ERIC EJ964437: Preference for Fluent versus Disfluent Work Schedules

2011·0.26 MB·English

by  ERIC

JOURNALOFAPPLIEDBEHAVIORANALYSIS 2011, 44, 847–858 NUMBER4 (WINTER2011)
PREFERENCE FOR FLUENT VERSUS DISFLUENT WORK SCHEDULES
DANIEL M. FIENUP
THEMAYINSTITUTEANDQUEENSCOLLEGEANDTHE
GRADUATECENTEROFTHECITYUNIVERSITYOFNEWYORK
AND
ASHLEY A. AHLERS AND GARY PACE
THEMAYINSTITUTEANDNORTHEASTERNUNIVERSITY
Twostudieswereconducted thatexamined thepreference ofastudentdiagnosed withabrain
injury.InStudy1,apreferenceassessmentwasfollowedbyathree-choiceconcurrent-operants
reinforcerassessment.Twochoicesresultedinaccesstopreferredactivitiesforcompletingwork,
andathirdchoiceresultedinaccesstonothing(i.e.,noactivity).Unpredictably,theparticipant
consistently chose the no-activity option. Study 2 examined why this student preferred work
associated with no activity over preferred activities. Through a variety of concurrent-operants
procedures,itwasdeterminedthatshepreferredfluentworkfollowedbyreinforcersratherthan
workthatwasbrokenupbyaccesstopreferredactivities.Implicationsforresearchonpreference
arediscussed.
Keywords: braininjury, fluency, preference,reinforcer assessment
_______________________________________________________________________________
One possible outcome of research is that the student’sbehaviorwasunexpectedandprompted
independent variable fails to exert control over follow-up analyses. The following introduction
behavior in the way it was intended. When this provides the background for the study that
outcome occurs, an experimenter has to make a produced the unexpected findings (Study 1).
decision regarding how to continue. At least two Study 2 describes the follow-up analyses that are
optionsexist.Theexperimentercan(a)claimthat the primary focus of this investigation.
the independent variable was not potent enough Measuring response allocation within a
or that some extraneous variable was controlling concurrent-operants arrangement is a popular
behavior or (b) further evaluate the ‘‘failure’’ to method for assessing the efficacy of reinforcers
determine why unexpected results occurred. The (e.g., DeLeon, Iwata, Goh, & Worsdell, 1997).
latter avenue is the type of process described Results of research utilizing this approach have
by Roane, Fisher, and McDonough (2003), who identifiedavarietyoffactorsthatinfluenceprefe-
sought to research the overjustification effect but rence for reinforcers, including response effort
ended up conducting detailed follow-up analyses (DeLeon et al., 1997; Tustin, 1994), reinforcer
when they obtained unexpected results. The choice(Tiger,Hanley,&Hernandez,2006),and
current evaluation describes this type of research. rateandqualityofreinforcers(Mace,Neef,Shade,
Initially, we sought to examine the effects of &Mauro,1996).Forinstance,Tigeretal.(2006)
various schedules of reinforcement on the behav- conducted an analysis of the value of choice us-
iorofstudentswithbraininjury.Fortwoofthese ingaconcurrent-operantsprocedure.Participants
students, the results were consistent with pub- were given three work options that differed only
lished and expected outcomes. However, one in the programmed consequences: single edible
item, multiple identical edible items, and choice
All correspondence should be addressed to Daniel M. ofaspecificedibleitemornoedibleitem.Results
Fienup,DepartmentofPsychology,QueensCollege,65-30 indicated that participants were more likely to
Kissena Blvd., Flushing, New York 11367 (e-mail:
allocate responding to the operant that allowed
[email protected]).
doi:10.1901/jaba.2011.44-847 choice of the edible reinforcer.
847
848 DANIEL M. FIENUP et al.
Mace et al. (1996) examined the effects of such as in school (e.g., Mace et al., 1996).
work difficulty and schedule of reinforcement Student responses of interest typically include
on choice. Participants chose between two labelingstimuli,completingmathematicswork-
stacks of math cards. In two studies, the groups sheets, and finding states on a map, to name a
of cards varied by difficulty and programmed few examples.Originally, wesoughtto examine
reinforcement. Participants allocated their time reinforcer preference using more typical school-
to math problems associated with a richer related work requirements to examine whether
schedule of reinforcement or higher quality of similar results would emerge. Second, we
reinforcement, regardless of whether the math sought to enhance experimental control by
problems were the same or different difficulty adding a control condition (e.g., Tiger et al.,
level. 2006) to the typical two-choice arrangement.
