Procedures for Teaching Appropriate Gestural Communication ... · Procedures for Teaching Appropriate Gestural Communication Skills to Children with Autism Dawn M. Buffington,1,2

Journal of Autism and Developmental Disorders, Vol. 28, No. 6, 1998

Procedures for Teaching Appropriate GesturalCommunication Skills to Children with Autism

Dawn M. Buffington,1,2 Patricia J. Krantz,1'2 Lynn E. McClannahan,1'2 andClaire L. Poulson2

Four children with autism were taught to use gestures in combination with oral communi-cation. Using a multiple-baseline across-responses design, intervention was introduced suc-cessively across three response categories containing gestures representative ofattention-directing/getting, affective, and descriptive behavior. Although none of the partici-pants displayed appropriate gestural and verbal responses during baseline, all participantsacquired this skill with the systematic implementation of modeling, prompting, and rein-forcement. Generalization measures indicated that the children learned to respond in thepresence of novel stimuli and a novel setting. Social validity measures revealed that theparticipants' behavior appeared more socially appropriate at the completion of the studythan at the start of the study, and that the participants' behavior was indistinguishable fromthat of their typically developing peers.

INTRODUCTION

The use of gestures is one mode of communi-cation that develops early in life and is associatedwith the development of language skills (Bates,Camaioni, & Volterra, 1975; Mundy, Sigman,Ungerer, & Sherman, 1987). Although the use ofgestural communication develops prior to the devel-opment of speech (Ricks & Wing, 1975), it is laterused in coordination with spoken language duringcommunicative interactions (Morford & Goldin-Meadow, 1992).

Although gestural communication developsearly and in sequence in typically developing indi-viduals, this is not true of individuals with autism.Even though individuals with autism have difficultyspeaking, rarely do they use gesture as an alternatemeans of communication (Loveland, Landry,Hughes, Hall, & McEvoy, 1988; Ricks & Wing, 1976;

1Princeton Child Development Institute, Princeton, New Jersey.2City University of New York, Graduate School and University

Center, New York, New York.

Wetherby & Prutting, 1984). Numerous studies havefound that individuals with autism differ from othersboth quantitatively and qualitatively in their use ofgesture. The number of gestures used by individualswith autism is significantly lower than that used byothers (Buitelaar, van Engeland, de Kogel, de Vries,& van Hooff, 1991; Loveland et al., 1988; Landry &Loveland, 1989; Mundy, Sigman, & Kasari, 1990; Sig-man, Mundy, Sherman, & Ungerer, 1986). More im-portantly, those gestures used by individuals withautism are typical of a lower level of development(Attwood, Frith, & Hermelin, 1988; Carr & Kemp,1989; Landry & Loveland, 1988; McHale, Simeons-son, Marcus, & Olley, 1980; Mundy, Sigman,Ungerer, & Sherman, 1986). Gesture is seldom usedas a stimulus to alter another person's orientation oreye contact, to describe objects, or to obtain socialconsequences. These differences are observed whenindividuals with autism are compared to normal con-trol subjects (Kasari, Sigman, Mundy, & Yirmiya,1990; Landry & Loveland, 1989; Landry et al., 1988;Mundy et al., 1986; Sigman, Mundy, Sherman, &

535

0162-3257/98/1200-0535$15.00 C © 1998 Plenum Publishing Corporation

KEY WORDS: Gestural communication; teaching procedure.

536 Buffington, Krantz, McClannahan, and Poulson

Table I. Participant Characteristics Including PPVT-R Score and PreschoolLanguage Scale Score

Name

AnneOscarKevinNick

CA

6-56-44-54-5

Months intreatment

48363

18

PPVT-Ra

SS

4347

<4280

AE

3,03,22,03,2

Preschool language

scale score (AE)

3,23,43,22,2

ass = standard sore; AE = age-equivalent score (years, months).

Ungerer, 1986; Wetherby & Prutting, 1984), mentallyretarded control subjects (Attwood et al,, 1988;Buitelaar et al., 1991; Kasari et al., 1990; Mundy etal., 1986, 1990; Sigman et al., 1986), and language-delayed control subjects (Landry & Loveland, 1988,1989; Landry et al., 1988; Loveland & Landry, 1986).

