USING JOINT ACTIVITY SCHEDULES TO PROMOTE PEERENGAGEMENT IN PRESCHOOLERS WITH AUTISM

ALISON BETZ, THOMAS S. HIGBEE, AND KARA A. REAGON

UTAH STATE UNIVERSITY

We assessed the use of a joint activity schedule to increase peer engagement for preschoolers withautism. We taught 3 dyads of preschoolers with autism to follow joint activity schedules thatcued both members of the pair to play a sequence of interactive games together. Results indicatedthat joint activity schedules increased peer engagement and the number of games completed forall dyads. Schedule following was maintained without additional prompting when activities wereresequenced and when new games were introduced for 2 of the 3 dyads.

DESCRIPTORS: activity schedules, autism, peer play, social interaction

_______________________________________________________________________________

Children with autism often play alone orbecome dependent on adult prompts to interactwith peers (McClannahan & Krantz, 1999).Independent play skills of students with autismhave been shown to improve with the use ofphotographic activity schedules, which arecomposed of a sequence of pictorial cues forchildren with disabilities to complete complexchains of behavior independently (McClanna-han & Krantz). In the research that hasexamined the effects of activity schedules onsocial interactions (e.g., Krantz & McClanna-han, 1993, 1998; Stevenson, Krantz, &McClannahan, 2000), interactions have in-volved 1 individual with autism following anactivity schedule that prompts him or her tointeract with a partner (adult or peer) who isnot following an activity schedule and is simplyavailable as a conversation partner. To date, noresearch studies have been published in whichboth members of a social pair follow the sameactivity schedule and complete a sequence ofinteractive games together. A joint activityschedule would potentially have the samebenefits of an independent activity schedule,namely increasing independence and decreasing

the need for adult prompts, plus the addedfeatures of promoting peer engagement inactivities, turn taking, and cooperative play. Ajoint schedule may also prompt both studentsto move from one activity to another together,which could facilitate the following of classwideschedules used in typical educational environ-ments. Thus, the purpose of the current studywas to investigate whether inclusion of a jointactivity schedule would result in higher levels ofengagement for pairs of children with autismduring interactive games.

METHOD

Participants and Setting

Three dyads of preschool-aged childrenparticipated. All were between the ages of 4and 5 years, had diagnoses of autism, and werereceiving intensive behavioral intervention ineither university-based (Brady and David) orpublic-school-based (Ali and Dillon, Jacksonand Nathan) preschools. With the exception ofAli, all participants were male. All children wereidentified as fluent followers of independentactivity schedules, and each child could alsoselect his or her own activities from a choiceboard (see McClannahan & Krantz, 1999, forspecific fluency criteria). All research sessionswere conducted in the play area of participantsclassrooms, which contained a shelf with all therelevant games as well as the activity schedule

We thank Henry Roane for his helpful comments on anearlier version of this manuscript.

Address correspondence to Thomas Higbee, Utah StateUniversity, 2865 Old Main Hill, Logan, Utah 84322 (e-mail: tom.higbee@usu.edu).

doi: 10.1901/jaba.2008.41-237

JOURNAL OF APPLIED BEHAVIOR ANALYSIS 2008, 41, 237241 NUMBER 2 (SUMMER 2008)

237

(except in baseline and reversal probes), table,chairs, and a board that displayed the gamechoices for the choice pages. The board hung ona wall near the toy shelf and activity schedule.

Measurement

During all sessions, independent observersused a 20-s momentary time-sampling proce-dure to estimate the duration of peer engage-ment and to record prompts. Dyads were scoredat each interval as being engaged, prompted, orunengaged. Dyads were scored as engaged ifboth participants were taking turns, using thegame materials in the manner in which theywere designed, cleaning up game materials,setting up game materials, obtaining or puttingthe game on the shelf, choosing a picture of thegame from the choice board, initiating play(either independently or by following theappropriate script without additional prompt-ing from an instructor), verbally interactingwith a peer, turning the page of the activityschedule, attending to the picture depicted onthe activity schedule, or walking between theschedule, game shelf, choice board, and tablewithout prompting. An interval was scored asprompted when the instructor had his or herhands on one or both participants or when theinstructor was shadowing one or both partici-pants within a distance less than 1.5 m.Participants were scored as unengaged if theywere engaged in behaviors other than thosespecified for engagement and were not beingprompted by an instructor.

