Superhero social skills for children with autism spectrum disorders: comparing video-based instruction to traditional didactic instruction

The current study compared the efficacy of video-based social skills instruction for children with autism to traditional didactic instruction of social skills. The study consisted of two separate social skills groups. Two children with Autism Spectrum Disorder (ASD) and two typically developing same-aged peers were included in an experimental group that received video-based social skills instruction. Two other children with ASD and 2 additional typically developing same-aged peers were included in the group that received traditional didactic social skills instruction. The social skills lessons were taught twice per week for 8 weeks. Generalization probes of social interaction during free play time were conducted during analog free time intervals for each child with autism. At the end the program, effect sizes and PND were calculated to examine differences in the amount of social interaction during free time periods for the two groups, and pre- and postmeasures of social responsiveness were compared.

SUPERHERO SOCIAL SKILLS FOR CHILDREN WITH AUTISM SPECTRUM DISORDERS: COMPARING VIDEO-BASED INSTRUCTION TO TRADITIONAL DIDACTIC INSTRUCTION by Keith Radley III in partial fulfillment of the requirements for the degree of Master of Science Department May 2010 SKJLLS BASED C A thesis submitted to the faculty of The University of Utah fu lfi llment Deparunenl of Educational Psychology The University of Utah May2010 Copyright © Keith C Radley III 2010 All Rights Reserved to 20 I 0 Righls The G r a d u a t e S c h o o l T H E U N I V E R S I T Y O F U T AH S U P E R V I S O R Y C O M M I T T E E A P P R O V A L thesis Keith C Radley III by vote be Date m i * > | i o illiam Date laine C l a rk Date The Graduate School THE UNIVERSITY OF UTAH SUPERVISOR Y COMMITIEE APPROV AL of a thesis submitted by ______ ~K~el~·tb C~.d~y ~II~I ______ _ This thesis has been read by each member of the following supervisory committee and by a majority vOle has been found to be satisfactory. L\ b..(o(ro Chair: iIIiam R. Jenson /" r ,~ ___ Dale Peter Nicholas The G r a d u a t e S c h o o l T H E U N I V E R S I T Y O F UTAH F I N A L R E A D I N G A P P R O V A L the Keith C I I I m its f m a l f ° r m and have found that 1) format, figures, satisfactory committee and is ready for submission to The L\ | 2^(p I \ Q (^jfyL^~-" ~~~ ' Chair Willirfm R. Jenson Department M f Z f e A o <f S . , / Date C l a rk Request for Publication Embargo I request that The Graduate School, the University of Utah, delay the publication of my thesis or dissertation through ProQuest/UMI and the University of Utah Libraries Institutional Repository (USpace) for a period of six months. The chair of my supervisory committee joins me in this request for an embargo. I am requesting this embargo because (check one): 1~11 have a patent pending through the University of Utah Technology Transfer Office. LZII have a copyright issue with the Journal of I understand that the embargo may be lifted at my written request. I also understand that any extension to the approved embargo must be submitted in writing before the embargo is lifted and that the extension will be for an additional six months. , Student Date Committee Date The embargo is for a period of six months beginning with the date this form is approved by the Dean of The Graduate School. Date G r a d u a t e Council tjlf^" . Dean of The Graduate School Charles A. Wight r The Graduate School THE UNIVERSITY OF UTAH FINAL READING APPROVAL To the Graduate Council of the University of Utah: I have read lhe thesis of ____- --'K=ei"t'h"-C"- Radley Ul in its fmal foml and have found that ( I) its fannat. citations, and bibliographic style are consistent and acceptable; (2) its illustrative materials including figures. tables, and charts are in place; (3) the final manuscript is sati sfactory to the supervisory LJ/z..CP/IO Date e Graduate School. • Chair of the Supervisory Committee Approved for the Major Department · Chair I Dean Elaine Clark Please DO NOT complete "this boxed section UNLESS you are requesting a publication embargo. EmbargO' requesl lhat Schoo], Uni versity Utah. ProQuestlUMI Ulah Libr.:tries U5pace) supervisory committee joins me in this request for an embargo. I am requesting this embargo because ~check one): o I lhrough Uni versity Tr::lI)sfel" o I i~sue loumal I undeNtand that the embargo may lifted at my written reques t. T also undeNland lha! any ex tension to the approved embargo submiued lhe theexlension will be · Student Dale , Chair of the Supervisory Commjttee Dare The embargo is for a period of six monlhs beginning with the dale this form is approved by the Dean of The GrJduate 5::';;h:o::o:;I;.,---, Approved for the Graduate Council -1.JJ);~'iI- . Dean or The Graduale School Cbarles A4if{:.~ , ABSTRACT video-instruction autism instruction skills. children with Autism Disorder (ASD) and two typically developing same-aged peers were included in an experimental in the group that received traditional didactic social skills instruction. The social skills Generalization interaction during free play time were conducted during analog free time intervals for child autism. calculated to examine differences in the amount of social interaction during free time periods for the two groups, and pre- and postmeasures of social responsiveness were compared. The current study compared the efficacy of vidl!o-based social skills in struction for children with auti sm to traditional didactic in struction of social skill s. The study consisted of two separate social skills groups. Two chi ldren wi th Auti sm Spectrum develo ping pee rs ex:perimental group that received video-based social skills instruction. Two other children with ASD and 2 additional typically developing same-aged peers were included th at didact ic ski lls lessons were taught twice per week for 8 weeks. Generali zatio n probes of social (ilne interva ls each c hild with auti sm. At the end of the program, effect sizes and PND were calcu lated socia l in teraction TABLE OF CONTENTS iv LIST viii 1. 1 Evidence-Based Practice 3 Meta-Analyses 10 Meta-Analyses of General Interventions for ASD 12 Meta-Analysis of Language Interventions for ASD 14 Meta-Analysis of Social Skills Interventions for ASD 15 Current Social Skills Programs 19 Social Skills Programs for Autism Spectrum Disorders 24 Video Modeling Interventions 24 Peer-Mediated Interventions 27 Self-Management Interventions 27 Social Stories 28 Superhero Social Skills 29 34 39 Setting 44 Measures 44 System 44 Social Scale 45 Autism Social Skills Profile 46 Power Cards 47 ABSTRACT .......................................................................... ........................... IV LIST OF TABLES .... .... ....... ... ................................ ......... .... ... ........ ............. ..... vi ii LIST OF FIGURES ............................................................................................ ix Chapter I. INTRODUCTION ...................................... ..................... ................... ..... I Evidence-Based Practice ................. ............................................. 3 Meta-Analyses ..................... ........... .......... ....................... .......... 10 Meta-Analyses of General Interventions for ASD ..................... 12 Meta-Analysis of Language Interventions for ASD ............ ...... 14 Meta-Analysis of Social Skills Interventions for ASO .............. 15 Current Social Skills Programs ... ............. ... .. .. .... .... ... ................ 19 Social Skills Programs for Auti sm Spectrum Disorders ......... ... 24 Video Modeling Interventions ................................................... 24 Peer-Mediated Interventions .................................................... .. 27 Self-Management Interventions ...................... .............. .... ......... 27 Social Stories ............ .... ....... .......... .... ... .......... ................. ... ... .... 28 Superhero Social Ski lls .............. .......... ...... ............... ...... ....... .... 29 Summary .................................................................................... 34 Statement of Purpose ................................................................. 35 Research Questions .................................................................... 36 2. METHODS ..... ....... ........ ....... .................. ............................. .. ............... . 38 Participants ............................................................................... .. 38 Inclusion Criteria for Children with ASD .................................. 39 Sening ........................................................................................ 44 Dependent .................................................................. 44 Observation System .. .... ............................................ ..... 44 Social Responsiveness Scale ............. ....... ...................... 45 Autism Social Ski lls ......................................... .. 46 Power Cards .................................. ....... .......................... 47 Scooter Hole Cards 47 Behavioral Intervention 52 Skills Implementation Data Analysis 56 3. RESULTS 58 Research Question 1 58 Research Question 2 63 Research Question 3 71 Research Question 4 75 Research Question 5 83 Research Question 6 90 Research Question 7 92 Research Question 8 95 Research Question 9 96 Research Question 10 101 Research Question 11 101 Reliability 102 DISCUSSION Practice A. SOCIAL SKILLS PLACEMENT CHECKLIST 117 B. SOCIAL OBSERVATION CODES 120 CHILD CONSUMER SATISFACTION SURVEY 123 D. SOCIAL VALIDITY CHECKLIST 125 E. PARENTAL PERMISSION DOCUMENT 127 vi ScoOicr and Black HoleCards ....................................... Rating Scale ............................ 48 Child Consumer Satisfaction Survey ............... .. ........ .... 48 Social Validity Checklist ..... " ........................................ 48 Intervention Fidelity ........................... ...... .. ............. ...... 49 Design ........ ..... ......... .................................................................. 49 Procedures ............................ ............ ................................. .. .... ... 51 Baseline .................... ........................... ............. ..... .... ... ....... .. ..... 52 Superhero Social Skills Program ................................. ............. . 52 Superhero Social Ski lls Group- Implemen tatioll .. .............. ........ 53 Didactic Group-Implementation ...... ........................ ...... ........ .... 54 Observation of Social Initiation and Responses .................... .... 54 Data Analysis ... .. .. ..... .. .............. .. .. .. .... .... .. .. .... ...... ............. ..... 56 3. R.ESU L TS ...... ....... ... ........................................ ... ... .... .. ... ....... .... .. .. ..... .. 58 Research Question I ...... ............... ... .................................... ...... 58 Research Question 2 .......................... .......... ..... ... .. .. .. ... ...... ....... 63 3 ........................ ............. ............... ... ..... .. .. .. 71 4 .... .. .................. .................. ................ ...... .. 75 5 .... .... ................ ................................... ....... 83 6 ...................... ...... ......... ................. ............ 90 7 .................................................................. 92 8 .................... .... ................................ ........ .. 95 9 .................................................... ......... ..... 96 10 ................................................ ....... ....... 101 I I ....................... ....................................... Reli ability .. .. ............................ ...................... ............ .......... .... 102 4. DiSCUSSiON ............................ ............... ......... .. ... ... ....... ... .. .. ............ 104 Limitations and Future Research ....... .... ... ....................... ........ 112 Implications for ............................................ ..... .. .. ..... 116 Appendices SKJLLS ..................................... .......................................... ................ C. CH ILD .................................. CHECKLI ST ..... ....... .. .. , ........................................ PERM ISS ION ............................................... v, F. DETAILED PROGRAM DESCRIPTION 132 G. SAMPLE LESSON 148 SKILLS REFERENCES vii PROGRAM DESCRiPTION ....................................... .... .. .... .................................................. .............................. .. .. H. AUTISM SOCIAL SKJLLS PROFILE ........................................... .......... 154 I. POWER CARD EXAMPLE (ENLARGED) ................................. ............. 159 REFER.ENCES ... .. ............................................................................................ 161 VII LIST Programs Youth with ASD 2. Participant Characteristics 43 Rated 5. Pearson Produce-Moment Correlation Coefficients 91 BIRS Item LIST OF TABLES 1. Social Skills Programs for Youth with ASD .. ................ ___ ........................... 25 Characteris tics ......... .. .. .. .. .. .. .. ...... ........ .............................. .... ... . 3. Mean Effect Sizes by Group ................................... ... ... .. .................. ....... .. .. 75 4. Average Pre- and Postintervention SRS Scores, as Raled by Teachers and Parents .... ............................ _.. ......................................... 76 S. .................................. .. 6. SIRS Means ................... .. .. .. ....................................................... . 97 1. Initiations, Participant 1 2. Participant 1 3. Total Social Engagement, Participant 1 62 Initiations, 5. Social Responses, Participant 2 65 Total Engagement, Initiations, Participant 8. Social Responses, Participant 3 69 Total Participant 3 10. Social Initiations, Participant 4 72 11. Social Responses, Participant 4 73 12. Total Social Engagement, Participant 4 74 1 14. Teacher SRS, Participant 1 78 15. Parent SRS, Participant 2 79 16. Teacher SRS, Participant 2 79 17. Parent SRS, Participant 3 80 18. Teacher SRS, Participant 3 81 LIST OF FIGURES I. Social Initiations. Participant I ............................... .............. ... .................... 60 1. Social Responses, Participant I ....................... ........................................... 61 I ................ ..................................... . 4. Social lnil imions. Participant 2 ...................... .............................................. 64 S. Responses. Part icipant ................................................................... 6. Tota l Social EngagemclH. Participant 2 ...................................... .................. 66 7. Social lnl li alions, Parti cipant 3 ... .................. ...................................... .. ........ 68 Socia l Pan icipant ............................... ..................................... 9. Tota l Social Engagement, Panicipant J ................... ................................... .. 70 lni ti atiolls, Partic ipant .......................... ......................................... I I. Soc ia l Part ici pant .. .............. .. .......... ........ .................... .......... Tota l Part icipan l ..................................................... 13. Parent SRS, Participant I .................................................. ............ .............. 