Because gestural behavior plays an importantrole in the establishment and maintenance of socialinteractions, difficulties in this area might contributeto the difficulties individuals with autism have duringsocial interactions (Garfin & Lord, 1986; Koegel &Frea, 1993). Because individuals with autism do notuse conventional gestures during social interactions,they (as well as their social partners) have difficultyin participating in a reciprocal social situation(Attwood et al., 1988; Buitelaar et al., 1991; Mundyet al., 1986; Prizant & Wetherby, 1987).

It is well known that individuals with autismhave difficulty in learning through observation andimitation (Smith & Bryson, 1994). In one study con-ducted to assess the imitation of gestures, Sigmanand Ungerer (1984) found that individuals withautism performed significantly more poorly than in-dividuals with mental retardation or individuals oftypical development on the Uzigiris-Hunt vocal andgestural imitation subtests. This deficit in imitationskills may contribute to the difficulties individualswith autism have in using gestural communication(Wetherby & Prutting, 1984). Additionally, individu-als with autism may not learn to use gestures thatlead to social consequences, because social conse-quences may not often serve as reinforcers for themin the same way that they do for individuals of typicaldevelopment.

Although extensive research has defined the ex-tent of the impairment in gestural communication,little research to date has demonstrated that gesturalcommunication can be taught to individuals withautism. Due to the paucity of research in the litera-

ture on the systematic teaching of gestural commu-nication to children with autism, the current studyfocused on teaching three gestural response catego-ries to these children. The three responses categorieswere chosen to contain gestures that would be ap-propriate for requesting items or directing the behav-ior of another, for obtaining social consequences, andfor describing characteristics of objects.

METHOD

Participants

Four children with autism participated in thisstudy. All of the children attended classes at thePrinceton Child Development Institute (PCDI) andhad previously received diagnoses of autism by inde-pendent agencies according to the criteria estab-lished in the DSM-III-R (American PsychiatricAssociation, 1987). The children who participated inthis study were between 4 and 6 years of age andhad been at PCDI from 3 months to 4 years. All ofthe children had some oral language. None of thechildren used gestures appropriately to request items,to obtain social consequences, or to describe charac-teristics of objects. See Table I for a complete de-scription of participant characteristics.

Setting and Apparatus

All sessions took place in a small classroom atPCDI. One desk was placed in the center of theroom, with two chairs next to it. Stimulus materialswere placed on the desk or on one of the items foundat the periphery of the room. To assess generalizationof trained responses to a new setting, pre- and

Gestural Communication 537

Table II Gestural and Verbal Responses with Examples of the Nonverbal and Verbal Stimuli Presented During Training andProbe Trials

Category

Attention

Affective

Reference

Gestural response

Point

Hand request

Raise hand

Arms up

Shake head

Hands on face

Tiny

Huge

Fast

Verbal response

Look!

Can I have that?

I do!

I won!

No way

Uh oh

It's tiny/tiny

It's huge/huge

It was fast/fast

Nonverbal stimulus"

PinwheelGlobe (P)BookSparkle gun (P)DinosaurXylophone (P)Pinball gameDart game (P)MustardBuzzer (P)Break crayonDrop puzzle (P)Mini bookMini gun (P)Large balloonLarge pencil (P)HelicopterExploder (P)

Verbal stimulusa

Let's talk about something on theFind something new on theLook at thisSee what I have (P)Who would like ?Anyone want ? (P)You're the winner!Wow, you did it! (P)Want this ?Do you want ?Look what happenedOh no! (P)Look at this littleCheck out this littleWhat do you think of this bigWow, this is a big (P)Tell me how that movedHow did move? (P)

a(p) indicates probe stimulus.

postintervention measures were taken in the chil-dren's regular classrooms at PCDI.

Response Definitions

The dependent measure consisted of two com-ponents, a gestural and verbal response (e.g., point-ing and saying "Look!"). The gestural and verbalresponses are presented in Table II. Responses werescored immediately following the first presentation ofthe stimuli during a session. The responses werescored by independent observers and were scoredcorrect only if they occurred within 5 seconds of thetherapist-presented stimuli that signaled the start ofeach interactive episode.

