children with ASD may exhibit behavior problems such as a

high level of motor activity, aggression (e.g., hitting, kicking), self-injurious behavior









(SIB) (e.g., head banging, hand biting), temper tantrums, stereotypy (e.g., hand-flapping,

spinning), and intense restricted patterns of interest/behavior (e.g., lining up objects,

playing with only one toy) (APA, 2000; Simpson & Zionts, 2000). Problem behaviors are

widely reported in the clinical and research literature (Fisher, Ninness, Piazza, & Owen-

DeSchryver, 1996; Richman, Wacker, & Winborn, 2001).

Social withdrawal and deficits in communication. Social withdrawal and a lack

of communication, both verbal and nonverbal, are also characteristics of children with

ASD (APA, 2000). Children with ASD may not make eye contact with adults or peers

and may withdraw from peers in order to engage in solitary play. These children may not

spontaneously initiate to peers (e.g., bring an item of interest to show a peer), may not

respond to their name or other peer attempts to gain their attention (e.g., tapping shoulder,

asking the child to play with them), and if the child does play with a peer, the play may

be stereotypic (e.g., lining up toys) rather than interactive (e.g., building a tower with

Lego's) or imaginative play (e.g., playing pretend school) (APA, 2000; Simpson &

Zionts, 2000). For example, McGee, Feldman, and Morrier (1997) compared the social

behavior of young children with ASD and young typically developing children, all

attending an integrated preschool. The results suggest specific peer-related social skill

deficits in young children with ASD in comparison to their typically developing peers,

including deficits in initiations, responses, and staying in proximity (i.e., 3 ft) of peers. In

addition, the results of this study suggest that these social skill deficits do not improve

with maturation over one year or with placement in an integrated preschool without

treatments focused on their social behavior.









In summary, children with ASD demonstrate deficits in the areas of problem

behaviors, social skills, and communication. Therefore, the early identification of ASD

and development of treatments that can address these children's deficits is critical.

Several types of treatments have been developed with the goal of increasing the

appropriate peer-related social behavior of young children with ASD (e.g., Laushey &

Heflin, 2000; Pierce & Schreibman, 1997b). However, to date treatments to increase the

social behavior of young children with ASD have primarily been based on a priori

decisions (Brown, Odom, & Holcombe, 1996; Spence, 2003) rather than stemming from

individual assessments. In the following section, one approach to treatment, applied

behavior analysis (ABA), that has been highly successful for use with children with ASD

will be reviewed.

Applied Behavior Analysis (ABA)

The field of ABA includes several strategies to individually assess problem

behaviors and to develop assessment-based treatments. Applied behavior analysis has

been demonstrated to be a successful treatment approach in assessing and treating

problem behavior (Asmus, Ringdahl, Sellers, Call, Andelman, & Wacker, 2004; Iwata et

al., 1994; Wacker, Berg, Harding, Derby, Asmus, & Healy, 1998). Baer, Wolf, and

Risley (1968), first, described ABA as an approach to treatment that focuses on socially

important behaviors and includes reliable measurement and procedures that can

differentiate variables that control behavior change, and can be replicated. Since that

time, considerable research has been conducted on the effectiveness of various ABA

strategies to decrease problem behavior (e.g., aggression, self-injury, destruction,

stereotypy) and increase alternative appropriate behaviors (e.g., communication, social

behaviors).









Experimental analyses of consequences (functional analysis) and antecedents

(structural analysis) are two common types of strategies used in the field of ABA to

individually assess problem behaviors and to develop assessment-based treatments. Both

functional and structural analyses lead to the identification of functional relationships that

exist between environmental events (e.g., levels of attention provided, availability of

preferred toys) and target behaviors (e.g., problem behavior, peer-related problem social

behavior) (Axelrod, 1987; Carr & Durand, 1985; Fisher et al., 1996; Peck, Sasso, &

Jolivette, 1997). For example, Richman et al. (2001) conducted a functional analysis on

the aggressive behavior of a 3-year-old child diagnosed with ASD. The findings of the

functional analysis indicated the participant engaged in aggressive behavior to gain

access to tangible items. Based on this information, a treatment was developed that taught

the participant an appropriate alternative behavior to gain access to tangibles (i.e.,

handing a card or signing please to request the tangible item). This assessment-based

treatment successfully reduced the aggressive behavior of the young child with ASD

while simultaneously increasing his appropriate communication.

Peck et al. (1997) conducted a hypothesis driven structural analysis on the peer-

related problem social behavior of five children with disabilities (two with ASD) between

the ages of 9 and 11-years. Based on the individual findings of the structural analysis,

two peer-mediated treatment protocols were developed and presented in an alternating

treatment design. One treatment included the antecedents identified in the structural

analysis that resulted in the highest percentage of peer interactions or the highest rate of

on-task behavior and lowest rate of problem behavior. In contrast, the second treatment

protocol included antecedents identified in the structural analysis that resulted in low









percentages of peer-related interactions or the lowest rate of on-task behavior and highest

rate of problem behavior. The structural analysis assessment in this investigation was

successful in helping to develop an effective treatment that resulted in a higher rate of on-

task behavior, lower rate of problem behavior, and higher percentage of peer-related

interactions in comparison to a treatment that was not developed based on the findings of

the hypothesis driven structural analysis.

Summary of ASD

In summary, children with ASD demonstrate deficits in social skills and excesses in

problem behavior. Research has demonstrated the potential negative effects social

behavior deficits can have on individuals with ASD from preschool through adulthood.

However, even with and without intensive treatments, gains in peer-related social

behaviors remains problematic for this population. Applied behavior analysis appears to

be a promising approach for developing individual treatments to promote peer-related

social behaviors in children with ASD. The remainder of this chapter will review the

literature in two areas. First, assessment and treatment of problem behaviors will be

discussed. Next, a review of the social skills treatment literature will be presented.

Assessment and Treatment of Problem Behavior in Young Children with ASD

Introduction of Assessment Methodology

One of the primary approaches for ameliorating problem behaviors in young

children with ASD is through experimental analysis. Experimental analysis methods are

based on Carr's (1977) theoretical model in which operant mechanisms are hypothesized

to maintain problem behavior. Carr hypothesized that different individuals engage in

similar topographies of problem behavior for different environmental reasons. Some

common topographies of problem behavior include aggression, destruction, SIB, and









disruption. Carr proposed three potential operant mechanisms underlying the motivation

for problem behavior: positive reinforcement, negative reinforcement, and sensory

reinforcement. Carr's proposal that problem behavior can have an operant basis has led to

substantial research and the subsequent development of assessment techniques to identify

the operant mechanisms underlying problem behaviors.

Axelrod (1987) discussed two types of experimental analyses: (1) manipulation of

antecedent variables, and (2) manipulation of consequence variables. Axelrod defined a

functional analysis as an assessment of factors that reinforce the occurrence of a behavior

(e.g., escape from demands maintains SIB). Axelrod defined a structural analysis as an

assessment of antecedent variables that occasion the behavior (e.g., demands evoke SIB).

Both functional and structural analyses may lead to the identification of functional

relationships that exist between these environmental events and target behaviors. The

influence of consequence events on problematic behavior has been studied extensively

via functional analyses, whereas the influence of antecedent events has been studied via

structural analyses.

Functional Analysis

In ABA, the term "function" refers to the purpose the behavior serves (i.e., the

reinforcer maintaining the behavior) for the individual (Carr, 1977). Carr identified

several reinforcers that maintain problem behavior: positive (e.g., obtaining attention

following problem behavior), negative (e.g., escaping demand following problem

behavior), and self-stimulatory (or automatic) reinforcement (i.e., problem behavior not

maintained by social-mediation).

In a seminal study based on Carr's work, Iwata, Dorsey, Slifer, Bauman, and

Richman (1982/1994) developed the functional analysis methodology to identify the









function of SIB by experimentally evaluating the consequences maintaining that

behavior.

Iwata et al. (1982/1994) developed the first operant methodology for identifying

functional relationships between specific environmental variables and SIB. Nine adults

with developmental disabilities who engaged in high levels of SIB participated in the

study. Self-injurious behavior was measured across analogue conditions in which

consequence events were manipulated. The analogue condition that resulted in the

highest percent of SIB identified the class of consequences (positive or negative

reinforcement) that served to maintain problem behavior. Results indicated that specific

environmental variables were identified that maintained the occurrence of SIB for 6 of 9

individuals.

Iwata et al. (1982/1994) individually assessed participant functions of SIB by

experimentally evaluating four distinct environmental conditions: social disapproval,

academic demand, free play, and alone. A free play condition was also evaluated as a

control condition.

In the social disapproval condition, the participant had access to tangibles and was

told to play with the toys because the experimenter had to do some work. The

participant's engagement in SIB resulted in statements of concern and disapproval along

with non-punitive physical contact. An increased percent of SIB in this condition

indicated the participant's SIB was maintained by positive reinforcement in the form of

attention.

In the academic demand condition, participants sat at a table next to the

experimenter and were presented with difficult tasks (i.e., tasks that were not completed









spontaneously) using a three prompt sequence. The demand was first presented verbally.

If the demand was not completed after 5 seconds, the experimenter repeated the verbal

demand and modeled how to complete the demand. If the demand was not completed

after another 5 seconds, the experimenter repeated the verbal demand and physically

guided the participant to complete the demand using the least amount of physical contact

necessary. Praise was provided for completion of the demand, regardless of modeling or

physical guidance. The next demand was then presented. If the participant engaged in

SIB at any point during the demand, the experimenter ended the demand and turned away

from the participant for 30 seconds before presenting the next demand. Repeated SIB

resulted in an additional 30 second delay to presentation of the next demand. An

increased percent of SIB in this condition indicated that the participant's SIB was

maintained by negative reinforcement.

In the free play condition the individual had access to tangibles and no demands

were presented. Attention (social praise, physical contact) was provided noncontingently

for appropriate behavior (absence of SIB) at least every 30 seconds. Attention for SIB

was not provided. If SIB were socially maintained, the lowest percent of SIB would be

observed in the control condition.

In the alone condition, the participant was in a room alone, and without any

materials. An increased percent of SIB in this condition indicated the participant's SIB

was maintained by self-stimulation.

The length of assessment averaged 8 days across participants. Eight sessions were

conducted per day and conditions lasted 15 minutes each. The order of conditions was

predetermined by random drawings. Data were collected using continuous 10 second









interval recordings. A multielement design was used to analyze the data. Data collection

continued until the data were stable, or the data were unstable in all conditions for 5 days,

or until sessions had continued for 12 days.

Results indicated that 4 of the 9 participants displayed the highest percent of SIB in

the alone condition. Two participants displayed the highest percent of SIB in the

academic demand condition, one displayed the highest percent of SIB during the social

disapproval condition, and two displayed undifferentiated patterns of SIB. Six of the 9

participants consistently displayed a higher percent of SIB in one specific condition,

indicating the participants' function of SIB varied both between and within individuals,

similar to Carr's (1977) proposal. That is, the topography of the behavior (i.e., SIB) did

not match the function of the behavior. This led to the conclusion that conducting

individual assessments to determine the function of SIB would lead to more effective

treatments that are based on the variables that are maintaining or reinforcing problem

behavior rather than the topography of the behavior.

The majority of research studies published in the literature have used the functional

analysis methodology developed by Iwata et al. (1982/1994) to determine the maintaining

contingencies of problem behavior; that is, they have focused on the manipulation of

consequence events during assessment. The results of Iwata et al. have been consistently

replicated by different clinical research teams and across different topographies of

problem behavior (McComas, Hoch, Paone, & El-Roy, 2000; Steege, Wacker, Berg,

Cigrand, & Cooper, 1989). In every case, the functional analysis methodology was used

to identify the effects of environmental variables on the occurrence of problem behavior.

The functional analysis methodology has been demonstrated to be successful in









identifying the function for a variety of topographies of behavior (e.g., SB, aggression,

destruction, stereotypy) and across a range of settings (e.g., home, school, clinic)

(Hanley, Iwata, & McCord, 2003). Several epidemiological studies have also

demonstrated the success of functional analysis in identifying reinforcers for a variety of

problem behaviors of adults and children with and without developmental disabilities,

and the successful reduction of problem behavior utilizing assessment-based treatments

(Asmus et al., 2004; Hagopian, Fisher, Sullivan, Acquisto, & LeBlanc, 1998; Iwata et al.,

1994; Wacker et al., 1998). Although extensive replication has occurred, one behavior

that has not been investigated using functional analysis methodology is peer-related

social behavior.

Structural Analysis

The results of a functional analysis are restricted to only one category of

environmental variables: consequences. Other environmental variables that occur prior to

or concurrently with behavior may also influence the occurrence of behavior due to their

previous association with reinforcement. A variety of contextual factors have been

assessed to evaluate their effect on problem behavior (e.g., off-task behavior, aggression)

and peer-related social behavior (e.g., aggression toward peers, negative vocalizations).

To study these variables, a slightly different methodology is used, referred to as a

structural analysis. A structural analysis is an assessment procedure that identifies the

antecedent variables that set the occasion for problematic behavior; for example, demand

situations may result in an increased likelihood of SIB (Axelrod, 1987; Carr & Durand,

1985). Thus, the purpose of a structural analysis is to more precisely identify antecedent

events that reliably occasion the target behavior.









Carr and Durand (1985) first developed the structural analysis methodology, which

was conducted with four children with ASD (ages 7 to 14-years). This assessment

methodology differed from Iwata et al. (1982/1994) in that behaviors were shown to

correlate with changes in antecedent conditions (high versus low demands and high

versus low social attention) rather than with contingent access to consequences (escape

and social attention). Despite this difference, the results were equally positive in

identifying a specific operant basis for each participant's problem behavior.

Peck et al. (1997) extended the structural analysis methodology to include

assessment of antecedents to examine the antecedent variables that occasion appropriate

social behavior. Peck et al. conducted hypothesis driven structural analyses with 5

participants in elementary school (ages 9 to 11-years) with developmental disabilities

(two with ASD), mild to moderate mental retardation, and peer-related negative social

behavior. Hypothesis driven structural analyses were conducted with each participant to

identify environmental factors that increased the likelihood that appropriate peer-related

social behavior would occur.

In this investigation, the special education teachers and researchers developed

hypotheses regarding the antecedents likely to increase the likelihood of each

participant's appropriate and negative peer-related social behavior. There were a total of

seven antecedent events identified: noise level, peer group size, amount of interaction

required for the task, task structure, gender of peer, level of social interaction, and

transition warnings.

The participants were in two self-contained classrooms for the majority of the

school day. Four typically developing peers that were identified by their general









education teachers as having exemplary social skills and were the same age as the

participants participated in the study. Each peer participated in the analyses of 2

participants.