77 14 . ......................................................................... .............................................. .............. ................ Partic ipa nt ........................ ....................................... ........ .. Part icipant .......... ................ .......................... ........................ 80 Part icipant ......................................................................... 19. Parent Participant 4 4 21. ASSP Participant 1 Participant Hole Hole Participant Hole Participant Cards 19. Pan~1lI SRS, Participant 4 ............ .... ........................................................... . 82 20. Teacher SRS, Participant 4 .... .......................... .... ............................... ...... 82 2 1. Average ASSP Score ............ ....... ................... ......................................... 84 22. ASSP, Participanl I .................................................................................... 86 23. ASSP, Participant 2 ............................................. ............................. .......... 87 24. ASSP, Participant 3 ................. ................... ........................................ .. ..... . 88 25. ASSP, Participant 4 ................. ....................... .. .......... ... .. .......... .. ... ......... .. 89 26. Scooter Cards and Black Hole Cards, Average .... .. ............................... ... .. 93 27. Scooter Cards and Black Hole Cards, P~mi c ipant 1 .......................... ... ...... 93 28. Scooter Cards and Black Hole Cards, Participant 2 ............. ...................... 94 29. Scooter Cards and Black Hole Cards, Participant 3 ... ............. .................. _ 94 30. Scooter Ca rds and Black Hole Cards, Participant 4 ................................... 95 x CHAPTER 1 ASD) American often & abnormal observed-aversion of gaze, and failure to respond to vocal cues. Before 48 months, children with ASD demonstrated abnormal social play and impaired imitation skills. Children also demonstrated a lack of awareness of others, impaired friendships, decreased imaginative play, and nonverbal communication skills (Stone, Hoffman, Lewis, & Ousley, 1994). INTRODUCTION Autism is a disorder marked by impairment of social interaction and communication. Individuals with Autism and other Autism Spectrum Disorders (ASD) struggle to create and maintain relationships, understand societal cues, take others' perspective, and demonstrate appropriate nonverbal communication skills (American Psychiatric Association, 2000). So great is the impairment in social interaction that it has been identified as the central feature of autism (Carter, Davis, Klin, & Volkmar, 2005). Due to their inability to create and maintain friendships, individuals with Autism often experience increased social rejection, anxiety, and depression (Bellini, Akullian, Hopf, 2007). Current research indicates that deficits in social development in children with ASD are apparent from the early months of life (Sigman, Dijamco, Gratier, & Rozga, 2004). From these early months, abnoDnal social interactions can be observed-such as 2 interaction with peers (Bellini et al., 2007). Abnormal social play, restricted interests, and impaired conversational skills limit opportunities to establish social relationships with peers. Children with ASD receive fewer social initiations, respond to fewer initiations, and when interacting, do so for shorter periods (McConnell, 2002). These limited opportunities to interact and establish relationships with peers can restrict the development of intelligence, language, and other related skills essential to normal childhood development (Garrison-Harrell, Kamps, & Kravits, 1997). Essentially, deficits in the ability to interact and communicate with others create a domino-like effect of negative consequences. Often, impairments in interaction and communication abilities early in life lead to social isolation and difficulty functioning in everyday life in adulthood. Due to poor social skills, adults with ASD often experience mental health problems, higher rates of unemployment, poor school achievement, and cognitive deficiencies (Howlin, Mawhood, & Rutter, 2000; Strain & Schwartz, 2001). Because of the negative outcomes associated with poor social skills, much research has been devoted to the creation and evaluation of social skills interventions for individuals with ASD-as well as other individuals with social skills deficits (e.g., Bellini Akullian, 2007; Cotugno, 2009). More and more, social skills development is becoming a primary concern of practitioners working with children with disabilities. In order to increase the probability that individuals engage in functional activity, it is essential that effective interventions be developed to increase the use of social skills. As children with ASD reach preschool, social skills deficits impede typical aI., AbnoTInal nonnal ASD-& essential that practices be employed based on empirical research. The implementation of evidence-based practices has been emphasized by educational, psychological, and independent associations. (Kratochwill & Shernoff, 2003) for the development and the integration of evidence-based practices into educational settings. First, there is a need for shared responsibility evidence-practices use manuals and other procedural guidelines to improve and operationalize specific Additionally, training of practitioners is fundamental in the integration of evidence-based practices to everyday practice. Both professional development opportunities and training in a scientist-practitioner model function to establish a strong relationship between research and practice. Practitioners are then able to provide additional outcome evaluations based on implementation under typical conditions. 3 Evidence-Based Practice While many therapies and interventions exist for increasing social skills, it is The National Association of School Psychologists (NASP) has supported definitions of evidence-based practice as given by Hoagwood and Johnson (2003) and Cournoyer and Powers (2002). Through stringent use of inclusion criteria and research designs, evidence-based practices are found to produce "predictable, beneficial, and effective results" in empirical studies. NASP has also set forth several guidelines evidence­based for developing interventions that are effective. Researchers and practitioners must work together to design and evaluate interventions. Second, it is important that evidence-based evidence-based practices. Kratochwill and Shernoff also emphasize the importance of guidelines that allow the practitioner to know how and when to use a specific intervention, in order to ensure effective implementation of the intervention. evelyday evidence-based practice (American Psychological Association, 2006; Chambless et al., 1996), similar to medical guidelines for evidence-based practice that encourage improving patient outcomes through informing practicing clinicians of current research. APA has defined evidence-based practice as "the integration of the best available research with clinical expertise in the context of patient characteristics, culture, and preferences" (American Psychological Association, 2006, p. 1). In addition to improving overall patient outcomes, APA defines the goals of evidence-based practice as making mental health care more cost effective, as well as increasing the accountability of practitioners. communication, research-based treatments. 4 The American Psychological Association has also provided guidelines for aI., infonning AP A AP A APA also defines the types of research designs that best contribute to the development of evidence-based practice. While traditional random controlled trials and meta-analyses provide stringent measures of treatment efficacy, clinical observations, qualitative research, single-subject designs, case studies, and process-outcome studies also contribute to the empirical evaluation of practices. APA also highlights the importance of not assuming interventions not yet evaluated through randomized trials to be ineffective, but stresses the importance of timely empirical evaluation. Overall, APA views evidence-based practice as designed to improve the mental health care of individuals through an environment of respect, conununication, and collaboration among those involved in the treatment process. Similarly to NASP, APA views collaboration between researchers, practitioners, and patients as essential to the advancement of 5 categories based on the amount of research demonstrating positive outcomes, using criteria presented by Chambless et al. (1996). For a practice to be identified as well-established, a minimum of two well-conducted group design studies must be conducted by different researchers. As per Division 12's recommendations, practices evaluated through single subject research may also be considered well-established. However, nine well-conducted single-subject studies must demonstrate substantial positive outcomes. Practices may also be determined to be probably efficacious. Two group design studies conducted by the same investigator and demonstrating positive outcomes may be considered probably efficacious. In the case that at least three single-subject studies have found a practice to be efficacious, the practice may be considered probably efficacious. The guidelines provided by Division 12 are useful in that they provide a concrete definition of evidence-based practice against which practices may be measured. The guidelines also provide researchers with clear directions for establishing evidence-based practices through empirical analysis. APA's criteria for selecting evidence-based practices in daily practice. First, studies determined to be well-established are often chosen due to statistical significance, not clinical significance. Clinical significance refers to a positive clinical outcome, instead of focusing solely on whether changes observed in the dependent variable were due to the introduction of the independent variable. O'Donohue and Ferguson suggest that while some therapies may be statistically significant, they fail to produce meaningful results. Division 12 of APA has attempted to sort empirically evaluated practices into ofresearch a1. well­established, efficacious. Using the definition of evidence-based practice provided by APA and Division 12, O'Donohue and Ferguson (2006) suggest that several weaknesses exist in using 6 al. efficacy, the extent to which a treatment is beneficial, while effectiveness, the successful exportation of an efficacious treatment to a community or private practice setting, is more valued in the clinical setting. In other words, the inclusion of a treatment on the EBP list often does not determine if the treatment is generalizable to other settings. Third, participants included in studies on the EBP list often exclude participants with comorbid conditions, such as intellectual disabilities and depression. While this is useful in research, participants in clinical settings often present with more than one condition, rendering the findings on many interventions irrelevant. Lastly, O'Donohue and Ferguson criticize the criteria for bias for group design studies and inferential statistics. Such bias often eliminates studies utilizing single-subject designs, as well as withdrawal and multiple-baseline designs. interventions 2001), and a lack of current research using ABA principles is apparent when examining evidence-based practice lists using the Chambless et al. (1996) criteria. Because O'Donohue and Ferguson suggest that ABA has drifted from its empirical roots, they suggest that new research focus on "new-school" behavioral principles, such as response deprivation, matching, and behavioral momentum. The authors also suggest that additional research be conducted on ABA treatments for individuals other than those with Second, the criteria proposed by APA (2006) and Chambless et a1. (1996) evaluate successful While studies evaluating applied behavior analysis (ABA) -based interventions often are excluded from evidence-based practice lists due to the weaknesses of EPB criteria proposed by O'Donohue and Ferguson (2006), several treatments have been determined to be well-established or probably efficacious. However, many of these treatments originate from "first generation behavior analysis" (O'Donohue & Ferguson, examining a1. Because 7 in ABA treatment evaluations. Odom, Brantlinger, Gertsen, Horner, Thompson, Harris 2005) have also contributed to the current emphasis on evidence-based practices. Due to changes in federal law (No Child Left Behind, 2001), teachers are mandated to use scientifically validated practices. While randomized clinical trials are often considered to be the highest standard of research due to their ability to control for threats to validity, they are rarely implemented for education research. While CEC has recommended that quality indicators be created for special education research, they have yet to be identified. Other research synthesis organizations have provided information on evidence-based practices in education, but practices are often limited to those evaluated in randomized clinical trials or rigorous quasi-experimental designs. The CEC Divisions of Early Childhood has adopted a system of literature reviews and focus groups to identify recommended practices, while the CEC Division of Learning Disabilities and the Division of Research rely on literature reviews by an expert in the field. The Department of Education, through the Institute of Education Sciences (IES), has created its own criteria for determining whether a practice is evidence-based (Institute of Education Sciences, 2003). Interventions found to be effective in well-designed and implemented randomized controlled trials in two or more typical school settings are determined to be backed by "strong" evidence. "Possible" evidence of effectiveness is comprised of randomized controlled trials that fall short of "strong" evidence and comparison-group studies with close matching of participant characteristics. Studies developmental disabilities, expanding research to populations that have not been included Special education groups, such as the Council for Exceptional Children (CEC, Homer, & Ranis detennined "probable" evidence of effectiveness. Overall, the IES places heavy emphasis on randomized controlled trials, stating that a properly designed and implemented randomized controlled trial is superior to other research designs. The IES also proposes that pre-, post- and comparison-group studies often produce erroneous results due to either a lack of or an unclear comparison group. Therefore, the IES proposes that the key to evaluation of treatments is through randomized controlled trials, in which the intervention and participants are clearly described. Detailed description of the control group and explanation of how the intervention is supposed to affect outcomes is also important in conducting research on an intervention's effectiveness. Studies contributing to strong evidence also need to demonstrate that there were no systematic differences between intervention groups. IES also highlights the importance of using outcome measures that are able to accurately assess the true outcomes. Studies with low-attrition rates and those that evaluate long-term outcomes are also important in contributing to strong evidence. Lastly, IES suggests that statistical tests be conducted in order to determine that the effect is not likely to have been produced by chance, with large sample sizes being better able to obtain such findings. determining Speech- Language-Hearing Association, 2004). Level I practices are those that have been shown to be effective through meta-analyses, including a minimum of one randomized 8 using