|USFDC Home | USF Electronic Theses and Dissertations||| RSS|
This item is only available as the following downloads:
Effects of Social Skills Tr aining on the Interpersonal Behaviors of Elementary School Students in an After-School Program by Robert Caples A thesis submitted in partial fulfillment of the requirements for the degree of Education Specialist Department of Psychologi cal and Social Foundation College of Education University of South Florida Major Professor: George M. Batsche, Ed.D. Michael Curtis, Ph.D. John Ferron, Ph.D. Date of Approval: October 18, 2005 Keywords: social skills, after-school program, multiple baseline, single-case, intervention Copyright 2005, Robert Caples
i Table of Contents List of Tables iv List of Figures v Abstract vi Chapter One Introduction 1 After School Programs 1 The Efficacy of After School Programs 3 Social Skills Training 6 Theoretical Framework 7 Purpose of this Study 7 Research Questions 8 Hypotheses 8 Definitions 8 Chapter Two Literature Review 10 Overview 10 ASPs 10 ASPs in the Educational Context 11 Afterschool Programs as an Educational Tool 12 Social Skills Training 18 Social Skills Training Effectiveness 20 Social Skills Training with Upper Elementary School Students (Grades 3-5) 21 Social Skills Training in Small Groups 26 Social Skills Training Using Modeling and Role Playing 28 Social Skills Training with Urban, At-Risk, or Minority Children 32 Chapter Three Methods 41 Participants 41 Setting 41 Dependent Measures 41 Behavior Observations 41
ii Selected Social Skills Rating In strument 45 Design and Procedures 46 Step 1: Site Selection 47 Step 2: Selection of Participants 48 Step 3: Baseline Data and Staff Training 49 Step 4: SST Introductory Meeting 50 Step 5: Session 1 and Continued Baseline Data Collection 51 Steps 6-8: Continued Intervention and Continued Baseline Data Collection 52 Data Analysis 52 Chapter Four Results 54 Introduction 54 Significance Levels and Experiment al Control 54 Hypotheses 58 Hypothesis 1 58 Hypothesis 2 62 Hypothesis 3 66 Comparison of Relative Effectiveness of Reinforcement, Direct Instruction, and the Combination of Bo th Interventions 68 Chapter Five Discussion 79 Introduction 79 Discussion of the Findings 79 Possible Explanations for the Present Findings 80 Social Skills Training Factors 81 Measurement of the Dependent Variables 84 Consistencies with Established Re search 86 Limitations and Threats to Valid ity 87 Sampling Procedures 87 Reactivity 88 Observer Drift 88 Failure to Include a Peer Comparison 89 Participant Variability 89 External Events 89 Staff Training 90 Implications and Future Directions 90 References 94 Appendices 103 Appendix 1: Social Skills Rating Form 104
iii Appendix 2: Behavior Observation Form 105 Appendix 3: Interview Guide: Behavior Identification 106 Appendix 4: Interview Guide: Partic ipant Selection 108
iv List of Tables Table 1 Mean Level of Sitting Properly Data Within and Between Phases 69 Table 2 Range of Sitting Properly Data Within and Between Phases 69 Table 3 Percentage of Nonoverla pping Points Between Phases for Sitting Properly Data 70 Table 4 Mean Level of Raising Hand Data Within and Between Phases 70 Table 5 Range of Raising Hand Data Within and Between Phases 71 Table 6 Percentage of Nonoverla pping Points Between Phases for Raising Hand Data 71 Table 7 Mean Level of Attendi ngData Within and Between Phases Baseline1 72 Table 8 Mean Level of Attendi ngData Within and Between Phases Baseline2 72 Table 9 Range of Attending Data Within and Be tween Phases Baseline1 73 Table 10 Range of Attending Data Within and Between Phases Baseline2 73 Table 11 Percentage of Nonoverla pping Points Between Phases for Attending Data Baseline1 74 Table 12 Percentage of Nonoverla pping Points Between Phases for Attending Data Baseline2 74
v List of Figures Figure 1 Target Behaviors 42 Figure 2 Example of Multiple Baseline Across Behaviors 47 Figure 3 Raising Hand, Sitting Properly, a nd Attending Data 57 Figure 4 Sitting Properly Data 59 Figure 5 Attending Data 60 Figure 6 Raising Hand Data 67 Figure 7 Participant One Data 75 Figure 8 Participant Two Data 76 Figure 9 Participant Three Data 77 Figure 10 Participant Four Data 78
vi Effects of Social Skills Tr aining on the Interpersonal Behaviors of Elementary School Students in an After-School Program Robert Caples ABSTRACT Social skills training was investigated in an after-school program setting with four sevenand eight-year-old males. Two were Hispanic and two were African-American. Social skills training consisted of a direct instruction, behavioral learning model of skillstreaming as described by McGinnis & Go ldstein (1997). There were four major components to each social skills training sess ion: (1) an explanation of the skill being taught; (2) modeling by th e researcher of the sk ill being taught; (3) ro le play by each of the participants; and (4) perfor mance feedback regarding the role plays. Sessions lasted approximately 30 minutes and were held weekly throughout the intervention phases of the study. The behaviors taught were raising one s hand before leaving the seat, sitting properly in ones seat, and attending to homework or staff instructions. Participants also received reinforcement for performance of the social skills in homework sessions at the after school program as is consistent with the literature regarding social skills traini ng. However, the reinforcement and behavior learning (direct instruction) components were introduced both in combination and at separate times to experimentally control for the influence of each intervention component. This research
vii design allows for the investigation into the relative effectiveness of direct instruction versus reinforcement in social skills training. Experimental control was demonstrated through the use of a multiple baseline across behaviors design. Dir ect instruction and reinfor cement for behaviors were systematically introduced at separate tim es, keeping some behaviors under baseline condition while moving others into intervention conditions. Visual analysis of the results indicates that social skills training was effective in improving the three target be haviors of all four stude nts. Direct instruction, reinforcement, and the combination of the tw o presented together all were effective in improving the target behaviors. Possible interv ention effects not relate d to social skills training may have influenced the behavior of attending.
SST in ASPs 1 CHAPTER ONE Introduction After School Programs The education of children is a large ta sk. Schools do much toward that end, but there is more to be undertaken. After school programs (ASPs) have been one response by society toward the ever increasingly complex task of education. ASPs have been defined broadly as a wide range of program offe rings for young people that take place before school, after school, on weekends, and during the summer and other school breaks (Peter, 2002). ASPs vary in their program areas (recreation to academic instruction), time of operation (from a few hours to six days a week), number of staff (1 to over 50), theoretical focus (e.g., youth development, acad emic improvement.), cost of services, and location (e.g., school-based, c hurch-based.) (Fashola, 2002; McComb & Scott-Little, 2003, National Youth Development Information Center, 2003). In Hillsborough County, Florida, the B oys and Girls Clubs of Tampa Bay (BGCTB), Young Mens Christian Associ ation (YMCA), and Hillsborough Country Recreation and Conservation alone list more th an 95 separate after-s chool program sites (BGCTB, 2003; YMCA,2003; Hillsborough Count y Recreation and Conservation, 2003). In addition, there are 162 public elementary a nd middle schools in th e School District of Hillsborough County (SDHC), most of which have after-school programs (SDHC, 2003). There are two primary ways in which ASPs can assist in education. First, ASPs can work to meet the specific needs (e.g., providing a safe physical environment, offering opportunities to engage in structured activit ies, providing positive adult supervision and guidance of children during after school hours). Second, ASPs can also be tools for
SST in ASPs 2 meeting the broad educational needs of ch ildren (e.g., reading, math, social skills) (National Youth Development Information Center, 2003). Once the school bell rings to end the da y, a host of issues co nfront children and their families. Greater numbers of single-parent families and the need for these single parents to work have left many children without adequate supervision in homes and neighborhoods (Lipsitz, 1984). This trend also leaves children lacking opportunities to engage in activities promoting academic en richment, socializati on, cultural awareness, and so forth. From a safety perspective, adolescents are most likely to be the victims or perpetrators of violent crim es between the after school hours of 2:30 PM and 8:30 PM (Chaiken, 1997). In addition, children who ar e not supervised during after school hours are more likely to engage in risk-taking be haviors such as alcohol and drug use (Chung, 2000). Beyond supervision, children have a variety of broad educational needs. In particular in reading, the 2003 National A ssessment of Educational Progress (NAEP) Report, documenting evidence toward the De partment of Educations Goals 2000, indicated that onl y 32% of 4th graders and 32% of 8 th graders were at or above proficiency in grade-level reading standards (US Depart ment of Education, 1995). The Florida Department of Educati on reports that only 60% of 4 th graders scored a level three (proficient level) or higher on the Florida Comprehensive Achievement Test (FCAT) reading in 2003. In 8 th grade, that number drops to 49%. In math, 54% of 4 th graders scored at or above a level three, while 56% of 8 th graders scored at or above a level three. In addition, the State of Florida now mandates retention for third grade students who score a Level 1 on the FCAT reading subt est. In 2003, The Florida Department of
SST in ASPs 3 Education reported that23% of students scored a Level 1 on th e FCAT reading subtest. In the behavioral realm, Nansel et al. (2001) re ported that more 16% of U.S. school children said they had been bullied in the current term. In addition, 6% said they had been both bullied themselves and bullied other children, sugge sting a strong need for behavioral support of children. After school programs have the propensity to address some of these broad educational needs and to foster more positive educational outcomes for students who participate in th em (Fashola, 2002). The past decade has brought an era of accountability to educational institutions. The passage of No Child Left Behind (NCL B) is but one example of this focus on educational outcomes and the efficacy of the schooling process. High stakes testing, the national focus on reading (e.g., Reading Firs t, NCLB), and school grading are all influences by the school accountability move ment. Schools are asked to meet higher expectations, and to show more evidence that these expectations have been met (FDOE, 2004; SDHC, 2004). School districts have been turning to ASPs to help address some of these needs. Over the last two years, over $25.5 million was awarded to school districts and other after school program organi zations in Florida through the 21 st Century Community Learning Centers program alone all for after school programs. The Efficacy of After School Programs Research evaluating the effectiveness of ASPs in meeting the broad educational needs discussed in the last section has been mixed (Barker, 1998; U.S. Department of Education, 2003). For example, the 21st Ce ntury Community Lear ning Centers (CCLC) program, funded through the federal govern ment, had a $1 billion budget in 2002. It funds about 7,500 ASPs in 1,400 communities. Th e programs Summary of First Year
SST in ASPs 4 Findings concluded that reading test scores, as well as grades in most subjects, were not higher among program participants. In addi tion, participating ASPs did not increase students feelings of safety during after sc hool hours. Fewer than 38% of middle school students reported that the ASP was a good place to get homework accomplished (US Department of Education, 2003 ) Possible reasons for the lack of effectiveness include insufficient accountability, participation ra tes, level of academic focus, and poor generalization of skills taught in after school programs to the school context. There were successful individual programs within the 21 st CCLC program. Research from several successful programs in Kansas yielded a list of factors that contributed to success: low staff:child ratios, constant supervision of children, support for family involvement, communication between staff, families, and schools, integration into the community, sufficient indoor and outdoor space, staff/child collaboration with activity planning, established policies and pr ocedures regarding safety and health, qualified and trained staff, staff support, effective organizational management, and policies and procedures that are responsive to the needs of children and families. Other studies also indicate positive results. For example, a 10-year study of LAs BEST after school enrichment program examined program participants in the second through fifth grades during the 1993-1994 school year, and followed them through the 1997-1998 school year. These stude nts were compared with non-participant peers using the Comprehensive Test of Basic Skills or the Stanford-9 Achievement Test in reading, mathematics, and language arts. The authors repo rted that higher participation in the ASP was correlated with higher sc ores on these measures. The au thors also reported that
SST in ASPs 5 greater program participation was correlate d with higher attendance rates at school (Huang, Gribbons, Sung Kim, Lee, & Baker, 2000). Baenen, Lindblad, and Yaman (2002) inve stigated an extended learning program in Wake County, North Carolina. Children in grades K-12 were given extra academic instruction by either their own teacher, anothe r teacher, or a volunteer. This instruction occurred after school and on Saturdays (year round schools offered the services during intersessions). Results indicat ed that students who participated in the program in all grades except grade 3 showed significantly greater academic improvement than students who did not participate. These results sugge st that after school programs can provide extra academic engaged time (AET), and that th is additional AET can result in increased academic performance. Barker (1998) examined the effects of a highly specialized ASP for juvenile delinquents. Compared with a control group, th ere were more than 50% fewer criminal convictions among program participants. On a self-report measure of aggressive behavior, self-discipline, and social control, participants evidenced improvement in selfcontrol. Similarly, Lamare (1997) cited teacher s reports that ASP participants had more positive social skills. One finding consistent thr oughout the research is that ASPs are quite diverse from the activities offered to the data supporting their effectiveness. Given this variability, the quality of specific strategi es implemented in, rather than the mere existence of, ASPs is of importance. Ther e are numerous educational and psychological strategies (e.g., assessments, interventions, modifications) that have been found to be
SST in ASPs 6 effective in schools for a wide range of behaviors (Shinn, 2002). However, these strategies have seldom been studied in ASP settings. Social Skills Training A childs ability to interact successfully with adults and peers is critical (Gresham & Lemanek, 1983). For example, children w ho experience less accepta nce by their peers are more likely to experience learning and adjustment problems both in and out of school (Hughes & Sullivan, 1988). Social skills training is a direct attempt to build a childs success in interacting with a dults and peers. While some ASPs have conducted activities in such areas as character and leadership and health and life skills training (BGCA, 2003), and some evaluations of ASPs include measures of social behaviors (Lamare, 1997), an extensive literature review has revealed no studies examining the explicit use and impact of a formal social sk ills training program in an ASP. Social skills have been de fined as learned behaviors th at are socially acceptable and which enable an individual to interact successfully with others (Gresham & Elliott, 1984). The presence of social skills has been shown to be connected with a variety of other significant educational variables. For example, the presence of social skills influences both the quantity and quality of academic engaged time in educational settings (DiPerna, Volpe, & Elliot, 2001). Social skills can be learned natura lly throughout the course of child and adolescent development; however, several fact ors warrant the explic it teaching of social skills. Societal changes, for example social mobility and the reduced role of the church, have made it less likely that social skills will be taught in a childs home or community environment. Children also spend a significan t amount of time engage d in activities that
SST in ASPs 7 do not directly teach real-life social skills, such as watching television. Given these and other factors, the explicit instru ction of social skills in schoo l settings has, in many cases, become a necessity (Cartledge & Milburn, 1986). Several methods for teaching social skill s have been developed. McGinnis and Goldstein (1997) describe a skillstreaming approach, which focuses on the following four direction instruction pr inciples of learning: modeli ng, role-playing, feedback, and transfer. The authors approach the instruction of social skills from a skill-deficit model, assuming that the child lacks certain behavioral skills which can be directly taught. Reinforcement and behavior strengthening are the primar y tools used for increasing social skills, rather than punishment for inappropriate behaviors. While the authors discuss how skillstreaming can increase sel f-esteem, the intervention does not focus on affect, but rather on skills (McGinnis & Goldstein, 1997). Theoretical Framework Social skills training stem s from Banduras (1977) so cial learning theory and Goldsteins (1981) psychological skill tr aining. Bandura (1977) described how individuals vicariously lear n social behavior through m odeling of behavior and its reinforcement, and imitating the observed be havior. Goldsteins psychological skill training (1981) extended contemporary psychological practice to include the direct instruction of psychological skills. Other schools of thought assumed that individuals already possessed these skills, but that the performance of these skills needed to be increased. McGinnis and Goldstein (1997) expa nded a social skills training curriculum from this concept of psychological skill training. Purpose of this Study
SST in ASPs 8 No studies have been found that have ev aluated the Stop & Th ink (or any other) social skills training program in an ASP setting. The purpose of this study, therefore, is to examine the effectiveness of the Stop and Think social skills training program on the interpersonal behaviors of elementary school -age children in an ASP setting. This study will contribute not only to the social skills training literature base, but also to the ASP literature in general. The edu cational significance of this st udy is its extension of the social skills research to the after school program setting. The number of after school programs is increasing (BGCA, 2003). Research investigating the effectiveness of practices within these programs, then, will b ecome more important, especially with the increased focus on the use of evidence-base d practices in school and school-related settings.. Research Questions The research question in this study is as follows: 1. In an after-school program setting, what is the relationship be tween social skills training and positive social behaviors among elementary school age children? Hypotheses 1. The direct instruction component of social skills training will result in a significant increase in positive social behaviors. 2. The reinforcement component of social sk ills training will result in a significant increase in positive social behaviors. 3. When combined, the direct instruction and reinforcement components of social skills training will produce a significant increase in positive social behaviors.v
SST in ASPs 9 Definitions After School Program : The National Institute on Outof-School Time defines outof-school time programs as encompassing a wide range of program offering for young people that take place before school, after school, on week ends, and during the summer and other school breaks (Peter, 2002). More detailed components of ASPs as discusse d in the literature will be presented in the next section. Social Skills : Social skills have been discu ssed as learned be haviors that are socially acceptable and which enable an indivi dual to interact successfully with others (Gresham & Elliott, 1984). Ladd and Mize (1983) described social skills as the purposeful organization of th inking and behavior to achie ve interpersonal goals. Social Skills Training : Social skills training often re fers to direct instruction of certain psychological skills which enable positive results in social situations (McGinnis & Goldstein, 1997).
