146(Autism Vorbire)

8/12/2019 146(Autism Vorbire)

http://slidepdf.com/reader/full/146autism-vorbire 1/11

http://foa.sagepub.com

DisabilitiesFocus on Autism and Other Developmental

DOI: 10.1177/10883576093383802009; 24; 146 originally published online Jun 17, 2009;Focus Autism Other Dev Disabl

Jessica H. Franco, Russell L. Lang, Mark F. O'Reilly, Jeffrey M. Chan, Jeff Sigafoos and Mandy Rispolifor a Child With Autism

unctional Analysis and Treatment of Inappropriate Vocalizations Using a Speech-Generating Device

http://foa.sagepub.com/cgi/content/abstract/24/3/146 The online version of this article can be found at:

Published by: Hammill Institute on Disabilities

and

http://www.sagepublications.com

can be found at:Focus on Autism and Other Developmental DisabilitiesAdditional services and information for

http://foa.sagepub.com/cgi/alertsEmail Alerts:

http://foa.sagepub.com/subscriptionsSubscriptions:

http://www.sagepub.com/journalsReprints.navReprints:

http://www.sagepub.com/journalsPermissions.navPermissions:

by Otilia Friedl on October 22, 2009http://foa.sagepub.comDownloaded from

http://www.hammill-institute.org/

http://foa.sagepub.com/cgi/alerts


http://foa.sagepub.com/subscriptions


http://www.sagepub.com/journalsReprints.nav



http://www.sagepub.com/journalsPermissions.nav


http://foa.sagepub.com/







http://www.hammill-institute.org/



Focus on Autism and OtherDevelopmental Disabilities

Volume 24 Number 3September 2009 146-155

© 2009 Hammill Institute onDisabilitie

10.1177/1088357609338380http://focus.sagepub.com

hosted http://online.sagepub.com

Functional Analysis and Treatmentof Inappropriate Vocalizations Usinga Speech-Generating Device for a ChildWith Autism

Jessica H. FrancoThe University of Texas at Austin

Russell L. LangUniversity of California, Santa Barbara

Mark F. O’ReillyJeffrey M. ChanThe University of Texas at Austin

Jeff SigafoosVictoria University of Wellington

Mandy RispoliThe University of Texas at Austin

The purpose of this study was to extend the research on functional communication training by examining the use of aspeech-generating device (SGD) for a 7-year-old child with autism and no spoken language who demonstrated inappropri-ate vocalizations that served multiple functions. An analog functional analysis was used to determine the function of the

behavior and the potential viability of the SGD as a treatment option. The child was taught to discriminate among multipleoptions on the SGD and then to choose an appropriate message in two generalization settings. When the SGD was available,the child reduced his inappropriate vocalizations across all settings and increased his engagement in both appropriateactivities and interactions with others.

Keywords: functional analysis; speech-generating device; autism; functional communication training

Inappropriate vocalizations may include combined orisolated sounds that are used in a repetitive manner.

When such vocalizations occur at a high frequency andare disruptive to the environment, they can be consid-ered challenging behavior. Frequently, these disruptivevocalizations are assumed to be stereotypic in nature

and therefore might not be conceptualized in terms oftheir operant function (Matson, Kiely, & Bamburg,1997). However, some studies have found that stereo-typic behaviors are often maintained by escape fromdemands, attention from others, or by sensory conse-quences (Ahearn, Clark, MacDonald, & Chung, 2007;Kennedy, Meyer, Knowles, & Shukla, 2000; Mace,Browder, & Lin, 1987).

One method for treating challenging behavior is func-tional communication training (FCT; Durand, 1990), in

which challenging behaviors are conceptualized ascommunication acts. With FCT, a child is taught to useappropriate communicative responses in lieu of the chal-lenging behaviors. For example, if a child is usingaggression to gain attention, an appropriate alternativeresponse might be to teach the child to say, “Come play

with me.” If the child is using challenging behavior toescape from demands, an appropriate alternative responsemight be, “I need a break.”

The treatment of inappropriate vocalizations is oftendifficult because individuals who engage in such behav-iors may have little to no ability to produce intelligiblespeech. This limits their ability to make alternative ver-

bal responses. One option for individuals with limitedspeech is the use of augmentative and alternative com-munication (AAC), including speech-generating devices







Franco et al. / Speech-Generating Device 147

(SGDs; Schlosser & Lee, 2000). An individual is taughtto push a key on such a device that produces a recordedspoken message.

Researchers have used AAC to reduce challenging behaviors in a number of studies, but only a few haveused SGDs (Durand, 1993, 1999; Northup et al., 1994;Peck et al., 1996; Steege et al., 1990; Winbon, Wacker,Richman, Asmus, & Geier, 2002). All of the researcherswho used SGDs for FCT focused on teaching a request asa replacement for challenging behavior. For example, achallenging behavior may have been reinforced by atten-tion, access to preferred items, or escape from demands(Durand, 1993), and therefore, the corresponding AACreplacement involved requesting attention, preferred items,or a break from the task. On the basis of the results of afunctional analysis, an SGD can be programmed with arequest that provides the same reinforcing outcome asthe challenging behavior.