pre- and postdesigns, comparison-group studies without close matching of characteristics, and meta-analyses including poorly completed studies do not comprise lES differences lES detennine Perhaps most similar to the guidelines set forth by APA Division 12, the American Speech-Language Hearing Association (ASHA) has adopted levels for detennining the amount of empirical support a practice has received (American Speech­Language- experimental design. Level II practices are those validated through controlled studies and quasi-experimental designs. Practices that have only received support through non-experimental designs, such as case studies, are considered Level III practices. Lastly, Level IV practices are comprised of practices supported only by expert committees and clinical experience of respected authorities. As with the standards supplied by Division 12, the ASHA standards provide clear guides for the development of evidence-based practices. Such guidelines can and should be used by researchers to develop and evaluate practices in order to achieve the stated goals of evidence-based practice-improved outcomes. In terms of evaluating evidence-based practices for specific disorders, the National Autism Center (NAC) has divided autism-specific treatments into three categories based on evidence of effectiveness (National Autism Center, 2009). For a treatment to be considered Established, the practice has to have produced beneficial effects for those involved in published research studies, and that access to an Established long-term harmful effects. The NAC notes that, while a treatment may be considered Established, it should not be assumed that one particular treatment would produce universal results for all individuals exposed to the treatment. Instead, it may be necessary to implement several different interventions to determine which is best suited for the individual. Treatments may also be classified as Unestablished, which suggests that a treatment has no research support or that current research does not allow for conclusions about efficacy. Unestablished Treatments may have been shown to produce no positive 9 non­experimental shldies, TIT supp0l1ed practice~improved Treatment is expected to produce Iong-tenn positive outcomes. Established Treatments are also defined as producing no hamlful 10 effective, ineffective, harmful. practices that lack substantial empirical evaluation only be considered after additional research has been conducted. Treatments and practices that have only limited empirical support have been classified by NAC as Emerging Treatments. NAC suggests that these treatments only be considered for use when Established Treatments have been deemed inappropriate by an informed decision-making team. The NAC has also added to statements from other organizations, factoring not only research findings for a particular intervention, but other critical factors as well. The NAC has identified professional judgment, values and preferences, and treatment provider capacity as playing a central role in the selection and use of evidence-based practices. These other critical factors are useful when making treatment decisions, especially when selecting a treatment that may qualify as Emerging due to limited empirical support. Analvses two 10% 10,000 subjects each treatment group to be able to detect an effect with 0% accuracy. Such large-scale studies are simply not possible for the majority of research. results, or simply lack enough research to determine if the treatment is effective, inefJective, or hannful. Unlike Division 12 of APA, NAC suggests that treatments and patiicular professional ill Meta-Analyses Often, single studies are not capable of detecting differences in effects of 1\'10 treatments (Egger & Smith, 1997). It is possible, due to low power, that a study may produce no significant effects when such an effect does exist. Large samples that provide adequate statistical power are often difficult to acquire. For example, Collins et al. (1992) suggest that a drug designed to reduce risk of disease by would require in 11 of analysis results from individual studies for the purpose of integrating the findings" (p. 3). Meta-analyses allow for several small, similar studies to be examined jointly, increasing the ability to detect statistical differences between treatment groups. Meta­analyses have additional advantages over traditional research (Davis & Crombie, 2001). First, meta-analyses allow for determining if, on average, an intervention produces significant benefits for participants. Second, meta-analyses have the ability to overcome bias of authors and reviewers, producing a more objective assessment of evidence than through traditional synthesis. Meta-analyses also offer more exact estimates of the benefits produced by a particular intervention. Due to higher statistical power, meta­analyses are capable of speeding the transition between research and practice often slowed due to false negative results. Lastly, meta-analytic research allows readers to determine the utility of decisions made during the analysis, and how said decisions have influenced final effect sizes produced. general in terms of magnitude and significance. for Glass (1976) defined meta-analysis as a "statistical analysis of a large collection often Additionally, meta-analyses have the added benefit of evaluating a priori hypotheses for patient subgroups, as well as generating future research questions (Egger & Smith, 1997). According to Blimling (1988), meta-analyses serve four general purposes: to describe existing studies of a treatment; to determine overall effectiveness of the treatment; to determine influences in the outcome of the treatment; and to quantify the outcome significance. Several standard steps exist for conducting a meta-analysis (Vegas, 2005). First, a question is defined and relevant studies are collected. Once the studies are selected for inclusion in the meta-analysis, a common metric is transformed from the outcomes of 12 each study. Last, an average is described from the outcome of the meta-analysis. Methodological soundness of studies is perhaps the most important criteria for meta­analyses, as poorly conducted studies will provide poor statistical results. Since the middle of the 20t h century, meta-analytical techniques have been employed to evaluate psychological studies. Meta-analytic studies have also evaluated the utility of various ASD treatments. A recent meta-analysis (Eldevik, Hastings, Hughes, Jahr, Eikeseth, & Cross, 2009) evaluated the utility of Early Intensive Behavioral Intervention (EIBI) for the treatment of ASD. Beginning with Lovaas (1987), behavioral interventions have been found to be effective in improving deficits associated with autism. EIBI has been found to be more effective than no intervention controls and other autism-specific special education interventions (Eikeseth, 2009), and Rogers and Vismara (2008) proposed that no other autism intervention presently qualifies as "well-established." Nine studies were included in the Eldevik et al. (2009) meta-analysis, with a total of 153 participants in EIBI groups. For participants in EIBI groups, an effect size of 1.103 was found for changes in IQ scores. Eldevik et al. were able to expand upon previous meta-analyses that determined EIBI produced smaller, but still substantial, effects (Reichow & Wolery, 2009) by inclusion of additional studies and through comparison between treatment groups-as opposed to within-subject comparison only. Additionally, Eldevik et al. examined EIBI effects of adaptive behaviors. This analysis provides additional information about the effect of EIBI on participants' daily life. The shldies Meta-Analyses of General Interventions for ASD 20th EIB! well­established." at. paIticipants a1. groups~as a1. moderate effect size. with autism, evaluating the effects of the intervention on cognitive, language, and adaptive skills. Spreckley and Boyd (2009) evaluated the utility of Applied Behavioral Intervention (ABI) in increasing cognitive, language, and adaptive behavior skills in preschool-age children with autism. Of the studies found in their literature review, only four studies were included in the meta-analysis. Of the four studies included, it was found that ABI produced an effect size of .38 for cognitive skills, or a moderate effect size. Similarly, ABI was found to produce an effect size of .37 for language skills. Lastly, adaptive skills were found to show the least improvement, with an effect size of .30. Due to the small number of studies included in their analysis, it is expected that a larger meta-analysis would be more representative of results produced through early behavioral interventions. Hourmanesh (2006) conducted a similar meta-analysis, but included a greater studies-thereby comprehensive interventions studies were included in the analysis. Similar to the Reichow and Wolery (2009) meta-analysis, an effect size of .64 was found in the 17 studies that evaluated cognitive skills. Analysis of 10 studies that included language skills as a dependent variable revealed that early comprehensive intervention produced an effect size of .61. The meta-analysis also found that, from the eight studies that evaluated adaptive skills as a dependent variable, an effect size of .68 was found. Overall, this meta-analysis found that EIBI produced an effect size of .660 for adaptive behaviors, a 13 An additional meta-analysis analyzed the effect of early intervention for children effect number of studies-thereby increasing the power of findings. A total of 16 early EIBI Hourmanesh behaviorally based intervention programs for children with autism. Nine articles were added to the 16 articles in the meta-analysis conducted by Hourmanesh. Overall, it was found that early behaviorally based interventions produced an effect size of .64 for cognitive skills and .28 for adaptive skills, with the adaptive skills effect size being substantially smaller than found by Hourmanesh. Early intervention was found to have a language effect size of .80, the only effect size found to be larger than those from the Hourmanesh meta-analysis. From this meta-analysis, it can be concluded that early behaviorally based intervention is effective in improving cognitive skills. Equally important is the finding that early behaviorally based interventions are effective in improving language skills, as language delays are common for individuals with ASD. sufficient communication skills. While early behavioral interventions have been found to have moderate to large effects on language skills, additional language interventions have been investigated via meta-analysis. Augmentative and alternative communication strategies have been explored in addition to EIBI for facilitating communication. Interventions using speech-generating devices, symbols and pictures-such as the Picture Exchange meta-analysis suggests that EIBT is effective in improving cognitive, language, and adaptive skills in children with ASD, with all effect sizes being in the moderate range. 14 Similar to Houmlanesh (2006), Backner (2009) evaluated the efficacy of early Houmlanesh. etTect to lan!,ruage Meta-Analysis of Language Interventions for ASD Because roughly half of individuals with autism will not develop sufficient language skills to meet their needs (Light, Roberts, DiMarco, & Greiner, 1998), meta­analyses have frequently evaluated the impact of interventions on language and conununication pictures-15 meta-analysis (Wendt, 2006). For this study, percentage of nonoverlapping data points (PND) was calculated for each intervention technique. Overall, it was found that highly iconic graphic symbols, such as those found in PECS, were associated with better outcomes-producing PND determined to be highly effective. Sign language communication strategies were also associated with improved receptive and expressive communication abilities, with a PND of 78%~or fairly effective. Graphic symbols and sign language were found to be equally effective in increasing functional requesting skills in individuals with ASD. Hopf s forms was also found that social skills programs show decreased efficacy when the social skills instruction is removed from the setting in which social interaction typically occurs. While traditional social skill instruction was shown to be relatively ineffective as a treatment for aiding in Communication System, and sign language and gestural acquisition were evaluated in a nonoverlap ping outcomes-producing conununication 78%--effecti ve. Meta-Analysis of Social Skills Interventions for ASD Bellini, Peters, Brenner, and Hopfs (2007) meta-analysis of social skill interventions investigated the effects that school-based social skills programs have on individuals with ASD. The meta-analysis included 147 students with ASD who received various fonTIs of social skills instruction in a school setting. The meta-analysis targeted acquisitions of skills such as group play, social initiations, and responding behaviors. The meta-analysis found that the social skills interventions included in the study produced questionable intervention effects due to the fact the skills did not generalize to other settings. Bellini, Peters, Brenner, and Hopf found that most programs produced only moderate maintenance effects and low generalization effects. It behaviors. successfully demonstrating the social skill to be acquired. In their meta-analysis, a total of 23 studies with a total of 73 evaluating the interventions were included. Overall, a PND of 80% was observed across all studies included, or a moderate intervention effect. PND was also calculated across three dependent variables: social-communication skills, functional skills, and behavioral functioning. Video-modeling and video self-modeling interventions produced the greatest effects in functional skills-such as self-help and purchasing skills. PNDs of 77% and 76% were observed in social-communication functioning and behavioral functioning, respectively. Unlike other social skill programs moderate maintenance effects (PND=83%) and generalization effects (PND=74%). Overall results suggest that video modeling and video self-modeling interventions are effective for improving social and functional skills in children with ASD. single-phases, as well as follow-up and skill generalization into additional settings. Overall, it 16 the acquisition of social skills for individuals with ASD, other forms of treatment, such as video modeling, have been successful in fostering acquisition and generalization of social Bellini and Akullian (2007) conducted a meta-analysis of video modeling and video self-modeling interventions for children with ASD. Interventions included in the study consisted of watching a video of another individual or the participant successfully functional self-modeling skills-such social-communication for children with ASD, video modeling and video self-modeling were found to have The utility of peers in the social skills training process for children with ASD has also been evaluated through meta-analysis. Zhang (2008) examined peer-mediated interventions for children with ASD under 8 years of age. Forty-five single-subject studies were included in the meta-analysis, which examined changes during intervention showed equally promising results for peer-mediated interventions, producing effect sizes of 1.49 and 1.51, respectively. It was also found that peer modeling was the most effective type of peer-mediated training, producing an effect size of 3.16. The majority of individuals included in the meta-analysis received some combination of peer modeling, peer initiation training, peer monitoring, peer networking, peer tutoring, and group-oriented contingencies, which was