Effects of a Token Economy System Within the Context of Cooperative Games on Social Behaviors of Adolescents with Emotional and Behavioral Disorders

THERAPEUTIC RECREATION JOURNAL Vol. 37, No. 2, 124-141, 2003

Effects of a Token Economy System within the Context of Cooperative Games on Social Behaviors of Adolescents with Emotional and Behavioral Disorders

Brent D. Wolfe, John Dattilo, and David L. Gast

A single subject A-B1-BC1-B2-BC2 design was used to examine effects of a cooperative games and token economy program on pro-social (statements of encouragement, statements of praise, statements of appreciation, hand shakes, or high fives) and anti-social behaviors (swearing, threatening physical harm, calling peers names, hitting or pushing) of three adolescents with emotional and behavioral disorders. Results are reported as group means. With the introduction of cooperative games, behaviors remained stable and unchanged. When a token economy system was introduced there was an immediate increase in the number of pro-social behaviors. Results for pro-social behaviors were replicated across conditions. Anti-social behaviors remained virtually non-existent throughout the study. Limitations and implications for research and practice are discussed.

Brent Wolfe is a doctoral student and John Dattilo is a professor in the Department of Recreation and Leisure Studies and David Gast is a professor in the Department of Special Education at the University of Georgia. This research was completed as a requirement for the first author's master's degree from the University of Georgia. Appreciation is extended to Dr. Douglas Kleiber, the Rutland psycho-educational center, Clay Crowder, and MC Tate for their assistance. We also thank the study participants and their families for their cooperation. Correspondence concerning this article should be addressed to Brent Wolfe, Department of Recreation and Leisure Studies, The University of Georgia, Athens, GA 30602. Email: bwolfe@coe. uga. edu

124 Therapeutic Recreation Journal KEY WORDS: Token Economy, Cooperative Games, Adolescents, Single-Subject Research, Therapeutic Recreation

Adolescents with emotional or behavioral (Walker et al., 1995), social skills are neces- disorders (E/BD) typically lack adequate so- sary for daily functioning (Hansen, Nangle, & cial skills (Walker, Colvin, & Ramsey, 1995). Meyer, 1998), and previous social skills pro- These adolescents' problems with social be- grams have demonstrated mixed results haviors are manifested in the form of aggres- (Moote et al., 1999; Quinn et al., 1999), addi- sion, acting out behaviors (i.e., not following tional research examining strategies designed directions), complacence, and noncompliance to promote social skills with youth with E/BD across a variety of contexts (Coie, Underwood, are needed. & Lochman, 1991; Phillips, Schwean, & Sak- lofske, 1997). Although social skills teaching Cooperative Games strategies have been developed and imple- Cooperative learning activities such as mented for adolescents with E/BD with mixed "warp speed" (passing a small ball around a results (e.g., Evans, Axelrod, & Sapia, 2000; circle that requires participants to work to- Loy & Dattilo, 2000; Mathur, Kavale, Quinn, gether) and "group knot" (standing in a circle Forness, & Rutherford, 1998; Moote, Smyth, and attempting to untangle the knot created by & Wodarski, 1999; Quinn, Kavale, Mathur, grasping another person's hands to allow par- Rutherford, & Forness, 1999), practitioners ticipants to demonstrate teamwork and coop- have traditionally considered these programs eration skills) can provide the context for so- to be effective (Berliner, 1992). cial skills to be encouraged and developed. Studies have demonstrated that individuals According to Cartledge and Cochran (1993), who lack effective social skills are more apt to individuals with E/BD are especially in need have difficulties succeeding in school (Hughes of "... activities that contribute to and rein- & Sullivan, 1988), have low self-esteem (El- force interpersonal social skills" (p. 10) such liott & Gresham, 1993), experience peer rejec- as cooperative games. tion and social maladjustment (Farmer & Hal- One benefit of cooperative learning activities lowell, 1994), become delinquent (Asher & is the positive effects they can have on people Wheeler, 1985; Elliott & Gresham, 1993), and who commonly experience failure (Harris & engage in alcohol and substance abuse Covington, 1993). Rutherford, Mathur, and (Wodarski & Feit, 1995). In addition, Swag- Quinn (1998) demonstrated that adolescents gart (1998) concluded that adolescents with with E/BD could be taught social skills in a E/BD engaged in more off-task than on-task cooperative learning context that included behaviors, were inattentive, uninvolved, im- face-to-face interactions and group processing. pulsive, and distractible. Adolescents with Cooperative games are a specific type of E/BD tend to demonstrate behavior patterns cooperative learning activity designed to in- that are problematic for successful interaction crease personal interaction, promote coopera- with society in settings requiring social skills tion with others rather than competition, and to (Farmer & Hollowell, 1994; Schonert-Reichl, foster a sense of teamwork while attempting to 1993; Steinberg & Knitzer, 1992). There is achieve a common goal (Carlson, 1999; Or- "Perhaps no greater challenge faces the clini- lick, 1981). "Cooperative games address the cal and educational communities than to help motivations of children for game and sport. troubled adolescents learn to cope with their They de-emphasize winning and competition, angry feelings in a socially appropriate man- are not aggressive, and are invested in fun ner" (Kellner & Bry, 1999, p. 645). Since experiences" (Carlson, 1999, p. 232). The fo- many adolescents with E/BD lack social skills cus of cooperative games on achieving goals