In all of the FCT studies in which researchers haveused SGDs, single response choices were programmedinto the devices. Activation of the devices produced ageneral requesting topography, such as “I want more.”Although this type of basic request may lead to an initialdecrease in challenging behavior, there is little evidencethat such interventions will teach generalized communi-cation skills. Furthermore, when an SGD is programmedwith only one basic request, the message might be mean-ingful in only a limited number of settings. In fact, gen-eralization of the target request has been addressed inonly a few studies (Durand, 1999; Northup et al., 1994;

Peck et al., 1996).Finally, in previous studies on SGDs in FCT, research-

ers have focused on only one target function. Only one participant in previous studies of SGDs in FCT wastreated for challenging behaviors maintained by multiplefunctions (Durand, 1993). In that study, the participantwas a 5-year-old girl with cerebral palsy who engaged inchallenging behavior for both escape from task demandsand attention. Her target response, “I want to be with thegroup,” was thought to allow her to both briefly escapefrom demands and also recruit social attention. One ofthe participants in Winbon et al.’s (2002) study demon-strated two functions of challenging behavior (attentionand escape) but was treated only for escape using an SGD.To date, none of the researchers using SGDs within FCTinterventions have addressed multiple functions of chal-lenging behavior with multiple message options or exam-ined teaching novel responses in generalized settings.

The purpose of this study was to extend the FCT andSGD research by examining the effects of an SGD pro-grammed with multiple messages on challenging beh aviormaintained by multiple functions (e.g., access and

escape). The participant, a child with autism, was firsttaught to use the multifunction SGD in a controlled set-ting to test the viability of such an intervention underexperimental control. The child was then given the SGDin two applied settings to evaluate generalization and theecological validity of such an intervention. In these gen-eralization settings (gymnasium and playground), theintention was to teach the child to (a) discriminate amongmultiple response options on the SGD, (b) choose anappropriate message on the basis of various conditions andsettings, and (c) reduce vocal stereotypy while increasinghis engagement in appropriate activities and with others.

General Method

Participant

Brent was aged 7 years, 6 months. He was diagnosed

by a neurologist as having developmental delays at 10months and was later diagnosed with autistic disorder at2 years by the same neurologist. At the onset of thestudy, Brent was nonverbal and requested preferreditems and activities using manual sign approximations,reaching, and hand leading. Parents and teachers reportedthat he occasionally used the signs for “movie,” “candy,”“swing,” and “music.” However, these signs were adaptedfrom standard American Sign Language because ofdelays in fine motor skills. As a result, these signs werenot easily understandable to anyone who was notspecifically trained in their meaning. Brent used these

communicative behaviors several times per day and onlyfor requesting. According to teacher reports, Brent

played independently with toys designed to providevisual and auditory stimulation but did not play with

peers. He receptively identified 10 specific objects whenthey were presented in an array of three. Brent followedone-step instructions and simple directions, such as “sitdown.” He also imitated a variety of actions with objectsand gross motor actions with his arms and legs. Brent didnot imitate any sounds or words. Brent scored a 40.5 onthe Childhood Autism Rating Scale (Schloper, Reichler,Devellis, & Daly, 1980), placing him in the “severeautism” category. Additionally, he scored a 32.8 on theVineland Adaptive Behavior Scales (Sparrow, Balla, &Cicchetti, 1984), for an age equivalent of 1 year, 9months, for daily living skills. Both assessments wereadministered to Brent’s teacher by an advanced doctoralstudent trained in the assessment procedures.

Brent attended a private school for children with autismspectrum disorder 5 days per week, 6 hours per day. Brent’sclassroom included seven students with autism spectrumdisorders, aged 3 to 8 years; four lead teachers; and two







148 Focus on Autism and Other Developmental Disabilities

assistant teachers. Throughout the day, he receivedapproximately 2 hours of one-on-one academic instruc-

tion, 2 hours of small-group instruction targeting playand motor skills, and 2 hours of leisure and self-helpactivities (e.g., lunch, music, recess). The staff usedinstructional practices based on applied behavior analy-sis. Brent also was receiving occupational and speechtherapy prior to and during the study. Goals for occupa-tional therapy focused on limb strengthening and coordi-nation; goals for speech therapy focused on teachingrequesting via manual signs, vocalizations, and pictureselection. He was not on any medication.

Setting

All sessions took place at Brent’s school. Specificlocations for instruction within the school included theclassroom, playground, and gymnasium. Functionalanalysis and initial SGD training took place in the class-room setting in Brent’s “work space” area, containing achild-sized desk. The area was separated from the rest ofthe room with partitions. Following training and analysisof SGD use across target functions, Brent was givenaccess to the device in two generalization settings, the

playground and the gymnasium. The playground was alarge, fenced-in outdoor area containing a playscape,

swings, hanging bars, and picnic tables. The gymnasiumwas a large indoor room with designated space for swings,crash mats, a trampoline, a trapeze, and a track for tri-cycles and scooter boards.