Responserequirement,orresponseeffort,also These two features guided our design of
hasbeenfoundtodifferentiallyaffectpreference Study 1. To examine the generality of results
fordifferentreinforcersassociatedwiththesame obtained by Tustin (1994) and DeLeon et al.
reinforcement schedule (DeLeon et al., 1997; (1997), we selected work requirements from
Tustin, 1994). One study compared preference students’everydayinstruction.Acontrol option
between two reinforcers and manipulated both was added to the reinforcer assessment to
enhance experimental control, in the event that
the reinforcement schedule and similarity of
a participant selected equally among stimuli.
reinforcers. In one manipulation, participants
This option, if chosen, resulted in access to no
chose between two identical operants (button
stimuli. It should be noted that the participant
presses)thatledtoeitherafoodortoyreinforcer
reported in Study 1 was only one of three
(dissimilarreinforcers).Thesameprocedurewas
participants who originally participated. The
conducted for two food reinforcers (similar
unexpected results that emerged warranted
reinforcers). In both procedures, the schedule
presenting an analysis of this student alone.
of reinforcement was increased systematically.
Participants responded differentially to one of
the two reinforcers only under larger schedules STUDY 1
(e.g., fixed-ratio [FR] 10) and with similar Method
reinforcers. However, they did not respond Participant. Eve was a 16-year-old girl who
differentially to dissimilar reinforcers under all attended a private day school that specialized in
schedules and to similar reinforcers under low neurorehabilitation. At 9 months of age, Eve
schedules (e.g., FR 1). had intractable seizures that were treated with
We sought to conduct research aimed at a right hemispherectomy. She later received a
addressing two limitations found in the rein- diagnosis of nonverbal learning disorder. At
forcer assessment and concurrent-operants lit- school, Eve worked on a variety of academic
erature on response effort. The first issue was programs such as math (e.g., greater than/less
the generality of published results. Previous than problems) and reading (fourth-grade level
research used simple responses to evaluate chapter books). She sat at her desk and worked
reinforcer effectiveness. Tustin (1994) required independently on worksheets and, after com-
participants to push a button on a joystick, and pletion, turned them in for feedback from her
DeLeon et al. (1997) required participants to teachers. Eve raised her hand and asked her
press a microswitch. Although these behaviors teachersforhelpwhensheexperienceddifficulty
may have been relevant for the participants in withherworkandengagedinpreferredacademic
those studies, a broader range of behavioral activities (e.g., reading or playing a game) for
outputs are required in more typical settings, approximately20to30min.Sheworkedonless
PREFERENCE FOR FLUENT WORK 849
preferred academic activities (e.g., math prob- to read. The ‘‘Great States with staff’’ card
lems)forapproximately5minbeforerequesting contained a picture of the Great States Junior
abreak.Eve’sclassroomstructureincluded5-to board game and signaled access to playing the
10-min breaks with preferred activities contin- gamewiththeexperimenter.Thisgameinvolved
gent on completing work in a given subject. rolling dice, moving a game piece around a
Shorter breaks were provided as needed. Eve board, and answering questions about states in
participated in this study to examine the impact America. The ‘‘Great States alone’’ card con-
of various contingent preferred activities on her tainedapictureoftheGreatStatesJuniorboard
completion of low, medium, and high work game and signaled access to playing the game
requirements. The high work requirements without the experimenter. The ‘‘go for a walk’’
exceeded what Eve was expected to complete in carddisplayedapictureofapersonwalkingand
her classroom. signaled taking a walk in the hall with the
Setting. All sessions were conducted in the experimenter. The ‘‘handshake and chat’’ card
library of Eve’s school. The large room was included a picture depicting two stick figures
divided into four sections, and sessions were talking and signaled access to a handshake with
conducted in a quiet corner that was surround- the experimenter and conversation based on
ed by bookshelves on three sides. This arrange- Eve’schoiceoftopics.The‘‘tidy’’carddisplayed
ment limited potential distractions from others apictureofsomeonecleaningandsignaledaccess
in the library. Sessions were conducted at a to sorting three colors of poker chips that were
square table that seated four. The experimenter spread out on the table.
