ERIC EJ889215: Increasing Response Diversity in Children with Autism

2010·0.2 MB·English

by  ERIC

JOURNALOFAPPLIEDBEHAVIORANALYSIS 2010, 43, 265–271 NUMBER2 (SUMMER2010)
INCREASING RESPONSE DIVERSITY IN CHILDREN WITH AUTISM
DEBORAH A. NAPOLITANO, TRISTRAM SMITH, JENNIFER R. ZARCONE,
KAREN GOODKIN, AND DAVID B. MCADAM
UNIVERSITYOFROCHESTERSCHOOLOFMEDICINE
Repetitive and invariant behavior is a diagnostic feature of autism. We implemented a lag
reinforcement schedule to increase response diversity for 6 participants with autism aged 6 to
10 years, 4 of whom also received prompting plus additional training. These procedures
appearedtoincreasethevarietyofbuilding-blockstructures,demonstratingthatanintervention
that includes differential reinforcement can increase response diversity for children with an
autismspectrum disorder.
Keywords: autism,behavioralvariability,lagreinforcement,stereotypy,responsevariability,
stimulusvariability
_______________________________________________________________________________
Althoughrestrictiveorrepetitivebehaviorisa may have collateral benefits such as enhancing
diagnostic feature of autism and other autism problem solving and creativity (Neuringer,
spectrum disorders (ASD), research on inter- 2004).
ventions for such behavior has been somewhat In one of the first applied studies of response
neglected (Bodfish, 2004; Turner, 1999). diversity, Goetz and Baer (1973) observed that
Repetitive behavior or a restricted range of 3 typically developing preschool children had a
interests may be a significant impediment to relatively invariant repertoire of block-building
improved functioning for persons with autism skills compared to their peers. Differential
(Langetal.,2009;Pierce&Courchesne,2001). reinforcement in the form of social praise for
To date, most studies have focused on conse- building novel block structures resulted in
quence-basedinterventionssuchasreducingthe increased diversity of block-structure construc-
repetitive behavior itself or the problem behav- tion for all 3 children. More recently, Cammil-
ior occasioned by repetitive behavior (e.g., leri and Hanley (2005) increased the selection
Napolitano, Tessing, McAdam, Dunleavy, & of novel classroom activities with 2 typically
Cifuni, 2006). A promising alternative, howev- developing children using a lag-differential
er, may be to apply findings from the reinforcement schedule. For both participants,
experimental analysis of behavior on systemat- differential reinforcement on a Lag 1 schedule
ically reinforcing variability in responses. Nu- produced immediate increases in novel activity
merous laboratory studies show that reinforce- choices and a subsequent increase in the
number of academic activity choices. In a
ment reliably increases the diversity of behavior
further demonstration of the utility of lag
emitted by both nonhumans and humans and
schedules of reinforcement for increasing re-
We thank the Strong Children’s Research Center for sponse diversity, Lee and Sturmey (2006)
funding this research, Troy Zarcone for his helpful
evaluated a similar reinforcement schedule for
commentsinthedevelopmentofthisstudy,andJonathan
Breidbord for his graphics support. We also thank the varied verbal responses to a standard question,
GreeceCentralSchoolDistrictandMonroe#1BOCESfor ‘‘What do you like to do?’’ in children with
allowing us to conduct the study in their schools and the
autism. Appropriate, novel responses that
childrenandtheirfamiliesforparticipatinginthestudy.
differed from the preceding trial were rein-
Address correspondence to Deborah A. Napolitano,
Division of Neurodevelopmental and Behavioral Pediat- forced, and response variability increased for 2
rics, University of Rochester School of Medicine, 601 ofthe3participantscomparedtoresponsesina
ElmwoodAve.,Box671,Rochester,NewYork14642(e-
condition during which appropriate but non-
mail:[email protected]).
doi:10.1901/jaba.2010.43-265 novel responses were reinforced.