Individualized hypothesis driven structural analyses were conducted in the

participants' self-contained special education classrooms during naturally occurring

classroom activities. The hypothesized "most effective" and "least effective" antecedents

identified were presented in 10 minute conditions using an alternating treatment design

(e.g., alternately presenting high and low noise conditions). A same-aged typically

developing peer was trained by the researchers and participated in the conditions with

each participant. Data were collected using 6 second partial interval recording on the

participants' initiations (e.g., verbal or gestural behavior to peer to indicate

communication when within at least 1 meter of a peer and physically oriented to peer),

responses (verbal or gestural response to peer within 3 seconds of that peer's initiation),

and continuations (third behavior in initiation, response within 3 seconds, continuation

within 3 second sequence). The participants' responses and continuations were calculated

as proportions of opportunities (i.e., the number of participant responses or continuations

divided by the number of peer initiations multiplied by 100%). In addition, data were

collected on the participants' on-task and off-task (not oriented toward or engaged with

activity or peer) behavior, negative behavior (e.g., aggression, swearing, SIB, escape

[attempts to leave activity or area without permission]) and purposeful (rather than off-

task) noncompliance. Peer initiations, responses, and continuations were scored using the

same definitions as the participant. In addition, peer behaviors were scored as cooperative









(conversation or interaction that was not about playing the game) or instructional

(conversation or interaction with or about the assigned task).

The data were analyzed by combining the percentage of the participants' initiations

and proportion of responses and continuations into one social interaction graph and the

percentage of participants' on-task behavior on another graph. The "most effective"

condition was identified using visual analysis, first evaluating only social interactive data.

If there was not a trend in the social interactive data, the "most effective" condition was

identified based on the condition with the highest percent of on-task behavior and the

lowest percent of problem behavior.

Individual treatments were developed based on the assessment findings. The

hypothesized "most effective" (Treatment A) condition was the combination of

antecedents assessed in the structural analysis that resulted in highest level of on-task and

effective communication. For example if high and low structure, interactive and social

behavior were individually assessed and high structure, low interaction, and high

socialization all resulted in the highest level of on-task and effective communication, all

of those antecedents (i.e., high structure, low interaction, and high socialization) were

combined into one condition, i.e., Treatment A. Conversely, Treatment B was the

combination of antecedents when the participant displayed the most off-task behavior and

problem behavior. Therefore, using the example above, Treatment B would be a

combination of low structure, high interaction, and low socialization. Antecedents that

did not result in substantial or any differences in behavior were not included in the

treatment evaluation (i.e., not in Treatment A nor Treatment B). The treatment conditions

were conducted for 10 minutes during naturally occurring activities and presented in an









alternating treatment design using a reversal sequence. Treatment A was presented first

and continued until a trend in on and off-task behavior and problem behavior was

established. Treatment B was then conducted until a trend was established. Finally,

Treatment A was repeated until a stable trend was again established.

The results indicated that all 5 participants demonstrated differential appropriate

on-task behavior in at least one antecedent condition. Three of the 5 participants

demonstrated inconclusive patterns of initiations, responses, and continuations across the

antecedents and of the two who displayed differential behavior, only one displayed

differential behavior for all three social interactive behaviors (i.e., initiations, responses,

and continuations). Treatments that were based on the combination of antecedent

variables that resulted in the most on-task behavior, least problem behavior, and for two

of the participants, also the most social initiation, response, and/or continuation behavior,

resulted in higher levels of on-task and lower levels of problem behavior than the

treatment based on the antecedent conditions with the most off-task and problem

behavior. All Treatment A interventions successfully increased on-task behavior, social

interactive behavior, and decreased problem behavior (except for one participant who did

not display problem behavior during treatment) in comparison to Treatment B. These

findings suggest hypothesis driven structural analyses led to the development of

successful peer-mediated treatments of peer-related social problems.

Summary of Assessment and Treatment of Problem Behaviors

Functional analysis is a strategy to use to assess consequences (i.e., reinforcers)

maintaining target behavior, and structural analysis is a strategy to use to assess

antecedent events preceding or co-occurring target behavior (Axelrod, 1987). Both

experimental analyses of behavior have proven successful in identifying the function of a









variety of problem behaviors (e.g., aggression, destruction) and development of effective

treatments (Carr & Durand, 1985; Iwata et al., 1994). Treatments that are based on the

function of problem behavior are more effective than treatments that are not function

based (Axelrod, 1987). To date, a functional analysis of peer-related problem social

behavior has not been conducted. Peck et al. (1997) conducted the first hypothesis driven

structural analysis of peer-related problem social behavior. Given the success of

functional analysis methodology to identify the functions of problem behavior and

develop effective treatments, an extension of the functional analysis to include the

assessment of peer-related problem social behavior appears to be a logical next step.

Review of the Literature

The purpose of this section is to provide a review of the literature on the use of

experimental analyses to assess and treat problem behaviors in young children with ASD

and to review the literature on social skills assessment and treatment.

Two bodies of literature will be reviewed in order to demonstrate their contribution

to the current study: (1) literature demonstrating the effectiveness of functional analysis

in identifying the function of problem behavior for young children with ASD, and (2)

literature addressing assessment and treatment of social skill deficits in young children

with ASD.

Review of Functional Analysis Literature

Method to select reviewed studies. Appropriate articles for this literature review

were identified by searching the following computerized databases: Academic Search

Premier, Journal of Applied Behavior Analysis, and PsyclNFO. Keywords used to search

the databases included, but are not limited to functional analysis, autism, children, young,

early, intervention, and behavior.









The functional analysis body of literature met the following criteria for inclusion in

the review: (a) peer-reviewed journal, (b) functional analysis conducted (including brief

functional analysis) using a single-subject design including a variety of designs

(including reversal, pair-wise, alternating treatment designs), (c) at least two analogue

conditions conducted, (d) procedures for each analogue condition described, (e)

individual results for participants presented in a line graph format, (f) at least one

participant with ASD 8-years of age or younger and/or at least one participant with a

diagnosis of developmental delays and characteristics of ASD (e.g., social withdrawal,

authors indicated the participant had autistic like behaviors and/or indicated the

participant was suspected of having autism).

Review of Functional Analysis Studies

This section of the review demonstrates the utility of the functional analysis

methodology in identifying the function of problem behavior for young children with

ASD. A functional analysis is an experimental procedure designed to evaluate the

environmental variables that are maintaining a target behavior(s). A total of 32 studies

met the criteria mentioned above. The functional analysis methodology according to

Iwata et al. (1982/1994) is relatively new (i.e., since 1982); therefore no studies prior to

1982 were included. Publication dates of the studies included in the review range from

1982 to 2005. One noteworthy finding to mention is there has been an increase in studies

that include a functional analysis with young children with ASD; only 9 studies were

conducted between 1982 and 1999, whereas 20 studies were conducted between 2000

and 2004.









Participants included in review of literature

A total of 47 participants were included across the 32 studies identified. These

participants are described in Appendix A (Review of Functional Analysis Literature). In

addition, the number of participants who met criteria in each study, age, gender, and

disability are included.

Age. All participants included in these studies were between the ages of 3 and 8-

years-old. Forty-five percent (21) of the participants were between 5 and 6-years-old,

34% (16) were between 7 and 8-years-old, and 17% (8) were between 3 and 4-years-old.

These numbers are not mutually exclusive, as two studies with more than one participant

included one participant between the ages of 3 and 8-years-old, however, their exact age

was not indicated.

Number. The majority (63%) of the 32 studies had only one participant that met

the criteria for inclusion. Twenty-five percent (8) of the studies had two participants, 6%

(2) had three participants, and one (3%) included four participants.

Gender. Of the 47 participants included in this review, 85% (40) were male and

15% (7) were female.

Disability. All 47 participants had either a diagnosis of PDD-NOS, autism, or if

under the age of 6 years, characteristics of ASD. Under the category of ASD, 83% (39)

had a diagnosis of autism, 13% (6) PDD-NOS, 2% (1) ASD, and 2% (1) was

characterized as having autistic-like behavior. Other diagnoses included mental

retardation (level unspecified) (4%), mild mental retardation (6%), moderate to severe

mental retardation (45%), and profound mental retardation (6%). For 36% (17) of

participants no level of mental retardation was indicated. In addition, 6% (3) of the

participants had a diagnosis of developmental delays, 6% (3) a diagnosis of attention









deficit hyperactivity disorder (ADHD), 4% (2) oppositional defiant disorder (ODD), 4%

(2) seizure disorder, and 2% (1) with a diagnosis of Fragile X syndrome.

Therapists in functional analysis

Of the 32 studies reviewed, 63% (20) of the therapists were a staff member or the

experimenter, 19% (6) were participant caregivers, and 13% (4) were participant

teachers. Twenty-two percent (7) of the studies did not explicitly report who served as the

therapist. The following are examples of how investigators have employed a variety of

therapists when conducting functional analyses. Sasso et al. (1992) first conducted a

functional analysis with experimenters serving as therapists and then, after providing

training to the teacher, the teacher conducted a functional analysis in the classroom,

serving as the therapist. In another functional analysis study of food refusal, a parent

served as the therapist throughout all functional analysis conditions (Najdowski, Wallace,

Doney, & Ghezzi, 2003). English and Anderson (2004) manipulated who served as the

therapists (experimenter or parent) while conducting a functional analysis. When the

parent was the therapist, the functional analysis identified an attention and tangible

function, however, when the experimenter served as the therapist, the functional analysis

only identified an attention function. Therefore, the results of the functional analysis were

similar across therapists in identifying an attention function, though differed across

therapists in identifying a tangible function. The findings of this study suggest the

identity of the therapist influenced the findings of the functional analysis.

Settings of functional analysis

The settings in which the functional analysis procedures were implemented varied

across the 32 studies, though the majority (56%) were conducted in a session room while

the participant was inpatient or attended an outpatient clinic. Twenty-two percent (7)









were conducted in the participant's classroom, 16% (5) in an unused room at the

participant's school, 16% (5) in the home, and 3% (1) was conducted in the classroom on

an inpatient unit in a hospital setting. In studies with more than one participant, there

were some instances of functional analyses being conducted in different settings (i.e., one

participant in the home and the other at school). When this occurred, both settings were

recorded; therefore, the categories used to report settings will exceed 100% when totaled.

Dependent measures

There were several behaviors measured in the functional analyses conducted across

the 47 participants. The categories of behaviors reported are not mutually exclusive and

therefore, will not add up to 100% when totaled. In addition, some researchers measured

one dependent variable (e.g., destruction) but included several behaviors (e.g.,

noncompliance, screaming, tantrums) within the operational definition of destruction. As

a result, the following should not be considered a comprehensive list of all behaviors the

participants in these studies displayed. Thirty-six percent (17) participants engaged in

aggression, 23% (11) in disruption, 19% (9) SIB, 11% (5) inappropriate mealtime

behavior, 9% (4) destruction, 9% (4) stereotypy, and 4% (2) of the participants engaged

in screaming behavior. One participant engaged in each of the following: tantrums, high

frequency vocalizations, pica, saliva play, food refusal, spitting, elopement, hair

manipulation, inappropriate language, and grabbing. Overall, 89% (42) of the

participants' dependent measures included one or more of the following behaviors:

aggression, disruption, destruction, SIB, or tantrums.

Experimental conditions

The experimental conditions included in Iwata et al. (1982/1994) included social

disapproval (e.g., attention), demand, play, and alone. Out of the 32 studies reviewed,









87% included each of those conditions (i.e., attention, demand, play, and alone/ignore).

Ninety-seven percent included attention, demand and play conditions, 92% included

these conditions with the addition of the tangible condition (i.e., attention, demand, play,

and tangible), and 85% with the addition of alone/ignore conditions (i.e., attention,

demand, play, tangible, and alone/ignore).

The demand condition was conducted in 100% of the studies reviewed. The

attention condition was conducted in 97% of the studies, the play condition in 94%,

tangible in 78%, and the alone/ignore condition was conducted in 56% of the studies

evaluated. The inclusion of the tangible condition and the absence of an alone/ignore

condition are the main differences of many of the functional analyses conducted in this

review and the functional analysis procedures originally described in Iwata et al.

(1982/1994).

Experimental condition modifications

Variations of experimental conditions were occasionally conducted in attempts to

clarify undifferentiated or inconclusive functional analysis findings.

Behaviors reinforced. Asmus, Franzese, Conroy, and Dozier (2003) varied the

behaviors reinforced (e.g., reinforcing stereotypy and/or disruption) following a

functional analysis with undifferentiated findings. These variations resulted in identifying

a tangible function for disruption (when only disruption was reinforced) and an automatic

function for stereotypy (when both stereotypy and disruption were reinforced).

Thompson, Fisher, Piazza, and Kuhn (1998) varied reinforcement of specific forms

of the same behavior (i.e., aggression) following an inconclusive functional analysis in

which all forms of aggression were reinforced. Based on descriptive assessments,

Thompson and colleagues hypothesized one form of aggression (i.e., chin grinding) was









not maintained by social attention, while all other forms of aggression were maintained

by social attention. This hypothesis was tested by conducting two attention conditions,

one in which all topographies of aggression were reinforced with attention and one in

which all topographies of aggression except chin grinding were reinforced with attention.

The findings confirmed the hypothesis generated from the descriptive assessments, other

forms of aggression (all except chin grinding) were maintained by social attention and

chin grinding was not sensitive to attention. This finding was significant as it

demonstrated the potential for different forms of the same category of behavior (e.g.,

aggression) to be maintained by different reinforcers. In addition, it demonstrated the

potential of aggression being maintained by a sensory function, a behavior not often

associated with automatic reinforcement.

Type of escape delivered. Following an inconclusive functional analysis,

Hagopian, Wilson, and Wilder (2001) conducted a modified escape condition, escape

from attention. Instead of providing the participant with a demand, attention was

provided, and contingent on problem behavior, attention (rather than the demand) was

briefly removed. This modification resulted in conclusive functional analysis findings,

indicating both an escape from attention and tangible function. In addition to modifying

the escape condition to include escape from attention, rather than escape from demands,

additional modifications included escape from food (Levin & Carr, 2001; Piazza et al.,

2003) and escape from noise (Tang, Kennedy, Koppekin, & Caruso, 2002).

Type of attention delivered. To identify specific variables of social attention

maintaining problem behavior, Fisher et al. (1996) evaluated the type of attention

delivered (verbal reprimand versus unrelated verbal statement) contingent on destructive









behavior. The findings indicated destructive behavior was maintained by attention in the

form of verbal reprimands and not by attention in the form of unrelated statements. Along

the same line, Hagopian, Contrucci Kuhn, Long, and Rush (2005) evaluated two types of

attention (physical and verbal) in the maintenance of aggressive behavior and found that

both forms of attention maintained the aggressive behavior.

Type of tangible delivered. Graff, Lineman, Libby, and Ahearn (1999) evaluated

the type of tangible provided (i.e., toy versus edible) contingent on screaming. In

tangible-toy condition, screaming resulted in brief access to a toy and in the tangible-

edible condition, screaming resulted in a small edible. The findings of this functional

analysis were inconclusive.

Treatment of problem behavior

Twenty-five percent (8) of the studies reviewed conducted only a functional

analysis and did not continue onto treatment. Thirteen percent (4) conducted a functional

analysis and then continued with further assessment such as antecedent, preference,

competing items, or choice assessments. The remaining 63% (20) conducted treatment

within the study.

