|USFDC Home | USF Electronic Theses and Dissertations||| RSS|
This item is only available as the following downloads:
Effectiveness of Social Story Interventions for Children with Asperger's Syndrome by Frank J. Sansosti A thesis submitted in partial fulfillment of the requirements for the degree of Education Specialist Department of Psychological and Social Foundations College of Education University of South Florida Major Professor: Kelly A. Powell-Smith, Ph.D. Member: Jonathan Greenstein, Ph.D. Member: John Ferron, Ph.D. Date of Approval: July 14, 2003 Keywords: autism spectrum disorders, social skills training, communication, peer interaction, visual supports Copyright 2003 Frank J. Sansosti
Acknowledgments I would like to express my appreciation to the many people who contributed to the completion of this project. I am extremely gr ateful to my major professor, Dr. Kelly A. Powell-Smith, for her patience, guidance, s upport, and countless hours of reviewing this document. I am also grateful to the member s of my committee, Dr. Jonathan Greenstein and Dr. John Ferron, for their support and addi tional guidance. I would also like to extend a special thanks to the students from the School Psychology Program at the University of South Florida who volunteered to assist with the data collection for this project. In addition, I am deeply thankful to the en tire staff at the Gray Center for Social Learning and Understanding for their funding of this research. Specifically, I would like to thank Bruce Mills for his cooperation and assistance with this project. Likewise, I thank Dr. Joseph McAllister and Dr. Karleen Pr eator who provided the vision for such a project, as well as countless hours of support and friendship over the years.
i Table of Contents LIST OF TABLES v LIST OF FIGURES vii ABSTRACT viii CHAPTER I: INTRODUCTION 1 Rationale 7 Purpose 8 Research Hypotheses 9 Definitions 9 Autism 9 Aspergers Syndrome 9 Pervasive Developmental Disorder 10 Higher-Functioning Autism 10 Theory of Mind 11 Social Engagement 11 CHAPTER II: REVIEW OF LITERATURE 12 Overview 12 History of Aspergers Syndrome 12 Contemporary Descriptions of Aspergers Syndrome 15 Social Impairments in Individuals with Aspergers Syndrome 17 Theory of Mind 19 Aspergers Syndrome in the Schools 20 Preschool 22 Elementary School Years 24 Middle and High School Year s 26 Directions for Intervention 28 Social Stories 28 Rationale Behind Social Stories 29 Developing a Social Story Intervention 30 Guidelines for Writing Social Stories 35 Guidelines for Successful Implementation of Social Stories 39 Effectiveness of Social Stories 41 Summary 53 Purpose of the Study 55
ii CHAPTER III: RESEARCH METHODS 56 Overview 56 Subjects 56 Darius 57 Francis 58 Angelo 58 Settings 59 Materials and Measures 60 Social Story Book 60 Observational Data Forms 61 Social Interaction Recording System 62 Observation of Appropriate Social Intera ction Skills 62 Pre-, Post-Test Measures 63 Social Skills Rating SystemParent Form, Elementary Level 63 Social Skills Rating SystemTeacher Form, Elementary Level 64 Childrens Atypical Development Scale 66 Social Story Journal 68 Dependent Measures 68 Joining In 69 Active Joining In (AJI) 69 Passive Joining In (PJI) 69 Absence of Joining In (AbJI) 70 Sportsmanship 70 Active Sportsmanship (ASP) 70 Passive Sportsmanship (PSP) 71 Absence of Sportsmanship (AbSP) 71 Procedure 72 Observer Training 72 Baseline Period 73 Intervention Phase 75 Follow-Up Phase 77 Inter-Rater Reliability 77 Data Analyses 78 Observational Data 78 Pre-, Post-Tests Data 81 CHAPTER IV: RESULTS 82 Overview 82 Participants Social Communi cation Progress: Observational Data 82 Active Display of Target Behavior 82 Darius 84 Francis 86 Angelo 89 Passive Display of Target Behavior 91 Darius 93
iii Francis 94 Angelo 96 Absence of Target Behavior 98 Darius 100 Francis 101 Angelo 103 OASIS Data 105 Participants Social Communication Progr ess: Peer Comparison 106 Active Comparison 108 Darius 108 Francis 108 Angelo 109 Passive Comparison 112 Darius 112 Francis 112 Angelo 113 Absence Comparison 115 Darius 115 Francis 115 Angelo 116 Participants Social Communication Progr ess: Follow-Up Data 117 Participants Social Communi cation Progress: Pre, Post-t est Data 118 Social Skills Rating System 118 Darius 118 Francis 118 Angelo 119 Childrens Atypical Development Scale 119 Darius 119 Francis 119 Angelo 120 Social Story Journal 120 Darius 120 Francis 120 Angelo 121 CHAPTER V: DISCUSSION 122 Overview 122 Impact of the Intervention 122 Observational Data 122 Preand Post-Test Data 128 Significance of Findings 129 Implications for Practice 132 Limitations 133 Recommendations for Future Research 135 Summary 138
iv REFERENCES 139 APPENDICES 149 Appendix A: DSM-IV Criteria for Au tistic Disorder 150 Appendix B: DSM-IV Criteria for Asperger Disorder 151 Appendix C: Examples of Social Stories 152 Appendix D: Social Story Information Form 153 Appendix E: Social Story Implementation Plan 155 Appendix F: Social Story Report Form 157 Appendix G: Social Story for Darius 158 Appendix H: Social Story for Francis 159 Appendix I: Social Story for Angelo 160 Appendix J: SIRS: Social Interacti on Recording System 151 Appendix K: OASIS: Observation of Appropr iate Social Interac tion Skills 163 Appendix L: Social Story Journal 164 Appendix M: Joining In Response Definitions 165 Appendix N: Sportsmanship Response Definitions 167 Appendix O: SSRS Pre/Post-Test Ratings 169 Appendix P: CADS Pre/Post-Test Ratings 170
v List of Tables Table 1: Steps in the Development of Social Stories 32 Table 2: Steps in the Development and Implem entation of Social Stories 33 Table 3: Social Story Sentences and Thei r Identified Purpose 35 Table 4: Steps in the Social-Skills Training Model 42 Table 5: Descriptive Statistics for Active S portsmanship for Darius 85 Table 6: Slope and Level Change for Active Sportsmanship for Darius 86 Table 7: Descriptive Statistics for Active Joining In for Francis 88 Table 8: Slope and Level Change for Active Joining In for Francis 88 Table 9: Descriptive Statistics for Active Joining In for Angelo 89 Table 10: Slope and Level Change for Active Joining In for Angelo 90 Table 11: Descriptive Statistics for Passive Sportsmanship for Darius 93 Table 12: Slope and Level Change for Passive Sportsmanship for Darius 94 Table 13: Descriptive Statistics for Passive Joining In for Francis 95 Table 14: Slope and Level Change for Passive Joining In for Francis 96 Table 15: Descriptive Statistics for Passive Joining In for Angelo 97 Table 16: Slope and Level Change for Passive Joining In for Angelo 98 Table 17: Descriptive Statistics for Absence of Sportsmanship for Darius 100 Table 18: Slope and Level Change for Absence of Sportsmanship for Darius 101 Table 19: Descriptive Statistics for Absence of Joining In for Francis 102 Table 20: Slope and Level Change for Absence of Joining In for Francis 103 Table 21: Descriptive Statistics for Absence of Joining In for Angelo 104 Table 22: Slope and Level Change for Absence of Joining In for Angelo 104 Table 23: OASIS Descriptive Data for Each Participant 105 Table 24: Mean Level of Active Sportsmanship for Darius with Peer Comparison 108 Table 25: Mean Level of Active Joining In for Francis with Peer Comparison 109 Table 26: Mean Level of Activ e Joining In for Angelo with Peer Comparison 110 Table 27: Mean Level of Passi ve Sportsmanship for Darius with Peer Comparison 112 Table 28: Mean Level of Passive Joining In for Francis with Peer Comparison 113 Table 29: Mean Level of Passi ve Joining In for Angelo with Peer Comparison 113 Table 30: Mean Level of Absence of Sportsmanship for Darius with Peer Comparison 115
vi Table 31: Mean Level of Absen ce of Joining In for Francis with Peer Comparison 116 Table 32: Mean Level of Absen ce of Joining In for Angelo with Peer Comparison 116
vii List of Figures Figure 1: Percent of Active Displays of Target Behavior Across Particip ants 84 Figure 2: Percent of Passive Di splays of Target Behavior Acro ss Participants 92 Figure 3: Percent of Absence of Target Beha vior Across Participants 99 Figure 4: Percent of Active Displays of Target Behaviors Across Participants with Peer Comparison 107 Figure 5: Percent of Passive Displays of Target Behaviors Across Participants with Peer Comparison 111 Figure 6: Percent of Absence of Ta rget Behaviors Across Participants with Peer Comparison 114
viii Effectiveness of Social Story Interventions for Children with Aspergers Syndrome Frank J. Sansosti ABSTRACT The purpose of this study was to investig ate the effects of individualized social story interventions on the soci al communication skills of thre e children with Aspergers Syndrome. Using a multiple-baseline across participants design, social stories were implemented and direct observations of the part icipants identified target behaviors were collected three times per week during unstructured school acti vities (e.g., recess). Data revealed an increase in the social communica tion skills of two of the three participants when the treatment was implemented. In addition, maintenance of treatment effects was observed in two participants. These data support recommendations for using social stories to teach social skills to children diagnosed with autism spectrum disorders.
1 Chapter I Introduction Aspergers Syndrome (AS) is a relatively new category of developmental disorder that has come into more genera l use during the last twenty ye ars despite its shared history with autism. AS is a term currently used to describe the mildes t and highest functioning end of autism spectrum disorders. Like autism, AS is a Pervasive Developmental Disorder (PDD) characterized by deficits in three broad aspects of development: (a) social interaction and use of social sk ills; (b) use of language for communicative purposes; and (c) repetitive and/or perseverat ive features and/or limited, but intense, range of special interests (Bauer, 1996). In addition to these clinical similarities, researchers, clinicians, and parents have suggested additi onal considerations regarding the presentation of children with AS. Speci fically, many have noted that children with AS possess higher cognitive abilities (oft en academically high-functioning), normal language function compared to other PDDs, and motor impairments and/or clumsiness (e.g., Bauer, 1996; Rietschel, 2000; Shery, 2000; Volkmar & Klin, 2000; Wallace, 2000; & Wing, 2000). Further, children with AS ar e often viewed as having an awareness of and need to actively seek friendships, an obser vation not generally seen in children with autism. Characteristic of all PDDs, AS shares a common deficit in social abilities that remain the greatest life challenge for these i ndividuals. In a review of research, Wing
2 (1988) categorized three genera l types of social impairments frequently observed in individuals with autism and PDD: (a) impa ired social recogniti on (i.e., apathy toward social interaction), (b) impaired social communication (i.e., expressive and receptive language deficits), and (c) impairment of social imagination and understanding (i.e., perspective taking). Specifically, children with AS often do not interact with peers, possess poor appreciation of social cues, and often make socia lly and emotionally inappropriate responses (e.g., laughing loudl y when another student gets hurt). To better understand such patterns, so me investigations have turned to understanding the capacity and nature for social cognition, which is ch aracteristic of all human social interactions (Mundy & Stella, 2001 ). Such social cognition is currently believed to follow a developmental progression and refers to the ability to infer the mental states of others (e.g., knowledge, perspec tives, beliefs, intenti ons) from behavior. This hypothesized component is what Leslie (1987) referred to as Theory of Mind (ToM). ToM is largely governed by indirect, socially medi ated cues that occur at a subconscious level. However, children with autism spectrum disorders are often blind to these social cues. Such mindblindness or the inability to mental ly depict the behavior of others, does not allow a child with AS to distinguish his/her ow n thoughts or feelings from the thoughts and feelings of othe rs (Baron-Cohen, 1995; Leslie, 1987, 1993). Following such logic, individuals with AS will have difficulty understanding social information, thereby compromising his/her ab ility to engage in reciprocal social interactions. Despite the unique profile of severe deficits in social and communication abilities in the presence of cognitive and language st rengths (Klin & Volmar, 2000), children with
3 AS are often considered to be functioning normally but with odd behaviors (Wallace, 2000). Often educators identify children with AS as too bright or too verbal (Klin & Volmar, 2000) and associate poor performance on nonverbal tasks to negativism or other intentional behaviors (Safran, 2001). As a result, these children may be viewed as behaviorally disordered and exhibiting conditions similar to Attention Deficit Hyperactivity Disorder (ADHD) (Volkmar, Klin, Schultz, Rubin, & Bronen, 2000). In fact, Church, Alisanski, and Amanullah (2000) found that the most fr equent diagnosis or educational label was ADHD (20%) (p. 14). In addition, children with AS are viewed by their peers as eccentric and peculiar. Th e inappropriate social skills and obsessive interests in obscure subjects of the child with AS only cause them to be victims of continued ridicule and further alienation, desp ite their attempts for friendships (Williams, 1995). Such ridicule and torment, combined w ith the inability to cope with change may also cause children with AS to be easily stressed and emotionally vulnerable. Combing the knowledge of current diagnosti c criteria, clinical observations, and recent developments in theory, Bauer (1996) provided a clinical outline regarding the developmental course of children with AS. Such an outline augmented Aspergers (1944) original description of symptoms while aligning the empirical progress in diagnostic trends since Wings (1988) influentia l review. However, due to the lack of empirical literature regarding the developmen tal experiences of children with AS, Bauer could only provide hypotheses regarding the developmental progression of AS. Church, Alinsanski, and Amanullah (2000) later conf irmed many of Bauers original hypotheses through descriptive analysis.
4 Both the theory laden hypotheses of Baue r (1996) and the descri ptive analysis of Church, Alisanski, and Amanullah (2000) illust rate that although soci al skill abilities are highly variable in individuals with AS, they remain the greatest challenge to overcome. Specifically, children with AS desperately want to fit in, but lack the skills and social knowledge necessary to do so (Bauer, 1996; Church, Alisanski, and Amanullah, 2000). In both reviews, the developmental experien ces can be better understood by the following themes: (a) during primary school years, soci al skill deficits become more prominent and observable by parents and teachers; (b) as children with AS enter middle school, they may begin feeling their differences and gene rally express heightened sadness, anxiety, and rejection; and (c) teenager s with AS face increased anxiet y and ever-increasing social skill deficiencies. Since individuals with AS look perfectly typical, their atypical social behaviors and odd use of language are rare ly understood by those around them. Instead, these individuals are often identified as odd or different (Bauer, 1996; Church, Alisanski, & Amanullah, 2000; Klin & Volk mar, 2000; Volkmar, et. al., & Wallace, 2000), and these children struggle to fit in. Taken together, these reviews provide edu cators with the information necessary to build interventions that are more comprehensiv e and address the critical deficit areas. Despite this burgeoning information base of ch ildren with AS in sc hools, little research has been conducted on the effectiveness of in terventions to determine whether and how specific strategies to encourage social integr ation with children with AS are successful. Nevertheless, several suggestions have been provided for practitioners and researchers that offer useful starting points (Atwood, 2000) Many of these suggestions are offered more for their heuristic value rather than as definitive strategies (e.g., teaching theory of
5 mind skills, encouraging friendship skills). Fu rther, recent reviews have suggested more direct interventions aimed at facilitating socialization in children with autism spectrum disorders (e.g., social skills training, structured teaching, pe er-mediated approaches, peertutoring, use of social stories and comi c strip conversations) (Atwood, 2000; Rogers, 2000). A strategy that is increasi ng in popularity within school s is the use of social stories (Gray, 1998). A social stor y describes social situations in terms of relevant social cues and often defines appropria te responses. The specific st rategy is designed to provide the individual with the necessary perspect ives and skills to successfully manage a particularly troublesome social situation. The norms for behavior in the targeted context, the perspective(s) of others, and the specific steps to implementing th e appropriate social skill(s) are instructed and modeled through a short visual and writt en story regarding specific content. Specifically, a social stor y is written to provide information on what people in a given situation are doing, thinki ng or feeling, the sequence of events, the identification of significant social cues and th eir meaning, and the scri pt of what to do or say; in other words, the what, when, who a nd why aspects of social situations (Atwood, 2000, p. 90). The social stories approach be gan in 1991, and was formally introduced to the field of special education in 1993 with sp ecific rules and procedures for development and implementation (Gray, 1994; Gray & Garrand, 1993). Grays social stories interventions, in th eory, are supported by the research that children with autism spectrum disorders suffer from an inability to establish ToM skills, and rely heavily on rule-based, governed behavior (Atwood, 2000; Baron-Cohen, 1995; Gray & Garand, 1993; Leslie, 1987, 1993; Mund y & Stella, 2001; Twachtman-Cullen,
6 1998). However, traditional teaching invol ves interaction between the teacher and student, creating a social situ ation whenever instruction oc curs (Gray & Garrand, 1993). Considering that children with AS will fail to interpret the social cues accurately, the lesson may hold little relevant meaning (Frith, 1989). Therefore, the use of social stories to increase awareness and understanding of the what, when, who, and why may prove to be more beneficial since the child is using the story as a script for further social interactions. Unfortunately, a review of the availabl e published research on social stories yielded only six empirical studies relating to the eff ectiveness of social story interventions used with children with autism. The effectiveness of utilizing social stories to teach individuals with autism have been de monstrated on a wide range of social skills, such as how to monitor volume when singi ng with others (Fulle rton, Stratton, Coyne, & Gray, 1996), how to greet people appropriately and share toys (Swaggart, et. al., 1995), how to reduce tantrum behavi or (Kuttler, Myles, & Carl son, 1998; Lorimer, Simpson, Myles, & Ganz, 2002), improving social intera ctions that occurred during lunch (Norris & Dattilo, 1999), and how to learn new routin es, and to respond to changes in routines (Gray & Garrand, 1993). In each of the studies that have used social stories as an intervention, positive trends in data were observed. However, it is necessary to point out that in all of these studies, the primary participan t(s) carried a primary diagnos is of autism and/or other comorbid disorders. In addition, in most of these studies the topic of the intervention was to increase prosocial behaviors (e.g., social awareness, positive social intergration skills). However, their focus was to reduce pervasiv e problem behaviors (e.g., tantrum behaviors,
7 spitting, yelling) in participant(s). Theref ore, examining the effectiveness of social stories as a positive support in a variety of settings is a relatively new endeavor. Further, it is likely that the increased intellectual, language, social awareness, and behavioral characteristics that typify children with AS will allow for increased outcomes of social story interventions as a means of teaching prosocial behaviors. Although the research regardi ng social skills interventions for children with autism has been well documented, very little lite rature is available regarding the efficacy of interventions for purely AS samples. Notably, many researchers have begun suggesting strategies for improving the soci al integration of children with AS (e.g., Attwood, 2000; Rogers, 2000; Safran, 2001). Of the many suggestions that are presented as being efficacious in peer-reviewed journals and professional conferences (i.e., social skills training, social stories, structured t eaching), none have been empirically validated with children diagnosed with AS. Therefor e, it is imperative that research begins focusing on demonstrating that such sugge stions are evidence-based approaches. Rationale Given that the clinical and diagnostic pa ttern of dysfunction fo r children with AS appears to be less pervasive th an that of classic autism, AS may be more common than autism, and perhaps more amenable to inte rvention. In fact, recent prevalence rates suggest that AS occurs at rates as high as 63 per 10,000 births, as compared to rates of 21 to 31 for autism (Hyman, Rodier, & Davids on, 2001). Elevated prevalence rates suggest a large unserved student population, making AS an auspicious challenge for special education.
8 The most appropriate method to incor porate social skills training in the classrooms for children with AS has received little attention in c linical literature. Although there has been a growi ng body of research on the treat ment efficacy of higherfunctioning individuals with auti sm that offers recommendations for educators, there is a deficit of studies that systematically addre ss the efficacy of social skills interventions with purely AS samples. Furthermore, the cu rrent increase in the number of referrals for special educational services for these children necessitates an immediate response by the research community to conduct efficacy studies related to special educational practices. Given such urgency, it may be impractical to wait for la rge comparative studies. A reasonable alternative is to re ly on a combination of integr ative approaches and singlesubject design studies to ma ke crucial educational deci sions. Therefore, a small comparative study of a currently recommended social skills intervention is desirable because it would fill the gap in the research literature as well as contribute knowledge in the area of program design and interventi on implementation for children with AS. Purpose Further research on the efficacy of using social stories designed to train social skills to individuals with AS is necessary. This study will examine the effectiveness of the use of a social stories intervention designe d to increase identifie d target behaviors in three children with AS as measured by direct observations and preposttest data. The present study serves to e xpand the current body of res earch in this area by (a) incorporating specific (i.e., fre quency of appropriate behavio r(s)) measures of treatment efficacy, (b) programming and measuring for gene ralization of treatment effects, and (c) employing a means of experimental control.
9 Research Hypotheses The following hypotheses will be examined:Partici pants in the social story intervention will show a significant increase in the occurrence of identified target behavior(s) in unstructured natural settings followi ng treatment, as compared to baseline condition. 2. Participants in the social story interventi on will demonstrate clinically significant increases in social intera ctions/social communication. 3. Participants in the social story interv ention will maintain positive treatment effects at a 2-week follow-up. 4. Participants will demonstrate improved soci al behaviors beyond identified target behaviors as rated by pare nts and teachers on pre and post-test measures. 5. Participants will demonstrate a decrease in unusual behavior patterns as rated by parents and teachers on pre and post-test measures. Definitions Autism The current definition of autism is marked by the presence of three categories of behavioral impairments: (a) a qualitative impairme nt of reciprocal soci al interaction; (b) a qualitative impairment in the development of language and communication; and (c) a restricted range of activitie s or interests (American Ps ychiatric Association, 1994). Appendix A provides the diagnos tic criteria for autism. Asperger Syndrome (AS) AS (also called Asperger Disorder) is a te rm generally used to describe children who exhibit severe and sustained impairment in social interaction and who display
10 restricted, repetitive patterns of behavior, interests, or act ivities (American Psychiatric Association, 1994). These impairments are sim ilar to, but usually mild er than, those seen in autism. However, such disturbances must cause significant distur bance in the childs social/occupational functioning (American Psyc hiatric Association, 1994 ). In addition, children with AS generally have normal to high levels of intelligence, exhibit no clinically significant genera l language delay, and possess poor motor abilities. See Appendix B for diagnostic criteria. Pervasive Developmental Disorder (PDD) PDDs are a broad diagnostic category th at describe a syndrome of behaviors characterized by severe and pe rvasive behavioral impairment s in three general areas of development: (a) social interaction skills; (b ) communication skills; and (c) restricted range of activities or interests (American Psychiatric Association, 1994; Bauer, 1995). Under this broad category are Autisti c Disorder, Retts Disorder, Childhood Disintegrative Disorder, Aspergers Disord er, and PDD-Not Otherwise Specified (PDDNOS). According to the Diagnostic and Stat istical Manual For Mental Disorders-Fourth Edition (DSM-IV), PDD is typically first evident during the firs t five years of life and is often associated with some le vel of cognitive impairment se parate from deficits noted above. Higher-Functioning Autism (HFA) There are currently no explicit diagnostic guidelines for HFA, and it remains a controversy whether AS and HFA actually di ffer or only differ by severity. Currently, areas of controversy appear to be that: (a) deficits in motor skills only appear in children with AS; (b) language is impaired in children with HFA, but spared for children with AS;
11 and (c) children with HFA and AS perform c ognitively at different levels (Gillberg & Ehlers, 1998; Kugler, 1998; Miller & Ozonoff, 2000). Theory of Mind (ToM) ToM is the ability to appreciate and mentali ze that other people have mental states based on cues from external behavior and th e circumstances of th at behavior (BaronCohen, 1995; Twachtman-Cullen, 2000). ToM is necessary for interpreting human behavior and offers a mental representation of reality. Without such a representational view of the world, communication and language development are affected. Social engagement Social integration is a broad category that can be used to define social interactions. Interactions include identifying so cial cues (visual or auditory), reciprocity in conversations, and sharing enjoyment, intere sts, or achievements with other people. Other characteristics include using eye contact, facial e xpressions, and body posture and gesture to regulate and unde rstand social interactions.
12 Chapter II Review of Literature Overview Chapter II is dedicated to a review of the literature relevant to this study and covers the history and charac teristics of children with As pergers Syndrome (AS), how AS manifests in school-age children, recommended ave nues for social skills interventions, a description of social stories, the clinical e ffectiveness of social stories, and the predicted effects of the use of social stories with children with AS. History of Aspergers Syndrome Although officially recognized in the Diagnostic and Statisti cal Manual of Mental Disorders fourth edition (DSM-IV; American Psychi atric Association, 1994) for the first time in 1994, AS has a history nearly as l ong as autism (Volkmar, Klin, Schultz, Rubin, & Bronen, 2000). Prior to the inclusion of AS in the DSM-IV, much of what was known relating to disorders of seve re social withdrawal in ch ildren, was understood through the description of autism. It has not been until recent years that AS has received heightened attention in English literature, and has now become a battleground for researchers and other mental health professionals to incr ease our understanding a nd awareness of its developmental course, as well as study eff ective interventions aimed at decreasing the symptoms of such a peculiar disorder. In 1943, Leo Kanner, a Baltimore child psyc hiatrist at Johns Hopkins, made his historic publication describi ng 11 young children with seve re social, language, and
13 cognitive deficits. Kanner emphasized the solitariness of these children, despite the presence of others, as a pervas ive lack of interest in soci al relationships, including their parents. In addition to thei r aloofness, he described these children as (a) being caught up in restricted, repetitive, st ereotypic movements, noises, and routines; (b) having an inability to adapt to novel environments and insist on sameness; and (c) displaying language characterized by echolalia (the ec hoing of others speech), pronoun reversal, and literalness, if language developed at all (Klin & Volkmar, 1999). Captivated by the childrens all-absorbing fascination w ith the inanimate environment, Kanner conceptualized that these children suffered from a possible congen ital disturbance that affected the childs capacity to relate emotionally to ot hers (Klin & Volkmar, 1999). According to Kanner (1943), such a disturba nce resulted in social withdrawal and aloneness, or, infantile autism Only one year after Kanners publication, Hans Asperger, a Vie nnese pediatrician specializing in remedial e ducation, published his indepe ndent description of boys suffering from severe social isolation, despite having what appeared to be good language and cognitive skills. Much like Kanners de scription, Asperger identified these children as having social and comm unication problems that made it difficult for them to participate in group activities and develop fr iendships (Klin & Volkmar, 1999). Unaware of Kanners work in the United St ates, Asperger chose the label autistic psychopathy emphasizing that these children suffered more from a personality disorder. Although Asperger did not provide a list of essent ial diagnostic criteria, he emphasized the following behavioral and clinical de scriptions (as listed in Wing, 1998b):
14 1. The children were socially odd, nave, inappropriate, emotionally detached from others. 2. They were markedly egocentric an d highly sensitive to any perceived criticism, while being oblivious of other peoples feelings. 3. They had good grammar and extensive vocabularies. Their speech was fluent but long-winded, literal and pedantic used for monologues and not for reciprocal conversations. 4. They had poor nonverbal communication and monotonous or peculiar vocal intonation. 5. They had circumscribed interests in specific subjects, including collecting objects or facts connected with these interests. 6. Although most of the affected children had intelligence in the borderline, normal, or superior range on tests, they had difficulty in learning conventional schoolwork. However, they were capable of producing remarkably original ideas and had skills connected with their special interests. 7. Motor coordination and organization of movement was generally poor, although some could perform well in areas of special interest to them, such as playing a musical instrument. 8. The children conspicuously la cked common sense (p. 12-13). Both Kanner and Asperger highlighted the sa me types of deficits in their patients, and although the two descriptions were quite si milar, they were not completely identical. In addition to their commonalities, Asperger noted other features present in his descriptions. These included idiosyncratic areas of special interest, that were much of
15 a focus in the childs life, awkward and clum sy motor skills, as well as difficulties in understanding social cues, and odd, eccentric patt erns of repetitive interests. Although Kanners work became of particular clinical interest in English literature, Aspergers concept remained limited to Germanic writi ngs until Lorna Wings influential review in 1981. Contemporary Descriptions of Aspergers Syndrome Wings (1981) review of Aspergers work dramatically increased awareness and interest in Aspergers concepts. She also us ed such concepts to br oaden the perception of the autism spectrum to include those childre n who did not meet criterion for autism but still presented with clear social impairme nts (Kugler, 1998). Volkmar and Klin (2000) note that since that time, more than 100 pub lications have been devoted to the further study of AS. Generally, these reports focus on the neuropsychological distinctions and/or commonalities of AS and autism but offers litt le to the field of education. Studies reviewing the disparity in neurocognitive aspe cts and profiles between children with AS and Higher-Functioning Autism (HFA) have appeared with increasing frequency in published reports over the past several years (e.g., Klin, Volkmar, Sparrow, Cicchetti, & Rourke, 1995; Manjiviona & Prior, 1999; Miller & Ozono ff, 2000; Ozonoff, Rogers, & Pennington, 1991; Ozonoff, South, & Miller, 2000; and Szatmari, Tuff, Finlayson, & Bartolucci, 1989). Review of these studies presents a mixed bag of results and only promotes greater confusion ove r the differential diagnosis of AS and HFA and how one can better facilitate learning. Althou gh the advancements in neuropsychology are important, they are beyond the scope of this review, and whether AS and HFA can be viewed as separate disorders remains to be answered within that domain.