Stimulus Materials

Nonverbal Stimuli. A total of 144 nonverbal stim-uli were used. These stimuli were an assortment offood items, toys, common objects, and activities thatwere appropriate discriminative stimuli for each tar-get response. Examples of the nonverbal stimuli canbe found in Table II. These stimuli were arranged infour stimulus sets, each of which contained 36 items.Within a stimulus set, four stimuli were assigned toeach target response. Three of these items were des-ignated as training stimuli and one of these items was

designated as a probe stimulus. Training stimuli wereassociated with treatment, whereas the probe stimuliwere never associated with treatment. All stimuli fora specific target response were randomly assigned tostimulus sets, as well as to the training and probecategories.

Verbal Stimuli. A total of 36 verbal stimuli wereused. As shown in Table II, these statements wereappropriate discriminative stimuli for each target re-sponse. Four verbal discriminative stimuli were as-signed to each target response. Three of thesediscriminative stimuli were randomly assigned astraining discriminative stimuli and one was randomlyassigned as a probe discriminative stimulus.

Experimental Design

A multiple-baseline across-responses design wasused. Three response categories were trained. Thesecategories were attention-directing/getting responses,affective responses, and reference responses. Withineach response category, three gestural and verbal re-sponses were trained. These gestures were selectedafter observing 12 preschool and kindergarten-agechildren in their school classroom and at home. Inaddition, gestural responses were defined after re-questing preschool and kindergarten-age children toshow the therapist specific responses with their hands

538 Bufflngton, Krantz, McClannahan, and Poulson

(e.g., the therapist said "show me fast with yourhands").

Intervention was introduced successively acrossthe legs of the design after the mastery criterion wasmet. This criterion required that accurate gesturaland verbal responses remained at 89% or higher forfour consecutive training sessions. The order of in-troduction of treatment across response categorieswas varied for each child to control for order effects.

Procedure

General Format

At the start of each session, the child was seatedin front of a desk across from the therapist. Oncethe child was attending to the therapist, a nonverbaland verbal discriminative stimulus was presented. Atrial consisted of the therapist presenting a nonverbaland a verbal discriminative stimulus to initiate theinteractive episode and waiting for a maximum of 5seconds for a gestural and verbal response to occur.Within each session, there were 27 training and 9probe trials.

Baseline

During baseline sessions, if the correct verbaland gestural target response occurred within 5 sec-onds of the start of the interactive episode, the thera-pist provided verbal feedback. If an incorrectresponse or no response occurred within 5 secondsof the start of the interactive episode, the trial wasterminated. Token reinforcement was provided alongwith verbal praise for on-task behavior after approxi-mately every third trial.

Treatment Sessions

Two types of trials were presented during treat-ment sessions. Training trials were associated withtreatment and probe trials were not associated withtreatment. For training trials, if the correct responseoccurred within 5 seconds of the start of the inter-active episode, the therapist provided token rein-forcement and verbal praise to the child. If, however,an incorrect response occurred or no response oc-curred within 5 seconds of the start of the interactiveepisode, the therapist modeled the correct gestural

and verbal response. If the child did not imitate bothcomponents of the response, the therapist physicallyprompted him/her to emit the gestural response andverbally prompted him/her to emit the verbal re-sponse. Following this correction procedure, the trialwas presented again. If the child made the correctgestural and verbal response, token reinforcementwas provided along with verbal praise and the trialwas terminated. If, again, an incorrect response oc-curred, the correction procedure was introducedagain. This process continued until the child inde-pendently emitted the correct gestural and verbal re-sponse within 5 seconds of the start of the interactiveepisode.

Within-session generalization of the trained ges-tural and verbal responses from the training stimulito a novel set of stimuli was measured by the pres-entation of probe trials. During these trials, the non-verbal and verbal stimuli that were used to initiatethe interactive episode were never associated withthe teaching procedure or reinforcement. Duringprobe trials, the therapist initiated the interactiveepisode, waited for a response to occur, and termi-nated the trial by removing the stimulus materials.Reinforcement was never provided for correct re-sponses and the correction procedure was neverused.