determined to have an effect size of 1.89. peer-mediated of 2.37 participants' increased for participants receiving peer-mediated interventions. These results suggest 17 was found that peer-mediated interventions for children with ASD were effective during intervention, producing an effect size of 1.46. Results for follow-up and generalization majority Peer-mediated interventions were also evaluated in Miller's (2006) meta-analysis of interventions for social interactions in children with ASD. A total of 30 studies were included in the meta-analysis, which evaluated collateral skill, child-specific, and peer­mediated interventions. Collateral skill interventions attempt to increase social interaction by training other skills that assist in social interactions, such as play and academic behaviors. Child-specific interventions are defined as interventions that directly teach and reinforce social behaviors of individuals with ASD. The findings of the meta-analysis revealed large effect sizes for all intervention types. Collateral skill and child-specific interventions produced effect sizes of2.37 and 2.19, respectively. The effect size for peer-mediated interventions was 3.27. As with results of Zhang's (2008) meta-analysis, peer-mediated interventions were found to produce large effects for individuals with ASD. Miller also found that, as pal1icipants' age increased, effect sizes decreased for collateral skill interventions. For peer-mediated interventions, effect size 11 articles were included in the meta-analysis, with 34 total participants. Studies included evaluated self-monitoring, self-assessment, self-observation, self-recoding, self-evaluation, self-instruction, and self-reinforcement. An overall PND of 81.9% was found for interventions using self-management strategies, which represents an effective treatment. Results of the study suggest that self-management is an effective treatment option for individuals with ASD. It was also found that interventions that incorporated monitoring by coparticipants produced greater effects. interventions-especially intensive-effective modeling, peer-mediated instruction, and self-management strategies have all been found to produce large improvements in social interaction skills of children with ASD. Therefore, it is expected that the inclusion of these components in a social skills program would produce greater results than currently existing social skills curricula. 18 that while collateral skill interventions are useful for young children, school age children with ASD may benefit more from peer-mediated interventions. The utility of self-management techniques for children with ASD has also been evaluated through meta-analysis (Lee, Simpson, & Shogren, 2007). A total of II alticles self­instruction, overalJ effects. Meta-analytic research has provided much information about what steps can be taken to improve treatment efficacy for individuals with ASD. Overall, behavioral interventions-especially those that are early and intensive-have been found to be useful in improving cognitive, adaptive, and language skills of individuals with ASD. Additionally, techniques providing children with additional communication skills have been effective in increasing receptive and expressive language skills. While social skill interventions have been found to produce little effect, other strategies have been effective in increasing social interactions of children with ASD. Through meta-analyses, video found (Lane, 1999; Mathur & Rutherford, 1991; Quinn, Kavale, Mathur, Rutherford, & Forness, 1999). In general, social skills instruction has been found to be of little use in producing behaviors that generalize across settings. In the Quinn et al. (1999) meta­analysis, social skills programs for children with Emotional and Behavioral Disorders were found to have an effect size of .199. According to Cohen (1988), an effect size of .199 falls in the small effect size category. In other words, social skills curricula were found to produce little to no effect. It is also important to note that, in addition to pro-social behaviors, problem behaviors showed little improvement after social skills training. An overall effect size for measures of disruptive behavior was found to be .131, suggesting social skills training is ineffective for changing behaviors more resistant to change-such as disruptive behaviors. In another meta-analysis of 49 studies evaluating social competence training (Beelmann, Pfingsten, & Losel, 1994) in 3- to 15-year-olds, social skills training was found to have a slightly higher effect size for children with Externalizing Disorders, .48. Again using Cohen's standards, this effect would be classified as a moderate effect. While moderate effects were found for improvement in social skills for children with Externalizing Disorders, no long-term improvements were observed. Follow-up studies included in this meta-analysis demonstrated declining effectiveness over time, with no permanent improvement. Cook, Gresham, Kern, Barreras, Thornton, and Crews (2008) also completed a meta-analysis on social skills training for students with emotional and behavioral disorders. Seventy-seven studies were included in their meta-analysis. Overall, it was 19 Current Social Skills Programs Much research has been conducted on the efficacy of social skills instruction \999). pro­social change-20 effect according to Cohen's metric. This effect size suggests that social skills training would produce noticeable differences in participant's social competency. It was also found that social skills instruction was most effective for preschool-age children and adolescents, with elementary-age children being less responsive to social skills training. While medium effect sizes were found through this meta-analysis, the authors note that the studies failed to evaluate social skill implementation in real-world settings. Therefore, it is suggested that greater emphasis be placed on the evaluation of social skills training in the school setting in order to better evaluate its effectiveness. & meta-individuals with emotional and behavioral disorders, and learning disabilities, Quinn et al. (1999) did find that social skills interventions produced larger effect sizes for found that social skills training produced an effect size of .32, or a medium effect effectiveness. Other meta-analyses have analyzed the utility of social skills programs for children with learning disabilities. Similar to individuals with ASD, individuals with learning disabilities have difficulties establishing relationships with peers (Swanson Malone, 1992). In their meta-analysis, Swanson and Malone found that children with learning disabilities were found to be less liked and more often rejected than their typical peers. However, existing social skills programs have demonstrated little utility in remediating social skill deficits in children with learning disabilities. In a meta-analysis of 53 studies (Forness & Kavale, 1996), social skills programs for children with learning disabilities were found to have an effect size of .211, or a small effect size. While the children rated themselves as improving substantially, peer and teacher observations of social skills did not. Although social skills interventions have not shown very promising results for a1. evaluated anxiety. Additional studies conducted after Quinn's meta-analysis have also evaluated the efficacy of social skills training as a treatment for anxiety. Spence, Donovan, and Brechman-Toussaint (2000) evaluated the efficacy of cognitive-behavioral therapy and social skills training as a treatment for school-based anxiety. The combination of therapies was found to be effective in decreasing school anxiety, with reductions in anxiety being maintained at a 12-month follow-up. Overall, social skills training appears to be beneficial to individuals with anxiety. pharmacotherapy pharmacotherapy, outperformed pharmacotherapy. Participants in the social skills plus placebo group were also found to have the lowest attrition rate of all groups in the study. 21 individuals with anxiety. Although the observed effect size of .422 suggests a moderate effect, it is important to note that only eight of the studies included in Quinn's meta­analysis shldies cognitive­behavioral Social skills training has also been investigated as a potential treatment for depression. Social skills have long been evaluated as treatment of depression due to the ability of positive social skills to enhance the ability of the individual to receive positive reinforcement, thereby decreasing depression (Lewinsohn & Clark, 1984). Bellack, Hersen, and Himmelhoch (1981) compared the efficacy of social skills training to phannacotherapy and psychotherapy as treatment for depression. The 72 participants in the study were randomly assigned to one of four groups: pharmacotherapy, social skills play phannacotherapy, social skills plus placebo, and psychotherapy plus placebo. Overall, it was found that patients in the social skills plus placebo group outperformed participants in all other treatment groups. The authors of the study state that, overall, social skills training may be a more useful intervention for treatment of depression than In Haley, 1991). Of 66 adolescents randomly assigned to either a social skills training group or a support group, those in the support group showed greater initial decreases in depressive symptoms. At a 9-month follow-up, these differences were no longer evident. Overall, the study found social skills training and the support group to be equally effective in reducing depressive symptoms over time. benefit improvements modification internentions Peer-mediated interventions training of typically developing peers to respond to and encourage social interactions of individuals in the social skills group. Child-specific interventions, perhaps the most 22 Tn addition to being compared to pharmacotherapy, social skills training has also been compared to support groups for treatment of depression (Fine, Forth, Gilbert, & supp0l1 This finding suggests that while social skills programs are not effective in increasing maintenance and generalization of skills taught in all populations, some populations do benefit from social skills interventions. Although externalizing disorders appear to me more resistant to change through social skills instruction, individuals with internalizing disorders such as anxiety and depression appear to receive more benefit from social skills training. Even though social skills training has produced improvements in skill demonstration in these populations, gains have been found to be only moderate. The majority of social skills programs currently available include several similar features. McConnell (2002) proposed that social skills programs could be divided into five categories. Environmental modt/ication strategies attempt to modify environmental characteristics in order to facilitate social interactions between program participants and peers. Collateral skills interventions aim to teach skills that facilitate social interactions instead of teaching actual social skills. Skills commonly taught in these interventions include play behaviors and social language. mediated entail the spec?fic the participant how to execute the target skill. Skills commonly taught in social skill programs of this variety include appropriate responding and appropriate question asking skills. Lastly, comprehensive interventions combine two or more of the previously discussed social skill intervention types. instruction 2001). al. instruction 10-12 skills training-continue to be developed and implemented without any empirical basis. 23 commonly used variety of social skills instruction, involves directly teaching the pal1icipant ill/erventions Several other commonalities exist in social skills programs. Skills deemed by the program designer to be most important for the target population are selected for inclusion in the program, and these skills are subsequently taught. Depending on the social skills program, skills may be demonstrated, explained, or modeled. Typically, social skills programs employ a didactic model of instmction in which an adult facilitator verbally instructs program participants. Although social skills programs are often delivered in the participant's natural environment (i.e., a classroom), instruction frequently occurs in pullout social skills groups (i.e., school counselor's office). Although frequently used in both clinical and school settings, pullout groups have been criticized as being "contrived, restricted and decontextualized" (Gresham, Sugai & Horner, Gresham et a1. also concluded that the average social skill program consists of 30 hours of instmction spread across weeks. Overall, social ski lls programs have struggled with a lack of research supporting their efficacy. While some disorders have been shown to benefit from inclusion in social skills training-such as anxiety and depression, others have not. While some social skills programs have adopted what has been learned through research, social skills programs 24 Due to Autism Spectrum Disorder's central impairment of social interaction, social skills programs for individuals with ASD have been widely developed (see Table 1). The majority of social skills programs currently available employ a traditional didactic method of instruction in which an adult clinician orally presents the steps to successfully demonstrate various social skills to the individuals with ASD. While each individual program varies with its design and content, most social skills programs adhere to this traditional model of instruction. While the different social skills programs currently available for individuals with ASD have all attempted to increase performance and generalization of social skills, meta-analysis results (Bellini, 2007) have shown that traditional social skills programs are relatively ineffective. produce little to no effects, especially for individuals with ASD, several interventions have demonstrated utility in increasing social interaction of individuals with ASD. Video modeling and video self-modeling have shown promise as a social skills video of someone demonstrating a discrete behavior or skill to be acquired without error, whereas video self-modeling is the process of watching oneself demonstrate a behavior or skill without error. Repeated viewings of videos of others and video self-modeled skills have proven effective in increasing social skills and having these skills generalize to other environments. Bellini and Akullian (2007) found that video self-modeling was Social Skills Programs for Autism Spectrum Disorders performance ineffective. While meta-analyses have suggested that most social skills programs generally Video Modeling Interventions intervention for individuals with ASD. Video modeling is the process of watching a 25 1 Social Training Social- Communication Problems (Baker Myles, 2003) Skills Skills Autism Navigating Social with Syndrome, Functioning Autism, Related McKinnon & Teaching Autism Difficulties Bellini, High-Autism Syndrome Dunn, for 2S Table I Socia l Skills Programs for Youth with ASD Social Skills Train ing for Children and Adolescents with Asperger Syndrome and Social ­Communication & Social Ski lls Solutions: A Hands-on Manual for Teaching Social Ski lls to Children with Auti sm Navig<l ting the Soc ial World: A Curriculum for Individuals wi th Asperger's Syndrome. High FUllctioning Auti sm, and Re laied Disorders (McK innon Krempa, 2005) Building Social Relationships: A Systematic Approach to Teach ing Social Interaction Skills to Children and Adolescents with Auti sm Spectrum Disorders and Other Social Difticulties (Bell ini, 2006) S.O.S. Social Skills