Materials

During treatment sessions, Brent used an SGD (GoTalk; DysphagiaPlus LLC, Shreveport, LA) with a staticdisplay of 12 choices. This device was chosen on the

basis of the following selection factors: (a) photographs

of actual stimuli from Brent’s environment could be usedin the device, as opposed to the more abstract computer-

generated graphic displays; (b) the device cost less than$300; (c) Brent’s parents indicated a preference for thedevice; (d) the device’s activation keys were large enoughto be used easily despite fine motor delays; (e) 12 mes-sages could be recorded per overlay; and (f) the deviceallowed easy navigation between multiple overlays.Review of Brent’s academic and communication skills

prior to the study suggested that he was not able to dis-criminate icons or symbols but did well with discrimina-tion of real pictures of items. Although he demonstratedweak receptive identification skills (i.e., he was unableto consistently demonstrate understanding of more than

10 object names), he was able to match identical objectsand pictures to sample when presented in an array of 10and could complete inset puzzles with 12 pieces by trialand error. Thus, it was hypothesized that Brent’s difficul-ties with receptive identification were more likely due toweak auditory processing or comprehension skills ratherthan a lack of the ability to scan a large array.

Four separate overlays were used with Brent: (a) tan-gible condition in the classroom, (b) escape condition inthe classroom, (c) multiple functions in the gymnasium,and (d) multiple functions on the playground. For eachoverlay, the device was programmed with single-wordmessages that corresponded to the names of various pre-ferred activities and items specific to the setting and/or ashort phrase requesting a break. Specific messages foreach overlay are listed in Table 1. The phrase “I want a

break” was available across all settings. The phrase“Come with me” was available in both generalization set-tings. Each overlay consisted of real photographs of itemsand activities or meaningful symbols that representedmessages. For example, a stop sign was used for the mes-sage “I want a break.”

Table 1Items and Phrases Available on the Speech-Generating Device

Grouped by Setting and Function of Challenging Behavior

Classroom (Tangible) Classroom (Escape) Gymnasium (Multiple Functions) Playground (Multiple Functions)

Puzzle I want a break. I want a break. I want a break.Book Come with me. Come with me.

Movie Rainbow swing SwingDuck game Frog swing SlideChip Bike Rock wallMarble Crash mats Glider Music toy Trapeze BridgeBubbles Big ball StairsKeyboard Scooter board Shovel Sensory toys Teeter totter Basketball Toys









Study 1: Functional Analysis

Target Behavior

Challenging behavior in this study was defined as an“ee” sound sustained longer than 2 s. This behavior wastargeted because teachers and parents reported that the

“ee” sound disrupted instruction and served as a distrac-tion for others. This behavior also interfered with Brent’sacquisition of new skills. When Brent engaged in thisrepetitive vocalization, he oriented his body away fromthe teacher and his volume exceeded that of the teacher’svoice. As a result, Brent frequently failed to respond cor-rectly to teacher directives. This noise also made it dif-ficult for teachers and other students to communicatewith one another.

Data Collection

All analysis and intervention sessions were 5 min inlength. A 10-s partial-interval data collection system wasused to record challenging behavior during all sessions.This interval length was chosen because the duration ofthe behavior was highly variable and because initial

probing of different interval lengths determined this to be manageable and accurate.

All sessions were conducted by the first or secondauthor. Interobserver agreement (IOA) data were col-lected by two of Brent’s classroom teachers. Theseteachers were trained in all data collection procedures

prior to the study by the second author. Both teachers wereasked to observe Brent in vivo and discuss with the trainerwhether the behavior they were observing met the opera-tional definition of challenging behavior. Once a col-lective understanding of the target behavior was reached,the procedures were practiced by role-playing. The trainer

pretended to be the child, and the teachers collected data.After the data were collected, scores were compared anddiscussed. This process was repeated in role-play until100% agreement was reached. Finally, the teachers wereasked to code data while observing Brent in vivo.Initially, they were instructed to stand next to each other,

discuss, and compare observations while they coded.When the teachers reported feeling competent, they wereseparated and asked to code independently. This wasdone until 90% agreement was obtained for three con-secutive 5-min intervals. This entire training proceduretook two sessions lasting approximately 45 min each andoccurred over a 2-day period. IOA was calculated bydividing the number of agreements (occurrence plusnonoccurrence) by the total number of intervals and thenmultiplying by 100%. IOA was collected during 61% of

the functional analysis sessions. Mean IOA was 96%(range = 90% to 100%).