Across Setting and Stimuli Generalization: RegularClassroom Probes

Generalization of the trained gestural and verbalresponses to a different setting and in the presenceof novel, verbal, and nonverbal discriminative stimulipresented by the therapist was assessed during theregular classroom probes. During these probes, thechildren followed their regularly scheduled activities.A novel set of stimuli was used during these probes.The set contained nine nonverbal and nine verbaldiscriminative stimuli that had never been associatedwith treatment. The therapist initiated an interactiveepisode by presenting the stimuli. No reinforcementor teaching was provided in the classroom setting.Three generalization probes were conducted in thechildren's classrooms during baseline conditions andthree were conducted in the children's classroomsimmediately after treatment was completed for thefinal response category.


Fig. 1. Percentage of training and probe trials in which Anne produced a correct verbal and gestural response.

Social Validity

Measure 1. To assess change in the students'communication skills from baseline to treatment, thefirst measure of social validity consisted of presenting72 videotaped interactive episodes to a group of 11graduate students. Two episodes for each target re-sponse were presented, one was a baseline episodeand one was a treatment episode. This was done foreach of the four children in the study. Each graduatestudent was asked to answer the following question:"In which of the two interactive episodes did thechild appear more expressive in his/her communica-tion?"

Measure 2. For the second social validity meas-ure, the same group of graduate students was pre-sented with 72 videotaped interactive episodes toassess whether the gestures used by the children inthis study were similar to those used by their age-matched peers. These interactive episodes containedthe four children in the study and four typically de-veloping children (matched on chronological age).

Each graduate student was asked to evaluate eachscene and answer the following question: "Was anappropriate gesture and verbalization used by thestudent?"

Interobserver Agreement

Two observers (the primary therapist and an-other teacher at PCDI) independently scored a mini-mum of 40% of the sessions in each of the baseline,treatment, and regular-classroom generalization con-ditions. Sessions for interobserver agreement (IOA)were randomly selected on a weekly basis from thetotal sessions for the week. Interobserver agreementwas obtained on the percentage of trials containingan appropriate gestural and verbal response. To en-sure the integrity of the independent variable, IOAwas calculated for the accurate presentation of thenonverbal stimuli, the verbal stimuli, the modelingprocedure, and the reinforcement contingencies. In-terobserver agreement was calculated on a point by


Fig. 2. Percentage of training and probe trials in which Oscar produced a correct verbal and gestural response.

point basis. The number of agreements was dividedby the number of agreements plus disagreements andmultiplied by 100 to obtain the percentage of IOA.

For the dependent measures, the mean percent-age of IOA was maintained at 94% or betterthroughout all conditions of the study. The percent-age of IOA on the correct delivery of the nonverbaland verbal stimuli, on the contingent use of the mod-eling procedure, and on the contingent delivery ofreinforcement was invariably 100% across all condi-tions and all children.

RESULTS

The individual results for each participant aregraphically presented in Figs. 1 through 4.

For each figure, the dashed vertical line repre-sents the point at which treatment was introducedfor each response category. Also in each figure, theclosed circles represent the data obtained duringtraining trials and the open circles represent the dataobtained during probe trials.

The percentage of training and probe trials inwhich Anne produced an appropriate gestural andverbal response is presented in Fig. 1 by session forall three response categories. During the baselineconditions for all three response categories, the per-centage of trials containing a correct gestural andverbal response was 0% throughout all sessions. Withthe successive introduction of treatment across thethree response categories, systematic increases incorrect responding in the presence of both the train-ing and probe stimuli were observed. The increasesin the percentage of trials containing an appropriategestural and verbal response were observed only inthe response category under treatment. Specifically,the percentage of training trials in which Anne pro-duced an appropriate attention-getting/directing ges-tural and verbal response increased from 0% duringbaseline to an average of 100% during the last foursessions of treatment. A similar increase was seen inthe probe trials for the attention category. That is,the percentage of probe trials in which Anne pro-duced an appropriate gestural and verbal responseincreased from 0% during baseline to an average of


Fig. 3. Percentage of training and probe trials in which Kevin produced a correct verbal and gestural response.