Second Quarter 2003 125 rather than defeating an opponent can lead to clined toward baseline levels. The authors pro- positive experiences for individuals who par- posed that "Perhaps the effectiveness of inter- ticipate (Roberts & Treasure, 1992). Ramsey ventions similar to the one implemented in this and Rank (1997) observed that cooperative study may be enhanced if... activities de- games can promote socialization such as shar- signed to decrease anti-social behaviors are ing, taking turns, and helping others and included" (p. 138). One way to structure a stated, "[r]esearch on the use of cooperative context, as suggested by Loy and Dattilo games has shown increased personal self-es- (2000), and to increase pro-social behaviors teem, decreased aggressive behaviors, in- and decrease anti-social behaviors, as sug- creased cooperation, and enhanced positive gested by McKenney and Dattilo, is to imple- socialization. In each case, cooperative games ment a token economy system. allowed youth to build social skills and self- images that permitted them to progress as Token Economy effective contributing members of society" (p. 35). A token economy is "an exchange system that provides individuals whose behavior is According to Loy and Dattilo (2000), it being changed with near-immediate feedback does not appear that presentation of a positive cues on the appropriateness of their behavior" context to develop social skills, such as that (Bauer & Shea, 1999, p. 225). Tokens, other- created by cooperative games, is sufficient for wise known as "exchangeables," are reinforc- accelerating social skills. Using an alternating ers that people earn for accomplishing a target treatments design, Loy and Dattilo compared behavior for a pre-determined period of time effects of free play, cooperative games, and (Rusch, Rose, & Greenwood, 1988). These competitive games, on social interactions of a tokens hold no real value except that they can boy with Asperger's Syndrome. Although co- be exchanged for a tangible reinforcer or priv- operative games were found to be the "best ilege reinforcer (e.g., candy, clothing, trading treatment" evidenced by high frequencies of cards, and television viewing time). Milten- positive interactions, these games were also berger (1997) defined a reinforcer as "a stim- associated with an increase in negative social ulus or event that will increase the future interactions. Therefore, Loy and Dattilo (p. probability of a behavior when it is delivered 208) stated that "Although cooperative games contingent on the occurrence of the behavior" provided a context for positive social interac- (p. 566). If a particular stimulus or event does tions, structuring such a context so that nega- not result in an increase in a target behavior, tive interactions are discouraged and positive the stimulus is not considered to be a rein- ones are promoted is advised." forcer (Maag, 1999). In a similar study McKenney and Dattilo (2001) used a single subject multiple baseline across behaviors design to observe pro-social Steps in Developing a Token and anti-social behaviors of five adolescents Economy during a sport context intervention. Through- Identify target behaviors. The first step in out the study participants were observed for establishing a token economy is to identify the 20-minute basketball scrimmages and pro-so- behavior to be changed, otherwise known as cial and anti-social behaviors were recorded. defining the "target" behavior (Martin & Pear, The intervention involved social skills instruc- 1999). According to Miltenberger (1997), a tion where the researcher taught the partici- target behavior is the specific behavior that is pants specific pro-social skills. With interven- to be modified through reinforcement in the tion initiation, immediate increases in pro- token economy. For an action to be identified social behaviors were observed; however, as a target behavior it must be observable, during follow-up, pro-social behaviors de- measurable, and changeable (Flick, 2000).

126 Therapeutic Recreation Journal Identify tokens. The second step in devel- of 100 on an intermittent schedule, averaging oping a token economy is to identify the to- five days, for any of or a combination of four kens to be used. Miltenberger (1997) stipu- back-up reinforcers" (p. 13). These steps in lated that the tokens must be tangible, be developing a token economy have been docu- provided immediately, and be easily dispens- mented in the literature: establishing back-up able and accumulated. The tokens must also be reinforcers (Pestle, Card, & Menditto, 1998), appropriate for the age of participants (Maag, identifying tokens (Hupp & Reitman, 1999), 1999). and identifying target behaviors (LePage, Establish backup reinforcers. The third 1999). step in establishing a token economy is to determine items for which participants will Relevant Studies of Token exchange their tokens, ("backup" reinforcers) (Miller, 1997; Maag, 1999). "Backup" rein- Economies forcers should only be available when partic- Pestle et al. (1998) examined effects of a ipants engage in the target behavior and should social learning program on appropriate behav- be meaningful to participants. iors of three women and three men (ages Establish rate of token presentation. An- 37-57) diagnosed with schizophrenia. A Cer- other step in development of a token economy tified Therapeutic Recreation Specialist con- system is establishing the rate of presentation ducted activity groups (40 min/sessions, 3 X of tokens when participants are to exchange wk) and participants received two tokens for tokens for backup reinforcers and how often attendance and two tokens for participation. this exchange occurs (Miltenberger, 1997). Tokens were exchanged for items from the Participants received and lost tokens on a con- hospital store, off campus outings, and later tinuous schedule every time they engaged in sleep times. Group time consisted of the social one of the pro-social or anti-social behaviors learning program used during physical (vol- (De Martini-Scully, Bray, & Kehle, 2000; leyball, kickball, and soccer) and non-physical Maag, 1999). Removal of tokens was included (trivia, arts and crafts, and board games) ac- in the design in concordance with suggestions tivities. The authors reported that "structured by McKenney and Dattilo (2001) that social recreational activities produced more appro- skill interventions include a component "de- priate behavior..." (p. 38). While this study signed to decrease anti-social behaviors" (p. examined effects of a social learning program 138). on appropriate behaviors in a leisure context, it did not examine inappropriate behaviors. Establish rate of token exchange. Variation exists in the literature concerning when and In a similar study, Hupp and Reitman how often the exchange of tokens for backup (1999) used a combined multi-component reinforcers should occur. Cyclical trends relat- skills and behavioral management program to ing to target behaviors may occur due to when improve sports skills and sportsmanship in participants are allowed to exchange tokens. three youth (8-11) with E/BD. A multiple Morisse, Batra, Hess, Silverman, and Corrigan baseline across participants' design was used (1996) developed a system where participants to examine effects of the training component exchanged tokens for backup reinforcers twice and an A-B design to evaluate the behavior a day, while Reese, Sherman, and Sheldon modification sportsmanship component. Dur- (1998) created a system where an exchange ing the sportsmanship phase, participants were occurred every time a participant earned ten awarded "B-ball bucks" for engaging in posi- tokens. Swain and McLaughlin (1998) and tive behaviors defined as "verbal praise and Truchlicka, McLaughlin, and Swain (1998) encouragement directed toward another player constructed an exchange program for partici- or coach, non-directive positive statements, pants where "points were exchanged in blocks and physical signs of encouragement" (Hupp

Second Quarter 2003 127 & Reitman, p. 42). There was no contingency zines, games, or special attention. The authors for negative behavior. The authors found that selected these reinforcers by observing the the token system immediately increased par- participant's preferences during free time. Re- ticipants' positive behaviors and improved ese et al. suggested that allowing the partici- their dribbling, but shot-making performance pant to earn different rewards served an im- was not improved. While a token economy portant role in reducing disruptive behaviors system was used as the primary intervention and allowing the token economy system to during part of this study, due to time con- function effectively. straints the authors were unable to use a reversal design that could have allowed for identi- fication of stronger connections between Theoretical Framework and behaviors and intervention. Statement of Purpose In addition, Franco, Galanter, Castaneda, The study was grounded in behavioral the- and Patterson (1995) combined a token econ- ory (Watson, 1913) as it relates to develop- omy and self-help approaches to influence be- ment of reinforcement strategies. "The process haviors of adults with dual diagnoses who of reinforcement can be conceptualized as a resided in an acute care ward. The token econ- sequence of two behaviors: (1) the behavior omy was conducted in conjunction with ther- that is being reinforced, followed by (2) the apeutic group meetings and recreational activ- behavior that is the reinforcer" (Powell, Sym- ities. While no statistical information was baluk, & McDonald, 2002, p. 257). The rein- presented, Franco et al. stated that according to forcement system used in this study was spe- staff documentation "as the token economy cifically based on the Premack principle became an integral part of the ward routine, (1959, 1962, 1965) stipulating that a high violence on the ward declined substantially, probability behaviors can be used to reinforce and at the same time, patient participation in a low probability behaviors. The Premack group activities doubled in less than a month" principle requires that observations occur in a (p. 228). The authors stated that the token situation where all relevant behavior can occur economy system was removed for one week without restriction and then these observations and inappropriate behaviors dramatically in- are used to predict that any higher frequency creased. Since little data were reported, it is behavior may serve as reinforcement for any not possible to determine what accounted for lower frequency behavior (Pierce & Epling, behavior changes. The authors suggested that 1999). before programs such as this are instituted, Through observations in the classroom set- more studies must be conducted to evaluate ting used in this study, it was determined that the impact of token economies on social skills. behaviors such as playing with and exchang- Reese et al. (1998) used a reversal single- ing Pokemon® cards, eating snacks, and subject design to examine effects of program drinking soft drinks were high frequency be- designed to reduce disruptive behaviors of a haviors while pro-social behaviors were low man (age 26) with severe to moderate mental frequency behaviors. Following the Premack retardation and autism who demonstrated most principle, these high probability behaviors disruptive behaviors during his free time. Us- were identified as reinforcers for the lower ing a differential reinforcement schedule of probability pro-social skills and, thus, engage- other behaviors (DRO), rewards were pro- ment in these activities was made contingent vided when the participant did not engage in on pro-social behaviors. Therefore, the pur- one of the target behaviors for a pre-deter- pose of this study was to examine effects of a mined time, and he lost a token when he 12-week cooperative games program used exhibited a target behavior. Tokens were ex- with a token economy system applied during a changed for gum, writing materials, maga- regular school day on verbal and physical