The second author collected treatment fidelity data on30% of functional analysis sessions. Data collectors useda checklist to score implementation. The checklist con-tained the following three components: (a) the target

behavior was followed by the appropriate consequence,(b) the consequence was never delivered independent ofthe target behavior (except in the play condition), and(c) the consequence was delivered for 10 s. A tally wasmade under a column labeled “correct” next to the sec-ond and third components of the checklist each time thatcomponent was correctly implemented. When an erroroccurred, a tally was made under a column labeled “error.”Treatment fidelity was calculated for each observed ses-sion by dividing the number of correctly completed steps

by the total number of steps (correct plus error) and mul-tiplying by 100%. Mean treatment fidelity was calculated

across each of these sessions by summing the individualsession percentages and dividing by the number of ses-sions. Mean treatment fidelity for the functional analysiswas 99% (range = 97% to 100%).

Experimental Design

Functional analysis sessions were conducted in a combined multielement and reversal design. In the first phase,a standard analog functional analysis was conducted.This analysis consisted of four conditions: tangible,attention, demand, and play (Hanley, Iwata, & McCord,

2003). Five sessions of each of these conditions wereimplemented in a multielement treatment design format.Brent did not have access to the SGD during this phase.In the second phase, two conditions were compared. Afunctional analysis tangible condition in which Brent hadaccess to an SGD that was programmed to request tan-gible items was compared with the same tangible condi-tion without access to the SGD. The third phase (reversal)was identical to the first phase, except that only threesessions were conducted for each of the four conditions.This reversal phase was included to demonstrate that theconsequences maintaining the challenging behavior hadnot changed. In the final phase, two more conditions werecompared. In one condition, the SGD was programmedwith a message indicating the desire to escape from taskdemands. In the other conditions, the SGD was not present.

Procedure

Functional analysis . During the standard functionalanalysis sessions in Phases 1 and 3, four conditions(demand, tangible, attention, and play) were randomly








selected for order and alternated in 5-min sessions. Thefirst author implemented all assessment sessions.

During the demand condition, matching and receptiveidentification tasks were presented to Brent. These dem-ands were chosen in collaboration with the classroomteacher and were consistent with Brent’s individualizededucation plan. Upon the occurrence of challenging

behavior, the task materials were removed, and theresearcher turned away for 10 s. New trials were pre-sented immediately following correct responses or the10-s task-removal consequence. A correct response wasdefined as completion of the presented matching orreceptive ide ntification task. If Brent did not attempt tocomplete the task, a least-to-most prompting system wasused. The correct response was prompted and thedemand was then immediately presented again with aless intrusive prompt.

In the tangible condition, Brent was first allowed brief

access to several preferred items (marbles, a toy castle,and electronic storybooks). Then these items wereremoved and placed out of Brent’s reach but still in hissight. Brent was then allowed 10 s of access to theseitems contingent upon challenging behavior. Otherwise,access to these items was denied. No demands were pla cedon Brent during tangible condition sessions, and atten-tion was limited to the minimum required for the thera-

pist to deliver the toys.During the attention condition, the therapist first

informed Brent that the therapist had work to do and thathe needed to play by himself for a little while. The thera-

pist sat at a table and pretended to read a magazine. Brentwas allowed constant access to the same preferred tan-gible items listed above. Upon occurrence of the chal-lenging behavior, the therapist turned to Brent and deliveredattention in the form of verbal praise and redirection for10 s (i.e., “You’re such a good boy, but I need you to stopmaking that noise, please”). During this condition, nodemands were placed on Brent.

Finally, in the control condition (play), Brent wasallowed constant access to preferred items. Attention inthe form of praise was delivered noncontingently every10 s. No demands were placed and all instances of chal-lenging behavior were ignored.

SGD training . Prior to implementation of the firstSGD-present session, Brent was taught to use the SGDto select three different tangibles during practice trialsusing a prompt-fade procedure with a most-to-least

prompting hierarchy. On the first trial, physical guidancewas used to show Brent the correct response. Then, animmediate gestural prompt (pointing) was used. If Brentresponded to the gesture prompt, the tangible item was

delivered. If he did not respond, physical guidance wasused, and the tangible item was delivered. After deliveryand removal of the tangible item, the SGD was presentedagain, along with whatever prompt was used immedi-ately prior to the last delivery of the tangible item. OnceBrent successfully responded to a gesture, the promptwas faded to a time delay. The time delay consisted of a5-s waiting period. If no response occurred after 5 s, theinstructor returned to a gestural prompt. Independentuse of the SGD was defined as activation of the appropri-ate message within 3 s of presentation of the device. Thecriterion for mastery was independent use of the SGD forfive consecutive trials for at least three different tangibleitems. Mastery for the first item was obtained within 30min and mastery of the second and third tangible itemswithin an additional 20 min.

Brent was taught to use the SGD to accomplish theescape function using the same procedures described

above. However, for escape function practice trials, whenhe was given a demand involving work materials, Brentwas prompted to push the “stop” key on the device, andthe materials were removed for 10 s.

SGD within the functional analysis . To systematicallyassess the effect of the SGD on only one function at atime, the SGD was introduced only into the tangiblecondition in Phase 2 and then only in the demand condi-tion in Phase 4.