100% during the last four sessions of treatment. Forthe affect category, the percentage of training trialsin which Anne produced an appropriate gestural andverbal response increased from 0% during baselineto an average of 97% during the last four sessionsof treatment. Again, this increase in appropriate re-sponding was also observed during the probe trialswith the percentage of appropriate gestural and ver-bal responding increasing from 0% during baselineto an average of 100% during the last four sessionsof treatment. Finally, for the reference category, thepercentage of training trials in which Anne producedan appropriate gestural and verbal response in-creased from 0% during baseline to an average of100% during the last four sessions of treatment. Dur-ing probe trials in the reference category, the per-centage of trials in which Anne produced anappropriate gestural and verbal response increasedfrom 0% during baseline to an average of 92% dur-ing the last four sessions of treatment.

The above findings were replicated with threeadditional children. That is, for Oscar (Fig. 2), Kevin

(Fig. 3), and Nick (Fig. 4) low, stable rates of gesturaland verbal responding were observed in all three re-sponse categories during baseline. With the succes-sive introduction of treatment across each responsecategory, systematic increases in the percentage oftraining and probe trials containing an appropriategestural and verbal response were observed for eachstudent.

To assess generalization of the gestural and ver-bal responses from the stimuli presented duringtraining in the training setting to the presentation ofnovel stimuli in a novel setting, generalization probeswere conducted in the children's classrooms. Thepercentage of trials in which an accurate gestural andverbal response was observed during the regular-classroom generalization sessions was measured.During the pretreatment sessions, none of the stu-dents demonstrated accurate gestural and verbal re-sponding in any of the response categories. Followingthe teaching conducted in the study, the percentageof accurate gestural and verbal responding increasedin the presence of untrained stimuli in a different set-


Fig. 4. Percentage of training and probe trials in which Nick produced a correct verbal and gestural response.

ting. By the end of the third posttreatment generali-zation session, each student made at most one error.

To assess changes in gestural and verbal re-sponding, social validity Measure 1 was calculated.The mean percentage of treatment episodes thatwere rated as more expressive than the baseline epi-sodes for Anne was 98%, with a range of 89 to 100%.The mean percentage of treatment episodes thatwere rated as more expressive than the baseline epi-sodes for Oscar was 98% (range 89-100%). Themean percentage of treatment episodes that wererated as more expressive than the baseline episodesfor Kevin was 98% (range 89-100%). Finally, themean percentage of treatment episodes that wererated as more expressive than the baseline episodesfor Nick was 96% (range 89-100%).

To ascertain whether the gestural and verbalperformances of the children in this study were typi-cal of the gestural and verbal performances of chil-dren in their peer group, social validity Measure 2was collected. The mean percentage of Anne's vide-otaped episodes scored as containing an appropriate

gesture and verbalization was 91% (range 56-100%).The mean percentage of her age-matched peer's epi-sodes scored as containing an appropriate gestureand verbalization was 100%. An independent / testrevealed that there was no significant difference be-tween the number of videotaped episodes scored asappropriate between these two individuals, t(20) =-2.043, p > .025. The mean percentage of videotapedepisodes scored as containing an appropriate gestureand verbalization for Oscar was 93% (range 78-100%). The mean percentage of episodes scored ascontaining an appropriate gesture and verbalizationfor his age-matched peer was 96% (range 89-100%).There was no significant difference between thenumber of episodes rated as appropriate betweenthese two individuals, t(20) = -1.074, p > .025. Themean percentage of Kevin's episodes rated as con-taining an appropriate gesture and verbalization was100%. The mean percentage of episodes rated ascontaining an appropriate gesture and verbalizationwas 97% for his age-matched peer. There was no sig-nificant difference between the number of vide-


otaped segments rated as appropriate between thesetwo individuals, t(20) = 1.936, p > .025. Finally, themean percentage of videotaped episodes scored ascontaining an appropriate gesture and verbalizationfor Nick was 91% (range 67-100%). The mean per-centage of videotaped episodes scored as containingan appropriate gesture and verbalization for his age-matched peer was 89% (range 56-100%). There wasno significant difference between the number of epi-sodes rated as appropriate for these two individuals,t(20) = .363, p > .025.

