ERIC EJ889217: Developing Stimulus Control of the High-Rate Social-Approach Responses of an Adult with Mental Retardation: A Multiple-Schedule Evaluation

2010·0.14 MB·English

by  ERIC

JOURNALOFAPPLIEDBEHAVIORANALYSIS 2010, 43, 285–289 NUMBER2 (SUMMER2010)
DEVELOPING STIMULUS CONTROL OF THE HIGH-RATE
SOCIAL-APPROACH RESPONSES OF AN ADULT WITH MENTAL
RETARDATION: A MULTIPLE-SCHEDULE EVALUATION
LAURA L. GROW
WESTERNMICHIGANUNIVERSITY
AND
LINDA A. LEBLANC AND JAMES E. CARR
AUBURNUNIVERSITY
We evaluated a multiple schedule in which the extinction (S2) components were signaled
overtlybyablacklanyardandthereinforcement(S+)componentswerenotcorrelatedwithany
programmedstimuliindevelopingstimuluscontroloverthehigh-ratesocial-approachresponses
ofanadultwithmentalretardation.RespondingwasconsistentlylowinthepresenceoftheS2
and consistently high when the lanyard was absent (i.e., the S+ condition). Component
durationswere thinnedsuccessfully toalevel thatwas manageablefor caregivers.
Keywords: multiple schedules, social-approach responses, stimulus control
_______________________________________________________________________________
Recent studies have evaluated multiple- Tiger, Hanley, and Heal (2006) compared
schedule variations for reducing high-rate the efficacy of three different multiple-schedule
social-approach responses. In such studies, signaling arrangements in gaining stimulus
periods (or components) of reinforcement and control over the social approaches of preschool
extinction are alternated based upon a time children. These arrangements involved provid-
schedule, and each period is associated with ing overt signals during (a) both S+ and S2
some programmed, salient discriminative stim- components, (b) the S+ component only (i.e.,
ulus. For example, Hanley, Iwata, and Thomp- the S2 component was signaled by the absence
son (2001) used a multiple schedule to reduce of a programmed discriminative stimulus), and
high-rate manding during functional commu- (c) neither during the S+ or S2 component
nication training for the problem behavior of 3 (termedamixedschedule).Bothofthemultiple
individuals with intellectual disabilities. The schedules with signaled reinforcement compo-
authors presented one colored card on a table nents wereefficacious;however, theauthorsdid
during reinforcement periods (S+) in which not evaluate a variation with overtly signaled
each response was reinforced on a fixed-ratio S2 components and no programmed discrim-
(FR) 1 schedule and a different colored card inative stimuli during S+ components.
during extinction (S2) periods. This procedure Tiger et al. (2006) excluded this S2 only
acquired stimulus control over mands, in that option based on the rationale of Terrace(1971)
mandingoccurredinthepresenceoftheS+but that the S2 may gain aversive properties after a
not the S2. history of pairing between the S2 and
extinction. However, we believe the dismissal
of an S2 only arrangement to be premature.
First, in some circumstances an S2 only
WethankCandiceJostadandCarlyCorneliusfortheir
assistancewithdata collection. arrangement may be more practical than the
AddresscorrespondencetoLindaLeBlanc,Department S+ only arrangement described by Tiger et al.
of Psychology, 226 Thach Hall, Auburn University,
For example, a caregiver may find it easier to
Alabama36849(e-mail: [email protected]).
doi:10.1901/jaba.2010.43-285 present an S2 duringa20-minphonecallthan
285
286 LAURA L. GROW et al.
topresentanS+fortheother23hrand40min scales) indicated that his problem behavior was
of the day. Second, following certain learning sensitive to social positive reinforcement in the
histories, extinction-correlated stimuli can ac- form of attention.
quire reinforcing properties (i.e., those associ-
ated with increased response efficiency; Case, Response Measurement and
Fantino, & Wixted, 1985). Furthermore, many Interobserver Agreement
stimuli in the environment signal the unavail- The therapist collected primary data on
ability of reinforcement; avoidance of those paper-and-pencildatasheetsthatwereprecoded
stimuli is adaptive (e.g., walking away from a with three columns (i.e., one for each target
locked door, driving by a closed store). response) under separate sections for reinforce-
Signaling the unavailability of reinforcement ment and extinction periods. Data were
mayreducethenumberofresponsesexposedto collected by making tally marks on the
extinction and maximize available reinforcers, appropriate column when a target response
negating the aversive aspects of this learning occurred. A social approach was defined as any
history. vocal(e.g.,askingabouttheexperimenter’sday,
Based on the potential utility of the S2 only saying ‘‘hi’’) or nonvocal (e.g., attempting to
multiple schedule, we believe evaluation of this shake the experimenter’s hand, waving) behav-
variation is warranted. Therefore, the purpose ior directed toward the experimenter or staff
of this study was to evaluate the effects of a member. Problem behavior was defined as
multiple schedule with FR 1 and extinction hittingorkickingsurfacesandthrowingobjects.
