O R I G I N A L P A P E R

Effects of a Perseverative Interest-Based TokenEconomy on Challenging and On-Task Behaviorin a Child with Autism

Amarie Carnett • Tracy Raulston • Russell Lang •

Amy Tostanoski • Allyson Lee • Jeff Sigafoos •

Wendy Machalicek

Published online: 19 March 2014

� Springer Science+Business Media New York 2014

Abstract We compared the effects of a token economy intervention that either didor did not include the perseverative interests of a 7-year-old boy with autism. An

alternating treatment design revealed that the perseverative interest-based tokens

were more effective at decreasing challenging behavior and increasing on-task

behavior than tokens absent the perseverative interest during an early literacy

activity. The beneficial effects were then replicated in the child’s classroom. The

results suggest that perseverative interest-based tokens might enhance the effec-

tiveness of interventions based on token economies.

Keywords Autism � Perseverative interest � Token economy � Challengingbehavior � Alternating treatment design

A. Carnett (&) � J. SigafoosVictoria University of Wellington, Karori Campus, PO Box 17-310, Wellington, New Zealand

e-mail: [email protected]

T. Raulston � W. MachalicekUniversity of Oregon, Eugene, OR, USA

R. Lang � A. LeeClinic for Autism Research Evaluation and Support, Texas State University, San Marcos, TX, USA

R. Lang

Meadows Center for the Prevention of Educational Risk, The University of Texas at Austin, Austin,

TX, USA

A. Tostanoski

Vanderbilt University, Nashville, TN, USA

123

J Behav Educ (2014) 23:368–377

DOI 10.1007/s10864-014-9195-7

Introduction

Token economy interventions involve delivering small tangibles (e.g., tokens)

contingent on the presence or absence of target behaviors and then providing an

opportunity to exchange a preset number of these tokens for backup reinforcers.

Previous research has demonstrated that behaviors can be established, decreased,

and/or maintained using token economy systems (Hackenberg 2009; Matson and

Boisjoli 2009). Research has also investigated several variations of this intervention

including the use of a response cost (i.e., losing tokens for inappropriate behavior),

pairing tokens with praise, and delivering tokens on a variety of intermittent

reinforcement schedules. These variables have been shown to influence the

effectiveness of token economy interventions in some cases (Maggin et al. 2011;

Matson and Boisjoli 2009; Mottram and Berger-Gross 2004).

One aspect of the token economy that has received relatively little attention is the

token itself. Traditionally, tokens are considered to be neutral stimuli (e.g., tickets)

that gain reinforcing power by being paired with the backup reinforcers. Charlop-

Christy and Haymes (1998) investigated the effectiveness of incorporating the

idiosyncratic perseverative interests of children with autism within tokens in an

effort to increase the reinforcing power of the token. Charlop-Christy and Haymes

(1998) defined such intense interests as preoccupations or obsessions that an

individual continually seeks. Results from that study indicated that making use of

tokens that reflected the child’s perseverative interests (e.g., using a small picture of

a train as a token for a child who had a perseverative interest in trains) improved

intervention outcomes. To date, this appears to be the only study to have

demonstrated the potential value of individualizing tokens based on a child’s

perseverative interest.

The purpose of this current study was to replicate and extend the work of

Charlop-Christy and Haymes (1998). Specifically, we compared the effects of a

token economy intervention that either did or did not make use of tokens that

reflected a child’s perseverative interest. We examined the effects of this

manipulation on the challenging and on-task behavior of a 7-year-old boy with

autism during an early literacy activity in a public school special education

classroom and an inclusion classroom.

Method

Participant, Setting, and Materials

Troy was a 7-year-old boy who had been diagnosed with autism. He resided at home

with his father, mother, and three older siblings and attended a local public school.

He scored a 31 on the Childhood Autism Rating Scale (CARS; Schopler et al.

1980), which is indicative of mild-moderate autistic symptoms, and a 99 on the

Behavior Assessment System for Children-II, which indicates an overall clinically

significant range (BASC-II; Reynolds and Kamphaus 2004). Troy spent the majority

of his school day in a special education life skills classroom with four to eight other

J Behav Educ (2014) 23:368–377 369

123

children with developmental disabilities, a special education teacher, and a teaching

assistant. Troy’s individualized education plan (IEP) called for him to spend 1 h of

his school day included in activities with students without disabilities. However,

Troy’s challenging behavior (i.e., screaming, falling, and/or lying on the floor)

occurred too frequently to be acceptable in an inclusion classroom (i.e., a classroom

with a combination of students with and without disability).