16 Overall, AS and HFA ma y involve the same fundamental symptomatology, differing only in degree or se verity. AS may represent th at portion of the Pervasive Developmental Disorder (PDD) continuum that is charac terized by higher cognitive abilities and by normal language function (Bau er, 1996). Most profe ssionals agree that the presence of early normal basic language skil ls is one of the defi ning characteristics of AS, and this combined with higher cognitive sk ills and a desire to form relationships (but in awkward ways) distinguishes AS from other forms of autism and PDD. AS is currently understood as a developm ental disorder characterized by children who: (a) have significant difficulties in social interactions and relationships, (b) display lack of empathy that is similar to but usually milder than that seen in autism, and (c) engage in unusual patterns of interest and unique stereotyped behavi ors, especially the tendency to over focus on certain topics or su bjects of interest (e .g., trains or vacuum cleaners) (Bauer, 1995, 1996; Church, Alisansk i, & Amanullah, 2000). Children with AS generally have normal to above normal cognitiv e abilities (sometimes in the superior range), and language function that is generally stronger than th at in other PDDs, but often is unusual in pragmatic ways (Bauer, 1995, 1996; Church, Alisanski, & Amanullah, 2000; Klinger & Dawson, 1996), a diagnostic patter n that is different from that observed in classic autism. In fact, both DSM-IV and ICD-10 specify that for a diagnosis of AS to be made, a history of normal language and cogn itive development must be present. Both DSM-IV and the International Classification of Diseases (ICD-10; World Health Organization, 1992) criteria for AS require th e social deficits and repetitive narrow interests as seen in autis m (Miller & Ozonoff, 2000).
17 Researchers, clinicians, and parents ha ve suggested additional considerations regarding how children with AS present. Many have noted that children suspected of AS exhibit some degree of motor impairment, or clumsiness, as well as actively seeking friendships not seen in autism (e.g., Ba uer, 1996; Volkmar & Klin, 2000; Wing, 2000; parent essays by Rietschel, 2000; Shery, 2000; & Wallace, 2000). Further, children with AS display deficits in the pragmatic use of language, characterized by perseveration on a topic of interest, irrelevant us e of detail in conversations (e .g., using dates and ages when discussing an event or person), abno rmal prosody (e.g., abnormal intonation, inappropriate laughing), and ignoring of conversational initiations (Klinger & Dawson, 1996). Although these symptoms are apparent, th ey do not appear to be overly atypical, and the behaviors of individuals with AS generally are less understood by those around them (Church, Alinsanski, & Amanullah, 2000). AS can be characterized by deficits in so cial interaction and stereotyped behavior patterns, yet is not associated with clinically significant de ficits in language or cognitive development. Although consid eration of the language a nd behavioral issues are important for any intervention of children with AS, the focus of this review will be a discussion of the social sequelae of AS, and th e research that is av ailable to support the sequence of social dysfunction. Social Impairments in Individua ls with Aspergers Disorder As is the case with all disorders under the PDD continuum, AS shares a common characteristic of poor, or absent, social rela tedness and erroneous use of social skills (Bauer, 1996). In a review of the research literature on the social impairments of individuals with autism spectrum disorder s, Wing (1988a) summarized the ways social
18 interactions are impaired. Impairment in social interactions ca n vary, depending on the severity of the dysfunction, but Wing (1988a ) divided these impairments into three separate most frequently observed categories: (a) impaired social r ecognition (i.e., apathy toward social interaction), (b) impaired social communication (i.e., expressive and receptive language deficits), and (c) impairme nt of social imagination and understanding (i.e., perspective taking). These impairments paired with communication deficits and restricted, stereotypical range of interests makes up the triad of impairments (Wing & Gould, 1979). Such impairments are essential in the diagnosis of any disorder within the autism spectrum and remain much of the focu s of intervention activ ities for individuals who fall along this continuum. As is the case of children with autism, i ndividuals with AS display the same triad of impairment describe above. Specifically, children with AS often do not interact with peers, possess poor appreciation of social cues and often make socially and emotionally inappropriate responses (e.g., laughing loudly when another student gets hurt). To understand these social interaction patterns, so me have turned to theory and research on the nature of social cognition (Mundy & Stella, 2001). The ca pacity for social cognition, and more importantly, social relationship s is a defining charac teristic of human neurobehavioral evolution (Cosmides, 1989). In keeping with this view, a perspective on cognition has been suggested th at the capacity to understa nd the intention of others follows a developmental course (Baren-C ohen, 1995; Leslie & Thaiss, 1992). This hypothesized component is based on what Le slie (1987) called the Theory of Mind (ToM).
19 Theory of Mind (ToM) ToM is the cognitive ability to infer the mental states of others (e.g., knowledge, perspectives, beliefs, intentions) from beha vior. Specifically, ToM employs a specific type of cognition called metarepresentation Such ability allows one to mentally depict the psychological status of othe rs (i.e., the thoughts and beliefs of others). It is called metarepresentation because it involves the capacity of one individual to mentally represent the mental representations of another individual (Mundy & Stella, 2001, p. 58). Baron-Cohen (1995) coined the term mindblindness to characterize this inability to read the behavior of others in mental stat es. Simply stated, a deficit in ToM does not enable a child with AS to distinguish his or her own t houghts or feelings from the thoughts and feelings of ot hers (Leslie, 1987, 1993). According to this view, a disturbance in ToM gives rise to the social and pragmatic deficits of individuals with AS (Baron-Cohen, 1995; Leslie, 1987). The logic is that these individuals will have difficulty conceptualizing and appreciating the thoughts and feelings of another person (Atwood, 2000). In addition, individuals with AS will exhibit deficits in identifyi ng the communicative intents of others, understanding figures of speech (e.g., idioms), and following the conventions of topic maintenance when speaking (Mundy & Stella, 2001). These difficul ties have a major impact on the childs social reasoning skills and behavior. For ex ample, a child with AS may have difficulty identifying whether another childs running into him in physical education class was intentional or acciden tal. Along the same lines, a child with AS may have difficulty understanding how their own actions (e.g., taking excessively about tr ains) may affect the thoughts of others around them.
20 Under normal circumstances, the ability to read social situations and people and to adjust ones communicative behavior is a task that is performed effortlessly (Twachtman-Cullen, 1998). The ability to hold the capacity for ToM is governed largely by the indirect, socially mediated cues that typically are recognized below the level of consciousness. Thus, because children w ith AS have a specific difficulty in understanding social information, they are part icularly compromised in their ability to engage in reciprocal social in teractions, an integral part of communication in any society. Taken together, the clinical and diagnos tic pattern of dysfunction for children with AS appears to be less pervasive than th at seen in classic autism. Less pervasive symptoms suggest that AS may be consider ably more common than autism, as well as more amenable to intervention. Notably, Baue r (1996) reports the prevalence of AS to be as high as 20-25 per 10,000 children, with generally a better pr ognosis than classic autism; however, it often goes undiagnosed. In a recent Journal of the American Medical Association (JAMA) report, Hyman, Rodier and Davidson (2001) suggest prevalence rates for AS to be as high as 63 per 10,000 bi rths, as compared prevalence rates of 21 to 31 for autism. Moreover, Safran (2001) reports that prevalence rates that appear to be several times higher than classic Kanne r childhood autism, [and] suggests a large unserved student population in North America (p. 151). With this in mind, AS becomes an auspicious challenge for special educa tion, and one that is ripe for research. Aspergers Syndrome in the Schools Due to recent inclusion in both the DSM-IV and the ICD-10 diagnoses of AS and other higher functioning autism spectrum disorder s have increased. This increase has led to increased referrals for special educati on services and has forced educators to
21 restructure services to better cat er to children with this unique profile of severe social and communication disabilities in the presence of cognitive and language strengths (Klin & Volkmar, 2000). Unfortunately, the resources an d services available to such individuals are still limited. To often, children with AS are seen as too bright, or too verbal, (Klin & Volkmar, 2000) or considered to be functioning normally but with odd behaviors (Wallace, 2000). Additionally, educators may be misled by th e enhanced verbal abilities of children with AS and may attribute poor social skills and poor perf ormance on nonverbal tasks to negativism or other voli tional behaviors (Safran, 2001). As a result, these individuals are viewed as behaviorally disordered or socia lly maladjusted and placed into special classes for children with behavior problems (Vol kmar, Klin, Schultz, Rubin, & Bronen, 2000). Children with AS may typically be viewed as unusual or just different, and miscatogorized as exhibiting conditions sim ilar to Attention Deficit Hyperactivity Disorder (ADHD) or other beha vioral or emotional disturba nces. In an attempt to describe the experiences of children with AS between the ages of 3 and 15, Church, Alisanski, and Amanullah (2000) found that the most frequent diagnosis or educational label was ADHD (20%) (p. 14). Placing childre n with AS in classes for children with behavioral impairments may lead to their c ontinued experiences of social failures with peers. Often, the child with AS may be ridiculed and tormented for his/her odd and eccentric behaviors in classes designated prim arily for children with behavior problems. As Volkmar et al. (2000) explai n, this approach might lead to the placement of a child as a perfect victim, with perf ect victimizers (p. 266).
22 Using knowledge from recent diagnostic criteria, clinical observations, and current theoretical models, Bauer (1996) provi ded a clinical outline of children with AS as they progressed through pr eschool, elementary, and middle/high school. Such an outline combined the empirical progress in di agnostic trends while integrating the many concerns and ideations of what Asperger had originally described nearly 60 years ago. Much of what Bauer (1996) had hypothesized clinically was confirmed through further descriptive analysis by Church, Alisanski, and Amanullah (2000). Combined, these reviews provide educators with critical info rmation, giving professi onals in education a more comprehensive model to build interventions. Preschool Key elements for educators to be aware of during the preschool years for children suspected of having AS include: (1) a tendency to avoid spontaneous social interactions; (2) impairments in the pragmatic use of social communication; (3) perseverance on particular objects or subjects, with a tendency to only want to talk about such interests; (4) difficulty in regulating social/emotional re sponses; and (5) an a ppearance of being in ones own little world (Bauer, 1996). Many of these characteristics do not appear to deviate much from what is seen in childre n within the autism spectrum. However, the child with AS is more likely to show some soci al interest in other children and with their elevated use of language and conversationa l ability, they may not present as being obviously different from other, more ty pical, children (Bauer, 1996; Safran, 2001). Volkmar and Klin (2000) indicat e that individuals with AS experience social isolation, but are not withdrawn or devoid of social interest (p. 59). Thus, children with AS show
23 an interest in having friends, a finding that contrasts with the patte rn found in individuals with autism. Church, Alisanski, & Amanullah (2000) conducted a retrospective qualitative review of 40 (39 boys, 1 girl) children diagnos ed with AS at a university medical center child development program between 1986 and 1998. For each child, at least two investigators reviewed all of the informati on available for review in the childs chart (e.g., parent, school, and heath-car e provider information) to ensure that the child met the DSM-IV criteria for AS. All charts were an alyzed qualitatively with emphasis placed on parent comments and observations, observe d child behaviors and interactions, and teacher reports. At the time of the data extraction, the children fell into the following groups: (a) preschool aged children (n = 1), el ementary aged children (n 26), and middle and high school aged children (n = 13). However, because the children were followed over several years, data were available on 40 preschoolers, 39 elementary aged children, and 18 middle and high school aged children (m iddle and high school aged children were not collapsed). Although this study did not employ rigor ous observational and quantitative measures, it does provide the reviewer with a developmental progression of what Bauer originally hypothesized. Because so little is known regarding the experiences of children with AS, this study illustrates specific problem areas that need to be addressed and are critical to the child with AS success. Elements of the ke y findings are reported to emphasize Bauers (1996) hypotheses. At the preschool level, children in Church, Alisanski, and Amanuallahs (2000) descriptive study did not stand out as being differe nt enough from their peers to come to
24 the attention of primary care providers as needing comprehensive assessment. As a group, the children had specific difficulties in social interactions play skills, and exhibiting odd and unique behaviors. Specifi cally, the preschoolers appeared to get along with teachers and other adults, but had the most difficulty in in itiating, sustaining, and maintaining relationships with other children. Mo st of the children preferred to be on the periphery of activities, failed to read the social cues of teachers and other students, desired the adherence to specific routines/ritua ls, and wanted to talk about their area of interest to anyone at any time. Interes tingly, 95% of the parents of these children suspected that something was seriously wr ong with their child pr ior to the start of kindergarten (p. 13); however, 72% did not recei ve a formal diagnosis by the end of this period (Church, Alisanski, and Amanuallah, 2000). Elementary school years As the child with AS enters kindergar ten, educators may pinpoint behavioral concerns (e.g., hyperactivity, inattention, aggression) and more often express concern over immature social skills and peer interactions (Bauer, 1996). When these concerns are severe enough, special education services may be suggested, but generally, most children with AS will enter the mainstream setting. Academically, children with AS typically perform well, especially in the ar eas of rote reading a nd calculation skills. However, the childs obsessive interests ma y interfere with the classroom setting and at the same time amaze the teacher. Consistent w ith the preschool years, elementary level children with AS express interest in others, but are unable to make and keep friendships. Bauer (1996) noted that childre n with AS may tend to associ ate with a few members of the class, but usually these relationships are relatively shallow and superficial.
25 Additionally, Prior et al. (1998) noted that ch ildren with AS actively look for friendships, but often their method of doing so is clumsy and unsuccessful. Developmentally, children at this age are beginning to seek meaningful friendships. These friendships are those in which trust and assistance become important components, both of which provide positive support for the development of prosoc ial and altruistic behaviors. Data available on the 39 elementary aged children (ages 6 to 11 years) in the Church, Alisanski, and Amanullah (2000) study demonstrated similar patterns. Parents, teachers, and physicians reported that children within this age group demonstrated serious developmental problems. In particular, social skills began to emerge as key areas of deficit. Through their review of elementary -aged childrens charts Church, Alisanski, and Amanullah (2000) reported that none of th e children in this sample had reciprocal relationships with other children, but some di d manage to form superficial relationships (e.g., usually to share information on topics of interest). Another key finding was that elementary-aged children did not have the ability to take the perspective of others. These children often misread so cial cues (e.g., body language, gest ures, facial expressions), and often acted inappropriately (e.g., blurting out socially inappropriate comments and not understanding the impact of those comments on others). In addition to these social problems, these children were extremely ru le-based and desired sameness and specific order to be maintained. Language was often used inappropriately, more often to discuss the childs topic of interest and perseverating on the same topic day after day. Although all of the reports by parents a nd other professionals indicated serious social deficits, only 25 of the 39 children (64%) were formally diagnosed with AS by the time they reached 11 years of age.
26 Middle and High School years Bauer (1996) emphasized that the most diffi cult areas for individuals with AS as they move through middle school and high school are those areas rela ted to socialization and behavioral adjustment. Often, these diffi culties only increase, leading a child with AS to become isolated and perhaps asha med of his/her eccentric behaviors. Subsequently, children with AS may be left out, misunderstood, or teased and persecuted (Bauer, 1996). Their inability to make friends and fit in may result in further withdrawal, resulting in some degree of depression. In fact, comorbidity of AS and depression are common in late childhood and adolescence due to the individuals increasing awareness of personal inadequacy in social interactions a nd repeated failures in making and/or maintaining relationships (Klin & Volkmar, 1997). This awareness is quite different from children with autism w ho are seemingly unaware of their social deficits. During adolescence, pressure to c onform is great and tolerance for differences are at a minimum. It is at this stage that the child with AS ma y become an outsider, undetermined to find and maintain friends, desp ite their wishes to hold friendships with peers. Church, Alisanski, and Amanullah (2000) de monstrated that for individuals with AS, social skills deficits continued to be the major area of difficulty during the middle and high school years. Not only were these deficits in social skills apparent, but also these children stood out socially among their peer group. For example, of the 18 students in their study (13 children followed thr ough middle school and 5 individuals followed through high school), all were described as having an inability to read the social cues of their peers, awkward body postu re, awkward use of gestures annoying habits such as
27 making noises or drumming desks, highly va riable eye contact, and odd body language (Church, Alisanski, & Amanullah, 2000, p. 17). Behaviorally, parents reported that many of these children adhered to routines and rituals during middle sc hool, which progressed to obsessive-compulsive behaviors during the hi gh school years. In addition, one third of reports by both parents and teachers describe d these children as very inappropriate, rude, or silly. Notably, 38% of these children had a secondary diagnosis of a behavior disorder (e.g., oppositional defian t disorder, conduct disorder) (Church, Alisanski, & Amanullah, 2000). Both the theory laden hypotheses of Baue r (1996) and the descri ptive analysis of Church, Alisanski, and Amanullah (2000), illu strate the social specific problems that need to be addressed and are critical to the childs success. Despite this burgeoning information base of children with AS in schools, there has been scant attention paid to AS from the special education literature (Safran, 2001). New information has sparked only limited interest among professionals in educa tion in considering e ffective interventions for children with AS. Williams (1995) offere d strategies for teachers, pinpointing the areas of known deficit. However, many of the suggestions Williams provided are similar to those methods used with children with auti sm and may not provide optimal benefits for children with AS. Attwood (2000) also made suggestions for improving peer relations for children with AS, outlining many of the techni ques used to facilitate social learning in high functioning children with autism. Many of his suggestions have been confirmed by research to be effective for children with HFA (e.g., opportunities to interact with normal children, labeling specific emotions) and ma y hold promise for children with AS. Further, recent reviews have suggested more direct interventions aimed to facilitate
28 socialization in children with autism spect rum disorders (e.g., struct ured teaching, social skills training, peer-tutoring, peer-mediated approaches, use of social stories and comic strip conversations). However, efficacy data are still limited pertaining to purely AS samples (Rogers, 2000; Safran, 2001). In es sence, there appear to be few support services for mainstreamed children with AS a placement that many of these children encounter. Directions for Intervention Given the critical information on AS reporte d in the past few years, it is now necessary to formulate suggestions to be considered when planning and implementing educational interventions for i ndividuals with AS. With the increase in the use of terms such as Higher Functioning Autism, hi gher functioning Pervasive Developmental Disorder-Not Otherwise Specified (PDD-NO S), or, more importantly, Aspergers Syndrome, it is imperative that educators and other service provi ders be cognizant of interventions that will benefit these children with severe social weaknesses despite their cognitive strengths. Social Stories One suggestion that has appeared repeatedly in recent years to improve the social integration of children with AS is the use of Social Stories. Social stories are an approach to teach social skills by teaching social understanding through the extensive use of visual materials (Gray, 1994, 1995). A social story is an individuali zed short story that can be used to assist individuals with autism spectrum disorders in interpreting challenging or confusing social situations (Gray, 1997). According to Gray (1995), social stories focus on providi ng the student with descriptiv e information regarding what
29 is occurring in a given situati on and why. Social stories serv e a wide variety of purposes, and they appear to be particularly helpful in facilitating the incl usion of students with autism spectrum disorders in mainstr eamed classrooms (Gray & Garand, 1993). Examples of social stories from The Social Story Book (Gray, 1994) appear in Appendix C. Rationale behind social stories. Social stories are based on the growing consensus among researchers that children with autism spectrum disorders suffer from an inability to read and understand social cues and situations, perspectives of others, as well as formulating appropriate responses to such social events (Atwood, 2000; BaronCohen, 1995; Gray & Garand, 1993; Les lie, 1987, 1993; Mundy & Stella, 2001; Twachtman-Cullen, 1998). Unfortunately, trad itional teaching, whethe r it be a classroom lesson or social skills intervention, genera lly involves face-to-f ace interaction between the student and the teacher/facilitator. Therefore, using a traditional teaching approach to teach social behaviors presents the student w ith what Gray and Garand (1993) identify as a compounded challenge: (a) understanding the lesson, and (b) accurately interpreting the social cues used in its presentation ( p. 2). Social stories seek to minimize these concerns by providing the student with information regarding (a) the norms for behavior in the context of the specific story, especia lly the unspoken social e xpectations in given situations; (b) the persp ective(s) of others involved in the social situation; (c) the specific behaviors required for the specifi ed situation; and (d) choices for the individual to make to participate successfully in different soci al situations (Fulle rton, Stratton, Coyne, & Gray, 1996; Gray 1997; Gr ay & Garand, 1993).
30 Social stories are offered as a more pow erful intervention compared to traditional analog social skills interventions because they dismiss the possibility of a child becoming confused when directly confronted with anot her individual. Traditional social skills interventions generally involve the process of role-plays and modeling with same-aged peers. However, due to their inability to mind-read, children with AS may not fully understand the content of these lessons, or, more commonly, become threatened by the presentation of the lesson. As previously discussed, children with AS hold unique perceptions of people and events and re spond according to their own rule-based cognitions. This rule-governed behavior affects their understanding of communication and interaction with others re sulting in the childs failure to interpret the interaction accurately. The shortcomings of traditional so cial skills lessons are that they do not include descriptive information regarding how people respond and why, as well as use materials/instruction that will be most eas ily understood by the child with AS. Under these circumstances, the child with AS may be unable to process, and later use, the information provided because of the highly soci al nature of its pres entation. This is perhaps the rationale behind the limited gene ralization of social skills training for children with any autism spectrum disorder. Developing a social story intervention. Gray (1994, 1995) outlined four basic steps, (see Table 1) in the development of a social story intervention. This process involves first targeting a specific problematic so cial situation that th e social story will have its focus. Once a target situation is identified, all of the salient features of the context or setting need to be identified (e .g., where a situation occurs, who is involved, how long it lasts, how it begins and ends, what occurs, and why). Gray (1995, 1998)
31 states that such information generally can be identified through in terviews and extensive observations of the target situation. Observ ations prior to implementing a social story should be of particular importance because th ey allow the observer not only to witness the target childs interests and abilities, but also to gain insight into what may be motivating his/her current responses. Gray (1994) provides a sample form ( Social Stories Information ) that can be used to structure the obser vations, as well as as sist in gathering information that is needed to write a successful social story. A copy of this form appears in Appendix D. Following this initial information gather ing process, the educator, clinician, and/or facilitator can then identify a new soci al skill to be taught using a social story. When considering a new social skill to be taugh t, consideration of (a) the target situation; (b) the perspective(s) of all individuals present in the cont ext (i.e., child him/herself, others); and (c) the behavioral expectation(s) in the context of the targeted situation (i.e., what should and should not be done and wh ere) must be considered (Gray, 1995).
32 Table 1. Steps in the Development of Social Stories (Gray, 1995) Identify and Target a Social Situation Identify a problematic or pot entially problematic social situation. Gather Information Gather data relative to the targ et childs interests, abilities, impairments, and factors motivating the current behavior. Use videotaped reviews to identif y the salient features of the social situation and the target ch ilds perspective of the social situation. Determine the focus of the social story to be developed by merging the salient contextual information with the target childs specific perspective and interests Share Observations Review the information gathered above with the target child and other relevant individuals to determine the accuracy and content relevance of the materials. Identify and Support New Social Skill Generate the specific social skil l, and related behavioral steps necessary to successfully ma nage the target situation. Generate a short social stor y which describes the target situation, including the identif ication of the relevant information regarding the contex t, perspectives of self and others, and desired behavior. Swaggart et al. (1995) el aborated Gray (1994) and Gray and Garands (1993) suggestions to developing soci al stories, and proposed a 10-step process for the creation and implementation of social stories. Their pr ocess involved all of the original aspects of social story creation (i.e., id entifying a target behavior for change, defining target behavior, writing a social stor y), but expanded on the process of creating a social story book (e.g., presentation of sentences on a page use of icons and/or pictures), and methods of implementing a social story. A brie f outline of Swaggart et al.s (1995) social story process follows, and is presented in Table 2.
33 Table 2. Steps in the Development and Implementation of Social Stories (Swaggart et al., 1995) 1. Identify a target behavior or problem situation. 2. Define the target behavior of data collection. 3. Collect baseline data on the target behavior 4. Write a short social story using the four sentence types. 5. Present one to three sentences on each page 6. Use photographs, hand-drawn pictures, or icons. 7. Read the social story to the stude nt and model the desired behavior. 8. Collect intervention data. 9. Review the findings and relate d social story procedures. 10. Plan for maintenance and generalization. 1. Identify a target behavior or problem situation At this stage, the developer of the social story should focus on a beha vior that will result in increased positive social interaction or a social learning opportunity. 2. Define the target behavior of data collection. The purpose of this stage is to describe the target behavior clearly and concisely. Behaviors should be defined so examples and nonexamples of that behavior are included. The purpose of this is to ensure th at anyone reading the definition could precisely identify the desired behavior. 3. Collect baseline data on the target behavior. Collecting data over a period of time will allow the developer to recogni ze any trends. Data on the target behavior should be collected for a period of 3 to 5 days. Baseline data allow the developer to tally the freq uency of the behavior, as well as provide a basis for comparison for th e individuals behavior during and following the social story intervention.