sat at a right angle to Eve. Evecompletedworksheetsduringthereinforc-
Materials. Cards (12.7 cm by 7.6 cm) indi- er assessment phase. The worksheets were pieces
cated the available preferred activities during ofpaper(21.6by27.9cm)thatcontainedtwoto
preference and reinforcer assessments.The cards 20 problems, with a maximum of 20 problems
included text and pictorial representations of in two columns of 10 problems. Each problem
preferred activities that were identified by Eve’s contained two one- or two-digit numbers with a
teacher. In the leftmost 5.1 cm of the card, text space in between for Eve to write in a . or ,
wasdisplayedinTimesNewRoman30fontthat symbol. These worksheets were identical to work
represented each choice. In the remaining space she completed in her classroom.
ontherightwasacenteredpictureoftheobject. Procedure. Study 1 involved two phases. The
Ten cards were created. The ‘‘draw on white- firstphasewasapreferenceassessmentdesignedto
board’’ card included a picture of a dry-erase identify a ranked preference of various activities.
board and signaled access to a dry-erase board The second phase was a reinforcer assessment
(21.6 cm by 27.9 cm) and marker. The ‘‘make designed to compare the reinforcing effects of
copies’’carddisplayedapictureofamanusinga stimuli chosen from the preference assessment
copy machine and signaled access to making when a work requirement was present.
copies in a copy room down the hall from the Preference assessment. A multiple-stimulus
library. The‘‘movie’’ carddisplayed a picture of without replacement (MSWO, DeLeon &
amoviecameraandsignaledaccesstowatchinga Iwata, 1996) preference assessment was con-
preferred movie on a television in the library. ducted with 10 stimuli. Five sessions were
The ‘‘computer’’ card included a picture of a conducted, each consisting of 10 choices. To
computer and signaled access to a laptop begin a session, the experimenter arranged the
computer on which Internet games could be 10 choice cards 2.5 cm apart in a semicircle
played. The ‘‘book’’ card contained a picture of array. Prior to any choices, the experimenter
a book and signaled access to a preferred book stated, ‘‘I am going to review your choices. You
850 DANIEL M. FIENUP et al.
can do each activity for 3 min. Follow along problems, FR 2 involved completing four
with your finger.’’ The experimenter then problems, and FR 20 involved completing two
pointed to each card and read the verbal worksheets (20 problems 3 2 worksheets 5 40
description. After all 10 choice cards were problems). Unique worksheets were created for
identified, the experimenter said, ‘‘Which activ- each of the FR requirements that contained the
ity would you like to do first? Point to the card number of problems specified by the schedule.
with the activity you would like to do.’’ When Within an FR requirement, several versions of
Eve pointed to a card, the experimenter said, the worksheet were created that varied by the
‘‘You have 3 minutes for —. Once I start the problems presented and order of problems.
timeryoucanbegin.’’Thetimerwasstarted,and ThestimuliusedinthisphaseweretheHPand
Eve was allowed 3 min of free access to the LP activities identified in the preference assess-
activity she chose. When the timer went off, the ment. During this assessment, three identical
experimenter removed the activity and said, worksheets,separatedby15.2cm,wereplacedin
‘‘Your time is up. Have a seat atthe table.’’ The front of Eve. The HP card and LP card were
experimenter removed the card Eve chose on placed above two of the worksheets, leaving one
the previous trial. To vary the order of stimuli, worksheet without a card above it (referred to as
theexperimenterthenremovedthefurthestcard theno-Sr+option).Theplacementofcardsabove
to the left and placed it farthest to the right on worksheets was determined randomly. Prior to
the semicircle. Then the experimenter said, making a selection, the experimenter said, ‘‘You
‘‘Whatactivitywouldyouliketodonext?Point can choose whichever worksheet you want to
to the card with the activity you would like to do. Each worksheet is the same. If you chose
do.’’ These procedures continued until the thisworksheet,youcanearn—.Ifyouchoosethis
participant made 10 choices. worksheet, you can earn —. If you choose this
Afterfivesessions,theorderinwhichastimulus worksheet, you will not earn an activity.’’ Then,
waschosenwasaddedacrossthesessions,resulting the experimenter asked Eve to repeat the
in a number that represented relative rank. The contingencies. The experimenter pointed to a
stimulus with the lowest number was designated worksheet, and Eve named the associated pre-
thehigh-preference(HP)optionforthereinforcer ferred activity. Eve called the no-Sr+ option
assessment.Thestimuluswiththehighestnumber ‘‘nothing.’’ Next, Eve was asked to choose a
wasdesignatedthelow-preference(LP)optionfor worksheet. She chose a worksheet by either
the reinforcer assessment. naming the associate activity (HP, LP, no Sr+) or
Reinforcer assessment. A three-choice concur- touchingaworksheet.Evethenbegantheworkand
rent-operants procedure was conducted to eval- wasgiven3minofaccesstotheassociatedpreferred
uate the reinforcing efficacy of the stimuli activitywhenshecompletedtheworkrequirement.