Design of functional analysis

The functional analysis procedures described by Iwata et al. (1982/1994) was an

extended functional analysis conducted utilizing a multielement design. Of the 32 studies

reviewed, 88% (28) utilized a multielement design, 13% (4) a multielement design with a

brief functional analysis, 13% (4) reversal, and 3% (1) utilized a pair-wise (test-control)

design. These categories are not mutually exclusive, therefore the percentages will not

add up to 100% when totaled. The following discussion illustrates how investigators have

used several designs to identify functions of problem behavior. For example, Marcus,









Vollmer, Swanson, Roane, and Ringdahl (2001) demonstrated a procedure to use the

least amount of time necessary to conduct a functional analysis. Marcus and colleagues

conducted brief functional analyses with 3 participants. If a function was not identified in

the brief functional analysis, an extended functional analysis was conducted. The brief

functional analysis was successful in identifying functions for 2 of the 3 participants and

the extended multielement functional analysis successfully identified the function of

aggression for the third participant. These findings suggest an extended functional

analysis may not always be necessary to identify the maintaining variables of aggression.

Along the same line, Brown et al. (2000) conducted a brief multielement functional

analysis with rapidly changing conditions with one participant and an extended

multielement functional analysis was a second participant. For both participants,

functions of their problem behaviors were identified. This was further confirmed when

Finkel, Derby, Weber, and McLaughlin (2003) conducted a brief functional analysis of

elopement and disruption in the classroom with the teacher serving as the therapist. The

brief functional analysis met the time limitations required in the natural setting and

successfully identified an escape function. One exception is a brief functional analysis of

saliva play conducted by Luiselli, Ricciardi, Schmidt, and Tarr (2004) in the participant's

classroom. The findings of the brief functional analysis were undifferentiated, further

supporting Marcus et al. (2001) findings that brief functional analysis may not always be

sufficient for identifying maintaining variables of target behavior for all participants. In

addition to extended multielement and brief functional analysis designs, reversal designs

were also employed to identify functions of problem behavior.









Richman, Wacker, Asmus, Casey, and Andelman (1999) varied topographies of

behavior reinforced to evaluate a potential hierarchy of behaviors. A brief functional

analysis was conducted using a reversal design to evaluate the topography of behavior

reinforced. A reversal design was also employed when evaluating the effect of therapists

(experimenter and parent) on the findings of an extended multielement functional

analysis (English & Anderson, 2004). Piazza et al. (2003) conducted a functional analysis

of inappropriate mealtime behavior utilizing a pair-wise comparison (test condition

compared to control condition) in a reversal design.

Additional variations include following an extended multielement functional

analysis with a series of consecutive ignore conditions (Sidener, Carr, & Firth, 2005) and

only including two experimental conditions (escape preferred food and escape

nonpreferred food) (Levin & Carr, 2001) in a reversal design.

Findings of functional analysis

The functional analysis methodology successfully identified the maintaining

variable of the target behavior(s) for 92% (43) of the 47 participants. The majority (74%)

of the participant's target behaviors were maintained by a single function. When only one

maintaining variable was identified, 28% (13) were maintained by an escape function,

13% (6) attention, 13% (6) tangible, and 13% (6) an automatic function. Thirty-two

percent (15) of participants' target behaviors were maintained by multiple functions.

When multiple maintaining variables were identified, 17% (8) were maintained by escape

and tangible functions, 9% (4) by attention and tangible, 2% (1) attention and escape, 2%

(1) tangible and escape from attention, and 2% (1) escape, attention, and tangible

functions. The functional analysis methodology did not successfully identify maintaining

variables for 8% (4) participants. Four percent (2) of the participants' functional analysis









findings were undifferentiated and 4% (2) were inconclusive. The categories are not

mutually exclusive and therefore will not add up to 100%. For example, one participant

had two target behaviors, each behavior with a separate, single function. Each function

was categorized separately, as a single function, not as multiply maintained. A participant

was indicated as having multiply maintained functions if the same target behavior(s) had

multiple functions. In addition, some studies conducted multiple functional analyses in

which more than one finding was indicated from each analysis. Rather than choosing

between which findings to report, both were included.

Summary of functional analysis review of literature

The review of these 32 studies has demonstrated the effectiveness of the functional

analysis methodology in identifying the variables maintaining a variety of problem

behaviors displayed by young children with ASD. To date, a functional analysis of peer-

related social behavior, particularly peer-related withdrawn and peer-related negative

social behavior has not been conducted with young children with ASD.

Strengths. The functional analysis literature has demonstrated good success in

identifying the functions of a variety of problem behaviors (e.g., SIB, aggression,

destruction, disruption, noncompliance, food refusal), across a variety of settings

(schools, homes, inpatient units, outpatient clinics), and therapists (teachers, parents,

experimenters) with young children with ASD. The increasing trend in the number of

studies conducted in the past four years is promising to the continuing demonstration of

the effectiveness of functional analysis methodology in the assessment of problem

behavior in young children with ASD. Despite the increase in the number of studies

conducted in the past four years, the vast majority (88%) only included 1 or 2 participants









per study, which does not provide a large participant pool from which to base the studies

findings.

The functional analysis methodology has been implemented across a variety of

therapists, though the majority (63%) were staff or experimenters and the majority (56%)

of studies were conducted in a session room located in an inpatient or outpatient clinic.

Additional studies utilizing parents, teachers, and peers in natural settings such as home

and school would provide additional support of the utility of this methodology in young

children with ASD.

Limitations. Out of the 32 studies reviewed, no functional analysis included an

assessment of any peer-related social behaviors, including peer-related withdrawn and

peer-related negative social behavior. In addition, peers were not included in any of the

studies as therapist or change agents in the assessment of target behaviors with young

children with ASD. However, some researchers did find that the functional analysis

identified different functions of problem behavior for the same child with ASD when

different therapists conducted the functional analysis conditions (English & Anderson,

2004). In addition, Thompson et al. (1998) found that different forms of the same

behavior (i.e., aggression) were maintained by different environmental variables.

Future directions. Given the success of the functional analysis methodology to

identify environmental variables of a variety of behaviors across a variety of therapists,

utilizing the functional analysis methodology to identify environmental variables of peer-

related withdrawn and peer-related negative social behaviors is a logical next step.

Review of Social Skills Literature

The purpose of this section is to review the literature on assessment and treatment

of social behavior in young children with ASD. Similar to the literature on problem









behavior with young children with ASD, the literature on social skills with young

children with ASD suggests target behaviors (e.g., appropriate social behavior) can be

manipulated (i.e., increased or decreased) based on modifications made to environmental

variables within the child's setting (e.g., peer-training).

Method to select reviewed studies. Appropriate articles for this literature review

were identified by searching the following computerized databases: Academic Search

Premier and PsycINFO. Keywords used to search the databases included, but are not

limited to social skills intervention, social skills assessment, social competence, social

skills, and autism.

The social skills body of literature met the following criteria for inclusion in the

review: (a) peer-reviewed journal, (b) published on or after 1985, (c) assessment and/or

treatment of peer-related social behavior (at least one of the peer-related social behaviors

had to be observable) conducted using a single-subject design including a variety of

designs (e.g., multiple baseline across participants, multiple baseline across settings,

reversal), (d) at least one participant with ASD 8-years of age or younger.

Review of Social Skills Studies

This section reviews the literature on social skills training and treatment of young

children with ASD. A total of 16 studies met the criteria mentioned above. The studies in

the review ranged from 1992 to 2005. One noteworthy finding to mention is there has

been a decrease in studies that include assessment and/or treatment of peer-related social

behaviors using a single subject design with young children with ASD; 10 studies were

conducted between 1992 and 1997, whereas 6 studies were conducted between 1998 and

2005. The most studies conducted in one year were five, in the year 1997. Since that year,

the most studies conducted in one year were two.









Participants included in review of literature

There were a total of 50 participants across the 16 studies identified. The number of

participants that met the criteria for inclusion are described in Appendix B (Review of

Social Skills Literature). In addition, the number of participants who met criteria in each

study, age, gender, and disability are reported in Appendix B.

Age. All participants included in these studies were between the ages of 3 and 8-

years old. Sixty percent (30) of the participants were between 3 and 5-years old and 40%

(20) were between 6 and 8-years old.

Number. Of the 16 studies included in this review, 44% (7) had three participants

that met the criteria for inclusion. Twenty-five percent (4) of the studies had two

participants, 25% (4) had four participants, and 6% (1) had five participants that met

criteria for inclusion in this review.

Gender. Of the 50 participants included in this review, 96% (46) were male and

8% (4) were female.

Disability. All 50 participants had either a diagnosis of PDD-NOS, autism, or

Asperger's disorder. Under the category of ASD, 90% (45) had a diagnosis of autism, 8%

(4) a diagnosis of Asperger's, and 2% (1) a diagnosis of PDD-NOS.

Therapists in studies

Of the 16 studies reviewed, peers were involved in each study. In some studies

peers were just present during the assessment and/or treatment and in other studies the

peers served as therapists to the participants. Specifically, peers were therapists in 69%

(11) of the studies reviewed. Thirty-eight percent (6) of the therapists were participants'

teachers, 31% (5) were experimenters, and 6% (1) were participants' caregivers. There

was more than one therapist (e.g., a peer and teacher) in 50% of the studies reviewed.









Specifically, 25% (4) of the studies included peer and teacher therapists, 19% (3)

included peer and experimenter therapists, and 6% (1) included the participants'

caregiver and experimenter as therapists. In studies that included more than one therapist,

the data from both therapists were reported individually (e.g., peer) and together (e.g.,

peer and teacher); therefore, the categories used to report therapists will exceed 100%

when totaled. The following are examples of how investigators employed a variety of

therapists in the assessment and/or treatment of social skills. Garfinkle and Schwartz

(2002) investigated the effects of peer imitation training on the social interaction of

young participants with ASD. Teacher therapists led small group peer imitation training

with the participant and 4 to 5 peers, all attending an integrated preschool. Training

consisted of the participant and peers imitating the play of the peer leader. The leader

switched throughout each session, so that during each session, each participant and peer

was a leader at least once. Therefore, peers served as models for the participant to imitate

and the teacher served as the trainer, providing least to most prompts to the participant

and peers to ensure imitation of the peer leader. In another study, Thiemann and

Goldstein (2004) utilized peers more directly as therapists. In this study, the investigators

trained peers to implement written text treatment with participants and examined the

effects of this intervention on the social behavior of young participants with ASD. This

investigation employed both the experimenter and peers as therapists. The peers were

trained by experimenters to use written text cues to facilitate social behavior between the

peers and the participant. After training, treatment began with the peer therapists and the

participant. The peer therapists were given scripts and instructions to follow throughout

the treatment sessions. The experimenter's involvement was limited to pre-session









directions, nonverbal feedback (i.e., circling a happy face for a correct prompt) during

sessions, and post-session feedback to peer therapists. Therefore, in this investigation, the

peers and experimenters served as therapists to the participant, though the peers were the

primary change agents, with the experimenter mainly serving as a trainer to the peers and

the participant. Alternatively, in an investigation conducted by Shabani, Katz, Wilder,

Beuchamp, Taylor, and Fischer (2002), peers were present during treatment; however,

peers were not provided with any training or instruction during treatment. Instead, the

experimenter provided initiation training to the participant using a tactile prompt (i.e.,

vibrator) and then evaluated the effects of the tactile prompt on the frequency of

participant initiations during free play with peers. Therefore, in this investigation, the

experimenter was the therapist and the peers were simply present during the baseline and

treatment assessment.

Settings of studies

All 16 studies were, in part, conducted in a school setting. Fifty percent (8) were

conducted in an integrated classroom, 31% (5) were conducted in a classroom when the

majority of the classmates were not present, 25% (4) in a general education classroom,

19% (3) in a special education classroom, and 13% (2) in the participants' home. In some

studies assessment and/or treatment was conducted across more than one setting (i.e.,

participant's special education classroom and general education classroom). When this

occurred, both settings were recorded; therefore, the categories used to report settings

will exceed 100% when totaled.

Dependent measures

There were several social behaviors measured across the 16 studies. The categories

of behaviors reported are not mutually exclusive and therefore, will exceed 100% when









totaled. For example, one study measured initiations and responses. Both initiations and

responses were included as categories of behavior reported. Therefore, when totaled, the

number of categories of behavior recorded (i.e., 2) divided by the number of studies (i.e.,

1) multiplied by 100%, would exceed 100%. In addition, some researchers reported

measuring the same dependent variable (e.g., initiation) but included different behaviors

(e.g., request for attention, request for action, helping) within their own operational

definitions of initiation. As a result, the following should not be considered a

comprehensive list of all peer-related social behaviors measured in these studies. A

noteworthy point is that direct measures of observable behaviors, rather than scores from

standardized achievement, intelligence, behavioral, or other measures were included in

this section. Sixty-nine percent (11) studies measured participant initiations, 50% (8)

participant responses, and 44% (7) measured interactions. Thirteen percent (2) of the 16

studies measured unprompted/spontaneous initiations, 6% (1) of studies measured

initiations and responses in the form of a compliment, verbal initiations, use of

augmentative communication to initiate, respond, and interact, social behavior (including,

initiations, responses, and negative behavior), and appropriate social skills. Six percent

(1) of studies measured use of language during interactions (i.e., number of words per

sentence), peer imitation, proximity to peers, engagement with materials while

interacting, and disruptive behavior.

Design of studies

The majority (75%) of the 16 studies utilized a multiple baseline design in the

measurement and treatment of social skills. Fifty-six percent (9) utilized a multiple

baseline design across participants, 25% (4) a reversal, 13% (2) a multiple baseline across

settings, 13% (2) a multiple baseline across peer trainers, 6% (1) a multiple baseline









across activities, and 6% (1) a multiple baseline across behaviors. The categories are not

mutually exclusive as some studies included multiple designs (e.g., multiple baseline

across participants and settings), therefore the percentages will exceed 100% when

totaled. For example, Garrison-Harrell, Kamps, and Kravits (1997) evaluated the effects

of peer networks on the duration of social interactions using a multiple baseline design

across settings within a multiple baseline across 3 participants. Therefore, peer networks

were systematically implemented across three settings for each of the 3 participants.

Assessment of social skills

Unlike the functional analysis studies reviewed in the previous section, a separate

assessment of social behaviors prior to treatment was not conducted. However, all studies

included a baseline phase that provided assessment information on the levels of social

behaviors of the participants prior to treatment. For example, Wert and Neisworth (2003)

measured the frequency of spontaneous requests made by four young children with ASD.

These measurements were obtained during 30 minute play sessions prior to the

implementation of treatment. The frequency of spontaneous requests made during these

baseline observations were then compared to the frequency of spontaneous requests made

during treatment (video self-modeling). Similarly, Laushey and Heflin (2000) obtained a

percent of appropriate social skills for two young children with ASD over a four week

period, prior to the implementation of treatment (i.e., active peer tutoring). The

percentage of appropriate social skills from baseline (i.e., pre-treatment) was then

compared to the percentage of appropriate social skills during treatment. One noteworthy

difference between the functional analysis studies reviewed in the previous section and

the social intervention studies reviewed in this section, is that although baseline measures









provided an assessment of the participants' social behaviors, the treatments were not

driven by the results of these assessments.