34 4. Write a short social story using the four sentence types. Always use the first person when writing social stories, an d make sure that story describes a situation that is likely to occur in the future. 5. Present one to three sentences on each page. Presentation of the story is directly dependent upon the abilities of the individual. However, the format should always be simple and should avoid overloading of information. 6. Use photographs, hand-drawn pictures, or icons Pictures are used to enhance student understanding (this is espe cially true for lower functioning individuals), but should not define the social situ ation too narrowly as this could result in limite d generalization (Gray, 1994). 7. Read the social story to the stude nt and model the desired behavior This step is perhaps most crucial in the so cial story intervention, and it should become a consistent part of the st udents daily rou tine/schedule. A social story should be presented several times throughout the day and should occur at the same time everyday. 8. Collect intervention data Data on the target beha vior should be recorded throughout the intervention process, a nd should be collected in the same manner as baseline data. 9. Review the findings and relate d social story procedures There should be a period of at least two w eeks to determine if the social story is producing the desired behavior(s). If the stor y is not producing the desired results, then the social story must be altered. However, it is imperative that only
35 one variable of the social story be changed at a time. For example, if one sentence is changed in the story, the time of day that the story is being presented to the indivi dual should not be changed. 10. Plan for maintenance and generalization Once behavior change has been established, then the process of fadi ng the social story can begin. This can be accomplished in several ways (e.g., increasing time between readings, placing more responsibility on the child to read the stories), but should assist in generalizing the content of the story across persons, settings, and situations. Guidelines for writing social stories. Social stories need to be written with regard to the students comprehension level (Gra y & Garand, 1993), and should be comprised of sentences of four basic types: (a) descriptiv e, (b) directive, (c) perspective, and (d) control, with each serving a specific pur pose (Fullerton et al ., 1996; Gray, 1994, 1995; Gray, 1997, 1998). These types of sentences a nd their identified purpose are summarized in Table 3. Table 3. Social Story Sentences and Their Identified Purpose Descriptive Sentences Begins the Social Story. Objectively defines where a situation occurs. Identifies the salient features. Directive Sentences Individualized statements of desired behavior(s). Perspective Sentences Describes the reaction and feeli ngs of others in the target situation. Control Sentences Aids in understanding abstract situations through use of nonhuman analogies.
36 Descriptive sentences are used to descri be a setting and explain what occurs, who is involved, what they are doing and why (G ray, 1998). Gray and Garand (1993) indicate that descriptive sentences paint the social b ackdrop of the targeted situation (p.3). An example of a descriptive sentence would be Some children ride to school on a bus. Such sentences immediately give information regarding the most relevant characters in a story and/or their role. Directive sentences represent the desired social skill or behavior (Gray, 1998). Often, these sentences begin with phrase like I can I will try or I will work on (e.g., I will try to sit quiet ly when someone is talking I will try to stay calm if a friend wins a game.), and follow descrip tive sentences. Collectively, these phrases provide a sequential list of e xpected responses to the target ed situation, as well as focus on what the student should do in order to be successful in that target situation (Gray, 1998; Gray & Garand, 1993). It is important th at directive sentences be stated positively, and emphasize a statement that is intended to assist in learning and practicing new responses. Such a practice minimizes the pressu re of a student to get it right the first time or comply exactly to be regarded as successful (Gray, 1998). At times, social stories may contain perspe ctive sentences. Perspective sentences describe the reactions and f eelings of others in a give n social situation (Gray, 1994, 1995, 1998; Gray & Garand, 1993), as wells as complete a social story by providing information regarding the impact of the soci al skill on ones self and others (e.g., My friends like it when I say hello to them in the morning.).
37 Control sentences are written by the stude nt with autism spectrum disorders that enable the student to identify strategies the student may use to remember the information contained within the social story (Gra y, 1994, 1998). Because these sentences are formulated by the individual, they contain in formation that reflects the interest of the individual. Therefore, it is im perative that such sentences ar e used as analogies utilizing nonhuman subjects (Swaggart, et al ., 1995). For example, a student who has a interest in insects may write the following control sentence about people when changing their minds: When someone says I changed my mind, I can think of an idea becoming better like a caterpillar changing into a butterfly (Gray, 1998, p. 179). Social stories generally comprise of two to five sentences. Gray (1998) suggests using the Social Story Ratio which defines the proportion of descriptive, directive, perspective, and control sentence s appear in the entire social st ory. As a rule of thumb, it is customary for each social story to have two to five descriptive, perspective, or control sentences for every directive sentence in the story (Gray & Garand, 1993; Swaggart, et al., 1996). The essence is to describe more than direct. Aside from sentence structure, social st ories should contain one concept per page, and should be presented with the perspective of the student in mind. As Gray and Garand (1993) state, such a practice emphasizes each point in a story and allows for variations to a basic social story (p. 4) Variations to the story are essential to keep events as realistic as possible (Gray, 1994, 1995, 1997) For example, using the word usually is often more accurate than always The need to portray possible variation is especially true if describing an event that calls for greeting someone at th e door. Sometimes it may be a girl, and sometimes it may be a boy. Using th is example, the developer would want to
38 show that the person could be someone diffe rent every time. C onsideration of such flexibility, and more importantl y, wording, is essential for th e generalization of the skill that is the focus of the story. Along the same lines, it is imperative that the developer of the social story does not present any misleading information to th e individual regarding the content of the story. Children with autism spectrum disord ers often make highly literal interpretations of statements (Kanner, 1943), regardless of the absurdity that may be apparent. Gray and Garand (1993) provided an excellent example. Using the statement I will not talk in the library in a story could be misleading. Wh at if the child is asked something by his teacher? Stating the desired behavior, I will wh isper in the library or I can whisper in the library, helps a student understand expectat ions (p. 4). In the same way, the use of pictures must be considered carefully when creating a social storybook. Often, pictures can be distracting, but even more alarming is that children may make an inaccurate interpretation of the situ ation based on the illustrati on (Gray & Garand, 1993). For example, if an illustration depicts a boy shaking another boys hand on the playground, next to a swing set, on a sunny day, the student may interpret this to mean that you only shake a boys hand on the playground, next to a swing set, on a sunny day. It has been suggested that several photogra phs depicting variations of the same theme may be more effective (Gray, 1995; Gray & Garand, 1993). The creation of a social story is not a sc ience, but is more of an art (Gray, 1994, 1997). Due to the highly idiosyncratic nature of autism spectrum disorders, social stories must reflect the students indivi dual understanding a nd situation in order to be effective.
39 Creativity and strength-based story creation are essential to the proper implementation of a social story. Guidelines for successful implementation of social stories Gray and Garand (1993) indicate that there are three basic ways for implementing social stories. Selection of the most appropriate tec hnique is highly dependent upon the individual abilities and needs of the target child (Gray & Gara nd, 1993). First, for a student who reads independently, an adult (e.g., car egiver, teacher, clinician) introduces the story by reading it to them twice. Gray and Garand (1993) sugge st that the adult sit behind and to the side of the child. The adult then r eads the social story to the child first, followed by the child reading it back. The adult may read the story with the child several more times in this manner, or until the child is acquainted with th e story. At this point, the child reads the story once a day, independently. The second manner in which a social st ory can be implemented is through audio equipment. Audio implementation is reco mmended for those individuals who cannot read (Gray & Garand, 1993). The story is firs t recorded on a cassette tape. The story is recorded onto the tape, as well as a beep that is used to indicate when the child is to turn the pages of the social story. Once the child is taught to use the ca ssette and to turn the pages when hearing the cue, the target chil d then should read the story no more than once a day (Gray & Garand, 1993). A third approach to implementing social stories is through the use of videotaped social stories and video mode ling (Charlop & Milstein, 1989). This approach is used either with students who can read independent ly or with those who need assistance (Gray & Garand, 1993). On the videotaped versions the social story is read aloud on the
40 videotape with one page appearing on the scr een at a time. Videotaped social stories make it possible for the story to be read by th e target child (volume on) or for the student to read the story himself or hers elf (volume off) (Gray & Garand, 1993). Regardless of the method that is used to implement any given social story, it is necessary for comprehension of the story to be checked. Gray and Garand (1993) recommend two ways in which this can be done : (a) the student completes a checklist or answers questions in writing at the end of the story, or (b) the student role plays demonstrating what he or she will do th e next time the situation occurs. Once comprehension has been checked, it is sugge sted that an implementation schedule be created (Gray, 1994, 1998). Gray (1994) recommends using the Social Story Implementation Plan which is used to indicate when instruction and reviewing will occur, as well as provides a record of the methods (i.e., story format, monitoring responses) and support materials (i.e., reminder sign, Social Calender) used for the story. A copy of this plan appears in Appendix E. The most essential factor that needs to be considered when implementing a social story is monitoring of student progress on the ne w social skill being ta ught. Gray (1994) provides the Social Story Report (which appears in Appendix F), which keeps a running record of the students progress and/or diffi culties with the story. Understanding where a social story may be weak and needs to be im proved is essential for the intervention to be successful. Therefore, such a form provides information regarding such events, as well as provides and efficient means of communi cation between the person implementing the social story and those directly involved with the target child.
41 Effectiveness of social stories. Unfortunately, a review of the available published research on both PsychINFO and ERIC database s regarding social story interventions for children with AS yielded no findings. Ho wever, six empirical studies were found relating to the effectiveness of social story in terventions used with ch ildren with autism. Only three of these studies employed expe rimental control, and none referenced consideration of programming for generaliza tion (e.g., natural contingencies, multiple environments) as suggested by Stokes and B aer (1977). Furthermore, none of these studies used methods to assess the social valid ity of the treatment. However, all of the studies that appear in this review demons trate preliminary efficacy of social story interventions with a variety of child ren with autism spectrum disorders. Swaggart, et al. (1995) used a social stor y intervention to teac h appropriate social behavior to three children with autism. A ll participants attended a self-contained laboratory educational setting for children and youth with autism housed in a large university medical center. Each of the participants was included in this restrictive selfcontained setting due to their in tense aggressive behavior(s). Participants in this study fell within the spectrum of having moderate to se vere autism. In addi tion, all participants manifested significant impairments in cogni tive ability and expr essive communication skills. Interventions were based on a combination of social stories a nd the social-skill intervention model for children with autism advocated by Simpson and Regan (1988). This approach, described in Table 4, incorpor ated the use of reinforcers to be used to develop and maintain appropriate responses. Environmental and task analyses of the aggressive behaviors by the three participants revealed that the first participants behavior
42 was related to an absence of appropriate greeting skills (i.e., ve rbalizing Hi, waving, keeping hands to herself), while the other tw o participants demonstrated a deficit in sharing skills (i.e., giving a requested item to another student, parallel play). Table 4. Steps in the Social-S kills Training Model 1. Identify and define social deficits to be modified. 2. Identify and analyze related environments and situations. 3. Identify and analyze continge ncies related to targeted behavior. 4. Accurately measure the deficit response and visually display its occurrence. 5. Conduct a task analysis of behavioral observations. 6. Select and apply appropriat e reinforcers and training procedures to develop the desired response. 7. Carry out program maintenance and follow-up activities. Following the identification of the targeted behaviors for the participants, social stories were developed. With the exception of one social story that was created for the first participant, the development of each social story followed the suggestions outlined by Gray and Garand (1993). Once the social stories were developed, the programs classroom teacher and paraprofessional, along w ith four graduate students working in the classroom, implemented the social story interv entions with each of the three students. Each of the implementers had known the studen ts for a minimum of two months prior to the study. The social stories were individually read to the target participants at the beginning of each day. Furthermore, verbal prompting was used by the staff members throughout the school day to reinforce the use of the skill outlined in the participants respective social story.
43 Baseline data were collected for nine days prior to the implementation of the intervention program for the first part icipant. Data on 119-baseline greeting opportunities indicated that 7% were judged as appropriate (i .e., involved verbally saying Hi), 82% involved inappr opriate touching (i.e., huggi ng), and 9% were overly aggressive (e.g., involved hitting, pushing, sc ratching). There were no occasions when this participant ignored others. Results of the combined social stories and social skills instruction intervention (which occurred for nine days) indicated a decrease in touching and aggressive behaviors. Specifically, 57% of the greeting behaviors were judged appropriate, while touching decreased to 26%, and no occasions of aggressive behavior occurred. Baseline data collected for ten days for the second participant indicated the presence of aggressive behavior 30% of the time. In addition, data collected on sharing revealed that the participant did not shar e in 100% of the sessions, and screaming occurred during 100% of the sessions. Follo wing a 18 day intervention phase, aggressive behavior was reduced to 6% of the sessions, sharing increased to 22% of the sessions, and screaming was reduced to 56% of the sessions. Relative to the third participant, baselin e data collected over 10 days indicated the presence of aggressive behavi or in 20% of the sessions a nd grabbing another toy from a child during 100% of the sessions observed. Furthermore, the participant did not demonstrate any sharing behaviors during th e sessions observed (absence of sharing 100%). Intervention phase lasted 17 days, a nd postintervention observations revealed a complete decrease in aggressive behavior and a 35% reduction in grabbing toys from another child.
44 The case studies explained here provide so me support for the use of social stories with children with autism, but should be cons idered with caution. Specifically, this study demonstrates support for the utility of social stories in reducing the inappropriate social behaviors demonstrated by the participants while increasing the frequency of more socially appropriate behavior. However, the study did not employ any means of experimental control. In addition, the sample size limits the generalizability. Furthermore, there was no measure utilized to assess the impact of the treatment on teachers or parents perspectives of the beha vior of the participants (social validity). Perhaps the most problematic issue is that Sw aggart et al. (1995) combined the use of social stories with a more traditional behavi oral social skills tr aining program and never controlled for the effects of eith er. Therefore, it is not possibl e to determine the effects of the social stories alone. Kuttler, Myles, and Carlson (1998) used a social story intervention to reduce tantrum behavior in an individual with auti sm. The participant was a 12-year-old boy diagnosed with autism, Fragile X, and intermittent explosive disorder with moderate cognitive impairment and severe expressive language deficits (i.e., only using oneand two-word utterances). The participants ta ntrum behavior was observed and analyzed to see what factors may be contributing to th e maintenance of the behavior. Tantrum behaviors were operationally defined as sc reaming, making inappropriate vocalizations, and dropping to the floor and thrashing. Envi ronmental and task analysis indicated that the tantrum behavior occurred most frequen tly during transitions, wait time, and free periods. Subsequent social stories were designed to reflect these concerns.
45 Two social stories were created for the participant. Both stories were based on the two most difficult periods of the school da y for the participant: (a) lunchtime and (b) independent work time. Each story id entified these difficu lt situations through descriptive sentences, desc ribed the target/replaceme nt behavior (e.g., waiting appropriately, moving from one task to anot her), and indicated th e reinforcement earned for engaging in the desired behavior. In addition, each social story was designed using picture icons to accommodate for the receptive communication deficits of the participant. Using an ABAB plus alternating trea tments experimental design, observational data were collected on the fr equency of the participants tantrum behaviors both during treatment and in the absence of the intervention. Initial baseline observations collected over a 5-day period indicated th at tantrum behaviors occurred, on average, 15.6 times per day during the morning work time. During l unchtime, the mean frequency of tantrum behavior was 11.6 times per day. During the first treatment phase, tantrum behaviors occurring during morning work time reduced to 0 times per day, and 2.0 times per day during lunchtime. When the return to baseline condition wa s presented, tantrum behaviors increased to an average of 15.33 tim es per day during morning work time, and 18 times per day during lunchtime. During th e second treatment phase, tantrum behavior again reduced for morning work time a nd lunchtime, 0 and 1.0 times per day respectively. Clearly, these findings suggest that imple mentation of the social stories were effective in reducing the frequency of tantru m behaviors of the participant. This study also provides a measure of control through its use of an alternati ng treatment-baseline design. However, a return to baseline frequency of tantrums upon removal of the
46 intervention suggests a possible threat to intern al validity. Due to th e teaching nature of the social story intervention gains in target behaviors should be maintained following discontinuation of treatment. As with the Swaggart el al. (1995) study, the generalizability of these results are limited bot h by its single subject design and failure to include any social validity measures re garding the impact of the intervention. Hagiwara and Myles (1999) investigated the effects of computer-based social stories for three elementary-aged boys with autism. Functional assessments were used prior to the development of any social st ory to identify target behaviors for each participant. For the first tw o participants, completion of washing hands was chosen as the target behavior. The target behavior identified for the third participant related to average duration of time spent on-task (e.g., reading aloud, writing, answering questions). Computer-based social storie s were designed as educative approaches for the identified behavior of each participant. Each participant was able to operate the social story program once a day immediately prior to enteri ng a specified setti ng. Effectiveness of these social stories was assessed using a multiple-baseline across settings design for each participant. Specifically, hand washing behavi ors for participant I were examined before snack, prior to lunch, and afte r recess. Likewise, hand washing behaviors for participant II were examined before going to resource room before lunch, and after recess. Settings for participant III included lunch, resource room, and a general education classroom. Results of social story implementati on revealed only minor improvements in targeted behaviors. Participant Is hand-wa shing behavior increased 17% during morning snack, 8% before lunch, and 9% after recess respectively. Data for participant II revealed similar patterns (8% increase before resource, 9% increase before lunch, and 8% increase
47 after recess respectively). The average durat ion of participant IIIs on-task behavior across lunch, resource room, and genera l education classroom settings did not demonstrate any consistent effects of the social story intervention. While the results of this study demonstrat ed limited support for the effectiveness of social story interventions the results need to be considered cautiously. First, in relation to the first two participants, increases in hand washing behaviors acr oss settings were not dramatic. In fact, baseline levels of completing hand washing averaged around 80%, and already high ceiling for any behavior, especi ally the percentage of times young children complete this behavior. Second, data were onl y collected for a period of 16 days due to restrictions beyond the control of the investigators. Thes e factors alone significantly impact the interpretation of the findings. However, the delivery method of the social stories (computer-based) was unique and represen ts an area for continued research in the future. Norris and Dattilo (1999) examined th e relationship between a social story intervention and the inappropriate social intera ctions of a girl with autism who attended an inclusive second-grade classroom. The participant was an 8-year-old AfricanAmerican girl diagnosed with mild to modera te autism with severe deficits in receptive and expressive language skills. The partic ipant functioned in th e average range of general mental abilit ies with strengths in reading recognition, sp elling, and daily living skills. Inappropriate social interactions (e.g., singing or talking to herself) occurred frequently during lunchtime, and were the focus of the social stories. Three different social stories addressing the participants social interactions during lunchtime were created according to Grays (1994, 1998) recommendations. One
48 of the three stories was read each day, im mediately prior (10-15 minutes) to lunch (priming strategy), with the participant wh ile she read the story aloud. A video camera, with an attached microphone, was then used to record 8 to 10 minutes of each lunch period. Videotaping occurred prior to the study to increase the participants exposure to being videotaped. This vi deotaping process allowed fo r monitoring of targeted inappropriate and appropriate social interactions, as well as the absence of social interactions, during baseline and inte rvention conditions of the study. Dependent measures in this study include d the participants estimated frequency of (a) appropriate social in teractions (i.e., initiating or responding to other students verbally or gesturally); (b) inappropriate social interactions (i.e., verbalizations with bizarre content, making noises); and (c) absen ce of social interac tions (i.e., no verbal, physical, or gestural initiations ). Videotapes were observed by the classroom teacher and a graduate student. Prior to the study, these observers rece ived training on sample tapes until they reached 80% agreement. During baseline and intervention conditions, interobserver reliability checks occurre d during 20% and 25% of the sessions respectively. Occurrence reliability was calcu lated on all three types of interactions and ranged from 89% to 100%. Using an A-B design, data were collected on the frequency of the participants social interactions during lunchtime. Du ring the initial 5-day baseline condition, the mean level of inappropriate social interacti ons was 34%. Level stability indicated that only 40% of the data points fell within 20% of the mean. However, the trend data of the last 3 data points was countertherapeutic, and the intervention was initiated. During the treatment condition, there was no immediate cha nge in either the level or the trend of
49 inappropriate social interactions. Data were variable and appeared to be moving in a contratherapeutic direction. However, on the fifth day of the interv ention, inappropriate social interactions began decreasing and contin ued to move in a decelerating trend for the remainder of the intervention (one exception poin t). Visual analysis of the data indicated that the mean level of inappropriate social interactions was reduced by 48% between the first (44%) and the last (23%) data points. Unfortunately, the level of appropriate social interactions did not show a clinically signi ficant change (occurring 6% of the time during baseline, and 4% of the time during the treatment condition). It appeared that the soci al story intervention had an effect on the participants social interactions during lunchtime. Howe ver, these results must be considered with caution for several reasons. First, due to the lack of experimental control associated with an AB design, there was no repli cation of the effect of the in tervention. Therefore, only a correlational relationship can be inferred. S econd, three social stories were used to address one behavior using varying content. Because numerous ideas were introduced in the stories, the participants ability to focu s on any one or two key points may have been jeopardized, resulting in possibl e confusion of what desired behaviors should have been practiced during lunchtime. These factors, combined with the qausi-experimental design utilized in the study, subj ect threats to both the internal a nd external validity of the study and, therefore, should be interpreted with caution. Theimann and Goldstein (2001) used a series of social stories and video feedback to increase the social communication skills of five children with autism. Each of the target children were paired with two normal developing peers as a s ource for comparison. These triads met twice a week for 30 minutes during each session. Each of the treatment
50 sessions were broken down into three segments : (a) a 10-minute social story instructional period, (b) followed by a 10-minute observation of social interac tions, (c) and a 10minute video feedback session. All of these sessions were audio and video recorded. These data were coded to measure increases in four primary dependent social measures: (a) securing attention, (b ) initiating comments, (c) initiati ng requests, and (d) contingent responses. Using a multiple baseline acro ss behaviors, data were collected on the estimated frequency for each of the depende nt social measures. These data were examined across the five students during the1 0-minute social intera ction. Inter-rater reliability was sufficient with a range acro ss subjects between 85% and 100%. This reliability occurred during 30% of all experimental sessions. Results of this study demonstrated incr eases in the social communication skills for all five participants. During intervention phases, each of the participants mean level of the four dependent measures increased. Overall, these findings support the use of social stories to increase the social development of children with ASD. However, theses results must be viewed cautiously for two reasons. First, this study employed multiple treatments (e.g., social stories, script-fad ing, self-evaluation) without examining the effects of each of these factors separately. Perhaps, results may have been better attained using a changing criterion design. Second, th e observation period of 10 minutes appears to be relatively brief given the social defic its of children with ASD. Longer observation intervals may have been more salient. Taken together, these limitations should not discriminate against the outcome data of the study. Specifically, this research did demonstrate the benefits of using visuals s upports within embedded text as an educative approach for children with autism.
51 Lorimer, Simpson, Myles, and Ganz (2002) examined the efficacy of a social story intervention implemented in a home setting to decrease the precursors to tantrum behaviors in a child with au tism. The participant was a 5-year-old boy diagnosed with mild to moderate autism. The participant was estimated to function in the average to above-average range of cogniti ve abilities with st rengths in his ability to communicate orally. Behavior problems in the home included tantrum behaviors characterized by screaming, hitting, kicking, and throwing obj ects. A functional assessment of these behaviors revealed that they were motivated by attention and attempts to gain a tangible reinforcer. The participant attempted to ve rbally express his wants and needs before displaying the challenging behaviors. Ho wever, the manner of communication the participant utilized was ineffective, characterized by shrill, loud, and perseverative commands (e.g., Listen to me! and Stop talk ing). These precursors to the tantrum behaviors were described as interrupting voc alizations, and if they were ignored or reprimanded, the participant escalated to tant rum and aggressive behaviors. The parents of the participant reported that the tantru ms occurred at least five times per day, and lasted from 45-minutes to 1- hours. Two social stories were created for the pa rticipant. Both stories were based on the hypothesis that reducing ineffective and in appropriate verbalizations would decrease the frequency of tantrums. Accordingly, the so cial stories addressed the participants (a) need for appropriately gaining attention from others who ar e talking and (b) waiting for the appropriate time to talk or ask for some thing. Each story identified these difficult situations through descriptive sentences, desc ribed the target/repl acement behavior (e.g., waiting appropriately, not yelli ng or hitting), and indicated the reinforcement earned for
52 engaging in the desired behavi or. In addition, each social story was designed using picture icons to accommodate for communi cation deficits of the participant. Dependent measures in this study included the participants fr equency of tantrum behaviors and frequency of inte rrupting vocalizations. Data were collected in the home by the participants parents, as well as during 45-minute home therapy sessions. The researchers made reliability checks during 33% of the observations. Prior to the study, these observers received training on how to identify and measure the two target behaviors. Average inter observer reliability was 96.10% throughout the intervention. Using an ABAB plus alternating treatm ents experimental design, observational data were collected on the fr equency of the participants tantrum behaviors both during treatment and in the absence of the intervention. Initial baseline observations collected over a 7-day period indicated th at tantrum behaviors occurred on 5 of the 7 days. During the first treatment phase, the participant e xhibited no tantrum beha viors on 6 of the 7 days. When the return to baseline conditi on was presented, the par ticipant had tantrums on 2 out of 3 days. During the second treatm ent phase, tantrum behavior again reduced. On 6 of 7 days, the partic ipant exhibited no tantrums. Changes also were observed in the part icipants precursor behaviors. These behaviors occurred several times a day duri ng the initial baseline period, but began to decrease when the social stories were introdu ced. Similar to the pattern observed in the participants tantrum behaviors, precursor behaviors dramatically increased during the second baseline period, when the participant did not have access to the social story. When the intervention was reintroduced, the participants precursor behaviors demonstrated a downward trend.