in Our Schools: A Social Skills Program for Children with Pervasive Developmental Disorders, Including I-ligh-Functioning Auti sm and Asperger Syndrome and Their Typical Peers (Dunn , 2005) Think Social: A Social Thinking Curriculum lo r School-aged Students (Winner, 2006) ASD. More importantly, treatment effects were maintained over time. In a separate study, Nikopoulos (2007) found that the use of 35-second video models of typically developing peers produced significant gains in time spent in social play using toys, resulting in increased interaction time with researchers and peers by children with autism. Nikopoulous also found that video models resulted in generalization of play skills to new toys, settings, and teachers-generalization that was maintained after 1- and 3- month follow-up probes. Video self-modeling has also been found to increase spontaneous requesting of children with ASD (Wert & Neisworth, 2003). Video self-modeling has rapidly become one of the most promising treatments for individuals with ASD. Although al. (Pierce & Schreibman, 1994). In addition to the Sherer et al. (2001) study, Bellini and Akullian (2007) conducted a meta-analysis of 23 studies, and this showed a similar result. to high effect sizes in functional skills, social-communication functioning, and behavioral 26 effective in increasing the rate of social engagement of 2 preschool-aged children with Nikopolliolls teachers-generalization AI though the efficacy of video self-modeling has been demonstrated in a number of studies, other studies comparing video self-modeling to video modeling have found no differences in treatment outcome (Sherer, Pierce, Paredes, Kisacky, Ingersoll, & Shreibman, 2001). Interestingly, Sherer et aL found that individuals who showed improvement using video modeling treatments, regardless of the format of watching others or oneself, also demonstrated higher visual learning skills than those who did not show improvements. This finding is especially important for the treatment of individuals with ASD since research has suggested that individuals with ASD are visual learners aL Bellini and Akullian (2007) found little difference in the efficacy of video modeling and video self-modeling, with both forms of intervention producing moderate 27 self-modeling which included 30 studies on social skills interventions for children with ASD, found that very young children with ASD benefit most from programs emphasizing the acquisition of collateral skills, such as language. As children with ASD grow older, peer-mediated interventions become the most effective in promoting demonstration of social interaction. This is especially important because relatively few social skills programs currently exist that incorporate peers into the teaching process. Interventions behavior, managing behavior, and empowering individuals to control their own behavior (Lee et al., 2007). Meta-analysis of single-subject research on self-management revealed that self-management interventions are effective in increasing the rate of appropriate behavior in individuals with ASD. functioning. Bellini and Akullian also found video modeling and video self-modeling interventions to have moderate to high maintenance and generalization effects. Peer-Mediated Interventions In addition to video-modeling, peer-mediated interventions have been shown to be effective in social skills training for children with ASD (Miller, 2006). Miller's meta­analysis, peer­mediated Self-Management Interventions Self-management, the process of monitoring and recording ones' own behavior, has also been examined as a technique to increase the use of social skills in individuals with ASD. Self-monitoring attempts to increase social skills by increasing awareness of aI., se If-with ASD. Social stories allow individuals with ASD to easily access social information that provides clear, order-based definitions of the desired behaviors. Social stories provide participants with information in a visual format, playing on strengths of individuals with ASD. Social stories targeting social communication skills, from basic skills such as maintaining eye contact to more the more complex perspective taking, have shown promise in increasing the display of said skills. Quirmbach, Lincoln, Feinberg- Gizzo, Ingersoll and Andrews (2009) evaluated the use of social stories for increasing play skills, finding social stories to produce significant changes in play behaviors as compared to a control group. Quirmbach et al. also found that basic verbal comprehension skills (ex., WISC VCI above 68) are prerequisite for producing behavioral change due to social stories interventions. Most importantly, social stories hold promise as an efficient, cost-effective intervention that can be designed to encourage generalization and maintenance of learned social skills. However, it has been found that, when utilized as the only intervention component, social stories are not effective in producing long-term changes in social skill demonstration (Sansoti, Powell-Smith, & Kinkaid, 2004; Crozier & Tincani, 2007). 2001, 2002), four times more than a sample of typically developing peers, with approximately 7 5% of 28 Social Stories In addition to interventions previously mentioned, social stories have been evaluated as a technique for increasing social communication difficulties of individuals information clear , defini tions palticipants Feinberg­Gizzo, aL Deficits in social skills have also been associated with special problems for children with ASD. Bullying has been found to be a consistent problem for children with ASD. Little (2001,2002), found that adolescents with Asperger's Disorder are bullied 75% impaired social interaction and communication abilities. Van Roekel, Scholte, and Didden (2009) found that students with ASD placed in special education or in ASD-specific schools experienced a lower rate of bullying-between 7 and 30%. Analysis of bullying in children with ASD revealed that students with ASD could accurately identify bullying situations, but that misinterpretation of nonbullying situations as bullying increased as experience with bullying increased. It was also found that approximately 26% of adolescents with ASD were classified as bullies. Clearly, bullying is a problem that needs to be addressed in social skills programs for children with ASD. of recent peer-mediated environment for young children is supported by current research findings. Using the evidence-based practice (EBP) criteria proposed by Reichow and Volkmar, both social skills groups and video modeling can be classified as EBP. It is these findings, along with meta-analyses conducted by Miller (2006), Sherer et al. (2001), and Bellini et al. (2007) that provide much of the basis for the development of the current study examining the Superhero Social Skills program for children with ASD. Disorder-29 individuals with ASD being victimized. This high rate of bullying can be attributed to ASD­specific bullying-between identify approximately In a review ofrecent social skills research for individuals with ASD, Reichow and Volkmar (2009) found empirical support for the use of video-modeling and peer­mediated interventions. It was also found that social skills intervention in the naturalistic supp0l1ed 200 I), a1. Superhero Social Skills Superhero Social Skills is intended for elementary-aged children with high functioning Autism, Asperger's Syndrome, or Pervasive Developmental Disorder-Not Otherwise specified. It has been designed to address the shortcomings of other social 30 The program employs a number of evidence-based practices in order to increase social skill acquisition, namely video modeling and self-modeling, inclusion of nondisabled peers, and the use of self-management strategies. Superhero Social Skills is comprised of 18 social skills (classified as foundational, intermediate, or advanced skills) presented via DVD by animated superheroes. The superheroes address key deficits in social skills functioning in children with ASD. The superhero characters and comic book style of the program make Superhero Social Skills of high interest to children while incorporating evidence-based practices that have been shown to aid in the acquisition, maintenance, and generalization of social skills. A more complete description and example of Superhero Social Skills can be found in Appendix F. of the participant; and generalization, which demonstrates the researchers attempt to demonstrate successful spread of effects across settings, time, responses, and stimuli. al., skills programs for children with ASD, including lack of maintenance and generalization. Superhero Social Skills has been designed to employ components of interventions that meet the criteria of evidence-based practice. The evidence-based practice standards provided by the NAC (2009) categorize several of the components of Superhero Social Skills as Established Treatments. The analysis employed by NAC for determining treatment evidence included the following: research design, which suggested the degree of experimental control; measurement of the dependent variable, or accuracy and reliability of the data; measurement of the independent variable, which expresses the extent of treatment fidelity; participant ascertainment, or correct inclusion and eligibility As previously discussed, modeling and video modeling have been found to be effective in increasing social interaction in individuals with ASD (Bellini et ai., 2007). successfully increase communication, play, personal responsibility, and interpersonal skills. Participation in a modeling intervention was also found to increase higher cognitive functions. Modeling was also found to produce a decrease in problem behaviors and improve sensory and emotional regulation. Effects were found for individuals with ASD between the ages of 3 and 18. Thirty-three of 21 self-management strategies. Overall problem behaviors were decreased for children with autistic disorder ages 3 through 18. Also found to be an Established Treatment by the NAC, Social Stories are an integral component of the Superhero Social Skills program. Through the Additionally, the NAC has classified modeling as an Established Treatment. In their review of 50 live modeling and video modeling studies, the intervention was found to 31 The inclusion of peers in the social skills intervention has also been found to be useful in increasing social engagement in children with Autism (Miller, 2006). Similarly, Peer Training Package interventions were classified as an Established Treatment. Thirty­three studies included in the NAC review found that peer-mediated interventions were effective in increasing communication, interpersonal, and play skills in children with autistic disorder and PDD-NOS. Peer-mediated interventions were also found to produce decreases in restricted, repetitive, nonfunctional patterns of behavior, interests, or activity. Peer-mediated intervention effects were found in children ages 3 to 14. Self-management strategies have been found to be effective in increasing appropriate behavior for children with ASD in meta-analyses (Lee et al., 2007) and in a review of2! studies by the NAC. In addition to increasing appropriate behaviors as found by Lee et al., studies included in the NEC National Standards Report found that self-regulation skills were increased through implementation of self-management self-regulation skills through story-based interventions. learning readiness, and play skills for children ages 0 to 9. These effects were observed in children with both autistic disorder and PDD-NOS. Adams Martella, formats skills use of cueing and unison responding are employed to encourage rapid acquisition of new skills. Modeling, guided practice, and independent practice are also central instructional 32 review of 21 studies on story-based intervention packages, children with autistic disorder and Asperger's Syndrome ages 6-14 were found to increase interpersonal and self­regulation Superhero Social Skills also emphasizes the importance of teaching social skills in the child's natural setting, such as a classroom. The NAC has classified Naturalistic Teaching Strategies and an Established Treatment. In 32 studies, teaching skills in the natural environment was associated with increased communication, interpersonal, POD-Superhero Social Skills also incorporates techniques of Direct Instruction (Adams & Carnine, 2003, Marchand-Martella, Slocum, & Manella, 2004). Direct Instruction is designed to increase the rate of learning while promoting generalization of learned skills. Direct Instruction aims to achieve this by the following: appropriate sequencing of skills, or teaching of prerequisites before the strategy is taught; use of clear instructional formats and expectations; and teaching skilIs across multiple lessons to promote understanding and use of the skill. Direct Instruction also encourages that students be heavily engaged in instructional activities in which they experience high levels of success, and that assessment of learned skills should be continuous in order to provide feedback about the instructional process. The types of teacher-student interactions are also a focus of the Direct Instruction method. Increasing motivation through high levels of success with the skill, teaching to mastery, error correction, increasing opportunities to respond, and the 33 has compared the Direct Instruction model to other instructional strategies for students in special education, with Direct Instruction producing better outcomes (Adams & Engelmann, 1996; Forness, Kavale, Blum, & Lloyd, 1997; White, 1988). Overall, Direct Instruction has been found to produce effect sizes between .84 and .90, or large effect sizes. Superhero Social Skills also establishes clear response expectations of the participants and reinforces learned skills across multiple lessons. Participants in the Superhero Social Skills program are provided with ample opportunities to use the learned skill in each and skills retaught when necessary, cues are provided, and students are engaged in instaictional Overall, the Superhero Social Skills program includes many components that meet criteria as evidence-based practice. Additionally, the NAC has determined that these components qualify as Established Treatments for ASD, having proven their utility in many studies. Meta-analyses conducted independent of the NAC National Standards Report have confirmed the usefulness of peer-mediated interventions, video modeling, and self-management in improving skill deficits typically associated with ASD. procedures in the acquirement of new skills in the Direct Instruction model. Research instmctional and. 90, effect Skill sequencing, curriculum that promotes generalization, and continual assessment of learned skills have been incorporated into Superhero Social Skills. SkHls lesson through engagement with peers and the program facilitator, errors are corrected activities where they may experience a high level of success. As with Direct Instruction, Superhero Social Skills employs modeling, guided practice, and independent practice as primary instmctional procedures. detemlined con finned 34 learning learned packages were effective in increasing interpersonal skills of children with ASD, but no substantial improvements were found in communication skills, personal responsibility skills, play skills, and self-regulation skills. Likewise, no decreases in undesired behaviors were observed after implementation of a social skills package. Superhero Social Skills differs from other existing social skills program in its utilization of a variety of evidence-based practices. Components of Superhero Social Skills have been repeatedly found to increase play behaviors, improve