Interobserver agreement data were obtainedin at least 30% of the sessions across allconditions. For an interval to count as anagreement, both observers had to agree onwhich game was being played; if the partici-pants were engaged, prompted, or unengaged;and if they were engaged in the activity depictedon the schedule in each interval. Meanagreement was 91% (range, 84% to 100%)for Brady and David, 88% (range, 63% to100%) for Ali and Dillon, and 92% (range,71% to 100%) for Nathan and Jackson.

Procedure

Sessions lasted 20 min during baseline andreversal probe phases. In all other phases,sessions ended when all pages of the jointschedule were completed and varied in length,depending on how long it took to completeeach game. Each session began with the childrenstanding in front of the shelf containing therelevant games and when the data collector gavethe instruction, These are the games you canplay with. Go play. Six interactive games wereconcurrently available, which we chose due toavailability, having a clear beginning and end,and allowing two people to play at the sametime (e.g., Dont Break the Ice; Hungry,Hungry Hippos; Dont Spill the Beans; Croc-odile Dentist). An adult instructor taught eachchild how to play appropriately with all of thegames, and each participant demonstratedproficiency with all of the games prior to thebeginning of the research. Children wereconsidered proficient at the games when theycould play and complete the game with noprompts to take turns, follow the rules of thegame, and set up and clean up the game.

Baseline and reversal probe. The joint activityschedule was not present. Inappropriate behav-ior or attempts to interact with anyone otherthan the participating peer were ignored.Children were given the standard instructionto play, and no additional manual or verbalprompts were delivered. After the initialbaseline, a baseline probe with the schedulepresent was conducted.

Teaching. A joint photographic activityschedule was displayed in a three-ring binderfor each dyad. Each schedule book containedtwo prechosen activity pages followed by twochoice pages (two games were designated for theprechosen activities and the remaining fourgames were used as choices). Each member ofthe dyads was responsible for one choice pageand one prechosen activity page. The prechosenactivity page included a picture of the respon-sible participant at the top and a picture of the

238 ALISON BETZ et al.

game that was to be completed. The participantwhose picture was depicted on a particular pagewas responsible for initiating play by readingthe script Lets play . Scripts werepresented in full at the beginning of theteaching phase. The scripts were systematicallyfaded from back to front by removing one wordat a time until no words were on the page andthe children continued to independently initiateplay (scripts were eliminated for all children butAli by the end of the evaluation). The choicepages also contained a picture of a participant atthe top and a hook-and-loop dot where thepicture of the chosen activity was to be placed.

In the beginning of the teaching condition,one of two instructors stood behind theparticipants. After the standard instruction wasgiven, instructors used graduated guidance(MacDuff, Krantz, & McClannahan, 1993) toprompt the participants to engage in interactiveplay (e.g., taking turns during a board game orspinning a spinner) and following the schedule(e.g., pointing to the picture of the activity,obtaining the activity, turning the page).During this condition, the schedule pagesremained in the same order with the sameprechosen activities for each participant, and thesame four games were available for theparticipants to choose for their choice pages.

Maintenance. During maintenance, the in-structors were present but remained at least1.5 m away from the participants. Prompts toengage were provided only if the participantsdid not initiate play within 5 s or emitted self-injury or aggression toward the peer.

Resequencing. To demonstrate that the spe-cific sequence of the activities was not control-ling student responding, the order in which thefour pages were displayed was different eachsession (the prechosen activities remained thesame for each participant, and the same fourgames were available for choice pages in thiscondition).

Generalization. Two of the original gameswere rotated in with four novel games in thiscondition. The two original games were

available on the choice page, two of the novelgames were selected for the prechosen pages,and the remaining two novel games wereavailable as the other choices. This phase wasimplemented to test for generalization ofengagement across novel games. Instructorprompts were reintroduced to increase peerengagement for Jackson during the generaliza-tion phase. Full manual prompts were usedinitially and were faded by introducing a briefdelay and moving away from Jackson after thepercentage of engagement was at or above 80%for two consecutive sessions.

A nonconcurrent multiple baseline designacross dyads was used to assess the effects of ajoint activity schedule on peer engagement.