53 These findings suggest that implementation of the social stories were effective in reducing the frequency of tantru m behaviors of the participant. This study also provides a measure of control through its use of an alternating treatment-baseline design. However, a return to baseline frequency of tantrums upon removal of the intervention suggests a possible threat to in ternal validity. Due to the teaching natu re of the social story intervention gains in target beha viors should be maintained following discontinuation of treatment. As with the studies previously described, the generalizability of these results are limited bot h by its single subject design and failure to include any social validity measures rega rding the impact of the intervention. Summary A fundamental component of AS is having difficulty in social integration, yet researchers have only just begun to describe the nature of th e characteristic difficulties, let alone develop remedial programs to en courage greater competence in this area (Atwood, 2000). Clearly, there is an increasing demand for the development of intervention programs for children with AS due to the increase of the practice of mainstreaming in schools. However, little re search has been conducted to develop such strategies. It appears that although quite similar, individuals with AS appear to be more socially interested than th eir autistic counterparts. Th eir interest in maintaining friendships, having girlfriends /boyfriends, and social cont act may be quite striking (Volkmar & Klin, 2000), and is apparent through the years of schooling (Baur, 1996; Church, Alisanski, & Amanullah, 2000). Unfortuna tely, due to their repeated exposure to ridicule and torment for their odd and eccen tric behaviors, many individuals with AS
54 eventually withdraw and set aside their desire for friendship. Such experiences are one of the major contributors to a secondary diagnosis of mood disorders, most notably depression. However, motiva tion to acquire social rela tionships may be a powerful resource for interventionists who seek to impr ove the social competency of children with AS. Although many techniques apply to the treatment of individuals with autism, resulting effects tend to be more limited becau se of the more withdr awn and disinterested nature of the presentation of individu als with autism (Vol kmar & Klin, 2000). The presence of language and elevated c ognitive skills seen in children with AS can be effectively used to foster their social skill development. This allows for the future development of interventions to be more ri gorous in design, but addresses the area of deficit more directly. Through the use of soci al stories, children w ith AS can socialize and model appropriate behavior(s), increas ing their interactions with others and beginning to actively seek social relationships. Social stories have been reported to be an effective means of modifying the behavior of individuals with autism, such as how to greet people appropriately and how to share to ys (Swaggart, et al., 1995), how to monitor volume when singing with others (Fullert on, et al., 1996), how to reduce tantrum behavior (Kuttler, Myles, & Carlson, 1998; Lorimer, Simpson, Myles, & Ganz, 2002) and how to act appropriately during luncht ime (Norris & Dattilo, 1999). Unfortunately, all of these studied have emphasized the use of social stories on i ndividuals with autism (and other comorbidities, see Kuttler, Myles, & Carlson, 1998) with conc urrent deficits in language processes necessary fo r understanding social storie s. Therefore, the use of social stories with individuals with AS, who express no deficits in language abilities may have an even greater impact when systematically controlled.
55 With the increased use of the practice of mainstreaming, and with the prevalence of AS being forseeably higher than that of autism, it is necessary to begin researching approaches to help children with AS develop greater social interac tion skills. Throughout history, children who were percei ved as just being different or weird, have confused and bewildered clinicians, parents, and, mo re importantly, educators. Generally, the outcomes have been poor, with improper pl acements and misunderstood reasons for the behaviors seen in children with AS. With greater awareness of how AS may differ from autism comes greater knowledge in developing appropriate strategies of intervention. Thus far, the social story paradigm appears to hold promise as a logical and beneficial intervention to diminish the qualitative di fferences that make children with AS so conspicuous. However, research on social ski lls interventions with children with AS is scant, and research on social stories with children with AS currently is nonexistent Purpose of the Study The purpose of the current study was to examine the effectiveness of social story interventions for children with AS. This study expands the current body of research in this area by: (a) employing a means of experimental control; (b) probing for generalization through a follow-up phase; (c) inco rporating indices of treatment efficacy; and (d) examining the social validity of the intervention
56 Chapter III Research Methods Overview This chapter provides an outline of the method used to conduct the study. Specific descriptions of the participants, instruments, treatment protocol, and data analysis procedures are outlined below. Subjects The participants for this study were sele cted from the Child Development Center (CDC) located at St. Josephs Childrens Ho spital in Tampa, Florida. Prior to participating in this study, a sample pool of five children who were not currently receiving any kind of social skill treatment were targeted for participation. This sample pool was based on those individual s who had previously received social skills training at the CDC. To be included in this sample pool, potential participants were not to have received social skills training from staff at the CDC for a period of at least one year. Information about these potential participants was gathered through cl inical chart reviews and discussions with the lead clinician at the CDC. Information including the level of the childs social skill development and cognitive and verbal abilities, as well as parent involvement was used to narrow the sample pool to three children. Once these three children were identified, their parents were contacted and briefed about the study by the director of psychological services at the Child Development Center. If the parents
57 expressed interest in this study, the primary i nvestigator contacted them to provide more details of the study and answer any questions. Three elementary-age participants were selected based on multiple inclusionary criteria. The three target pa rticipants were boys ranging in age from 9 years 9 months to 11 years 6 months (X = 10 years 5 months). For inclusion in this study, the participants (a) had a current diagnosis of Aspergers according to DSM-IV, with current cognitive functioning levels in the aver age to above-average range, (b) possessed the ability to communicate orally with others, and (c) demons trated evidence of a potential benefit for the intervention program as indicated by a c linician, teacher, related support staff, or parent report. Prior to final inclusion in this study, acquisiti on of parental permission for participation and child assent was warranted. The decision of who to include in this study was based on the notion that social stor ies are most likely to benefit students functioning intellectually in th e trainable mentally impaired range or higher who possess basic language skills (G ray & Garand, 1993; p. 2). Darius Darius (age 10 years 1 month) attended fourth grade in a fully integrated private school. Information gathered through functional assessments revealed that Darius is a very compassionate child who enjoys playi ng video games, soccer, and football. Across all academic domains, Darius demonstr ated above average performance with no academic concerns. His teachers described Da rius as a good reader and gifted in math. Aside from these strengths, Darius was descri bed as having difficulty in group situations. Specifically, both parents and teachers stated that Darius become s extremely oppositional and negative when playing games (e.g., calls other child names, yells at teammates).
58 Such behaviors were viewed as socially immature and would not be conducive to maintaining friendships at school. Dire ct observations condu cted on the playground verified this information, and a social st ory to reflect this weakness was created (Appendix G). Francis Francis (age 11 years 6 mont hs) attended the fourth grad e in a private Catholic Diocesan school dedicated to meeting the need s of students with lear ning disabilities. Both the parents and teacher described Franci s as a sweet and sens itive boy with great creativity. Specifically, Francis enjoyed art, writing stories, a nd creating adventure games. Aside from his many talents, Franci s was at times socially immature and often said (e.g., inappropriate jokes) or did things (e.g., walk away from a conversation) that offended others at school. Observations conduc ted at Franciss school revealed difficulty with joining in activities. Specifically, Franci s was observed to want to join in with many other children on the playground, but he would often ask to pl ay then walk away. In addition, his eye contact was highl y infrequent and he did not abide to the hidden rules of being aware of another persons body space. A subsequent social story was designed to reflect these difficulties (Appendix H). Angelo Angelo (age 9 years 9 months) attended th e fourth grade in an integrated private school. Both his family and teachers described Angelo as a very scientific child who enjoyed experimenting. In particular, A ngelo was spending a great deal of time developing a juice modeled after ancient Egypt ians. Other interests included soccer, video games, and various cartoon characte rs (e.g., Spider Man, Sponge Bob Square
59 Pants). Main areas of difficulty for Angelo appeared to be spont aneously joining in activities. His parents stated that it takes a great deal of effort to get Angelo to join in any activity with other children and they often re sort to making deals with him. Angelos teacher verified this information, stating th at Angelo spends most of his time during recess alone. Both the parents and the teacher felt that Angelo would benefit from an intervention that focused on how to politely as k to join in an already existing game or conversation. Observations conducted duri ng recess at Angelos school verified the information gathered from Angelos family and teacher. Angelos social story focused on ways to join in activities and the benefits of making friends associat ed with joining in (Appendix I). Settings Observations of the three participants we re conducted at each of the participants schools during regular school hours. The prim ary setting in which these observations occurred was directly related to the identif ied behaviors targeted for the social story interventions. For example, in stances of joining in play ac tivities or conversations with other children were observed on the school playground. Respect for all of the school personnel was regarded and the primary investigator discusse d the procedures with those involved prior to any observations. For Darius the environmental setting wa s a fenced in area on the side of the school where there were several jungle gyms and an area to play an active sport (e.g., football). Around this area were sidewalks and breezeways were children could sit and play games or engage in conversation. During recess time, there were multiple grades on
60 the playground at once. Typical behaviors included a variet y of playing sports, climbing on the swing sets, and engaging in conversation under the breezeways. For Francis, the environmental setting wa s a courtyard in the middle of the school where there were several picn ic tables. Around this area was a sidewalk with an overhang roof. This area was used during the morning hours for recess/snack time. During this time, there were multiple classr ooms in the courtyard (first through sixth grade). Typical behaviors during the r ecess/snack time included walking around the courtyard engaging in conversation, eati ng the morning snack, or playing tag. For Angelo, the environmental setting was a large fenced in soccer field behind the school. This area also had a swing/slide set that the children were permitted to play on. Recess time occurred immediately following lunch and it was only Angelos classmates and one other fourth grade cla ss. During this time, the children were permitted to play soccer, play on the swing/s lide set, or walk around and talk. Typical behaviors included playing soccer. In fact, mo st of the children played soccer each day during recess. Angelos cla ss would often pick teams in the morning prior to going to recess. Materials and Measures Social story book Three social stories were designed addressing an identi fied target behavior for each participant. Each social story book wa s designed according to the criteria used by Norris and Dattilo (1999). Thes e social stories were indivi dualized print books that the students could carry to and from school a nd use in multiple settings (e.g., library, cafeteria, playground). Each book was constructe d on 6 inch (width) by 8 inch (length)
61 pieces of paper. Including the cover page, there were be five to nine pages in each story that was laminated and fastened together along the left marg in. The cover page of each social story contained only the title, whic h was placed inch from the bottom of the page and typed in 14-point Times New Roman font. The remaining pages contained one or two sentences typed in 14-point Times Ne w Roman font and prin ted near the bottom of each page. There were inch margins on th ree sides of each page (the margin on the bottom of the page was set at 0 to allow for a ppropriate room for the sentences). This spacing left approximately a 5 x 7 area above the sentences where color 1 x 1 Mayer-Johnson picture symbols were placed (Mayer-Johnson, 1994). The Mayer-Johnson symbol library is a nationally accepted and fundamental component of educational programming fo r children with communication disorders across the United States of America. The system, available in either hard copy or through computer database (Board Maker A pplications, 1989), provides visual symbols for a variety of types of words (e.g., nouns, ve rbs, adjectives) in the English language. Mayer-Johnson symbols were used to empha size and communicate the message of each social story. Although children with AS po ssess average language skills, the MayerJohnson symbols were used to make each social story more appealing for the reader. Observational data forms Two separate observational data forms (i.e ., frequency of joining in, percentage of appropriate social interaction skills) were used to gather da ta to test hypotheses. Both forms were developed by the primary invest igator and were modified based on expert opinion and sample (pilot) observations.
62 Social Interaction Recording System (SIRS). Data on the frequency of identified target behaviors were gathered using a part ial interval recording procedure to measure estimated frequency (i.e., number of intervals of occurrence) and per centage of intervals of active and passive occurrence of the targeted social engagement be havior (i.e., joining in, initiating a conversation). A sample of th e SIRS is provided in Appendix J. The SIRS provides for 100-observational sequences, each of which are divided into a 10-second observation and a 5-second recording period, re sulting in a total documentation period of 20-minutes per form. Observers using the SIRS first observed for a 10-second interval. The observers then had 5-seconds to record the behavior th at the target child was engaged in at the conclusion of the 10-second interval. The obser vers recorded the target childs behavior as active, passive, or absent. The number of total intervals in which each type of behavior occurred was summed for the entire observation period. This number was then be divided by the total number of observati on intervals (50-100), ge nerating a percentage of intervals in which the target child or peer comparison was actively or passively socially engaged. The percentage of interv als calculated provides an estimate of the frequency of social engagement. Observation of Appropriate Socia l Interaction Skills (OASIS). Additional observational data were collected by observers using the OASIS. The OASIS is a simple 15-item checklist that the observer used to ga ther data for each participant relative to the frequency of desired soci al interaction sk ills (e.g., eye contact, body basics, conversational skills). A copy of this checkli st is provided in Appe ndix K. The form was completed by placing a check or tally mark in the appropriate column when one of the
63 stated behaviors occurred. Totals and rati os of yes/no responses were calculated during each observation to demonstrate any change in frequency for each participants use of appropriate social interaction skills. This checklist also had space to document the students name, where the child goes to school the students gender, and the date of the data collection. Pre, Post-Test Measures Social Skills Rating System Parent Form, Elementary Level (SSRS-PE; Gresham & Elliott, 1990). The SSRS-PE is a 55-item social sk ills rating scale that measures two domains of social competence : (a) social skills and (b) problem behaviors for children between Kindergarten and sixth grades. The Social Skills Scale is comprised of five subscales: (a) cooperation; (b) assertion; (c) responsibility; (d) em pathy; and (e) selfcontrol. Items are rated first on how often th e behavior occurs (ranging from 0 = never to 3 = very often). Items are then rated as to how important these behaviors are to the rater (ranging from 0 = not important to 3 = critical). The Pr oblem Behavior scale contains three subscales: (a) externalizing problems, (b) internalizing problems, and (c) hyperactivity. Reliability and validity of the SSRSPE indicates adequate to excellent psychometric properties when used for indi vidual educational a nd home interventions with students suspected of having significant so cial behavior problems. Across all forms and levels, the median coefficient alpha reli ability for the Social Skills Scale was .90, while it was.84 for the Problem Behaviors Sc ale. The subscale internal consistency estimates for the parent-elementary form ranged from .65 to .87 for Social Skills and from .71 to .87 for Problem Behaviors. Due to the small number of items on each
64 subscale (typically 6 to 10 items), the intern al consistency coeffici ents may have been compromised. Test-retest reliability of the SSRS-PE demonstrated high stability over a 4-week period with 45 subjects. Parent correl ations were .87 for Social Skills and .65 for Problem Behaviors. Social Sk ills subscale reliability coeffi cients ranged from .77 to .84, while the Problem Behaviors subscale reliabilities ranged from .48 to .72. The SSRS-PE was administered with the Child Behavior Checklist-Parent Report Form (CBCL-PRF) to a sample of 45 elementary-aged children drawn for the standardization sample. The Problem Beha viors scale on the SSRS-PE correlates .70 with the corresponding scale on the CBCL-PRF. However, the Social Skills scale on the SSRS-PE correlates .58 with the corresponding scale (Socia l Competence) on the CBCLPRF. However, this lower estimate is likely when considering the relative distinctiveness and ease when observing disruptive behaviors in children as opposed to observing less distinctive interna lizing behaviors. For purposes of this study, the SSRS-PE se rved as a broad measure to assess any change in behaviors that parents could reco rd. Because this study dealt with higherfunctioning children, it was hypot hesized that they may alr eady possess some appropriate social skills. However, the degree to whic h the participants used these skills was unknown. The SSRS-PE was used as a measure to examine if the intervention has any effect on other social skills beyond those identified target behaviors. Social Skills Rating SystemTeacher Form, Elementary Level (SSRS-TE; Gresham & Elliott, 1990). The SSRS-TE is a 57-item so cial skills rating scale that measures three domains of social competence: (a) social skills, (b) problem behaviors and (c) academic competence for children between for Kindergarten and sixth grades. The
65 Social Skills Scale and the Problem Behavior scale are identical to those described in the SSRS-PE. The Academic Competence domain is unique to the teacher form and is concerned with the students academic f unctioning. The Academic Competence Scale consists of 9-items on a five-point Likert sc ale that corresponds to percentage clusters (i.e., 1 = lowest 10%, 5 = highest 10%). Reliability and validity of the SSRS-TE also indicates ade quate to excellent psychometric properties when used for indi vidual educational deci sions with students suspected of having significant social be havior problems. The subscale internal consistency estimates for the teacher-element ary form ranged from .86 to .94 for Social Skills, from .78 to .88 for Problem Behaviors, and .95 for Academic Competence. Testretest reliability of the SSRS-TE demonstrat ed high stability over a 4-week period with 288 subjects. Teacher correlations were .85 fo r Social Skills, .84 for Problem Behaviors, and .93 for Academic Competence. Social Ski lls subscale reliability coefficients ranged from .75 to .88, while the Problem Behaviors su bscale reliabilities ranged from .76 to .84. The Academic Competence Scale demonstrat ed test-retest reliability of .93. The SSRS-TE was administered with th e Child Behavior Checklist-Teacher Report Form (CBCL-TRF) to a sample of 99 elementary-aged children drawn for the standardization sample. The SSRS-TE and the CBCL-TRF externalizing scores were highly correlated ( r = .75), as were the SSRS-TE Probl em Behaviors Total score and the CBCL-TRF Total score ( r = .81). Internalizing scores on both measures correlated .59 with each other. However, the hyperactivity scales of both measures demonstrated a high correlation ( r = .77).
66 Like the SSRS-PE, the SSRS-TE was used as a broad measure of prosocial development from the teachers perspective. The SSRS-TE was used as a measure to examine if the intervention had any additional effects on the development of participants prosocial behaviors in both classroom and school settings. Childrens Atypical Development Scal e (CADS; Guevremont & Dinklage, see Barkley, 1990). The CADS is a 53-item rating scale designed to measure four areas of unusual behaviors in clinic-referred childre n from ages 4 to 16. The domains rated include: (a) Communication Defi cits, (b) Lability, (c) Soci al Relatedness Deficits, and (d) Preoccupation. Items are rated as to how relevant these behaviors are to the rater (ranging from 0 = not true to 2 = very true of often true). The CADS is a useful instrument for distinguishing atypical beha viors associated with PDD, and has been recommended for use as both a screening and a preand post-intervention measure. Although this scale was never standardize d, it appears to be well constructed and has been demonstrated in a p ilot study to be a useful s cale as a means of assessing clinically significant symptoms of affective disorders in clinic-referred children. The item content of the CADS was generated by f our experienced clinicians who compiled an extensive list of unusual behaviors ba sed on Cohen, Paul, and Volkmars (1986) classification criteria for PDDs and the me dical records of 40 ch ildren diagnosed with PDDs. A pilot study was conducted with parent s of 23 children clinic ally diagnosed with PDD, and 20 children with ADHD. Both groups had normal intelligence and were referred to the same medical center. Two groups of student s with learning disabilities (LD) served as a control. Parents were asked to rate each item on a 3-point scale. Results of this pilot study i ndicated that items were most characteristic of the PDD group,
67 endorsed to a lesser extent by parents of ADHD children, and ra rely reported for the LD and typical control groups (Barkley, 1990). More recently, the psychometric properties and factor structure of the CADS were assessed. Reliability and validity of the CADS indicates adequate to excellent psychometric properties when used with cl inic-referred childre n. The Communication, Lability, and Social Relatedness domains dem onstrate internal consistency with alphas ranging from .80 to .87. However, the Preoccu pation domain does not show high internal consistency (Cronbachs alpha of .53). It is important to note that this domain only consists of two items which may compromise the alpha level. Estimates of internal consistency corrected for the scale length using the Spearman -Brown formula resulted in a corrected Cronbachs alpha of .82 for the Preoccupation domain. Concurrent validity of the CADS was asse ssed with standardized parent ratings. Parents of the children were asked to comp lete the CBCL and the Conners Parent Rating Scale (CPRS). CADS scores correlated signifi cantly with the parent report measures including virtually a ll the subscales on the CPRS (rangi ng from .22 to .57), as well as on both the Internalizing (co rrelations ranging from .38 to .50) and Externalizing (correlations ranging from .21 to .63) scales of the CBCL. Although a number of rating scales and questionnaires are available which assess behaviors associated with autism (e.g., Child Autism Rating ScaleCARS; Schopler, Reichler, & Renner, 1988), these scales may be less applicable fo r higher-functioning children. In addition, most of these scal es do not adequately sample odd or unusual behaviors commonly seen in higher-functioning ch ildren with autism spectrum disorders. The CADS is a scale that appears to be particularly applicable to sample unusual
68 behaviors observed in higher-functioning childr en with autism spectrum disorders and, in this study, served as a preand post-interv ention measure for both parents and teachers. Social Story Journal Due to the highly individualized nature of social stories implementation, a journal was kept by both the participati ng children and thei r parents. This journal allowed the primary investigator to assess if the social story had been implemented according to the recommendations of Gray (1994) and Gray a nd Garand (1993). This form was used to keep a running record of the participants progress and/or any difficulties with the story. Specifically, each journal entry assessed where and with whom the social story was read and the childs reaction to th e social story (see Appendix L) Such information provided a level of treatment integrity for the social story intervention. Dependent Measures The primary behaviors that were identified for each participant related to social engagement (e.g., sportsmanship, joining in). Specific target beha viors were identified through functional assessments with each target child. First, interviews with parents and teachers were conducted. The focus of these interviews related to areas of concern or weakness for each child, as well as inform ation on the relative frequency of the behaviors. Once a specific behavior was id entified for each participant, detailed information regarding those behaviors was gathered prior to the study through observations of the situations in which the be haviors were likely to occur and interviews with relevant individuals (e.g., parent(s), teacher(s)). Information collected included relevant cues of the behavior, the typical sequence of events th at occurs prior to and after the behavior, and descriptions of the beha vior gathered from those involved in the
69 situation. This information was essential to understanding what would be important to include in the social stories, as well as what information was irrelevant. Dependent measures included the estimat ed frequency (i.e., percentage of intervals) of active and passive target be haviors for each individual, as well as the absence of the identified target behavior. Specifically, the dependent measures for this study were joining in (Francis and Angelo) and sportsmanship (Darius). Joining in Active Joining In (AJI). AJI was defined as instances in which the target child was contributing to a reciprocal conve rsation or is actively partic ipating in some play activity with one or more children (see Appendix M). Examples of AJ I were defined as children playing next to each other and using a vari ety of social exchanges to show their recognition of peers, such as talking to a pe er about a game that is being played on the playground. During AJI, the child was obser ved borrowing or lending toys, using each others toys, or sharing accomplishments (e.g., successfully building a tower with blocks). AJI also included any type of organized group game (e.g., tag, hide-and-seek) that involved taking turns and sharing identification with th e group. AJI did not include any sort of aggressive act such as curs ing, shouting, pushing, name calling, hitting, and making forceful bodily contact with someone else during a play activity. Passive Joining In (PJI). PJI was defined as instances in which the target child was playing next to peers, using the same toys or engaged in the same activity, but playing independently of those peers (similar to parallel play). Th is also included times when students were engaging in conversation, ra ther than play. Examples of PJI included listening to a group of peers while sitting among them, or watching a group of peers
70 playing a game and following the group format alone. PJI did not include aimlessly walking around a peer group talking to himself or herself, silently reading while peers were actively playing around the target child, or the target child not sharing any of the toys with those around him/her. Absence of Joining In (AbJI). AbJI was defined as instances in which the target child did not display any verbal physical, or gestural initiati ons or responses to peers. AbJI included the target child engaged in a game or other activity by himself, not responding to his or her name being called by another peer, refusing to share toys on the playground, and any other behaviors that were observed as ignoring pe ers and/or adults. The child was not involved with other chil dren and appeared distant from activities occurring around the participan t. Any occurrence of fighting, name calling, and making forceful bodily contact with someone else during a play activity was labeled as AbJI. In addition, data regarding the qualitative aspect s of such negative behaviors was monitored. Sportsmanship Active Sportsmanship (ASP). ASP was defined as inst ances in which the target child actively treated teammates, opponents, an d/or coaches with respect (see Appendix N). Active displays of sportsmanship demons trated an awareness of encouraging group play through verbal, physical, or gestural means. ASP incl uded instances when the target child offered positive encouragement during a play activity to other children on the playground (e.g., saying good job, way to go, awesome, good luck, maybe next time, good effort). In addition, ASP include d times when the target child cheered for opponents or own team members after a good play or score (e.g., clapping when a team
71 scores, telling the othe r team good play). ASP did not include any behaviors such as yelling at teammates or ot her players during the game. Passive Sportsmanship (PSP). PSP was defined as instances in which the target child attended to the parameters of a game/activity with a full commitment to participatory cohesion. Passive displays of sportsmanship demonstrated aspects of fairness and playing in an a ppropriate manner. For example, following and/or playing within the rules of the game (e.g., not cheating, not intentionally hurting an opponent or other team member) and accepting the loss of the game without making negative comments to self or others would be exam ples of PSP. PSP woul d not include any acts of verbal or physical aggre ssion, nor any instances in whic h the child made excuses for the game progressing poorly. Absence of Sportsmanship (AbSP). AbSP was defined as instances in which the target child did not display any verbal, physical or gestural sportsmanship initiations or responses to peers nor display equality a nd fairness in his play. AbSP included any negative comments or sarcastic remarks to opponents or own team members (e.g., complaining about team members performan ce, talking down to other players, laughing when someone gets hurt) or any aggressi ve actions towards opponents or own team members (e.g., running after opponent who just scored, throwing the ball away from players when other team scores). Instances in which the target child was quiet and played within the rules of the game would not be an example of AbSP. AbSP essentially was poor sportsmanship (e.g., engaging in nega tive behaviors during a sport activity).