128 Therapeutic Recreation Journal pro-social and anti-social behaviors of adoles- spectful behaviors (e.g., calling out during cents with E/BD. Tokens that represented the class), and impulsive behaviors, especially low probability pro-social behaviors were used with female authority figures. When on the by participants to acquire the high probability playground, Paul pushed and verbally threat- behaviors. ened to physically harm other students. He engaged in physical activities in which he is Methods proficient, but quits or becomes verbally and physically aggressive when someone is better Participants than he, or when he loses. Paul has had recent Data were collected on three adolescents outbursts involving hitting, tripping, and from Northeast Georgia (ages 12-14) identi- throwing objects at his peers. fied by the school system as having E/BD. John. John was a 12-year-old male who Also included in the study were six other exhibits verbally and physically aggressive be- youth whose Individualized Education Pro- haviors towards classmates and teachers. John grams (IEP) contained information suggesting could be polite as indicated by him saying they had characteristics consistent with E/BD "please" and "thank you," asking for help, and and that they may benefit from being involved accepting help from other students; however, in structured activities. The six additional at times, John showed disrespect toward teach- youth were included in the study to allow for a ers by refusing to follow instructions and by sufficient number of people to be involved in using excessive profanity. John reported that the cooperative games (more than three were he enjoys team sports and has demonstrated needed). The nine youth were involved in all teamwork during free time. conditions of the study. Bill. Bill was a 14-year-old male who dem- Design onstrated marked changes in behaviors in short A single subject A-BI-BC1-B2-BC2 with- periods of time. He could be quietly working drawal design with baseline (free play) (A), or playing and quickly become physically and cooperative games (B^, cooperative games verbally aggressive. Bill stated that he heard and token economy system (BCj), cooperative voices (specifically of his dead grandmother) games (B2), and cooperative games and token and has verbally responded to these voices. He economy system (BC2) replicated across par- has been involved in episodes of sexual mis- ticipants was used to measure effects of an conduct (revealing himself to others), non- intervention on pro-social and anti-social be- compliance (walking from the classroom and haviors. The "ABAB design permits the most refusing to return), verbal and physical aggres- powerful demonstration of experimental con- sion, and making obscene gestures. Bill stayed trol because it requires the repeated introduc- to himself during most activities but partici- tions and withdrawal of an intervention strat- pated in one-on-one activities with adults. In egy" (Tawney & Gast, 1984, p. 200). Threats the classroom he refused to respond to direc- to internal validity of a single subject with- tives, was often off task, and disrupted the drawal design include: history, testing, instru- class by yelling. Bill has received speech ser- mentation, and multiple treatment interference vices to correct a slight stutter. (Dattilo, Gast, Loy, & Malley, 2000). Paul. Paul was a 12-year-old male who has Since single subject designs occur over a history of being severely socially aggressive time, they are likely to be influenced by out- as evidenced by numerous physical assaults on side events in participants' lives. Often with a students. He has demonstrated verbally and token economy system there is a "pay day" physically threatening behaviors (e.g., invad- when participants spend tokens they earned. ing others' personal space with fists raised Participants' behaviors may improve on days while cursing at them), disruptive and disre- when they receive back-up reinforcers on this

Second Quarter 2003 129 "payday." To reduce the chance of cyclical Cooperative Games Condition (Bj) changes such as those that occur with a "pay The first author was the instructor for the day," paydays occurred in this study each day cooperative games. Each day when partici- participants engaged in the token economy pants arrived, the planned activity was ex- system. plained and questions were answered. Safety When tests are administered numerous considerations were delineated and partici- times, there is likely to be improvement due to pants verbally committed to understanding familiarity with the test as opposed to actual safety precautions as well as the session plan. material comprehension. Since single subject During the activity minimal interaction oc- designs are essentially repeated pretests, they curred between the instructor and participants. are apt to be influenced by testing. To limit this The instructor interacted with participants to threat, the least intrusive measurement and praise appropriate behaviors, keep them safe, observation methods were used. Before the answer their questions, and did not intervene study began, the first author visited the class- unless it was for reasons of physical or emo- room 3 X week (30-45 min) for two weeks. tional safety. Upon completion of the activity, During this time he observed participants from the instructor asked different versions of the a distance as the teacher conducted regularly following questions: (a) what did we just do, scheduled activities. These observations helped (b) why did we just do it, and (c) how does it participants become accustomed to the pres- apply to me. All cooperative games were phys- ence of the first author and were designed to ical in nature and selected for their ability to elicit social interactions between participants. reduce participants' perception of a testing situation during data collection. An additional threat to internal validity, Cooperative Games and Token multiple intervention interference, exists when Economy System Condition (BC!) more than one intervention is presented within Procedures followed during BC, were a single study. When multiple interventions identical to those followed in other conditions are present, it is impossible to determine (e.g., the length of sessions, activities pre- whether change resulted from the intended sented) except for initiation of the token econ- intervention or other factors such as informa- omy system. Participants arrived and were tion learned in the classroom relating to social informed that they were starting a new part of skills. Within this ABAB withdrawal design, the program. The instructor had in his pocket, there were two interventions, cooperative token bills similar in appearance to dollar bills. games (B) and the token economy system in On a nearby bench there was one clear plastic conjunction with cooperative games (BC). To Tupperware® container for each participant prevent multiple intervention interference, the labeled with his name that functioned as a bank. At the beginning of the intervention, two interventions were examined individually each participant had five one thousand-token and then combined. bills in his container to serve as a visual reminder of the reinforcer and prevent partic- Experimental Conditions ipants from creating a "negative balance" if they engaged in negative behaviors before Free Play Condition (A) they engaged in positive behaviors (to earn During baseline, participants engaged in 35 more token dollars). Based on suggestions by minutes of unstructured free play when they Naughton and McLaughlin (1995), each time participated in various recreation activities the instructor observed a participant engaging (e.g., basketball, Frisbee, 4-square, kite flying) in one of the pro-social behaviors, he handed a bill to the participant. Similarly, each time the in an outside play area connected to the school.