identified in the preference assessment. Each However,ifshechosetheworksheetwiththeno-Sr+
reinforcer assessment session consisted of six option, the next trial started immediately. This
trialsinwhichthreeidenticalworkrequirements pattern continued until she completed six work-
werepresentedtoEve,eachresultinginaccessto sheets for a session. Once she chose an option, she
different preferred activities. The work require- wasnotpermittedtoreceiveadifferentconsequence,
ment was increased whenever responding was although she never requested to do so. For each
stable with a given work requirement. A unit of session, the number of choices of each worksheet
workwasdefinedasthecompletionoftwomore andassociatedactivitywerecalculatedandconverted
than/less than math problems. The following topercentageoftrialschosen.
scheduleswereevaluated:FR1,FR2,FR5,FR Design. During the reinforcer assessment,
10, and FR 20. FR 1 involved completing two experimentalcontrolwasdemonstratedthrough
PREFERENCE FOR FLUENT WORK 851
Table 1
Preference Assessment Results
Sessionranking
Stimuli 1 2 3 4 5 Total
drawonwhiteboard 8 2 1 1 1 13
makecopies 10 4 5 2 2 23
movie 2 5 2 10 6 25
computer 3 1 3 9 10 26
book 4 6 9 5 3 27
GreatStateswithstaff 1 9 6 4 7 27
goforawalk 9 3 4 6 8 30
tidy 5 7 8 3 9 32
handshakeandchat 6 8 7 7 5 33
GreatStatesalone 7 10 10 8 4 39
Note.EachcolumnrepresentsEve’swithin-sessionrankofthestimuli.TheTotalcolumnrepresentstheadditionof
ranks across all five sessions for a given stimulus. Thus, lower numbers represent more preferred stimuli, and higher
numbersrepresentlesspreferredstimuli.Dataarepresentedintheorderofpreference,withthemostpreferredstimuliat
thetop ofthe table andleast preferred stimuliat the bottom.
theuseofaconcurrent-operantsprocedure.The sessions and 32% of the reinforcer assessment
no-Sr+ option was considered a control choice sessions. For all preference assessments and
because it was not expected that Eve would reinforcer assessments, the procedures were
select this option with the availability of HP implemented with 100% integrity.
and LP activities.
Interobserveragreementandtreatmentintegrity. Results and Discussion
For the purpose of calculating interobserver Table 1 displays the results of Eve’s prefer-
agreement, an independent second observer ence assessment.‘‘drawonwhiteboard’’hadthe
collected data during 20% of the preference lowest overall rank and ‘‘Great States alone’’
assessment sessions and 32% of the reinforcer had the highest overall rank. Thus, for the
assessment sessions. Agreements and disagree- reinforcer assessment, ‘‘draw on whiteboard’’
ments were assessed on each trial during a served as the HP option and ‘‘Great States
session. An agreement entailed both observers alone’’ served as the LP option.
recording identical choices (e.g., HP, LP, or no Figure 1 displays the results of the reinforcer
Sr+). Per session, interobserver agreement was assessment.Duringthefirstthreesessionsofthe
calculated by dividing the number of agree- FR 1 phase, Eve chose work associated with
ments by the total number of observations and each of the available options (HP, LP, and no
multiplying that number by 100%. Agreement Sr+). In Sessions 4 and 5, she exhibited
for all preference assessment and reinforcer preference for work associated with the HP
assessment sessions was 100%. activity, but reverted back to even distribution
A treatment integrity data sheet was created of choice during Session 6. During Sessions 7,
for both preference and reinforcer assessment 8, and 9, she demonstrated exclusive preference
sessions. Each data sheet contained all steps of for the work associated with no Sr+. When the
the assessment, including placement of choice schedule was increased to FR 2 and FR 5, she
cards, delivery of instructions, reviewing con- demonstrated exclusive preference for the no-
tingencies with Eve, and allowing access to Sr+ work option. During the first session of FR
chosen stimuli. Treatment integrity data were 10, she chose the HP option on five of six trials
collected on 20% of the preference assessment and the no-Sr+ work option for the remaining