265
266 DEBORAH A. NAPOLITANO et al.
The current study extends the work of Goetz building responses were possible per session. A
and Baer (1973) on increasing the variability in variant form response differed from the prior
play of children with ASD. We used procedures form based on placement direction. Ten form
described by Goetz and Baer and incorporated variations were possible including a step (block
changesbasedonrecentstudiesthathavetargeted placed in the same direction on half of the pegs
verbal responses of individuals with ASD. The of the previous block) and a cross (block placed
changes included adding tangible reinforcement perpendiculartopreviousblock).Avariantcolor
to social praise as a consequence for response response was one that differed from the
diversity, incorporating brief sessions of direct previous block (e.g., red on a blue block). An
instruction if participants initially failed to invariant color response was scored if the
respond to the intervention, and targeting participant repeated a color (e.g., blue, blue)
responses that differed from the prior response or color pattern (e.g., red, blue, red, blue), and
(i.e., response variability) rather than all novel an invariant form was scored if the form (e.g.,
responses.Generalizationwasassessedwithanew step, step) or form pattern (e.g., step, cross,
task (block building), and maintenance was step, cross) repeated. The number of variant
evaluated over a 2-month follow-up period. responsesandthenumberofinvariantresponses
were each divided by 23 to calculate the
METHOD percentage per category. Although observers
scored both form and color for each response,
Participants, Settings, and Materials
we present information for only the targeted
Werecruited1girland5boysfromauniversity
dimension.
school-based consultation program. Participants
We videotaped all sessions for later scoring.
metthefollowingcriteria:(a)diagnosisofASDby
We collected interobserver agreement data for
a developmental pediatrician or psychologist, (b)
44% of the experimental sessions across all 6
chronologicalageof6to10years,and(c)IQ#70
participants (range, 32% to 67%). We defined
on a standardized intelligence test or composite
an agreement as both observers scoring a
standard score # 85 on the Vineland Adaptive
response identically on both form and color.
Behavior Scale (Sparrow, Cicchetti, & Balla,
We calculated percentage agreement for each
2005). Results of the Repetitive Behavior Scale–
session by dividing the number of agreements
Revised (Bodfish, Symons, Parker, & Lewis,
by the number of disagreements plus agree-
2000) completed by the participant’s parent or
ments and converting to a percentage. Agree-
teacher were available for 5 of 6 participants and
mentwas92%acrossallsessions(range,50%to
indicated that these 5 participants had at least
100%). Mean agreement was 85% for Ric
moderate problems with restricted or sameness
(range, 50% to 100%), 99% for Jason (range,
behavior.Weconductedallexperimentalsessions
97% to 100%), 83% for Oliver (range, 77% to
inaroomwithminimaldistractionsatthechild’s
90%), 96% for Alyssa (range, 87% to 100%),
school. Materials included 24 eight-peg plastic
buildingblocks(sixblocksoffourprimarycolors) 98%forZach(range,96%to100%),and92%
and various reinforcers identified through a for Barry (range, 86% to 100%). We collected
paired-stimuluspreferenceassessment(dataavail- procedural fidelity data on the accuracy of
ablefromthefirstauthor). implementation of the general procedures (e.g.,
number and color of blocks provided, session
Response Definitions and Interobserver Agreement length) and the intervention procedures (e.g.,
Observers scored participant responses for reinforcer delivery) for 34% of sessions across
form (variant or invariant) and for color all participants. The mean fidelity was 96%
(variant or invariant). With 24 blocks, 23 (range, 88% to 100%).
RESPONSE DIVERSITY 267
Design and Procedure trials,boththeparticipantandtheexperimenter
We used an ABAB withdrawal design to had a set of the blocks. The experimenter gave
evaluate the effects of the intervention on the initial instruction ‘‘build something,’’ then
response diversity for the dimension with the modeled building something different and
leastvariabilityduringbaseline.Theexperiment- verbally prompted the participant to imitate
er conducted two to three experimental sessions her model (e.g., ‘‘now you build something
weeklyduringconsecutiveweeks(exceptfortwo different’’). All other procedures were identical
timeswhenaparticipantwasnotavailabledueto to the Lag 1 reinforcement intervention.