130 Therapeutic Recreation Journal instructor observed an anti-social behavior, a opportunity to record any target behavior ob- token was removed from the participant's served during any portion of the interval (Dat- Tupperware container. tilo & Wolfe, 2002).

Withdrawal of Intervention #2 Dependent Variables (B ) 2 Pro-social verbal behaviors. Based on the Participants engaged in cooperative games work of Miller and Cole (1998) and Richman, in the same manner that they did during the Wacker, Asmus, Casey, and Andelman previous cooperative games condition without (1999), pro-social verbal behavior was defined the token economy system. as (a) making statements of encouragement to a peer or adult (e.g., "good try," or "better luck Reintroduction of Intervention #2 next time"), (b) making statements of praise to (BC2) a peer or adult (e.g., "nice job," "good work"), During the final condition, the token econ- and (c) making statements of appreciation to a omy system was reintroduced and participants peer or adult (e.g., "thank you," "I really again received tokens for pro-social behaviors enjoyed this"). Also vital to pro-social verbal and had tokens removed for anti-social behav- behavior is the manner in which words are iors. Because the intervention was conducted spoken. According to Hughes (1999), a re- during the second semester of the school year, sponse is "scored appropriate if the volume, the length of the study was determined by the tone, and quality of voice, pitch, intensity, school schedule. intonation, rate, and topic approximate stan- dards established by social comparison to Data Collection peers" (p. 44). For the observer to score a Data were collected using a partial interval participant as having engaged in a pro-social recording method of observing behaviors that verbal behavior, the participant needed to ver- required the 15-minute session be divided into balize appropriate words and adhere to the 45 15-second intervals with 5-second periods guidelines established by Hughes during the when the observer prepared for the next ob- observation time interval. servation interval. If a participant engaged in Pro-social physical behaviors. Based on the target behaviors during his assigned obser- Gunter, Shores, Jack, Denny, and DePaepe's vation period, a mark was made for that be- (1994) work, pro-social physical behaviors havior in that time interval. The primary ob- were defined as "purposeful contact that dis- server examined participants on a 20-second plays affection (e.g., pats on the back, "high rotating basis, allowing each participant to be fives," or hand shakes) that is not. . . teasing observed approximately 10 times per session. or aggressive" (e.g., bear hugs, loud smacks on Along with recording the occurrence of a be- the back, missing a "high five," squeezing the havior, the primary observer recorded the to- other person's hand when shaking it) (p. 438). pography of the behaviors. In the space on the Anti-social verbal behaviors. Anti-social observation form next to the participant's in- verbal behaviors were defined based on sug- terval, the primary observer wrote the specific gestions by Zanolli, Dagget, Ortiz, and Mul- behaviors that occurred to provide detail about lins (1999), Harding, Wacker, Berg, Cooper, the behaviors. Data points on the graphs rep- Asmus, Mlela, & Muller (1999), and Posavac, resent mean scores across the three partici- Sheridan, and Posavac (1999) and include (a) pants for each day. While the interval record- swearing (e.g., hell, damn, shit, asshole) (b) ing system has the potential for the observer threatening physical harm to other participants not to record some behaviors that occur, this (e.g., "I'm gonna beat the shit out of you" or system of observation provided observers the "I'm gonna kick you I your face if you don't

Second Quarter 2003 131 get away from me"), and (c) calling peers were conducted. Based on suggestions by names (e.g., "idiot," "moron," or "faggot"). Tawney and Gast (1984), inter-rater reliability Anti-social physical behaviors. Based on data were collected on eleven separate occa- suggestions by Durand (1999), Richman et al. sions (34% of the conditions) (sessions #2, #3, (1999), and Piazza, Bowman, Contrucci, De- #5, #10, #11, #12, #13, #16, #20, #23, and lia, Adelinis, and Goh (1999), anti-social #32) with overall reliability of 91%. During B2 physical behaviors were defined as making (withdrawal), reliability data were not col- contact with another person without their per- lected. Reliability scores were calculated using mission by (a) hitting another individual or (b) the point-by-point, nonoccurrence agreement, pushing another individual. During the pilot- and occurrence agreement methods. ing of the observational recording system, par- The point-by-point method was used to ticipants primarily engaged in the above anti- determine reliability between the primary ob- social physical behaviors. While anti-social server and the secondary observer. This physical behaviors could include more con- method was conducted by examining data cepts than the two listed, for purposes of this from each observer, summing all instances study, anti-social physical behaviors were only where observers agreed, dividing that number denned as hitting or pushing. These two be- by the sum of the number of agreements and haviors were selected based on problems ob- the number of disagreements, and multiplying served by the first author prior to initiation of the quotient of those two numbers by 100 to data collection and confirmed by the class- calculate percent of agreement. room teacher. The nonoccurrence and occurrence methods were used to examine instances where Inter-Rater Reliability behaviors occurred more than 75% of the time and less than 75% of the time, respectively. Training. Two observers (primary and sec- The nonoccurrence method examined only in- ondary) were given a verbal description and tervals where the observers viewed a nonoc- task analysis for each activity, were trained to currence, while the occurrence method only recognize all participants by name, and were examined intervals where the observers taught to recognize identified pro-social and viewed an occurrence. These two methods anti-social behaviors. The primary observer allowed for the most conservative reliability recorded behaviors during every session while scores on the basis of occurrence or nonoccur- the secondary observer recorded behaviors to rence. provide reliability data during 34% of the sessions. Both observers were trained in the behavioral recording procedure and asked to Procedural Reliability observe 20-minute segments of participants in Based on by Wolery and Holcombe (1993), the classroom and record behaviors on the procedural reliability data were collected behavioral recording forms according to the across all conditions in the study. Overall target definitions. The observers achieved pro- procedural reliability across all conditions was ficiency when they achieved an 80% agree- 95%, and data were collected during the fol- ment of behaviors on three consecutive occa- lowing sessions: #3, #6, #8, #9, #14, #16, #22, sions (Dattilo & Hoge, 1999; Kazdin, 1982). #23, #25, #26, and #30. Different procedural Observations. A potential threat to internal reliability forms existed for different study validity is instrumentation and observation conditions and included criteria such as: (a) (Dattilo et al., 2000). Over the course of time introducing the session, (b) explaining the ac- it is possible for the observer to drift in the tivity, (c) handling of questions, (d) initiating interpretation of definitions within the study. the activity, (e) providing positive feedback, To assess observer drift, reliability checks (f) asking concluding questions, (g) introduc-