self-regulation and communication skills, increase personal responsibility, and improve higher cognitive functions in addition to increasing interpersonal skills. The components and instructional strategies of Superhero Social Skills have been thoroughly researched and meet the criteria for evidence-based practice, as well as Established Treatments as per NAC guidelines. While existing social skills programs have shown some promise and are currently categorized as Emerging, Superhero Social Skills incorporates a number of empirically-based practices designed to ameliorate social skill deficits not currently addressed by other social skills programs. summary, Superhero Social Skills also employs instructional procedures found to increase leaming rate and generalization of leamed skills. In the National Standards Report, the NAC categorized Social Skills Packages as an Emerging Treatment. Their review of 16 studies found that existing social skills communication empirically­based Summary In summaty, a wide variety of social skills interventions for children with ASD currently exist, though very few have been shown to be successful in fostering the 35 instruction for individuals with ASD. Superhero Social Skills has been designed to overcome this problem through the use of evidence-based practices, such as video modeling, self-modeling, and incorporation of nondisabled peers in treatment groups. The present study aimed to evaluate the relative efficacy of Superhero Social Skills video-based approach compared to a traditional didactic approach, the most commonly used form of social skills instruction, by comparing effect sizes and percentage of non-overlapping data points obtained from two different groups receiving the two types of social skills instruction. instruction format, instruction instruction instruction instruction. generalization of social skills. Because of the negative consequences associated with poor social skills, it is important that additional research be conducted on social skill instmction lise Skills instmction, non­overlapping Statement of Purpose The purpose of the present study was is to examine the effects of social skills instmction for children with autism presented through video fomlat, compared to traditional didactic instruction. The efficacy of video-based instmction compared to didactic instmction will be evaluated with preschool and kindergarten students in an Autism classroom. It is hypothesized that video-based instmction using animated characters will increase interest in the social skills lessons while reducing anxiety experienced during sessions, leading to increased social skills demonstration and generalization in comparison to didactic instmction. 36 1. What is the effect size and PND of Superhero Social Skills for Children with Autism when presented in its video-based format, as measured during a social interaction observation during analog free time? a. An observation system designed by Bellini (2007) was used to evaluate the effect of adult facilitator social skills instruction on social engagement of children with ASD. 2. What is the effect size and PND of Superhero Social Skills for Children with Autism when presented in a traditional didactic format, as measured during a social interaction by Bellini (2007) was used to evaluate the effect of Superhero Social Skills on social engagement of children with ASD. 3. Using Cohen's (1988) effect size construct, does traditional didactic instruction or a. Results obtained using the Bellini observation system (2007) for the Superhero Social Skills group and the didactic instruction group were compared. 4. Comparing preintervention and postintervention scores, what are the changes in participant's participant's Autism Social Skills Profile scores, as rated by parents? cards and their percentage of social interaction during analog free time? Research Questions Au/ism Skillsfor Childrell Autism observation during analog free time? a. An observation system designed SkW,\' video-based instruction produce a greater effect size? palticipant's Social Responsiveness Scale scores? a. Social Responsiveness Scale 5. Comparing preintervention and postintervention scores, what are the changes in parents'? 6. What is the correlation between number of circles filled in on each participant's power b. Bellini observation system following rules, Scooter Cards and Black Hole cards distributed during each social skills lesson? Cards teachers, BIRS? a. Behavior Intervention Rating Scale (BIRS) validity to measured BIRS? a. Behavior Intervention Rating Scale (BIRS) satisfaction Superhero Skills for Children with Autism, Satisfaction Survey 11. validity to measured from Bellini's social validity scale (unpublished)? Bellini's validity a. Power Cards Belli ni observa lion 7. What is the improvement in followi ng group rul es , as measured by the number of ca rds di stributed a. Scooter Cards b. Black Hole cards 8. What is the social validity according to leachers, as measured by the S IRS? <l. Rat ing SIRS) 9. What is the social va lidity according [0 parents, as meas ured by the SIRS? Interve ntion Rat ing S IRS) 10. What is the participant sati sfaction of SlIperhero Social ,S'kills.!ur Childrtf,1l wit/! Allfism, as measured by the Child Consumer Sati sfac tion Survey? a. Child Consumer Satisfaction Survey 37 I J. What is the social validi ty according 10 parents, as meas ured by a scale adapted fi'om Bellini 's (a. Scale adapted from Be llini 's social va li di ly scale CHAPTER 2 participants, Mental Health Research Committee. The Valley Mental Health Research Committee reviewed the proposed study and the Superhero Social Skills program to determine whether the purpose of the proposed study was congruent with the services Valley Mental Health provides. Consent from Institutional Review Board (IRB) at the University of Utah was also obtained prior to the start of this study. IRB evaluated the design of the study, and ensured that participant's rights are protected during the study. Autism-Carmen participants were recruited for inclusion in this study-4 children with ASD and 4 normally developing peers. Recruitment consisted of the following: first, teacher recommendation of appropriate participants; second, notes sent home to parents METHODS Prior to recruitment of pal1icipants, consent to conduct the current study at the Carmen B. Pingree Center for Children with Autism was obtained from the Valley stal1 Participants Participants in this study were recruited from a center specializing in the treatment of children with Autism-the Cannen B. Pingree Center for Children with Autism. Such specialized schools offer programming for children with and without ASD. Eight palticipants study-describing die study and inviting participation; third, random selection through number assignment of those consenting to participate and lottery drawing to determine participant placement. To qualify for inclusion in the study, participants met the following criteria. Inclusion Criteria for Children with ASD In order to be included as a participant, children had to meet the following criteria: 1. Have a current diagnosis of Autism, Asperger's Syndrome, or Pervasive 2. Obtain scores on the Autism Diagnostic Observation Schedule (ADOS) that off IQ administrator. language Be currently enrolled in the Pingree Center's preschool or kindergarten program. for this study to screen participants was administered to teachers and parents to aid in the selection of participants. Placement checklist criteria consisted of the following: 39 the pm1icipants following Developmental Disorder-Not Otherwise Specified from a qualified clinician (i.e., MD, psychologist). meet or exceed the cut-off for Autism Spectrum Disorders 3. Obtain a score on the Social Responsiveness Scale (SRS) that meets or exceeds the cut-off for Autism Spectrum Disorders. 4. Obtain an 1Q score of 70 or higher on a standardized intelligence test, administered within the past 3 years by a qualified administrator. 5. Possess and demonstrate use of sufficient expressive and receptive lan,6'1lage so as to be able to participate in the social skills group. 6. In addition to meeting these criteria, a placement checklist (Appendix A) designed following: 40 interactions 8. Memory abilities staff at the Pingree Center were asked to identify students with is conducive to small group social skills instruction. After a list of students with and without ASD who meet inclusion criteria for inclusion was compiled, a letter was sent to each student's parents(s)/guardian(s) in order to obtain consent to participate in the study. Because only 4 children with Autism and 4 1. Developmental level of child 2. Language abilities of child 3. Parent description of cognitive abilities 4. Current diagnosis of ASD 5. Unusual behaviors that interfere with social interactions 6. Motivational level 7. Attentional abilities 9. Anxiety 10. Other factors deemed important by parent. Teachers and identi fy ASD who meet the inclusion criteria for children with ASD. Because the Pingree Center houses a developmentally appropriate preschool for integration purposes, not all children at the Pingree Center have ASD. Four nonautistic same-aged peers who possess developmentally appropriate social skills were recruited from the Pingree Center to participate in the lessons. Typically developing peers were recruited from an on-site preschool that allows children with ASD to have access to nondisabled peers. Teachers at the school were asked to nominate children in their classes that regularly demonstrate developmentally appropriate social skills without teacher prompting and whose behavior 41 child may or may not be a participant. The 4 students with autism and their peers included in the study were randomly selected from those that obtain consent to participate in the study. Of all potential participants who returned consent forms, each was assigned a number. After 2 weeks of allowing return of consent forms, a lottery was conducted in which random numbers were selected from a container. Participants whose assigned number matched the drawn number were selected for inclusion in the study. The 4 participants with ASD were then randomly assigned to one of the two treatment conditions in a similar fashion: video-based instruction and didactic instruction. The typical peers were also randomly assigned to one of the two groups in the same manner. The following are more detailed descriptions of characteristics of participants with ASD. Participant 7, 1 1 12. Participant 1 was capable of using phrase speech, but was not considered verbally fluent. Previously, Participant 1 had obtained a Full Scale IQ score 101 on the Wechsler Preschool and Primary Scale of Intelligence-Third Edition (WPPSI-III). Participant l 's mother described her as having an abnormal adherence to routine, often becoming upset when placed in new situations or frustrating activities. Participant 1 was also described as having difficulty calming herself when upset. developmentally normal peers were included in the study, the letter explained that their mayor J, a 4.9-year-old Caucasian female, was assigned to the video-based social skills group. Participant I had previously received a clinical diagnosis of autism from the Clinical Medical Director of the Pingree Center. Prior to intervention, an ADOS was administered to confirm a diagnosis of an ASD. Participant I obtained a Communication + Social Interaction Total score of 14, exceeding the autism cut-off of Wechsler lntelligence-1 abnOlmal 42 Participant video-based social skills group. Participant 2 had previously received a clinical diagnosis of PDD-NOS from the Clinical Medical Director of the Pingree Center. An ADOS was administered to confirm a diagnosis of ASD. Participant 2 obtained a Communication Social Interaction Total score of 11, exceeding the autism spectrum cut-off of 8. Participant 2 exhibited increased difficulty with the Reciprocal Social Interaction section of the ADOS, receiving a score of 8, exceeding the autism cut-off of 6. Prior to recruitment to participate in the study, a Stanford-Binet 5 was administered. Participant 2 received a Nonverbal IQ score of 128, a Verbal IQ score of 111, and a Full Scale IQ score of 120. Participant 2's mother described him as rigid and inflexible, often engaging in behavior inappropriate for the situation. During the study, Participant 2 continued his normal use of Prozac-1.25 mg/day. Participant of PDD-community physician originally diagnosed Participant 3 with autism, the diagnosis was also confirmed by the Clinical Medical Director of the Pingree Center. Confirmation of Communication Social Interaction Total score of 10, exceeding the autism spectrum cut-off of 8. A WPPSI-III was also administered to determine inclusion criteria. Participant 3 received a Verbal IQ score of 110, a Performance IQ score of 105, and a Full Scale IQ score of 109. Participant 3's mother described several behaviors that Participallt 2, a 4.8-year-old Caucasian male, was also assigned to the video­based + off of] 11, ofProzac-l.25 Participallt 3, a 3.8-year-old Caucasian male, was assigned to the didactic social skills instruction group. Prior to inclusion in the study, Participant 3 had received a clinical and educational diagnosis of Autistic Disorder. The mother of Participant 3 reported that he first received a diagnosis ofPDD-NOS at age 15 months. While a the diagnosis was obtained through administration of an ADOS. Participant 3 received a + Ill 3 'thaL 43 3'strings around things). Since Participant 3' s first diagnosis, he has continually received interventions to remediate his difficulties. received a clinical and educational classification of Autistic Disorder. While the Clinical Medical Director of the Pingree Center assigned the clinical diagnosis, the origin of the educational diagnosis is unknown. Confirmation of this diagnosis was obtained through administration of an ADOS. Participant 4 received a Communication Social Interaction Total of 13, exceeding the Autism cut-off. Participant 4 had previously been administered a Stanford-Binet 5, on which he obtained a Full Scale IQ score of 95. Participant 4 was described as having particular difficulty in frustrating activities. Participant inflexible focusing. Information on all participants is also presented in Table 2. Preintervention SRS 10 109 103 95 82 88.7 interfered with Participant 3 's social interaction with others: rigidity in play behaviors (i.e., "play my way"), and abnormal play behaviors (i.e., lining objects up and wrapping difficulties. Participant 4, a 5.2-year-old African American male, was also assigned to the didactic social skills instruction group. Prior to inclusion in the study, Participant 4 had + PaI1icipant 4 was also described as being int1exible and having difficulty focusing. Table 2 Participant Characteristics Participant 1 Participant 2 Participant 3 Participant 4 Total Child's Age 4.9 4.8 3.8 5.2 4.6 ADOS Total 14 1 1 13 12 FSIQ Score 101 120 106 GADS Total 78 92 Total Parent 75 69 90 70 76 Teacher 55 56 47 50 52 autism-The with Autism, but also educates normally developing children as well. The Pingree Center bases treatment of children with ASD on an Applied Behavior Analysis model. Initial assessment, social skills instruction, observation of skills, and follow-up assessment took place at the Pingree Center. Lessons were taught in a kiva type conference room located at the center. The conference room was equipped with a projector, with which videos were shown. Participants sat on carpet squares on the floor of the room. The room was sufficiently large as to play a social skills game as a group after each lesson. Free time observations occurred in the same room as social skills instruction. During free play, carpet squares were removed from the play area. For coding and reliability, free time sessions were videoed. Six toys were available for participant use during free time sessions. Initially, participants had access to toy cars (Mattel), LEGOS (LEGO), Ants in the Pants Spongebob Edition (Hasbro), Transformers (Hasbro), and Don't Break the Ice (Hasbro). After the first baseline session, Don't Break the Ice was replaced with Ring Around the Nosy (Pressman Toy) due to participant difficulty in assembling the game. by participants with ASD. Bellini's Social Observation codes allow for coding of social 44 Setting Children were recruited for inclusion in the current study through a school specializing in the treatment of children with autism-The Pingree Center for Children playa Matte!), lee Dependent Measures Observation System Bellini's Social Observation System (2007, see Appendix B) was used during baseline and treatment phases to evaluate the amount of social interaction demonstrated an object; and requesting interaction or participation. In order for a behavior to be coded as a social initiation, the behavior must be the beginning of a new behavioral sequence. This may be accomplished through changing play activities, partners, or discontinuing a behavior for at least 5 seconds. Social responses are defined as responding to a request for assistance or information, joining an activity upon request, accepting an object when offered, and appropriately continuing an interaction. 