852 DANIEL M. FIENUP et al.
Figure 1. Percentage of trials in which Eve chose work associated with a high-preference (HP) activity, low-
preference (LP)activity, andnoSr+during the reinforcerassessment.
trial. However, in the remaining FR 10 and all completed academic work independently in her
FR 20 sessions, Eve demonstrated exclusive classroom.However,anumberofothervariables
preference for the no-Sr+ work option. could have accounted for the obtained results.
Originally, the no-Sr+ option was conceptu- First,thebehaviorofchoosingtheno-Sr+option
alized as a control condition. Using a two- and saying ‘‘nothing’’ could have been more
choice concurrent-operants procedure, DeLeon reinforcing than choosing one of the other
et al. (1997) demonstrated that participants options, regardless of the work and presence or
allocated responding equally between two absence of preferred activities that followed.
choices under low response requirements. The Although Eve was not required to state which
no-Sr+ option was added to enhance experi- choiceshewasmaking,shetendedtovocalizeher
mental control in the current study, because it choice in addition to touching the work sample
was expected that the participant would not she chose. To test this, Eve could be presented
select this option if her behavior was under with the same choices as during the reinforcer
control of the various choices. Nonetheless, Eve assessment, but without the work requirement.
demonstrated preference for the no-Sr+ work Thus, she could be asked to choose among the
option under all response requirements. following choices: ‘‘nothing,’’ ‘‘draw on white-
One interpretation is that the preference board,’’ and ‘‘Great States alone.’’ Under these
assessment did not identify a stimulus that was conditions, choosing ‘‘nothing’’ would indicate
a reinforcer for academic responding. This that this behavior was more automatically
outcome may not be surprising, given that Eve reinforcing than choosing the other options.
PREFERENCE FOR FLUENT WORK 853
Eve also could have chosen the no-Sr+ work et al., 1997), and thus did not permit this type
option because completing worksheets was of finding. It appeared to be worthwhile to
more reinforcing than engaging in the activities conduct follow-up analyses to investigate fur-
identifiedfromthepreferenceassessment.Ifthis ther the results obtained in Study 1. Study 2
was the case, she chose the no-Sr+ work option was a three-phase study designed to evaluate
to maintain or receive quicker access to another potential variables that may have influenced
trial with the relatively more preferred activity Eve’s choice behavior. Phase 1 evaluated the
(i.e., academic work). Thus, choosing the no- value of choosing no Sr+ relative to other
Sr+ option minimized the latency to access socially mediated stimuli. Phase 2 evaluated the
another worksheet. To test this, Eve could be value of work versus activities identified by the
asked to choose among completion of work- preference assessment. Phase 3 evaluated the
sheets, drawing on a dry-erase board (HP), and effect of work schedule and time in session on
playing Great States Junior by herself (LP). the efficacy of reinforcers.
Under these conditions, choosing worksheets
Method
would indicate that completing worksheets was
a more reinforcing activity than the activities Participant and setting. The participant and
identified from the preference assessment. setting were identical to Study 1.
AthirdreasonwhyEvechosetheno-Sr+work Materials. Some of the same materials from
optioncouldhavebeenduetohowthereinforcer Study 1 were used in Study 2, including the
assessment procedure allocated work within a ‘‘draw on whiteboard’’ (HP) and ‘‘Great States
session. It is possible that the scheduling of work alone’’ (LP) cards. The cards were created using
had an abolishing effect on the reinforcers (e.g., the same specifications as those in Study 1. The
Laraway, Snycerski, Michael, & Poling, 2003). no-Sr+ card had ‘‘nothing’’ written on the card
Eve may have chosen the no-Sr+ option because and displayed an empty space instead of a
the reinforcing value of the HP and LP activities picture. The ‘‘worksheet’’ card displayed a
wasdiminishedbybreakingupworkintosixtrials pictureof a math worksheet (described in Study
thatinvolvedthecompletionofoneworksheetat 1)andsignaledaccessto20-problemworksheets.