SST in ASPs 10 CHAPTER TWO Literature Review Overview This chapter will present a review of the literature of afterschool programs (ASPs) and social skills training. First, the general characteristics of ASPs will be discussed and related to ASPs as an educational initiativ e. Next, the research regarding efficacy and quality of ASPs presented. Finally, social learning theory and psychological skills training will introduced. Social skills training, an interv ention extended from these theories, will then be explored. ASPs ASPs vary on several dimensions. A few of these dimensions include: Scope of activities offered: some ASPs focus on one particular activity (e.g., reading), whereas others offer a range of activities, including recreation, arts and crafts, fields trips, and so forth. (Fashola, 2002; McComb & Scott-Little, 2003, National Youth Development Information Center, 2003). Time in operation: some ASPs run for only a brief time after school, whereas some are open before school, until late in the evening, and on weekends (Fashola, 2002; McComb & Scott-Li ttle, 2003, National Youth Development Information Center, 2003). Staff: from as few as 1 to more than 50 (Fashola, 2002; McComb & ScottLittle, 2003, National Youth Development Information Center, 2003). Theoretical focus: youth development focus (views such as focusing on the positive, proactive, mobilizing the pub lic as well as all youth-serving
SST in ASPs 11 organizations in a community, viewi ng youth as resources); educational focus (academic improvement); other, specific skill focus (e.g., career planning, job placement) (Fashola, 2002; McComb & Scott-Little, 2003, National Youth Development Information Center, 2003). Cost of services: from free to w eekly fees exceeding $100 (Fashola, 2002; McComb & Scott-Little, 2003, National Youth Development Information Center, 2003). Location: school-based (e.g., public sc hool ASPs), private building (e.g., many Boys and Girls Clubs), public r ecreation centers, churches, etc. (Fashola, 2002; McComb & Scott-Li ttle, 2003, National Youth Development Information Center, 2003). ASPs in the Educational Context Schools in Florida are being held increasingly more accountable for educating their students (FDOE, 2004). First, educational expectations are increasing, especially in light of the No Child Left Behind (NCLB) Legislation (US Department of Education, 2004). Second, state and federal departments of education are mandating that schools collect more data to demonstrate that they are meeting these exp ectations (FDOE, 2004; US Department of Education, 2004). As a result of this increase in educ ational accountabilit y, schools are being provided with a variety of evidence-base d ideas and resources proven to increase educational performance (FDOE, 2004). One variable that has consistently been demonstrated to improve student performance is the amount of academic engaged time (AET) schools offer students (DiPerna, Vo lpe, & Elliot, 2001). Many schools have
SST in ASPs 12 shifted their schedules to include more time de voted to core academic areas, with less time being allotted to non-instructional time (SDHC, 2003). In addition, many schools have begun to view afterschool hours as addi tional opportunities to pr ovide this increased academic time (Chung, 2000; Fashola, 2002; McComb & Scott-Little, 2003). Afterschool Programs as an Educational Tool As described before, ASPs vary widely in their characterist ics. A plethora of research also has suggested that ASPs vary widely in their effectiveness to improve the academic success of their participants. (Chung, 2000; Fashola, 2002; McComb & ScottLittle, 2003; US Department of Education, 2003). In light of higher standards of accountability, if schools are to use ASPs as a venue for additional AET, ASPs must be able to demonstrate that they both offer a dditional AET and can relate that additional AET to increased student performance. The literature supporting the efficacy of ASPs has been mixed (Barker, 1998; U.S. Department of Education, 2003). Th e 21st Century Community Learning Centers (CCLC) program, for example, funded thr ough the federal government, had a $1 billion budget in 2002. It funded about 7,500 ASPs in 1,400 communities. The programs Summary of First Year Findings concluded that reading test scores, as well as grades in most subjects, were not higher among program participants. In addition, participating ASPs did not increase students feelings of safety during after school hours. Fewer than 38% of middle school students reported that the ASP was a good place to get homework accomplished (US Department of Education, 2003 ) Possible reasons for the lack of effectiveness include insufficient accountabilit y, participation rates, level of academic
SST in ASPs 13 focus, and poor generalization of skills ta ught in after school pr ograms to the school context. The 21st CCLC federal funds are filtered th rough state level offices. Many of these state offices (and even some local level funding recipients) conduct their own evaluations of 21st CCLC programs. The Massachusetts Department of Education reports more positive results of thei r programs (Resnick, 2004). Mo re than 12,800 children were reportedly served statewide. More than 4,300 of these children were tested to measure the efficacy of the after-school programs. Re snick reports that 61-100% of the students tested demonstrated statistically signifi cant gains in readi ng, language, and math depending on the measures given. The Boys and Girls Club of Broward C ounty partnered with the School Board of Broward County to provide 304 children an opp ortunity to participate in the Youth Educational Success (YES) program (Albright, 2002). In its third year of operation, an evaluation was conducted indicating an improve ment in school attendance with children who attended the YES program at least 50% of the time. Children in a comparison group exhibited a decline in school attendance. However, there were no significant gains in reading scores, math scores, or a NRT test used in relation to the comparison group. This evaluation also included a program-school co mmunication component in which club staff were expected to solicit a certain number of forms used for communication as to academic needs. None of the clubs met this minimum requirement, with two clubs not collecting any forms. The San Francisco Beacon Initiative, m odeled after the New York Beacons, was begun in 1994 with the mission of transf orming urban schools in low-income
SST in ASPs 14 neighborhoods into centers of community action, including before and after school programs. Each site was well funded, with c onsistent staff and or ganizational structure common among all five of the initial program s. An evaluation by Walker and Arbreton (2004) compared participants and non-partic ipants on school grades and scores obtained from the Stanford Achievement Test Nint h Edition (SAT-9). Results demonstrated no differences between participants and non-part icipants. The authors suggest two reasons for these results. First, while there were academic components in the program, these components were not sufficiently rigorous, cons istently primarily of homework help and general educational activities. Second, while attendance in the program was consistent throughout the study, children did not attend sufficiently on a da ily basis to benefit from the programs results. These hypotheses have been supported elsewher e in the literature as significantly influencing the educationa l outcomes in other after-school programs. Wahlstrom, Sheldon, Anderson, and Zorka (2001) studied a specific 21st CCLC Project in Minnesota. The program target ed both students with low academic achievement and students who exhibited at least one of several risk factors (e.g., poverty). The project was implemented across ei ght school sites in St. Paul, MN. Overall, no gains in reading or math test scores (using the Metropolitan Achievement Test 7th Ed.) among program participants were presen t when compared with children who did not participate. In addition, there was no differen ce between participants and non-participants in their school grades. School attendance for both groups was high no improvement was achieved for either group. Participants did experience a maintenance of school behavioral referral rates comp ared with non-participants.
SST in ASPs 15 Other studies include more supportive results. For exam ple, a 10-year study of LAs BEST after school enrichment program examined program participants in the second through fifth grades during the 1993-1994 school year, and followed them through the 1997-1998 school year. These student s were compared with non-participant peers using the Comprehensive Test of Basic Skills or the Stanford-9 Achievement Test in reading, mathematics, and language arts. The au thors reported that hi gher participation in the ASP was correlated with hi gher scores on these measures The authors also reported that greater program particip ation was correlated with hi gher attendance rates at school (Huang, Gribbons, Sung Kim, Lee, & Baker, 2000). Bissell, Dugan, Ford-Johnson, Jones, As hurst, J. (2002) evaluated the YS-CARE after school program in California, implement ed in 28 elementary schools. There were 567 YS-CARE children who par ticipated in the study, as well as 350 matched control participants who did not participate. Depe ndent measures included the SAT-9 Reading and Math and a local reading measure. When compared to national norms on the SAT-9 Reading and local norms on the local reading measure, program participants made statistically significant improvements. Howe ver, while there were gains on these measures when compared with the control group, these gains were not statistically significant. Baker and Witt (1996) investigated two after school programs in Austin Texas using both academic and behavioral depende nt measures. The researchers included both after school particip ants and control group of child ren who did not participate. Participants in the after school programs had statistically significan tly higher post-test scores in math, science, reading, and la nguage. The degree of participation in the
SST in ASPs 16 programs was included as variable, and was st atistically significantly correlated with greater academic improvement than those chil dren who participated less in the program. On the behavioral measures (Behavior Rating Profile Second Edition), there were no differences between participants and non-partic ipants. A measure of self-esteem indicated significantly higher scores for participants. Finally, both teachers a nd parents rated the program as effective, with many parents (80 %) remarking that they would enroll their children again the program. The Transition to Success Pilot Project in Boston, MA was conducted to investigate the effectiveness of remedial tuto ring along with other activities in one of six Boston after-school programs (Massachuset ts 2020, 2004). There were 116 students who participated in the after-s chool program. The researchers included a control group of Boston Public Schools students. Results indicate d that those particip ating in the program were statistically significantly more likely to move on to the next grade than those in the control group. In addition, younge r participants, and childre n who demonstrated higher attendance rates in the program, were more lik ely to show increased standardized test scores. A variety of survey and qualitativ e measures evidenced the programs validity with its constituency, with many parents and teachers finding the program to be successful in areas such as math, readi ng, and organizational skills. In addition, participants reported increased effort at school, carrying over from their experience with the after school program. The TASC program in New York City has been funded since 1991 and serves predominantly low-income, minority individuals (Welsch, Russell, Williams, Reisner, & White, 2002). When compared with non-particip ants, children in these programs showed
SST in ASPs 17 statistically significant increases in achiev ement in only select areas. Children who participated in the program for at least one year showed statistica lly significant gains in math. Children who participated for at le ast two years did no t show substantive achievement increases, nor did children who participated for three years. There seems to be a ceiling effect, then, on the effects of particip ation in this afte r-school program. No differences in the reading abilities of pa rticipants and non-par ticipants were found. Nance, Moore, Lewis (2000) conducted a study with seven elementary schools participating in a 21st CCLC grant in St. Louis. There we re three general intervention foci of the after school programs: (1) academic tu toring, (2) recreational activities, and (3) social/behavioral issu es. The authors analyzed pre and post math exam scores from 278 participants. Results indicated a statistically significant increase in math scores with students participating in the program. Barker (1998) examined the effects of a highly specialized ASP for juvenile delinquents. Compared with a control group, th ere were more than 50% fewer criminal convictions among program participants. On a self-report measure of aggressive behavior, self-discipline, and social control, participants evidenced improvement in selfcontrol. Similarly, Lamare (1997) cited teacher s reports that ASP participants had more positive social skills. The Foundations After-School Enrichme nt Program operated in 19 different schools in three different stat es (Klein & Bolus, 2002). St aff planned activities that promoted academic, physical, and emotional de velopment. Staff ratios were low (10:1) and all teachers and coordinators had college degrees. The researchers used the CTB/McGraw-Hill Terra Nova reading/language arts and mathematics tests both at the
SST in ASPs 18 beginning of the school year and at the end of the school year. Results indicated that posttest scores were statistically significantly high er for those participati ng in the after-school program. However, the post-test scores were only slightly above what should have been expected compared to the national normativ e sample. These results suggest that while this after-school program did not lead to significant gains wh en compared with a national normative sample, it at least did help participants achieve normal academic progress. Blanton, Mooreman, and Zimmerman (year unknown) studied the effects of an alternative academic improvement program in an afterschool program component called the Fifth Dimension. The goal of this program was to increase academic ability by offering activities that promoted childrens development as active learners in their environment. There were 52 children who pa rticipated in the study. The sample was divided into two for a experimental and cont rol group. Statistical analysis showed no pretest score differences on a measure of fo llowing written directions, but there was a statistically significant post-test difference on the same measure. While no reliability or validity estimates are given for this measure, the results do indicate an increased ability of program participants at le ast to perform better on the sp ecific dependent measure of the evaluation. In a follow-up study, Blanton, Moorman, Hayes & Warner (1997) investigated the effects of participation in the Fifth Dimension on scores on the North Carolina End-of-Grade Test, published by the North Carolin a Department of Public Instruction. Fifth Dimension pa rticipants demonstrated stat istically significantly higher performance on the test than the control group. Social Skills Training
SST in ASPs 19 There are numerous ways to increase A cademic Engaged Time. One way is to improve social behavior in school, including such behavior s as self-control, ignoring distractions, paying attention, and following directions (DiP erna, Volpe, & Elliot, 2001). The quality and amount of academic engaged time (AET), then, is influenced by these various academic and social behaviors, includi ng social skills (DiP erna, Volpe, & Elliot, 2001). Bandura (1977) described the processes by which individuals observe models of particular behaviors and vicar iously learn these behaviors. Applied to social skills, children often grow up obser ving adults and peers in their natural environment demonstrating various social skills. Th rough observation of the performance and reinforcement of these behaviors by others, as well as imitation of these models, children learn many necessary social skills (Bandura, 1977). Societal changes, including social mobili ty and the changing role of the church, have made it less likely that social skills wi ll be taught in a childs natural environment (home, church, neighborhood, etc.). Children also spend an increasing amount of time engaged in activities, such as watching tele vision, that may not model desirable social skills. Given these and other factors, the explic it instruction of social skills has, in many cases, become a necessity (Cartledge & Milburn, 1986). The roots of social skills training can be found in the psycho logical skill training movement in the 1970s (Goldstein, 1981). Th e psychological skill training movement developed as a response to the lack of a ttention to teaching behaviors. All other psychological paradigms of the time had as a basic assumption the inclusion of psychological skills in the individuals beha vioral repertoire. The psychological skills
SST in ASPs 20 movement, on the other hand, took the perspectiv e that individuals may not have certain behaviors in their re pertoire. The focus of interven tions, then, was to teach these behaviors. Using the theoretical background of ps ychological skill training, McGinnis and Goldstein (1997) describe a process for explici tly teaching social skills. They make use of the principles of social learning theory (Bandura, 1997), yet make the process more explicit. Each social skill is taught in a se parate lesson, and includes several components. The skill is first introduced, explained, and di scussed. Then, the skills is modeled by an adult facilitator. Next, child ren get the opportunity to role play these skills. Finally, students receive performance f eedback regarding their role plays (McGinnis & Goldstein, 1997). In addition to the explicit instruction of social skills, McGinnis and Goldstein (1997) also discuss ways in which prosocial behaviors can be generalized to times and settings outside the training sessions. Reinforcement schedule s for the performance of the appropriate skills, antecedent stimuli promp ting the performance of social skills, and review sessions are examples of some of th e strategies used to promote generalization. Social Skills Trai ning Effectiveness There are numerous studies which su pport the effectiveness of SST across a variety of conditions. Only one article was f ound describing the effects of an after school program on social skills training (Riley, 1994) This evaluation did not empirically measure social skills, nor did it include a fo rmal social skills training program. It did, however, qualitatively investig ate parent and teacher percep tions as to kids prosocial
SST in ASPs 21 interactions with each other. Overall, the evaluation includes positive remarks, indicating positive impacts on childrens social skills. Other studies, while not in an after sc hool program setting, have more directly supported social skills tr aining programs. Ang and Hughes (2001) conducted a metaanalysis of 38 studies involving SST with ch ildren described as antisocial. An overall effect size of .62 was found, indicating a sta tistically significant advantage to those receiving SST. The researchers also ex amined the difference between SST groups composed of all children labeled as devi ant and groups composed of both deviant and model peers. A statistically significant difference was f ound between the groups, with mixed groups yielding an average effect si ze of .15 higher than all-deviant groups. In addition, follow-up data suggest that the effects of SST on mixed groups maintained and generalized more than with the all-deviant groups. With all groups, though, SST was shown to be an effective intervention. Beelmann, Pfingsten, and Losel (1994) conducted a meta-analysis of social competence training, involving SST. Overall, th ey found social competence training to be an effective intervention, at least in the short term. However, generalization and maintenance components of the studies include d did not support long -term effectiveness. In addition, effects were more pronounced fo r specific outcome variables (e.g., direct observation of the target skills) as opposed to general outcome variables (e.g., measures of broad concepts such as social competence). Social Skills Training with Upper El ementary School Students (Grades 3-5) SST has been conducted with all age gr oups, from very young children to adults. Numerous studies, however, have concentrat ed on elementary-aged students. Berler,
SST in ASPs 22 Gross, and Drabman (1982) studied the effects of social skills trai ning on three children ages 8 10 who were labeled as learning disabled. Sessions began with an initial explanation of the social skill. Children were presented with a variety of scenes to role play. Instructors and peers provided corrective feedback immediately after each role play. Several role plays were videotaped and im mediately shown to the students along with feedback. Assertive skills of eye contact and appropriate verbal cont ent in reference to certain social situations were the targeted social skills. Dependent measures (role plays, behavi oral observations, an d peer sociometric ratings) indicated that the part icipants rates of behavior improved significantly during the course of treatment when measured in the analogue treatment settings. However, the researchers failed to find a ge neralization of the behaviors to other treatment settings. Results also failed to maintain during the maintenance and follow-up data collection periods. The authors mention the lack of treatment integrity by teachers in providing prompts and feedback to children in the na tural settings. These results demonstrate the importance of programming for generalization and ensuring treatment integrity when conducting social skills training. Elardo and Caldwell (1979) conducted SST with 34 students in grades four and five in an inner-city elementary school. Homeroom teachers led discussions around a variety of topics related to social skills such as childrens understanding of other childrens thoughts and feelings. Teachers presented various so cial situations and asked participants to discuss potential ways of handling the problems presented in the situations. Role plays were frequently conduc ted, although the resear chers did not specify
SST in ASPs 23 how many times. Participants receiving the inte rvention showed more gains in dependent measures of social skills th an those in the control group. Yu, Harris, Solovitz, and Franklin ( 1986) conducted a social problem-solving training with 35 boys ages seven to 12. Social problem-solving training is a category of interventions designed to improve an indi viduals cognitive skill s in processing and generating solutions to a vari ety of social problems. Soci al problem-solving training is often a component of social skills traini ng, but can also be conducted separately. The purpose of this particular training was to bot h decrease antisocial behaviors and increase corresponding prosocial behaviors. Activitie s involved in the pr ogram included role playing, group discussion, a nd other supplemental activitie s. Children were presented with topics such as understanding feelings recognizing problems, generating problem solutions, and implementing problem solutions. Parents were encouraged to use aspects of the program at home. The researchers f ound an improvement in social-cognitive skills, a reduction in the targeted behavior problems, and an increase in selected prosocial skills compared with the control group. Bierman and Furman (1984) investigated three different types of social skills training using peer acceptance as th e dependent variable. Fifty-six 5th and 6th grade students participated in the study. The fi rst type involved eac h child in the group receiving individual coaching a bout one of three target conve rsational skill areas. Over the course of training, participants in this group role played and received feedback on their performance. The second type, focu sing on the same skills, involved a group of three students. These children also engaged in the same activities as the first group, but received no feedback, coaching, or reinforcement for any par ticular skill. A third type