The Questions About Behavior Function (QABF) Scale (Matson et al. 2012)

suggested that Troy’s challenging behavior was maintained by escape from

demands. As a result, the inclusion time specified in his IEP was met by

nonacademic activities with fewer demands (e.g., lunch, recess). Troy’s school

counselor referred him to this study in an effort to identify a strategy that could be

used to increase Troy’s inclusion during academic instruction in the general

education classroom. Additionally, Troy had previous experience in using a

traditional token economy within a discrete-trial format, and thus did not require

additional training to use the token economy system for this study.

The baseline and intervention sessions were conducted in Troy’s life skills

classroom and in his inclusion classroom. A video camera on a tripod was used to

record the participant and the researcher during all sessions. The inclusion

classroom included one teacher, 14 students without disabilities, two students with

learning disabilities who spent 100 % of their time in that classroom, and two

students with developmental disabilities who divided their time between the

inclusion and life skills classrooms. Both classrooms had a regularly scheduled early

literacy activity, which lasted 10–12 min and occurred three or four times per week.

During the activity, the teacher sat in a chair and read a story to the children as they

sat on the carpet with a teacher assistant and researcher observing. The children

were expected to sit quietly, look at the teacher or book, listen, and answer

occasional reading comprehension questions.

Response Measurement and Interobserver Agreement

Data were collected on Troy’s challenging behavior and on-task behavior.

Challenging behavior was defined as screaming (i.e., loud vocalizations lasting

3 s or more that were considered disruptive in the classroom), falling, and/or lying

on the ground (i.e., collapsing head and body to the ground). Screaming and falling

often occurred in tandem, and the QABF suggested both were maintained by escape

from demands, so these two topographies, whether they occurred alone or in

combination, were recorded as challenging behavior. On-task behavior was defined

as sitting with buttocks on the ground, head oriented toward the teacher, and having

an absence of challenging behavior. Challenging behavior was scored using 10-s

partial interval recording, and on-task behavior was scored using 10-s whole-

interval recording (Kennedy 2005). The on-task behaviors were selected due to their

incompatibility with Troy’s challenging behavior; thus, challenging behavior and

on-task behavior could not be scored in the same interval. Interval data were

converted to a percentage by dividing the number of intervals with each dependent

variable by the total number of intervals, then multiplying by 100 to convert into a

percentage.

370 J Behav Educ (2014) 23:368–377

123

Data on interobserver agreement (IOA) were collected from videos for both

dependent variables during 30 % of the baseline and intervention sessions by two

trained independent coders. IOA was calculated by dividing the number of intervals

with agreement (i.e., both data collectors scored the presence or absence of

challenging behavior/on-task behavior for the interval) by the total number of

intervals (i.e., agreements plus disagreements), then multiplying by 100 to convert

into a percentage. Mean agreement for both dependent variables was 98.5 % (range

95–100 %).

Treatment integrity was assessed for 30 % of the sessions. A procedural checklist

of intervention procedures (available upon request) was used to record the accuracy

of intervention implementation. The mean of treatment integrity was 96.9 % (range

84.6–100 %).

Procedure

Research Design

The two token economy interventions (i.e., with and without embedded persever-

ative interests) were compared using alternating treatments with an initial baseline

design (Gast 2010). The alternating treatments phase was conducted in the life skills

classroom, and the intervention was implemented by the researcher. Generalization

from the life skills classroom to the inclusion classroom was assessed by conducting

a probe in the inclusion classroom during baseline and by adding a third phase, best-

treatment phase, in which the intervention associated with less challenging behavior

and more on-task behavior was implemented in the inclusion classroom (Gast

2010). Across all phases of the study, the following conditions were held constant:

(a) session duration (10 min), (b) time of day when sessions were conducted, (c) the

types of backup reinforcers that were available, (d) the number and timing of

opportunities to exchange tokens for the backup reinforcers, and (e) the reading

level of the stories. During the intervention and generalization phases, the reading

activity was led by the classroom teacher. A teaching assistant was also present, and

the researcher implemented the intervention.