Treatment of social skills

All 16 studies reviewed included a treatment phase that focused on increasing the

appropriate social behavior of the participants. A variety of treatments were implemented

across these studies, though all treatments were determined a priori (i.e., treatments were

not assessment-based). In the majority, 63% (10), of treatments peers played a major role

in the delivery and implementation of the treatment (i.e., peer pivotal response training,

peer networks, peer training, peer incidental teaching, peer activity tutoring, peer

imitation, peer video modeling, written text treatment, peer self-evaluation). In 33% (5)

of the studies teachers were a part of the treatment (i.e., teacher led social skills training

with and without feedback and contingent rewards, teacher incidental teaching, teacher

facilitation of participant peer interactions). Finally in 33% (5), treatments that included

only the participants' were implemented (i.e., tactile prompt, priming, self-video

modeling, self-management, positive reinforcement). The purpose of these treatments

was to increase the occurrence of appropriate social behavior from baseline levels. For

example, Pierce and Schreibman (1997a) investigated the effects of peer pivotal response

training (PRT) on the use of language during interactions by young children with ASD.

The participants increased their frequency of language during interactions from baseline

to intervention. Similarly, in another example, the duration of social interactions for three

young children with ASD increased from baseline following the implementation of peer

networks in an investigation by Garrison-Harrell et al. (1997).









Findings of social skills assessment and treatment

All 16 studies reviewed reported increases in the occurrence of social behavior

following treatment, in comparison to baseline measures. For 88% (14) of the studies, all

participants in the studies displayed increases across all target appropriate social

behaviors. For 12% (2) of the studies, all participants displayed an increase in at least one

target social behavior. Five of the 16 studies reviewed included a generalization

component in the treatment analysis (e.g., generalization across settings, peers, time). For

4 (80%) of those 5 studies, all participants successfully generalized the target social

behavior. For example, Zanolli, Daggett, and Adams (1996) evaluated the effects of

priming on the spontaneous initiations of two young children with ASD. Spontaneous

initiations increased for both participants, however priming only increased the

spontaneous initiations for both settings for one participant. The second participant

increased spontaneous initiations in one, but not both settings.

Summary of social skills review of the literature

The review of these 16 studies has illustrated the success of a variety of treatments

to increase the appropriate social behaviors displayed by young children with ASD. A

variety of topographies of social behavior have been targeted for assessment and

treatment. To date, a treatment of social behavior driven by the findings of a functional

analysis of peer-related social behavior has not been conducted with young children with

ASD.

Strengths. The social skills literature has demonstrated good success in developing

treatments that increase a variety of appropriate social behaviors (e.g., initiations,

responses, interactions, peer imitation, language used during interactions), across a

variety of settings (e.g., general education classrooms, integrated classrooms, special









education classrooms, home) and therapists (e.g., peers, teacher, experimenter) with

young children with ASD. All 16 studies reviewed were conducted, at least in part, in a

school setting and with peers present. This demonstrates the feasibility of implementing

treatments focused on increasing participants' appropriate peer-related social behavior in

the participant's natural environment with their peers.

Treatments were evaluated using single subject design methodology, with the

majority (75%) of studies employing multiple baseline design and the remaining 25% of

studies employing a reversal design. The use of these single subject designs provided

experimental control to demonstrate that the implementation of the treatment components

rather than an unknown variable resulted in the increases in social behavior (Kazdin,

1982).

Limitation. One of the main limitations in this literature is the lack of assessment-

based treatments. Despite the success the pre-determined treatments have had in

increasing social behavior and demonstrating experimental control of these successes

through the use of multiple baseline and reversal designs, the reasons young children with

ASD engage in social interactions with their peers is unknown (i.e., the outcomes of their

social interactions). Without this information it is unknown whether treatments linked to

the functions or outcomes of peer-related social behaviors would result in more robust

and durable changes in behaviors, (e.g., increases in social behavior across participants,

behaviors, and settings, and improvements in generalization and maintenance).

One interesting findings is that even though all 16 studies reviewed were conducted

in applied settings, teachers served as therapists in less than half the studies (38%). In

order for these treatments to continue over time it may be necessary to increase the use of









teachers as change agents. It is unknown if the low percent of teachers serving as

therapists is due to the willingness of the teachers to participate, feasibility of teacher

implementation of treatment procedures or other reasons. However, if the treatment is not

feasible for a teacher to implement, even if it is successful in increasing social behavior

during the course of the investigation, the treatment may not be continued, resulting in

diminished treatment gains.

Future directions. Given the demonstrated effectiveness of the functional analysis

methodology to identify variables maintaining a variety of behaviors, across a variety of

settings and therapists, a logical next step in the social skills literature is to examine the

use of functional analysis methodology to develop assessment based treatments.

Conclusions and Future Directions for Research

In this chapter, two bodies of literature were reviewed to (a) demonstrate the

effectiveness of functional analysis methodology in identifying the function of problem

behavior for young children with ASD, and (b) illustrate assessment and treatment of

social behaviors in young children with ASD. The review of the functional analysis

literature demonstrated the success of this methodology to identify maintaining variables

of a variety of problem behaviors of young children with ASD. Despite the variety of

problem behaviors assessed, the functional analysis literature does not include an

assessment of peer-related social behavior for young children with ASD. The review of

the social skills literature illustrated a variety of treatments successfully implemented to

increase the appropriate social behavior of young children with ASD. However, not all

treatments were demonstrated to be successful across all participants, settings, or time.

The social skills literature does not include treatments based on functions or outcomes

that maintain participants' social behaviors. There is a paucity of research utilizing









functional analysis of peer-related social behavior with young children with ASD. Given

the success of the functional analysis to identify variables maintaining problem behavior

and develop robust treatments, the extension of the functional analysis methodology to

assess peer-related social behavior appears to be the next step.

This study seeks to expand the literature by applying functional analysis

methodology as a means of assessing peer-related social behaviors of young children

with ASD. To that end, the study has the following purposes: (a) to evaluate the

effectiveness of the use of the functional analysis methodology to determine the function

of peer-related withdrawn and peer-related negative social behavior for three young

participants with ASD, (b) to compare the results of a functional analysis of peer-related

withdrawn and peer-related negative social behavior across participants with ASD to

determine if the functions differed across participants, (c) to determine if peers were able

to implement the functional analysis procedures with acceptable integrity, and (d) to

examine rate per minute of participant peer-related positive initiations and percentage of

participant peer-related positive and peer-related no response social behavior across

functional analysis conditions.














CHAPTER 2
METHOD

Overview. This study is an extension of an OSERS funded project (Asmus &

Conroy, 2003). The project is a 3-year study that examined the use of ABA techniques in

the assessment and treatment of peer-related withdrawn and peer-related negative social

behavior displayed by young children (between the ages of 18 months and 5-years) with

ASD in their natural settings.

Participants

Participants were the first three children enrolled in an OSERS funded project that

reached the functional analysis phase of assessment (Asmus & Conroy, 2003). The

project involved the assessment and treatment of peer-related withdrawn and peer-related

negative social behavior for participants diagnosed with ASD. The participants were

referred to the project by their parents or the University of Florida Center for Autism and

Related Disabilities (CARD). Enrollment criteria for inclusion in the project included: (a)

the participant was between the ages of 18 months and 5-years of age, (b) the participant

had a diagnosis that fell under the category of ASD, (c) the participant was identified and

directly observed to engage in peer-related withdrawn and/or peer-related negative social

behavior in the school setting, and (d) the parents, teacher, and principal/director agreed

to participate in the project. The participants included in this study were Shane, Shannon,

and Colin. Pseudonyms were used to protect participants' anonymity.

Shane. Shane was a 5-year-old boy diagnosed with Pervasive Developmental

Disorder, Not Otherwise Specified (PDD-NOS). He attended a pre-school varying









exceptionalities classroom at a local public school where he received speech/language

and occupational therapy (OT) services. After school, Shane attended a private general

education pre-school with 18 students and one teacher, where all assessments were

conducted. Shane had mild cognitive delays and significant expressive language delays

with echolalia. Shane could follow simple requests. He rarely used speech to

communicate spontaneously, though he would repeat phrases such as, "Help please"

when repeatedly asked to do so by an adult. He mainly used gestures such as handing

items or leading adults to communicate his wants and needs. Shane had limited play

skills (i.e., played with Lego's by banging them together or on the floor instead of

building with them) and spent the majority of his time engaging in stereotypic behavior.

Shane's peer-related withdrawn social behavior was leaving the area in which peers were

present. Shane's peer-related negative social behavior was aggression (i.e., pinching,

biting, hitting, kicking, pushing, throwing toys at peers). His stereotypic behavior was

pacing back and forth in the classroom, repeatedly shaking a small toy with his hand, and

making nonfunctional vocalizations.

Shannon. Shannon was a 4-year-old girl diagnosed with autism. She attended a

private general education pre-school classroom with 15 students and one teacher, where

all assessments were conducted. Previously, Shannon had participated in a separate,

unrelated study in her home environment that evaluated treatment for problem behavior

(e.g., noncompliance, tantrums). Concurrently with this study, Shannon participated in a

separate, unrelated study that focused on increasing her verbalization skills. Shannon

could follow simple requests and spoke in full sentences, though her speech was often

difficult to understand as she often spoke very softly. She displayed independent play









skills (e.g., pushed buttons to open and close the drawer of a play cash register) and

independent pretend play (e.g., pretended she was making pizza with play food/oven and

pretended Barbie's were a family with a mother and father). Shannon's peer-related

withdrawn social behavior was leaving the area in which peers were present. Her peer-

related negative social behavior included yelling and/or whining, "No" and "I want the

(e.g., toy)." Shannon did not engage in stereotypic behavior.

Colin. Colin was a 3-year-old boy diagnosed with PDD-NOS. He attended a public

special education pre-school classroom with five students, one teacher, and two

paraprofessionals (one was assigned just to Colin). Colin received 30 minutes of

speech/language services twice a week and 30 minutes of OT services once a week. Colin

had extremely limited expressive communication and displayed immediate and delayed

echolalia. He inconsistently used gestures such as handing and leading adults to obtain

his wants and needs. Colin displayed independent play skills (e.g., putting together

puzzles) and some parallel play skills (e.g., putting together a puzzle while sitting next to

a peer). Colin's peer-related withdrawn social behavior was leaving the area where peers

were present. His peer-related negative social behavior included aggression (i.e.,

pinching, scratching, pulling hair, biting). Colin's stereotypic behavior was hand/arm

flapping.

Settings and Therapists

This study included three phases: (1) descriptive assessment, (2) preference

assessment, and (3) functional analysis. Each phase was conducted in the participant's

school setting. Phase 1 was conducted in the participant's classroom with all other

students present. Phase 2 and 3 were conducted in the classroom (except Colin) either

when all of the students were present (Shane) or when other students were at recess









(Shannon). For Shane, assessments were conducted during classroom activities that were

student directed (e.g., centers) to avoid interrupting the class schedule. For Colin, Phase 2

and 3 assessments were conducted in the atrium directly next to his classroom. During

Phase 3, one part of the classroom (designated by the teacher) was used to conduct the

sessions, with the exception of Colin, whose assessment was conducted in a designated

area of the atrium. Masking tape was placed on the floor to separate the assessment area

from the general classroom or atrium. The assessment area was then further split into

equal halves, by placing tape on the floor, to identify the in and I ithdh,. ii/ areas. The in

area was the area where peers were present and the withdrawn area was the area where

only the participant could go to be alone. The dimensions for both Shane's in and

withdrawn areas were 5 ft by 3 ft, Colin's were both 3 ft by 3 ft, and Shannon's were each

4 ft by 3 ft. For each participant, the dimensions for the in area were the same as the

dimensions of the withdrawn area. Additional physical borders (e.g., bookshelf) were

used for Shane and Shannon, as needed, to block off the assessment area (e.g., to

decrease opportunities for the participant to leave the entire assessment area).

Phase 1 (descriptive assessment) involved naturalistic classroom observations and

interviews with teachers; therefore there was no therapist. A graduate research assistant

with training in ABA served as the primary therapist for Phase 2 (preference assessment).

During Phase 3 (functional analysis), a graduate research assistant with training in ABA

trained the peers, provided coaching and assistance to the peers as needed in each

condition, and oversaw all sessions. In addition, during Phases 2 and 3, a second

therapist, either a graduate research assistant or undergraduate assistant, was present at









each session to operate the video camera, and in Phase 2 to serve as a second data

collector.

In Phase 3, a pool of five peers were selected from the participant's class (except

for Colin) based on their parent/guardian consent, consistent attendance, willingness to

participate, compliance and skills to follow one step verbal instructions, and intelligibility

of speech (volume and clearness). There were two peers present who participated in each

functional analysis condition. The two peers were chosen prior to each condition based

on availability (e.g., presence in the classroom) and willingness to participate. Attempts

were made to utilize the same two peers as often as possible. If two peers were not

available or were not willing to participate, sessions were cancelled for that day. Colin's

classmates did not participate due to lack of compliance and skills to follow one step

verbal instructions, and/or intelligibility of speech (volume and clearness). His peers were

chosen, using the same guidelines stated above, from a general education kindergarten

classroom at his school, with peers he had opportunities to interact with on previous

occasions.

Materials

The materials used in Phase 1 were materials from the participants' school setting.

Materials for use in the functional analysis (Phase 3) were identified for each participant

via a preference assessment (Phase 2). Seven stimuli were selected for the preference

assessment based on experimenter observations of materials the participants' played with

in the classroom during Phase 1; and materials that were conducive to socialization (e.g.,

peer-related social behavior could occur when playing with Lego's though would be more

difficult when listening to music on headphones). The experimenter reviewed all sessions

from Phase 1 and selected from a running tally of items the participant played with. Only









materials that could be played with by more than one child were selected for use during

the preference assessment. In addition, because there needed to be more than one item

available for three children to play with, only materials that had at least one identical

match (e.g., two Thomas the Engine electronic books, multiple wooden blocks) in the

classroom or that could be easily purchased were selected for inclusion in the preference

assessment.

Preference assessment. Prior to the functional analysis, a multiple stimulus

without replacement (MSWO) preference assessment was conducted with each

participant based on the procedures described in DeLeon and Iwata (1996). From this

assessment, high preferred, neutral, and low preferred tangibles were identified for each

participant, for use in the functional analysis (Phase 3). The materials differed for each

participant and varied across functional analysis conditions. Neutral tangibles were

available in all functional analysis conditions, except the ignore conditions (which had no

materials). The participant's high preferred tangible was available only during the

tangible condition of the functional analysis.

Response Definitions

Preference Assessment

During Phase 2, the participant's choice of a stimulus was scored for each trial. A

choice was defined as physically touching a stimulus. If two or more stimuli were

touched, the stimulus touched first was scored as the stimulus chosen. If two or more

stimuli were touched simultaneously, no stimulus was scored and the trial was repeated.

Functional Analysis

Several participant and peer behaviors were scored during Phase 3. The primary

dependent variables were the participants' peer-related withdrawn and peer-related









negative social behaviors. In addition, other participant peer-related social behaviors were

scored as appropriate initiationss, responses, no response).

The primary peer behaviors scored in Phase 3 were (1) presentation of social

demands to the participant in the escape condition, and (2) delivery of relevant

consequences following participant peer-related withdrawn and/or peer-related negative

social behavior in the escape, attention, and tangible conditions. Correct delivery of

social demands and consequences were recorded as indicators of procedural integrity.

Additional peer behaviors recorded were positive initiations in the free play condition.

Participant peer-related withdrawn behavior

Peer-related withdrawn social behavior was defined as more than 50% of the

participant's body crossing the vertical plane of the tape into the withdrawn area (the tape

separated the withdrawn and in areas). This included entering the withdrawn area and

leaving the entire assessment area (i.e., more than 50% of the participant's body crossing

the vertical plane of the tape that led out of the in area, though not into the withdrawn

area).