SST in ASPs 24 involved children working in groups, but also receiving coach ing, feedback, and reinforcement. Results suggest that simply spending time engaging with peers does not produce meaningful changes in prosocial competenci es. Both groups receiving coaching showed increases in prosocial skills that were sustained at follow-up (six weeks). The intervention involving children meeting with the group but receiving no coaching showed temporary improvements in status and social interaction, but no increase in prosocial skills like the childre n receiving coaching. Parent ratings were included in a st udy by Pepler, King, Craig, Byrd, and Bream (1995). Teachers nominated students they perceived as aggressive. Seventy-four aggressive children (63 boys) received direct so cial skills training in nine specific skills. Teacher ratings upon conclusion of the interv ention revealed a statistically significant difference between those receiving interven tion and those who did not. These results maintained over a nine-month period, although th ey weakened over time. Parent and peer ratings, however, showed little difference be tween the intervention and control groups. Nelson and Carson (1988) conducted two studi es in which they used a variety of techniques, including self-monitoring, disc ussion, paired work on assignments, and games, to teach social problem-solving skills. Participants were thirdand fourth-grade children. Outcome measures fell into two categ ories: (1) measurement of social problemsolving skills, and (2) measurement of behavior al skills thought to be related to social problem-solving skill. Results in both studies revealed that participants receiving the interventions did make gains in social-problem solving abilities. However, participants in both studies showed no gains in teacher rati ngs of behavioral outcome variables. The
SST in ASPs 25 acquisition of greater so cial-problem solving skills alon e was not found to be related to changes in classroom behavior as measured by teachers. These results were reported as consistent with previous studies. The author s conclude that increased social-problem solving skills do not increase social skills. However, it is possible that social-problem solving skills may be a helpful or even n ecessary, but not sufficient, component of an intervention package that would effectively increase social problem skills. The effects of positive reinforcement a nd punishment on social skill acquisition and problem behaviors were investigated by Bierman, Miller, and Stabb (1987). Children in grades 1-3 were randomly assigned to one of four conditions: (1) positive reinforcement for targeted prosocial skills, (2 ) prohibition with a response-cost system for targeted antisocial behaviors, (3) a combina tion of the first two, and (4) a control group. The same activities promoting prosocial peer interaction across thr ee different categories of skills (questioning others, helping, and sharing) were used in all intervention groups. Participants in the positive reinforcement group were instructed in the targeted prosocial skills, as well as reinforced for these behaviors using a token economy. In the responsecost group, only general reinforcement for positive peer interaction was given. In addition, participants lost the ability to earn this general rein forcement if certain antisocial behaviors were observed. The combination of the reinforcement with the response-cost was indicated as the most effective intervention. The response-co st intervention alone produced immediate declines in antisocial behavior. Howeve r, the effects did not maintain, and no corresponding increases in prosocial behavior were observed. The positive reinforcement intervention alone produced sustained increase in prosocial behavior but these behaviors
SST in ASPs 26 did not stabilize until after the follow -up observations. The combination of both interventions resulted in the addi tive benefits of each intervention. Tanner and Holliman (1988) conducted asse rtiveness SST in order to increase cooperative behaviors and decr ease aggressive behaviors. Participants included 24 secondand third-grade stude nts. Behavioral observati ons were conducted across a variety of settings. Teacher ra ting forms were also administered. Results moderately support the effectiveness of the program, with some data to suggest that cooperative behaviors increased and aggr essive behaviors decreased. Social Skills Training in Small Groups The number of participants receiving social skills tr aining at one time varies by study. One on end, some researchers have conduc ted social skills training with one child at a time (Cooke & Apolloni, 1976). On the other end, others have instructed entire classrooms during the same session (Nelson & Carson, 1998). The number of participants receiving the intervention at one time affect s a variety of variables in a study. The small group is a frequently used condition in the literature. For example, in a study conducted by Vaughn & Lancelotta (1990), the effects of pairing low-status peers with high-status peers was investig ated. This study originated from the theory that simply increasing pros ocial behaviors is not sufficient in improving the acceptance rates of low-stat us peers. Results suggest that the type of interpersonal interaction provided to participants did not im prove the status of pa rticipants receiving the intervention. The researchers point to already high gains in outcome variables attributable to other intervention component s. In addition, the authors hypothesize that the nature of the interactions between th e low-status and highstatus peers was not
SST in ASPs 27 effective. They suggest that different, and mo re intense, types of interaction may produce increases in status. By conducting social skills training in using the small group condition, the researchers provided children with this opportunity for more intense interpersonal interaction. This type of inte raction would likely not have been available in either a training condition of one student or an entire classroom of students. Coats (1979) employed a cognitive self-ins truction technique with sixteen thirdgrade boys. A small group condition was also us ed in this study (four students per group). Participants were taught to ve rbally guide themselves through a variety of tasks. First, simple, non-social tasks were the target of intervention. In the final phase, the selfinstruction skills were applied to social situ ations. Verbal self-instruction was overt at first, then faded to covert. The researchers discovered that the intervention was successful in decreasing deviant behavior s, but was less successful in increasing cert ain prosocial behavior. Thus, increased self-control by the participants was achieved. Vitaro and Tremblay (1994) held social skil ls training sessions as part of a larger prevention package also offering parent tr aining programs. Participants were 46 boys (ages 8-9). The social skills groups were comp rised of 4-6 boys per group. For the parent training component, parents were trained in a variety of basic pa renting techniques, ranging from appropriate use of time-out to using a problem-solving process during family conflict. Children participated in social skills training groups at their school. During the first year, nine prosocial skill were selected as the targets for intervention. During the second year, self-control and pr oblem-solving skills became the targets.
SST in ASPs 28 The researchers collected data for three ye ars. Those participating in the program showed significant gains at th e end of the data collection pe riod. Specifically, participants in the intervention group were rated as less aggressive by teachers (who were blind to research group status). Additionally, participan ts associated with less deviant peers, as measured by self-reports of delinquent conduct by participants self -identified friends. Small groups were also used by Dubow, Huesmann, and Eron (1987). Participants receiving intervention in this study received cognitive social-probl em solving training, behavioral social skills tr aining, or a combination of th e two. In addition, there was a control group which received the same amount of attention from adults and peers, and participated in similar activities. The most effective intervention type was the combination of behavioral social skills training with social problem-solving training. However, these results did not maintain at the six-month follow-up. The researchers point to the need for more long-term programs a nd support for generali zation. Finally, unlike most studies, the control group not only improved on measures of prosocial competencies, but demonstrated sustained effects at the six-month follow-up. Dubow, Huesmann, and Eron suggest that children w ith antisocial behavioral patterns may respond more favorably to free play setti ngs rather than in structional settings. Social Skills Training Usi ng Modeling and Role Playing The literature describes many techniques for teaching social skills. Modeling and role playing are two specific st rategies, frequently used in combination, to teach social skills. The use of modeling stems from soci al learning theory, pos iting that children can learn behaviors vicariously (Bandura, 1977). The use of role playing follows from a behavioral, direct instructi on approach emphasizing repetitiv e practice of behaviors with
SST in ASPs 29 explicit instruction (Cooper, Heron, & Hewar d, 1987). Many studies also make use of additional strategies such as board games, contingent reinforcement, or group discussion. However, modeling and role playing ar e the central and most commonly used components in many studies evaluated. Kendall and Zupan (1981) used modeling and role playing along with a selfinstruction and response-cost procedure. They gave the same interventions to children in two different groups. One group consisted of children receiving th e intervention in a small group format. The other group involve d children receivi ng the intervention individually. A third nonspecific treatment group, originally create d as a control group (but receiving some level of treatment), wa s also established. Sessions followed a specific format, teaching and fading different psyc hoeducational tasks de pending on the session. While all three groups evidenced improvement the treatment groups, who received the modeling and role playing techniques (in ad dition to the self-instruction and responsecost procedures), showed stat istically significantly higher sc ores on certain measures of generalization. Forty-one children, ages seven through 12, were the participants in a study by Kettlewell and Kausch (1983). Children who displayed aggressive behavior during the first week of camp were selected as partic ipants. Four weeks of treatment, including modeling and role playing, as well as a self -instruction technique, were used with the children in small groups at a summer day cam p. Children also discussed situations in which they had applied the learned techniques in between sessions. The 12 outcome measures (ranging from self-report forms to analogue measures of behavioral responses requiring self-cont rol), when examined together, support the
SST in ASPs 30 effectiveness of the social skills training. However, unifo rm performance differences between the intervention and c ontrol groups were not observed. Several of the measures had fatal limitations, and othe rs indicated no difference betw een the two groups. Overall, however, the data supports th e effectiveness of modeli ng, role playing, and selfinstruction in improving the beha vior of aggressive children. Lochman, Lampron, Gemmer, Harris, and Wyckoff (1989) investigated the use of a package of interventions including modelin g and role playing with 32 boys (average age=11.0). The intervention components were co ntingent reinforcement of rules, selfinstruction, social problem-s olving training, activities enco uraging social perspectivetaking, videotaped models, discussion, and ro le play. Another treatment group received these interventions as well as teacher consul tation. This intervention component involved six hours of training in over th e course of four to six mee tings. Each meeting had several teachers, and was facilitated by the social skills training leaders. Behavioral observations we re conducted to measure offand on-task behavior, specifying for the type of off-task behavior In addition, teachers completed rating forms relating to levels of aggression of the participants. Results indicated that the social skills treatments improved scores in all depe ndent measures. However, there was no statistically significant difference between the group receiving additional teacher consultation and the group that did not re ceive this added component. The authors concluded that the specific type of consulta tion (dialoguing and pr oblem-solving training) was ineffective. It is also possible that the quality of the social sk ills component made the additional teacher consultation unnecessary.
SST in ASPs 31 Modeling, role playing, and feedback we re the central components in a social skills study by Mize and Ladd (1990). Eighteen child ren were taught four targeted social skills across an eight-week period. Hand puppets were used as models. Following the models, children were asked to use the models to role pla y. Then, the participants role played the targeted behavior with their part ner. These role plays were videotaped, and were then shown to the partic ipants with corrective feedb ack. To promote generalization, both the instructors and other ad ults spent time in the classrooms with the participants encouraging the use of the skills in the natural environment. The researchers were interested in meas uring both the ability of participants to verbalize social-problem solving skills, as well participants skill-related behavior in the natural environment. Results di d not support the ability of chil dren to verbally relate the content of the social skills training. The aut hors suggest that these skills might not have been operationally defined. In addition, the auth ors hypothesize that th e participants were not effectively trained in this skill. Child ren did, however, significantly increase the number of targeted prosocial skills perf ormed in their natural classroom setting, suggesting that the interv entions were effective. Spence and Spence (1980) investigated th e use of modeling and role playing, in combination with instructions, discussi on, social reinforcement, and homework. Participants included 44 adolescent males who were randomly assi gned to the social skills training. There were 12 targeted skills, selected based on deficiencies found during initial assessment. In addition to this inte rvention group, there was an attention-placebo group as well as a control group. The attention-placebo group received the same amount
SST in ASPs 32 of time and attention as the intervention gr oup, but no social skills training. The control group received nothing. Only measures of locus of control an d self-esteem were used as dependent measures. There were no behavior observations or any other direct measure of the skills that were taught during training. Dependent measures indicated that increases in locus of control were statistically si gnificantly higher in the intervention group than in both the attention and control groups. Ho wever, participants in both the intervention group and the attention group improved on the self-esteem measures, indicating that social skills training was not necessary to improve childre ns self-esteem. Both dependent measures, though, demonstrated a lack of maintenance of any effects for any group. In fact, losses were noted in some case. If the researchers had included dependent m easures that directly measured the dimensions of the target beha viors, more differen ces between the groups might have been noticed. Social Skills Training with Urban, At-Risk, or Minority Children Many social skills training studies analy zed included primarily participants from majority, middle-income, or suburban populatio ns. The results of these studies do not necessarily generalize to child ren from difference geographical, socioeconomic, or racial backgrounds. There were several studies f ound, however, which did include at least a substantial number of participants with one or more of these background characteristics. Huey and Rank (1984) studied the effect s of an assertiveness training program with 48 Black males designated as aggressive Participants were put into two groups one receiving the intervention from a prof essional counselor, the other receiving the
SST in ASPs 33 intervention from a peer counselor. Both groups of counselors received the exact training, and were told to follow the treatment regimen closely. Results indicate that the assertiveness tr aining was effective with the participants. However, there was no statistically signifi cant distinction between those receiving the intervention from the peer counselors a nd those receiving it from the professional counselors. There are two possibl e indications of these results. The first is the strength of the intervention was in the design of the assertiveness training materials, not in the skills of those presenting the materials. The second pos sible indication is that the skills of the counselors were important, but that the training afforded to the peer counselors was sufficient to instill these skills. Fifty-three percent of the sample in Lochman, Burch, Curry, and Lampron (1984) was African-American. Anger-coping and goal-se tting were the two specific social skill sets defined as intervention components. Four groups were created one each for angercoping and goal-setting, one co mbining the two, and a fourth receiving no intervention. There were 76 boys aged 9-12 involved in the study. Children in the anger-coping group demonstrated significant reducti ons in aggressive off-task and disruptive behaviors. The addition of the goal-setting piece resulted in greater transfer of the results to the classroom setting. Teacher and parent percep tions of the participants did not change, however, despite meaningful decreases in the presence of behavior. The authors point to the rigidity of perceptions of aggressive boys, and the rela tive lack of integration of teachers and parents into the treatment components. Eighteen students from an inner-city elementary school were randomly assigned to a cognitive restructuring, response cost, or control condition in Forman (1980). The
SST in ASPs 34 purpose of the interventions was to decrease aggressive behaviors. The cognitive restructuring component involved identifying th oughts and feelings th at lead to anger arousal, learning how people control their though ts and feelings, and learning scripts to help calm oneself down. The response cost proc edure involved basic loss of privileges for aggressive behaviors. Dependent measures included behavior al observations, teacher ratings, and teacher behavioral records of aggression. Both the cognitive restructuring piece and the response cost procedure were shown to be e ffective in reducing aggr essive behavior. The response cost procedure was found to be slig htly more effective, however. The authors suggest that the greater involvement of teach ers in the response cost procedure might have influenced their responses and observations recorded in the teach er rating scales and records of aggressive behavior. Another po ssible reason for the greater effect of the response cost procedure could be a stronger generalization component involved in the procedure. La Greca and Santogrossi (1980) conducted a study with thirty children (15 males and 15 females) in grades 3-5. Each particip ant was assigned to one of three groups a social skills group, an attention-placebo group, or a control group. Participants were selected based on low peer ratings of who the peers would like to both play with and work with. Children in both the social skills a nd placebo groups met for 90 minutes after school once per week for four weeks. Child ren in both groups were given identical introductions in terms of what the progr am would contain. The social skills group consisted on instruction of eight social ski lls selected from a literature review by the
SST in ASPs 35 researchers. Treatment procedur es were also selected based on a literature review, and included modeling, coaching, and behavioral re hearsal with videot aped feedback. All procedures were used in each session. Each session began with participants viewing a video of children performing the social sk ills and encountering positive consequences. A variety of situations, with different models, were s hown. Following the videotapes, children discussed what behaviors were observe d, the importance of these behaviors, and how the participants could engage in the beha viors in their daily lives. Children then had an opportunity to role play each skill, combined with coaching from the trainers. Situations for role plays were based on part icipant suggestions. These role plays were videotaped. Participants were immediately shown these tape s and given both corrective and positive feedback. Each session ended with a homework assignment, which was reviewed at the beginnin g of the next meeting. The placebo group engaged in similar activities, yet unrelated to social skills. They viewed unrelated video segments, role played charades-type scenarios, and given homework assignments that were not peer-o riented. The control group only participated in pre and post assessments. There were four pre and post measures Behavior observations of two sample behaviors were conducted. A sociometric m easure was administered. Children were asked to assess a videotape of peers engaging in situations (verbally described situations, and remark what they would do in a similar situation). Finally, ch ildren were rated on role plays. All observations were quantifie d, and various analyses were run. Several ANOVAs indicated a main effect for treatm ent when analyzing both the participants verbal responses to the videotapes as well as
SST in ASPs 36 their reconstruction of the skills during role play. Other analyses of the data indicated similar results (a main effect for treatment) with the exception of the sociometric ratings. Finally, no differences were found between th e attention-placebo group and the control group, indicating that simply increasing the amount of social in teraction opportunities alone for students with social skills deficits may not be enough to improve social skills. This calls into question the practice of mainst reaming students with social skills deficits without providing sufficient social skills instruction. Gresham and Nagle (1980) conducted a study with 40 third and fourth grade students from a predominantly middle-cla ss school. Similar to the previous study, participants were selected based on ratings by same-sex peers on measures of the degree to which children preferred to work or play with each other. Part icipants were randomly assigned to one of four groups: modeling, co aching, modeling and coaching, and control. Those in the modeling group were shown vi deotaped models for six sessions over a three week period. Videotapes involved peer models of certain social skills with a narration by an adult female. Sessions were conducted in dyad s or triads, with participants having different partners each se ssion. Each session lasted approximately 20 minutes. Participants in the coaching group receiv ed instruction in the same social skills as the modeling group. Each coaching session in volved three component s: presentation of steps and standards for each behavior; behavi oral rehearsal; and feedback on performance as well as discussion and suggestions. The mi xed group received abbr eviated versions of both modeling and coaching. The control group wa tched an unrelated video for the same amount of time as each of the experimental gr oups were engaged in social skills training.