Baseline

Four of the five baseline sessions were conducted in the life skills classroom. Due to

high rates of challenging behavior, only one baseline session was conducted in the

inclusion classroom. The duration of the reading activity was always between 10

and 12 min. To keep session duration constant, data were recorded during the first

10 min only. During baseline, all teachers and assistants were told to conduct the

reading activity as they would normally. During baseline, the teachers in both

classrooms verbally prompted on-task behavior (e.g., ‘‘Troy, please be quiet and sit

up.’’), provided praise contingent upon on-task behavior, and occasionally ignored

challenging behavior or delivered a mild reprimand (e.g., ‘‘Troy, stop that.’’).

However, none of these components were consistently implemented, and despite

this effort, the participant’s challenging behavior had persisted for over 6 months.

J Behav Educ (2014) 23:368–377 371

123

Preference and Backup Reinforcers

Backup reinforcers were selected by first asking Troy’s teachers to identify potential

reinforcers that would be appropriate in their classrooms. The teachers suggested

small edibles (e.g., bite-sized candy or cracker) because they were inexpensive and

could be consumed quickly without causing distraction. A pairwise preference

assessment was then conducted to identify preference of bite-sized edibles (Fisher

et al. 1992). Prior to each session, Troy selected a backup reinforcer from his top

three preferences (i.e., M&M, fruit snack, and chip). The researcher reviewed on-

task behaviors with Troy using a visual support that included pictures and words of

targeted on-task behaviors (i.e., sitting down, staying quiet, and looking at the

teacher) prior to the start of all sessions. The visual support remained present and

was used to redirect challenging behavior if it occurred, at the end of each 20-s

interval (i.e., the researcher pointed to the picture that represented the desired

behavior instead of delivering a token) throughout each session.

Token Economy without Perseverative Interest

The token economy system that did not include Troy’s perseverative interest used

pennies with a small patch of Velcro�

on the back that could be fastened to a token

board. Penny tokens were delivered by the researcher sitting near Troy, contingent

on 20-s of consecutive on-task behavior. A maximum of 30 tokens per 10 min

session could be earned. Backup reinforcers (i.e., bite-sized candy) could be

obtained for every 10 tokens earned, and an opportunity to exchange was presented

within sessions at each moment in which Troy had earned 10 tokens. For data

collection purposes, the exchanges were coded as on-task behavior. The token board

included circles drawn in groups of 10 as a visual representation of the number of

tokens needed to earn a backup reinforcer. Upon earning a token for targeted on-

task behaviors, Troy was handed a token to place on the board (also coded as on-

task).

Token Economy with Perseverative Interests

The token economy system used in this condition differed from the previously

described condition in that the pennies and token board were replaced by tokens and

a board related to Troy’s perseverative interest in jigsaw puzzles. Specifically, the

tokens were small foam puzzle pieces, and the token board was a thin cardstock

frame into which the pieces fit. This token board mirrored the traditional token

economy, in that it included 10 outlined locations for each puzzle piece. The same

procedures, response requirements, exchange rate, and backup reinforcers were used

in both token economy conditions (i.e., with and without perseverative interests).

Troy’s perseverative interest in puzzles was determined by interviews with teachers

and a free operant preference assessment, in which a puzzle was made available

alongside other toys and activity options (Roane et al. 1998). All of Troy’s teachers

agreed that he perseverated on a specific puzzle, and he devoted 100 % of his time

in the free operant preference assessment touching, holding, and manipulating the

372 J Behav Educ (2014) 23:368–377

123

puzzle pieces. Further, Troy always selected this specific puzzle when other puzzles

were available.

Generalization

Troy’s behavior during the group reading activity in the inclusion classroom was

measured in one baseline session using the same procedures as the other four

baseline sessions conducted in the life skills classroom. In the final best-treatment

phase of the study, three sessions were conducted in the inclusion classroom using

the perseverative interest token economy system.

Results

The top panel of Fig. 1 displays the percentage of intervals during which Troy

engaged in on-task behavior during the entire 10-s interval. During baseline, Troy

was on-task in the life skills classroom for a mean of 11 % of the intervals (range

8–18 %). In the inclusion classroom, he was on-task during 13 % of the intervals.