132 Therapeutic Recreation Journal ing the token economy, and (h) providing or agreed with was in support of incorporating a removing tokens as necessary. program similar to the one used in this study into daily school programs. Social Validity It is important that individuals other than Results the primary researcher validate aspects of the current study. As posited by Baer, Wolf, and Pro-social Verbal Behaviors Risley (1968), it is vital to examine "behaviors Although data, as reported by mean scores, which are socially important, rather than con- were variable during BC, and BC2, there were venient for study" (p. 92). Social validation immediate level changes with initiation of data were collected on appropriateness of ob- BCi, B2, and BC2. Lower means were objectives, importance of intervention, and effec- served for Bj and B2 when compared to the tiveness of intervention (see Table 1). higher scores of BC! and BC2 and no overlap Social appropriateness of objectives was between these conditions existed. Therefore, assessed via Likert scale (range 1-5, more during baseline and B^ the group's pro-social positive results associated with lower num- verbal behaviors were low, with initiation of bers) instruments completed by participants' BCl5 these behaviors increased, when BC! parents, and teachers. Teachers rated and com- was removed during B2, behaviors returned to mented about the cooperative games compo- baseline levels and when the tokens were re- nent, the token economy intervention, and initiated in BC2 behaviors again increased dependent measures. Mean scores for profes- demonstrating replication of effects during sionals' questionnaire on the intervention ap- baseline and intervention (see Figure 1). When propriateness (range 1.33-1.88) were all fa- observing each individual's scores, data points vorable toward the use of the intervention in were found to mimic the group means with no context of cooperative games on the assigned obvious outliers. dependent variables. Parents were asked to provide feedback on the social importance of the program by com- Pro-social Physical Behaviors pleting an 11-item, five point Likert scale with Although data, reported as mean scores, one being the highest rating. The questionnaire were variable during BCi and BC2, there were assessed current behaviors of participants and immediate level changes with initiation of whether those surveyed felt the need for sim- BCj, B2, and BC2 and lower means were ilar continued programming. Parents' re- observed for Bj and B2 when compared to the sponses to the questions relating to importance higher scores of BC! and BC2 and low overlap varied but were mostly favorable with means between these conditions. Therefore, during A between 1.4 and 3. The lowest score was and Bj, the group's pro-social physical behav- related to their children mentioning some as- iors were low, with initiation of BCj, these pect of the program when at home and the behaviors increased, when BCi was removed highest scores were related to importance for during B2, behaviors returned to baseline lev- children getting along with one another and els and when the tokens were re-initiated in the use of a token store to purchase items. BC2 behaviors again increased demonstrating In response to questions relating to effec- replication of effects during baseline and in- tiveness of the intervention, parents' responses tervention (see Figure 2). It is important to were less favorable than questions relating to note that during BC2 the group mean may not importance (range 1.6-3). The lowest score have presented an accurate picture of all par- was associated with parents feeling their chil- ticipants. During this condition, Bill engaged dren had not improved as a result of the in a total of two pro-social behaviors while program (M = 3), and the question most John engaged in nine.

Second Quarter 2003 133 Table 1. Social Validity Questions Asked and Mean Participant Responses

Mean Appropriateness Scores

Objective 1: Use of cooperative games 1.33* Objective 2: Use of a token economy system 1.83 Objective 3: Promoting pro-social behaviors 1.83 Objective 4: Reducing anti-social behaviors 1.83 Importance My child needs to be taught social skills. 2.6 Leisure/recreation is important in my child's life. 1.6 Leisure/recreation activities can teach my child how to behave better. 2.2 Providing rewards is a good way to teach my child social skills. 2.4 Having a store where my child can exchange his rewards for real items is effective. 1.4 My child could benefit from a program that uses games to teach social skills. 1.6 My child could benefit from a program that uses physical activity to teach social skills. 2.0 My child mentioned some aspect of this program at home. 3.0 It is important that my child have opportunities in school to play with other children his age. 1.6 It is important that my child learn how to get along with other children. 1.4 Effectiveness My child acts in a more socially acceptable manner after this program. 3.0 My child learned social skills as a result of this program. 2.6 Leisure/recreation is more a part of my child's life now. 2.6 Leisure/recreation activities taught my child how to behave better. 3.2 My child gets along with his friends better now as a result of this program. 2.6 I will try to provide rewards to continue to teach my child social skills. 1.8 My child benefited from this program. 2.4 My child mentioned some aspect of this program when at home. 2.6 My child is better behaved when at home. 2.8 Given the opportunity I would have my child participate in a program similar to this if it were offered. 1-8 I would like the psycho-educational center to incorporate this type of a program into the everyday schedule at the school. 1.6

*based on a 5 point Likert Scale (1 = Strongly Agree and 5 = strongly disagree).

Anti-SOCial Verbal Behaviors changes associated with condition changes and hi her means were Although data, reported as mean scores, g observed for B, and B2 were stable during BCX they became variable when compared to lower scores of BCj and during BC2, there were no immediate level BC2. Therefore, during A and B,, the group's

134 Therapeutic Recreation Journal Pro-social and Anti-social Verbal Behaviors

BC, BC,

6 -

4 -

2 '

10 15 20 25 30 35 Consecutive Sessions A Anti-social Verbal • Pro-social Verbal

FIGURE 1. MEAN NUMBER OF PRO-SOCIAL AND ANTI-SOCIAL VERBAL BEHAVIORS OF THREE ADOLESCENTS WITH E/BD. anti-social verbal behaviors were higher, with changes were observed with initiation of BQ, initiation of BCj, these behaviors decreased, B2, and BC2. Therefore, during baseline, the when BCj was removed during B2, behaviors group's anti-social physical behaviors were high, returned to baseline levels and when the to- with initiation of B^ these behaviors increased, kens were re-initiated in BC2 behaviors again when BC! was initiated, anti-social behaviors decreased demonstrating replication of effects dropped to zero and remained there during BQ, during baseline and intervention (see Figure B2, and BC2 demonstrating that replication of 1). Anti-social verbal behaviors were virtually effects did not occur during baseline and inter- non-existent. vention (see Figure 2). Anti-social physical behaviors occurred more frequently during B, than Anti-social Physical Behaviors any other condition. It should be noted that Although data, reported as mean scores, were during this condition, Bill did not engage in any variable during A and B1( no immediate level anti-social physical behaviors.

Second Quarter 2003 135 Pro-social and Anti-social Physical Behaviors

10 15 20 30 35 Consecutive Sessions

Anti-social Physical Pro-social Physical

FIGURE 2. MEAN NUMBER OF PRO-SOCIAL AND ANTI-SOCIAL PHYSICAL BEHAVIORS OF THREE ADOLESCENTS WITH E/BD.