SST in ASPs 37 Dependent measures included behavior al observations, peer nominations, and peer ratings. The behaviors selected for obs ervation were initiati ng and receiving both positive and negative interactions (four in tota l). These behaviors were selected based on literature suggesting to the researchers that th e behaviors are highly co rrelated with social acceptance. Peer nominations and ratings i nvolved peers listing and rating peers on the degree to which they would like to work with or play with them. Results suggest that there was a main effect for treatment (as opposed to no treatment), but not for any one group placem ent. The combination of modeling and coaching, therefore, did not produce more bene fits than either of the two conditions alone. In addition, peer ratings of wanting to play with group particip ants increased, but not ratings of wanting to work with group participants. In a larger study, Weissberg et al. (1981) studied the e ffects of a 52-lesson, classtaught social problem solving training pr ogram. Participants were 243 third grade students in both urban and suburban settings Participants were split between receiving intervention (n=122) and control (n=121). Th e treatment groups were comparable on several demographic variables, including sex and race. The lessons came from a common book writt en by the researchers, with explicit descriptions of 52 lessons of 20-30 minutes in length. There we re five major categories of skills: recognizing feelings; problem identifi cation; generation of alternative solutions; consideration of consequences ; and integration of problem solving behaviors. Activities included small group role play, videotap ed models, cartoon workbooks, competitive games, and class discussion.
SST in ASPs 38 Dependent measures included four problem-solving assessments given to children, behavior observations during an analogue situation, and participant interviews (measuring their ability to describe the so cial problem-solving process). In addition, teacher rating scales, peer rating scales and self report forms were used. Those participating in the training showed gains across several skills, but not all skills. These results were mediated by the stat us of participants as urban or suburban. In both groups, participants gained skills. Howeve r, teacher ratings indicated a decline in social problem solving skills with the urban participants. The researchers offer several explanations, ranging from teacher apprehensi on of the program to unusually high pretest ratings by a teacher. Finally, there were no linkages found between social problem solving skills and behavioral outcomes as measured by rating scales, indicating that social-problem skills training alone may not be enough to modify classroom behavior. Overall, results varied highly depending on th e specifics of the participants in each classroom. Further research, as well as mo re narrowly tailored programs, are likely necessary supplements to the interventions in this study. Cooke and Apolloni (1976) used a smalle r sample (n=4) of children in a withinsubject multiple baseline across treatment components design. Students came from a small class of children labeled as learning di sabled. The four children with the lowest levels of the target social skills were se lected as participants. The remaining three children were used during asse ssment following each session. There were various stages of sessions. First, five days of baseline data were collected. Then, instruction, modeling, and soci al praise were used to increase one behavior of the four selected. The next phase involved using the same techniques to
SST in ASPs 39 increase both the first behavior and an adde d second behavior. In each subsequent phase a new behavior was added. Immediately fo llowing each session, participants were brought into a room with their three othe r classmates in a free-play condition. The instructor left the room and onl y the data collectors remained. Behavior observations were conducted using a whole interv al momentary timesampling technique. Interobser ver agreement was conducted to ensure at least 85% agreement. Results suggest that the social sk ills training was effective for three of the behaviors (smiling, sharing, and positive physical contact), but not for the fourth behavior (verbal compliments). The authors hypothesize that verbal compliments may be unlikely, even given explicit inst ruction and praise, in a setting free of adult contingencies. The increases in the three behaviors were main tained during the four weeks of follow-up observations. In addition, behavioral obs ervations of the th ree children who did not receive instruction revealed increased levels of behaviors in a similar fashion to the four participants who received the intervention. Th is finding suggests that even children who do not receive social skills training may be nefit vicariously through those who do receive training. The researchers attrib ute this finding to the effect s of modeling and/or social reciprocity. In another multiple baseline across treatment component design, Bornstein, Bellack, and Hersen (1977) investigated the e ffects of social skills training with four children identified as not being assertive. Pa rticipants were selected on the basis of teacher ratings and behavioral observations. Pa rticipants ranged in age from eight to 11. Each session involved one participant a nd two adult models. There were three 15 30 minute sessions per week. A new behavi or was introduced each week. During each
SST in ASPs 40 session, various social scenario s requiring assertive respons es were narrated over an intercom by an observer in a nother room separated by a one-w ay mirror. The participant was then asked to respond to the social si tuation. This activity was followed by corrective feedback and modeling by the tw o adults. Then, the participan t engaged in the role play for a second time. This process continued until the outside observer determined that the participant had reached mast ery level for that skill. The dependent measure in this study was an observers rating of participant performance as a role player in an analogue setting similar to th e intervention setting previously described. The behaviors selected were ratio of eye contact to speech duration, loudness of speech, and requests for new behavior. In addition, separate observers rated each participants overall asse rtiveness using a five-point Likert scale. Interobserver agreement was conducted for each measure. Results were similar across all four participants. All three target behaviors were significantly increased. In addition, the global rating of assertiveness also increased as training sessions progressed. Two and four-week follow-up obs ervations suggested that these changes generalized. However, no obser vations were conducted in natural settings, decreasing the strength of the generalization of the results. Overall, though, these results add support for the use of social skills traini ng with both specific populations of children, and with specific behavioral concerns.
SST in ASPs 41 CHAPTER THREE Method Participants The sample consisted of four English-speaking male participants aged 7-8. Two were African-American and two were Hispan ic. Participants were selected based upon information provided by ASP staff. Participan ts attended the after-school program in a low socio-economic neighborhood. Setting This study was conducted at an after-schoo l program in an urban area in Tampa, Florida. The site served approximately 100 children, ages six to 14 years. The site conducted activities from 2:30 9:00 P.M. Activities ranged from educational programming to recreation, w ith the primary purpose being positive youth development in a safe, supportive environment. The student-staff ratio was approximately 25:1. There were approximately six staff on site every day. Educational backgr ound of staff varied from high school to college education. Some staff were empl oyed full time and some part time. The background of staff varied on other dimensions including ethnicity (several White staff, several Hispanic staff, and several African-A merican staff) and experience (two months to over ten years). Dependent Measures The dependent measures in this study c onsisted of behavior observations and a social skills rating instrument. Behavior Observations
SST in ASPs 42 Behavior observations of the social skills used by participants in the study during normal ASP activities were c onducted by the researcher and outside observers who had experience in behavior observations, and ha d been trained by the researcher. Behavior observations offered an opportunity for direct assessment of prosocial behaviors of the participants in the after-s chool setting. Observations were conducted prior to the beginning of the study in order to identify a ppropriate participants and again throughout the study on all four students selected as the final part icipants. Observations were conducted before intervention, three times per week during intervention, and once upon conclusion of all interventions. Three skills were chosen by the program staff as target beha viors for observation. Staff were asked to identify common behavi oral concerns. From these concerns, three undesirable behaviors and their subsequent replacement behaviors were identified. The target behaviors that were selected and their definitions in behavioral terms (so as to be observable and measurable), were as follows: Figure 1 Target Behaviors Raising Hand Before Leaving Seat The child raised his hand before l eaving assigned seat or seating area unless instructed otherwise by an adult in the room. Examples included raising ones hand to go to the bathroom and raising ones hand to get additional materials. A student did not need to raise his hand to stand up in assigned seating area. Sitting Properly In Seat The student had his buttocks on the seat completely, had his torso facing forward, had both legs facing fo rward, had all four legs of the chair on the floor, and had the chair squarely facing the table in front of him. Sitting in seat was not scored when the child was out of the seating area. Attending The child was looking at approved materials. The child may also have been looking at an adult w ho was speaking to him or his group. Examples of attending included l ooking at a book and looking at a teacher giving directions directly to that student. Non-examples included looking at a paper airpla ne, and looking at an adult who was talking to another student and not the target student.
SST in ASPs 43 The fact that these behaviors were be haviorally defined also increased the potential reliability of the obs ervations. In addition, the res earcher used a standardized event recording procedure and momentary time sampling procedure which may have increased the reliability of the observations A sample observation form is included in Appendix 2. Students from the University of South Fl orida were chosen as observers. Each of the students that was selected had been trai ned and received superv ision in the use of behavior observation methods and technology. The observers we re instructed in the use of the specific observation forms developed fo r this study. The selected target behaviors were described in detail. The researcher engaged the observers in a discussion of examples and non-examples of the target behaviors. The coding system was then explained. Once the researcher observed th at the observers understood the target behaviors and the coding system, modeling techniques were used to demonstrate the specific target behaviors, as well as to further demonstrate proper observation and recording procedures. Finally, the observers and the researcher ensured inter-observer agreement using the following procedure. Inter-observer agreement consisted of two data collectors observing a child for 20 minutes using the same procedure as regular observation described below. Inter-observer agreement was conducted both during training to ensure competency, and during the study to ensure on-going data collection in tegrity. There are no set guidelines on the frequency of conducting inter-observer agr eement during low-N st udies (Kazdin, 1982). Kazdin (1982) suggests that inter-observer agreement should be conducted at least once
SST in ASPs 44 per phase of the study. As such, during the study, inter-observer agreement was conducted for each individual once during most phases (including baselines and each of the three intervention phases). Then, the following formula was used to calculate interobserver agreement: Inter-observer agreement = Total Agreements x 100% Total Agreements + Total Disagreements The criterion for inter-observer agreement was set at 90%. If this percentage would have been lower either during training or during th e study, additional review would have been conducted with the observer. This review woul d have included a discussion of the target behaviors, examples, and non-examples. It w ould also have included a review of the coding process and use of the behavior obser vation instrument. Inter-observer agreement then would have been conducted again. Furthe r review would have continued until the observer and researcher reached the 90% in ter-observer agreement criterion. No data from the observer would have been used unt il the 90% criterion had been met. There were no instances when the criteria for inte r-observer agreement was not met. A review training was conducted with one observer, however, upon request by that individual. Observations involved an observer sitting in the back or side of the homework room, away from and not par ticipating in the main activit y. Each observer put only the initials of the participant to enhance conf identiality. When the child raised his hand before getting out of his seat, a mark was put in the appropriate space on the observation form. When the child did not raise his hand befo re getting out of his seat, a mark was put in the appropriate space on the observation form. When the child was attending, an a was marked in the appropriate interval box. When the child was not attending, no mark
SST in ASPs 45 was put in the interval box to indicate no at tending. When the child was sitting properly in his seat, an s was put in the appropriate interval box. Wh en the child was not sitting properly in his seat, an n was put in the appropriate interval box. A sample behavior observation form is included in Appendix 2. A behavior observation form was created specifically for use in this study. The observation forms utilized both a momentary time sampling and event recording procedure as described above. The same obser vation form was used throughout the entire study. Thus, those not involved in the intervention phase of the study had no knowledge of what phase the study was in, limiting any tendency an observer might have had to react to participants behavior on th e basis of the phase of the study. Selected Social Skills Rating Instrument Because only three behaviors were target ed in this study, a published rating scale was not used. Published rating scales, such as the Social Skills Rating System (SSRS) (Gresham & Elliott, 1990), include items across many behaviors. Composite scores on these scales, then, would not have been sens itive to changes in onl y a select number of behaviors. Instead of a published rati ng scale, each staff member at the ASP was given a rating scale comprised of items designed speci fically for this study. Multiple items were crafted for each behavior. These items were then presented in a random order using a Likert-type scaling procedure. Each staff member was asked to fill out the form as described in the participant selection secti on and instructions on the form (Appendix 1). Acopy of this rating scale is included in Appendix 1.
SST in ASPs 46 For all measures, there were no known et hical issues with the data collection methods. They were unobtrusive and required very little time of program staff. They required no time of study participants, a nd resulted in no known adverse outcomes for anyone involved. In regard to confidentialit y, no participant was identified or discussed with anyone except site staff, the participants parent(s ) or guardian(s), and the researcher. Design and Procedures A multiple baseline across behaviors was used. There were three baselines. Different social skills were in troduced at different times to e ach of the four children. In addition, the differential effects of two major co mponents of social skil ls training direct instruction and reinforcement on prosocial behaviors were explored using a multiple baseline format. Specifically, direct instruction and reinforcement were introduced in different phases for different behaviors to determine whether direct instruction, reinforcement, or both influenced prosocia l behavior of the participants the most. Experimental control was achieved for both th e direct instruction intervention and the direct instruction/reinforcement combinati on. Experimental control was not, however, achieved for the reinforcement intervention.
SST in ASPs 47 Figure 2. Example of Multiple Baseline Across Behaviors (with three behaviors) This study was implemented in a series of steps that made up the procedures. Step 1: Site Selection The researcher approached the Boys a nd Girls Club of Tampa Bay and requested permission to conduct this study at one of thei r sites. However, no sites were available from the the organization. As a result, th e researcher approached Hillsborough County Department of Parks, Recreation, and C onservation, who approved the study to be conducted at one of their recr eation facilities. The resear cher presented the following criteria for a site to be selected: urban at least four staff members daily attendance of at least 40 children at least 25 children who regularly attend are in grades 1-3 Behavior 1 Behavior 2 Behavior 3 Introduction of Behavior 1 Introduction of Behavior 2 Introduction of Behavior 3 S1 . S4 S1 . S4 S1 . S4
SST in ASPs 48 The site selected was then contacted, and permission was re quested for the study to be conducted at that site. The site administrator and personnel from Hillsborough County Department of Parks, Recreation, and Cons ervation granted permission for the study to take place at the proposed site. Step 2: Selection of Participants Participants were selected through severa l steps designed to identify children who exhibit specific social skills deficits. First, staff member s, as a group, identified three target social skills commonly deficient am ong children in their program. Staff then nominated eight English-speaking boys who they believed most lacked the target behaviors. Only English-speaking boys we re selected to eliminate the potentially confounding effects of language and gender in the study. Also, the age range of the participants was narrowed down to a 2-year range (ages 7-8-year-olds). The purpose of this criterion was to reduce the potentially confounding effects of age differences in the study. Second, once these eight boys were nomina ted, parental consent was obtained for each participant. Third, each staff member th en completed the rating scale included in Appendix 1 on each child. The six children with the most deficient social skills were identified. Fourth, three behavior observ ations were conducted on each of these six children. Fifth, the four children (of the six) who exhibited the greatest skill deficits during behavior observations were id entified as potential participants. Each potential participant was required to exhibit each desired target be havior 50% or less of the time when an antecedent of the behavior occured. If any of the four potential pa rticipants would not
SST in ASPs 49 have met this criterion, they would have b een replaced by the child from the previously identified group of six who had exhibited the ne xt lowest skill deficit, assuming that child would have met the criterion. If a total of f our children would not have been able to be identified, behavior observa tions would have been conduc ted on the other two children (of the original eight) who were not sele cted previously after the rating scale was administered. If neither of these children w ould have met the criterion, all staff would have selected three more children (in the sa me manner as was originally done) on which to complete rating scales and behavior observations. Step 3: Baseline Data and Staff Training Baseline data was collected across all three behaviors three times per week for more than two weeks and the behavior had stabilized. Stabilization of behavior was defined as a trend line which was flat or worsen ing in slope for at least 3 data points, as assessed by the researcher using visual anal ysis procedures. Data collected during the participant selection phase of the study also was used as the initial baseline data. Upon completion of the last baseline point before any intervention was implemented, staff training was conducted. Staff needed to be trained prior to intervention implementation so that they we re ready to give social skills prompts immediately following the first interventi on phase. However, staff training was coordinated so that it did not occur substa ntially in advance of the intervention. The purpose of this timing coordination was to ensu re that staff did not begin using prompts before the intervention began. Staff were also told not to use any of the prompts until the intervention began. The exact time of in tervention implementation was clearly communicated to staff.