In Phase 2, the SGD was examined in the tangible con-dition. In the SGD-present sessions, access to preferred

items was withheld, and the SGD was presented to Brent.The device was programmed with the overlay representedin column 1 of Table 1, which contains preferred itemsidentified by teacher and parent reports. If Brent used theSGD, he received access to the requested item for 10 s. Ifhe did not use the SGD within 5 s after the SGD was

presented, he was prompted with the question, “What doyou want?” During this phase, no additional promptingwas used, and any occurrence of challenging behaviorwas ignored. The SGD-present sessions and the no-SGDsessions were otherwise identical.

In Phase 4, the SGD was examined in the demandcondition. During the SGD-present sessions, Brent wasgiven the SGD while working on task demands. Thedevice was programmed with the overlay represented incolumn 2 of Table 1. If Brent activated the device, thetherapist withdrew the task for 10 s. Challenging behav-ior was ignored. The no-SGD sessions were identical tothe demand condition of the functional analysis, in whichtask demands were removed contingent on the occur-rence of challenging behavior. The same demands wereused in the SGD-present and no-SGD sessions.









Results

The results of the functional analysis are displayed inFigure 1. During Phase 1 (Sessions 1 to 5), inappropriatevocalizations were higher in the tangible, demand, andattention conditions relative to the control (play) condi-tion, suggesting a multiply controlled behavior. However,the percentages of time Brent engaged in inappropriatevocalizations in the demand and tangible conditions(64% to 93% and 46% to 100%, respectively) were sub-stantially higher than in the attention condition (0% to28%). Additionally, the levels of behavior in the demandand tangible conditions were never lower than in the playcondition.

In Phase 2 of the functional analysis (Sessions 6 to 10),the no-SGD condition replicated the tangible conditionfrom Phase 1, with high levels of challenging behavioroccurring during all sessions (range = 37% to 80%). Theoccurrence of challenging behavior in the SGD-present

condition was substantially lower, ranging from 3% to20%.

In the reversal phase (Sessions 11 to 13), the results ofthe tangible and demand conditions were comparablewith the findings of the first phase. Challenging behavioroccurrence ranged from 47% to 86% of intervals in thetangible sessions and from 63% to 70% of intervals in thedemand sessions. The results of the attention conditionwere indistinguishable from those of the play condition.

In Sessions 14 through 18, the results of the compari-son between demand conditions with and without the

SGD are displayed. The results of the no-SGD conditionwere similar to those of the demand conditions of the

previous functional analysis, with challenging behavioroccurring in 60% to 80% of intervals. However, with theSGD present, the level of inappropriate vocalizationswas between 0% and 3%.

Study 2: Intervention in Applied Settings

Target Behaviors

Challenging behavior in this study again was definedas the “ee” sound sustained for longer than 2 s. Use ofthe SGD was defined as the activation of the device bythe participant when it was presented without promptsfrom an adult. Engagement was defined as active par-ticipation in appropriate activities or interactions withothers. Examples of engagement with activities includedBrent working with instructional materials in the gymna-

sium (e.g., identifying heavy vs. light objects) or usingthe equipment on the playground (e.g., going down theslide). Because Brent was capable of making the “ee” soundwhile participating in appropriate activities (e.g., swingingwhile producing the sound), challenging behavior andengagement were not always mutually exclusive.Examples of interactions with others in the gymnasiuminclude following directions from the teacher (e.g., liningup) and sharing an item with a peer. Examples of interac-tions with others on the playground include playinggames (e.g., chase) and following teachers’ directions.

Figure 1Percentage of 10-s Intervals With Challenging Behavior (CB) During Four Phases of Functional Analysis

0102030405060708090

100

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

P e r c e n

t o

f I n t e r v a l s W i t h C B

Session

Functional Analysis With and Without SGD

Attention

Play

No SGD

SGD present

Demand

Tangible

Play

Attention

No SGD

SGD present

Tangible Demand

Phase 1 Phase 2:Tangible Only

Phase 3 Phase 4:Demand Only

Note: SGD = speech-generating device.









Behavior was not coded as engagement when Brent wasinteracting with peers or teachers while producing the“ee” sound. Nonexamples of engagement included stand-ing still and the use of materials in an inappropriate way(e.g., rubbing face on swing).

Data Collection

Occurrence of the challenging behavior was recordedduring generalization and follow-up using a 10-s partial-interval method, as in Study 1. However, engagement wasscored using a 10-s whole-interval data system. This sys-tem was chosen because engagement occurring for lessthan 10 s was difficult to identify and potentially lessmeaningful than engagement sustained for at least 10 s.During the follow-up sessions, observers collected data onthe specific messages Brent selected on the SGD. Thesedata were recorded by making a tally under a columnlabeled “escape” or “tangible” after each correct use of theSGD. A tally was made in the escape column if the “I wanta break” button was pressed and in the tangible column ifa button corresponding to a tangible item was pressed.