10-gained 45 initiations, social responses, total social engagement, and play behaviors. Social initiations are defined as the following: requesting assistance or information; joining in a play activity or interaction; giving a greeting or compliment; giving, sharing, or showing The observation system uses a 1 O-second time sampling method of observing behaviors, in which the observer watches the behavior for 5 seconds, and then records the behavior during the next 5 seconds. During each interval, only one behavior was coded: a social initiation, a social response, or no behavior. The observational data gained provide the needed information to calculate effect sizes for the video-based and didactic instruction groups. At the end of each session, observations of a 10-minute analog free time period were conducted. Free time periods were videotaped to facilitate coding of social engagement. Analog free time periods were conducted at the end of each lesson, with social engagement codes being calculated for each participant with ASD during all free time periods. Social Responsiveness Scale The Social Responsiveness Scale (SRS; Constantino, 2002) is a 65-item questionnaire that is designed to assess several characteristics of autism spectrum 1-true to almost always true). The SRS was administered to parents and teachers of participants with ASD prior to intervention implementation. Upon the completion of the social skills program, the Social SRS was again completed for the children with ASD by parents and teachers. Pre- and postintervention scores will be compared to determine any changes in social responsiveness. designed to assess social engagement abilities of children with ASD (Bellini & Hopf, 2007). Items may be answered 1-never/almost never, 2-sometimes/occasionally, 3- often/typically, or 4-very often/always. The ASSP was designed to aid in the planning of social skills interventions, as well as assess the effect of developed interventions. The ASSP yields a total score, as well as three subscale scores-Social Reciprocity, Social Participation/Avoidance, and Detrimental Social Behaviors. On the ASSP, higher scores suggest less impairment in social functioning. The ASSP was completed by parents of participants with ASD prior to beginning intervention, as well as at the conclusion of intervention. 46 disorders. The SRS provides an assessment of social impairments, social awareness, social information processing, ability for reciprocal social communication, anxiety and avoidance of social situations, and autistic traits. Items on the SRS are rated from \-4 (not pat1icipants Autism Social Skills Profile The Autism Social Skills Profile (Appendix H) is a 49-item likert-type scale I-scores-Social Participation! A voidance, successful demonstration of social skills taught in the program. Power Cards are distributed during the first social skills lesson of a new skill. The Power Cards have a picture of one of the superheroes, along with the steps to the current skill being learned. To track demonstrations of the current skill, the Power Cards are bordered with circles that are filled in when the skill is demonstrated correctly. Each week, the number of circles each student has filled in was recorded. At the end of the program, the number of circles completed each week was correlated with frequency of social engagement as measured by Bellini's Social Observation System during analog free time sessions. rule-following following behavior for individuals in the group. 47 Power Cards The Superhero Social Skills program uses Power Cards (Appendix I) to track the cOlTelated Scooter and Black Hole Cards Scooter Cards and Black Hole Cards (Appendix I), built-in components of the Superhero Social Skills program, were used to evaluate change in rule-following behavior over the course of the social skills program. During social skills lessons, Scooter cards were distributed to children based on their following group rules (i.e., Get Ready, Follow Directions, Be Cool, and Participate), while failing to follow group rules earned participants Black Hole cards. The number of Scooter cards and Black Hole cards passed out during each social skills lesson was charted to determine changes in rule Behavioral Intervention Rating Scale The Behavioral Intervention Rating Scale (BIRS; Von Brock Elliott, 1987) was administered to teachers and parents upon the completion of the intervention. The BIRS item Likert-type scale that is effective in assessing treatment efficacy and iatrogenic treatment effects. Responses to items are based on a six point scale ranging from strongly agree to strongly disagree. The BIRS will be used to evaluate teacher and parent judgments as to the efficacy of the Superhero Social Skills program. Consumer Satisfaction Survey In order to evaluate participant satisfaction in this study, the Child Consumer Satisfaction Survey CCSS) was developed ( Block, Hood, Radley, unpublished; see CCSS simple Likert-determine acceptability of interventions, as assessed the participant. These will used evaluate participant experience of the Superhero Social Skills Social Validity Checklist The Social Validity Checklist (Appendix D) adapted from Bellini (unpublished) was completed by parents upon completion of the intervention. The social validity checklist is a 7-question Likert-type scale that is used to assess parental perceptions of the efficacy and of the Superhero Social Skills program. 48 Beha vioral Rat ing Sca le S IRS: & admin istered pa rents cOlllplelion orlile S IRS is a 24-Liken-e ffective s ix J"fUlging stro ngly wi ll lIsed teac her parcntjudgmcms lO e rticacy or til e Ski lls Child Consllmer Sati s t:1ction thi s Sati s facti on (CeSS) & Appendix C). The cess is a s impl e 5-question Liken -type scale that is used to dete nni ne by partic ipant. scores wi ll be to eva luate Soc ial Ski lis program. Socia l Va lidity Checkli st Checkli st A ppendix Bell ini unpubli shed) tile in tervention. valid ity effi cacy worth tile Superhe ro Socia l was video-based social objective form 49 Intervention Fidelity To assess the fidelity with which the Superhero Social Skills program was implemented, an intervention fidelity checklist W(lS created (Appendix E). The checklist was created by compiling a detailed, sequential list of the steps of the Superhero Social Skills program. At the conclusion of each social skills lesson, the social skills instructors completed the checklist. A mean percentage of treatment fidelity was then calculated. Design A replicated AB single-subject design was used to compare the efficacy of video­based instruction to traditional didactic instruction (Hayes, 1981). Baseline observations of soci(ll interaction during analog free time periods were conducted three times for each participant with ASD. Observations of free time social interaction were conducted during the experimental phase at the conclusion of each social skills lesson. The replicated single-subject AB design involves the establishment of stability during the baseline phase (A), and the introduction of change in conditions affecting the subject (B). Although several threats to internal validity present themselves when using AB single-subject research design, Kazdin (1982) and Kratochwill (1992) have concluded that AB designs are valid when they meet strict conditions. Kazdin argued that AB designs are valid when: 1. The data are objective 2. Assessments occur on multiple occasions 3. The target behavior being treated is stable 4. When participants fonn a heterogeneous group and effects. single-subject Assessments of social interaction during analog free time periods are conducted multiple 5. When the intervention produces immediate and marked effects. Kratochwill added several criteria that a study must meet before it can be considered valid: 1. The study must be planned 2. Maintain a high level of treatment integrity 3. Consist of a standardized treatment and 4. Produce a large effect size. 50 If the trend observed in phase A suddenly changes with the introduction of phase B, one can be confident that changes in phase B are responsible for changes in behavior (Hayes, 1981). However, results from the AB design can be confounded by factors such as maturation, the effect of measurement, and other external events (Kratochwill, 1978). To control for confounding variables, the multiple baseline design is commonly used in single-subject research. Essentially, the multiple baseline design involves implementation of phase B at different times for each subject by employing staggered baseline phases. This technique allows for control of reaction to assessment and maturation, as well as reducing the probability that external factors are producing the changes in the target behavior (Hayes, 1981). Because the social skills programs evaluated in the current study are designed to be conducted in groups of children, a multiple baseline design could not be used. This study meets both with Kazdin's and Kratochwill's criteria for valid single­subject research. The problem being treated, ASD, is a stable trait. The participants are heterogeneous in diagnosis, age, inclusionary abilities, and level of education. 51 structured to be included in a group receiving Superhero Social Skills via video-Participant 1 and Participant 2 were assigned to the video group. After teachers nominated children to participate in the program, parents were contacted through a note sent home in order to obtain consent for their child to participate in the current study (see Parental Consent Form, Appendix E). The purpose of the current study was explained and any questions that parents/guardians may have were answered. After consent was obtained, 4 children with ASD and 4 typically developing peers were randomly selected for inclusion in the numbers were drawn at random to determine the participants in the video-based ASD-Participant 4-and instruction format in which the primary researcher and a trained graduate assistant presented each lesson. times during the study, and the treatment consists of highly stmctured lessons, delivered by trained researchers. Results of this study also suggest that the current study meets the criteria for producing large changes in the target behavior. Procedures Two children with ASD and 2 normally developing peers were randomly selected video-Participant I cunent study. Each child who returned a consent form was assigned a number, after which detennine Superhero Social Skills group. When those participants had been determined, the remaining 2 children with ASD-Participant 3 and Participant 4-and their peers were assigned to the group that received Superhero Social Skills through a traditional didactic phase of the current study was conducted. During baseline phase, participants were observed in an analog setting for three 10-minute free time sessions to establish a baseline measure of social interaction. Observations were conducted in the same room in which the lessons were taught. During free time observations, six toys/games as previously described were provided for the children. A time-sampling procedure will be used to determine the number of 10-second intervals in which the child with ASD initiates interaction or responds to their peers during the 10-minute interval. superheroes-Interactor of children demonstrating the skill. After viewing several video models of the skill, the viewers are invited to role-play the skill. The second weekly lesson reviews the acquisition of the skill through a repeated viewing of the entire social skills video and additional role- 52 Baseline After participants had been randomly assigned to treatment groups, the baseline I O-IO-IO-Superhero Social Skills Program The program is designed to be taught twice weekly over an 18-week period (see program description in Appendix F). One new social skill is presented each week to each group in the study, with each skill consisting of two lessons. Lessons from the social skills program, Superhero Social Skills, are intended to be taught to children with Autism when accompanied by their peers. During the first session of each week, the social skill is presented by animated superheroes-The Initiator and lnteractor Girl, and their sidekick Scooter the Robot. The superheroes introduce the skill, provide rationale for use of the skill, and outline steps for correct demonstration of the skill. The superheroes then invite the children with ASD and their peers to watch video models participate learned provided for the home, using similar lessons to those taught at school or in a clinic. In addition to the use of DVDs to present social skills, Superhero Power Cards are used. Power Cards are used for self-monitoring, and are marked by a facilitator or parent when the child successfully demonstrates the learned social skill. The Power Card is also used as a homework mechanism, allowing empty spaces on the card to be filled in by the parent when the social skill is demonstrated in a generalized setting. Self-monitoring occurs at the end of each social skills lesson, when the participant transfers the marks on the Power Card to a Power Poster, which is displayed in the treatment setting. Each week, participants in the program receive a new Power Card, encouraging the demonstration of the current week's social skill. Social Stories in the form of a comic book are also given as homework. Each week, participants are assigned a new comic book chapter in which the superheroes demonstrate the social skill. Group-study, only the Superhero Social Skills eight foundational skills were taught. These lessons included Introduction to the Social Skills Group, Get Ready, Following Directions, Anxiety Reduction, Participating, Generalized Imitation, Body Basics, Expressing Wants and Needs, and Joint Attention (see Appendix F for complete description of the skills). A new lesson was presented at the beginning of each new 53 playing. At the end of each social skill lesson, participants in the program paI1icipate in social games that allow them to practice newly leamed skills in an analog setting. In order to foster generalization, a DVD of the video modeled social skill is Self-monitoring Superhero Social Skills Group-Implementation Social skills lessons