SST in ASPs 50 Staff training consisted of the researcher giving program staff an overview of the research project, of each staff members role, of the nature and purpose of social skills training and its relevance in the present st udy. Each staff member was instructed on the importance of prompts in the implementation of social skills in natural settings. They were also instructed as to their responsibil ity in prompting a participant, as much as possible, when presented with potential stimuli of the target behavior. A prompt, in this study, was described as a verbal cue to the chil d to engage in social problem solving steps (e.g., Stop and think do you want to make a good choice or a bad choice?) that the child learned in SST along with the target behaviors. The researcher modeled the appropriate use of a prompt. The researcher then observed site staff practice giving prompts to ensure competency in this skill area. Step 4: SST Introductory Meeting Before the first SST session began, the ch ild participants were introduced to the researcher. The training (and all subsequent sessions) was conducted in a computer room with no other individuals present at the tim e of training. The researcher conducted the session, which lasted appr oximately 30 minutes. During the introductory meeting, the resear cher introduced SST and explained the process to the four participants. Input from the part icipants was sought as to the importance of social skills, when they shoul d be used, what happens when someone uses social skills, and example social skills. Social skills prompts, as desc ribed in the site staff training section above, were ta ught to the participants. Spec ifically, participants were taught that staff use of a prompt is their si gnal to engage in the desired behavior. The researcher then modeled a desirable respons e to the prompt (completion of target
SST in ASPs 51 behavior). After the researcher modeled re sponding to a prompt, each participant was asked to practice responding to a prompt Feedback regarding the strengths and weaknesses of each participants performan ce was given, and practice continued until the researcher observed the participants successf ully responding to prompts. The researcher concluded by informing the participants that they would be meeti ng once a week at a prearranged time. Step 5: Session 1 and Continued Baseline Data Collection During this step, the firs t social skills training sessi on was conducted and the first behavior introduced. This session, and all others that will fo llow, lasted approximately 30 minutes. Session 1 begin with a review of the na ture of social skills and their importance. Then, the skill was introduced. Next, the res earcher asked the participants for a common situation when this skill may be useful. The researcher then modeled the skill using externally verbalized self-talk. The researcher then gave each of the participants a chance to role play the behavior. Each participant had the opportunity to role play at least once during each session. After each role play, the ma in actor (child participant exhibiting the target behavior) had the opportunity to rece ive performance feedback from his peers. Each session concluded with any remaini ng comments or questions. These sessions closely followed McGinnis and Goldsteins (1997) skillstreaming curriculum, and utilized the Stop and Think social pr oblem solving steps introduced by Knoff and Batsche (1995). Data was collected on all three beha viors following Session 1. Data collection continued three times per week until the intervention target behavior stabilized. Stabilization of behavior was defined as a tr end line which was flat or worsening in slope
SST in ASPs 52 for at least 3 data points, as assessed by the researcher using visual analysis procedures. Each week, the SST group convened at the same time and place. If data were stable, a new behavior was introduced. If, after one or two weeks, behavior had not stabilized, a review of the previously introduced skill wa s conducted. This review followed the same format as a session when a skill was intr oduced. However, greater emphasis was placed on child role plays. In addition, hindrances en countered to successful use of the target behavior were discussed. Steps 6-8: Continued Intervention an d Continued Baseline Data Collection Steps 6-8 followed the same procedures as Step 4, with a new skill being introduced upon stabilization of the beha vior. The same randomization procedure described before were again used during thes e steps. At the beginning of each session after Session 1, the skill taught the previous week was reviewed. Baseline data collection continued. However, Step 7 did not include baseline data collecti on because all three skills already had been introduced. Step 7 c oncluded when the thir d behavior introduced stabilized. Data Analysis Upon completion of all phases of the st udy, the data were analyzed. Because the design was a low-N design, visual inspection was used to dete rmine if a behavior change was significant. Visual inspecti on is the most often used form of analysis; and, except in certain circumstances, is sufficient in determin ing the effectiveness of the intervention in a multiple baseline design (Kazdin, 1982). Visu al inspection involves an analysis and comparison of data within and between pha ses according to several criteria: mean (comparing the means of data in each phase), level (change from last data point in one
SST in ASPs 53 phase to first data point in next phase), trend (slope of data within a phase), variations in the latency of change (how long it takes for ch ange to take effect), variability of data (homogeneity of data), and overlapping points (Kazdin, 1982). Significance of the intervention effects was also determined by computing the percenta ge of non-overlapping data (PND) points between ad jacent phases. A PND of 90 and above indicates a highly effective intervention. A PND of 70-90 indica tes a moderately effective intervention. A PND of 50-70 indicates a mildly effectiv e intervention, and a PND of 50 and below represents an ineffective intervention and/ or non-significant effects (Mathur et al., 1998).
SST in ASPs 54 CHAPTER FOUR Results Introduction The purpose of this chapter is to repo rt the results of the present study as described in Chapter Three. The research hypot heses are restated and addressed with the results of the behavior observation data. Significance Levels and Ex perimental Control A multiple baseline across behaviors design was used in this study to address the effects of social skills traini ng on prosocial behaviors. In a ddition, the differ ential effects of two major components of soci al skills training direct instruction and reinforcement on prosocial behaviors were explored using a multiple baseline format. Specifically, direct instruction and reinforcement were introduced in different phases for different behaviors to determine whether direct inst ruction, reinforcement, or both influenced prosocial behavior of the part icipants the most. Experiment al control was achieved for both the direct instruction intervention and the direct instruction/reinforcement combination as evidenced by the data in Figure 1. Experimental control was not, however, achieved for the re inforcement intervention. Reinforcement was delivered separate from direct instruction for both the behaviors of sitting properly and attending. Pr ior to the implementation of reinforcement for sitting properly, the data indicated that th e performance of the participants had begun to improve. The participants were sitting prop erly an average of 89% of the time. While the data did indicate some improveme nt in behavior upon implementation of reinforcement, the data already had begun to show improvement, limiting experimental
SST in ASPs 55 control. The data for attending showed a si milar pattern. Attending behavior began to improve significantly during baseline conditions limiting the degree to which it could be asserted that reinforcement, as opposed to another variable, was responsible for the change. Significance for each intervention was determined using several procedures. Data were analyzed in terms of m ean, level, slope, variabilit y, latency, and overlapping points for each phase of the study. Visual analysis of trend was conducted using trend lines created through Microsoft Excel These trend lines were compared across phases and participants of the study. A st able trend line indicates no in tervention effects, while an increasing trend indicates an improvement in behavior. A decreasi ng trend line indicates worsening behavior. Visual an alysis of level was conducted by comparing the last data point in the first phase with the first data point in the second phase of comparison. Visual analysis of mean was conducted by comparing the mean level of behavior within each phase of the study. Visual analysis of late ncy was conducted by examining the amount of time between the implementation of the interv ention and the presence of any intervention effects. Visual analysis of variability was conducted by comparing the range of data points within each phase with the range of data points from other phases. Significance of the intervention effects was also determ ined by computing the percentage of nonoverlapping data (PND) point s between adjacent phases. A PND of 90 and above indicates a highly effective intervention. A PND of 70-90 indicates a moderately effective intervention. A PND of 50-70 indica tes a mildly effective intervention, and a PND of 50 and below represents an ineffectiv e intervention and/or non-significant effects (Mathur et al., 1998).
SST in ASPs 56 The specific phases along with the data are presented in Figure 3 below:
SST in ASPs 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Part. 1 Part. 2 Part. 3 Part. 4Baseline SST and Rf+: SST: Attending and Rf+: Raising Hand and Sitting in Seat SST: Seated Pro p erl y Raising Hand 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Part. 1 Part. 2 Part. 3 Part. 4 Sitting Properly 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%3/3 3 /9 3 /3 0 4 / 5 4/7 4 /1 3 4/20 4/26 5 /2 5 /4 5 / 9 5 /1 1 Part. 1 Part. 2 Part. 3 Part. 4 A ttending 57
SST in ASPs 58 Hypothesis: The hypotheses in this study were as follows: 1. The direct instruction com ponent of social skills training will result in a significant increase in positive social behaviors. 2. The reinforcement component of social sk ills training will result in a significant increase in positive social behaviors. 3. When combined, the direct instruction and reinforcement components of social skills training will produce a significant increase in positive social behaviors. The direct instruction component of social skills training will re sult in a significant increase in positive social behaviors. This hypothesis was confirmed. There wa s a positive relationship between the direct instruction component of social skil ls training and prosocia l behaviors. Direct instruction was delivered separate from rein forcement for two of the three behaviors sitting in seat properly and atte nding. Visual analysis indicate s that the dire ct instruction component produced significantly higher rates of sitting in seat properly. However, direct instruction had varying results on attending. Figure 4
SST in ASPs 59 Sitting Properly0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%3 /3 3/ 1 0 3/ 1 7 3/ 2 4 3 /3 1 4 /7 4/14 4/21 4/ 2 8 5/ 5 5 /1 2 Part. 1 Part. 2 Part. 3 Part. 4 Baseline DI and Rf+: Raising Hand DI: Sitting Properly and Rf+: Attending and Raising Hand DI: Attending and Rf+: Raising Hand and Sitting Properly Rf+=Reinforcement; DI =Direct Instruction Figure 2 presents the data for all four part icipants for the positive social behavior of sitting properly. There was a large level change between the last data points in baseline and the first data points in th e direct instruction interventi on phase. Specifically, all data points were under 20% on the last baseline da ta points and jumped to over 80% for the first intervention data points. There was also a large increase in the means between the two phases. Across all four participants, ther e was an 89% increase in the mean of sitting properly behavior with the implemen tation of direct instruction. There does not appear to be a difference in slope between the two phases. However, the rate of sitting properly could not have significantly gone down any more during baseline, and could not have significan tly gone up during the direct instruction intervention phase. There was no latency of cha nge from baseline to the direct instruction phase the increases occurred immediately.
SST in ASPs 60 The PND data indicate a highly effective in tervention according to the significance criteria described previously. A PND of 100% was computed for each participant between the baseline and direct instruction pha ses. Finally, the variability for two of the students (Participant One and Participant Thre e) decreased substantia lly as indicated by substantially diminished ranges of data be tween the baseline a nd direct instruction phases. Specifically, Participant Ones range decreased by 34 percentage points, while Participant Threes range decr eased by 23 percentage points. The variability appeared to be low, yet unchanged for the other two student s (Participant Two and Participant Four). Participant Twos range increased from 17 to 20 percentage points, and Participant Fours range decreased from 16 to 13 percentage points. Figure 5 Attending0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%3 /3 3 /10 3 /17 3 /24 3 /31 4/7 4/14 4 /21 4 /28 5 /5 5 /12 Part. 1 Part. 2 Part. 3 Part. 4 Baseline DI and Rf+: Raise Hand DI: Sitting in Seat and Rf+: Attending and Raise Hand DI: Attending and Rf+: Raising Hand and Sitting in Seat The direct instruction com ponent of social skills tr aining on attending behavior had varying results across the four partic ipants. The data for Participant Two and Participant Three indicate a positive level ch ange from the third (baseline for direct instruction) to fourth (direct instruction intervention) phases. However, the data indicate a negative level change for Participant One a nd Participant Four. The only level change
SST in ASPs 61 exhibiting moderate significan ce is the positive level change exhibited in the data for Participant Two. The means of all four participants increased from the reinforcement phase to the direct instruction phase for attending. Th e lowest percentage point increase was experienced by Participant Two at a 13 percen tage point increase be tween the third and fourth phases. On average, participants attending incr eased 18 percentage points between the two phases. This suggests that th e direct instruction co mponent did increase attending behavior. Visual analyses of the slopes indicate that during phase three, the slopes of all participants data were increas ing. During the direct instruct ion intervention phase (phase four), however, all of the slopes began decrea sing with the exception of Participant Four. Attending data remained high consistent ly throughout the fo urth phase, though. The variability of attending data decr eased substantially during the direct instruction phase for Participants Three and Four. Specifically, Participant Threes range decreased by 28 percentage points. Participan t Fours range decreased by 44 percentage points. Particpants One and Two saw no large changes in range with the implementation of direct instruction. PND data indicate that direct instru ction, overall, was not an effective intervention with a PND of 41%. However, dire ct instruction was a moderately effective intervention for Participant One with a PND of 75%, and a mildly effective intervention for Participant Two with a PND of 50%.
SST in ASPs 62 It should be noted, as di scussed previously, that there was a low degree of experimental control with attending data. Th e data already had begun to demonstrate an improvement prior to intervention, limiting the extent to which changes in data can be attributed to direct instruction. Overall, the direct instruction com ponent does appear to have improved significantly the behavior of sitting in seat properly. In addition, the differences in means between the third and fourth phases of attending suggest that direct instruction had an impact on attending behavior as well. Variabil ity and PND data also suggest that direct instruction had a stronger impact on certain in dividual participants than others. However, inconsistent level changes across individuals, worsening slopes from the third to fourth phases, and inconsistent PND and variability data suggest that th e direct instruction component was not as effec tive for attending behavior. The reinforcement component of social skills training will result in a significant increase in positive social behaviors. This hypothesis was partially confirmed. Reinforcement was delivered separate from direct instruction for two of the th ree behaviors sitting in seat properly and attending. For the behavior of sitting in seat properly, reinforcement served to maintain an already high level of behavior. Ther e were varying results for attending. Figure 2 (above) indicates that the direct instruction component of social skills training produced consistent and high le vels of sitting properly across all four participants. Reinforcement did not improve si tting properly behavior above what direct instruction already had accomplished. The aver age increase in mean of sitting properly after the implementation of reinforcement wa s only 4 percentage points across all four
SST in ASPs 63 participants. The average reduction in range of sitting properly data was one percentage point after the implementation of reinforcemen t. In addition, the PND data indicate that reinforcement did not produce any significant bene fit to direct instruction as the average PND was 41% across all four pa rticipants. Participants One and Two, however, did have PNDs of 75% of 88% respectiv ely, indicating moderately st rong interventions. The mean data for Participants One and Two, though, did not indicate that a si gnificant change in behavior actually occurred. The PND data for in this situation, then, should be interpreted with caution. Although reinforcement did not significantly change sitting properly behavior, no further increase in sitting pr operly behavior would have been possible. Dir ect instruction served to improve sitting properly data to such a high degree that there was a ceiling effect on the data no additional benefit could have occurred. It is not possible, then, to determine whether reinforcement could have further strengthened this behavior if the direct instruction component had not been as effective. Reinforcement, however, did serve to maintain a consistently high level of behavior even after the direct instruction component had been discontinued for sitting properly in ones seat. Figure 3 (above) indicates that the rein forcement component of social skills training affected participants attending behavi or to varying degrees. For data analysis purposes, two different sets of baseline data wi ll be considered. The first set of baseline data is the traditional baseline data set, or all data collected before the implementation of any intervention. The second, alte rnative, set of baseline da ta is all data collected on March 31 and after, when direct instructi on and reinforcement were introduced for raising ones hand. This distinction is made because there appears to be a significant
SST in ASPs 64 difference between the attending data be fore and after March 31 when the first intervention for a different be havior (raising ones hand) began. Mean, variability, and PND data for both baselines are summarized in Tables 7-12. However, only the data calculated using the modified baseline data are discussed, as the modifi ed baseline data is considered to be a more accurate representati on of the data just prior to implementation of reinforcement. Before discussing the data for attending, it is important to rest ate that there was a low degree of experimental control for a ttending data. The data demonstrated an improvement in attending behavior before any intervention was conducted for attending. There were several trends in the data indicating that attend ing behavior was, in fact, better after the implementation of both reinfo rcement and direct instruction. However, it is not possible to conclude th at this improvement was the result of reinforcement of or direct instruction in attending. The individual participants responded so mewhat differently over the course of reinforcement for attending. Participant One responded very positively to reinforcement without prior direct in struction. There was a 16 percentage point increase be tween the last baseline data point and the first reinforcem ent intervention data point. In addition, there was an increase in the mean of attendi ng behavior from 34% to 67% between the modified baseline phase and reinforcement phase s. There is also a higher slope indicating a higher rate of improvement of attendi ng behavior after the introduction of reinforcement. There was no significant differe nce in the variability of attending behavior after the introduction of reinforcement. PND da ta suggest a mildly effective intervention at 68%.