During the alternating treatment phase, both token economy interventions resulted

in an increase in on-task behavior relative to baseline. However, Troy was on-task

more often during the perseverative interest token economy condition

(M = 59.7 %, range 48–70 %) than in the token economy condition that did not

involve tokens reflecting his perseverative interest (M = 45 %, range 32–55 %).

The increase in on-task behavior in the perseverative interest condition was then

replicated during the final best-treatment phase in the inclusion classroom

(M = 64 %, range 52–72 %).

The bottom panel of Fig. 1 displays the percentage of 10-s interval during which

at least one instance of challenging behavior occurred. During baseline conditions in

the life skills classroom, challenging behavior occurred during a mean of 89 % of

intervals (range 82–92 %) and in 87 % of intervals in the inclusion classroom.

Challenging behavior decreased from baseline levels in both token economy

conditions; however, a lower percentage of intervals had challenging behavior in the

perseverative interest condition (M = 40 %, range 30–52 %) compared with the

condition where the token did not coincide with Troy’s perseverative interests

(M = 55 %, range 45–68 %). The reduction in challenging behavior in the

perseverative interest condition in the life skills classroom was replicated during the

final best-treatment phase in the inclusion classroom (M = 36 %, range 28–48 %).

Discussion

The results of this study replicate previous research demonstrating the utility of

token economy interventions for children with autism (Matson and Boisjoli 2009)

because both token economy interventions (i.e., with and without the perseverative

interest) resulted in decreased challenging behavior and increased on-task behavior.

Further, the superiority of the condition involving tokens reflecting Troy’s

J Behav Educ (2014) 23:368–377 373

123

perseverative interest is consistent with the findings of Charlop-Christy and Haymes

(1998). Finally, these data extend previous research by demonstrating the benefit of

interest-based tokens in a special education classroom with generalization to an

inclusion classroom.

The perseverative interests inherent to an autism spectrum disorder (ASD)

diagnosis often impede appropriate classroom behavior and learning (e.g., Rispoli

et al. 2011; Lang et al. 2010) and can be associated with serious challenging

behavior (e.g., Hausman et al. 2009; Matson et al. 2009). Thus, interventions have

primarily sought to address challenging behavior associated with such restricted and

repetitive behaviors and interests (RRBI) with antecedent manipulations to enrich

0

10

20

30

40

50

60

70

80

90

100

Per

cent

age

of 1

0-s

Who

le I

nter

vals

w

ith

On-

Tas

k B

ehav

ior

Token Economy Comparison in Baseline PI Token Economy

in Inclusion

Life SkillsClassroom

With PI

Without

0

10

20

30

40

50

60

70

80

90

100

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Per

cent

age

of 1

0-s

Par

tial

Int

erva

lsw

ith

Cha

llen

ging

Beh

avio

r Inclusion Classroom

Probe

Sessions

Fig. 1 The top panel displays the percentage of 10-s whole interval during which Troy was on-task, andthe bottom panel displays the percentage of 10-s partial interval during which Troy engaged inchallenging behavior. The closed circles represent baseline in the life skills classroom, triangles representthe inclusion classroom, open diamonds represent the token economy without the perseverative interest(PI), and closed squares represent the token economy with the PI

374 J Behav Educ (2014) 23:368–377

123

the environment and prevent challenging behaviors, and consequence-based

interventions that involve interrupting the repetitive behavior (see Boyd et al.

2012 for a recent review). However, other researchers have demonstrated the utility

of capitalizing on perseverate interests by incorporating them into the intervention

procedures or making access to RRBI contingent on targeted appropriate behavior

or the absence of target challenging behavior (Baker et al. 1998; Charlop-Christy

and Haymes 1996, 1998; Vismara and Lyons 2011). This study, considered in

tandem with Charlop-Christy and Haymes (1998), suggests idiosyncratic persev-

erative interests can be utilized to improve intervention efficiency and effectiveness.