were taught two times per week for 45 minutes. During this Imitatioll, In these cases, social skills instruction was presented on the next available day. The the group receiving only traditional didactic instruction. Group-same content as those in the video-based group, except that the group facilitators, the primary researcher and another School Psychology graduate student at the University of Utah, presented all the information, rationale, and social skills steps. While the didactic as part of the weekly social skills lessons and homework. Skills group and the didactic group over the 8 weeks. Probes were conducted in an identical fashion to baseline social interaction probes: each participant with ASD was observed during each free time period of 10 minutes in the analog setting. These periods were filmed, and coding of social initiations and responses were completed at a later time. Social engagement was coded using Bellini's Social Interaction Codes (Appendix B; Bellini, unpublished). Bellini's codes allow for coding of seven different types of social initiation and six types of social responses. While codings were completed using the 54 week. Occasionally, regular social skills instruction was interrupted by school holidays. instlUction social skills lessons follow a set schedule for the group viewing the animated videos and Didactic Group-Implementation Those participants assigned to the didactic teaching group received exactly the teaching group did not view the animated videos, they viewed the peer modeling videos Observation of Social Initiation and Responses Following the social skills lesson, probes of social interactions during the analog free time period were conducted 18 times for each child in the Superhero Social Ski lls see Appendix B. These codes allowed for detailed descriptions of participant's social engagement during the free time periods. graph, allowing for an analysis in change of rule-following behaviors over time. The recording was completed by the participant as they transferred marks from the Power Card to the Power Poster. This recording, which also functioned as a self-monitoring facilitator. nonoveiiapping 55 specific initiation and response codes, data analysis only considered the more general categories of initiation and response. For a more detailed description of Bellini's codes, In addition to filming analog play periods and coding for social interactions, the number of Scooter Cards and Black Hole cards distributed during the social skills lesson was recorded at the close of each lesson. This number was recorded and plotted on a number of Power Charges earned was also recorded at the end of each lesson. This self-monitoring technique for the participant, was verified by the group facilitator. Upon conclusion of the social skills lessons, participants with ASD in both groups were given the Child Consumer Satisfaction Survey (CCSP). The CCSP items were presented verbally to participants to ensure understanding of each question. Parents of children with ASD and their teachers were given the BIRS and the SRS, while only parents of children with ASD were asked to complete Bellini's social validity checklist. These measures were collected by the primary researcher and scored. Effect sizes and percentage of nonoverlapping data points were also calculated by the primary researcher using data collected through the Bellini Observation System. 56 calculated for each group. Percentage of observed 10-second intervals in which social initiation and social response occurred were calculated for each participant with ASD for baseline and treatment phases. Using the Busk and Serlin (1992) No Assumptions model, effect sizes were calculated for the video instruction group and the traditional didactic instruction group. Effect sizes were interpreted using the Cohen's (1988) standards for interpreting effect sizes. Cohen defined a small effect size as one between 0.1 and 0.3, a medium effect size as between 0.3 and 0.8, and a large effect size as a correlation of 0.8 and above. rule-following behavior over time, as measured by Scooter and Black Hole cards. Pre- and posttest scores obtained from SRS were compared. Constantino suggests that changes in Data Analysis In order to compare efficacy of the two forms of instruction, effect sizes were In conjunction with effect sizes, percentage of nonoverlapping data points (PND; Scruggs, Mastropieri, & Casto, 1987) was calculated for social interactions of children with autism during the observed free time probes. PND is computed by determining those intervention data points whose value is higher than the highest baseline data point value. Once these points have been determined, the number of nonoverlapping data points is divided by the total number of intervention data points. When PND has been determined for each subject, PND scores for each participant were averaged to obtain one PND score for each group. The number of Power Card circles completed by each participant and the percentage of free time spent in social interaction were correlated using a Pearson Correlation Coefficient. Descriptive statistics were used to determine changes in rule­following scores of measures significant Constantino's conservative recommendation, to significant changes. Teacher and parent treatment acceptability scores gathered from the BIRS and Bellini's Social Validity Measure, as well as participant ratings of acceptability on the CCSS, were also analyzed using descriptive statistics. 57 SRS T-sco res by one to two standard error ofl1leasures should be considered sign ificam treatment effects (Constantino, 2002) This study follows Constantino' s more conserva tive recommenda tion, assuming scores that change by two or more SEM be Bellini 's Vali dity acceptabili ty cess, a lso descript ive stati sti cs. CHAPTER 3 RESULTS study to skills to preschool-age children with ASD in a specialized setting, and to compare the efficacy of video-based presentation of social skills training to a didactic method of instruction. The program is based on a superhero theme with animation and comic books, and incorporates research-validated components of previously developed social skills programs, including video modeling, social stories, direct instruction, and self-monitoring. Research Question 1 What the effect size and PND of Superhero Social Skills for Children with Autism when presented in its video-based format, as measured during a social interaction observation during analog free time? Analysis of treatment fidelity checklists revealed that the Superhero Social Skills video-based intervention was delivered with 100% integrity. Overall, participants in the video-based group initiated social interactions an average of 4.72% of intervals during baseline, and initiated interactions during 4.7% of The purpose of this siudy is to pilot test a program developed teach social ski lls ch ildren sett ing. (0 efficacy prescntalion ski lls meihod in structio n. all anil11mion boo ks, ofpreviollsly socia l sk ills sIories, instruct ion, self­monitoring. I 11'Iw( is fhe '4tJecl sbt a/Superhero Skills/or ('/Ji!drell willi Au/isl/I whell pre.)'el1led ill ilS llOsedformal, mcaslwed il/ferae/ioll ohsen'ulioll ill/rilig Clllalogfree lime'! Anal YS IS revea led thutlhe Ove mll , participa nts interac tio ns avemge 4. 72% initialed int eractions 4 .7% 59 baseline, and 11.95% of intervals during treatment. Participants in the video-based group averaged social engagement during 6.35% of baseline intervals, and averaged social engagement during 16.75% of intervals. Using Cohen's metric to interpret average effect sizes for participants in the video-based group, a small effect size was observed in social initiations (£5=.295). Large effect sizes were observed in both social responses (ES=\.05) and total engagement (£5= 88). Skills skill 1.1% treatment. of 6.3% 1, 2, forjudging 5=(£"5=0.82), (£5=0.71). presented. At the end of the intervention, follow-up forms were sent to Participant 2's parents, but were never returned. While incomplete, the data show progress made before treatment intervals. Participants responded to others during 1.35% of intervals during treatment effect smal1 (E5"=.295). (£S=1.05) 8=.Participant 1, assigned to the Superhero Social Skil1s video-based instruction group, was present for 15 of 16 social skil1 lessons. During the 10-minute baseline observations, Participant 1 initiated social initiations an average of 0.05% of intervals. During treatment, Participant 1 averaged social initiations during 1.5% of intervals. Participant 1 responded to others during 1.1 % of intervals of baseline, and 4.7% during treatment Social engagement was observed in an average of 1.6% of intervals during the baseline phase, and an average of6.3% of intervals during treatment (Figures 1,2, and 3). Using Cohen's metric for judging effect sizes, large effect sizes were observed in social initiations (£8=0.80) and total social engagement (£S=0.82), while a medium effect was observed in social responses C};,,')'=O. 71). Percentage of nonoverlapping data points was calculated at 40%. Participant 2, assigned to the Superhero Social Skills video-based instruction group, was present for 9 of 16 social skill lessons. Participant 2 was unexpectedly removed from the Pingree Center just after half the social skills lessons had been P a r t i c i p a n t 1 - S o c i a ! I n i t i a t i o ns Session Figure 1. Initiations, Participant 1 o Participant l-Sociallnitiations 10.0% 9.0% c 8.0% 0 '0 • '0 •• 7.0% • '0 ~ 6.0% ;; .•. 5.0% •< ..E. 4.0% w• • " 3.0% • "• ~ 2.0% 1.0% L 0.0% 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Figure I. Social Initiations, Participant 1 j P a r t i c i p a n t 1 - S o c i a l R e s p o n s es 2.4.0% Session Figure 2. Social Responses, Participant 1 ON Participant l -Social Responses 26.0% 24.0% 22.0% •• 2180..00%% ••& ~ 16.0% .~. 14.0% ~ • • 12.0% ..§. ~ t 10.0% ~•0 8.0% • ~ 6.0% t "" ~ 1\ 1,'" 1\ 0,0% • 1 3 , 5 6 7 8 9 10 11 12 13 " 15 16 17 18 Session I P a r t i c i p a n t 1 - T o t a ! S o c i a l E n g a g e m e n t 27.0% 24 .0% 21.0% "• E 18.0% ~ •~ '0 15.0% • ] c 0: 12.0% 6.0% l .O% 00% 2 l , Participant i-Total Social Engagement 5 6 7 8 , 10 11 12 Il " 15 16 17 18 Session Figure 3. Total Social Engagement, Pal1ieipant I 9.4% Participant 2 averaged social initiations during 7.9% of interactions. Participant 2 responded to others during 1.6% of intervals in the baseline phase, and responded to others an average of 19.2% of intervals during the treatment phase. Social engagement was observed in an average of 11.1% of intervals of baseline, and an average of 27.2% of intervals of treatment (Figures 4, 5, and 6). A small negative effect was observed in social initiation (ES=-0.21). Large effect sizes were observed in both social responses (ES= 1.28) and total social engagement (ES= 1.05). Percentage of nonoverlapping data points was calculated to be 55.6%. Research Question 2 What is the effect size and PND of Superhero Social Skills for Children with Autism when presented in a traditional didactic format, as measured during a social interaction obsemation during analog free time? Superhero Social Skills group was delivered with 100% treatment integrity. 63 removal from the school. During the 10-minute baseline observations, Participant 2 initiated social interactions an average of 9 A% of intervals. During the treatment phase, 11.1 % 1.28) Overall, data collected suggest that video-based Superhero Social Skills is effective in increasing social responding and total engagement in individuals with ASD. Both participants made gains in social responses and total engagement, while only Participant 1 improved in social initiations. This research question was satisfied with the data collected. a/ld PNf) (~lSuperhero with presellted ill social observation analogfree lime? Fidelity checklists from the didactic instruction group revealed that the didactic P a r t i c i p a n t 2 - S o c i a l I n i t i a t i o ns Social Initiations, Participant 2-Sociallnitiations 20.0% 18.0% 16.0% 0 0 0 • 14.0% ·0 ,:~. .~ 12.0% ~ ,·. .0 0% .s•< ~ ~ 8.0% • E • 6.0% • ' .0% 2.0% 0 .0% 1 2 3 5 6 7 8 9 10 12 Session Figure 4. Initiat ions, Participant 2 Participant 2-Social Responses 55.0% 50_0% 45.0% • 40.0% 0 0 :"•~!! 35.0% .•. 30.0% ~ ..••<.~ 25.0% • ~ 20.0% "• • ~ 15.0% 10_0% ~ 5.0% ~ I 0.0% +- , 3 , 5 6 7 s 9 >0 11 12 Session Figure 5. Social Responses, Part icipant 2 Session Figure 6. Total Social Engagement, Participant 2 Participant 2-Total Socia l Engagement 55.0% 50.0% 45.0% ~ • 40.0% E • ~ • ~ •< 35.0% ~ •~ 30.0% "~ ~ 25.0% .5 "• 20.0% ~ ~ e •• 15.0% 10.0% '.0% 0.0% , 3 , , 6 7 8 9 10 11 12 Overall, participants in the didactic instruction group initiated social interactions during an average of 3.25% of intervals during baseline, and 6.3% during treatment. The 2 participants responded to others during an average of 7.75% of intervals of baseline, and 16.65% during treatment. Baseline social engagement was observed during 11.1% of intervals, with social engagement during 22.95% of treatment intervals. Using Cohen's metric to interpret effect sizes, a small effect size was observed for social initiations (£5=.46), and large effect sizes were observed for social responses (£5-1.48) and total engagement (£5=1.54). Participant 3, assigned to the didactic instruction group, was present for 15 of 16 social skill lessons. During the 10-minute baseline observations, Participant 3 initiated social interactions an average of 2.7% of intervals during baseline, and 10.6% during phase, and 18.1% during treatment phase. Social engagement during baseline was (£5-1.76), (£5=2.76), (£5=3.09). effect sizes are classified as large by Cohen's metric forjudging effect sizes. Percentage of nonoverlapping data points was calculated at 100%. social skill lessons. During the 10-minute baseline observations, Participant 4 initiated social interactions an average of 3.8% of intervals, and initiated interactions during 2.0% of treatment intervals. Participant 4 responded to others during 13.3% of intervals during baseline, and in 15.2% of treatment intervals. Social engagement was observed in an 67 of7.75% 11.1 % (ES=.(1:'S=1.48) (ES= 1.54). I O-of2.7% treatment. Participant 3 responded to others during 2.2% of intervals during baseline 18.1 % observed in an average of 5% of intervals, and in 28.7% of intervals during treatment (Figures 7, 8, and 9). Effect sizes were calculated for changes in social interactions (ES=1.76), social responses (FS=and for total social engagement S=All for judging 100%. Participant 4, assigned to the didactic instruction group, was present for 16 of 16 PaI1icipant of3.8% Participant 3-Social Initiations 26.0% 24 .0% 22.0% 20.0% c 0 :•'0•5 18.0% ~ 16.O'K. "0 • '0 14.0% ",• • 12.0% .S '0 ~ 10.0% • f • ..... • ~ 6.0% ' .0% J 2.0% 0.0% 2 , • 5 6 7 • 9 10 11 12 l' 15 16 17 I. Session Figure 7. Social Initiations, Participant 3 P a r t i c i p a n t 3 - S o c i a I R e s p o n s es 40.0% 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Session Figure Responses, Participant Participant 3-Social Responses 40.0% 35.0% ••• 30.0% ••& 1 25.0% .. "• 20.0% < .,~. N 15.0% •• •" 10.0% 5.0% 0.0% 1 , 3 , 5 6 7 8 , 10 11 12 13 15 16 17 18 Senioo Figurc 8. Social Responses, Partic ipant 3 55.0% P a r t i c i p a n t 3 - T o t a l S o c i a l E n g a g e m e n t 50.0% 0.0% 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Session Engagement, Participant Participant 3-Total Social Engagement 55.0% 45.0% E • 40.0% •••~ •• 35.0% Ji! .~". 30.0% ~ .1 25.0% .. ••• 20.0% E••• 15.0% ~ 10.0% 5.0% V 0.0% 2 3 , 5 6 7 8 9 10 II 12 13 15 16 18 Session Figure 9. Total Social Engagcmcnt Participanl 3 71 average of 17.2% of baseline intervals, and in 17.2% of intervals duri