Limitations collection before the end of the school year, the secondary observer recorded behaviors for Participants were observed for 10 15-second intervals each session for a total of 150 these sessions. While inter-rater reliability was seconds of actual observation time. When previously established at 90%, this inconsis- compared to the remainder of the participants' tency in observers is a concern, especially school day, this is a very small period of time. since inter-rater agreement could not be exam- Caution is advised when attempting to gener- ined for B2. alize these results due to the limited duration The study was initially designed to phase of observation. out the intervention during the final condition During withdrawal of the token economy using a series of variable ratio schedules of (B2), the primary observer was unable to at- reinforcement; however, this did not occur due tend two of three sessions. To complete data to time constraints. As a result, the token

136 Therapeutic Recreation Journal system was abruptly terminated when the simply the interaction of the instructor with school year ended. However, findings were participants during cooperative games did not shared with the classroom teacher and sugges- influence behaviors as much as the initiation of tions provided on ways to improve the behav- the token economy. Therefore, the researcher ior of students in the classroom. interacted with participants during four of the conditions including the reversal conditions. Implications for Research Decisions to move to the next experimental Findings of this study are consistent with condition were based on group scores. Blue- previous research demonstrating positive ef- chardt and Shephard (1995) suggested the fects of token economies. Results demon- need for individualized research to explore strated that pro-social verbal and physical be- differences that are lost when performing sta- haviors increased with initiation of a token tistical computations on an aggregate. Because economy system in the context of cooperative effects of the intervention varied slightly games. across the group, future researchers might ac- Due to the complexities of the dependent centuate individualized analysis to learn more variables and participants' behaviors, observa- about how interventions such as the one im- tions using a video recorder are advised. The plemented in this study influence individual current research is potentially limited by the behaviors. use of in vivo observations as opposed to When working with youth with E/BD, video taping interactions during data collec- there is debate concerning teaching individuals tion. The design was used for its simplicity of new social skills versus encouraging them replication for classroom teachers and due to to exhibit previously acquired social skills restrictions placed on videotaping by the (Miller & Cole, 1998). In the current research, school district; however, videotaping sessions since participants possessed the skills neces- rather than relying on in vivo observations sary to engage in pro-social behaviors, they would permit a more thorough examination of were rewarded for demonstrating these behav- dependent variables. In addition, observing iors. Continued examination of programs de- participants for longer periods of time in dif- signed to teach and reward social skills is ferent settings would help to obtain a more warranted. Determining methods to teach new thorough sampling of participant behaviors. social skills as well as reward demonstration Murphy, Hutchinson, and Bailey (1983) of previously learned skills is important. suggested that it is important to demonstrate that decreases in incidents of inappropriate Implications for Practice behaviors during an intervention are not a If practitioners are to use an intervention, it result of increased contact between adults and must be simple enough for them to implement participants. During baseline the researcher within the confines of their day (Phillips et al., did not interact with participants; he stood off 1997). In the classroom where the study was to the side with the primary and secondary conducted there was a token economy system observers. When the cooperative games inter- in place; however, according to the teacher, vention began, the instructor interacted with the existing system was not effective. The participants while data were collected. How- intervention developed for this study involved ever, contrary to concerns of Murphy et al., facilitation by two people and could have been during this condition behaviors did not implemented by one facilitator. TR specialists change. Since behaviors did not change be- could implement a token economy system tween baseline and the cooperative games in- where rewards are distributed to participants tervention and behaviors were similar during (as in the current research) and collected and all conditions that did not employ a token totaled at pre-determined times. In addition, it economy system, support was provided that was necessary to provide back-up reinforcers

Second Quarter 2003 137 to participants that were effective, meaningful, receive one token for every pro-social behav- and age appropriate (Maag, 1999), but these ior until the study ended. If practitioners at- back-up reinforcers were not limited to store- tempt to use a token economy system to bought items. Classroom privileges served as change behaviors permanently, implementa- effective back-up reinforcers (Hail, 2000). tion of a fading system should help to produce Therefore, the intervention was simple enough this result. to be conducted by one person, inexpensive Another implication of the findings of this enough to fit within the school budget and study relates to the use of cooperative games powerful enough to change behaviors. as opposed to competitive games when at- Authors (e.g., Loy & Dattilo, 2000; Quinn tempting to teach social skills. Roberts and et al., 1999) have discussed the importance of Treasure (1992) suggested that competitive combining interventions such as a social skills games fostered the importance of defeating an training program or cooperative games pro- opponent while cooperative games provided gram with behavior management programs. an opportunity for socialization, sharing, and Behavior management programs "in which taking turns (Ramsey & Rank, 1997). When specific behaviors are identified, defined pre- TR specialists develop activities for their par- cisely, and changed using interventions based ticipants with emotional and/or behavioral dis- on principles of applied behavior analysis" orders, they may consider the advantages of (Maag, 1999, p. 536) are used in many set- using cooperative games when developing tings. TR specialists may find that using a strategies to encourage pro-social behaviors systematic reinforcement system (e.g., a token and promote friendships (Dattilo, 2002). economy as detailed by Dattilo & Murphy, 1987) can help them achieve desired client Conclusion outcomes. During behavior observations in the current Adolescents with E/BD are characterized study a trend was noticed in the token econ- as experiencing problems with demonstration omy and cooperative games conditions—pro- of social skills. Because traditional competi- social behaviors were exhibited in direct pro- tive games can lead to increased aggressive portion to the proximity of the first author to behaviors (Anderson & Morrow, 1995; Mar- the participant. For instance, if the group was lowe, 1979), cooperative games offer an alter- standing in a circle working on a project, native method of play. However, simply pro- pro-social comments and actions were en- viding cooperative games does not appear to gaged in at a higher rate by participants nearest be sufficient when attempting to increase pro- the first author. As the first author moved social behaviors and decrease anti-social be- around the circle, pro-social behaviors ceased haviors (Loy & Dattilo, 2000; Quinn et al., by those farthest from him and increased by 1999). Therefore, a token economy system in those now closest to him. A concern voiced by the context of cooperative games was exam- a teacher was that participants may not be ined in this study and this intervention was learning behaviors but simply learning how to associated with an increase in pro-social be- receive rewards. One method to help increase haviors and low levels of anti-social behav- the chance that participants learn the pro-so- iors. cial behaviors is to change the reward system from a fixed ratio exchange of one token for one pro-social behavior to a variable ratio of exchange (e.g., one token for two, four, or References eight pro-social behaviors). Due to time constraints in this study, this system of withdraw Anderson, C. A., & Morrow, M. (1995). Com- petitive aggression without interaction: Effects of was not used and participants continued to competitive versus cooperative instructions on ag-