components in which only the extinction Attention delivery was defined as the experi-
component (S2) was signaled overtly to menter vocally or nonvocally acknowledging a
develop stimulus control over the high-rate social approach within 5 s.
social-approach responses of an adult with A second observer simultaneously but inde-
mental retardation. A secondary purpose of pendently scored all behavior during 29% of
the study was to thin the values of the sessions. Total agreement on social-approach
component durations to a level that could be responses, problem behavior, and attention
reasonablysupportedbybehavior-changeagents deliveries was calculated by dividing the smaller
in the natural environment. number of responses by the larger number of
responsespersessionandconvertingtheratioto
a percentage. Mean total agreement was 96%
METHOD
for social approaches (range, 93% to 100%),
Participants and Setting 100% for problem behavior, and 98% for
Tim was a 43-year-old man who had been attention deliveries (range, 95% to 100%).
diagnosed with moderate mental retardation. Attention was delivered accurately on 99%
He resided in a group home and attended a (range, 95% to 100%) of the opportunities.
shelteredworkshopduringtheday.Hespokein
full sentences and completed most daily living Procedure
skillsindependently.Experimentalsessionswere All sessions (15-min duration) were conduct-
conducted at his home during leisure periods ed in Tim’s living room, which contained a
(e.g., after dinner). Tim was included in the variety of leisure items (e.g., television, board
study based on a history of high-rate social- games). We did not arrange any programmed
approachresponses.Heengagedinlowlevelsof consequences for problem behavior during the
problem behavior that were not specifically study.
targeted during the intervention. Previous FR 1 baseline. Each social approach resulted
functional assessments (i.e., observation, rating in an experimenter-delivered attention state-
STIMULUS CONTROL 287
ment on an FR 1 schedule. No discriminative of the treatment evaluation as well as a follow-
stimuli were presented during this condition. up session 6 months later. In addition, staff
Multiple schedule plus rules. Similar to Tiger members at his workplace conducted five
and Hanley (2004), the experimenter initiated sessions following successful schedule thinning
sessions by presenting a black lanyard and at his residence.
stating,‘‘WhenIamwearingthisnecklace,Iam
doingworkand cannot talkto you. When Iam RESULTS AND DISCUSSION
not wearing the necklace, I can talk to you and
Figure 1 depicts Tim’s social-approach re-
answer your questions.’’ Each session consisted
sponses and problem behavior during the
ofanalternationbetweenextinction(alwaysthe
analysis. The FR 1 (baseline) condition was
first component presented) and FR 1 compo-
associatedwithhighratesofapproachresponses
nents.Theexperimenterdeliveredapproximate-
(M 5 4.8 responses per minute) during the
ly 5 to 10 s of attention following each social-
entire 15-min session. However, when the
approach response during FR 1 components. If
multiple schedule plus rules was implemented,
Tim engaged in a social-approach response
Tim engaged in very low rates of social
immediately prior to the completion of the FR
approaches during the signaled extinction
1 component, the reinforcer interval was
components (M 5 0.4) and maintained high-
terminated at the end of the component, and
rate responding during the FR 1 components
the extinction component was initiated as
(M 5 3.4). From these data, it is evident that
planned. No attention (i.e., attention extinc-
the combination of the signaled S2 multiple
tion) was provided following social approaches
schedule and the delivery of presession rules
during extinction components. The final com-
demonstrated effective control of behavior, in
ponent was terminated early if the 15-min
thatsocialapproacheswereseeminglyturnedon
session duration elapsed.
and off both within and across sessions of this
The duration of each component started at
evaluation. This differential response pattern
1 min and was increased systematically across
persisted throughout schedule thinning, during
sessions to reach a terminal goal that was
sessions in which Tim’s home and work staff
acceptable to home and work staff. The
conductedsessions,andata6-monthfollow-up
duration of the extinction component was
assessment. Marginally higher levels of social-
increased systematically by 1 min and the FR
approach responses during both components
1 component was increased by 30 s to terminal were observed when staff members from Tim’s
durations of 10 min for the extinction compo- workplace conducted the sessions. Across all
nent and 5 min for the FR 1 component. The condition types, intervention agents, and set-
durations were increased when sessions pro- tings, problem behavior remained low (M 5
duced two or more social-approach responses 0.02 responses per minute; range, 0 to 0.3) and
per minute during the reinforcement compo- occurred most often during the extinction
nents and less than 0.5 responses per minute components with staff members from home or
during the extinction components for a session. work.
Component response rates were determined by In certain situations, signaling the unavail-
dividing the total number of responses during a abilityofattentiononly(asinthecurrentstudy)
component by the total duration of that maybemorepracticalthansignalingtheS+and
component. S2 or the S+ only, as described by Tiger et al.