a time. This would indicate that disfluent work Phase 3 of Study 2 included choices among
abolishedtheefficacyofreinforcers.Furthermore, different schedules of work. Cards associated
it is possible that Eve learned that the sooner she with each schedule had text in Times New
completed her work, the sooner she escaped the Roman 24 font. At the top center of each card
sessions and returned to her classroom. Thus, was the word ‘‘schedule.’’ Two spaces below
choosingnoSr+ decreasedthesessionlength.To began the schedule activities. Each activity was
testbothhypotheses,Evecouldbepresentedwith written on a single bulleted line. The disfluent
schedules that vary in the temporal allocation of work and HP schedule had the activities in the
work(fluentanddisfluent)andinthenumberof following order: worksheet, draw on white-
activities contained in each schedule. board, worksheet, draw on whiteboard, work-
sheet, draw on whiteboard, worksheet, draw on
whiteboard, worksheet, draw on whiteboard,
STUDY 2
worksheet, draw on whiteboard. The fluent
In Study 1, Eve unexpectedly chose no-Sr+ work and HP schedule had the activities in the
work over work choices that were associated following order: worksheet, worksheet, work-
with HP and LP activities. Previous work on sheet, worksheet, worksheet, worksheet, draw
progressive-ratio reinforcer assessments includ- on whiteboard, draw on whiteboard, draw on
ed choices between various reinforcer options whiteboard, draw on whiteboard, draw on
without a no-Sr+ control option (e.g., DeLeon whiteboard, draw on whiteboard. The fluent
854 DANIEL M. FIENUP et al.
work and no-Sr+ schedule had the activities in to preferred activities. Two of the schedules
the following order: worksheet, worksheet, included access to the HP activity from the
worksheet, worksheet, worksheet, worksheet. preference assessment. One schedule did not
Procedure. A series of concurrent-operants include access to a preferred activity. The
procedureswasused.Inallcases,Evewaspresented disfluent work and HP activity schedule began
withthreechoicesandinstructedtoselectanoption. with the completion of a worksheet followed by
In Phase 1, Eve was presented with three free access to the HP activity for 3 min. This
choices that compared access to preferred pattern continued until six worksheets were
activitieswitha‘‘nothing’’optionintheabsence completed and18min ofaccesstothedry-erase
of a work requirement. Specifically, Eve was board elapsed. The fluent work and HP activity
presented with cards that represented the schedule began with the completion of six
following stimuli: access to the HP activity, worksheets followed by free access to the HP
access to the LP activity, and no Sr+. Selecting activity for 18 continuous minutes. The fluent
theHPcardproduced3minoffreeaccesstothe workandno-Sr+schedulebeganandendedwith
dry-erase board and marker. Selecting the LP the completion of six worksheets. All schedules
card produced 3 min of free access to the Great hadidenticaltotalworkrequirements.Schedules
States Junior board game that could be played thatinvolvedaccesstotheHPactivityproduced
alone(i.e.,notwiththeexperimenter).Selecting the same total amount of access to the activity.
the no-Sr+ card terminated the trial. The The fluent work and no-Sr+ schedule was
contingency for the no-Sr+ card was the same associated with the shortest session because it
as during the Study 1 reinforcer assessment. only included work and then the session was
Each session was composed of six trials. On terminated.
eachtrial,Evewaspresentedwiththethreechoice Each session involved a single choice trial.
cards, horizontally arranged with approximately Only one trial was conducted on a given day
2.5cmbetweeneachcard.Theorderofthecards because of the amount of work involved. For
was randomized across trials. For each trial, the each session, a single schedule was placed in
experimenter pointed to each card and asked, front ofEve,andsheandtheexperimenterread
‘‘Whatdoyougetifyouchoosethiscard?’’Once the list of activities, which involved stating that
Eve correctly labeled the stimuli (draw on white- ‘‘draw on whiteboard’’ was for 3 min. Then,
board, Great States alone, nothing), the experi- Eve was asked to read the schedule, but she was
menter asked, ‘‘Which one do you choose?’’ Eve required to read only the activities (worksheet
selectedacardbytouchingit.Sessionscontinued and draw on whiteboard) without stating time
untilaclearpreferencewasdemonstratedacrossa variables.Afterreviewingallthreeschedules,the
minimum of three sessions. experimenter placed the schedules horizontally
In Phase 2, Eve was presented with cards infrontofEveinarandomlydeterminedorder.
representing the following stimuli: access to the The experimenter asked, ‘‘Which one do you
HPactivity,accesstotheLPactivity,andaccess choose?’’ Eve selected a schedule by touching
to math worksheets. All other procedures were that schedule. A session was terminated after
identical to those in Phase 1. completing all activities listed on the schedule.