72 Procedure A multiple baseline across subjects experimental design was used to analyze the data and test the stated hypotheses. In addition, a follow-up observation occurred two weeks after the completion of the interventions. Such a design allowed for: (a) initial demonstrations of the controlling effects prior to an intervention; (b) multiple demonstrations of the controlling effects of the intervention; and (c) an assessment of the intervention effects at follow-up. Parental permission for participation in this study was obtained for these three individuals identified in the process explaine d previously. A meeting in the families home was then scheduled to begin the proce ss of identifying potential targets for the intervention. In addition, parents were given a consent form. which provided a description of the study, the benefits and risks of the st udy, information regarding new therapies, costs of th e intervention, and the rights of th e parents and children involved in this study. Parents who agreed to partic ipate were given the SSRS-PE and CADS pretests to complete and mail to the prim ary investigator within two weeks of the orientation. In addition to parental consent, teachers who were dir ectly involved with the participants were debriefed about the intervention, asked for their consent, and also informed to complete the SSRS-TE and CADS pretests. Observer training Following obtaining consent information from the parents and teachers, each of the observers that volunteered to assist with data collection were trained. The primary and secondary observers were graduate students who have received training on observational methods, as well as behavioral de finitions. The observers all had at least
73 one year of experience in behavioral observa tions. In addition, obs ervers were trained using the observational record ing device designed by the primar y investigator prior to the initiation of the study. Unfortunately, vide os were unable to be used for training purposes due to wishes expressed by the pare nts. Therefore, examples and nonexamples of the target behaviors were provided to th e observers in training, as well as role-plays that the observers coded. Specifically, observe rs were trained with written, verbal, and modeled examples of the dependent measures until they reached 80% agreement using the following formula: Agreements Agreements + Disagreements X 100 Baseline period After obtaining the appropriate agreemen t criterion during training, the observers began to collect baseline data for all of the participants in this study. During the baseline condition, observational data were recorded for each participants targeted behaviors prior to the introduction of the social story. No intervention occurred prior to or during this period. The SIRS was used to measure the frequency (i.e., the number of intervals of occurrence) and percentage of intervals of occurrence for each of the identified behaviors (e.g., joining in with others). If any of the targeted behaviors occurred during a 10second interval, the observer recorded the appropriate res ponse (e.g., active or passive) on the data collection sheet. If there was an absence of the target ed behavior (fighting, hitting, screaming) during any of the 10-sec ond intervals, the obser ver recorded a mark on the accompanying data sheet. In addition, any observational notes (e.g., what the
74 target child said) were provide d on the data sheets. All baseline observations occurred three times per week and were, on average, 15 minutes in length. Baseline observations occurre d for three weeks (or 9 da ta points). Such an extended baseline allowed sufficient time to ensu re a stable baseline relative to the first participant. Stability was defined as a relativ ely constant rate of behavior, or behavior that clearly demonstrated a countertherapeutic effect (consecutive decelerating or flat trend), allowing for the proper implementati on of the intervention phase. In addition, those individuals involved with the particip ant (i.e., parents, teachers) were shown the baseline results and asked if such trends were representative of these behaviors each day. Prior to implementing the intervention, observations also were conducted during unstructured activities (e.g., school playground) to ascertain th e median level of social interactions in which typical peers engage. Colle ction of peer comparison data used a Round Robin approach for every fifth interval on the SIRS data sheet. That is, during every fifth interval, data collectors selected the first comparison peer they were able to observe and recorded his/her behavior. This means that during each comparison peer interval a different child was observed. Th ese data were collected during both baseline and intervention phases. Peer comparison data were collected because little is know n about how often social engagement occurs for a typical child. When examining higher-functioning individuals, who may already possess some appropriate skills, it becomes difficult to determine the effectiveness of the intervention. This information provided the comparison for future observations of targ eted students and allowed for a direct
75 comparison of the effects of the social stor ies on increasing the so cial engaged time of children with AS. Intervention phase Three social stories were formatted accord ing to Grays (1994) recommendations. Ideas for the content of the social stories were derived from base line observations and parent and teacher rating scales. In additi on, input from both the parent(s) and teachers was gathered through a semi-structured inte rview using the Social Story Information Form (see Appendix D). This form consists of opened ended questions designed to gather general academic information (e.g., grad e, reading level, comprehension), as well as information on the strengths of the child (e.g., interests and speci al abilities). In addition, this form allows for a functional beha vioral assessment of ta rgeted situations to occur. All of this information was used to increase the relevancy of the information presented in the social storie s with target behaviors. After the social stories were constructed for each of the participants, information regarding the most appropriate manner to implement the social story was gathered. Such information came from the Social Story Im plementation Plan (Gray, 1994) (see Appendix E). This form allowed the family and the pr imary investigator to: (a) identify methods to introduce the social story; (b) identify support materials and/or activities that can be used for the social story; and (c) identify factor s that may affect the implementation of the social story. The primary investigator took the responsib ility of establishing and maintaining rapport with parents and teachers, as well as organizing social stories around the stated deficits and recommendations for each child. Ea ch social story was re ad and/or reviewed
76 two times per day. Intervention data were coll ected in the same manner as baseline data. Observations continued to occur three tim es per week and were approximately 15 minutes in length. Implementation of the social story inte rventions progressed according to Ferron and Joness (2002) recommendations. Specifica lly, the primary investigator plotted the baseline data for each participant and decided that when the data had stabilized, intervention with the first participant would begin. The observers continued to collect data. However, the observers were not told which participant was selected for the intervention. The primary investigator cont inued to monitor the data, and implemented the initiation of treatment with the second pa rticipant when increasing data trends were observed and maintained for at least three data points (one week) among the first participants data. Again, the observers we re not told which participant entered the treatment phase. The same process was used for the implementation of the intervention for the third participant. That is, once the data for th e second participant had shown increasing trends or stabilized, as defined by at least 3 data points (one week), the primary investigator initi ated the intervention. During the intervention phase of the st udy, the child and the parent were the primary persons responsible for reviewing and reading the social stor y each day. At the end of each review session, the child and/or parent placed a sticke r indicating that the social story was read and re viewed (e.g., date and time) on the journal sheet provided by the primary investigator (see Appendix I). In addition, every two weeks, the parents and the primary investigator discussed how the intervention was progressing and if a ny modifications were needed. The parent
77 provided the investigator with information on how the child reacted to the social story, any problems with the story, and made suggestions to improve the story by using the Social Story Report Form (Gray, 1994) (see Appe ndix F). This form was used to keep a running record of the particip ants progress and/or any di fficulties with the story. Specifically, this form allowed for documentation of the participants reaction to the social story, the childs current resp onses in the targeted situation, and suggestions/recommendations for revising the so cial story. Such information provided a better understanding of where a social story was weak and in need of improvement. Such conditions help improve the likelihood of having a successful intervention. Follow-up phase Following the intervention phase, all social stories were faded over a period of one week. That is, each part icipant was instructed to read his or her re spective social story with less frequency until the story was not read at all. Each family developed their own system of fading the social stories. At two-weeks following the intervention (one week after the social story was faded), observa tions were conducted at each participants school. Follow-up data were collected in th e same manner as base line and intervention data. Observations occurred three times during the week and lasted approximately 15 minutes. Inter-rater reliability Inter-rater agreement for observations was calculated to assess reliability of the data. The same formula to cal culate agreement during initia l observer training was used. Inter-rater reliability was calculated by di viding the number of ra ter agreements (the occurrence or nonoccurrence of each behavi or) by the number of agreements plus
78 disagreements (the two raters did not agree on the occurrence of a particular behavior) and then multiplied by 100. These reliability checks occurred during 20% of the baseline condition and 25% of the intervention conditi on. Observations were considered reliable if at least 80% inter-rater agreement was achieved for each observation. Inter-rate agreement was consistently above 80% for a ll participants. Agreement ranged from 87% to 100% for participant 1 (Darius) (X = 94%), 82% to 94% for partic ipant 2 (Francis) (X = 87%), and 80% to 84% for partic ipant 3 (Angelo) (X = 82%). Data Analyses Observational data. Observational data collected by the SIRS were used to test the first, second, and third hypotheses. With regard to the first hypothesis: Participants in the social story interven tion will show an increase in the occurrence of identified target behavior(s) in unstructured natural se ttings following treatment, as compared to baseline condition descriptive visual analyses were conducted. Three visual inspection methods were employed for analyzing changes in the frequency of each targeted social engagement behavior for each participant. A ll visual inspection analyses considered both the changes in individual partic ipants data, as well as comp aring changes across each of the participants. First, the average daily pe rcentage of active, passive, and absence of social engagement was graphed for each par ticipant for the entire c ourse of the study. Graphs display the day-to-day performance of each participant. Data relative to percentage of social engagement was analyzed to examine changes in each participants mean daily percentage of active, passive, and absence of social engagement during baseline, intervention phases, and at follow-up. A mean trend line for each phase was plotted as a solid line in each phase for each participantproviding a summary line.
79 Percentages of participants so cial engaged time were also compared to peer comparison data. If participants mean data approximated the average social perf ormance of peers, the mean trend data were considered clinically significant. Second, data were analyzed to examine changes in the level and slope (i.e., trends) of the daily percentages for each participant during baseline, intervention, and follow-up phases. Changes in level were co mputed using the least squares technique within each phase (calculated by Excel). Fo r each phase, celeration lines were drawn. The slope of each line and the initial and final level of each phase was expressed numerically. The change in level was estimat ed by comparing the last data point in the baseline and the first data point in the inte rvention phase. The smaller value was then subtracted by the larger value yielding the change in level across conditions. Similarly, for a change in slope, the smaller slope wa s subtracted by the la rger slope yielding a change in slope across conditions. Examini ng changes in level and slope summarized the differences in performance across ph ases for each of the participants. Slopes that indicated positive improvement in behavior were considered significant. Because the range of social functioning in children with AS is generally low, any improvement in social engagement was cons idered a success. However, the range of success was hypothesized to be hi ghly variable. Therefore, data that revealed a clear indication of an increasing tre nd was considered significant. Finally, observational data were anal yzed to examine the percentage of overlapping data points for each participant be tween baseline, intervention, and follow-up phases. Overlap was computed by comparing data points in the intervention phase with baseline standards and with follow-up to inte rvention phases. A level of less than 70%
80 overlapping data points was consid ered significant. This level was chosen due to the fact that it was expected that much of the data would be highl y variable. Because social engagement and interactions are such a broad sk ill, they were not expect ed to be linear. In addition to these data, information from the OASIS was used to examine any collateral effects of the social story interventions. Data from the OASIS was used to determine the mean percentage of increase in socially appropriate be haviors for each of the participants. These data came from th e yes/no ratio on the OASIS. Data were presented by demonstrating the mean change in percent of appropriate skills, range of the percent of appropriate skills, and correlation with the occurrence data for each participant. Peer comparison data was used to test the second hypothesis: Participants in the social story intervention will demonstrate c linically significant increases in social interactions/social communication Each participants da ta was compared with a comparison peer during the observations using the SIRS. This information was then used to graph the percentages of target behavior s for each participant compared to peers. Percentages of participants so cial engaged time were compared to peer comparison data. If participants mean data approximate the av erage social performance of peers, the mean trend data were considered clinically significant. With regard to the third hypothesis: Participants in the social story intervention will maintain positive treatment effects at a 2-week follow-up the visual analyses described in detail previously were perf ormed. Specifically, follow-up phases for each participant were examined for change in mea n, change in level and slope, as well as the percentage of overlapping data points.
81 Pre-, post-test data Data pertaining to the four th and fifth hypotheses will be tested by investigating for a statistically relia ble change in pre and post-test data. The SSRS and CADS preand post-te st composite and domains sc ores for each participant are presented in a table. Sp ecifically, the fourth hypothesis: Participants will demonstrate a decrease in unusual behavior pat terns as rated by parents and teachers on pre and post-test measures was examined by calculating the reliable change index (Jacobson & Truax, 1991) for each subjects da ta on the SSRS to determine whether the magnitude of the change is statistically relia ble. The test-retest reliability data for the SSRS composite scores was used to calcula te the reliable change index for each participant. In addition, the fifth hypothesis: Participants will demonstrate a decrease in unusual behavior patterns as rated by parents and teachers on pre and post-test measures was examined by calculating the reliable change index for each subjects data on the CADS to determine whether the magnitude of the change is st atistically reliable. The test-retest reliability data for the CADS composite scores was used to calculate the reliable change index fo r each participant.
82 Chapter IV Results Overview This chapter presents the data relative to each research hypothesis in both written and visual presentation format (Figures 1 thr ough 6). Specifically, for the first, second, and third hypotheses, visual an alyses of observational data ar e presented. For the fourth and fifth research hypotheses pre and post-test data analyses are presented. The chapter concludes with a summary of the findings from the data analyses. Participants Social Communicati on Progress: Observational Data With regard to the first hypothesis: Participants in the soci al story intervention will show an increase in the occurrence of id entified target behavior(s) in unstructured natural settings following treatment, as compared to baseline condition the following results were found through visual analysis of data obtained by the Social Interaction Recording System (SIRS). Because there were three behaviors that were examined, the results are separated into categories ba sed upon the dependent measures for each participant. Active Display of Target Behavior Figure 1 provides a graphic display of th e percentage of active target behaviors (sportsmanship and joining in) for each pa rticipant across baselin e, intervention, and follow-up phases. According to the multiple baseline design, these data show increases
83 in social engagement for two of the three participants following the initiation of the social story interventions. While the mean data relative to Darius and Francis during the intervention phase demonstrated substantial improvement, data for Angelo are highly variable and should be interpreted with caution. In addition, the initial effects of the data demonstrate that Darius and Francis mainta ined an elevated performance of active behaviors across time. A detailed descripti on of the visual analys is conducted on active displays of target behaviors for each participant follows.
84 Figure 1. Percent of Active Displays of Targ et Behavior Across Participants 0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 1001 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 Days Baseline Intervention Follow-Up Darius Francis Angelo Percent S p ortsmanshi p Percent Joinin g In Percent Joinin g In
85 Darius. Prior to intervention phase, Darius displayed a somewhat variable, but relatively consistent baseline of low rates of active displays of sportsmanship as indicated by the decreasing trendline (see Figure 1). The overall mean percentage of active displays of sportsmanship for Darius duri ng baseline was 8.67%. Upon implementation of the intervention, a prompt increase in the daily percen tage of active displays of sportsmanship emerged. Specifically, from the last data point during the baseline period (13% active sportsmanship) to the initial da ta point during the inte rvention phase (20% active sportsmanship) there is a 7% increase in active rates of behavior. This demonstrates the relative rapidity of change in behaviors with the implementation of the social story intervention for Darius. Although this degree of change was not replicated with successive data points, this increase pe rsisted variably thr oughout the intervention phase. Specifically, Darius averaged 17.95% of active sportsmanship during the intervention phase, an increase of approximately 9.28% compared to baseline. At followup, Darius maintained an elevated perfor mance of active behaviors across time. Specifically, mean percentage of active spor tsmanship for Darius during follow-up phase was relatively stable at 11.33% (2.66% higher than baseline). Descriptive statistics relative the changes in mean are provided in Table 5. Table 5. Descriptive Statistics for Ac tive Sportsmanship for Darius X SD Min. Max. Baseline 8.67% 4.21 0 15 Intervention 17.95% 5.56 5 26 Follow-Up 11.33% 2.52 9 14 Celeration lines, or trendlin es, are indicated by solid li nes through data points for each phase as shown in Figure 1. Trendlines depict the rates of behavior increase or
86 decrease over time. Prior to the implementation of the soci al story intervention, Darius demonstrated a decreasing trend of active di splays of sportsmanship (slope = -.04). During the intervention phase, the direction of Dariuss behavior changed. Specifically, the trend of active displays of sportsmanship for Darius maintained a consistent, slightly increasing trend (slope = .04) across the interv ention period. When compared to baseline, this trend was both elevated and relatively c onsistent across time. Trendlines were not calculated for follow-up conditions because not enough data points were available to calculate an accurate trend. Table 6 highlight s the slopes of each trendline, as well as calculations of slope and le vel change across conditions. Table 6. Slope and Level Change for Active Sportsmanship for Darius Condition Slope Change in Slope Level Change in Level Baseline -.04 13 (@ last day) Intervention .04 .08 20 (@ first day) 7 Intervention --18 (@ last day) Follow-Up ----9 (@ first day) 9 Although data for Darius during the inte rvention phase was highly variable, the percentage of overlap of in tervention with baseline was only 25%. The percent of overlapping data points during the follow-up period was 100% when examining overlap with the intervention and baseline phases. This suggests that duri ng the follow-up phase, mean performance remained elevated, but d ecreased to levels that were previously observed during the baseline period. Francis. Francis showed a somewhat increasin g baseline trend of active displays of joining in prior to treatm ent with an overall mean percentage of 30.14% (Figure 1).
87 However, determination to begin implementa tion of the social st ory intervention was based upon the decreasing number of days availa ble to collect data prior to the end of the school year. Upon implementation of the inte rvention, an abrupt in crease in the daily percentage of active displays of joining in occurred. Specifically, from the last data point during the baseline period (28% active joini ng in) to the initial data point during the intervention phase (67% active joining in) there is a 39% increase in active rates of behavior. This replicates the rapid change in behaviors following the implementation of the social story intervention demonstrated with Darius. However, th is rapid change was followed by a rapid decrease in rates of active responding only to be followed by consecutive increasing data. An alysis of the anecdotal data revealed that during this second observation it was raining and the chil dren were only permitted to play on the covered sidewalk between classrooms. Th erefore, environmental factors beyond the control of this study may have contributed to the low rates of behavior seen on this day. During the intervention phase, Francis averaged 58.29% of active joinin g in. This change represents an increase in the use of activ e joining in behaviors of 28.15% compared to baseline. In addition, Fran cis demonstrated maintenance of skill acquisition during follow-up as indicated by a mean percentage of active joining in behavior of 41.5% (11.36% higher than baseline). However, it s hould be noted that fa irly high variability was present throughout all phases (see Figure 1). Descriptive statistics relative the changes in mean across time for Francis are provided in Table 7.
88 Table 7. Descriptive Statistics for Active Joining In for Francis X SD Min. Max. Baseline 30.14% 13.18 11 50 Intervention 58.29% 21.87 12 83 Follow-Up 41.50% 13.44 32 51 Changes in trend are depicted in Figure 1 as solid lines draw in each phase. For Francis, a steadily increasing trend was dem onstrated during the ba seline phase (slope = .51). Although the trend of the data for Fr ancis was increasing, im plementation of the intervention progressed due to time constraint s. During the intervention phase, the trend of active displays of joining in for Franci s maintained an incr easing trend across the intervention period (slope = 1.10). In additi on, the slope of the intervention phase trend was greater than that demonstrated during base line. No trend data were examined for the follow-up phase due to the lack of data. The slopes of each trendline, as well as calculations of slope and level change acro ss conditions for Francis are displayed in Table 8. Table 8. Slope and Level Change for Active Joining In for Francis Condition Slope Change in Slope Level Change in Level Baseline .51 28 (@ last day) Intervention 1.10 .59 67 (@ first day) 39 Intervention --60 (@ last day) Follow-Up ----51 (@ first day) 11 Although data for Francis during the inte rvention phase was highly variable, the percentage of overlap of in tervention with baseline was only 36%. The percent of overlapping data points during the follow-up period was 100% when examining overlap
89 with the intervention phases. During the fo llow-up phase, the percent of overlap was 50%. These data suggest that during the follow-up phase, mean performance remained above baseline for one data point, but decreased to levels that were previously observed during the baseline period. Angelo. Angelo demonstrated a highly vari able baseline with high and low rates of active joining in behaviors averaging 18.32% Baseline data demonstrated a countertherapeutic trend as indicated by the negativ e slope (see Figure 1). Upon implementation of the intervention, an increase in the daily percentage of active displays of joining in emerged. That is, from the last data point during the baseline phase (36% active joining in) to the initial data point during the interv ention phase (45% activ e joining in) there was a 9% change in level in the positive dire ction. During the inte rvention phase, Angelo averaged 23.83% of active engagement, an increase of 5.51%. Follow-up effects were not as pronounced for Angelo as they were with Darius and Francis. The mean percentage of active displays of joining in for Angelo during follow-up was 1.67% (range 0% to 4%). This was a decrease of 16.65% from the baseline condition. Descriptive statistics relative the changes in mean acro ss time for Angelo are provided in Table 9. Table 9. Descriptive Statistics for Ac tive Joining In for Angelo X SD Min. Max. Baseline 18.32% 18.77 0 73 Intervention 23.83% 30.10 0 83 Follow-Up 1.67% 2.08 0 4 As can be observed in Figure 1, there is some change in the slope of Angelos data across time. Specifically, during th e baseline condition, Angelo demonstrated a decreasing trend (slope = -.40). Simila rly, during the intervention phase, Angelo
90 demonstrates a slightly increasing trend (s lope = .21). During both the baseline and intervention conditions there were instances of high rates of active joining in behaviors. It was noted on the data sheets that these rates of active joini ng in behavior were precipitated by Angelo be ing selected as a team captain for a game during recess or asked to participate in a game activity. However, during the majority of the observation days, Angelo was not picked as a captain nor asked to participate in activities and rates of his active joining in behavior appeared to be va riable, but lower than these several points. During the follow-up phase, Angelo demonstrated decreasing rates of active joining in behaviors. Table 10 highlights the slopes of each trendline, as well as calculations of slope and level change acro ss conditions for Angelo. Table 10. Slope and Level Change for Active Joining In for Angelo Condition Slope Change in Slope Level Change in Level Baseline -.40 36 (@ last day) Intervention .21 .61 45 (@ first day) 9 Intervention --5 (@ last day) Follow-Up ----4 (@ first day) 1 Due to the high variability of Angelos data, th ere was a high level of overlap in the data across phases. The percent of overlapping data points the intervention period with baseline was 92%. The only point that did not overlap was on e of the days when Angelo was selected as a team captain. The number of overlapping data points during the followup period was 100% when examining overlap with the intervention and baseline phases.
91 Passive Display of Target Behavior Figure 2 shows the mean percentage of passive displays of target behaviors (sportsmanship and joining in) for each pa rticipant across baselin e, intervention, and follow-up phases. These data show an increas e in the percent of passive sportsmanship for Darius following the initiation of the social story interventions. Specifically, the data show that Darius engaged in significantly more passive sportsmanship (e.g., following the rules of the game, accepting the loss of a game without making a negative comment) during the intervention and follow-up phases than during baseline. This effect was not replicated in either Francis or Angelo. In fact, the opposite trend was observed. Detailed description of the data relative to passi ve engagement for each participant follows.
92 Figure 2. Percent of Passive Displays of Ta rget Behavior Across Participants 0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 1001 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 Baseline Intervention Follow-Up Darius Francis Angelo Percent S p ortsmanshi p Percent Joinin g In Percent Joinin g In Days
93 Darius. Prior to the intervention, Darius di splayed a steadily increasing baseline (see Figure 2). The overall mean percentage of passive displays of sportsmanship for Darius during baseline was 51.22%. Upon impl ementation of the in tervention, there was a very slight decrease in the percentage of passive displays of sportsmanship. That is, from the last data point during the baseline phase (55% passive sportsmanship) to the initial data point during the intervention phase (52% passive sportsmanship) there was a 3% change in level in the negative direct ion. However, 3 consecutive increasing data points followed this initial starting point. Although the data did not reveal a prompt display of passive sportsmanship for Darius at the beginning of the intervention phase, overall mean percentage of behavior dur ing the intervention pe riod was 73.5% (range 42% to 86%). Overall, this is an increas e in mean data for Darius of 22.28% from baseline to intervention phase. Similarly, these behaviors demonstrated maintenance during the follow-up period with a mean percent of passive s portsmanship of 75% (range 68% to 79%). This suggests that Darius engaged in more passive sportsmanship behaviors across time, a trend that is consider ed positive. Descriptive statistics relative the changes in mean across time fo r Darius are provided in Table 11. Table 11. Descriptive Statistics for Passi ve Sportsmanship for Darius X SD Min. Max. Baseline 51.22% 4.58 44 57 Intervention 73.5% 10.76 42 90 Follow-Up 75% 6.08 68 79 Changes in trend are depicted in Figure 2 as solid lines drawn in each phase. For Darius, a steadily increasing trend was demons trated during the base line phase (slope = 1.19). Although the trend of the data for Da rius was increasing, implementation of the
94 intervention progressed due to time constraints. However, the change of mean for Darius from baseline to intervention phases supports the effectivenes s of social stories. The trend continued to increase during the intervention period (s lope = .32). However, this trend was not as steep as orig inally seen during the baseline condition. Trend data were not examined for the follow-up phase due to insufficient data. Table 12 highlights the slopes of each trendline, as well as calc ulations of slope and level change across conditions for Darius Table 12. Slope and Level Change for Passi ve Sportsmanship for Darius Condition Slope Change in Slope Level Change in Level Baseline 1.19 55 (@ last day) Intervention .32 -.87 52 (@ first day) 3 Intervention --80 (@ last day) Follow-Up ----79 (@ first day) 1 The percentage of overlap of interven tion compared to baseline was only 10%, which reveals that most of the intervention data were significantly above percentages of engagement established during baseline. Th e percent of overlappi ng data points during the follow-up period was 100% when examining overlap of intervention with follow-up. These results suggest that during the follo w-up phase, mean performance remained at similar levels to the intervention phase. Over lapping data points for a comparison of the follow-up and baseline conditions revealed no overlap. All follow-up data were above baseline percentages. Francis. Mean percent of passive joining in decreased across time for Francis. Figure 2 displays an increasing trend in ba seline for Francis and a mean percent of
95 passive engagement of 35.93% prior to im plementation of the intervention. Upon implementation of the social story interv ention, an abrupt decrease in the daily percentage of passive joining in emerged. Specifically, from the la st data point during the baseline phase (62% passi ve joining in) to the init ial data point during the intervention phase (21% passive joining in) a 41% decrease in leve l occurred. Although this degree of change was not replicated w ith successive data poi nts, this decrease persisted variably throughout the intervention phase. Overal l, mean data for Francis during the intervention phase was 27.14% (range 13% to 59%). These data represent a decrease of 8.74% compared to baseline. Howe ver, this decline in percentage of passive engagement is expected given the nature of soci al interaction. That is, with an increase in active engagement (discussed previously) it s hould be expected that passive engagement would decrease. Likewise, during the fo llow-up phase, similar results were found. Specifically, mean percent of passive join ing in for Francis during follow-up was 29% (range 20% to 38%). De scriptive statistics relative the changes in mean across time for Francis are provided in Table 13. Table 13. Descriptive Statistics for Pa ssive Joining In for Francis X SD Min. Max. Baseline 35.93% 18.29 3 62 Intervention 27.14% 13.44 13 59 Follow-Up 29% 12.73 20 38 Changes in trend are depicted in Figure 2 as solid lines draw in each phase. For Francis, a steadily increasing trend was dem onstrated during the ba seline phase (slope = .73). However, during the intervention phase, the trend decreased (s lope = -1.14). Trend data were not calculated for the follow-up period due to insufficient data. Table 14
96 highlights the slopes of each tr endline, as well as calculations of slope and level change across conditions for Francis. Table 14. Slope and Level Change for Passive Joining for Francis Condition Slope Change in Slope Level Change in Level Baseline .73 62 (@ last day) Intervention -1.14 -1.87 21 (@ first day) 41 Intervention --17 (@ last day) Follow-Up ----38 (@ first day) 21 The percentage of overlap for all levels wa s 100%. This reveal s that most of the intervention data was highly variable, and th e data for interventi on and follow-up phases were within the range of baseline behaviors. However, visual inspection of the data (see Figure 2) demonstrates the d ecrease of variability during the intervention period. The overlap of 100% for all levels is attributed to the wide variability of baseline data (see Figure 2). Angelo. Prior to intervention phase, Ange lo demonstrated a variable, but somewhat increasing, base line of passive joining in behavi ors (see Figure 2). The overall mean percentage of passive displays of jo ining in for Angelo during baseline was 9.26% Upon implementation of the in tervention, there was an abru pt decrease in the daily percentage of passive joining in behaviors. Specifically, from the last data point during the baseline period (31% passi ve joining in) to the initial data point during the intervention phase (13% passive joining in) there is an 18% d ecrease in Angelos passive rates of joining in behavior. Throughout the intervention phase, Angelo demonstrated relatively stable, but variable data. Mean pe rcent of passive joining in for Angelo during
97 the intervention phase was 7.25%. Thus, a sma ll decrease of 2.01% from intervention to baseline phases occurred. As was the case with Francis, a decrease in the percent of passive engagement may signify increased use of active behaviors when interacting with others. During follow-up, Angelo did not enga ge in any passive joining in behaviors Descriptive statistics relative the changes in mean across time for Angelo are provided in Table 15. Table 15. Descriptive Statistics for Pa ssive Joining In for Angelo X SD Min. Max. Baseline 9.26% 8.74 0 31 Intervention 7.25% 9.01 0 24 Follow-Up 0% --0 0 As can be observed in Figure 2, there is some change in the slope of Angelos passive joining in data acro ss time. Specifically, during th e baseline condition, Angelo demonstrated an increasing trend (slope = .41). On the other hand, during the intervention phase, Angelo demonstrated a s lightly decreasing trend (slope = -.06). Based on decreasing rates of passive behavi or, there appears to be some change. However, these data should be considered cau tiously due to the highl y variable data, as well as the lack of significant ch ange demonstrated in Angelos active rates of joining in. Table 16 highlights the slopes of each trendline, as well as calculations of slope and level change across conditions for Angelos passive joining in behaviors.