SST in ASPs 65 Participant Twos attending behavior al so increased, but not with the same strength as Participant One. First, there was a drop in level upon the introduction of reinforcement. However, this appears to have just been a temporary decrease as there was an increase in mean of attending behavior fr om 55% to 72%. The slope of the data after the introduction of reinforcement, though, does ap pear to be lower than during baseline, indicating that attending behavi or increased at a slower ra te after the introduction of reinforcement. In addition, although there was a nine-point decrease in the range of the data after the implementation of reinforcement, the data do not appear to be significantly more stable as there was still a 33-point range in the data. PND data suggest that reinforcement was a mildly effective intervention for Participant Two at 50%. Participant Threes attending behavior appears to have improved when looking at the level change between the baseline a nd reinforcement conditions. There was a 39 percentage point increase between the last point of the baseline phase and the first point of the reinforcement phase. A nother indication of the effec tiveness of reinforcement on Participant Threes attending be havior is the change in sl ope between the two phases. During the baseline phase, the slope indicated a decline in attendi ng behavior. After the introduction of reinforcement, however, the slope indicated an increase in attending behavior. Despite these data trends, however the mean of attending behavior actually declined from 69% to 68% between phases. Fu rthermore, the data were no more stable after the introduction of reinforcem ent with a five-point increase in the range of data after the implementation of reinforcement. PND da ta suggests an ineffective intervention as 100% of the data points during the reinforcem ent phase overlapped with data points from the modified baseline phase.
SST in ASPs 66 Reinforcement appears to have been slight ly more effective with Participant Four than Participant Three, but not significantly. There was a mean increase in attending behavior from 52% to 59%, and there was a pos itive change in slope from baseline to reinforcement conditions. However, there wa s a 36 percentage poi nt drop in level between the two phases, and the data during the reinforcement phase was more variable than during baseline as evidenced by an in crease in range by 7 pe rcentage points. PND data do suggest that reinforcement was a m ildly effective intervention, though, at 50%. Reinforcement, overall, appears to have had varying effects across participants. According to visual analysis procedures, Participant One and Pa rticipant Two responded more positively to reinforcement than did Participant Three or Participant Four. All participants did appear to have benefited to some degree from reinforcement. However, there were significant limitations to the effec tiveness of reinforcement with three of the four participants (excluding Participant One). Across all four participants, there was a 14 percentage point increase in mean from the modified baseline phase to the re inforcement phase. There was an increase by two percentage points in the range of the data. In addi tion, PND data suggest that, overall, reinforcement was not an effective intervention for attending with only 52% of the data in the reinforcement phase not overl apping with data from the modified baseline phase. When combined, the direct instruction and re inforcement components of social skills training will produce a significant incr ease in positive social behaviors. It is not possible to make a direct co mparison between the effects of direct instruction alone, reinforcement alone, and th e two conditions combined within a single
SST in ASPs 67 behavior because all three conditions were not experimentally presented for any single behavior. However, the effectiveness of th e combination of direct instruction and reinforcement was examined with the behavior of raising ones hand before leaving the seat, referred to as raising hand. Figure 6 Raising Hand0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%3/3 3/10 3/17 3/24 3/31 4/7 4/14 4/21 4/28 5/5 5/12 Part. 1 Part. 2 Part. 3 Part. 4 Baseline SST and Rf+: Raise Hand SST: Seated Properly and Rf+: Attending SST: Attending and Rf+: Raising Hand and Sitting in Seat Figure 6 presents the data for rais ing hand. Upon the presentation of the combination of reinforcement and direct in struction, all domains of visual analysis indicate an improvement in raising hand. Ther e was a large increase in the means of each participant between baseline and interv ention phase. Specifically, there was a 94 percentage-point increase from the baseline to interventi on phase. There was a significant level change between the two phases. PND da ta suggest a strong intervention with 100% of the data points during the intervention phase overlapping with data from the baseline phase. A positive change in slope was noted for all participants during intervention phase except for Participant Three, who raised his hand 100% of the time during all intervention phases. There was an increase in th e variability of the data, with an increase
SST in ASPs 68 in range of data by 30 percentage points acro ss all participants dur ing intervention phase. However, data during the baseline phase was hi ghly consistent because of the 0% level of raising hand data dur ing baseline phase. While it is not possible to make a statement regarding the relative effectiveness of the combination of direct instruction and rein forcement when compared to either element presented independently, it is possible to confirm that this combination of intervention components produced powerful and lasting results across all in tervention phases. Comparison of the Relative Effectiveness of Reinforcement, Direct Instruction, and the Combination of Both Interventions Both direct instruction and the comb ination of direct instruction with reinforcement seem to have produced the grea test results in the present study. There was an 89 percentage point increase on average in the means of sitting pr operly data after the implementation of direct inst ruction. PND demonstrate that this intervention was highly effective at 100%. For attending behavior the range of data decreased from 52 percentage points during reinforcement phase to 35 percentage points during direct instruction phase. The combination of direct instructio n and reinforcement resulted in a 94 percentage point increase on average in the means of raising hand data after the implementation of direct inst ruction and reinforcement. PND data suggest a highly effective intervention at 100%. Reinforcement, at its best, resulted in a 25 percentage point increase in attending behavior. However, this statistic is cannot be experimentally attributed to reinforcement as the data had begun to increase before re inforcement was implemented. Overall, when
SST in ASPs 69 compared to direct instruction, reinforcem ent produced fewer desirable results. When reinforcement did appear to produce greater resu lts than direct instruction, the difference was not large. Table 1 Mean Level of Sitting Properly Data Within and Between Phases Participant Baseline DI1 Change Rf+2 Change Net Change3 1 9% 95% 86 97% 2 88 2 5% 93% 88 99% 6 94 3 5% 100% 95 100% 0 95 4 6% 93% 87 99% 6 93 Average 6% 95% 89 99% 4 93 1DI = Direct Instruction 2Rf+=Reinforcement 3Net Change=Total change betw een Baseline and Rf+ phases. Table 2 Range of Sitting Properly Data Within and Between Phases Participant Baseline DI Cha nge Rf+ Change Net Change 1 421 8 -34 25 17 -17 2 17 20 3 5 -15 -12 3 23 0 -23 1 1 -22 4 16 13 -3 6 -7 -10 Average 25 10 -14 9 -1 -15
SST in ASPs 70 1Percentage Points Table 3 Percent of Nonoverlapping Points Between Phases for Sitting Properly Data Participant Baseline to DI DI to Rf+ Baseline to Rf+ 1 100%1 75% 100% 2 100% 88% 100% 3 100% 0% 100% 4 100% 0% 100% Average 100% 41% 100% 1Interpreted as 100% of the da ta in the DI phase does not overlap with data in the Baseline Phase. Table 4 Mean Level of Raising Hand Data Within and Between Phases Participant Baseline DI and Rf+1 Change 1 0% 87% 87 2 0% 98% 98 3 0% 100% 100 4 0% 90% 90 Average 0% 94% 94 1Direct Instruction and Reinforcement were delivered simultaneously for Raising Hand
SST in ASPs 71 Table 5 Range of Raising Hand Data With in and Between Phases Participant Baseline DI and Rf+ Change 1 0 40 40 2 0 30 30 3 0 0 0 4 0 50 50 Average 0 30 30 Table 6 Percent of Nonoverlapping Points Betw een Phases for Raising Hand Data Participant Baseline to DI and Rf+ 1 100% 2 100% 3 100% 4 100% Average 100%
SST in ASPs 72 Table 7 Mean Level of Attending Data With in and Between Phases Baseline11 Participant Baseline1 Rf+ Change DI Change Net Change 1 29% 67% 38 82% 15 53 2 35% 72% 37 85% 13 50 3 61% 68% 7 89% 21 28 4 42% 59% 17 85% 26 43 Average 42% 67% 25 85% 19 44 1Baseline1 = traditional baseline data set, or all data collected before the implementation of an intervention for attending behavior. Table 8 Mean Level of Attending Data With in and Between Phases Baseline21 Participant Baseline2 Rf+ Change DI Change Net Change 1 34% 67% 33 82% 15 48 2 55% 72% 17 85% 13 30 3 69% 68% -1 89% 21 20 4 52% 59% 7 85% 26 33 Average 53% 67% 14 85% 18 32 1Baseline2 = modified baseline data set, or a ll data collected on Ma rch 31 and after, but before the implementation of any intervention for attending behavior.
SST in ASPs 73 Table 9 Range of Attending Data Within and Between Phases Baseline1 Participant Baseline1 Rf+ Change DI Change Net Change 1 31 33 2 38 5 7 2 70 34 -36 33 -1 -37 3 56 61 5 33 -28 -23 4 71 78 7 34 -44 -37 Average 57 52 -6 35 -17 -23 Table 10 Range of Attending Data Within and Between Phases Baseline2 Participant Baseline2 Rf+ Change DI Change Net Change 1 31 33 2 38 5 7 2 43 34 -9 33 -1 -10 3 56 61 5 33 -28 -23 4 71 78 7 34 -44 -37 Average 50 52 2 35 -17 -15
SST in ASPs 74 Table 11 Percent of Nonoverlapping Points(PND) Betw een Phases for Attending Data Baseline1 Participant Baseline1 to Rf+ Rf+ to DI Baseline1 to DI 1 67% 75% 100% 2 50% 50% 88% 3 0% 38% 38% 4 50% 0% 50% Average 42% 41% 69% Table 12 Percent of Nonoverlapping (PND)Points Betw een Phases for Attending Data Baseline2 Participant Baseline2 to Rf+ Rf+ to DI Baseline2 to DI 1 67% 75% 100% 2 50% 50% 88% 3 0% 38% 38% 4 50% 0% 50% Average 42% 41% 69%
SST in ASPs 75 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 3/33/93/304/54/74/134/204/265/25/45/95/11 A ttending Participant One0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Baseline SST and Rf+: Raise Hand SST: Seated Properly and Rf+: Attending SST: Attending and Rf+: Raising Hand and Sitting in Seat Raising Hand 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Sitting Properly Figure 7. Participant One Data
SST in ASPs 76 Figure 8. Participant Two Data. Participant Two0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Baseline SST and Rf+: Raise Hand SST: Seated Properly and Rf+: Attending SST: A ttending and Rf+: Raising Hdd Raising Hand 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Sitting Properly 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%3/3 3/ 10 3/17 3 / 24 3/ 3 1 4 / 7 4/ 1 4 4/21 4/ 2 8 5/5 5 / 12 A ttending
SST in ASPs 77 Figure 9. Participant Three Data. Participant Three0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Baseline SST and Rf+: Raise Hand SST: Seated Properly and Rf+: Attending SST: A ttending and Rf+: Raising Raising Hand 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%3/3 3/ 1 0 3/17 3/ 24 3/31 4/7 4/14 4/21 4/ 2 8 5/5 5/ 12 Sitting Attendi
SST in ASPs 78 Figure 10. Participant Four Data. Participant Four0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Baseline SST and Rf+: Raise Hand SST: Seated Properly and Rf+: Attending SST: A ttending and Rf+: Raising Hdd Raising Hand 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Sitting Properly 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%3/3 3/ 1 0 3/17 3/ 24 3/31 4/7 4/14 4/21 4/ 2 8 5/5 5/ 12 A ttending
SST in ASPs 79 CHAPTER FIVE Discussion Introduction The present study investigated the extent to which social skills training improved the prosocial behaviors of elementary male s in an after school program. The present study also examined the differential effects of the direct instruction and reinforcement components of social skills training. This discussion will address the findings of the present study, consistency of the finding with previous research, and limitations in the design of the present study that affect inte rpretation and generalization of these findings to other populations. This discussion will also address implications for implementing social skills training in an after-school program, as well as implications for future research. Discussion of the Findings The data presented in Chapter Four indicat e several trends. First, the combination of the direct instruction and reinforcement co mponents did have a si gnificant affect on the behavior of raising hand. As discussed in the previous chapter, the data indicate a large increase between baseline and in tervention phases for that behavior. Second, the direct instruction component of social skills training did increase participants prosocial behavior of sitting properly. Data af ter the implementation of the direct instruction component in dicate significantly higher leve ls of sitting properly than during baseline phase. Third, there is not sufficient data to indicate whether the reinforcement component alone influenced any of the behavior s in the study for participants as a whole.
SST in ASPs 80 However, there were some effects noted with individual participants. Reinforcement was part of an intervention package that mainta ined consistently high levels of behavior across all three behaviors. In addition, wh en reinforcement alone was added to the behavior of attending, a desi rable change was noted. Howe ver, it is not possible to conclude that these results are attributab le to the reinforcement component as the behavior of attending had already begun to change during baseline. Fourth, some general intervention effects we re likely an influencing factor with at least the behavior of attending. Attending began to increase during baseline phase. Although attending continued to be stre ngthened after the implementation of interventions, it did not improve at a faster rate than it had during baseline. While social skills training may have influe nced attending, then, it was not the only factor involved. Overall, social skills training as a whole was successful in improving the behaviors of raising hand and sitting properl y. It also may have had an influence over attending. However, the results with attending are less certain. Strong evidence exists that the direct instruction component was a powerful part of social skills training. Limited data prevent a similar conclusion for rein forcement, though some effects were noted. Nevertheless, the social skills intervention pa ckage as a whole resulted in a clinically significant effect on behavior in that participants, for the most part, were on-task over 70% of the time at the end of the study. Possible Explanations for the Present Findings Several factors may have contributed to the outcomes in th is study. The factors can be separated into two categories: (1) social skills training factors, and (2) factors related to the measurement of the dependent variables.
SST in ASPs 81 Social skills training factors. There may be several reasons why the improvement and maintenance of the targeted behaviors o ccurred. Before the implementation of social skills training, participants may have lacked important behavioral and cognitive skills needed to produce high frequencies of the targ et behaviors. In addi tion, participants may not have known sufficiently where and when to use the target behaviors. For children that might have had these skills, they may have l acked sufficient practice or feedback that would have facilitated competent performance of each skill. The participants also might have been experiencing high ra tes of interfering behaviors or thoughts that prevented the use of the target skills. The participants may have acquired compet ency in the target behaviors and their knowledge of where and when to use them duri ng the social skills training sessions held weekly (Elliot & Gresham, 1993; Sheridan, Hungelmann, & Maughan, 1999). Participants were taught the distinct behavioral and cogni tive steps of each expected behavior. Participants were then given the opportunity to practice eac h step and skill both in the social skills training session and during homework time at the after school program. In order to further generalization, two f actors were included in the social skills training intervention. First, the Stop & Thi nk language was used in both the social skills training sessions and during homewo rk time. More specifically, staff and volunteers in the homework help area were inst ructed to use the same language to prompt the use of newly acquired social skills as the language being used during the social skills training sessions. Second, the model and role play scenarios were selected to match actual scenarios that occurred during homewo rk time at the after school program. These
SST in ASPs 82 procedures were intended to further gene ralization into the homework help setting (Stokes & Osnes, 1989). Participants previously may not have found it reinforcing to use the targeted social skills. They may not have understood the connection between social skills and reinforcement, or the environment may not have been naturally reinforcing of the targeted social skills. For example, prior to implementation of the interventions, there was no reinforcement system during homework time. In addition, verbal praise was seldom given for performance of the targ eted behaviors. Implementation of the reinforcement procedures may have modified the environment to be more reinforcing of the targeted social skills. Other trends in the data we re also observed. At the beginning of the study, the behavior of attending was cons istently at or below 50% for all four participants, with no visible trend in either direction. Over the c ourse of the investiga tion, rates of attending became consistently high (e.g., all data point s above 65% during direct instruction of attending). However, these increases began to occur before any inte rvention targeted at attending was implemented. Specifically, as soon as social skills training was implemented for raising hand, attending data also began to increase. This increase occurred steadily throughout the study and di d not appear to be affected by the addition or subtraction (reinforcement was withdrawn for attending during th e last phase) of any condition. There may be several reasons for this trend. First, the beginning of intervention imple mentation for raising hand might have served as a signal to the staff that higher le vels of behavior in general were expected. Staff might have increased the frequency or quality of other unrelated behavior
SST in ASPs 83 management practices as a result of the knowledge that the researcher was expecting behavioral improvement. This might have o ccurred because they knew they were being watched and wanted to improve their own perf ormance as staff members. It is a general trend that people tend to cha nge their behavior when they know that they are being observed (Kazdin, 1982). In addition, it could have occurred because the staff members were interested in improving the behavior of the participants and subsequently increased their efforts with behavior management. The possibility that staff members might ha ve made a concerted effort to improve overall behavior seemingly would have affect ed the behavior of sitting properly as well. Sitting properly, though, showed little improvement duri ng the time that attending began to rise. It may be that atte nding was considered by staff as a more general behavioral index, and was the focus of unrelated interv entions more frequently. For example, it might have been more common for staff or volunteers to approach a child and encourage him to get back to work more frequently than to sit in ones seat properly. Informal observations during data colle ction confirm that the children, at least, had this perspective. Throughout the study, the researcher fre quently discussed expectations and appropriate behavior with the participants All four participan ts frequently overgeneralized expectations of th eir behavior. For example, when asked what was needed to earn reinforcement for a particular day, a pa rticipant might have said be good and do my work, when all that was requi red according to the study was to sit properly in ones seat. The participants were always corrected, but continued to over-generalize what was expected of them during homework time.