RESULTS AND DISCUSSION

The percentage of intervals of engagementwas low during baseline and the schedule probefor all dyads (Figure 1). Peer engagementrapidly increased during teaching and persistedabove 80% during the maintenance phase withno prompting (Ali and Dillon and Nathan andJackson) or a single prompt (Brady and David).Engagement persisted during resequencing andgeneralization phases for Ali and Dillon and forBrady and David. A brief reversal probeconducted with Brady and David, during whichthe joint activity schedule was removed, showeda decrease to 0% engagement by the secondsession; engagement increased immediatelywhen the schedule was reintroduced. Duringresequencing with Nathan and Jackson, peerengagement became variable and began todecrease rapidly as the phase progressed. Whenthe generalization phase began, engagementremained low. Graduated guidance with Jack-son produced an immediate increase in re-sponding that was maintained throughout theremainder of the phase.

Our study extends the research on activityschedules by showing the efficacy of a jointactivity schedule to increase peer engagement.All 3 pairs of children learned to follow the

TEACHING PEER PLAY 239

Figure 1. Results of the joint activity schedule intervention for Ali and Dillon (top), Brady and David (middle), andNathan and Jackson (bottom). The asterisk represents the session in which manual prompts were reintroduced forNathan and Jackson. The double asterisks represent the session in which those prompts were completely faded.

240 ALISON BETZ et al.

joint activity schedule and 2 of the 3 dyadsmaintained high levels of peer engagementwhen activities within the schedule wereresequenced and new activities were introduced.

Although the activity schedule was effectivein increasing peer engagement, the precisebehavioral mechanisms responsible for itseffectiveness are unclear. It is most plausiblethat the activity schedule functioned as adiscriminative stimulus by signaling availableactivities for which engagement would bereinforced. Because programming reinforce-ment at the end of a schedule is generallyrecommended, it was interesting that theparticipants in this study completed the activityschedules without any programmed reinforce-ment mediated by an instructor. It is possiblethat the persistent engagement was governed byrules, but it seems more likely that playinggames with the designated peers according tothe schedule became a reinforcing event.Nonetheless, the behavioral mechanisms re-sponsible for the effectiveness of activityschedules should be isolated in future research.

Although all 3 dyads were successfully taughtto follow a joint activity schedule that promotedengagement, there were several limitations tothis study. First, data on procedural fidelitywere not taken on prompting and prompt-fading procedures. It is possible that variationsin prompting occurred, which may account forsome of the response patterns observed in thedata. Second, levels of engagement deterioratedduring the resequencing phase for Nathan andJackson, necessitating the reintroduction ofgraduated guidance of engagement. Althoughengagement eventually increased and persistedafter the prompts were faded, the reason for thedecrease cannot be determined from the presentdata. Third, all the participants in this studywere fluent followers of independent activityschedules, and most had an extensive verbal

repertoire; these factors likely mediated theeffects of our joint activity schedule. Fourth,our definition of peer engagement encompasseda variety of interactive and play behavior.Therefore, future researchers should examine ifjoint activity schedules are successful withstudents who do not have a history withindependent activity schedules or who have lessextensive verbal repertoires. Future researchersshould also examine the extent to which specificsocial behaviors are affected by joint activityschedules.

Nevertheless, our preliminary data suggestthat a joint activity schedule may be apromising tool to increase peer engagementand game play among children with autism.This strategy should eventually be compared toother effective means for increasing interactiveplay and engagement such as video modeling,discrete-trial training, and social stories.

REFERENCES

Krantz, P. J., & McClannahan, L. E. (1993). Teachingchildren with autism to initiate to peers: Effects of ascript-fading procedure. Journal of Applied BehaviorAnalysis, 26, 121132.

Krantz, P. J., & McClannahan, L. E. (1998). Socialinteraction skills for children with autism: A script-fading procedure for beginning readers. Journal ofApplied Behavior Analysis, 31, 191202.

MacDuff, G. S., Krantz, P. J., & McClannahan, L. E.(1993). Teaching children with autism to usephotographic activity schedules: Maintenance andgeneralization of complex response chains. Journal ofApplied Behavior Analysis, 26, 8997.

McClannahan, L. E., & Krantz, P. J. (1999). Activityschedules for children with autism: Teaching indepen-dent behavior. Bethesda, MD: Woodbine House.

Stevenson, C. L., Krantz, P. J., & McClannahan, L. E.(2000). Social interaction skills for children withautism: A script-fading procedure for nonreaders.Behavioral Interventions, 15, 120.

Received January 9, 2007Final acceptance May 21, 2007Action Editor, Gregory Hanley

TEACHING PEER PLAY 241