98 Table 16. Slope and Level Change for Passive Joining for Angelo Condition Slope Change in Slope Level Change in Level Baseline .41 31 (@ last day) Intervention -.06 -.47 13 (@ first day) 18 Intervention --10 (@ last day) Follow-Up ----0 (@ first day) --Due to the high variability of Angelos data, there was a high level of overlap when comparing each condition to baseline. The percent of overlapping data points across all conditions was 100%. These data de monstrate the highly va riable nature of Angelos data, as well as suggest that no si gnificant change occurred across time. This lack of significant change is clearly displayed in Figure 2. Absence of Target Behavior Figure 3 displays the mean percent of absence of target behaviors across participants during three treatment phases. The data show that Darius and Francis engaged in significantly less negative beha viors during the inte rvention and follow-up phases of the study than during th e baseline phases. This change in the rates of absence of target behavior was not observed for Angel o. Detailed summaries of the data in terms of changes in level, slope, and percen t of overlapping data points follow.
99 Figure 3. Percent of Absence of Target Behavior Across Participants 0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 1001 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 Days Baseline Intervention Follow-Up Darius Francis Angelo Percent S p ortsmanshi p Percent Joinin g In Percent Joinin g In
100 Darius. Prior to the intervention, Darius displayed a decreasing percentage of absence of sportsmanship (see Figure 3). Although data reveal ed a trend in the hypothesized direction, initiati on of the intervention proceed ed due to the decreasing rates of active sportsmanship for Darius duri ng baseline. The overall mean percentage of absence of sportsmanship for Darius duri ng baseline was 36.44%. Upon implementation of the intervention, there was a slight decrease in the percentage of absence of sportsmanship behaviors for Da rius. Specifically, from the last data point during the baseline period (24% absence of sportsmans hip) to the initial data point during the intervention phase (17% absence of sportsmans hip) there is a 7% decrease in Dariuss absence of sportsmanship behavior. During the intervention phase, mean percentage of the absence of sportsmanship behavior was 7.1%. Overall, this is a decrease in mean data for Darius of 29.34% from baseline to interven tion phase. Followi ng the withdrawal of the social story intervention, mean level fo r the absence of sportsmanship behavior increased. Specifically, mean level of the absence of sportsmanship behavior during follow-up was 13.67%. However, the mean le vel during the follow-up condition was still far below the baseline rates of absence of s portsmanship. Such a change in means across conditions demonstrates maintenance of be havior change for Darius. Descriptive statistics relative the changes in mean acro ss time for Darius are pr ovided in Table 17. Table 17. Descriptive Statistics for Absen ce of Sportsmanship for Darius X SD Min. Max. Baseline 36.44% 7.00 24 44 Intervention 7.1% 6.46 0 22 Follow-Up 13.67% 6.66 8 21
101 Changes in trend are depicted in Figure 3 as solid lines drawn in each phase. For Darius, a steadily decreasing trend was demons trated during the baseline phase (slope = -1.94). Similarly, during the intervention phase, Dariuss behavior continued to demonstrate a decreasing trend (slope = -.30) Slope was not calculated for the follow-up period due to insufficient data. Table 18 highl ights the slopes of each trendline, as well as calculations of slope and level change across conditions for Darius. Table 18. Slope and Level Change for Absence of Sportsmanship for Darius Condition Slope Change in Slope Level Change in Level Baseline -1.94 24 (@ last day) Intervention -.30 1.64 17 (@ first day) 7 Intervention --2 (@ last day) Follow-Up ----12 (@ first day) 10 The percentage of overlap of intervention with baseline was 0%. Suggesting that all of the intervention data were significantly below percentages of absences of joining in behavior established during baseline. The pe rcent of overlapping da ta points during the follow-up period was 67% when examining overl ap with the intervention phase. This result suggests that during the follow-up phase mean performance remained at levels similar to the intervention phase. Overla pping data points for a comparison of the follow-up and baseline condi tions revealed no overlap. Francis. Prior to intervention phase, Franci s displayed a highly variable, but decreasing baseline of absence of joining in behaviors (see Figure 3). The overall mean percent of the absence of joining in beha vior for Francis duri ng baseline was 31.86%. Upon implementation of the social story interv ention, there was a very slight increase in
102 the percent of Franciss absence of joining in behavior. Spec ifically, from the last data point during the baseline phase (10% absence of joining in) to the initial data point during the intervention phase (13% abse nce of joining in) there was a 3% increase in level. This change in level was in contra st to the hypothesized direction. However, overall percent of absence of joining in behaviors fo r Francis decreased to 14.07% during the intervention period, a change from baseline of 17.79%. In addition, the variability decreased substantially (see Figure 3). Duri ng the follow-up phase, there was an increase in the mean level of absence of joining in be haviors. Specifically, mean percent of absent behaviors for Francis during follow-up was 29%. This change represents an increase of 14.93% from intervention to baseline conditions. However, the follow-up mean performance remained lower than baseline. Descriptive statistics relative the changes in mean across time for Francis are provided in Table 19. Table 19. Descriptive Statistics for Absence of Joining In for Francis X SD Min. Max. Baseline 31.86% 19.38 5 68 Intervention 14.07% 6.29 2 28 Follow-Up 29% 25.46 11 47 As can be observed in Figure 3, there is change in the slope of Franciss data across time. Specifically, during the baseline conditi on, Francis demonstrated a decreasing trend (slope = -1.13). Similarl y, during the interventi on phase, absence of joining in behaviors demonstrated a decreasi ng trend (slope = -.20). A trendline for the follow-up condition was unable to be determined due to the lack of a data. Therefore, it is difficult to determine the direction of Franciss absence of joining in behavior
103 following the intervention. Table 20 highlights the slopes of each trendline, as well as calculations of slope and level chan ge across conditions for Francis. Table 20. Slope and Level Change for Absence of Joining In for Francis Condition Slope Change in Slope Level Change in Level Baseline -1.13 10 (@ last day) Intervention -.20 .93 13 (@ first day) 3 Intervention --13 (@ last day) Follow-Up ----11 (@ first day) 2 During the intervention phase, the percenta ge of overlap of intervention with baseline was 100%. However, visual inspec tion of the interventi on period reveals a distinct pattern of decreased variability (see Figure 3). Co mparison of the intervention and follow-up conditions revealed an overl ap of 67%. Similarly, the percent of overlapping data points during the follow-up period was 100% when examining overlap with the baseline condition. However, it is di fficult to determine the maintenance of the behavior change due to the lack of su fficient data during the follow-up phase. Angelo. Prior to intervention phase, Angelo di splayed variable ra tes of absence of joining in behaviors (see Figure 3). The overall mean percentage of absence of joining in behaviors for Angelo during baseline was 72.16% Following baseline, there was little change in the rates of absen ce of joining in behaviors dur ing the intervention condition (see Figure 3). Specifically, Angelo averaged 68.42% of absence of joining in during the intervention phase, a decrease of 3.74% compar ed to baseline. In addition, there was very little change in level from baseline to intervention. At follow-up, Angelos mean percentage of absence of jo ining in increased to 98.33%. These data demonstrate a
104 substantial increase when compared to bot h intervention and baseline conditions. Overall, these data suggest that the percen t of the absence of joining in behavior demonstrated little change acro ss conditions. Descriptive statistics relative the changes in mean are provided in Table 21. Table 21. Descriptive Statistics for Abse nce of Joining In for Angelo X SD Min. Max. Baseline 72.16% 21.97 33 100 Intervention 68.42% 34.59 17 100 Follow-Up 98.33% 2.08 96 100 Celeration lines are indicated by solid lines through data points for each phase as shown in Figure 3. For Angelo, a relatively flat trend was dem onstrated during the baseline (slope = -.11) and intervention phases (slope = -.003). No trendlines were calculated for the follow-up condition due to the few number of data points available. Table 22 highlights the slopes of each trendline, as well as calculations of slope and level change across conditions for Angelo. Table 22. Slope and Level Change for Absence of Joining In for Angelo Condition Slope Change in Slope Level Change in Level Baseline -.11 33 (@ last day) Intervention -.003 .107 36 (@ first day) 3 Intervention --85 (@ last day) Follow-Up ----96 (@ first day) (-) 9 The data for Angelo are highly variab le throughout both the baseline and intervention phases (see Figure 3). During the intervention phase, the percentage of overlap of intervention with baseline was 75%. The percent of overlapping data points
105 during the follow-up period was 100% when exam ining overlap with the intervention and baseline phases. The overlap data suggest that the intervention had little impact on decreasing Angelos rates of abse nces from joining in activities. OASIS Data Observational data collected from the OASIS form was used to examine any collateral effects of the interv ention for each participant. That is, information regarding social interaction skills (e.g., eye contact, calling peers by their name, smiling) for each participant was compared across conditions to examine mean changes in the use of appropriate interaction skills and if there were any changes in the rates of these behaviors. Descriptive data for each particip ant are detailed in Table 23. Mean changes that were in the hypothesized dire ction appear in italics. Table 23. OASIS Descriptive Data for Each Participant X SD Min. Max. Darius Baseline 49.00% 11.03 33 67 Intervention 56.53% 8.74 40 67 Follow-Up 51.00% 3.46 47 53 Francis Baseline 47.93% 14.43 20 67 Intervention 64.58% 20.25 27 100 Follow-Up 57.00% 14.14 47 67 Angelo Baseline 41.50% 24.95 0 73 Intervention 25.22% 32.52 0 80 Follow-Up 4.33% 7.51 0 13 These data demonstrate that Darius and Fr ancis displayed more appropriate social interaction skills following baseline conditions In addition, there appears to be some maintenance regarding the use of these appr opriate social interactions following the
106 intervention. Specifically, both Darius and Fr ancis displayed mean percentages of these skills higher than baseline conditions. Howeve r, the same effects were not observed for Angelo. In fact, Angelo displa yed decreasing rates of appropria te social interaction skills across both intervention a nd follow-up conditions. Participants Social Communica tion Progress: Peer Comparison With regard to the second hypothesis: Participants in the soci al story intervention will demonstrate clinically significant in creases in social interactions/social communication the following results were found through analysis of peer comparison data for each of the identified target behavi ors across participants. Specifically, data from the participants and peer comparisons we re analyzed visually to determine if rates of active, passive, and absence displays of ta rget behaviors approached peer levels of these behaviors.
107 Figure 4. Percent of Active Displays of Targ et Behaviors Across Participants with Peer Comparison 0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 1001 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 Days Baseline Intervention Follow-Up Darius Francis Angelo Percent S p ortsmanshi p Percent Joinin g In Percent Joinin g In Peer Target
108 Active Comparison Darius. Figure 4 displays the percent of ac tive sportsmanship behavior for Darius and comparison peers. The data show that ove r time, Darius approach ed peer levels of sportsmanship behaviors. In addition, a correspondence between Darius and peer behavior is demonstrated. Th at is, on days when peers di splayed relatively low rates of active sportsmanship, the same effects are displayed by Darius during the intervention period. Although these effects also appear during baseline and follow-up conditions, the level of correspondence is not as pronoun ced as during intervention (see Figure 4). Mean levels of active sportsmanship for Darius with peer comparisons are displayed in Table 24. These data show th at during the interven tion period, Dariuss mean level of active displays of sportsmans hip approached the level of his peers. Specifically, during baseline there was a difference in rates of active sportsmanship between Darius and his peers of 19.22%. Howe ver, during the interv ention period this difference decreased substantially to 6.75%. Similar effects were demonstrated during follow-up. Specifically, Dariuss mean level of active sportsmanship differed from his peers by 7.67%. Table 24. Mean Level of Active Sportsmanship for Darius with Peer Comparison X SD Darius Peer Darius Peer Baseline 8.67% 27.89% 4.21 8.94 Intervention 17.95% 24.70% 5.56 8.98 Follow-Up 11.33% 19.00% 2.52 5.29 Francis. Figure 4 displays the percent of ac tive joining in behavior for Francis and comparison peers. The data show that ove r time, Francis approached peer levels of
109 joining in behaviors. In a ddition, a correspondence between Fr ancis and peer behavior is demonstrated. That is, on days when p eers displayed relatively low rates of active joining in, Francis displays the same effects during the intervention pe riod. These effects are similar across all pha ses of data collection. Mean level of active joining in for Fran cis and peers are displayed in Table 25. These data show that duri ng the intervention period, Fran ciss mean level of active displays of joining in appro ached the level of his peers. Specifically, during baseline there was a difference in rate s of active joining in betw een Francis and his peers of 48.43%. However, during the intervention pe riod this difference decreased to 22.06%. Effects during follow-up did not reveal sim ilar effects. Specifically, during follow-up there was a 39.50% difference between Franci ss active rates of joining in when compared with peers. While the difference in mean levels are not as pronounced as displayed in Dariuss data, visual inspec tion of the data (see Figure 4) display an increasing trend during the intervention phase. Table 25. Mean Level of Active Joining In for Francis with Peer Comparison X SD Francis Peer Francis Peer Baseline 30.14% 78.57% 13.18 12.45 Intervention 58.29% 80.35% 21.87 18.23 Follow-Up 41.50% 81.00% 13.44 11.31 Angelo. Figure 4 displays the percent of active joining in behavior for Angelo and comparison peers. The data show that over time, Angelo did not approach peer levels of joining in behaviors. Mean level of active joining in fo r Angelo and peers are displayed in Table 26. These data show that across all conditions, Angelos mean level
110 of active displays of joining in differed subs tantially from his peers. However, during times when Angelo was asked to be a team cap tain (first data point during baseline on Figure 4) or prompted to play a game (8 a nd 9 data point during intervention on Figure 4), his data approached peer levels of active joining in. Table 26. Mean Level of Active Joining In for Angelo with Peer Comparison X SD Angelo Peer Angelo Peer Baseline 18.32% 98.26% 18.77 4.20 Intervention 23.83% 98.42% 30.10 2.19 Follow-Up 1.67% 99.33% 2.08 1.15
111 Figure 5. Percent of Passive Displays of Targ et Behaviors Across Participants with Peer Comparison 0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 1001 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 Baseline Intervention Days Follow-Up Target Peer Percent S p ortsmanshi p Percent Joinin g In Percent Joinin g In Darius Francis Angelo
112 Passive Comparison Darius. The percent of passive sportsmanship behavior for Darius and comparison peers are displayed in Figure 5. Th ese data show that over time, Darius and comparison peers approached the same leve l of performance dur ing the intervention phase. In addition, there was some corre spondence between Darius and peers during both the intervention and follow-up cond itions, but not during baseline. Mean level of passive sportsmanship for Dari us with peer comparis ons are displayed in Table 27. These data show that during the in tervention period, Dari uss mean level of passive displays of sportsmanship was simila r to the level of his peers. During the intervention period Dariuss percent of passive sportsmanship increased to a level that was higher than, but consistent with the tre nd of his peers (see Figure 5). Both Darius and his peers demonstrated similar mean passive sportsmanshi p during the follow-up period. Table 27. Mean Level of Passive Sportsmanshi p for Darius with Peer Comparison X SD Darius Peer Darius Peer Baseline 51.22% 59.67% 4.58 15.17 Intervention 73.50% 69.60% 10.76 10.43 Follow-Up 75.00% 79.00% 6.08 7.00 Francis. Figure 5 displays the percent of pa ssive joining in behavior for Francis and comparison peers. The data show that over time, Francis demonstrated more passive joining in behaviors than comp ared to peers. Data also show a correspondence between Franciss behavior and comparis on peers. This trend is dem onstrated during baseline and
113 intervention phases. However, during the intervention phase, Franciss passive joining in behavior began to decrease and match that of his peers (see Figure 5). Mean level of passive joining in beha vior for Francis and comparison peers are displayed in Table 28. These data show that during the baseline and intervention periods, Franciss mean level of passive displays of joining in were higher than the comparison peers. During the follow-up phase, both Franci s and peers demonstrated the same mean percentage of passive joining in behaviors. Table 28. Mean Level of Passive Joining In for Francis with Peer Comparison X SD Francis Peer Francis Peer Baseline 35.93% 15.50% 18.29 12.43 Intervention 27.14% 13.43% 13.44 14.93 Follow-Up 29.00% 29.00% 12.73 12.73 Angelo Figure 5 displays the percent of pa ssive joining in behavior for Angelo and comparison peers. The data show that over time, Angelo and comparison peers differed in mean level of passive joining in. Peers demonstrated near zero levels of passive joining in behaviors whereas Angelo displayed variable rate s of passive joining in. Mean level of passive jo ining in for Angelo and peer s are displayed in Table 29. These data show that across a ll conditions, Angelos mean le vel of passive displays of joining in was higher than that of his peers. Table 29. Mean Level of Passive Joining In for Angelo with Peer Comparison X SD Angelo Peer Angelo Peer Baseline 9.26% 1.16% 8.74 3.56 Intervention 7.25% 0.83% 9.01 1.64 Follow-Up 0 0.67% 0 1.15
114 Figure 6. Percent of Absence of Target Behaviors Across Participants with Peer Comparison 0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 1001 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 Days Baseline Intervention Follow-Up Darius Francis Angelo Percent S p ortsmanshi p Pe r cent Joinin g In Percent Joinin g In Target Peer
115 Absence Comparison Darius. Figure 6 displays the percent of absence of sportsmanship behavior for Darius compared to his peers. The data s how that over time, Darius approached peer levels of negative behaviors. Mean level of absence of sportsmanship for Darius with peer comparisons are displayed in Table 30. These data show that during the intervention period, Dariuss mean level of absent displays of sportsmanship appr oached the level of his peers. Specifically, duri ng baseline there was a differe nce in rates of absence of sportsmanship between Darius and his peers of 28.11%. However, during the intervention period this differe nce decreased substantially to 1.8%. During follow-up, the difference between Darius and his peers incr eased. However, this increase did not approach the same level as displayed during baseline. Table 30. Mean Level of Absence of Sportsmans hip for Darius with Peer Comparison X SD Darius Peer Darius Peer Baseline 36.44% 8.33% 7.00 4.30 Intervention 7.10% 5.30% 6.46 6.04 Follow-Up 13.67% 1.67% 6.66 2.89 Francis. Figure 6 displays the percent of absence of joining in behavior for Francis and comparison peers. These data demonstrate that dur ing the intervention period, Francis displayed absences of joining in behaviors at rates si milar to his peers. Specifically, there was only a 0.64% differen ce between Francis and peers absence of joining in behaviors during the interven tion. In addition, a correspondence between Francis and peers was demonstrated.
116 Mean level of absence of joining in for Francis and peers are displayed in Table 31. As these data demonstrate, there was a s ubstantial decrease in the mean percentage of absence of joining in behaviors from ba seline to intervention. However, during the follow-up phase, Franciss absence of join ing in behaviors increased from the intervention level. This increase was si milar to that displayed during baseline. Table 31. Mean Level of Absence of Joining In for Francis with Peer Comparison X SD Francis Peer Francis Peer Baseline 31.86% 15.50% 19.38 12.43 Intervention 14.07% 13.43% 6.29 14.93 Follow-Up 29.00% 6.50% 25.46 3.54 Angelo Figure 6 displays the percent of absence of joining in behavior for Angelo compared to his peers. These data show that over time, Angelo and comparison peers differed in mean level of absence of join ing in. Mean level of absence of joining in for Angelo and peers are displayed in Ta ble 32. These data show that across all conditions, Angelos mean level of absent displa ys of joining in was higher than that of his peers. Peers demonstrated near zero levels of absence levels of joining in behaviors, whereas Angelo displayed relatively high ra tes of absent joining in (see Figure 6). Table 32. Mean Level of Absence of Joining In for Angelo with Peer Comparison X SD Angelo Peer Angelo Peer Baseline 72.16 0.58% 21.97 2.52 Intervention 68.42 0.75% 34.59 1.76 Follow-Up 98.33 0 2.08 0
117 Participants Social Communica tion Progress: Follow-Up Data Relative to the third hypothesis: Participants in the social story intervention will maintain positive treatment effects at a 2-week follow-up the following results were found. While descriptions of the follow-up data are discu ssed in detail in previous sections, a summary of the follow-up data for each of the participants follows. For Darius, data demonstrate that mainte nance of behaviors continued during the follow-up condition across active, passive, and absence of sportsmanship (see Figures 13). Specifically, Dariuss rates of active and passive behaviors were lower than those that occurred during the interventi on period, but remained above baseline levels. In addition, Dariuss rates of absence of sportsmans hip behaviors remained lower than during baseline, and only slightly above rates of be havior during the intervention phase. These data demonstrate that, for Darius, maintenance of the skill introduced in the social story intervention occurred. However, these data s hould be considered with caution due to the relatively few number of da ta points available. Similar results were found for Francis. Specifically, data demonstrate that maintenance of behaviors occurred for Francis for active, passive, and absence of joining in behaviors (see Figures 1-3). For both act ive and passive joining in behaviors, data suggest that Francis continued to demonstrat e improved social enga gement as compared to baseline. However, this effect was not as apparent when examining the rates of absence of joining in (see Figure 3). Also, due to the lack of data points for Francis during follow-up, it is difficult to determine wh at future performance of each behavior would be. Although indications of maintena nce are suggested, further data would be necessary to reveal a trend of performa nce across all three ta rgeted behaviors.
118 There was no indication of any maintenan ce of increased social engagement for Angelo during the follow-up period across each behavior (see Figures 1-3). In fact, during the follow-up condition, rates of each beha vior for Angelo actually moved in the opposite direction of the stated hypothesis and no clear effects were demonstrated to show neither improvement nor maintenance of any improvement. Participants Social Communication Progress: Pre-, Post-Test Data With regard to the fourth hypothesis: Participants will dem onstrate improved social behaviors beyond identifi ed target behaviors as rated by parents and teachers on pre and post-test measures, the following results were found through an examination of the reliable change index (Jacobson & Truax, 1991) for each subjects data on the on the SSRS to determine whether the magnitude of th e change was statistically reliable. Data from both parent and teacher preand post-tests are provided in Appendix O. Social Skills Rati ng System (SSRS) Darius. For the parent ratings, the reliabl e change index was demonstrated a significant change (RC = 1.98). When the reliab le change index is greater than 1.96, it is unlikely that the post-test score is not reflecti ng real change. Specifically, for Darius, this result suggests that parents i ndicated collateral improvement of broader social skills on a standardized social skills inst rument. In addition, these data reveal that change from preto post-test reflects more than the fluctuati ons of an imprecise measuring instrument. Teacher analysis of preand post-test SSRS di d not reveal a statistic ally reliable change (RC = 1.88) Francis. Both parent and teacher analyses of the SSRS did not reveal any statistically reliable change fr om preto post-test indicating that neither the parent nor the
119 teacher indicated significant improvement of collateral social skills adjustment. The reliable change index for the parent rati ngs was .88 and 1.50 for the teacher ratings. Angelo. Unfortunately, no parent preand post-test data were available for Angelo. Therefore, the reliab le change index for Angelos data, as rated by the parent, was not calculated. Results of the reliable ch ange index for the teacher ratings preand post-test revealed no clini cal significance (RC = 1.32). With regard to the fifth hypothesis: Participants will demonstrate a decrease in unusual behavior patterns as rated by parents and teachers on pre and post-test measures the following results were found through an examination of the reliable change index (Jacobson & Truax, 1991) for each subjec ts data on the on the Childrens Atypical Development Scale (CADS) to determine wh ether the magnitude of the change was statistically reliable. Data from both pa rent and teacher preand post-tests were examined. Data from both parent and teacher preand post-tests are provided in Appendix P. Childrens Atypical Development Scale (CADS) Darius. For the parent ratings, the reliabl e change index was demonstrated a significant change (RC = 2.97). This reve als that parents indicated collateral improvement of atypical behaviors on the CADS In addition, this reveals that change from preto post-test reflects more than the fluctuations of an imprecise measuring instrument. Teacher analysis of preand post-test SSRS did not reveal a statistically reliable change (RC = 1.19) Francis. Similar to the results for Angelo, the reliable change index for parent ratings on the CADS was statisti cally reliable (RC = 2.77), indi cating that the parent rated
120 Franciss atypical behaviors as having decr eased when compared before and after the social story intervention. However, teacher ra tings were not statisti cally reliable (RC = 1.19). Angelo. Unfortunately, no parent or teacher pr eand post-test data were available for analysis. Therefore, the reliable change index for Angelos data, as rated by the parent and teacher, was not calculated. Social Story Journal Darius. Social story journal information for Darius revealed a consistent pattern of reading his social story twice a day dur ing the intervention pe riod. Specifically, Darius read his social story on the way to school and during bedtime each day. Darius indicated that the social stor y made him feel upset (i.e., sad or mad) sometimes because he is ridiculed on the playground. However, as the intervention continued, Darius indicated that his social st ory made him happy and that he would practice the skills embedded in the social story. During the last two weeks of the intervention, Darius began to fade the social stor y, only reading it every other da y. Slowly, Darius and his family began fading the social story to read ing it only once during th e final week of the intervention. Darius did not read hi s social story duri ng the follow-up period. Francis. Francis was very consistent in r eading his social story. Specifically, Francis read his social story with his mother twice a day at home with his mother during the intervention period. Franci s indicated that reading his social story made him feel happy and several journal entries indicated th at he would practice the skill introduced in the story. During the last two weeks of th e intervention, Francis read his story three times per week. Following the intervention, Francis ceased to read his story while on his
121 Spring Break for two weeks. During the follo w-up period, Francis did not read his social story. Angelo. Unfortunately, neither Angelo nor hi s family completed the social story journals. Therefore, it is difficult to determin e at what level Angelo read his social story on a daily basis. In addition, it is difficult to determine if Angelo was able to acquire and practice the skills presented in the social story. Such information is vital in determining the effectiveness of the social story for Angelo.