SST in ASPs 84 The expectations inherent in each target ed social skill did not overlap. In other words, the same expectation could not be f ound in more than one behavior. An increase in raising hand would not automatically result in an increase in attending. However, there might have been an informal association be tween the targeted va riables. Specifically, once the participants began raising their hands be fore getting out of th eir seats, they spent significantly more time in their seats. This might have created additional opportunity for participants to engage in an appropriate task, thus increa sing their attending behavior. Measurement of the dependent variables. Several factors related to the methods of measurement of the target behaviors might ha ve accounted for the results in this study. First, direct observations occurred during the same time and activity (homework time) each day. However, homework time was run differently on different days. Staff sometimes held large group activities or helped the entire group with the same homework assignment. Other times, staff helped i ndividual children while the rest worked independently on homework. Staff also varied in their use of behavior management techniques. Some days staff w ould issue a large number of directives and punishment, for example, while other times the staff would be relatively unengaged in terms of behavior management. In addition, staff might have re acted to the researchers presence as an observer by increasing their effo rts in behavior management, compared with times when the researcher did not serve as the observer These variances might account for some of the variability in the data. Variability in the raising hand data may be attributed to the method of calculating the percentage of time raising hand. If the part icipant got out of his seat only one time, and raised his hand to get out of his seat no ti mes, then the participant was considered to
SST in ASPs 85 have raised his hand 0% of the time. However, if the participant got out of his seat once and raised his hand for that occu rrence, then the participant was considered to have raised his hand 100% of the time. This method of reporti ng the data is, at times, oversensitive to small changes, resulting in the presentation of the data as mo re variable than if it would have been presented in a different format. In the future, this type of variability might be reduced by expanding the number of behaviors contributing to each data point. This might be accomplished by collapsing multiple observation sessions into one data point, expanding observation periods, or increasing pa rticipants opportunitie s to engage in the behavior by artificially creating more stimuli for the behavior to occur. Inconclusive results regarding the beha vior of attending may be due to the researcher moving too quickly from baseline phase to interv ention phase. The data were not stable, but rather were moving in the direction of intended intervention effects. Ideally, the researcher would ha ve continued to collect baseli ne data until such data had stabilized. The researcher, however, had to move into intervention phase in order to complete data collection by the end of the time period allotted for the study. As a result, it is not possible to draw firm conclusions regard ing the differences in the data in different phases for this behavior. Each participant in the study was located in a different part of the room, which may have influenced the results of this study. Some participants were located closer to the staff. Some were located cl oser to other children in the room, while one was seated at his own table. Seating patterns also cha nged slightly throughout the study, possibly influencing results. Informal observations di d not reveal any cha nges that produced any
SST in ASPs 86 known trends in the data, yet the possibility sti ll exists that the resu lts were influenced by these changes. Consistencies with Established Research The findings in this study support the asser tion that reinforcement can be a part of an effective intervention package to increas e social skills (Bierman, Miller, & Stabb, 1987). These authors also investigated re sponse cost used independently and in combination with reinforcement and found th at a combination of the two produced the most effective results. Similarly, reinforcem ent was found to be effective when used in combination with another intervention com ponent in the present study. Bierman, Miller, and Stabb, however, were able to isolate re inforcement as a variable more effectively. Lochman, Lampron, Gemmer, Harris, and Wy ckoff (1989) investigated the use of modeling, role-playing, discussion, social pr oblem-solving training, and self-instruction to improve social skills. On all dependent measures, they found the interventions to be successful in increasing the occurrence of prosocial behaviors. These results are comparable to the results of the present st udy. In addition, many of the methods used to increase the social skills of the participants were also used in the present study. Mize and Ladd (1990) also found that mode ling and role playing in a pull-out session similar to the one used in the pres ent study increased the number of prosocial behaviors exhibited by participants. Mi ze and Ladd used puppets to model the appropriate behaviors, but childr en participated in role plays of a similar fashion to the ones used in the present study. In addition, the researchers videotaped the role plays and used them to provide performance feedback to the participants.
SST in ASPs 87 Social skills training can be conducted in a variety of formats and settings. There were many different studies that demonstr ated significant improvements in prosocial behaviors using different t echniques (Ang & Hughes, 2001). This trend of equifinality might suggest that there are certain core com ponents of social skills training, such as modeling and role-playing, used in many effective SST program s. However, the specific way in which these elements are applied in e ach study vary. This variability in specific intervention components, then, might suggest that consumers of th e literature should not rigidly interpret every element of a st udy, but rather look for underlying themes consistent across investigations. Consumers of this literature s hould then adapt these general themes to meet the specif ic demands of th eir environment. The results of this study, overall, are cons istent with the literature demonstrating that social skills training can increase prosocial behavior (Ang and Hughes, 2001). However, the results of this study extend the lite rature base in that social skills training has not yet been investigated in the literature in after school program settings. Limitations and Threats to Validity Several limitations that were inherent in the design of the study may have accounted for some of the findings, and jeopard ize the interpretabil ity and generalization of these results to other populations and settings. Th ese issues include sampling procedures, reactivity, presence of the researcher, observer drift, failure to include a peer comparison, participant variab ility, and external events. Sampling procedures The participants for the pres ent study were not selected on a random basis. Rather, they were selected th rough the systematic pr ocess of identifying children in most need of soci al skills training. The final ga ting procedures involving the
SST in ASPs 88 use of the social skills rating form and beha vior observations more objectively narrowed the range of participants down to four. However, the initial selection of th e eight potential participants was done using onl y staff recommendations. No m easures were in place to control for biased selection of participants. There is no way to determine, therefore, if the initial eight participants were selected base d on the criteria outlined by the researcher, or whether hidden variables (personality, academic performance) were used by staff to make their recommendations. The occurrence of so cial skills deficits, though, was ensured using the social skills rating fo rm and behavior observations. Reactivity The researcher served as one of th e data collectors in the present study. This may have influenced the behavior of bot h the staff and participants. Both the staff and participants might have modified thei r behavior because of the presence of the researcher. Upon implementation of the interv entions, participants may have associated the researcher with the interventions, and may have felt more compelled to engage in the target behaviors. However, the researcher was frequently present even when other data collectors were used. The presence of the rese archer likely had uniform effects, then, across all of the data. In addition, the re searcher had been attending the homework sessions for approximately four months on a dail y basis prior to the collection of baseline data in order to reduce reactivity. Both the st aff and participants had ample opportunity to acclimate to the presence of the researcher. Observer drift Several inter-observer agreement checks were conducted throughout data collection to measure any pot ential drift. One follo w-up training session was conducted at the request of one data colle ctor in the middle of study. Inter-observer
SST in ASPs 89 agreement was not conducted during each pha se of the study, however. Therefore, observer drift was a possibility. However, when measured, no drift was detected. Failure to include a peer comparison No control participan ts were included in the study, which has resulted in an inability to compare the behaviors of children receiving the intervention with children not receiving the intervention. However, experimental control was achieved using a multiple baseline across behaviors design as discussed previously. Participant variability. While efforts were made to match participants based on selected demographic variables, some importa nt differences between participants were still present. One participant, for exampl e, began learning English upon enrollment in school several years prior. Previously he had spoken primarily Spanish. Another participant was undergoing a medical ev aluation for the consideration of psychopharmacological interventions for behavi or problems at school, while the others were not. These effects might have had an uncontrolled effect on the findings. External events This study occurred over the cour se of several weeks, and did not control for the occurrence of events outside the study. Some of these results may have influenced the findings. Standardized testi ng occurred during two weeks of the study, for example. Stress and a lack of homework are two potentially present factors during these two weeks which may have influenced data collection. In addition, the after school program was evolving as well. Throughout the c ourse of data collection, staff attempted several different behavior mana gement techniques independent of researcher efforts. In addition, the composition of the staff cha nged during data collection several new individuals became staff and some move d on. However, the staff conducting the
SST in ASPs 90 homework sessions of the participants remain ed the same throughout the entirety of data collection. There were days, however, that subs titute staff members were present in the event of regular staff absence. This may have affected the findings. Staff Training. The presence of an individual wi th doctoral-level training in behavioral interventions might limit the external validity of the findings in this study. Social skills training may not be adapted a nd implemented as effectively in other after school programs without the same resources. Implications and Future Directions The present study was conducted in an era of uncertainly surrounding the effectiveness of after school programs. The proliferation of ASPs has been sudden and sure, yet their worth remains relatively unprove n in a time of increased accountability in education and youth development (BGCTB 2003; FDOE, 2004; Hillsborough County Recreation and Conservati on, 2003; SDHC, 2004; YMCA,2 003). Only recently have studies been conducted to determine whether th e public should continue to take stock in these new educational endeavors. The present study adds to this literature base, helps answer several questions, and raises severa l challenges to those who influence after school programs. The results of this study, while so mewhat variable, do indicate that a psychological intervention (social skills training) can be used effectively in an after school program. This study adds to the relativ e dearth of existing studies making such a statement. More broadly, this investigation helps science to respond affirmatively to the question of whether after school programs can work to educate children. The present findings help assert that after schoo l programs can experience success.
SST in ASPs 91 These findings also help assert that after school programs can measure this success using traditional research methods. This truth creates possibilities for after school programs in that it might enable its proponents to see yet another color in the spectrum of opportunities available in program evaluation. Ho wever, it also create s responsibility. As more sophisticated forms of program evalua tion are used in after school programs, less impressive forms will cease to suffice. Simply measuring the number of participants served and the average GPAs of participants soon will no longer bring the money it once did. Programs will be required to show that more is being done with the enormity of wealth being invested in programs. However, supporters of after school program s are also challenged by the results of this study in that, in order to fulfill expect ations, after school programs need resources. Most after school programs do not have the assistance of a resear cher with doctoral training in social skills training. If the expectations are to be raised, then the resources must be raised as well. Supporters are now mo re aware of the potential of such resources the benefit of investment in afte r school programs is less deniable. The necessity of resources in after school programs similar to those available in the present study does not mean that individuals with doctoral level training need to be present in every after school program. There are more crea tive and financially feasibly ways of addressing resource limitations, su ch as improved connections between the research base in education (and those who unde rstand this research ba se) with those that structure the content of after school programs. For example, if a school district sponsors a series of after school programs around the distri ct, it might be possible for a team of welltrained individuals in the district to infu se of set of guidelines regarding behavior
SST in ASPs 92 management into after-school programs. Anot her example would invol ve staff of after school programs being invited to teacher in-service trainings. The findings of this study also help establish the assertion that after school programs should stay focused on quality rather than quantity. While quantity is important (kids do need supervision after school), focu sing on rapid growth will likely result in a lower quality of programs. Basic goals such as supervision might be achieved. However, in an age of accountability, those examini ng the cost effectiveness of after school programs will want to see more. If expanded too rapidly, after school programs might not be able to deliver desired results. Cont rolled growth, with a concerted effort on establishing high quality, will systematically build trust in after school programs and open doors for future expansion and long-term stability. The findings of this study help support the premise that, if given appropriate resources, after school programs can be an inte gral partner in the de livery of educational and psychological services. The data also add to the growing awareness that school psychologists can be used in settings othe r than schools. School psychologists are beginning to establish th eir importance in settings such as hospitals and mental health centers. This expansion of service deli very does not need to stop there. Furthermore, if educational and psychol ogical services are to be expanded into new domains, there needs to be increased coordination between t hose providing these services. Mental health agen cies, schools, churches, afte r schools programs, hospitals, and all organizations participating in this collaboration will need to become more organized around their common mission of im proving the educational and psychological
SST in ASPs 93 well-being of their clients. School psyc hologists that understand the importance of collaborative consultation and systems theo ry can be integral in this process. As this is the first study that was found to investigate social skills training in an after school program, and one of the first to investigate evidence-based psychological interventions in after school programs, furthe r research is needed to confirm the findings in this study and build a more broad ev idence base regarding other psychological interventions in after school programs. The generalization of psyc hological interventions from school settings to after school program settings is not automatic, and must be investigated more systematically to further explore the possibilities present in using after school programs as settings for psychologi cal and educational interventions. More specific to social skills training, future research should further investigate practical yet effective ways of implementi ng social skills training in after school programs. Small pull-out sessions such as th e ones used in the present study may not always be feasible given the resources a nd configurations of after school programs. Future research also should investig ate more effective ways of measuring intervention integrity regarding social skills prompts by staff. Intervention integrity might be measured by observing the degree to which staff issue prompts following the occurrence of undesirable behavior. In addition, future research should investigate methodologi cal issues surround program evaluation and the establishment of evidence-based practi ces in after school programs. After school programs might need different technology, ha ve different goals than schools, and have different resources. Th ese distinctions will require an examination of effective yet feasib le research methods in after school programs.