To assess generality and maintenance of the (2006) and may be just as efficacious as these
multiple-schedule arrangement, staff from previously described procedures. The concern
Tim’s home conducted the last three sessions raised by Tiger et al. regarding the S2 only
288 LAURA L. GROW et al.
Figure1. Social-approachresponses(top)andproblembehavior(bottom)perminuteduringunsignaledFR1(open
squares) andsignaled extinction (filled squares) conditions.
procedure having a negative impact (i.e., The current study is also one of the few to
decreased preference for the environments demonstrate procedures to thin reinforcement
associated with these stimuli, avoidance behav- density via lengthening extinction-component
ior) was not addressed directly in the current durations and reducing reinforcement-compo-
study; however, no increases in problem nent durations (see also Hagopian, Toole,
behavior were associated with the signaled S2 Long, Bowman, & Lieving, 2004; Hanley et
condition. Future research should evaluate al., 2001; Neidert, Iwata, & Dozier, 2005) in
client preferences for S2 only multiple sched- an effort to make these procedures more
ules relative to other schedules of discontinuous practical for caregivers. Our study was also
reinforcement availability. relatively novel in that our multiple-schedule
STIMULUS CONTROL 289
evaluation incorporated typical caregivers and procedures to eliminate extinction-component
was assessed in multiple settings typically approachesandproblembehavior.Forinstance,
experienced by our participant (see also Cam- one could change the multiple-schedule ar-
milleri, Tiger, & Hanley, 2008). rangement to a chained-schedule arrangement
There are a few procedural limitations of the in which access to the FR 1 component is
current study that should be resolved in future withheld until a period passes without a social
research. First, the total agreement method approach.
worked well with our paper-and-pencil data-
collectionsystembutisarelativelynonstringent
REFERENCES
calculation for interobserver agreement. Thus,
these data should be interpreted with caution Cammilleri,A.P.,Tiger,J.H.,&Hanley,G.P.(2008).
Developing stimulus control of young children’s
despitethehighagreementcoefficients.Second,
requests to teachers: Classwide applications of
thestimuluscontrolexertedinthecurrentstudy multiple schedules. Journal of Applied Behavior
was not evaluated independently of control Analysis,41,299–303.
Case, D. A., Fantino, E., & Wixted, J. (1985). Human
exerted by the schedules of reinforcement. That
observing:Maintainedbynegativeinformativestimuli
is, responding may have ceased during extinc-
only if correlated with improvement in response
tion components due to direct contact with efficiency. Journal of the Experimental Analysis of
extinction rather than the presentation of the Behavior,43,289–300.
Hagopian, L. P., Toole, L. M., Long,E. S., Bowman, L.
S2 (such a phenomenon was demonstrated by
G.,&Lieving,G.A.(2004).Acomparisonofdense-
Tiger & Hanley, 2005). Future research should
to-lean and fixed lean schedules of alternative
conduct a mixed-schedule reversal (i.e., same reinforcement and extinction. Journal of Applied
schedules of reinforcement during each com- Behavior Analysis,37,323–337.
Hanley,G.P.,Iwata,B.A.,&Thompson,R.H.(2001).
ponent without programmed signals) as a
Reinforcementschedulethinningfollowingtreatment
control condition to evaluate the acquired with functional communication training, Journal of
stimulus properties of signals over time. Applied BehaviorAnalysis,34,17–38.
Similarly,thedeliveryofcontingency-specifying Neidert, P. L., Iwata, B. A., & Dozier, C. L. (2005).
Treatment of multiply controlled problem behavior
statements co-occurred with the implementa-
with procedural variations of differential reinforce-
tion of the multiple schedule in the current ment. Exceptionality,13,45–53.
study,sotheadditiveeffectsofthecontingency- Terrace, H. S. (1971). Escape from S2. Learning and
Motivation,2,148–163.
specifying statements are unknown. Future
Tiger, J. H., & Hanley, G. P. (2004). Developing
research may examine the additive effect, if
stimulus control of preschooler mands: An analysis
any, of contingency-specifying statements while of schedule-correlated and contingency-specifying
schedule thinning during multiple schedules in stimuli. Journal of Applied Behavior Analysis, 37,
517–521.
which only the extinction component is
Tiger, J. H., & Hanley, G. P. (2005). An example of
signaled.
discovery research involving the transfer of stimulus
Finally, it is worth noting that some level of control. Journal of Applied Behavior Analysis, 38,
extinction-component responding persisted 499–509.
Tiger, J. H., Hanley, G. P., & Heal, N. A. (2006). The
throughout the evaluation, which is indicative
effectiveness of and preschoolers’ preferences for
of incomplete stimulus control and may be variationsofmultiple-schedulearrangements.Journal
particularly problematic should caregivers in- of AppliedBehaviorAnalysis,39,475–488.
termittently or inadvertently reinforce ap-
Received March 11,2009
proaches during extinction periods. Future
Final acceptanceJune 4,2009
research should evaluate the use of additional Action Editor,Jeffrey H. Tiger