the school winter recess). Session duration Generalizationandfollow-up.Woodenblocks
varied.Eachsessionbeganwhentheexperiment- of similar size and color to the plastic blocks
er provided the student with 24 blocks and wereusedtoassessgeneralizationunderbaseline
ended when one of the following criteria were conditions. Follow-up data were collected 2 to
3monthsaftertheendofthefinalintervention,
met: (a) 10 min had elapsed, (b) the participant
using procedures identical to the Lag 1
used all the blocks, (c) the participant said ‘‘I’m
reinforcement condition.
done’’ or pushed the blocks away, or (d) the
participantdidnotinteractwiththeblocksfor2
consecutive minutes. RESULTS AND DISCUSSION
Baseline. The experimenter gave the partici-
Diverse responding increased for all partici-
pant the blocks and said, ‘‘build something.’’
pants with intervention. Ric’s diverse respond-
The experimenter delivered praise intermittent-
ing increased from a mean of 1% of responses
ly and at least once in each session for building,
duringbaselinetoameanof21%afterteaching
regardless of the form or color. Individual
sessions(Figure 1,top).Hereplicatedtheeffect
baseline data were used to select the dimension
inthereversal,withalargerincreaseindiversity
(i.e., form or color) with fewer novel responses
in the second intervention phase to a mean of
to target in intervention.
65%, with poorer performance during follow-
Lag1reinforcementschedule.Theexperiment-
up. Oliver (middle) displayed some diverse
ergave theparticipant theblocksand theinitial
responding in baseline (29% of responses) that
verbal prompt, ‘‘build something.’’ The exper-
increased to a mean of 84% in the first
imenterdeliveredtangiblereinforcers (e.g., 30-s
intervention phase after a teaching session. He
access to a preferred item) or edible reinforcers
replicated this effect during reversal and
(e.g.,smallbitofpreferredfood)followingeach maintained it at follow-up (M 5 68%). Barry’s
color (Jason, Alyssa) or form (Ric, Oliver, data(bottom)revealedverylittlediversityinhis
Barry, Zach) response that differed from the block building during both baseline conditions.
previous one. Preferred tangible and edible However, his percentage of diverse responses
items had been identified in a previous increased to a mean of 73% in the initial
preference assessment. Consumption time was intervention condition and 58% in the second
excluded from session time, and after consum- intervention condition. At follow-up, his mean
ing the reinforcer participants were told, ‘‘build diverse responding decreased from the second
differently.’’ intervention, but was maintained at a mean of
Teaching trials. The experimenter imple- 25%.
mented teaching trials for Ric (seven sessions), At baseline, Alyssa showed a rate of color
Oliver (two sessions), Alyssa (five sessions), and variation of only 13%, the minimum possible
Barry (four sessions) because the Lag 1 (Figure 2, top). She initially displayed little
reinforcement schedule did not result in improvement in the first intervention condi-
performance improvements. During teaching tion. After several reinforcer changes, Alyssa’s
268 DEBORAH A. NAPOLITANO et al.
Figure 1. The percentage of variant form responses (VR) displayed by Ric, Oliver, and Barry across all
experimentalconditions.
RESPONSE DIVERSITY 269
Figure 2. The percentage of variant form responses (VR) displayed by Zach (middle) and variant color responses
(VR) displayedbyAlyssa (top)andJason (bottom)across allexperimentalconditions.
270 DEBORAH A. NAPOLITANO et al.
diverserespondingincreasedtoameanof33%, differently’’intheintervention,whichmayhave
67%inthesecondinterventionphase,and51% increased the likelihood that the participants
in follow-up. Both Zach’s and Jason’s perfor- understood the contingencies. Finally, the test
mances (middle and bottom panels, respective- of generalization did not include a baseline
ly) improved in the first intervention phase comparison prior to the intervention, which
relative to their initial baselines (Jason from weakens any conclusions that can be drawn
13% to 100%, Zach from 46% to 78%). from these data. One further consideration is
However, they failed to return to baseline rates that increasing response diversity for play
in the second baseline condition and main- behavior may have not had any impact on
tained a high rate of diverse responses during otherstereotypicorinvariantbehaviorsforthese
the remaining phases of the study. Despite the participants.