SST in ASPs References Albright, L. (2002). Evaluation of the 21 st Century After Hours Program: Poudre School District Fort Collins, CO: Research and Development Center for the Advancement of Student Learning. American Psychological Association. (1985). Standards for educational and psychological testing. Washington, DC: Author. Ang, R. P., & Hughes, J. N. (2001). Differential be nefits of skills training with antisocial youth based on group composition: A meta-analytic investigation. School Psychology Review, 31 164-185. Baenen, N. R., Lindblad, M., & Yaman, K. (2002). Can Extended Learning Opportunities Improve Student Achievment? Paper presented at the Annual Meeting of the American Educational Re search Association, New Orleans, LA. Baker, D.W., & Witt, P.A. (1996). Evalution of the impact of two after-school recreation programs. Journal of Park and Recreation Administration, 14 23-44. Barker, N.C. (1998). Can specialized after-school progr ams impact delinquent behavior among African-American youth? Child welfare and juvenile justice Proceedings of the Tenth Annual Research Conferen ce, A System of Care for Childrens Mental Health: Expanding the Research Base, Tampa, Fl. 2-7. Beelmann, A., Pfingsten, U., & Losel, F. (1994). Effects of training social competence in children: A meta-analysis of recent evaluation studies. Journal of Clinical Child Psychology, 23, 260-271. Bierman, K. L., & Furman, W. (1984). The eff ects of social skills training and peer 94
SST in ASPs involvement on the social adjustment of preadolescents. Child Development, 55 151-162. Bierman, K. L., Miller, C. L., & Stabb, S. D. (1987). Improving the social behavior and peer acceptance of rejected boys: Effects of social skill training with instructions and prohibitions. Journal of Consulting and Clinical Psychology, 55 194-200. Bissell, J., Dugan, C., Ford-Johnson, A., Jones, P., & Ashurst, J. (2002). Evaluation of the YS-CARE After School Program fo r California Work Opportunity and Responsibility to Kids. Irvine, CA: Department of Education, University of California, Irvine, & Res earch Support Services. Blanton, W. E., Moorman, G. B., Hayes, B. A., & Warner, M. L. (1997). Effects of participation in the Fifth Dimension on far transfer. Journal of Educational Computing Research, 16 371. Blanton, W. E., Moorman, G. B., & Zimmerman, S. J. (year unknown). Ways of knowing, ways of doing, ways of transporting: Ma stering social practices in the Fifth Dimension Boone, NC: Laboratory of Lear ning and Technology, College of Education Appalachian State University. Boys and Girls Clubs of Tampa Bay (BGCTB ) Website (n.d.). Retrieved October 5, 2003 from www.bgatampa.org Boys and Girls Clubs of America (BGCA) Website (n.d.). Retrieved October 5, 2003 from www.bgca.org Cartledge, G., & Milburn, J.E. (1986). Teaching social skills to children: Inovative approaches (2 nd ed.). New York: Pergamon Press. 95
SST in ASPs Chaiken, M. (1997). Youth afterschool programs and law enforcement Washington, DC: US Department of Justice, Office of Justice Programs, National Institute of Justice. Chung, A. (2000). After-school programs: Keep ing children safe and smart. Washington, DC: US Department of Education, Partnership for Family Involvement in Education. Coats, K. I. (1979). Cognitive self-instructi onal training approach for reducing disruptive behavior of young children. Psychological Reports, 44 127-134. Cooke, T. P., & Apolloni, T. (1976). Devel oping positive social-emotional behaviors: A study of training and generalization effects. Journal of Applied Behavior Analysis, 9(1) 65-78. Cooper, J.O., Heron, T.E., & Heward, W.L. (1987). Applied behavior analysis Upper Saddle River, NJ: Prentice Hall. DiPerna, J. C., Volpe, R. J., & Elliott, S. N. (2001). A model of academic enablers and elementary reading/language arts achievement. School Psychology Review, 31 298-312. Dubow, E. F., Huesmann, E. R., & Eron, L. D. (1987). Mitigating aggression and promoting prosocial behavior in aggressive elementary schoolboys. Behaviour Research and Therapy, 25, 527-531. Elliot, S.N., & Gresham, F.M. (1993). Social skill interventions for children. Behavior Modification 17, 287-313. Fashola, O.S. ( 2002). Building effective a fterschool programs. Thousand Oaks, CA: Corwin Press, Inc. 96
SST in ASPs Florida Department of Education (FDOE ) Website (n.d.). Retrieved January 18, 2004 from www.fldoe.org Forman, S. G. (1980). A comparison of cognitiv e training and response cost procedures in modifying aggressive behavior of elementary school children. Behavior Therapy, 11, 594-600. Gay, L. R., & Airasian, P. (2003). Educational research: Competencies for analysis and applications (7 th ed.). Upper Saddle River, NJ: Merrill Prentice Hall. Goldstein, A.P. (1981). Psychological Skill Training. New York: Pergamon. Gresham, F. M., & Elliott, S. N. (1990). Social skills rating system manual Circle Pines, MN: American Guidance Service, Inc. Gresham, F. M., & Lemanek, K. L. (1983). Social skills: A review of cognitivebehavioral training procedures with children. Journal of Applied Developmental Psychology, 4 239-261. Gresham, F. M., & Elliott, S. N. (1984). Asse ssment and classification of children's social skills: A review of methods and issues. School Psychology Review, 13 292-301. Hillsborough County Recreation and Conservati on Website (n.d.). Retrieved October 5, 2003 from http://www.hillsboroughcounty.org/parks/after_school.html Huang, D., Gribbons, B. Sung Kim, K ., Lee, C., & Baker, E.L. (2000). A decade of results: The impact of LAs BEST Aftersc hool Enrichment Program on subsequent student achievement and performance Los Angeles: UCLA Center for the Study of Evaluation, with support from the Bandai Foundation and the City of Los Angeles. 97
SST in ASPs Huey, W. C., & Rank, R. C. (1984). Effects of counselor and peer-l ed group assertive training on black adolescent aggression. Journal of Counseling Psychology, 31 95-98. Hughes, J.N., & Sullivan, K.A. (1988). Outcome assessment in social skills training with children. Journal of School Psychology, 26 167-183. Kazdin, A.E. (1982). Single-case research designs: Methods for clinical and applied settings. New York: Oxford University Press. Kendall, P. C., & Zupan, B. A. (1981). Individu al versus group application of cognitivebehavioral self-control procedures with children. Behavior Therapy, 12 344-359. Kettlewell, P. W., & Kausch, D. F. (1983). The generalization of the effects of a cognitive-behavioral treatment pr ogram for aggressive children. Journal of Abnormal Child Psychology, 11, 101-114. Klein, S. P., & Bolus, R. (2002). Improvements in math and re ading scores of students who did and did not participa te in the Foundations A fter School Enrichment Program during the 2001 school year. Santa Monica, CA: Gansk & Associates. Knoff, H. M., & Batsche, G. M. (1995). Project ACHIEVE: Analyz ing a school reform process for at-risk and underachieving students. School Psychology Review, 24 579-603. Ladd, G.W., & Mize, J. (1983). A cognitive-social learning model of so cial skill training. Psychological Review, 90 (2), 127-157. Lamare, J. (1997). Sacramento START: An evaluation report Sacramento, CA: 98
SST in ASPs Sacramento California Neighborhoods Planning and Development Services Department. Lipsitz, J. (1984). Successful schools for adolescents. New Brunswick, NJ: Transaction Books. Lochman, J. E., Burch, P. R., Curry, J. F., & Lampron, L. B. (1984). Treatment and generalization effects of a cognitive-beh avioral and goal-setting intervention with aggressive boys. Journal of Consulting and Clinical Psychology, 52 915-916. Lochman, J. E., Lampron, L. B., Gemmer, T. C., Harris, S. R., & Wyckoff, G. M. (1989). Teacher consultation and cognitivebehavi oral interventions with aggressive boys. Psychology in the Schools, 26, 179-188. Massachusetts 2020. (2004). Research report: The Transition to Success Pilot Project Boston, MA: Author. Mathur, S. R., Kavale, K. A., Quinn, M. M., Fo rness, S. R., & Rutherford, R. B. (1998). Social skills interventions for students w ith emotional and behavioral problems: A quantitative synthesis of single-subject research. Behavioral Disorders, 23 193201. McComb, E.M., & Scott-Little, C. (2003). After-school programs: Evaluations and outcomes Greensboro, NC: Author. McGinnis, E., & Goldstein, A.P. (1997). Skillstreaming the elementary school child: New strategies and perspectives for teaching prosocial skills Champaign, IL: Research Press. Mertens, D. M. (2003). Mixed methods a nd the politics of human research: The 99
SST in ASPs transformative-emancipatory perspective. In Tashakkori, A. & Teddlie, C. (Eds.), Handbook of mixed methods in social and behavioral research (pp. 135 164). Thousand Oaks, CA: Sage Publications. Mize, J., & Ladd, G. W. (1990). A cognitive-social learning approach to social skills training with low-stat us preschool children. Developmental Psychology, 26 388397. Nance, E. E., Moore, D. H., & Lewis, C. F. (2000). 21 st Century Community Learning Centers: Do they affect student achievement? Community Education Journal, 27, 7-11. Nansel, T. R., Overpeck, M., Pilla, R. S., Ruan, W. J., Simons-Morton, B., & Scheidt, P. (2001). Bullying behaviors among U.S. yout h: Prevalence and association with psychosocial adjustment. Journal of the American Medical Association, 285 2094-2100. National Center for Educational Statistics (NCES) Website (n.d.). Retrieved January 18, 2004 from http://nces.ed.gov National Youth Development Information Cent er (NYDIC) Website (n.d.). Retrieved October 5, 2003 from http://www.nydic.org/nydic/ Nelson, G., & Carson, P. (1988). Evaluation of a social problem-solving skills program for thirdand fourth-grade students. American Journal of Community Psychology, 16, 79-98. Office of the Under Secretary (n.d.). Executive Summary; When Schools Stay Open Late: 100
SST in ASPs The National Evaluation of the 21 st -Century Community Learning Centers Program Retrieved October 5, 2003 at www.ed.gov/pubs/21cent/firstyear/summ.html Pepler, D. J., King, G., Craig, W., Byrd, B ., & Bream, L. (1995). The development and evaluation of a multisystem social skills group training program for aggressive children. Child & Youth Care Forum, 24 297-311. Peter, N. (2002). Outcomes and research in out-of-school time program design Philadelphia, PA: Best Practices Institute. Resnick, K. (2004). Building an outcome evaluation system: 21 st Century Community Learning Centers and after-school and out -of-school time programs (Report on Outcomes for FY 2003). Boston, MA: Massachuttes Department of Education. Riley, D. (1994). Preventing problem behavio rs and raising academic performance in the nation's youth: The impacts of 64 school age child care programs in 15 states Madison, WI: University of Wisconsin Cooperative Extension Service. School District of Hillsborough County (SDHC) Website (n.d.). Retrieved October 5, 2003 from www.sdhc.k12.fl.us Sheridan, S.M., Hungelmann, A., & Maughan, D. P. (1999). A contextualized framework for social skills assessment, intervention, and generalization. School Psychology Review 28, 84-103. Shinn, M.R., Walker, H.M., & Stoner, G. (Eds.). (2002). Interventions for academic and behavior problems II: Preventive and remedial approaches Bethesda, MD: National Association of School Psychologists (NASP) Publications. 101
SST in ASPs Spence, A. J., & Spence, S. H. (1980). Cognitiv e changes associated with social skills training. Behaviour Research & Therapy, 18 265-272. Sprafkin, R. P., Gershaw, N.J., & Goldstein, A.P. (1993). Social skills for mental health: A structured learning approach Needham Heights, MS: Allyn & Bacon. Stokes, T.F., & Osnes, O.G. (1989). An operant pursuit of generalization. Behavior Therapy, 20, 337-355. Tanner, V. L., & Holliman, W. B. (1988). E ffectiveness of asser tiveness training in modifying aggressive be haviors of young children. Psychological Reports, 62 3946. US Department of Education. (1995). Goals 2000: A progress report US Department of Education: Author. US Department of Education. (2003). Executive Summary; When Schools Stay Open Late: The National Evaluation of the 21 st -Century Community Learning Centers Program US Department of Education: Author. US Department of Education website (n.d.). Retrieved January 18, 2004 from www.nclb.org Vaugn, S., & Lancellota, G. X. (1990). Teachin g interpersonal social skills to poorly accepted students: Peer-pairing versus non-peer-pairing. Journal of School Psychology, 28, 181-188. Vitaro, F., and Tremblay, R. E. (1994). Impact of a prevention program on aggressive childrens friendships an d social adjustment. Journal of Abnormal Child Psychology, 22, 457-471. Wahlstrom, K., Sheldon, T., Anderson, R., & Zorka, H. (2001). 21 st Century Community 102
SST in ASPs Learning Centers: Pathways to Progress Project, St. Paul Public Schools. St. Paul, MN: Center for Applied Research and E ducational Improvement, University of Minnesota. Walker, K. E., Arbreton, A. J. (2004). After-school pursuits: An examination of outcomes in the San Francisco Beacons Initiative. Oakland, CA: Public/Private Ventures. Welsch, M. E., Russell, C. A., Williams, I., Reisner, E. R., & White, R. N. (2002). Promoting learning and school attendan ce through after-school programs: Student-level changes in educational pe rformance across TASCs first three years. Washington, D.C.: Policy Studies Associates, Inc. Young Mens Christian Associat ion (YMCA) of Tampa Bay We bsite (n.d.). Retrieved October 5, 2003 from www.tampaymca.org Yu, P., Harris, G. E., Solovitz, B. L., & Fr anklin, J. L. (1986). A social problem-solving Intervention for children at hi gh risk for later psychopathology. Journal of Clinical Child Psychology, 15 30-40. 103
Appendix 1 Social Skills Rating Form Childs Name: _____________________ Staffs Name: ______________________ Directions: Please fill in the following information on the child listed above. To fill out this form, you will only need to write one number in each of the grey boxes below. In each box, put a number 1 through 5 which describes how of ten the child shows the behavior listed. 5 = Always 4 = Frequently 3 = Sometimes 2 = Infrequently 1 = Never How often does the child above Pay attention. Sit in his seat properly Stays in assigned area (e .g., seat) when expected. W h en an adult speaks to the ch ild, the child looks at the adult W h en sitting down, has his buttocks completely on the seat and squarely faces the tab l e. During hom ework tim e, the child stays in his seat unless given perm ission to get up. Look at his hom ework during hom e work tim e. Keep all four legs of his chair on the ground. The child does not get up from his seat w ithout perm ission. W h en an adult speaks to a group containing the child, the child looks at the adult. Keeps both legs facing forwar d when sitting in his seat. W h en working on hom e work, the child does not get up out of his seat to pick up objects, talk to friends, or appro ach th e coach withou t raising his hand and asking first. Thank you for your participation. 105
Appendix 2 Data Collection Initials: Date: Observer:_____ A = Attending (quiet, looking at work, in seat) (momentary time sampling) N = Not seated corrected; S = S eated correctly (mark no letter if child is not in seat area) / = Not attending and not in seat :10 :20 :30 :40 :50 :00 :10 :20 :30 :40 :50 :00 = Left seat; + = Left seat after raising hand Raised hand before leaving area/seat. Attending: _____/_____. Seat ed correctly: _____/_____. Ra ising Hand: _____/_____. Data Collection Initials: Date: Observer:_____ A = Attending (quiet, looking at work, in seat) (momentary time sampling) N = Not seated corrected; S = S eated correctly (mark no letter if child is not in seat area) / = Not attending and not in seat :10 :20 :30 :40 :50 :00 :10 :20 :30 :40 :50 :00 Raised hand before leaving area/seat. = Left seat; + = Left seat after raising hand Attending: _____/_____. Seat ed correctly: _____/_____. Ra ising Hand: _____/_____. 106
Appendix 3 Interview Guide Beha vior Identification Date: 7 December, 2004 Interviewer: Robert Caples Interviewees: Various staff. This interview guide is to be used when interviewing staff regard ing the selection of target behaviors for this study. The interview will be conducted with the entire staff at one time. 1. Explain purpose of selecting behaviors spec ific to this partic ular site. Explain criteria for selection of behaviors. This includes social validity (importance to staff), frequency, opportunity to observe in as many settings within the program as possible, and behaviors which are easily observable and measurable. This was done. 2. Ask group to brainstorm behavi ors. List behaviors below: Not following directions, paying attenti on during homework, talking back, getting upset easily, self-control, getting out of seat, hands on others or hitting, lying, sitting in seat properly 3. Ask group to narrow down this list of behaviors according to criteria above. Interject observations about the ability of the selected behaviors to meet the criteria above. List the fi nal three behaviors below: 1. Attending 2. Staying in seat 3. Sitting in seat properly 4. Provide the group with an explanation of behavioral ly descriptive terminology. Ask the group to write the three behavi ors selected above in behaviorally descriptive terms. 1. Eyes on staff, mouth closed, body still. 2. The child keeps his buttocks on the seat. 3. The child keeps his whole buttocks on the seat, faces forward, chair aligned to desk, and all four chair legs on the floor. 5. Provide the group with an e xplanation of replacement be haviors. Ask the group to write the three replacement behaviors fo r the target behaviors above, providing both examples and non-examples for each behavior. 1. Attending child is looking at approved ma terials during homework time. If staff is speaking to the target child or to the group, the child is also considered attending if looking at th e adult. Examples include looking at homework, looking at coloring sheet if approved, looking at adult when 107
being given an individual instructi on, and looking at an adult when the adult is giving an instruction to the entire group. Non-examples include looking at a piece of paper that is not approved by the staff member and looking at an adult who is spea king individually to another child. 2. Stay in Seat child keeps buttocks in chair w ith the chair located in the assigned area; child stands next to chair in appropriate area. The child may leave with permission. The main goal is for the child to stay in his assigned area. It does not matter how the child sits in his chair or stands in his area. Examples include sitting in the chair at the desk and standing behind ones chair. Non-examples in clude standing behind another childs chair, sitting in ones chair when th at chair is placed in an inappropriate location, and getting up without permission. 3. Sitting in Seat Properly child keeps entire buttocks on seat with the chair facing forward, all four chair legs on ground, and legs in front of chair. Examples include sitting in seat wi th legs facing forward and chair facing forward with all four chair legs on the ground. Non-examples include sitting with legs hanging off the side of the chair, sitting with half of ones buttocks on the seat, leaning back in one s chair, and sitting in a chair not facing directly toward ones table. 108
Appendix 4 Interview Guide Participant Selection Date: 7 December, 2004 Interviewer: Robert Caples Interviewees: Staff at after-school program site. This interview guide is to be used when interviewing staff regard ing the selection of participants. The interview will be conducte d with the entire staff at one time. 1. Explain the criteria for selecting participants. This criteria involves the low frequency of the occurrence of the repl acement behaviors previously identified using the interview guide in Appendix 3. Part icipants also need to be of the same ethnic/racial group, within any 2 year span (for example, 9-10-year-olds) inside the parameters of the 8-11-year-old range, and of English-speaking background. This was done. 2. Remind the group of the three replacem ent behaviors selected. Ask group to brainstorm potential participants who th ey believe might exhibit low levels of these behaviors. List these names below. Ten names were recorded here. 3. Ask group to narrow down this list of poten tial participants acco rding to criteria above. Interject observations about the ability of the selected pot ential participants to meet the criteria above. List the final eight potential participants below: 1. 5. 2. 6. 3. 7. 4. 8. Eight names were recorded here. 4. Go over each selected potential participant a final time and confirm that staff agree that each potential participant is likely to meet the criteria listed above. In addition, ask staff a final time if there are any other likely potential participants. This was done. 109
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001709496
007 cr mnu|||uuuuu
008 060509s2005 flu sbm s000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001351
Effects of social skills training on the interpersonal behaviors of elementary school students in an after-school program
h [electronic resource] /
by Robert Caples.
[Tampa, Fla.] :
b University of South Florida,
Thesis (Ed.S.)--University of South Florida, 2005.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
Title from PDF of title page.
Document formatted into pages; contains 117 pages.
ABSTRACT: Social skills training was investigated in an after-school program setting with four seven- and eight-year-old males. Two were Hispanic and two were African-American. Social skills training consisted of a direct instruction, behavioral learning model of skillstreaming as described by McGinnis and Goldstein (1997). There were four major components to each social skills training session: (1) an explanation of the skill being taught; (2) modeling by the researcher of the skill being taught; (3) role play by each of the participants; and (4) performance feedback regarding the role plays. Sessions lasted approximately 30 minutes and were held weekly throughout the intervention phases of the study. The behaviors taught were raising ones hand before leaving the seat, sitting properly in ones seat, and attending to homework or staff instructions.Participants also received reinforcement for performance of the social skills in homework sessions at the after school program, as is consistent with the literature regarding social skills training. However, the reinforcement and behavior learning (direct instruction) components were introduced both in combination and at separate times to experimentally control for the influence of each intervention component. This research design allows for the investigation into the relative effectiveness of direct instruction versus reinforcement in social skills training. Experimental control was demonstrated through the use of a multiple baseline across behaviors design. Direct instruction and reinforcement for behaviors were systematically introduced at separate times, keeping some behaviors under baseline condition while moving others into intervention conditions.Visual analysis of the results indicates that social skills training was effective in improving the three target behaviors of all four students. Direct instruction, reinforcement, and the combination of the two presented together all were effective in improving the target behaviors. Possible intervention effects not related to social skills training may have influenced the behavior of attending.
Adviser: George Batsche.
x School Psychology
t USF Electronic Theses and Dissertations.