The putative mechanism of action responsible for the enhanced effectiveness is

likely the increased reinforcing value of the token itself. Compared with the use of

neutral stimuli as tokens, the reinforcement from perseverative interest-based tokens

may be more immediate, and thus more efficient, than relying only on the

reinforcing power of the backup edibles that were available only after a number of

tokens had been earned and exchanged. Although Troy was always willing to

exchange 10 tokens for the backup reinforcers, it is possible that some children may

value the perseverative interest-based tokens more than backup reinforcers. In such

cases, challenging behavior maintained by continued access to preferred tangibles

might be occasioned when the child is asked to exchange the high preferred token

for a less preferred item. Practitioners using this approach are therefore cautioned to

consider the reinforcing value of the perseverative interest token relative to the

backup reinforcers. If challenging behavior is observed during the exchange, it may

be preferable to use neutral stimuli as tokens or to merely use the preservative

interest tokens alone without additional backup reinforcers. As part of a larger effort

to better incorporate the characteristics of children with autism into intervention

approaches with the goal of improving educational outcomes, future research

designed to elucidate and then potentially address such a limitation remains

warranted.

These findings buttress the evidence supporting the use of token economy

systems with this population and align with the perspective that circumscribed

interests can be a unique strength of individuals with high-functioning ASD

(Mercier et al. 2000). Nevertheless, when children with ASD perseverate to the

exclusion of other activities, such RRBIs significantly restrict their social and

learning opportunities (Pierce and Courchesne 2001; Koegel et al. 1974; Lovaas

et al. 1971). Research could continue to investigate the effects of embedding

perseverative interests into other interventions, such as video modeling. However, it

is possible that the use of perseverative interests in this way may inadvertently lead

to a counterproductive increase in fascination with the perseverative interest.

Although we are not aware of this issue having been reported in previous research, it

would seem a plausible potential limitation that should be investigated as research

in this area continues.

The results of this current study should be considered in light of a few limitations.

First, we selected an alternating treatment design because teachers expressed

concern regarding a reversal to baseline conditions. Although this design facilitated

implementation in an applied setting, the lack of a reversal phase introduced the

potential of carryover effects. Second, to identify Troy’s perseverative interest, we

J Behav Educ (2014) 23:368–377 375

123

utilized teacher reports and a free operant preference assessment, which did not

capture a hierarchy of reinforcers (Roane et al. 1998). Further, there is not a well-

established procedure for distinguishing between high preferred stimuli and the

level of fascination indicative of a true perseverative interest, and our assertion that

puzzles were indeed a perseverative interest should be considered with caution. It is

possible that puzzles were merely highly preferred. Future research should further

investigate reinforcement hierarchies to determine more precise ways of identifying

perseverative interests. Third, the visual cues utilized to prompt on-task behavior

and redirect challenging behavior, although held constant in all intervention and

generalization sessions, were not evaluated as a separate intervention component.

Thus, we are uncertain to what degree they may have contributed to the effects on

the dependent variables. Finally, because on-task behavior increased and challeng-

ing behavior decreased with the use of both token systems, assessing the value of

the perseverative interest token proved difficult. It is possible that the effectiveness

of both systems might approach equivalence over time if challenging behavior

continued to decrease. Thus, future research should investigate the effects of

extended use of the two systems, as well as the effects of systematic fading

procedures of the embedded token system.

References

Baker, M. J., Koegel, R. L., & Koegel, L. K. (1998). Increasing the social behavior of young children with

autism using their obsessive behaviors. The Journal of the Association for Persons with Severe

Handicaps, 23, 300–308. doi:10.2511/rpsd.23.4.300.

Boyd, B. A., McDonough, S. G., & Bodfish, J. W. (2012). Evidence-based behavioral interventions for

repetitive behaviors in autism. Journal of Autism and Developmental Disorders, 42, 1236–1248.

doi:10.1007/s10803-011-1284-z.

Charlop-Christy, M. H., & Haymes, L. K. (1996). Using obsessions as reinforcers with and without mild

reductive procedures to decrease inappropriate behaviors of children with autism. Journal of Autism

and Developmental Disorders, 26, 527–545.

Charlop-Christy, M. H., & Haymes, L. K. (1998). Using objects of obsession as token reinforcers for

children with autism. Journal of Autism and Developmental Disorders, 28, 189–198. doi:10.1023/A:

1026061220171.

Fisher, W. W., Piazza, C. C., Bowman, L. G., Hagopian, L. P., Owens, J. C., & Slevin, I. (1992). A

comparison of two approaches for identifying reinforcers for persons with severe and profound

disabilities. Journal of Applied Behavior Analysis, 25, 491–498. doi:10.1901/jaba.1992.25-491.