IOA data were collected on challenging behavior bythe teachers trained in the first study. The teachers werealso trained to score engagement using the same training

procedures described above; training to criterion forengagement was completed during one 45-min session.IOA was collected on both engagement and challenging

behavior during 48% of the generalization sessions(inc luding baseline and follow-up). Mean IOA during

generalization in the gymnasium was 98% (range=

97%to 100%), and IOA during generalization on the play-ground was 100%.

Experimental Design

A multiple-baseline design across two settings (play-ground and gymnasium) was used. Baseline, interven-tion, and follow-up sessions were conducted in eachsetting. The playground and gymnasium were chosen astargets on the basis of teacher reports that challenging

behavior occurred most frequently in these settings.Instruction in social skills, self-help skills, and receptiveidentification of objects occurred in these environments.Therefore, in both settings, multiple functions were prob-able given the presence of demands and preferred activi-ties. Two different overlays on the SGD were used in eachsetting as represented in columns 3 and 4 of Table 1.

Procedure

Baseline . During baseline sessions, Brent was observedin the setting, and teachers were instructed to interactwith Brent as usual. The SGD was not present. Baseline

began on the same date for both settings, but introduc-tion of the SGD was staggered across settings.

Intervention . The SGD was presented to Brent approx-imately every 15 s, and he was allowed to select a mes-sage. The instructor then provided the appropriateconsequence for the message that was selected (i.e.,

break from task or access to a preferred tangible). Noadditional prompts to use the SGD were given, and alldisruptive vocalizations were ignored.

Follow-up . The researchers returned 3 months after thelast intervention-phase data point to collect maintenancedata. In that time, the teachers had voluntarily expandedthe use of the SGD across Brent’s entire day. However,follow-up data were collected only on the playground andgymnasium settings. Additionally, data were collectedregarding which messages were being used in each environ-ment. This was done to determine if the SGD was being usedexclusively for just one function or if it was used acrossmultiple functions. The same intervention procedures inwhich Brent was presented the SGD every 15 s were used.

Results

The results of Study 2 are displayed in Figure 2. During baseline in the gymnasium, engagement was low (range = 0% to 10%), and inappropriate vocalizations were high( M = 88%, range = 73% to 100%). After the SGD wasintroduced in the gymnasium, the level of inappropriate

vocalizations dropped ( M =

40%, range=

20% to68%), and engagement increased ( M = 59%, range = 33% to 68%).

During baseline on the playground, inappropriatevocalizations were high ( M = 94%, range = 85% to100%), and engagement was low ( M = 2%, range = 0%to 3%). After the SGD was introduced on the playground,inappropriate vocalizations dropped ( M = 28%, range = 10% to 50%), and engagement increased ( M = 68%,range = 50% to 83%).

Follow-up . In the gymnasium, challenging behavior

remained low ( M = 6%, range = 3% to 6%), and engage-ment remained high ( M = 73%, range = 63% to 87%).On the playground, challenging behavior also remainedlow ( M = 10%, range = 7% to 13%), with high engage-ment ( M = 77%, range = 70% to 87%).

Combined across all of the follow-up sessions in thegymnasium, Brent selected the escape message 10 timesand messages related to tangibles 32 times. On the play-ground, Brent selected messages related to the escapefunction 6 times and messages related the tangible func-tion 48 times.









Discussion

These studies examined the analysis and treatment ofinappropriate vocalizations of a 7-year-old child withautism. The results of a functional analysis suggested

that the vocal stereotypy was maintained by multiplesocial functions. The success of the intervention seems to provide verification of the findings of the functionalanalysis. Inappropriate vocalizations were reduced, andengagement was increased. These findings highlight sev-eral implications regarding the use of an SGD in an FCTintervention.

First, the results of this study suggest that the use ofan SGD may be an effective treatment for inappropriatevocalizations in individuals with autism. The occurrenceof inappropriate vocalizations was substantially reduced

during each of the conditions in which the participantused an SGD. Additionally, the participant was able togeneralize the use of the device to two different settings,resulting in a subsequent reduction of inappropriatevocalizations across environments.

In addition to reducing the rate of inappropriate behavior, the intervention resulted in an increase inengagement. Brent’s ability to request breaks and access

preferred items did not reduce his access to the curricu-lum. Instead, Brent spent more time interacting withteachers and peers as well as engaging in appropriateactivities. In Brent’s school, instruction is conductedacross all environments and not limited to the classroom.Therefore, an increase in engagement on the play-ground and in the gymnasium meant an increase inopportunities for ins truction. This is noteworthy, because

Figure 2Percentage of 10-s Intervals With Challenging Behavior and Engagement Across Generalization Settings

Note: SGD = speech-generating device.









some professionals may worry that an FCT interventionthat allows additional breaks could reduce the quantity ofinstructional time. Further research concerning this find-ing is warranted. A comparison between allowing breaksand placing such requests on extinction might provideadditional information regarding this finding.