138 Therapeutic Recreation Journal gressive behavior in video games. Personality and iors and applying consequences. In D. R. Austin, J. Social Psychology Bulletin, 21, 1020-1030. Dattilo, & B. P. McCormick (Eds.), Conceptual Foundations for Therapeutic Recreation (pp. 31— Asher, S. R., & Wheeler, V. A. (1985). Chil- 48). State College, PA: Venture Publishing Inc. dren's loneliness: A comparison of rejected and neglected peer status. Journal of Consulting and De Martini-Scully, D., Bray, M. A., & Kehle, Clinical Psychology, 53, 500-505. T. J. (2000). A packaged intervention to reduce disruptive behaviors in general education students. Baer, D. M., Wolf, M. M., & Risley, T. R. Psychology in the Schools, 37(2), 149-156. (1968). Some current dimensions of Applied Behav- ior Analysis. Journal of Applied Behavior Analysis, Durand, V. M. (1999). Functional communica- 1, 91-97. tion training using assistive devices: Recruiting natural communities of reinforcement. Journal of Ap- Bauer, A. M., & Shea, T. M. (1999). Learners plied Behavior Analysis, 32, 247-267. with emotional and behavioral disorders: An introduction. Upper Saddle River, NJ: Prentice-Hall. Elliott, S. N., & Gresham, F. M. (1993). Social skills and interventions for children. Special Issue: Berliner, D. C. (1992). The nature of expertise in Social-skills and intervention with children and ad- teaching. In F. K. Oser, A. Dick, & J. Patty (Eds.), olescents. Behavior Modification, 1, 287-313. Effective and responsible teaching (pp. 227-248). San Francisco: Jossey-Bass. Evans, S. W., Axelrod, J. L., & Sapia, J. L. (2000). Effective school-based mental health inter- Bluechardt, M. H., & Shephard, R. J. (1995). ventions: Advancing the social skills training pro- Using an extracurricular physical activity program gram. Journal of School Health, 70, 191-194. to enhance social skills. Journal of Learning Dis- abilities, 28, 160-170. Farmer, T. W., & Hollowell, J. H. (1994). Social networks in mainstream classrooms: Social affilia- Carlson, M. J. (1999). Cooperative games: A tions and behavioral characteristics of students with pathway to improving health. Professional School E/BD. Journal of Emotional and Behavioral Disor- Counseling, 2, 230-236. ders, 2, 143-155. Cartledge, G., & Cochran, L. (1993). Develop- Flick, G. L. (2000). Managing teens with ing cooperative learning behaviors in students with ADHD: Practical tools and strategies for dealing behavior disorders. Preventing School Failure, with difficult behaviors. West Nyack, NY: The Cen- 370), 5-11. ter for Applied Research in Education. Coie, J. D., Underwood, M., & Lochman, J. E. Franco, H., Galanter, M., Castaneda, R., & (1991). Programmatic intervention with aggressive Patterson, J. (1995). Combining behavioral and self- children in the school setting. In D. J. Pepler & K. H. help approaches in the inpatient management of Rublin (Eds.), The development and treatment of dually diagnosed patients. Journal of Substance childhood aggression (pp. 389-411). HiUsdale, NJ: Abuse Treatment, 12, 227-232. Erlbaum. Gunter, P. L., Shores, R. E., Jack, S. L., Denny, Dattilo, J. (2002). Inclusive leisure Services: R. K., & DePaepe, P. A. (1994). A case study of the Responding to the rights of people with disabilities effects of altering instructional interactions on the (2nd edition). State College, PA: Venture. disruptive behavior of a child identified with severe Dattilo, J., & Hoge, G. (1999). Effects of a behavior disorders. Education and Treatment of leisure education program on youth with mental Children, 17, 435-444. retardation. Education and Training in Mental Re- tardation and Developmental Disabilities, 34(1), Hail, J. M. (2000). A token economy in the fifth 20-34. grade. Social Education, 8, 5-7. Dattilo, J., Gast, D. L., Loy, D. P., & Malley, S. Hansen, D. J., Nangle, D. W., & Meyer, K. A. (2000). Use of single-subject research designs in (1998). Enhancing the effectiveness of social skills therapeutic recreation. Therapeutic Recreation Jour- interventions with adolescents. Education and nal, 34, 253-270. Treatment of Children, 21, 489-513. Dattilo, J., & Murphy, W. D. (1987). Behavior Harding, J. W., Wacker, D. P., Berg, W. K., modification in therapeutic recreation: An introduc- Cooper, L. J., Asmus, J., Mlela, K., & Muller, J. tory learning manual. State College, PA: Venture. (1999). An analysis of choice making in the assessment of young children with severe behavior prob- Dattilo, J., & Wolfe, B. (2002). Behavior mod- lems. Journal of Applied Behavior Analysis, 32(1), ification in therapeutic recreation: Observing behav- 63-82.