Participant peer-related negative behavior

Peer-related negative social behavior was individually defined for each participant.

Peer-related negative social behavior included the following behaviors: aggression (i.e.,

pinching, scratching, biting, hitting, kicking, pushing, pulling hair, throwing materials at

peers) and yelling/whining (i.e., "No," "I want the [toy]"). For Shane, negative

initiations and responses were also recorded. A negative initiation was defined as Shane

initiating any non-accidental verbal or gestural social behavior with the potential to cause

harm to the peer(s) in attempt to elicit a social response. It included destruction (e.g.,

destroying peers' objects/structures) and disruption (e.g., yelling). If Shane's negative









initiation also met the definition for peer-related negative social behavior, peer-related

negative social behavior was scored instead of negative initiation. A negative response

was defined as any non-accidental verbal or gestural social behavior with the potential to

cause harm to peer(s) when it overtly acknowledged an initiation within 3 seconds of the

end of the initiation (e.g., if the initiation was saying "Hi," a negative response was

scored if it occurred after the peer said, "Hi" and before 3 seconds had elapsed following

the termination of the peer initiation). This included physical destruction (e.g., destroying

peers' objects/structures) and disruption (e.g., yelling). If Shane's negative response also

met the definition for peer-related negative social behavior, peer-related negative social

behavior was scored instead of negative response.

Participant peer-related appropriate behaviors

Participant peer-related positive initiation. A positive initiation was defined as

the participant initiating any verbal or gestural social behavior directed toward a peer in

an attempt to elicit a social response. Positive initiations included verbal or gestural social

behavior that (a) provided assistance, comfort, or affection (e.g., gently putting hand on

peers back); (b) elicited peer attention or access to objects/activities (e.g., inviting peer to

join play activity, calling out peer name, offering materials, seeking information,

commenting, and complimenting, and/or seeking help/assistance); and (c) elicited

attention or access to objects including role assignment, directing an activity, and

coordinating pretend play and social contact.

Participant peer-related positive response. A positive response was defined as

any verbal or gestural social behavior that the participant engaged in to overtly

acknowledged an initiation within 3 seconds of the end of the peer's initiation (e.g., if a

peer initiated to the participant by saying their name, a positive response was scored if the









response occurred after the participant's name was stated and before 3 seconds had

elapsed following the name being stated). This included the participant helping a peer to

complete a task or other activity, following a peer's lead, helping behaviors, and

reciprocal social behaviors.

Participant peer-related no response. A no response was defined as the

participant either knowingly or unknowingly ignoring a peer(s) initiation or continuing to

engage in the same play behavior. No response was scored if the participant did not

engage in peer-related negative social behavior, and if the participant did not engage in a

peer-related verbal or gestural positive response (or negative response, for Shane only)

within 3 seconds of the termination of the peer's initiation (e.g., If a peer said, "Colin,

Colin, Colin" with no time gap, one no response rather than three no responses was

scored).

During the escape condition participant peer-related no response behavior was also

scored when the peer presented a social demand to the participant and the participant

either did not complete the social demand or did not complete the social demand

correctly (e.g., the participant was told to turn the page of a book and instead pressed a

button on the book).

Peer procedural integrity data

Peer social demands. Peer social demands were defined as any verbal or gestural

social demand presented to the participant during the escape condition. Correct peer

presentation of social demands was defined as the peer presenting a social demand

(independently or following a prompt from the experimenter) using a two (verbal,

gestural) or three step (verbal, gestural, physical) prompt sequence every 15 seconds

(excluding the 15 seconds the participant received contingent negative reinforcement). A









criterion rate of 4.0/minute that was consistent across escape conditions was identified as

a target to demonstrate adequate procedural integrity. Presenting social demands a mean

rate of at least 3.2/minute was considered to meet the integrity guideline.

Peer delivery of consequences. Peer delivery of consequences was defined as a

peer providing the participant with 15 seconds of the relevant consequence (attention,

escape, or access to the high preferred tangible) contingent on the participant's peer-

related withdrawn and/or peer-related negative social behavior. Consequence delivery

was defined for each condition as follows: 15 seconds of attention in the form of verbal

reprimands/requests to stay and play near peers in the attention condition; 15 seconds

access to high preferred tangible in the tangible condition, and a 15 second break (escape)

from social demands in the escape condition contingent on participant peer-related

withdrawn and/or peer-related negative social behavior. Correct peer consequence

delivery was defined as the peer providing the relevant consequence (independently or

following a prompt from the experimenter) within 5 seconds of the participant's peer-

related withdrawn and/or peer-related negative social behavior and providing the relevant

consequence for at least 15 seconds (the experimenter kept track of the 15 seconds using

a timer and told the peers when the 15 seconds had elapsed). A minimum criterion of

80% that was consistent across escape, tangible, and attention conditions was identified

as a target to demonstrate adequate integrity. Correct delivery of consequences an

average of 80% was considered to meet the integrity guideline. Peer consequences were

not provided for participant peer-related appropriate initiations and/or responses.

Peer positive initiations. A peer positive initiation was defined as the peer

initiating any verbal or gestural social behavior directed toward the participant in an









attempt to elicit a social response. This was recorded only during the free play condition.

Positive initiations included verbal or gestural social behavior that (a) provided

assistance, comfort, or affection (e.g., gently putting hand on participant's back); (b)

elicited participant attention or access to objects/activities (e.g., inviting participant to

join play activity, calling out participant's name, offering materials, seeking information,

commenting, and complimenting, and/or seeking help/assistance); and (c) elicited

attention. Correct presentation of a peer positive initiation was defined as the peer

positively initiating to the participant every 15 seconds (the experimenter kept track of

the 15 seconds using a timer and told the peers when the 15 seconds had elapsed). A

criterion rate of 4.0/minute that was consistent across free play conditions was identified

as a target to demonstrate adequate procedural integrity. Presenting positive initiations a

mean rate of at least 3.2/minute was considered to meet the integrity guideline.

Observation System and Interobserver Agreement

Observation System

During all three phases, all sessions and observations were videotaped using a

standard mini DV video camcorder. Audio was obtained from a wireless microphone

connected to the participant by hooking the microphone pack on the participant's pant

belt or waste band and clipping the microphone to the participant's shirt. An audio

receiver unit was connected to the camcorder to transmit audio from the wireless

microphone to the mini DV. This assisted in enhancing the participant and surrounding

peers' verbalizations and limiting background noise from the classroom and atrium. For

all phases, the mini DV tape was played back on the camcorder, in VCR mode, and

downloaded to Movie Star on an HP or Dell computer via DAZZLE or PINNACLE. In

Phase 1, all sessions were then viewed for data collection on a computer via Movie Star,









with video and audio occurring simultaneously. In Phase 2, data were taken both live and

via Movie Star. In Phase 3, all sessions were viewed for data collection via Movie Star or

via the camcorder (the DV was played on the camcorder on VCR mode and viewed on a

television screen by connecting the camcorder video and audio devices to a television

monitor).

During Phase 1, five (Shane and Colin) to six (Shannon) hours of naturalistic

observations across pre-specified contexts were collected. Data were collected using

paper and pencil to tally each material the participant engaged with for use in Phase 2. In

addition, a list of every participant and peer initiation was obtained to identify social

demands for use in Phase 3.

During the preference assessment (Phase 2) data were collected via paper and

pencil. A data sheet with five sessions and seven trials per session was utilized. The

stimulus chosen for each trial was written down in the corresponding session and trial.

The percentage of trials each stimulus was selected was calculated by dividing the

number of times a stimulus was chosen by the number of trials the stimulus was

presented. For example, if a stimulus was selected on the first trial of all five sessions the

formula was 5 / 5 x 100 = 100%. If a stimulus was selected on the second trial of all five

sessions the formula was 5 / 10 x 100 = 50%. The stimuli were then ranked, based on the

percentage of trials each stimulus was selected, as high preferred (stimulus chosen the

highest and/or second highest percentage of trials), neutral (1 to 3 stimuli chosen second

through fifth percentage of trials), and low preferred (2 to 3 stimuli chosen fifth through

seventh percentage of trials).









During Phase 3 (functional analysis), data were collected via Dell palm pilots using

the Tap It program (Tapp, 2003). Subsequent to scoring each day's sessions, the data

were downloaded from the palm pilots to a computer. The data were then analyzed using

the Multiple Option Observation System for Experimental Studies (MOOSES, Tapp,

2002) from which frequency, duration, and percentage reports of each behavior were

obtained. From these reports responses per minute were calculated by dividing the total

frequency of the behavior by the duration (i.e., minute) of the condition (i.e., 5 minutes).

Phase 3 behaviors. During Phase 3 participant peer-related positive initiations,

responses, no responses, withdrawn, and negative social behavior were scored as

frequency data. In addition, peer social demands, peer consequence delivery, and peer

positive initiations were scored as frequency data.

Frequency behaviors were scored at the start of the behavior. For example, each

time an initiation started (e.g., participant stated peer's name) the initiation (e.g.,

participant peer-related positive initiation) was scored. For each peer initiation a

participant response was recorded. Similarly, for each peer social demand, a participant

response was recorded.

The duration behavior scored in Phase 3 was peer delivery of consequences. For

example, when a peer started to provide contingent attention (in the attention condition)

following the participant's peer-related withdrawn and/or peer-related negative social

behavior, the peer delivery of consequences button on the palm pilot was turned on. The

delivery of consequences duration code was turned off 3 seconds after the peer stopped

providing the participant with the consequence (e.g., attention). Peer consequence

delivery was analyzed using occurrence, latency to onset, and duration measures.









Interobserver Agreement

Interobserver agreement was not collected during Phase 1 as it involved compiling

information from the descriptive observation to determine which materials to include in

the preference assessment (Phase 2) and what social demands peers would present

participants during the escape condition of the functional analysis (Phase 3). During

Phase 2, the primary and reliability observer independently collected data either live, via

the camcorder (the DV was played on the camcorder on VCR mode and viewed on a

television screen by connecting the camcorder video and audio devices to a television

monitor), or from a computer via Movie Star. The primary data collector, reliability data

collector, or an additional graduate research assistant conducted the preference

assessment with each participant. Data collectors consisted of the first author and/or

trained graduate research assistants. Interobserver agreement was calculated for each

session by dividing the number of agreements by the number of agreements plus

disagreements and multiplying by 100% (Kazdin, 1982). Interobserver agreement was

collected for an average of 63% of all preference assessment sessions. Interobserver

agreement was collected for 100% of Shane's sessions, 30% of Shannon's sessions, and

60% of Colin's preference assessment sessions. Interobserver agreement was 100%

across all sessions and participants.

During the functional analysis (Phase 3) the primary and reliability data collectors

consisted of the first author and/or trained graduate research assistants. Graduate research

assistants were trained to criteria prior to taking reliability data. A graduate research

assistant was considered trained after collecting data on three functional analysis

conditions with interobserver agreement of at least 80% when compared to the first

author. Both the primary and reliability observers collected data independently via a









camcorder (by playing the tape on VCR mode and connecting the camcorder to a

television screen) or Movie Star on a computer. Interobserver agreement on frequency

and duration behavior was calculated for each session using MOOSES (Tapp, 2002).

Interobserver agreement on frequency behaviors was calculated by MOOSES by dividing

the number of agreements by the number of agreements plus disagreements multiplied by

100% (Kazdin, 1982). An agreement for frequency behaviors was defined as both

observers recording the same behavior within 5 seconds before or after each other. A

disagreement for frequency behaviors was defined as the two observers not recording the

same behavior within 5 seconds before or after each other.

Interobserver agreement on duration behaviors was calculated by MOOSES by

dividing the smaller duration by the larger duration multiplied by 100% (Miltenberger,

1997). An agreement for duration behaviors was defined as both observers recording the

beginning of the same behavior within 5 seconds before or after each other and recording

the end of that same behavior within 5 seconds before or after each other. Interobserver

agreement was obtained on an average of 58% of sessions across participants. The

interobserver agreement for each behavior was averaged across conditions for each

participant to obtain an overall agreement for the functional analysis.

Interobserver agreement was obtained for 33% of Shane's functional analysis

sessions. Agreement averaged 85% (range 33% to 100%) for Shane's peer-related

withdrawn social behavior, 95% (range 80% to 100%) for peer-related negative social

behavior, 100% for peer-related negative initiations, 96% (range 80% to 100%) for peer-

related positive response, 100% for peer-related negative response, and 91% (range 78%

to 100%) for peer-related no response. Although positive initiations occurred,









interobserver agreement was not collected during those conditions and therefore was not

reported. Agreement averaged 76% (range 69% to 83%) for Shane's peer social demands

and 95% (range 89% to 99%) for Shane's peer delivery of consequences.

Interobserver agreement was obtained for 78% of Shannon's functional analysis

sessions. Agreement averaged 100% for Shannon's peer-related withdrawn social

behavior and 100% for peer-related negative social behavior. Agreement averaged 100%

for Shannon's peer-related positive initiations, 66% (range 0% to 100%) for peer-related

positive responses, and 87% (range 50% to 100%) for Shannon's peer-related no

response behavior. Agreement averaged 84% (range 65% to 95%) for Shannon's peer

social demands and 96% (range 95% to 97%) for peer delivery of consequences.

Interobserver agreement was obtained for 52% of Colin's functional analysis

sessions. Agreement averaged 100% for Colin's peer-related withdrawn social behavior

and 100% for his peer-related negative social behavior. Agreement was 100% for peer-

related positive initiations and for peer-related positive responses and 89% (range 68% to

100%) for peer-related no responses. Agreement averaged 95% (range 79% to 100%) for

peer social demands and 99% (range 99% to 100%) for peer delivery of consequences.

Design and Data Analysis

Design of Procedures

No design was used in Phase 1 (descriptive assessment) to select items for use in

the preference assessment (Phase 2) or to select social demands to be presented to the

participant in the escape condition (Phase 3). In Phase 2, (preference assessment) stimuli

were counterbalanced (placement in the line of items presented) across sessions and after

each trial each stimulus was moved one place to the left, to avoid possible sequence

effects (DeLeon & Iwata, 1996). Phase 3 (functional analysis) was conducted by









alternating between four to five conditions using a multielement design (Barlow &

Hayes, 1979; Iwata et al., 1982/1994). The functional analysis conditions were presented

in a random order and/or based on continuous analysis of data, to avoid possible

sequence effects. The order was determined by placing the name of each condition in a

hat and pulling out one at a time or based on ongoing analysis of conditions. For

Shannon, a second functional analysis was conducted with minor procedural variations,

though conditions and design remained the same as the first functional analysis.

Data Analysis

During Phase 3, data were analyzed via visual inspection of single-subject design

line graphs. Although data were collected on a variety of participant and peer behaviors,

decisions regarding what conditions to conduct and when to stop data collection were

based only on rate per minute of participant peer-related withdrawn and/or peer-related

negative social behaviors. Rate per minute of these behaviors was calculated by

MOOSES (Tapp, 2002) by dividing the total frequency of each behavior by five (i.e.,

total session time in minutes). To evaluate the function of both the participant's peer-

related withdrawn social behavior and peer-related negative social behavior, the

behaviors were graphed separately. Therefore, one graph depicted rate per minute of the

participant's peer-related withdrawn social behavior and a second graph depicted rate per

minute of the participant's peer-related negative social behavior. Decisions on when to

end data collection were made separately based on each graph. Each graph was visually

inspected for stability based on procedures described by Kazdin (1982). A three point

trend rule was used as a guideline to make decisions on when to stop data collection. The

conditionss, in which each behavior occurred most often, in a stable trend, were

considered the function of that behavior.