In Phase 3, Eve was presented with three Design. Similar to Study 1, all phases of
schedules. All schedules required the completion Study 2 used a concurrent-operants procedure
ofsixworksheets.Fluentworkschedulesinvolved to demonstrate experimental control.
completion of the worksheets consecutively. The Interobserveragreementandtreatmentintegrity.
disfluent work schedule broke up completion of As in Study 1, an independent second observer
the work requirements with periods of access collected data for 40% of the Phase 1 sessions,
PREFERENCE FOR FLUENT WORK 855
Figure 2. Percentage of trials in which Eve chose no Sr+, a high-preference (HP) activity, a low-preference (LP)
activity,andwork during concurrent-operants procedures.
33% of the Phase 2 sessions, and 33% of the RESULTS AND DISCUSSION
Phase3sessions.Agreementsanddisagreements
Figure 2displaystheresultsforPhase1.During
were assessed the same way as described in
the first two sessions, Eve chose each of the three
Study 1. Per session, interobserver agreement
optionsequally.Intheremainingthreesessions,she
was calculated by dividing the number of
demonstratedaclearpreferencefortheHPoption,
agreements by the total number of observations
and she infrequently chose the LP and no-Sr+
and multiplying that number by 100%.
options. Overall, these results demonstrated that
Agreement for all phases was 100%.
Eve had a clear preference for the HP option.
A treatment integrity data sheet was created
Figure 2 also displays the results for Phase 2.
for the three phases of Study 2. The data sheet
Overall,Evedemonstratedexclusivepreferencefor
included all steps of the assessment, including
the HP option in five of six sessions. Figure 3
placement of choice cards, delivery of instruc-
displays the cumulative choices in Phase 3. With
tions, reviewing contingencies with Eve, and
the exception of the second session, Eve always
allowing access to chosen stimuli. Treatment chosethefluentworkHPactivityoption.Shenever
integrity data were collected for 40% of the chosethedisfluentwork HP activity option.
Phase 1 sessions, 33% of the Phase 2 sessions, Taken together, a clear picture of Eve’s
and 17% of the Phase 3 sessions. Across all preferences was determined across the phases
phases, the procedures were implemented with of Study 2. Phase 1 ruled out automatic
100% integrity. reinforcement as an alternative explanation
856 DANIEL M. FIENUP et al.
Figure 3. The cumulative frequencyof Eve’s choicesamong different schedules during Phase3.
because Eve consistently chose the HP activity preference assessment procedure that is com-
over the LP activity or no-Sr+ choice. Phase 2 monly employed with individuals with intellec-
ruled out the possibility that Eve preferred to tual disabilities (Deleon et al., 1997; Tustin,
work rather than to engage in the activities, 1994). The original aim of Study 1 was to
because she chose to engage in the HP activity examine whether students, engaging in more
more than the LP activity or completing typical school activities, would demonstrate
worksheets. Phase 3 demonstrated that, when similar patterns of behavior to those in the
able to choose a work schedule and conse- earlier two studies. Due to Eve’s unusual
quence, Eve chose to complete her work all at pattern of responding, we conducted follow-
once (fluent work) and consume all of her up analyses to examine why she chose to work
reinforcing activities at once. This choice for no reinforcer instead of either the HP or LP
resulted in a relatively longer time spent in the activity that had been identified via the
session. Thus, Eve was not trying to escape preference assessment. A series of concurrent-
sessions, because she selected the fluent work operantsevaluationsrevealedthatEvechose the
option (no Sr+) only once. no-Sr+ option because she preferred to com-
plete all of her work at once. The HP activity
functioned as a reinforcer only when Eve was
GENERAL DISCUSSION
able to select fluent work schedules.
This set of studies produced unexpected GiventhefinaloutcomesofStudy2,onecan
findings that resulted from a well-researched reframe the contingencies that governed Eve’s