improvements in diverse responding by all Directions for future research include testing
participants during intervention and follow- interventions to promote response diversity in
up, only Jason displayed a high rate of personswithASDacrossawidevarietyofsocial
responding during generalization. skills and behaviors (e.g., play skills, vocal
Overall, the results indicated that an inter- responses), combining interventions to increase
vention package consisting of differential rein- responsediversitywithinterventionstodecrease
forcement on a Lag 1 schedule combined with repetitive or invariant behavior, and evaluating
additional teaching for some participants and a long-term global outcomes of such interven-
prompt to ‘‘build differently’’ increased diver- tions in larger samples of persons with ASD.
sity of block-building responses. This finding
replicates and extends previous basic and REFERENCES
applied behavior-analytic work and holds
Bodfish, J. W. (2004). Treating the core features of
promising potential for increasing the prob-
autism: Are we there yet? Mental Retardation and
lem-solving abilities and creativity of persons Developmental Disabilities Research Reviews, 10,
with ASD (Goetz & Baer, 1973; Lee, McCo- 318–326.
Bodfish,J.W.,Symons,F.J.,Parker,D.E.,&Lewis,M.
mas, & Jawor, 2002; Lee & Sturmey, 2006;
H.(2000).Varietiesofrepetitivebehaviorinautism.
Neuringer, 2004). Further demonstrations of Journal of Autism and Developmental Disabilities, 30,
theeffectivenessofdifferentialreinforcementon 237–243.
Cammilleri, A. P.,& Hanley, G.P. (2005). Use of a lag
diverserespondinginpersonswithASDusinga
differential reinforcement contingency to increase
variety of materials (e.g., blocks) and across a
varied selection of classroom activities. Journal of
varietyoftargetbehaviors(e.g.,vocalresponses) AppliedBehaviorAnalysis,38,111–115.
have important implications for the treatment Goetz, E. M., & Baer, D. M. (1973). Social control of
form diversity and the emergence of new forms in
of the core features of ASD (i.e., restrictive
children’s blockbuilding. Journal of Applied Behavior
repetitive behavior) and the development of Analysis, 6,209–217.
comprehensive treatment programs. Lang, R., O’Reilly, M., Sigafoos, J., Lancioni, G. E.,
Machalicek,W.,Rispoli,M.,etal.(2009).Enhancing
Limitations of the study include a lack of
the effectiveness of a play intervention by abolishing
reversal for 2 participants (Zach and Jason) and the reinforcing value of stereotypy: A pilot study.
only a partial reversal for Ric, which may JournalofAppliedBehavior Analysis, 42,889–894.
Lee,R.,McComas,J.J.,&Jawor,J.(2002).Theeffectsof
weaken the conclusions. The lack of reversal to
differentialandlagreinforcementschedulesonvaried
baseline for Zach and Jason, however, may also
verbalrespondingbyindividualswithautism.Journal
indicate that diverse responses acquired rein- ofAppliedBehavior Analysis, 35,391–402.
forcingvalue.Additionallimitationsmaybethe Lee,R.,&Sturmey,P.(2006).Theeffectsoflagschedules
and preferred materials on variable responding in
lackofaprogrammedscheduleforverbalpraise
studentswithautism.JournalofAutismandDevelop-
and the addition of the prompt to ‘‘build mentalDisorders,36,421–428.
RESPONSE DIVERSITY 271
Napolitano, D. A., Tessing, J. L., McAdam, D. B., Sparrow, S. S., Cicchetti, D. V., & Balla, D. A. (2005).
Dunleavy, J. J., III, & Cifuni, N. (2006). The Vineland Adaptive Behavior Scales: Parent/caregiver
influence of idiosyncratic antecedent variables on edition.CirclePines,MN:AmericanGuidanceServices.
problem behavior. Journal of Developmental and Turner, M. (1999). Repetitive behavior in autism: A
Physical Disabilities, 18,295–305. review of thepsychological research.Journalof Child
Neuringer, A. (2004). Reinforced variability in animals Psychology and Psychiatry,40,839–849.
andpeople.American Psychologist,59,891–906.
Pierce, K., & Courchesne, E. (2001). Evidence for a
cerebellarroleinreducedexplorationandstereotyped Received July3, 2008
behavior in autism. Biological Psychiatry, 49, Final acceptanceAugust 28,2009
655–664. Action Editor,Linda LeBlanc