Visual inspection techniques were used to determine which conditions) the

participant engaged in more peer-related withdrawn social behavior and/or negative

social behavior. The functional analysis was continued until a clear function (at least

three increasing, decreasing, or stable data points) was determined for the rate per minute

of each of these behaviors. The line graphs were compared to determine if the function of

peer-related withdrawn and/or peer-related negative social behaviors differed for each

participant. Following the analyses of rate per minute of peer-related withdrawn and

peer-related negative social behavior, averages of the peer integrity measures were

compared to determine the extent with which the peers implemented the functional

analysis procedures with acceptable procedural integrity.

Mean rate per minute of participant peer-related positive initiations, responses, and

no responses were compared post-hoc to determine if there were differences across

functional analysis conditions.

Procedures

Phase One: Descriptive Assessments

During the descriptive assessment seven stimuli were selected for use in the

preference assessment (Phase 2). Tangibles were identified through direct observation.

Social demands were identified for use in the escape condition of the functional analysis

(Phase 3). Social demands were selected through direct observation.

Direct observations were conducted in each participant's classroom. Five hours of

direct observation were conducted for Shane and Colin, and 6 hours were conducted for

Shannon. Observations were videotaped for subsequent viewing and scoring. These

observations were conducted across several days in order to obtain 30 minutes of data on

participant and peer social behavior in each of the following contextual factors: activity









(manipulatives/blocks, cognitive/books/pre-academic, art/sensory, pretend/socio-

dramatic, dance/music, snack/meal, computer, and games with rules), play format (adult

directed, child directed), adult engagement (adult active, passive, and disengaged), and

peer group size (one-on-one, small, and large). From these observations, a list of

tangibles and social demands were developed for use in the preference assessment (Phase

2) and functional analysis (Phase 3). The primary investigator or a research assistant

viewed the videotapes and made a list of all the tangibles the participant engaged with

during the direct observations. From this list and indirect assessments, tangibles were

selected for use in the preference assessment. The primary investigator or a research

assistant viewed the videotapes a second time and made a list of all the peer-related

initiations made by the participant and peer(s) during the direct observations. The exact

topography of each participant and peer initiation observed during the descriptive

observations was listed (e.g., peer initiated to participant by asking, "Will you hand me

the block?" or the participant initiated to the peer by holding up her doll in the direction

of the peer). From these peer-related initiations, social demands were selected for use in

the escape condition of the functional analysis.

Identification of tangibles. From information obtained from the direct descriptive

assessments, seven stimuli were selected for inclusion in the preference assessment. In

addition, stimuli were selected based on more than one person being able to play with the

stimulus (e.g., social behavior could occur when playing with Lego's though would be

more difficult when listening to music on headphones) and the availability of multiple

exact stimuli in the classroom setting. For Shannon, one additional stimulus (play cash

register) was included in the preference assessment that was not selected from the direct









descriptive assessment. The play cash register was included, because Shannon had been

observed to play with the cash register at other times and her teacher reported Shannon

would often play with this toy.

Identification of social demands. Social demands were determined individually

for each participant and based on the participant and peer initiations that were recorded

from the descriptive observations. Initiations that facilitated peer-related positive social

behavior (e.g., "Hand me the book") were selected as social demands in the escape

condition. Initiations that did not facilitate peer-related positive social behavior (e.g.,

"No!" "Stop it") were not selected to be used as social demands.

Phase Two: Preference Assessment

A preference assessment was conducted at each participant's school in either their

classroom or atrium (Colin). The seven stimuli that were selected for each participant

were evaluated for preference using procedures similar to DeLeon and Iwata (1996).

There were five sessions conducted, with seven trials per session. Additional preference

assessments were conducted as needed if there was concern that preference for items had

changed. The primary investigator or graduate research assistant served as the therapist

across all sessions. On the first trial of each session, all seven stimuli were presented in a

straight line across a table or on the floor, facing the participant. The participant sat at a

table (or on the floor) with the seven stimuli spread evenly across the table (or floor).

After being shown how to manipulate each stimulus the participant was asked to choose

what item s/he wanted to play with. The first stimulus the participant touched was

recorded as the chosen stimulus for that trial. The participant was then given the

opportunity to engage with the chosen stimulus for 30 seconds. Before beginning the next

trial the chosen stimulus was removed from the participant's view and the remaining









stimuli were moved one place to the left and evenly spread out in front of the participant.

The participant had 30 seconds on each trial to choose a stimulus. If no stimulus was

chosen, the session was terminated and the remaining trials were scored as "not chosen".

Attempts to choose two stimuli were blocked and resulted in the choice being

represented. Each stimulus was ranked by dividing the number of trials the stimulus was

chosen by the number of trials the stimulus was presented, then multiplying by 100%.

The stimulus (or stimuli) chosen the highest and/or second highest percentage of trials

was identified as the high preferred tangible item and was presented only during the

tangible condition. The stimuli chosen second through fifth percentages of trials were

ranked as the neutral tangibles and were present in all conditions except ignore. In

addition, the same ratio of neutral tangibles were present in the withdrawn and in areas.

For example, if Thomas the Train book was identified as a neutral stimulus, three

Thomas the Train books would be present in the in area and one Thomas the Train book

would be present in the withdrawn area in the attention, tangible, and escape conditions

of the functional analysis. This procedure was adapted for Shannon, as only one type of

stimulus (i.e., Barbie's) was ranked as neutral. Therefore, three Barbies were present in

the in area and one Barbie in the withdrawn area.

Phase Three: Functional Analysis

A functional analysis was conducted at each participant's school in the participant's

classroom (except Colin, whose functional analysis was conducted in the atrium outside

of his classroom). The procedures were based on Iwata et al. (1982/1994), though

differed in the following ways: (a) relevant reinforcement was delivered in test conditions

(i.e., attention, tangible, escape) contingent on peer-related withdrawn and/or peer-related

negative social behavior, rather than problem behavior directed toward adults, (b)









relevant reinforcement (i.e., access to attention, access to high preferred tangible, escape

from social demands) was delivered by peers rather than an adult, and (c) in the escape

condition, social demands were presented by peers rather than adults, and the demands

were to engage in peer-related positive responses rather than academic/vocational tasks.

In addition, the session area was split in half by placing masking tape on the floor and

entering one half (i.e., the withdrawn side) was designated as peer-related withdrawn

social behavior. Approximately three to five 5 minute conditions were conducted during

each 30 to 60 minute session. Sessions were conducted one to three times per week.

Reinforcement was provided for 15 seconds contingent on peer-related withdrawn and/or

peer-related negative social behavior in the attention, tangible, and escape conditions.

Training to Peers

Functional analysis conditions were presented to peers as games, each game

(condition) having different rules for how to play with the participant. Prior to each

condition, peers were told the rules for that condition (game) and the length (i.e., 5

minutes) of the condition. After the therapist verbally instructed the peers on the rules for

that condition, the peers were required to demonstrate understanding of the rules by role-

playing with each other and the participant. Each peer role-played as both the participant

and the peer. The condition only began after both peers demonstrated knowledge of the

rules by successfully role-playing the relevant responses to participant behavior with a

peer or the participant. Therefore, a peer was required to demonstrate 100% accuracy in

delivering the relevant consequence prior to participating in each condition. The

instructions provided to the peers are described below. The therapist provided reminders

and instructions to the peers as needed to ensure procedural integrity of the condition.









Free play condition

Peers were instructed to frequently talk, look at, and play with the participant. The

peers were also instructed to withhold consequences (e.g., attention, tangible, escape)

contingent on participant peer-related withdrawn or peer-related negative social behavior.

As needed, the therapist provided peers with reminders to talk, look at, and play with the

participant, and provided examples of things to talk to the participant about, and/or ways

to play with participant.

Ignore condition

Peers were instructed to talk, look at, and play only with each other and to not talk,

look at, or play with the participant. No tangible items (i.e., toys) were present in this

condition; therefore, peers were directed to pick a game they could play together without

toys (e.g., patty-cake) prior to the beginning of the condition. The therapist assisted, as

needed to identify an appropriate activity/game for the peers to play. As needed, the

therapist provided peers with reminders to talk, look at, and play only with each other,

examples of things to talk about with each other, and/or games to play with each other.

Attention condition

Peers were instructed to talk, look at, and play only with each other, unless the

participant engaged in peer-related withdrawn or peer-related negative social behavior.

Contingent on the participant engaging in peer-related withdrawn or peer-related negative

social behavior, peers provided the participant with 15 seconds of attention in the form of

verbal reprimands and/or requests to stay and play near them (e.g., "No!," "Stop leaving

all the time," "We want you to play by us"). The therapist kept track of the 15 second

contingent attention using a timer and told the peers when the 15 seconds had elapsed. As

needed, the therapist provided peers with reminders to talk, look at, and play only with









each other, to provide contingent attention to the participant, and when to stop providing

contingent attention.

Tangible condition

Peers were instructed to frequently talk about the participant's preferred tangible

item (selected in Phase 2 during the preference assessment), look at the participant, and

play with the preferred tangible, unless the participant engaged in peer-related withdrawn

or peer-related negative social behavior. Contingent on the participant engaging in peer-

related withdrawn or peer-related negative social behavior, peers provided the participant

with 15 seconds access to the preferred tangible. The therapist kept track of the 15

seconds contingent access to the preferred item using a timer and told the peers when the

15 seconds had elapsed. As needed, the therapist provided peers with reminders to

frequently talk, look at, and play with participant, to provide the participant with

contingent access to his/her preferred tangible, to provide noncontingenet access to

neutral tangibles, and when to stop providing the participant access to the preferred

tangible.

Escape condition

Peers were instructed to present the participant with social demands (selected by

the therapist based on information obtained in Phase 1) using a two or three step prompt

sequence every 15 seconds, unless the participant engaged in peer-related withdrawn or

peer-related negative social behavior. Verbal social demands (e.g., "Say hi to me") were

presented using a two step prompt sequence (verbal, model) and gestural social demands

(e.g., "Wave to me") were presented using a three step prompt sequence (verbal, model,

guide). Contingent on the participant correctly completing the social demand within 3

seconds of either a verbal or model prompt, the peers provided brief verbal praise (e.g.,









"Good") to the participant. Contingent on the participant engaging in peer-related

withdrawn or peer-related negative social behavior, peers told the participant "OK, you

don't have to" and provided the participant with a 15 second break from social demands.

The therapist kept track of the 15 seconds contingent escape from social demands using a

timer and told the peers when the 15 seconds had elapsed. The therapist provided peers

with the social demands to present to the participant and the sequence of prompts to use,

reminders to provide the participant with contingent escape from social demands, and

when to present the next social demand.

Functional Analysis

Prior to each condition the participant was directed/guided to stand over the taped

line (i.e., one foot in the in area and one foot in the withdrawn area) where the therapist

told the participant the rules for the relevant condition. The therapist then directed/guided

the participant to the in area prior to the start of the condition.

Free play condition

Prior to the free play condition the therapist reviewed the rules with the participant,

"You can play here (pointing to withdrawn area) or here (pointing to in area). Your

friends will talk and play with you no matter where you go or what you do." The therapist

then directed/guided the participant to sit in the in area next to the two peers. The peers

were seated next to each other so that when the participant sat, a semicircle was created.

Therefore, the condition began with the peers and the participant facing each other. The

participant had continuous access to neutral tangibles and access to peer attention at least

every 15 seconds. No peer consequences were provided for participant peer-related

withdrawn or peer-related negative social behavior.









Ignore condition

Prior to the ignore condition the therapist reviewed the rules with the participant,

"You can play here (pointing to withdrawn area) or here (pointing to in area), your

friends are only going to talk and play with each other no matter where you go or what

you do." The therapist then directed/guided the participant to sit in the in area, at the

point closest to the withdrawn area, sitting sideways to the peers. The two peers sat

across from each other in the in area, at the point furthest away from the withdrawn area.

No tangibles were available for the participant or peers and no peer consequences or

attention was provided for participant peer-related withdrawn, negative, or appropriate

social behavior.

Attention condition

Prior to the attention condition, the therapist reviewed the rules with the participant,

"You can play here (pointing to withdrawn area) or here (pointing to in area). Right now

it is your friends turn to talk only with each other, but if you go over here (pointing to

withdrawn area) or if you (peer-related negative social behavior) your friends will play

with you." Participant peer-related negative social behavior was individually defined for

each participant and stated specifically, such as, "You need to whine, play with me."

After the rules were given, the therapist then directed/guided the participant to sit in the

in area, at the point closest to the withdrawn area, sitting sideways to the peers. The two

peers sat across from each other in the in area, at the point furthest away from the

withdrawn area. Neutral tangibles were available to the participant and peers throughout

the condition. The neutral tangibles in the in area were split in half prior to the condition,

so that the participant and peers would have toys near them at the beginning of the

condition (i.e., toys were located by the peers and where the participant would be seated).









Peers provided 15 seconds of attention in the form of verbal reprimands and requests to

stay and play near them contingent on participant peer-related withdrawn or peer-related

negative social behavior. If the participant was in the withdrawn area after the 15 seconds

of contingent attention, the therapist neutrally directed/guided the participant to the in

area. No attention was provided for participant peer-related appropriate social behavior.

Tangible condition

Prior to the tangible condition the participant had access to the high preferred

tangible for 2 minutes. The participant was then directed/guided to the line while still

maintaining access to the high preferred tangible (i.e., holding on to it) as the therapist

reviewed the rules with the participant, "You can play here (pointing to withdrawn area)

or here (pointing to in area). Right now it is your friends turn to play with (high preferred

tangible), but if you go over here (pointing to withdrawn area) or (peer-related negative

social behavior) you can play with (high preferred tangible)." Participant high preference

tangible was defined individually based on preference assessment results and stated

specifically, such as "play-doh." Participant peer-related negative social behavior was

defined individually and stated specifically, such as, "Yell, I want to play with it." After

the rules were given, the therapist directed/guided the participant to the in area next to the

two peers. The condition began immediately following the peer removing the high

preferred tangible from the participant, while stating, "It is our turn." and then gave the

participant a neutral tangible. The participant had continuous access to neutral tangibles

and access to peer attention at least every 15 seconds. Peers played with the high

preferred tangible unless the participant engaged in peer-related withdrawn or peer-

related negative social behavior. Peers provided the participant with 15 seconds access to

the high preferred tangible contingent on peer-related withdrawn or peer-related negative









social behavior. After the 15 seconds access to the high preferred tangible, the peers

removed the high preferred tangible from the participant, while stating, "It is our turn,"

and then gave the participant a neutral tangible. If the participant was in the withdrawn

area after the 15 seconds of contingent access to the high preferred tangible, the therapist

neutrally directed/guided the participant to the in area after the 15 seconds. The peers

then removed the high preferred tangible and replaced it with a neutral tangible. No

access to the high preferred tangible was provided for participant peer-related appropriate

social behavior

Escape condition

Prior to the escape condition, the therapist reviewed the rules with the participant,

"You can play here (pointing to withdrawn area) or here (pointing to in area). If you play

here (pointing to in area) your friends are going to tell you how to play. If you don't want

to do what your friends tell you to do, you can (peer-related negative social behavior) or

go over here (pointing to withdrawn area) and play how you want to." Participant peer-

related negative social behavior was defined individually and stated specifically, such as,

"Yell, no!" After the rules were given, the therapist directed/guided the participant to the

in area between the two peers. The participant had continuous access to neutral tangibles

and access to peer attention at least every 15 seconds. Peers presented the participant with

social demands every 15 seconds using a two prompt sequence (verbal, model) for social

demands that required a verbal response and a three prompt sequence (verbal, model,

guide) for social demands that required a gestural response (e.g., "Turn the page").