Gast, D. L. (2010). Single subject research methodology in behavioral sciences. New York: Routledge.

Hackenberg, T. D. (2009). Token reinforcement: A review and analysis. Journal of the Experimental

Analysis of Behavior, 91, 257–286. doi:10.1901/jeab.2009.91-257.

Hausman, N., Kahng, S., Farrell, E., & Mongeon, C. (2009). Idiosyncratic functions: Severe problem

behavior maintained by access to ritualistic behaviors. Education and Treatment of Children, 32,

77–87.

Kennedy, C. (2005). Single-case designs for educational research. Boston: Pearson Education Inc.

Koegel, R. L., Firestone, P. B., Kramme, K. W., & Dunlap, G. (1974). Increasing spontaneous play by

suppressing self-stimulation in autistic children. Journal of Applied Behavior Analysis, 7, 521–528.

Lang, R., Regester, A., Rispoli, M., & Camargo, S. H. (2010). Rehabilitation issues for children with

autism spectrum disorders. Developmental Neurorehabilitation, 13, 153–155.

Lovaas, O. I., Schreibman, L., Koegel, R., & Rehm, R. (1971). Selective responding by autistic children

to multiple sensory input. Journal of Abnormal Psychology, 77(3), 211–222.

376 J Behav Educ (2014) 23:368–377

123

Maggin, D. M., Chafouleas, S. M., Goddard, K. M., & Johnson, A. H. (2011). A systematic evaluation of

token economies as a classroom management tool for students with challenging behavior. Journal of

School Psychology, 49, 529–554. doi:10.1016/j.jsp.2011.05.001.

Matson, J. L., & Boisjoli, J. A. (2009). The token economy for children with intellectual disability and/or

autism: A review. Research in Developmental Disabilities, 30, 240–248. doi:10.1016/j.ridd.2008.04.

001.

Matson, J. L., Dempsey, T., & Fodstad, J. C. (2009). Stereotypies and repetitive/restrictive behaviours in

infants with autism and pervasive developmental disorder. Developmental Neurorehabilitation, 12,

122–127.

Matson, J. L., Tureck, K., & Rieske, R. (2012). The questions about behavioral function (QABF): Current

status as a method of functional assessment. Research in Developmental Disabilities, 33, 630–634.

doi:10.1016/j.ridd.2011.11.006.

Mercier, C., Mottron, L., & Belleville, S. (2000). A psychosocial study on restricted interests in high-

functioning persons with pervasive developmental disorders. Autism, 4(4), 29–46.

Mottram, L., & Berger-Gross, P. (2004). An intervention to reduce disruptive behaviours in children with

brain injury. Developmental Neurorehabilitation, 7, 133–143.

Pierce, K., & Courchesne, E. (2001). Evidence for a cerebellar role in reduced exploration and

stereotyped behavior in autism. Biological Psychiatry, 49, 655–664.

Reynolds, C. R., & Kamphaus, R. W. (2004). BASC-II: Behavior assessment system for children (2nd

ed.). Bloomington, MN: Pearson Assessments.

Rispoli, M. J., O’Reilly, M. F., Lang, R., Machalicek, W., Davis, T., Lancioni, G., et al. (2011). Effects of

motivating operations on aberrant behavior and academic engagement for two students with autism.

Journal of Applied Behavior Analysis, 44, 187–192.

Roane, H. S., Vollmer, T. R., Ringdahl, J. E., & Marcus, B. A. (1998). Evaluation of a brief stimulus

preference assessment. Journal of Applied Behavior Analysis, 31, 605–620. doi:10.1901/jaba.1998.

31-605.

Schopler, E., Reichler, R. J., Devellis, R. F., & Daly, K. (1980). Toward an objective classification of

childhood autism: Childhood autism rating scale (CARS). Journal of Autism and Developmental

Disabilities, 10, 91–103. doi:10.1007/BF02408436.

Vismara, L. A., & Lyons, G. L. (2011). Using perseverative interest to elicit joint attention behaviors in

young children with autism: Theoretical and clinical implications for understanding motivation.

Journal of Positive Behavioral Interventions, 9(4), 214–228. doi:10.1177/10983007070090040401.

J Behav Educ (2014) 23:368–377 377

123

Reproduced with permission of the copyright owner. Further reproduction prohibited withoutpermission.