122 Chapter V Discussion Overview This chapter outlines the implications of the data relative to the hypotheses for the study. In addition, this chapter (a) outlines the significance of the current findings, (b) proposes implications for the de velopment of social story inte rventions for children with Aspergers Syndrome, and (c) discusses potentia l considerations for future research in this area. Impact of the Intervention Observational Data Social story interventions were effective in increasing specific social engagement skills in 2 out of the 3 children included in this study. Overall, the effects of the intervention appear to be most pronounced fo r Darius and Francis during the intervention condition. Following implementation of the social story, both Darius and Francis demonstrated improved and more consistent ra tes (less variability) of targeted social behaviors compared to baseline performan ce. In addition, Dariuss and Franciss improvements maintained at elevated rates co mpared to baseline. Maintenance of the target behaviors for Darius and Francis not only demonstrate the impact of their respective social stories, but also suggests th at social reinforcement for being included in activities on the playground may have been responsible for or c ontributed to their
123 maintenance. In addition to their individual data, Darius and Francis approached levels of performance that were similar to, or at sometimes greater than, their comparison peers across each target behavior. Interesti ngly, Darius and Francis demonstrated a correspondence between thei r rates of behavior and that of their peers. That is, if peer data were observed to decrease/incr ease, Darius and Franciss behavior decreased/increased. In addition to the co rrespondence in behavior, trends in the behavior for both the target participants a nd comparison peers demonstrate that for both Darius and Francis rates of beha vior were equivalent to or approaching the same level of peers, which demonstrates the clinical sign ificance of the social story interventions. Although marked increases in both Darius a nd Francis were observed, the same effects were not found for Angelo. Due to the indivi dualized nature of the target behaviors, a brief description of the results for eac h participant is hi ghlighted below. A close look at the percentage of spor tsmanship data for Darius revealed an increase in active and passive displays of sportsmanship during intervention, and at follow-up. In addition, Dariuss rates of absence of spor tsmanship behaviors steadily declined across all conditions. Overall, these data reveal positive effects for Darius across active, passive, and absence of spor tsmanship during intervention and follow-up conditions. Changes in mean, level, and slope were all in the hypothesized direction and indicated substantial improveme nt across conditions. In addition, percent of overlap revealed low overlap (below 70%) for the th ree targeted sportsmanship behaviors. Specifically, the percent of ove rlap for active and passive sportsmanship were low when comparing intervention with baseline. Simila rly, Darius displayed a continued decrease in overlapping data for absence of sports manship across all cond itions, increasing the
124 confidence that maintenance effects could be a ttributed to the intervention. Perhaps what is most revealing regarding the effectiveness of the intervention is that Dariuss level of sportsmanship behaviors matched the level of comparison peers. Performing at the same level of peers may demonstrate the ability fo r Darius to understand the perspectives of others and more readily read the social cues of others. Upon closer analysis of Dariuss data, it was observed that hi s passive rates of sportsmanship behavior occurred at rates higher than active s portsmanship. This result is not surprising considering the nature of the definitions of sportsmanship. Specifically, passive sportsmanship includes behaviors su ch as, following the rules of the game, accepting the loss of a game without making negative comments, and/or accepting a bad play without dissent. In many games or activities that involve reciprocal play, these may be the modal behaviors. In fact, excessive ly high rates of active sportsmanship (e.g., repeatedly yelling awesome, way to go, or good effort) may be perceived as eccentric behavior itself because it does not o ccur at high rates for typical peers. Peer comparison data support this hypothesis. Specifically, during the baseline and intervention periods, peers engaged in relativ ely low rates of active sportsmanship (30% actively engaged). However, peer rates of passive sportsmanship were relatively high (70% passively engaged) during the same conditions. Therefor e, observing an increase in passive rates of sportsmans hip behaviors for Darius was considered a significant improvement, as this demonstrated his improve d ability to read and understand the social cues of his peers and respond in socially appropriate ways. Similar results for Francis were found upon the implementation of the social story intervention. Specifically, mean data demons trated positive effects for Francis for each
125 targeted behavior across interv ention and baseline phases. Da ta relative to the percentage of active rates of joining in demonstrated improvement during the intervention period for Francis as indicated by change in mean, level, and slope. Closer in spection of Franciss data reveals a sharp decrease during the second day of the intervention period (see Figure 1). However, this decrease may have been due to rain during the recess time. During this time, all of the children were under a canopy and not permitted to walk around and play in the courtyard. Such environmental constr aints may have had an influence on Franciss joining in behavior. In addition, rates of active joining in for Francis appr oached the performance of comparison peers during the intervention peri od. These results suggest that Francis engaged in a greater number of social behavi ors in which he was actively contributing to a reciprocal conversation or act ively participating in some pl ay activity with one or more children. These increases in social engagement also maintained following the withdrawal of the intervention as indicated by the elevat ed rates of active joining in behaviors at follow-up. In addition, the percent of overla pping data points across phases for Francis provides an indication of the impact of th e social story interv ention on the target behaviors. Specifically, per cent of overlap was low (below 70%) for active and absence of joining in behaviors across all conditions. In contrast to Dariuss da ta, it was hypothesized that increasing active joining in behaviors would cause a decrease in the rates of passive joining in be haviors for Francis. Unlike sportsmanship behaviors, passive joini ng in behaviors would indicate occurrences of independent social behavior (e.g., playing next to peers, listening to a conversation without participating). Such be haviors are countertherapeutic to the targeted behavior for
126 Francis. Therefore, it was hypothesized that rates of passive joini ng in behaviors would decrease across time. Analysis of Franciss data revealed just th at. Over time, the percentage of time that Francis engaged in pa ssive joining in behavi ors steadily decreased during the intervention period. In additi on, these effects were observed during the follow-up condition, demonstrating that Francis engaged in higher rates of active joining in behavior across conditions. The results for Angelo were not as compelling. Specifically, the only positive effects that were found were changes in m ean across conditions during the intervention period. Specifically, active rates of joining in and rates of absence of joining in behaviors increased in the hypothesized direction. Ho wever, the increase for both behaviors was only equivalent to 4% and may have been attr ibuted to aggregating the highly variable data. Similarly, rates of pa ssive joining in behavior fo r Angelo decreased in the hypothesized direction, but only by 2%. Further visual analys es (i.e., slope and level), demonstrated little or no change for Angelo across time. In addition, percent of overlap data did not reveal any strong indication of change across time (i.e., less than 70% overlap). Data for Angelo were highly va riable and there was no visual indication of a decrease in the variabilit y of observational data. Several reasons for the lack of effectiven ess of the social story intervention for Angelo are offered. First, it is likely that the ineffectiveness may have been due to poor adherence to the treatment prot ocol. That is, the degree to which the intervention plan was implemented as intended (Gresham, 1989). Specifically, the parents were responsible for making sure that Angelo read his social story two times each day during the intervention period. Unfortunately, Angelo s parents did not complete any of the
127 social story journal entries. Therefore, it is difficult to determine whether Angelo engaged in learning the material from the social story from day to day. Second, due to his relatively low rates of joining in beha viors, Angelo may have responded well if a reinforcement paradigm was implemented to ju mpstart the effects of the social story. That is, by offering reinforcement for enga ging in active joining in behaviors during recess, Angelo may have been able to unde rstand the relationship between the social story and his subsequent behavi or. A closer look at A ngelos data demonstrates a significant disparity between his rates of target behaviors with comparison peers. This disparity is most pronounced for active joining in behaviors (see Figure 4). With such a large difference between Angelos and comp arison peers behaviors it may have been more difficult for Angelo to practice joining in. Third, Angelo may have benefited from additional practice of the social story content using role-plays That is, Angelos family may have set up practice opportunities for Ange lo to practice the content of the social story, as well as provide him with feedback (e.g., reinforcem ent) regarding his usage of the skill. Finally, there rema ins the possibility that identif ying soccer as the reinforcing activity within his social story may not ha ve been effective enough to initiate active joining in behaviors for Angelo. Choice assessments of recess activities may have revealed a more reinforcing situation (e.g., pl aying in the sandbox) as the focus of his social story. The overall results of the observational data demonstrat e that the social story interventions were effective for Darius and Fr ancis. Consistent with these findings were positive improvements in their use of appropriate social interaction skills as indicated by OASIS data. These collateral improvement s are noteworthy given the persistent and
128 significant difficulties reported for this population in using social-communicative behaviors to initiate, res pond, and maintain conversations and/or activities (Atwood, 2000; Bauer, 1995, 1996; Mundy & Stella, 2001). Overall, the observational data demonstrated that social st ory interventions were effec tive for improving the rates of social engagement for Darius and Francis, bu t showed little evidence of effectiveness for Angelo. Preand Post-test Data Analysis of the preand post-test data revealed a limited number of significant findings. Specifically, Dariuss parents demons trated clinically significant improvement on both the SSRS-PE and CADS. Such increases in collateral social skills and decreases in atypical behaviors suggest that the social story intervention may have produced change above and beyond the content of the social stor y for Darius. However, these results must be considered cautiously because there was no replication of these findings neither from Dariuss teacher nor across participants. These results are not surprising given the pervasiveness of social impairment in child ren with AS. Specifically, the hallmark of difficulty for children with AS remains so cial functioning across multiple domains (e.g., initiating and sustaining conversations, eye cont act, personal space, perspective taking), and extended time may be necessary for pr actice of targeted skills followed by introduction of additional target skills befo re long-term effects can be evaluated. In addition, the individualized na ture of social stories makes it difficult to observe more global changes in behaviors in such a short time frame. Therefore, preand post-test measures may not have been robust enough to demonstrate significant change across multiple domains of social functioning for this study.
129 Significance of Findings The findings of this study contribute in se veral ways to the effectiveness of social stories literature. First, this research demonstrates the poten tial benefits of using social story interventions to teach new prosocial behaviors to children with varied social behaviors. The results of this study were similar to those found pr eviously and this study contributes to the growing knowledge of social story research. Specifi cally, the results of this study replicate the previous findings of the efficacy of social stories used to teach appropriate social skill be havior to children with au tism (Noris & Dattilo, 1999; Theimann & Goldstein, 2001; Swaggart, et al., 1995). In addition, the results of this study replicate previous resear ch by Kuttler et al., (1998) a nd Swaggart, et al., (1995) by demonstrating a substantial change in be havior following the implementation of the social story. Similarly, two additional studie s (Kuttler, Myles, & Carlson, 1998; Lorimer, et al., 2002) demonstrated convincing evidence that the intervention was responsible for change through the use of an ABAB (Reversal) design. In each of these studies, behaviors returned to baseline following the removal of the so cial story interventions. However, the intended purpose of social story interventions is to attain permanent change even after the intervention is withdrawn. Th erefore, a loss of treatment gains following the removal of a social story demonstrates conc ern for the use of social stories in applied practice. Therefore, a fading paradigm to slowly eliminate the supports provided by the social story may prove beneficial for childre n to maintain the treatment effects. Aside from the previous research, the im pact of this study may be even more significant for the research literature on social story interventions because: (a) it employs experimental control, and (b) is the only know n research study to use a pure AS sample.
130 In many of the studies reviewed prior to c onducting this research, lack of experimental control was an issue. For example, Noris a nd Dattilo (1999) and Swaggart, et al. (1995) employed an A-B design in their respective studies. Such designs do not provide sufficient information to rule out the influence of a host of confounding variables (e.g., history, maturation) (Barlow & Hersen, 1984; K azdin, 1982). Therefore, it is difficult to determine the natural course that the behavi or(s) would take had no intervention occurred (Risley & Wolf, 1972). In addition, all of the prior research literature on the effectiveness of social story interventions has been conducted with samples of children with autism and severe social and behavioral difficulties. No known studies have focused specifically on samples of children with AS. Second, this study demonstrates the util ity of conducting functional assessments prior to designing and implementing a social story intervention. F unctional assessments allow for greater specification of the problematic social situation that the social story will have as its focus, as well as the salient feat ures of the context and setting for the social story. In essence, there is a direct lin k of assessment data with the design and implementation of the soci al story intervention. Third, this study offers a unique contri bution to the research literature by employing peer comparison data to demonstr ate the correspondence of participants target behaviors with that of comparison peers. To date, no known studies have employed such a comparison to examine the c linical effectiveness of the intervention. Comparing the rates of target behaviors w ith comparison peers has several important implications. First, no known informati on could be found on how often typical peers engage in the behaviors that were under study. However, interventions for children with
131 autism spectrum disorders primarily focus on in creasing social engagement (Brady, et al., 1987; Rogers, 2001; Twachtman-Cullen, 1998). The goal in most of this research is to reach functional levels of social engageme nt. However, without understanding how often typical peers engage in targeted behaviors, it is difficult to determine the level of what signifies clinical significance. This concer n is particularly salie nt when considering social communication and social behavior. Second, comparing th e rates of target behaviors with those of comparison peers in essential for demonstrating participants development of understanding social cues. That is, if target beha viors taught through a social story began to exceed the behaviors of typical peers, the child with ASD may be ridiculed and bullied more so than he or she al ready is due to not fitt ing in with the social norms of the environment. Finally, comparison data allows for closer examination of any direct correspondence or sequence effect betw een typical peer beha vior and the target child. That is, does the behavior of the typi cal peer (good or bad) have any impact on the behavior of the target individual? Results from this study revealed that there was a correspondence between the behaviors for both Darius and Francis and typical peers. The use of peer comparison data not only provi des a means for the evaluating the clinical significance of behavior change, but also provides a means of changing the perceptions of teachers regarding students with severe social deficits and peculiar behaviors. Lastly, this research contributes to th e literature base by providing evidence of collateral improvements of social skill behaviors and decrease s in atypical behaviors as indicated by preand post-test measures. The secondary changes observed in this study are consistent with Thiemann and Goldstein s (2001) recommendation that treating social behaviors within a similar response class wi ll ultimately lead to positive changes in
132 untreated social behaviors. Specificall y, by improving basic sportsmanship and joining in behaviors, either directly or indirectly, more opportunities to participate in day-to-day social interactions may arise that will ultim ately help children with ASD to be accepted members of their school social network. Implications for Practice This study extends previous investigations of the positive effects of social stories for children with autism. More importantly, this study offers the much needed empirical support for strategies when considering interv ention methods for students with AS. Such information is greatly needed considering the rapid increase in diagnoses and referrals for special education services for students with AS. This preliminary information may serve as a springboard for future program design and intervention implementation for children with AS in educational settings. Writing a social story and adhering to the specifics of creating a social story may take some practices for those with no previous traini ng. Specifically, educators/practitioners may re quire additional information on the types of sentences used, as well the ratio of sentences used in a social story. However, these skills can be easily acquired through Grays (1995) original guideline for developing and implementing social story interventions for children with au tism. A unique feature about social stories is the fact that no intensive training is nece ssary prior to using th em in practice. Overall, social stories are easy to implement across many environments, are cost effective, and can be individualized to m eet the specific needs of a student with difficulties interacting socially. Due to their ease of implementation, a variety of service personnel (e.g., teachers, school psychologist, guidance counselors) would be able to
133 conduct social story interventions. Having a myriad of people available to implement a social story is extremely bene ficial when prior implementation in home-settings has been unsuccessful. For example, for Angelo th ere was concern regarding the familys adherence to the intervention as was origina lly planned. With access to other treatment providers there is a greater li kelihood that social stories wi ll be implemented as planned. Limitations The small sample size and single-subject design of the current study limit the generalizability of the findings to other students. In addition, the generalizability of the treatment effects to children of other disability types (e.g., autism) are further limited due to the homogeneity of the participants. Theref ore, the results of th is study provide an example of an efficacious intervention prot ocol to improve and expand the social repertoires of individuals with AS. Another limitation of this study involves the lack of consistency in educational environments in which the interventions were conducted. Each of th e participants read their respective social story at home with a parent. Because of this, it was not possible to identify whether the manner and/or situation the social story was r ead has any impact on the effectiveness of the intervention. Howe ver, information regarding the individual implementation of the social story interventions is provided with the social story journals completed by the parents. For both Darius and Francis there appeared to be strict adherence to the protocol of reading their re spective social storie s twice a day during the intervention period. Unfortunate ly, social story information for Angelo was unavailable. Therefore, information regarding interventi on adherence for Angelo is questionable.
134 Related to the lack of consistency ac ross environments, the amounts of social consequences for each participant in their respective environments were not assessed. That is, the rates of social consequen ces coming from peers and teachers were not evaluated. Without such information, it is di fficult to identify whether the rehearsal of the social stories was more effective than the subsequent access to the natural reinforcement the participant(s) received for engaging in the target skill. Importantly, this study was not designed to determine the contribution of this social reinforcement apart from the social stories In addition, because there is no standa rd protocol for writing social stories (combining the art of social story construction with the sp ecificity of social story components), it may be difficult to ascertain the effectiveness of such interventions. That is, when a social story does not work fo r one participant (e.g., Angelo), it becomes difficult to identify whether it was an effect of the social story or the concept of the social story that resulted in ineffectiveness or some other extraneous variable or an interaction of these variables. Finally, there was a potential for multiple treatment interference effects due to the fact that the participants prev iously received social skill trai ning (e.g., social skills groups at the CDC). Although there was an attempt to control for such variables, there may have been the possibility that the parents were predisposed to be more effective treatment agents because of their commitment to assi sting their children with developing social skills. However, employing a multiple baselin e condition for each part icipant prior to the beginning of the social story interventions should have minimized this limitation.
135 Recommendations for Future Research The fundamental concern for future resear ch on the effectiveness of social story interventions should be the continued use of procedures and methods that employ experimental control. From a research st andpoint, the establishm ent of experimental control allows the researcher to conclude th at a functional relations hip exists between the independent and dependent variables. Furt her research employing experimental control would provide further validity to the practice of using social st ories for children with AS. More importantly, recent trends in edu cational practice are calling for increased evidenced-based approaches verified thr ough well-controlled research paradigms. Continued efforts using research designs that l ack the rigor of experimental control limits the acceptability of using social stories in clinical practice. Re plication of this study would prove useful for the overa ll research literature suppor ting the use of social story interventions. In addition to demonstrating experimental control, future research endeavors should further examine training for maintenance and generalization of skills following the implementation of social story inte rventions. Training for maintenance and generalization is especially important for populations of children with autism spectrum disorders. Many studies have demonstrat ed that students with autism often do not maintain or generalize behaviors (Kli nger & Dawson, 1996; Simpson & Regan, 1987; Wing, 1998b). These findings have also been reflected in re cent social story research literature (Kuttler, et al., 1998; Swaggart et al., 1995; Thiemann & Goldstein, 2001). Future research should begin the proce ss of delineating the effective components of social stories. That is, identifying whic h of the four steps (ide ntify a target social
136 situation, gather functional assessment data, share information, and develop a new target skill) outlined by Gray (1995) are essential for determining the ta rget skill, as well as the focus of the social story content? Resu lts of this study sugge st that the second component of Grays recommendations may be the most salient. It is important to note that this step is a data collection phase a nd is crucial to establishing the frequency of behavior(s), as well as providing a basis fo r comparing the indivi duals behavior during and following the social story interventi on. Without this component, evidence-based outcomes would be difficult to determine. Future research also needs to explore th e impact of social stories implemented alone versus the impact of social storie s used in combination with reinforcement strategies for engaging in appropriate target behavior in an identified problematic social situation. Specifically, some individuals us ing social stories ma y not be immediately reinforced socially for practicing/engaging in their target skill. In fact, some students may not even attempt to use the target skil l in the setting in which it occurs. By providing some initial reinforcement to the student, it may be possible to jumpstart the relationship between the social story and social behavior, as well as create ongoing social opportunities for practice of the target skill. Eventually, these initial opp ortunities may later promote more spontaneous initiations of the target behavior that are socially reinforced. Additional research endeavors may also wa nt to examine the differential benefits of participation in a similar intervention program with variable treatment durations received between groups. For example, varyi ng the length of time and frequency that the student has access to a social story. Some social skills may take a longer amount of
137 instructional time to bring a student to a pr oficient level, not to mention produce enough opportunities for practice and succe ss of the targeted skill. Therefore, studies examining the effects of varied instructional variables could verify the hypothe sis that increasing the duration and/or frequency of the social story intervention may result in larger treatment benefits, benefits to more pa rticipants, or greater maintena nce of treatment effects. However, the highly individualized nature of ASD would make creating large groups with homogeneity of variance virtually im possible. For group research, larger populations are necessary for increasing the pow er of findings. Gaining access to a large enough population of children with AS would be difficult since AS is still considered a low-incidence disability. In addition, the indi vidualized nature of so cial stories would not bode well with group design research. Because social stories are written specifically for individual cases, finding a meaningful sample of individuals who could benefit from the specific information of one social story woul d prove difficult. Aside from gaining access to large enough populations, as we ll as creating individual social stories, the feasibility of such an endeavor would be extremely limite d. A substantial amount of resources (both financial and man-hours) would be necessary. Finally, additional research is necessary on the differential effects of social stories with various ASD (e.g., Aspergers Syndrom e, Higher-Functioning Autism), as well as a variety of subtypes of typical and non-typical children (e.g., typical peers, EMH, TMH). Furthermore, as the results of this study indi cated, treatment effects were not consistent across all participants. In sp ite of their flexibility, social story research should examine the interaction of participants characteristic s and intervention success. Such that, the
138 characteristics of individuals that contribute to the success of a so cial story intervention are identified. Summary In summary, this study investigated the e ffects of social story interventions for three children with AS. T ogether, the results of this study support previous positive findings regarding the use of social story interventions for children with autism. In addition, the results of this study support clin ical recommendations for using social story interventions to teach prosoc ial skills in children w ith AS (Atwood, 2000; Gray, 1998, Rogers, 2000; Safran, 2001). Because this study represents the first empirical support for social story interventions with children with AS, this information should be used to assist with the development of social story interv entions, as well as pr ovide the foundation for future research. The present research provides no definitive claims of the effectiveness of social story interventions for children with AS Rather, it adds preliminary evidence that social stories may be a beneficial method of remediating social skill difficulties for many children and youth with AS.
139 References Asperger, H. (1944). Die Autistis chen Psychopathen im Kindesalter. Archiv Fur Psychiatrie und Nervenkrankhieten 117 76-136. Autistic psychopathy in childhood (U. Frith, Trans.) In U. Frith (ed.), Autism and Asperger syndrome (pp. 37-92). Cambridge: Cambri dge University Press. Attwood, T. (2000). Strategies for improving the social integration of children with asperger syndrome. Autism 4 (1), 85-100. American Psychiatri c Association (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: APA. Barkley, R. A., (1990). Attention deficit hy peractivity disord er: A handbook for diagnosis and treatment New York: Guilford Press. Barlow, D. H., & Hersen, M. (1984). Single case experimental designs: Strategies for studying behavior change (2nd edition). New York: Pergamon Press. Baron-Cohen, S. (1995). Mindblindness: An essay on autism and theory of mind Cambridge, MA: MIT Press. Bauer, S. (1995). Autism and the pervasive developmental disorders: Part 1. Pediatrics in Review, 16 (4), 130-136. Bauer, S. (1995). Autism and the pervasive developmental disorders: Part 2. Pediatrics in Review, 16 (5), 168-176. Bauer, S. (1996). Asperger Syndrome. Retrieved December 15, 2000. http://www.udel.edu/bkirby/asperger/.
140 Brady, M. P., Shores, R. E., McEvoy, E. D., & Fox, J. J. (1987). Increasing social interactions of severely ha ndicapped autistic children. Journal of Autism and Developmental Disorders 17 375-390. Charlop, M. H., & Milstein, J. P. (1989) Teaching autistic children conversational speech using video modeling. Journal of Applied Behavior Analysis 22 (3), 275285. Church, C., Alisanski, S., & Amanullah, S. (2000). The social, behavioral, and academic experiences of children w ith asperger syndrome. Focus on Autism and Other Developmental Disabilities 15 (1), 12-20. Cohen, D. J., Paul, R., & Volkmar, F. R. ( 1986). Issues in the classification of pervasive and other developmental disorders: Toward DSM-IV. Journal of the American Academy of Child Psychiatry 25 213-220. Cosmides, L. (1989). The logic of social exchange: Has natural selection shaped how humans reason?: Studies with the Wason selection task. Cognition 31 187-276. Ferron, J., & Jones, P. (2002, April). Visual tests for the analysis of multiple-baseline data Paper presented at the American Educational Research Association (AERA) Conference, New Orleans, LA. Frith, U. (1989). Autism and theory of mind. In C. Gillberg (Ed.), Diagnosis and Treatment of Autism (pp. 33-52). New York: Plenum. Fullerton, A., Stratton, J., Coyne, P., & Gray, C. (1996). Higher functioning adolescents and young adults with autism. Austin, TX: PRO-ED.
141 Gillberg, C., & Ehlers, S. (1998). High-functioning people with autism and asperger syndrome: A literature review. In E. Schopler, G. B. Mesibov, & L. J. Kunce (Eds.), Asperger syndrome or high-functioning autism? (pp. 79-106). New York: Plenum. Gray, C. & Jonker, S. et. al. (Eds.) (1994). The social story book 1994 Arlington: Future Education. Gray, C. A., (1994, May). Social interventions with high-functioning people with autism Paper presented at the Treatment a nd Education of Autistic Children and Other Communication Handicapped Childre n (TEACCH) Conference, Chapel Hill, NC. Gray, C. A., (1995). Teaching ch ildren with autism to read so cial situations. In K. A. Quill (Ed.), Teaching children with autism (pp.219-241). New York: Delmar. Gray, C. A., (1997, February). Social st ories and comic strip conversations: Unique methods to improve social understandi ng. Paper presented at the Autism 1997 Conference of Future Horizons, Inc., Athens, GA. Gray, C. A., (1998). Social stories and comi c strip conversations with students with asperger syndrome and high-functioning au tism. In E. Schopler, G. B. Mesibov, & L. J. Kunce (Eds.), Asperger syndrome or high-functioning autism? (pp. 167198). New York: Plenum. Gray, C. A., & Garrand, J. D. (1993). Soci al stories: Improving responses of students with autism with accurate social information. Focus on Autistic Behavior 8 (1), 1-10.
142 Gresham, F. M. (1989). Assessment of treat ment integrity in school consultation and prerefferral intervention. School Psychology Review 18 (1), 37-50. Gresham, F. M., & Elliot, S. N. (1990). Social Skills Rating System. Circle Pines: American Guidance Service. Hagiwara, T. & Myles, B. S. (1999). A mu ltimedia social story intervention: Teaching skills to children with autism. Focus on Autism and Other Developmental Disabilities 14 (2), 82-95. Hyman, S. L., Rodier, P. M., & Davidson, P. (2001). Pervasive developmental disorders in young children. Journal of the Amer ican Medical Association, 285(24), online resource. Jacob-Timm, S., & Hartshorne, T. S. (1998). Ethical-legal issues in the education of pupils with disabilities under IDEA In J. Wiley & Sons (Eds.), Ethics and law for school psychologists. (pp. 95-147). New York: John Wiley. Jacobson, N. S., & Truax, P. (1991). Clinical significance: A sta tistical approach to defining meaningful change in psychotherapy research. Journal of Consulting and Clinical Psychology 59 (1), 12-19. Kanner, L. (1943). Autistic distur bances of affective contact. Nervous Child 2 217-250. Kazdin, A. E. (1982). Singl e-case research designs: Method s for clinical and applied settings. New York: Oxford Press. Klin, A., & Volkmar, F.R., (1995). Autism and the pervasive developmental disorders Child and Psychiatric Clinics of North America 4 617-630.
143 Klin, A., & Volkmar, F. R., (1997). Aspergers syndrome. In D. J. Cohen & F. R. Volkmar (Eds.), Handbook of autism and pervasive developmental disorders (pp. 91-112). New York: Wiley. Klin, A., & Volkmar, F. R., (1999). Au tism and other pervasive developmental disorders. In S. Goldstein & C. R. Reynolds (Eds.) Handbook of neurodevelopmental and geneti c disorders in children (pp. 247-274). New York: The Guilford Press. Klin, A., & Volmar, F. R. (2000). Treatment and intervention guidelines for individuals with asperger syndrome. In A. Klin, F. R. Volkmar, & S. S. Sparrow (Eds.), Asperger syndrome ( pp. 340-366). New York: Guilford Press. Klin, A., Volkmar, F. R., Sparrow, S. S., Cicchetti, D. V., & Rourke, B. P. (1995). Validity and neuropsychological charact erization of asperger syndrome: Convergence with nonverbal lear ning disabilities syndrome. Journal of Child Psychology and Psychiatry 36 (7), 1127-1140. Klinger, L. G., & Dawson, G. (1996). Autistic di sorder. In E. J. Mash, & R. A. Barkley (Eds.), Child Psychopathology (pp. 311-339). New York: The Guilford Press. Kugler, B. (1998). The differentiation be tween autism and asperger syndrome. Autism 2 (1), 11-32. Kuttler, S., Myles, B. S., & Carlson, J. K. (1998). The use of social stories to reduce precursors to tantrum behavior in a student with autism. Focus on Autism and Other Developmental Disabilities 13 (3), 176-182. Leslie, A. (1987). Pretense and represenat ion: The origins of theory of mind. Psychological Review 94, 412-426.