Participant social demands were defined individually and based on descriptive

observations conducted in Phase 1. Contingent on participant peer-related withdrawn or

peer-related negative social behavior, peers stated, "OK, you don't have to" and provided









15 seconds of escape from social demands. If the participant did not complete the social

demand or did not complete the social demand correctly, the social demand was repeated

again (for a total of two presentations) prior to presenting the next social demand.

Participant compliance with social demands resulted in peers providing brief neutral

praise (e.g., "Thanks for talking to me," "OK," "Good") and continued presentation of

social demands every 15 seconds. If the participant was in the withdrawn area after the 15

seconds of contingent escape from social demands, the therapist neutrally directed/guided

the participant to the in area after the 15 seconds and the peers presented a social demand.

Escape was not provided for participant peer-related appropriate social behavior.

Second Functional Analysis

A second functional analysis was conducted with Shannon only. In the first

functional analysis, Shannon did not engage in peer-related withdrawn social behavior,

thus this behavior could not be evaluated. In addition, Shannon engaged in low to zero

rates per minute of peer-related negative social behavior, resulting in inconclusive

findings. As a result, a second functional analysis was conducted. The procedures in the

second analysis were similar to those in the first functional analysis, though altered in

attempts to reduce response effort to engage in peer-related withdrawn social behavior.

Shannon's second functional analysis was conducted in the same setting and with the

same peers as the first functional analysis. Approximately three 5 minute conditions were

conducted during each 30 minute session. Conditions included free play, escape, tangible,

and attention. The ignore condition was not conducted, as Shannon did not engage in

peer-related withdrawn or peer-related negative social behavior in the free play or ignore

conditions of the first functional analysis. Sessions were conducted one to three times per









week. Reinforcement was provided for 15 seconds contingent on peer-related withdrawn

or peer-related negative social behavior in the attention, tangible, and escape conditions.

The free play, attention, and escape conditions were conducted exactly the same

as described in the first functional analysis, with one exception. The only difference was

every 15 seconds the therapist directed/guided Shannon to stand over the taped line (i.e.,

one foot in the in area and one foot in the withdrawn area) and asked her, "Which side do

you want to go to?" Only Shannon's physical choice (i.e., which side she entered) was

recorded. When standing over the taped line, the response effort to enter the withdrawn

and in area was seemingly equal, as both required one step to either her right or left side.

This differed from the first functional analysis when the conditions began with Shannon

sitting in the in area, which required Shannon to stand up and walk at least two steps to

enter the withdrawn area. Practically, Shannon could have moved her body from the in

area to the withdrawn area without standing (e.g., crawling), though the number of motor

movements required for her to enter the withdrawn area would have remained higher than

the one step required in the second functional analysis. Shannon was only observed to

enter the withdrawn or in area by walking.

The tangible condition differed from the first functional analysis in two ways. First,

every 15 seconds the therapist directed/guided Shannon to stand over the taped line (i.e.,

one foot in the in area and one foot in the withdrawn area) with the high preferred

tangible in her hand, and asked her, "Which side do you want to go to?" Only Shannon's

physical choice (i.e., which side she entered) was the recorded. When standing over the

taped line, the response effort to either enter the withdrawn or in area was seemingly

equal, as both required one step to either her right or left side. Secondly, peers did not









deliver the consequence (i.e., access to high preferred item) for peer-related withdrawn

social behavior. If Shannon entered the in area, the therapist directed/guided her to give

the high preferred tangible to her peers. Peers then played with the high preferred

tangible. Contingent on peer-related negative social behavior, peers provided Shannon

access to the high preferred tangible for 15 seconds. However, if Shannon entered the

withdrawn area, she simply maintained access to the high preferred tangible. Therefore,

peers did not deliver the consequence (i.e., access to high preferred tangible) contingent

on peer-related withdrawn social behavior. This procedure was different from the first

functional analysis when Shannon was required to engage in peer-related withdrawn

social behavior prior to receiving the consequence (i.e., high preferred tangible) and the

peers delivered the consequence rather than Shannon maintaining access to the high

preferred tangible from standing at the line.














CHAPTER 3
RESULTS

Overview of Research Aims

The purpose of this investigation was to (a) evaluate the effectiveness of the use of

the functional analysis methodology to determine the function of peer-related withdrawn

and peer-related negative social behavior for three young participants with ASD, (b)

compare the results of a functional analysis of peer-related withdrawn and peer-related

negative social behavior across participants with ASD to determine if the functions

differed across participants, (c) to determine if peers were able to implement the

functional analysis procedures with adequate integrity, and (d) to examine rate per minute

of participant peer-related positive initiations and percentage of participant peer-related

positive and no response social behavior across conditions.

These research aims were addressed via three phases of assessment. During Phase

1, a list of participant and peer initiations and a list of tangibles (e.g., toys) was generated

for each participant based on descriptive assessment. Data from Phase 1 were obtained by

observing videotapes from 5 (Shane and Colin) to 6 (Shannon) hours of direct

observation and listing the topography of participant and peer initiations and the tangibles

the participants engaged with. The initial list of peer-related initiations was then

narrowed down, according to several criteria, to 4 to 11 initiations, to be used as social

demands for the escape condition of the functional analysis (Phase 3). The initial list of

tangibles was narrowed down to seven, according to the ability to be manipulated by

more than one child and the feasibility of obtaining at least two identical tangibles, for









use in the preference assessment (Phase 2). During Phase 2, seven tangibles were

assessed for preference and then rank ordered (high, neutral, and low) for use in the

functional analysis (Phase 3). The tangible(s) chosen on the highest percentages) of trials

was ranked as high preferred and used during the tangible condition in Phase 3. The

tangible(s) ranked as neutral were used during the free play, attention, tangible, and

escape conditions in Phase 3. During Phase 3, a functional analysis based on Iwata et al.

(1982/1994) of peer-related withdrawn and peer-related negative social behavior was

conducted. The functional analysis included free play, attention, tangible, escape, and

ignore conditions. Data from Phase 3 were analyzed using visual analysis.

Phase One: Descriptive Assessment Results

The descriptive assessment consisted of identifying 4 to 11 social demands and

seven tangibles for each participant for use in Phases 2 and 3. The results of Phase 1 are

presented separately for each participant.

Shane Descriptive Assessment Results

Social demands identified. During direct observation, Shane and his peers were

observed to engage in a total of 14 peer-related verbal initiations and 17 peer-related

nonverbal (gestural) initiations (see Appendix C-1, Shane's Verbal and Gestural

Initiations). Of those initiations, six were selected as social demands and were used in the

escape conditions during Phase 3 (as described in Chapter 2). The seven social demands

selected for use in the escape condition of Phase 3 were: 1) "Turn the page," 2) "Press the

button," 3) "Look at me," 4) "Touch my shoulder," 5) "Hold my hand," and 6) "Give me

a high five" (see Table 3-1, Social Demands Used in Escape Condition of Functional

Analysis).









Tangibles identified. The tangibles identified for Shane via direct observation are

listed in Table 3-2. One point that should be noted is Shane typically engaged with these

tangibles in a stereotypical rather than a functional manner. It was hypothesized that

when Shane engaged in stereotypy he was less likely to engage in social behavior

regardless of environmental consequences (i.e., stereotypy may have been automatically

reinforced and competed with social behavior). Therefore, the procedures to identify

tangibles for use in the preference assessment were supplemented with descriptive

assessments in attempts to identify tangibles Shane would engage with in a functional

manner. The additional assessment included Shane's teacher identifying tangibles Shane

had been observed to engage with in a functional manner. Shane's teacher identified the

ABC Phonics, sound-effect books, and a plastic ball. Of these tangibles, seven were

selected, using direct observation procedures (as described in Chapter 2) and information

obtained from supplementary descriptive assessments. The seven tangibles selected for

use in the preference assessment (Phase 2) were: 1) Scooby-do book, 2) foam fish, 3)

Thomas the Engine book, 4) plastic ball, 5) ABC Phonics, 6) Bob the Builder book, and

7) Lego's (see Table 3-3, Tangibles Included in Preference Assessment). Of note, the

books identified via direct observation did not include Scooby-do, Thomas the Engine, or

Bob the Builder, as those books were not present in the classroom. The classroom had a

limited number of books with sound effects and Shane had a history of bending and

breaking books. Therefore, based on the teacher's recommendation, these specific books

were purchased for use in the preference assessment.

Shannon Descriptive Assessment Results

Social demands identified. During direct observation, Shannon and her peers were

observed to engage in a total of 63 peer-related verbal initiations and 41 peer-related









nonverbal (gestural) initiations (see Appendix C-2, Shannon's Verbal and Gestural

Initiations). Of those initiations, 11 were selected as social demands and were used in the

escape conditions during Phase 3. The 11 social demands selected for use in the escape

condition of Phase 3 were: 1) "Pick a dinosaur," 2) "Show me what you are making," 3)

"Hand me the (e.g., block)," 4) "Look at (e.g., me)," 5) Put the (e.g.,

block) here," (where the peer points) 6) "Help me make a snake with (e.g., the play-

doh), you roll this side," (where the peer points) 7) "Put your hands on your lap," 8)

"Let's play patty cake, come sit by me," 9) "Say hi," 10) "Let's sing a song, sing with me,

Shannon," and 11) "Tell me (e.g., what you are making)" (see Table 3-1, Social

Demands Used in Escape Condition of Functional Analysis).

Tangibles identified. The tangibles identified for Shannon via direct observation

are listed in Table 3-2. Of these tangibles, seven were selected. The seven tangibles

selected for use in the preference assessment (Phase 2) were: 1) T-Rex dinosaur, 2)

Barbie dolls, 3) play cash register, 4) small dinosaurs, 5) wooden blocks, 6) play-doh, and

7) Dr. Seuss book (see Table 3-3, Tangibles Included in Preference Assessment).

Colin Descriptive Assessment Results

Social demands identified. During direct observation, Colin and his peers were

observed to engage in a total of 12 peer-related verbal initiations and three peer-related

nonverbal (gestural) initiations (see Appendix C-3, Colin's Verbal and Gestural

Initiations). Of those initiations, four were selected as social demands and were used in

the escape condition during Phase 3. The four social demands selected for use in the

escape conditions of Phase 3, were: 1) "Look at me," 2) "Hand me the (e.g., toy)," 3)

"Take this (e.g., toy)," and 4) "Give me a high five" (see Table 3-1, Social Demands

Used in Escape Condition of Functional Analysis).









Tangibles identified. The tangibles identified for Colin via direct observation are

listed in Table 3-2. Of these tangibles, seven were selected. The seven tangibles selected

for use in the preference assessment (Phase 2), were: 1) drawing, 2) Magnadoodle, 3) Mr.

Potato Head, 4) puzzle, 5) play-doh, 6) wooden blocks, and 7) blocks on a strings (see

Table 3-3, Tangibles Included in Preference Assessment).

Table 3-1. Social Demands Used in Escape Condition of Functional Analysis
Shane Shannon Colin
Turn the page Pick a dinosaur Look at me
Press the button Show me what you are Hand me the (e.g., toy)
making
Look at me Hand me the (e.g., Take this (e.g., toy)
block)
Touch my shoulder Look at (e.g., me) Give me a high five
Hold my hand Put the (e.g., block)
here (where the peer points)
Give me a high five Help me make a snake with
(e.g., play-doh), you roll
this side (where the peer
points)
Put your hands on your lap
Let's play patty cake, come
sit by me
Say hi
Let's sing a song, sing with
me, Shannon
Tell me (e.g., what you
are making)

Table 3-2. Tangibles Identified During Descriptive Assessment
Shane Shannon Colin
Foam alphabet letters Pretend kitchenware (e.g., Computer game
food, plates, spoons)
Bean bag Pretend physician tools Listening to music
(e.g., blood pressure
monitor)
Plastic tools (e.g., pliers) Art activities (e.g., gluing, Dancing to music
painting, drawing on paper
with pens, crayons, and
colored and pencils)
Play-doh Cutting with scissors Hair rollers









Table 3-2. Continued
Shane
Play stethoscope

Yellow spoon
Wooden blocks

Small green dinosaur

Mr. Potato Head arm
Small yellow glasses

Clip


Small plastic dinosaurs

Plastic snake
Plastic dragon fly
Beads
Books
Yellow foam fish
Lego's


Shannon
Matching cards by color,
insects, and shape
Sorting by colors
Puzzles

Play-doh

Plastic necklaces
Plastic and wooden blocks

Plastic animals (e.g., turtle,
giraffe, lion, horses, small
dinosaurs, T-rex)
Musical instruments (e.g.,
maracas, tambourine)
Looking at a book
Barbie's


Colin
Blocks on strings

Puzzles
Drawing with markers and
crayons
Large and small plastic
blocks
Wooden blocks
Books (looking at and
listening)
Magnadoodle


Play-doh

Blue toy car
Mr. Potato Head
Plastic bottle
Water
Barbie doll
Plastic cookware and food
Big Lego's
Plastic tools
Play phone


Table 3-3. Tangibles Included in Preference Assessment
Shane Shannon Colin
Scooby-do book T-rex dinosaur Drawing
Foam fish Barbie dolls Magnadoodle
Thomas the Engine book Play cash register Mr. Potato Head
Plastic ball Small dinosaurs Puzzle
ABC Phonic Wooden blocks Play-doh
Bob the Builder book Play-doh Wooden blocks
Lego's Dr. Seuss book Blocks on strings

Phase Two: Preference Assessment Results

During Phase 2, preference for the seven tangibles selected for each participant

during the descriptive assessment was evaluated. Criteria for identifying a high preferred

tangible(s) included having the highest ranking (i.e., highest percentage of trials chosen).

If there was a tie between two tangibles for the highest ranking or if the two highest









rankings were within 10% of each other, both tangibles were considered high preferred.

Criteria for identifying neutral preferred tangible(s) included having a ranking between

second and fifth. Criteria for identifying low preferred tangible(s) included having a

ranking between fourth and seventh. Tangibles were not ranked if they did not meet the

above criteria (e.g., if there was more than a 10% difference between the tangible chosen

the first and second highest percentage of trials). The results of Phase 2 are presented

separately for each participant.

Shane Preference Assessment Results

Five preference assessment sessions were conducted with Shane. Shane's high

preferred tangibles were the Scooby-do book and the foam fish, chosen 42% and 33% of

the trials, respectively (see Table 3-4, Shane's Rank Order of Stimuli Obtained from

Preference Assessment). His neutral preferred tangibles were the Thomas the Engine

book, the plastic ball, and the ABC Phonics, chosen 24%, 20%, and 18% of trials,

respectively. Shane's low preferred tangibles were the Bob the Builder book and Lego's,

chosen 14% and 11% of the trials, respectively.