This study is the first to demonstrate that an SGD cansuccessfully be used across multiple functions of challeng-ing behavior. Behaviors that are maintained by multiplefunctions are common; treatment of such behaviors oftenrequires multiple intervention components and may becomplicated by conflicting functions. For example, anintervention for challenging behavior maintained by

both attention and escape might contain opposing inter-vention components: If attention is denied, the childescapes; if escape is denied, the child may have access toadult attention. The use of an SGD to address both func-tions may be more efficient. A replication of this finding

across additional participants and combinations of main-taining consequences is needed.Although the functional analysis may have limited

ecological validity, its purpose in Study 1 was to demon-strate that Brent’s challenging behavior was maintained

by multiple functions and that the SGD could be usedeffectively to replace challenging behavior. Study 2 wasintended to be more ecologically valid. The best evi-dence for the ecological validity of the intervention is themaintenance of the intervention after the researchersended their involvement (Kennedy, 2002). Teachers con-tinued to use the intervention following the conclusion of

the study and reported to have expanded the use of theSGD across more of Brent’s environments (i.e., home,cafeteria, and one-on-one instruction).

This study also examined the generalizability of anFCT intervention using an SGD across settings. Each ofthe maintaining consequences, escape and access to tan-gibles, was first treated separately in a controlled setting(i.e., the classroom). Then responses for both functionswere made available on an SGD in two generalizationsettings. Brent successfully used the SGD in these gen-eralization environments without additional instruction.An SGD programmed with multiple functions may be

particularly suited for generalization. First, the physicaldevice itself may serve as a mediator between environ-ments (Stokes & Baer, 1977); the mere presence of thedevice serves as a prompt for its correct use and as adiscriminative stimulus for the availability of reinforce-ment. Second, by programming the device with multiplefunctions, the device can be used easily in multiple envi-ronments without the need for reprogramming. A chal-lenging behavior may be sensitive to one form of

reinforcement in one environment but have reduced sen-sitivity to that same reinforcer (in favor of another) in adifferent environment. For example, on the playground,access to tangibles may be more valuable and thereforeelicit more challenging behavior, whereas in the class-room, escape from task demands may become the most

powerful reinforcer (Lang et al., 2008, in press). Pro-gramming an SGD for multiple functions may compensatefor such discrepancies and help maintain treatment gainsacross environments in which different motivating opera-tions and reinforcement histories may influence behavior.

Finally, the messages selected and recorded on thedevice included specific messages in single words orshort phrases that could be understood easily by others.Specificity is important because specific communicationacts are more likely to be interpreted by parents or teach-ers (Keen, Woodyatt, & Sigafoos, 2002). The device alsooffered multiple specific choices. When only one choice

is programmed on a speech device, the individual usingthe device does not have the option to ask for somethingspecific. In this study, the participant was able to selectfrom 12 different messages in each setting. In addition to

basic options for requesting a break or access to a pre-ferred activity or item, each overlay included multiplespecific tangible activities within the setting (e.g., swingvs. slide on the playground).

Although the results of this study are promising, sev-eral limitations need to be considered. First, the partici-

pant in this study learned to use the SGD with very littleinstruction. However, individuals in this population

often require very extensive and systematic instruction.Although this study demonstrated the viability of such atreatment, it did not describe specific instructional strate-gies for teaching an individual to use a SGD with mul-tiple functions. Future research could focus on such aninstructional methodology. Second, only generalizationacross settings was demonstrated. In future studies, the

possibility of generalization across different communica-tive partners could be explored. Finally, data regardingwhich message was chosen and the corresponding func-tion of that message were collected only during follow-up. Therefore, it is possible that Brent may have selectedmessages related to one function during the initial gener-alization phases on the playground and in the gymna-sium. However, this concern is minor, because follow-updata demonstrate use across functions, and it is this typeof use that maintained over time. Regardless, additionalresearch examining how an individual uses a multi-function SGD during an FCT intervention is warranted.Ultimately, using a multifunction SGD to communi-cate with untrained people in the community and repair








communication breakdowns may eventually prove to bea viable way to increase the quality of life for individualswith severe disabilities.

References

Ahearn, W. H., Clark, K. M., MacDonald, R.P.F., & Chung, B.(2007). Assessing and treating vocal stereotypy in children withautism. Journal of Applied Behavior Analysis , 40 , 263–275.

Durand, V. M. (1990). Severe behavior problems: A functional com-munication training approach . New York: Guilford.

Durand, V. M. (1993). Functional communication training usingassistive devices: Effects on challenging behavior and affect.

Augmentative and Alternative Communication , 9, 168–176.Durand, V. M. (1999). Functional communication training using

assistive devices: Recruiting natural communities of reinforce-ment. Journal of Applied Behavior Analysis , 32 , 247–267.

Hanley, G. P., Iwata, B. A., & McCord, B. E. (2003). Functionalanalysis of problem behavior: A review. Journal of Applied

Beha vior Analysis , 36 , 147–185.Keen, D., Woodyatt, G., & Sigafoos, J. (2002). Verifying teacher

perceptions of the potential communicative acts of children withautism. Communication Disorders Quarterly , 23 , 133–142.