144 Leslie, A. (1993). What autism teaches us a bout metarepresentation. In S. Baron-Cohen, H. Tager-Flusberg, & D. Cohen (Eds.), Understanding other minds: Perspectives from autism (pp.83-111). New York: Oxfo rd University Press. Leslie, A., & Thaiss, L. (1992). Domain specificity in conceptual development: Neuropsychological evidence from autism. Cognition 43 225-251. Lorimer, P. A., Simpson, R. L., Myles, B. S., & Ganz, J. B. (2002). The use of social stories as a preventative behavioral in tervention in a home setting with a child with autism. Journal of Positive Behavior Interventions 4 (1), 53-60. Manjiviona, J., & Prior, M. (1999). Neuropsychological profile s of children with asperger syndrome and autism. Autism 3 (4), 327-356. Mayer-Johnson, R. (1981). The picture communication symbols book. Solona Beach, CA: Mayer-Johnson. Mesibov, G. B., (1992). Treatment issues w ith high-functioning adolescents and adults with autism. In E. Schopler and G. B. Mesibov (Eds.), High-functioning individuals with autism. (pp. 143-156). New York: Plenum. Miller, J. N., & Ozonoff, S. (2000). The extern al validity of asperger disorder: Lack of evidence from the domain of neuropsychology. Journal of Abnormal Psychology 109 (2), 227-238. Mundy, P., & Stella, J. (2001). Joint atte ntion, orienting, and nonverbal communication in autism. In A. M. Wetherby & B. M. Prizant (Eds.) Autism Spectrum Disorders: A transactional developemental perspective. (pp. 55-77). Baltimore: Brookes.
145 Nirje, B. (1969). The normalization principl e and its human management implications. In R. Kugel and W. Wolfensberger (Eds.), Changing patterns in residential services for the mentally retarded (pp. 181-195). Washington, DC: US Government Printing Office. Norris, C., & Dattilo, J. ( 1999). Evaluating effects of a social story intervention on a young girl with autism. Focus on Autism and Other Developmental Disabilities 14 (3), 180-186. Ozonoff, S., Rogers, S. J., & Pennington, B. F., (1991). Aspergers syndrome: Evidence of an empirical distinction fr om high-functioning autism. Journal of Child Psychology and Psychiatry 32 (7), 1107-1122. Ozonoff, S., South, M., & Miller, J. N. ( 2000). DSM-IV-defined asperger syndrome: Cognitive, behavioral and early history differentiation from high-functioning autism. Autism 4 (1), 29-46. Prior, M., Leekam, S., Ong, B., Eisenmajer, R., Wing, L., Gould, J., & Dowe, D. (1998). Are there subgroups within th e autistic spectrum? A cluste r analysis of a group of children with autistic spectrum disorders. Journal of Child Psychology and Psychiatry 39 (6), 893-901. Risley, T. R., & Wolf, M. M. (1972). Strate gies for analyzing behavioral change over time. In J. Nesselroade & H. Reese (Eds.), Life-span developmental psychology: Methodological issues (pp. 175-183). New York: Academic Press. Rogers, S. J. (2000). Interventions that facili tate socialization in ch ildren with autism. Journal of Autism and Developmental Disorders 30 (5), 399-409.
146 Rutter, M., Greenfield, D., & Lockyer, L. ( 1967). A fiveto fifteen-year follow-up study of infantile psychosis: Social and behavioral outcome. British Journal of Psychiatry 113 1183-1199. Rutter, M., & Schopler, E. ( 1992). Classification of pervas ive developmental disorders: Some concepts and prac tical considerations. Journal of Autism and Developmental Disorders 22 (4), 459-482. Rietschel, L. (2000). How did we get here? In A. Klin, F. R. Volkmar, & S. S. Sparrow (Eds.), Asperger syndrome ( pp. 448-453). New York: Guilford Press. Sacks, O. (1994). An anthropologist on mars New York: Vintage Books. Safran, S. P. (2001). Asperger syndrome: Th e emerging challenge to special education. Exceptional Children 67 (2), 151-160. Schopler, E., Reichler, R. J., & Renner, R. R. (1988). Child Autism Rating Scale Los Angeles: Western Psychological Services. Shery, L. S. (2000). A view from inside. In A. Klin, F. R. Volkmar, & S. S. Sparrow (Eds.), Asperger syndrome ( pp. 443-447). New York: Guilford Press. Simpson, R. L., & Regan, M. (1988). Manageme nt of autistic behavior. Austin, TX: PRO-ED. Stein, M. A., Szumowski, E., Sandoval, R., Na delman, D., OBrien, T., Krasowski, M., & Phillips, W. (1994). Psychometric properties of the childrens atypical development scale. Journal of Abnormal Child Psychology 22 (2), 167-176. Stokes, T. F., & Baer, D. M. (1977). An implicit technology of generalization. Journal of Applied Behavior Analysis 10 349-368. Swaggart, B. L., Gagnon, E., Bock, S. J., Ea rles, T. L., Quinn, C., Myles, B. S., &
147 Simpson, R. L. (1995). Using so cial stories to teach social and behavioral skills to children with autism. Focus on Autistic Behavior 10 (1), 1-15. Szatmari, P., Tuff, L., Finlayson, M. A., Bart olucci, G. (1989). Aspergers syndrome and autism: Neurocognitive aspects Journal of the American Academy of Child and Adolescent Psychiatry 29 (1), 130-136. Thiemann, K. S., & Goldstein, H. (2001). Soci al stories, written text cues, and video feedback: Effects on social communication of children with autism. Journal of Applied Behavior Analysis 34 (4), 425-446. Twachtman-Cullen, D. (1998). Language and communication in high-functioning autism and asperger syndrome. In E. Schopler, G. B. Mesibov, & L. J. Kunce (Eds.), Asperger syndrome or high-functioning autism? (pp. 199-225). New York: Plenum. Twachtman-Cullen, D. (2000). More able ch ildren with autism spectrum disorders: Sociocommunicative challenges and guideli nes for enhancing abilities. In A. M. Wetherby & B. M. Prizant (Eds.) Autism Spectrum Disorders: A transactional developemental perspective. (pp. 225-249). Baltimore: Brookes Publishing. Volkmar, F. R., & Klin, A. ( 2000). Diagnostic issues in asperger syndrome. In A. Klin, F. R. Volkmar, & S. S. Sparrow (Eds.), Asperger syndrome ( pp. 25-71). New York: Guilford Press. Volkmar, F. R., Klin, A., Schultz, R. T., Rubin, E., & Bronen, R. (2000). Aspergers disorder. American Journal of Psychiatry 157 (2), 262-267. Wallace, J. (2000). Walter. In A. Klin, F. R. Volkmar, & S. S. Sparrow (Eds.), Asperger syndrome ( pp. 434-442). New York: Guilford Press.
148 World Health Organization. (1992). International Classificat ion of Diseases: Tenth Revision Chapter V. Mental and behavioral disorders (including disorders of psychological development). Diagnostic cr iteria for research. Geneva: Author Williams, K. (1995). Understanding the student with aspergers syndrome: Guidelines for teachers. Focus on Autistic Behavior 10 (2), Wing, L. (1981). Aspergers syndrome: A clinical account. Psychological Medicine 11 115-129. Wing, L. (1998a). The continuum of autisti c characteristics. In E. Schopler & G. Mesibov (Eds.), Diagnosis and Assessment in Autism (pp. 91-110). New York: Plenum Press. Wing, L. (1998b). The history of asperger s yndrome. In E. Schopler, G. Mesibov, & L. J. Kunce (Eds.), Asperger Syndrome or High-Functioning Autism? (pp. 11-28). New York: Plenum Press. Wing, L. (2000). Past and future of research on asperger syndrome. In A. Klin, F. R. Volkmar, & S. S. Sparrow (Eds.), Asperger syndrome ( pp. 418-432). New York: Guilford Press. Wing, L., & Gould, J. (1979). Severe impairme nts of social interaction and associated abnormalities in children: Epidemiology and classification. Journal of Autism and Developmental Disorders, 9 11-29.
150 Appendix A DSM-IV Criteria for Autistic Disorder* A.) A total of at least six items from (1), (2 ), and (3), with at least two from (1), and one each from (2) and (3). (1) Qualitative impairment in social interacti on, as manifested by at least two of the following: a. Marked impairment in the use of multiple nonverbal behaviors, such as eye-to-eye gaze, facial expression, body postures, and gestures to regulate social interaction. b. Failure to develop peer relationships appropriate to developmental level. c. A lack of spontaneous seeking to share enjoyment, interests, or achievements with other people (e.g., by a lack of showing, brining, or pointing out objects of interest). d. Lack of social or emotional reciprocity. (2) Qualitative impairments in communication, as manifested by at least one of the following: a. Delay in, or total lack of, the de velopment of spoken language (not accompanied by an attempt to compensa te through alternative modes of communication such as gesture or mime). b. In individuals with adequate speech, marked impairment in the ability to initiate or sustain a c onversation with others. c. Stereotyped and repetitiv e use of language or idiosyncratic language. d. Lack of varied spontaneous make-believe play or social imitative play appropriate to developmental level. (3) Restricted, repetitive, an d stereotyped patterns of behavior, interests, and activities, as manifested by at least one of the following: a. Encompassing preoccupation with one or more stereotyped and restricted patterns of interest that is abnormal either in intensity or focus. b. Apparently inflexible adherence to specific, nonfunctional routines or rituals. c. Stereotyped and repetitiv e motor mannerisms (e.g., hand or finger flapping or twisting or complex whole body movements). d. Persistent preoccupation with parts of objects. B.) Delays or abnormal functioning in at least one of the follow ing areas, with onset prior to age 3 years: (1) social interacti on, (2) language as used in social communication, and (3) symbolic or imaginative play. C.) Not better accounted for by Rett diso rder or childhood disintegrative disorder. *American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders 4th ed. 1994.
151 Appendix B DSM-IV Criteria for Asperger Disorder* A.) Qualitative impairment in social interact ion, as manifested by at least two of the following: 1. Marked impairment in the use of multiple nonverbal behaviors, such as eye-to-eye gaze, facial expression, body postures, and gestures to regulate social interaction. 2. Failure to develop peer relationships appropriate to developmental level. 3. A lack of spontaneous seeking to share enjoyment, interests, or achievements with other people (e.g., by a lack of showing, brining, or pointing out objects of interest). 4. Lack of social or emotional reciprocity. B.) Restricted, repetitive, and stereotyped patterns of behavior, interests, and activities, as manifested by at least one of the following: 1. Encompassing preoccupation with one or more stereotyped and restricted patterns of interest that is abnormal either in intensity or focus. 2. Apparently inflexible adherence to specific, nonfunctional routines or rituals. 3. Stereotyped and repetitiv e motor mannerisms (e.g., hand or finger flapping or twisting or complex whole body movements). 4. Persistent preoccupation with parts of objects. C.) The disturbance causes clinically signi ficant impairment in social, occupational, or other important areas of functioning. D.) There is no clinically significant de lay in language (e.g., single words used by age 2 years, communica tive phrases used by age 3 years). E.) There is no clinically significant delay in cognitive development or in the development of age-appropriate self-help sk ills, adaptive behavior (o ther than in social interaction), and curiosity about the environment in childhood. F.) Criteria are not met for another pervasive development disorder or schizophrenia. *American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders 4th ed. 1994.
152 Appendix C Examples of Social Stories* Giving a Gift A gift is something you give someone. People give other people gifts. Some gifts are big. Some gifts are small. When I give someone a gift I might say, Heres a gift for you. It is polite to say, Heres a gift for you. People say, Heres a gift fo r you, because it is correct. Sometimes people give me a gift. When people give me a gift, I will try to say, Thank you. Saying, Thank you, is polite. People like to hear, Thank you, after they give someone a gift. Playing Fairly It is a good idea to play fairly with my friends. Sometimes my friend may wi n the game we are playing. I will try to stay calm if my friend wins a game. If my friend wins a game, I will ask them to play again. It is good to play fairly at games. Future Education The Social Story Book 1994
153 Appendix D Social Story Information Form* General Information Name ____________________________________ School Year _____________________ Grade______ Teacher________________ School General Academic Information: Interests/Special Abilities: Reading Level: Comprehension: Math: Day/Time for: Art: Music: PE: P.T.: O.T.: Speech: Other(s): Names of a few classmates/friends: Other general information: Observation Notes Targeted situation: Time: Day(s): General description of targeted situation: Childs current response: Always?: Desired response: Teacher(s) attribute the response to: Parents attribute the response to: Child attributes the response to:
154 Appendix D (Continued) Observation Date: / / Day: Time: Notes: Future Education The Social Story Book 1994
155 Appendix E Social Story Implementation Plan* Name: Date: Title of Story: Story Format: Printed Story: Story & Audio Cassette: Story on Video: Suggested Implementation: Begin Implementing story on / / 1. To introduce the story: 2. Review schedule: 3. Monitoring responses: Progress review dates: / / / / / / / / Suggested fading procedure: fading by writing re-vising review schedule ____decrease verbal or other cues ____ other Support Materials and Activities: revise posted classroom schedule story bookmark(s) revise, modify written classroom rules reminder sign Keep Me Posted Notes story passes story folder Daily Oral Langauge (DOL) social calender/goals other Describe:
156 Appendix E (Continued) Date/Notes of second observation: Date / / Day: __ Time: Notes: Factors which may fluctuate, change, etc.: Possible variations which may apply: fill in the blanks generalizing with other stories checklist story story to address fears curriculum story judgment story Ideas: Future Education The Social Story Book 1994
157 Appendix F Social Story Report* Name:_ Date: / / Story Title: Support Materials or Activities: Current Review Schedule: Do you feel the story is hav ing a positive impact? Childs Current Response(s) to the targeted situation: Childs reaction to the story: Any problems with the story? No Yes Explain: Suggestions for revisions, etc.: Please return report to: by: *Future Education The Social Story Book 1994
158 Appendix G Social Story for Darius Sportsmanship: Page 1: My Social Story Book: Sportsmanship Page 2: After lunch we go to recess. Sometimes, recess is on the playground. Page 3: When I get to the playground, I like to play football. There are other kids who like to play football with me too. Page 4: When I play football, I should remember to be a good sport, and never let losing the game bother me. Page 5: A good sport is someone who says good job or awesome during a good play or for winning the game, no matte r whose team they are on. A good sport would never yell at anybody wh ile playing football during recess. That hurts others feelings. Page 6: I will try to practice my sportsmanship sk ills when I play football during recess. I will try to say things like good job, nice pass, or awesome. Page 7: If I show all of these sportsmanship skills and dont get mad, my friends will want to play with me more often. Page 8: I want kids to like me and play with me, so I need to show good sportsmanship skills so they will think of me as a fair person.
159 Appendix H Social Story for Francis Joining In with Friends: Page 1: My Socials Story Book: Joining In with Friends Page 2: Usually, I have recess most of the days I go to school. Page 3: Recess is a time that I can go outside. I can walk and run around, or I can talk to other kids. Page 4: Most of the time during recess, I like to ta lk to other kids because I love to hear what other people have to say. Page 5: Looking at people is a VERY important part of talking to them, and it is a very nice thing to do. Page 6: I should never turn and walk away from anyone when I am talking, unless I am asked to do so. If I turn away, I will hurt their feelings, and they will think I dont care about them. Page 7: I would be hurt if my friends walked aw ay from me, and my friends may not want to talk to me again. Page 8: When I am talking to other kids at recess, I should try to look at a part of their face and listen to what they say. Page 9: I want to be a good friend, and Im proud to be a good friend.
160 Appendix I Social Story for Angelo Joining In with Friends: Page 1: My Socials Story Book: Joining In with Friends Page 2: I love playing with ot her kids during recess. Page 3: The most fun times I have are when I jo in kids who are already playing soccer on the playground. Page 4: I do this by asking in a nice gentle voice, Can I play with you? Page 5: If they say yes, then I ask, Show me how to play. Page 6: As long as I keep joining other kids, I will be popular and have lots of friends. Page 7: When I join other kids who are playing, I re ally feel like I have friends and that I belong to the group.
161 Appendix J SIRS Social Interaction Recording System Child Observed: School Name: Date: Observer: Reliability Observer: Time of Observation: Setting: PE Class Other: Recess/Playground Cafeteria/Lunch Partial 1 2 3 4 5* 6 7 8 9 10* S P ABx PBx AbBx Partial 11 12 13 14 15* 16 17 18 19 20* S P ABx PBx AbBx Partial 21 22 23 24 25* 26 27 28 29 30* S P ABx PBx AbBx Partial 31 32 33 34 35* 36 37 38 39 40* S P ABx PBx AbBx Partial 41 42 43 44 45* 46 47 48 49 50* S P ABx PBx AbBx ABx = Active display of target behavior PBx = Passive display of target behavior AbBx = Absence of target behavior
162 Appendix J (Continued) Partial 51 52 53 54 55* 56 57 58 59 60* S P ABx PBx AbBx Partial 61 62 63 64 65* 66 67 68 69 70* S P ABx PBx AbBx Partial 71 72 73 74 75* 76 77 78 79 80* S P ABx PBx AbBx Partial 81 82 83 84 85* 86 87 88 89 90* S P ABx PBx AbBx Partial 91 92 93 94 95* 96 97 98 99 100* S P ABx PBx AbBx Target Student *Peer Comparison S ABx: % ABx: P ABx: % ABx: S PBx: % PBx: P PBx: % PBx: S AbBx: % AbBx: P AbBx: % AbBx: Total Intervals Observed: Total Intervals Observed:
163 Appendix K OASIS Observation of Appropriate Social Interaction Skills Section I: Identifying Information Childs Name:_______________________________ Male______ Female______ School:_____________________________________ Grade:_____ Date:_____________________ Observers Name:________________________________ Reliability Observer:______________________________ Section II: Response Record Carefully read each item. Ask yourself if the child can do what the item says. Check either Yes or No by each item. If you are uncertain or doubt that the child can do what the item states, check No Check Yes for those items that the child can do right now or is beginning to do. Check No if the child cannot do what the item says. Remember, if you have not heard it or seen it, mark No Yes No ___ ___ 1. Smiles at a familiar person. ___ ___ 2. Calls peers by their name. ___ ___ 3. Ask questions using words such as who, what, and where. ___ ___ 4. Starts a conversation with his or her peers. ___ ___ 5. Refers to himself or herself by name. ___ ___ 6. Makes eye contact with peers cl ose to him or her for at least 5 seconds. ___ ___ 7. Uses age-appropriate langua ge to talk to other peers. ___ ___ 8. Responds to other peers verb ally, physically, or gesturally. ___ ___ 9. Engages in reciprocal conversations with peers. ___ ___ 10. Hands something to or receives something from peers ___ ___ 11. Invites others to join in activities. ___ ___ 12. Gives compliments to peers. ___ ___ 13. Cooperates with peers without prompting. ___ ___ 14. Joins ongoing activity or gr oup without being told to do so. ___ ___ 15. Accepts peers ideas for group activities. Total Appropriate Skills _____ Percentage Appropriate Skills _____
164 Appendix L Social Story Journal Subjects Social Story Journal Todays Date: Time: AM/PM I read my story called: I read it: At home In the bus/car On Playground At Lunch Other I read it with: A friend Just Me An Adult Other___ The story makes me: Happy Sad Mad Confused
165 Appendix M Joining In Response Definitions Actively Joining In (AJI): Definition : Instances in which the target child is contributing to a reciprocal conversation or is actively participating in some play activity with one or more children. Examples of AJI: Nonexamples of AJI: Playing next to each other and using a variety of social exchanges to show their recognition of peers (e.g ., talking to a peer about a game that is being played on the playground). Engaging in any sort of aggressive act such as cursing, shouting, pushing, name calling, hitting, and making forceful bodily contact with someone else during a conversation or a play activity (AbJI). Borrowing or lending toys, using each others toys, or sharing accomplishments (e.g., successfully bu ilding a tower with blocks). Making verbalizations with bizarre content, such as delayed echolalia consisting of dialogue from a television show (AbJI). Participating in any type of organized group game that involves taking turns (e.g., tag, hide-and-seek ) or fulfilling a group role (e.g., playing goalie in soccer). Walking up and observing without actively joining in (PJI). Making small talk with another peer (e.g., did you see the football game yesterday?). Playing independently and separately from peers (AbJI). Passively Joining In (PJI): Definition : Instances in which the target child is playing next to peers, using the same toys or engaged in the same ac tivity, but playing independently of those peers. This also includes times when students are engagi ng in conversation, rather than play*. Examples of PJI: Nonexamples of PJI: Mimicking the behavior(s) of other children who are close to them. Aimlessly walking around a peer group talking to himself (AbJI). Watching a group of peers playing a game and following the group format alone. Silently reading while peers are actively playing around the ta rget child (AbJI) Walking up to or sitt ing with a group of peers and listening to conversation without participating. Not using or sharing any of the toys with those around him (AbJI). Watching or listening to peers but participating in the reciprocal interaction (AJI)
166 Appendix M (Continued) Absence of Joining In (AbJI): Definition : Instances in which the target child do es not display any verbal, physical, or gestural initiations or responses to peers*. Examples of AbJI: Nonexamples of AbJI: Engaging in a game or other activity by himself. Actively participating in an activity/conversation with peers (AJI) Not responding to his or her name being called by another peer. Child walks up to a group of peers and listens (PJI) Refusing to share toys on the playground Fighting, name calling, and making forceful bodily contact with someone else. The primary distinction between PJI and AbJI behaviors is that in the instance of AbJI, the student is not physically near othe r peers during play or socialization.
167 Appendix N Sportsmanship Response Definitions Active Sportsmanship (ASP): Definition : Instances in which the target child actively treats teammates, opponents, and/or coaches with respect. Active displa ys of sportsmanship should demonstrate an awareness of encouraging group play through ve rbal, physical, or gestural means. Examples of ASP: Nonexamples of ASP: Offering positive encouragement during play activity (e.g., saying good job, way to go, awesome, good luck, maybe next time, good effort). Allowing a new child to play in the activity who does not understand the rules (PSP). Cheering for opponents or own team members after a good play or score (e.g., clapping when a team scores, telling the other team good play). Accepting the out performance of an opposing team member without making excuses and/or negative comments (PSP) Helping a player up off the ground. Refusing to shake hands (AbSP) Maintaining integrity during play activity (e.g., congratulating opposing team for winning, not accepting another players poor behavior during the play activity). Talking down to other players on opposing or own team (e.g., you suck, I was open.) (AbSP). Passive Sportsmanship (PSP): Definition : Instances in which the target child attends to the parameters of a game/activity with a full commitment to participatory cohe sion. Passive displays of sportsmanship should demonstrate aspects of fairness and plays in an appropriate manner. Examples of PSP: Nonexamples of PSP: Following and/or playin g within the rules of the game (e.g., not cheating, not intentionally hurting an opponent or other team member). Offering encouraging comments to opposing or own team (e.g., nice pass, good work, maybe next time) (ASP). Accepting a bad play or bad call without dissent (e.g., not argui ng with the team if ball is not caught; maintaining control when ball is dropped, walking back to play without saying a word). Going with the flow of the game. Kicking dirt/sand on the shoes of opposing team member who scores. (AbSP). Adjusting play to keep the game fun (e.g., allowing an extra player; taking turns sitting out of a game; Offering help to opposing or own team member. (ASP).
168 Appendix N (Continued) Examples of PSP: Nonexamples of PSP: Allowing the other team to start play first. Making excuses for lack of ability during the game (AbSP). Accepting the loss of the game without making negative comments to self or others. Absence of Sportsmanship (AbSP): Definition : Instances in which the target child do es not display any verbal, physical, or gestural sportsmanship initiations or responses to peers nor displays equality and fairness in his play. Examples of AbSP: Nonexamples of AbSP: Arguing with opponents or own team members (e.g., yelling at other players, making excuses; complaining about team members performance). Cheering all good plays throughout the game whether made by own or opposing team (ASP). Making negative comments or sarcastic remarks to opponents or own team members (e.g., cheering when his team scores, talking down to other players, laughing when someone gets hurt). Telling the opposing team or own team members good play during the game (ASP). Intentionally violating the rules of the game (e.g., cheating). Accepting the loss of a game graciously (PSP). Making aggressive actions towards opponents or own team members (e.g., running after opponent who just scored, throwing the ball away from players when other team scores). Following and/or playin g within the rules of the game (e.g., not cheating, not intentionally hurting an opponent or other team member) (PSP).
169 Appendix O SSRS Pre/Post-Test Ratings Pre-Test Raw Score Means Post-Test Raw Score Means Average Raw Mean Score Reliable Change Index Parent TeacherParentTeacher Parent Teacher Parent Teacher Darius 26 25 35 35 51.3 42.3 1.98* 1.88 Francis 52 39 56 44 51.3 42.3 .88 1.50 Angelo --32 --39 51.3 42.3 --1.32 indicates a clinical ly significant change
170 Appendix P CADS Pre/Post-Test Ratings Pre-Test Raw Score Means Post-Test Raw Score Means Reliable Change Index Parent TeacherParentTeacher Parent Teacher Darius 41 15 26 9 2.97* 1.19 Francis 32 17 18 23 2.77* 1.19 Angelo ------------* indicates a clinic ally significant change
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001681124
007 cr mnu|||uuuuu
008 060106s2003 flu sbm s000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0000621
Sansosti, Frank J.
Effectiveness of social story interventions for children with asperger's syndrome
h [electronic resource] /
by Frank J. Sansosti.
[Tampa, Fla.] :
b University of South Florida,
Thesis (EdS)--University of South Florida, 2003.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
Title from PDF of title page.
Document formatted into pages; contains 180 pages.
ABSTRACT: The purpose of this study was to investigate the effects of individualized social story interventions on the social communication skills of three children with Asperger's Syndrome (AS). Using a multiple baseline across participants desing, three social stories were implemented and direct observations of the participants' identified target behaviors were collected three times per week during unstructured school activities (e.g, recess). Data revealed an increase in the social communication skills of two of the three participants when the treatment was implemented. In addition, maintenance of treatment effects was observed in two participants. These data support recommendations for using social stories to teach social skills to children diagnosed with autism spectrum disorders.
Adviser: Powell-Smith, Kelly A.
Autism spectrum disorders.
Social skills training.
x Interdisciplinary Education
t USF Electronic Theses and Dissertations.