The high preference stimuli used in the tangible conditions of Shane's functional

analysis were the Scooby-do book and the foam fish. The Scooby-do book and foam fish

were rotated across tangible conditions so that only one high preferred tangible was

available in each tangible condition. The neutral tangibles available in Shane's free play,

attention, tangible, and escape conditions were the Thomas the Engine book and the ABC

Phonics. The plastic ball was identified as a neutral preferred tangible; however, each

time Shane chose the plastic ball during the preference assessment, he threw it out of the

assessment area. As a result of the distraction this caused to the ongoing classroom

activities, the plastic ball was not included in any of the functional analysis conditions.









Table 3-4. Shane's Rank Order of Stimuli Obtained from Preference Assessment
Stimuli Percentage of Rank Functional Analysis Conditions
Trials Chosen Stimuli Available
Scooby-do 42 High preferred Tangible (rotated)
book
Foam fish 33 High preferred Tangible (rotated)
Thomas the 24 Neutral Free play, attention, tangible,
Engine book preferred escape
Plastic ball 20 Neutral None
preferred
ABC Phonics 18 Neutral Free play, attention, tangible,
preferred escape
Bob the Builder 14 Low preferred None
book
Lego's 11 Low preferred None

Shannon Preference Assessment Results

Ten preference assessment sessions were conducted with Shannon; five sessions

were conducted in the first preference assessment and five sessions in the second

preference assessment.

First preference assessment results

In the first preference assessment, Shannon's high preferred tangible was the T-rex

dinosaur, chosen 83% of the trials (see Table 3-5, Shannon's Rank Order of Stimuli

Obtained from First Preference Assessment). Her neutral preferred tangibles were the

play cash register, small dinosaurs, and wooden blocks, chosen 28%, 23%, and 20% of

the trials, respectively. Shannon's low preferred tangibles were play-doh and the Dr.

Seuss book, chosen 18% and 16% of the trials, respectively. The tangible not ranked was

the Barbie dolls, chosen 56% of the trials.

In the first preference assessment, the high preference stimulus used in the tangible

conditions of Shannon's functional analysis was the T-rex dinosaur. The neutral tangibles

available in Shannon's free play, attention, tangible, and escape conditions were the small

dinosaurs and the wooden blocks. The play cash register, ranked high preferred, was not









included in the tangible conditions due to one of the two play cash registers breaking

prior to the beginning of the functional analysis (i.e., two play cash registers were needed

to conduct a tangible condition and only one was available).

The results of the first preference assessment were utilized in functional analysis

conditions 1 through 15. Toward the end of the first functional analysis there was concern

that the high preferred tangible was no longer high preferred. Therefore, a second

preference assessment was conducted prior to condition number 16.

Table 3-5. Shannon's Rank Order of Stimuli Obtained from First Preference Assessment
Stimuli Percentage of Rank Functional Analysis Conditions
Trials Chosen Stimuli Available
T-rex dinosaur 83 High preferred Tangible
Barbie dolls 56 Not ranked None
Play cash 28 Neutral None
register preferred
Small dinosaurs 23 Neutral Free play, attention, tangible,
preferred escape
Wooden blocks 20 Neutral Free play, attention, tangible,
preferred escape
Play-doh 18 Low preferred None
Dr. Seuss book 16 Low preferred None

Second preference assessment results

In the second preference assessment, Shannon's high preferred tangibles were the

play cash register and play-doh, both chosen 45% of the trials (see Table 3-6, Shannon's

Rank Order of Stimuli Obtained from Second Preference Assessment). Her neutral

preferred tangible was the Barbie dolls, chosen 29% of the trials. Shannon's low

preferred tangibles were the wooden blocks, the small dinosaurs, and the Dr. Seuss book,

all chosen 17% of the trials. The tangible not ranked was the T-rex dinosaur, chosen 42%

of the trials.









In the second preference assessment, the high preference stimuli used in the

tangible conditions of Shannon's functional analysis were the play cash register (the

second play cash register had been fixed and was available for use) and the play-doh. The

tangibles were rotated across conditions so that only one high preferred tangible was

available in each tangible condition. The neutral tangible available in Shannon's free

play, attention, tangible, and escape conditions was the Barbie dolls.

Table 3-6. Shannon's Rank Order of Stimuli Obtained from Second Preference
Assessment
Stimuli Percentage of Rank Functional Analysis Conditions
Trials Chosen Stimuli Available
Play cash 45 High preferred Tangible (rotated)
register
Play-doh 45 High preferred Tangible (rotated)
T-rex dinosaur 42 Not ranked None
Barbie dolls 29 Neutral Free play, attention, tangible,
preferred escape
Wooden blocks 17 Low preferred None
Small dinosaurs 17 Low preferred None
Dr. Seuss book 17 Low preferred None

Colin Preference Assessment Results

Five preference assessment sessions were conducted with Colin. Colin's high

preferred tangible was drawing, chosen 71% of the trials (see Table 3-7, Colin's Rank

Order of Stimuli Obtained from Preference Assessment). His neutral preferred tangibles

were the Magnadoodle and Mr. Potato Head, chosen 29% and 28% of the trials,

respectively. Colin's low preferred tangibles were the puzzle, play-doh, wooden blocks,

and blocks on strings, chosen 17%, 14%, 12%, and 9% of the trials, respectively.

The high preference stimulus used in the tangible conditions of Colin's functional

analysis was drawing. The neutral tangibles available in Colin's free play, attention,

tangible, and escape conditions were the Magnadoodle and Mr. Potato Head.









Table 3-7. Colin's Rank Order of Stimuli Obtained from Preference Assessment
Stimuli Percentage of Rank Functional Analysis
Trials Chosen Conditions Stimuli
Available
Drawing 71 High preferred Tangible
Magnadoodle 29 Neutral preferred Free play, attention,
tangible, escape
Mr. Potato Head 28 Neutral preferred Free play, attention,
tangible, escape
Puzzle 17 Low preferred None
Play-doh 14 Low preferred None
Wooden blocks 12 Low preferred None
Blocks on strings 9 Low preferred None

Phase Three: Functional Analysis Results

Functional analyses of peer-related withdrawn and peer-related negative social

behavior were conducted with each participant incorporating social demands in the

escape condition (see Phase 1 descriptive assessment results). The tangible(s) identified

as high preferred, via the preference assessment, were incorporated into the tangible

condition (see Phase 2 preference assessment results). The tangibles identified as

neutrally preferred via the preference assessment (with the exception of the plastic ball in

Shane's preference assessment and play cash register in Shannon's first preference

assessment), were incorporated into the free play, attention, escape, and tangible

conditions. Data were graphically displayed and analyzed for each participant. Peer-

related withdrawn and peer-related negative social behaviors were graphically displayed

using two different methods for each participant. First, graphs were created for each

behavior displaying the rate of that behavior per minute in each functional analysis

condition. Second, the rate per minute of peer-related withdrawn and peer-related

negative social behaviors were combined post hoc and graphed for each functional

analysis condition. Throughout each 5 minute functional analysis condition, participants









had the opportunity to engage in peer-related withdrawn and peer-related negative social

behaviors, with each behavior resulting in the same consequence in the given condition

(e.g., in the attention condition, attention was provided contingent on both peer-related

withdrawn and peer-related negative social behavior). At some point in the functional

analysis conditions, 2 of the 3 participant's (i.e., Shane and Colin) engaged in both social

behaviors (peer-related withdrawn and peer-related negative) during the same 5 minute

condition. Therefore, the behaviors were combined post-hoc to display the total rate per

minute of combined peer-related (withdrawn and negative) social behaviors displayed for

each functional analysis condition. These data addressed the first and second research

questions "Can the functional analysis methodology of evaluating problem behavior be

used to identify the function of peer-related withdrawn and peer-related negative social

behavior?" and "Are the functions of peer-related withdrawn and peer-related negative

social behaviors different across young children with ASD?" Tables were generated to

present the mean rate per minute and ranges of each peer-related social behavior across

functional analysis conditions and to provide a comparison between participants.

Next, peer integrity behaviors were analyzed. Tables were generated to present the

mean rate per minute and ranges of peer correct presentation of social demands in the

escape condition, average percentage and ranges of correct peer delivery of consequences

in the tangible, attention, and escape conditions, and mean rate per minute and ranges of

peer correct delivery of positive initiations in the free play condition. These data

addressed the third research question "Can peers implement functional analysis

procedures with acceptable integrity?"









Finally, tables were generated to present the mean rate per minute and ranges of

participant peer-related positive initiations, average percentage and ranges of participant

peer-related positive responses, and average percentage and ranges of participant peer-

related no response social behaviors. These data addressed the final research aim "To

examine rate per minute of participant peer-related positive initiations and percentage of

participant peer-related positive and no response social behavior across conditions." The

results from Phase 3 are presented separately for each participant.

Shane Functional Analysis Results

Shane peer-related withdrawn and peer-related negative behaviors

The rate per minute of peer-related withdrawn social behavior (top panel), peer-

related negative social behavior (middle panel), and the combination of peer-related

withdrawn and negative social behaviors (bottom panel) from Shane's functional analysis

are displayed in Figure 3-1. The mean rate per minute and ranges of peer-related

withdrawn, peer-related negative, and the combination of peer-related withdrawn and

negative social behaviors are displayed in Table 3-8. Shane exhibited a three point

increasing trend of peer-related withdrawn social behavior in the ignore and tangible

conditions. The ignore and tangible conditions ended on a three point upward trend,

though peer-related withdrawn social behavior was variable throughout the tangible

condition, with 4 of 10 conditions with zero rates per minute of withdrawn social

behavior. The attention condition also began on an increasing trend, though ended on a

three point downward trend. Peer-related withdrawn social behavior was exhibited in 3 of

the 4 escape conditions, though the rates per minute were low (0.2/minute to 0.4/minute)

and a stable three point trend (i.e., upward/downward/stable) was not established. With

the exception of one condition, peer-related withdrawn social behavior occurred at zero









rates per minute in the free play condition. Shane's peer-related withdrawn social

behavior was variable and overlapped between all functional analysis conditions. Visual

analysis suggests that Shane's peer-related withdrawn social behavior (Figure 3-1, top

panel) maybe multiply maintained by automatic reinforcement and positive

reinforcement in the form of access to preferred tangibles, though due to variable rates of

peer-related withdrawn social behavior across all conditions, the results may best be

described as inconclusive.

Shane exhibited a three point upward/stable trend of peer-related negative social

behavior in the ignore and tangible conditions, though the ignore condition ended on a

downward point (i.e., zero rates per minute of peer-related negative social behavior were

displayed in the final ignore condition). The tangible condition ended on a three point

upward/stable trend, though peer-related negative social behavior occurred at low to zero

rates per minute in the first seven tangible conditions, with 6 of 10 conditions with zero

rates per minute of negative social behavior. Peer-related negative social behavior was

displayed in 3 of 4 escape conditions, though the first three data points established a

downward trend. In the final escape condition, peer-related negative social behavior

increased to 0.4/minute, though a three point upward trend was not established and there

was overlap between ignore, attention, and tangible conditions. Low to zero rates per

minute of peer-related negative social behavior were observed in the attention and free

play conditions. Visual analysis suggests that Shane's peer-related negative social

behavior (Figure 3-1, middle panel) maybe maintained by positive reinforcement in the

form of access to preferred tangibles. However, due to variable rates of peer-related

negative social behavior in the ignore and escape conditions and low and variable rates of









peer-related negative social behavior in the tangible condition, the results may best be

described as inconclusive.

Shane combined peer-related withdrawn and peer-related negative behaviors

As noted previously, the opportunity to engage in both peer-related withdrawn and

peer-related negative social behaviors was available across all functional analysis

conditions and during the escape, tangible, and attention conditions, reinforcement was

delivered contingent on the occurrence of either peer-related withdrawn or peer-related

negative social behavior. For example, in tangible condition number 23, Shane engaged

in one instance of peer-related withdrawn social behavior and one instance of peer-related

negative social behavior. The occurrence of peer-related withdrawn social behavior was

reinforced with 15 seconds access to his high preferred tangible and the occurrence of

peer-related negative social behavior was reinforced with 15 seconds access to his high

preferred tangible, for a total of two reinforcement deliveries in that tangible condition.

Visual analysis of Shane's peer-related withdrawn (Figure 3-1, top panel) and peer-

related negative social behavior (Figure 3-1, middle panel) indicated both peer-related

withdrawn and peer-related negative social behaviors occurred in the same functional

analysis conditions. Specifically, when peer-related withdrawn or peer-related negative

social behavior occurred, both occurred in 100% (3 of 3) escape (i.e., peer-related

withdrawn social behavior and peer-related negative social behavior occurred in escape

condition numbers 9, 16, and 24; no peer-related withdrawn or negative social behavior

occurred in escape condition number 22), 67% (4 of 6) tangible, and 50% (2 of 4) of

ignore conditions. Therefore, when peer-related withdrawn (or peer-related negative)

social behavior occurred in the escape condition, peer-related negative (or peer-related

withdrawn) social behavior also occurred in 100% of escape conditions. Similarly, when









peer-related withdrawn (or peer-related negative) social behavior occurred peer-related

negative (or peer-related withdrawn) social behavior also occurred in 67% of tangible

conditions and 50% of ignore conditions. Peer-related withdrawn and peer-related

negative social behaviors did not occur in the same condition for attention and free play

conditions.

Shane exhibited a three point upward trend of combined peer-related (withdrawn

and negative) social behavior in the ignore and tangible conditions. The ignore and

tangible conditions ended on a three point upward trend, though combined peer-related

(withdrawn and negative) social behavior was variable throughout the tangible condition,

with 4 of 10 conditions with zero rates per minute of combined peer-related (withdrawn

and negative) social behavior. Combined peer-related (withdrawn and negative) social

behavior was initially elevated in the attention condition, though ended on a decreasing

and variable trend. Combined peer-related (withdrawn and negative) social behavior was

exhibited in 3 of the 4 escape conditions, though the rates per minute were variable

(2.4/minute to 0.6/minute) and a stable three point trend was not established. With the

exception of one condition, combined peer-related (withdrawn and negative) social

behavior occurred at zero rates per minute in the free play condition. Visual analysis

suggests that Shane's combined peer-related (withdrawn and negative) social behavior

(Figure 3-1, bottom panel) maybe multiply maintained by automatic reinforcement and

positive reinforcement in the form of access to tangibles. However, due to variable rates

of combined peer-related (withdrawn and negative) social behavior in the tangible,

attention, and escape conditions, the results may best be described as inconclusive.











-I- Attention
-- Tangible
-0- Ignore


-U- Escape
-- Free Play


2.5
-e
2

1.5

1

S0.5

0


2.4

S 2

S1.6

. 1.2

S0.8

0.4

0


1 3 5 7 9 11 13 15 17 19 21 23 25 27


1 3 5 7 9 11 13 15 17 19 21 23 25 27
Sessions


Figure 3-1. Rate per minute peer-related withdrawn (top panel), negative (middle panel),
and combined social behavior (bottom panel) for Shane.


1 3 5 7 9 11 13 15 17 19 21 23 25 27


2.5

2

1.5
g Z 1


a 0.5

S 0