Kennedy, C. H. (2002). The maintenance of behavior change as anindicator of social validity. Behavior Modification , 26 , 594–604.

Kennedy, C., Meyer, C., Knowles, T., & Shukla, S. (2000). Analyzingthe multiple functions of stereotypical behavior for students withautism: Implications for assessment and treatment. Journal of

Applied Behavior Analysis , 33 , 559–571.Lang, R., O’Reilly, M., Machalicek, W., Lancioni, G., Rispoli, M., &

Chan, J. M. (2008). A preliminary comparison of functionalanalysis results when conducted in contrived versus naturalisticsettings. Journal of Applied Behavior Analysis , 41 , 135–139.

Lang, R., O’Reilly, M., Lancioni, G., Rispoli, M., Machalicek, W.,

Chan, J. M., et al. (in press). Discrepancy in functional analysisresults across two applied settings: Implications for interventiondesign. Journal of Applied Behavior Analysis.

Mace, F. C., Browder, D. M., & Lin, Y. (1987). Analysis of demandconditions associated with stereotypy. Journal of BehaviorTherapy and Experimental Psychiatry , 18 , 25–31.

Matson, J. L., Kiely, S. L., & Bamburg, J. W. (1997). The effect ofstereotypies on adaptive skills as assessed with the DASH-II andVineland Adaptive Behavior Scales. Research in Developmental

Disabilities , 18 , 471–476. Northup, J., Wacker, D. P., Berg, W. K., Kelly, L., Sasso, G., &

DeRaad, A. (1994). The treatment of severe behavior problems inschool settings using a technical assistance model. Journal of

Applied Behavior Analysis , 27 , 33–47.

Peck, S. M., Wacker, D. P., Berg, W. K., Cooper, L., Brown, K. A.,Richman, D., et al. (1996). Choice-making treatment of youngchildren’s severe behavior problems. Journal of Applied Behavior

Analysis , 29 , 263–290.Schloper, E., Reichler, R. J., Devellis, R. F., & Daly, K. (1980).

Toward an objective classification of childhood autism: ChildhoodAutism Rating Scale (CARS). Journal of Autism and Developmental

Disabilities , 10 , 91–103.Schlosser, R. W., & Lee, D. L. (2000). Promoting generalization and

maintenance in augmentative and alternative communication:

A meta-analysis of 20 years of effectiveness research. Augmentativeand Alternative Communication , 16 , 208–226.

Sparrow, S. S., Balla, D. A., & Cicchetti, D. V. (1984). Vineland Adaptive Behavior Scales: Interview edition survey form . CirclePines, MN: American Guidance Service.

Steege, M. W., Wacker, D. P., Cigrand, K. C., Berg, W. K., Novak, C. G.,Reimers, T. M., et al. (1990). Use of negative reinforcement in thetreatment of self-injurious behavior. Journal of Applied Behavior

Analysis , 23 , 459–467.Stokes, T. F., & Baer, D. M. (1977). An implicit technology of gener-

alization. Journal of Applied Behavior Analysis , 10 , 349–367.Winbon, L., Wacker, D. P., Richman, D. M., Asmus, J., & Geier, D.

(2002). Assessment of and selection for functional communica-tion training packages. Journal of Applied Behavior Analysis , 35 ,295–298.

Jessica H. Franco , PhD, BCBA, CCC-SLP, is a graduate ofthe University of Texas at Austin program in Autism andDevelopmental Disabilities. Her interests include the earlyidentification of autism, communication intervention, andtreatment of challenging behavior.

Russell L. Lang , PhD, BCBA, is currently a researcher at TheEli and Edythe L. Broad Asperger Research Center, Universityof California, Santa Barbara. His research interests include theassessment and treatment of challenging behavior, behavioranalysis, and the use of assistive technology with persons withdevelopmental disabilities.

Mark F. O’Reilly , PhD, BCBA, is an associate professor atThe University of Texas at Austin. He currently coordinatesthe graduate training programs in autism and developmentaldisability. He is interested in the design, implementation, and

evaluation of evidence-based and culturally sensitive servicesfor students with autism and related developmental and physi-cal disabilities and their families.

Jeffrey M. Chan , MA, BCBA, is a doctoral candidate in specialeducation at The University of Texas at Austin. His currentinterests include autism spectrum disorders, social stories,challenging behavior, and teacher training.

Jeff Sigafoos , PhD, BCBA, is a professor at Victoria Univer-sity of Wellington. His research is focused primarily onteaching communication skills to individuals with develop-

mental disabilities, including the use of augmentative com-munication devices, and approaches for using video-basedinstruction in vocational and residential programs for adultswith developmental disabilities.

Mandy Rispoli , PhD, BCBA, is a graduate of the Department ofSpecial Education at The University of Texas at Austin. Her cur-rent research interests include the assessment and treatment ofchallenging behavior in young children with severe disabilities.






146(Autism Vorbire)

Documents

Transcript of 146(Autism Vorbire)