Second Quarter 2003 139 Harris, A. M., & Covington, M. V. (1993). The behavioral problems: A quantitative synthesis of role of cooperative reward interdependency in suc- single-subject research. Behavioral Disorders, 23, cess and failure. Journal of Experimental Education, 193-201. 61, 151-168. McKenney, A., & Dattilo, J. (2001). Effects of Hughes, C. (1999). Identifying critical social an intervention within a sport context on the proso- interaction behaviors among high school students cial behavior and antisocial behavior of adolescents with and without disabilities. Behavior Modification, with disruptive behavior disorders. Therapeutic Rec- 23, 42-60. reation Journal, 35(2), 123-140. Hughes, J. N., & Sullivan, K. A. (1988). Out- Miller, D. N., & Cole, C. L. (1998). Effects of come assessment in social skills training with chil- social skills training on an adolescent with comorbid dren. Journal of School Psychology, 26, 167-183. conduct disorder and depression. Child and Family Behavior Therapy, 20, 35-53. Hupp, S. D. A., & Reitman, D. (1999). Improv- ing sports skills and sportsmanship in children diag- Miller, L. K. (1997). Principles of everyday nosed with attention deficit/hyperactivity disorder. behavior analysis. New York: Brooks/Cole Publish- Child and Family Behavior Therapy, 21, 35-51. ing Co. Kazdin, A. (1991). Aggressive behavior and Miltenberger, R. G. (1997). Behavior modifica- conduct disorder. In T. Kratochwill & R. Morris tion: Principles and procedures. Albany, NY: (Eds.), The practice of child therapy (2nd ed.j. New Brooks/Cole Publishing Co. York: Pergamon Press. Moote, G. T., Smyth, N. J., & Wodarski, J. S. Kazdin, A. (1982). Single-case research de- (1999). Social skills training with youth in school signs: Methods for clinical and applied settings. settings: A review. Research on Social Work Prac- New York: Oxford University Press. tice, 9, 427-465. Kazdin, A. (1982). Single-case research de- Morisse, D., Batra, L., Hess, L., Silverman, R., signs: Methods for clinical and applied settings. & Corrigan, P. (1996). A demonstration of a token New York: Oxford University Press. economy for the real world. Applied and Preventive Kellner, M. H., & Bry, B. H. (1999). The effects Psychology, 5, 41-46. of anger management groups in a day school for Murphy, H. A., Hutchinson, J. M., & Bailey, emotionally disturbed adolescents. Adolescence, 34, J. S. (1983). Behavioral school psychology goes 645-651. outdoors: The effect of organized games on play- LePage, J. P. (1999). The impact of a token ground aggression. Journal of Applied Behavior economy on injuries and negative events on an acute Analysis, 16, 29-35. psychiatric unit. Psychiatric Services, 50, 941—944. Naughton, C. C, & McLaughlin, T. F. (1995). Loy, D. P., & Dattilo, J. (2000). Effects of The use of a token economy system for students different play structures on social interactions be- with behaviour disorders. British Columbia Journal tween a boy with Asperger's Syndrome and his of Special Education, 19, 29-38. peers. Therapeutic Recreation Journal, 34, 190- Orlick, T. (1981). Positive socialization via co- 210. operative games. Developmental Psychology, 17, Maag, J. W. (1999). Behavior management: 426-429. From theoretical implications to practical applica- Pestle, K., Card, J., & Menditto, A. (1998). tions. San Diego: Singular Publishing Group. Therapeutic recreation in a social-learning program: Marlowe, M. (1979). The games analysis inter- Effect over time on appropriate behaviors of resi- vention: A procedure to increase the peer acceptance dents with schizophrenia. Therapeutic Recreation and social adjustment of a retarded child. Education Journal, 32, 28-41. and Training of the Mentally Retarded, 14, 262- Phillips, D. R., Schwean, V. L., & Saklofske, 268. D. H. (1997). Treatment effect of a school based Martin, G., & Pear, J. (1999). Behavior modifi- cognitive-behavioral program for aggressive chil- cation: What it is and how to do it (6th ed.). Upper dren. Canadian Journal of School Psychology, 13, Saddle River, NJ: Prentice Hall. 60-67. Mathur, S. R., Kavale, K. A., Quinn, M. M., Piazza, C. C, Bowman, L. G., Contrucci, S. A., Forness, S. R., & Rutherford, R. B. (1998). Social Delia, M. D., Adelinis, J. D., & Goh, H. (1999). An skill interventions with students with emotional and evaluation of the properties of attention as reinforce-

140 Therapeutic Recreation Journal ment for destructive and appropriate behavior. Jour- M. M. (1998). Promoting social communication nal of Applied Behavior Analysis, 32, 437-449. skills through cooperative learning and direct in- struction. Education and Treatment of Children, 21, Pierce, W. D., & Epling, W. F. (1999). Behavior 354-369. analysis and learning (2nd ed). Upper Saddle River, NJ: Prentice Hall. Schonert-Reichl, K. A. (1993). Empathy and Posavac, H. D., Sheridan, S. M., & Posavac, social relationships in adolescents with behavioral S. S. (1999). A cueing procedure to control impul- disorders. Behavioral Disorders, 18, 189-204. sivity in children with attention deficit hyperactivity Steinberg, Z., & Knitzer, J. (1992). Classrooms disorder. Behavior Modification, 23, 234-253. for emotionally and behaviorally disturbed students: Powell, R. A., Symbaluk, D. G., & McDonald, Facing the challenge. Behavioral Disorders, 17, S. E. (2002). Introduction to Learning Behavior. 145-156. Belmont, CA: Wadsworth Group. Swaggart, B. L. (1998). Implementing a cogni- Premack, D. (1959). Toward empirical behav- tive behavior management program. Intervention in ioral laws: 1. Positive reinforcement. Psychological School and Clinic, 33, 235-239. Review, 66, 219-233. Swain, J. C, & McLaughlin, T. F. (1998). The Premack, D. (1962). Reversibility of the rein- effects of bonus contingencies in a classwide token forcement relation. Science, 136, 235-237. program on math accuracy with middle-school stu- Premack, D. (1965). Reinforcement theory. In dents with behavioral disorders. Behavioral Inter- D. Levine (Ed.), Nebraska Symposium on Motiva- ventions, 13, 11-19. tion (Vol. 13). Lincoln, NE: University of Nebraska Tawney, J. W., & Gast, D. L. (1984). Single Press. subject research in special education. Columbus, Quinn, M. M., Kavale, K. A., Mathur, S. R., Ohio: C. E. Merrill Publishing Co. Rutherford, R. B., & Forness, S. R. (1999). A meta-analysis of social skill interventions for stu- Truchlicka, M., McLaughlin, T. F., & Swain, dents with emotional or behavioral disorders. Jour- J. C. (1998). Effects of token reinforcement and nal of Emotional and Behavioral Disorders, 7, 54- response cost on the accuracy of spelling perfor- 64. mance with middle-school special education students with behavior disorders. Behavioral Interven- Ramsey, G., & Rank, B. (1997). Rethinking tions, 13, 1-10. youth sports. Parks and Recreation, 32, 30-36. Walker, H. M., Colvin, G., & Ramsey, E. Reese, R. M., Sherman, J. A., & Sheldon, J. B. (1995). Antisocial behavior in school: Strategies (1998). Reducing disruptive behavior of a group- and best practices. Pacific Grove, CA: Brooks/Cole. home resident with autism and mental retardation. Journal of Autism and Developmental Disorders, Watson, J. B. (1913). Psychology as the behav- 28, 159-165. iorist views it. Psychological Review, 20, 154—177. Richman, D. M., Wacker, D. P., Asmus, J. M., Wodarski, J. S., & Feit, M. D. (1995). Adoles- Casey, S. D., & Andelman, M. (1999). Further cent substance abuse: An empirical-based group analysis of problem behavior in response class hier preventive health paradigm. New York: Haworth. archies. Journal of Applied Behavior Analysis, Wolery, M., & Holcombe, A. (1993). Procedural 32(3), 269-283. fidelity: Description, measurement, and an example Roberts, G., & Treasure, D. (1992). Children in of its utility. Unpublished manuscript, Alleghany- sport. Sport Science Review, 1, 46-64. Singer Research Institute, Pittsburgh, PA.

Rusch, F. R., Rose, T., & Greenwood, C. R. Zanolli, K., Dagget, J., Ortiz, K., & Mullins, Jo. (1988). Introduction to behavior analysis in special (1999). Using rapidly alternating multiple treatment education. Englewood Cliffs, NJ: Prentice-Hall, Inc. schedules to assess and treat aberrant behavior in natural settings. Behavior Modification, 23, 358- Rutherford, R. B., Mathur, S. R., & Quinn, 378.

Second Quarter 2003 141