DISCUSSION

The four children with autism who participatedin this study used little or no gestural communicationprior to the present intervention. With the introduc-tion of an intervention package that contained mod-eling, prompting, and reinforcement, all four of thechildren learned to use gestural and verbal responsesin the presence of nonverbal and verbal discrimina-tive stimuli presented by the therapist during an in-teractive episode.

Fluently combined gestural and verbal responsesoccurred not only in the presence of the trainingstimuli but also in the presence of the probe stimuli.The occurrence of responding in the presence ofprobe stimuli suggests that the gestural and verbalbehavior of the children was not only under the con-trol of the training discriminative stimuli but also un-der the control of other stimuli that resembled thosestimuli (Bijou & Baer, 1961; Catania, 1992).

Generalization of the gestural and verbal re-sponses from the training setting to a novel settingwas also assessed in the current experiment. Thefindings suggest that generalization of the trained re-sponses did occur to some extent in the presence ofnovel stimuli presented in a novel setting. By the fi-nal generalization probe in the regular classroom,only one error occurred in one response category foreach child. These results are consistent with the find-ings in the training setting. That is, the greatest levelof generalization of the trained responses from thediscriminative stimuli presented by the therapist inthe training setting to novel discriminative stimulipresented in a different setting occurred for those re-sponses that the children had the least difficulty ac-quiring.

The social validity measures taken in this studysuggest that the participants' behavior appeared

more socially appropriate after treatment. Graduatestudents consistently rated posttreatment videoscenes as more socially appropriate than baselinevideo scenes. In addition, the ratings obtained fromcomparison of typical children to the participants inthis study suggest that the children with autism wereindistinguishable from their peers in their use of ap-propriate gestures and verbalizations in the scenariospresented.

The responses taught in this study were repre-sentative of various levels of gestural competence(Barten, 1979). That is, the gestures that the childrenlearned could be used to obtain tangible conse-quences in a socially appropriate way, to obtain socialconsequences, to display affective behavior, and todescribe characteristics of objects.

Through the gestures in the attention category,the children learned to request items in a more ap-propriate manner and alter the orienting behavior ofothers in their environment (Kasari et al., 1990;Loveland & Landry, 1986; Mundy et al., 1990). Byteaching children to use requesting gestures (e.g.,"hand request") that are more socially acceptablethan those gestures often used by individuals withautism (e.g., grabbing), it increases the probability oftheir gaining access to reinforcers in their environ-ment (Carr & Kemp, 1989; Prizant & Wetherby,1987). In addition, by teaching children with autismto point and say "look" rather than to only label anobject, they are learning a necessary joint attentionskill (Pierce & Schreibman, 1988).

The gestures that the children learned in the af-fective category provided the children with behaviorthat could be used to recruit social responses fromothers in their environment. Because children withautism have deficits in social interaction skills, teach-ing them to use gestures to recruit social responsesfrom others is a useful way to increase the probabilityof successful social interactions between themselvesand others in their environments (Loveland et al.,1988; McHale et al., 1980; Mundy et al., 1986; Ricks& Wing, 1975; Sigman et al., 1986; Wetherby & Prut-ting, 1984; Wetherby, 1986). These gestures wereconsidered to be affective responses because they areso often associated with the description of affectivestates, another large deficit area for children withautism (Attwood et al, 1988; Kasari et al, 1990).

Finally, the gestures that the children learned inthe reference category provided them with responsesto describe objects. The majority of the vocal andgestural behavior of children with autism is used to


label and request items (Landry & Loveland, 1988;Partington, Sundberg, Newhouse, & Spengler, 1994;Shah & Wing, 1986; Wetherby & Prutting, 1984).These responses, however, enable children withautism to increase the complexity of their communi-cation and, thereby, strengthen their overall commu-nication skills.

Vocal and gestural behavior are two importantcontributors to social behavior (Bijou & Baer, 1965;Garfin & Lord, 1986; Prizant & Wetherby, 1987). Byteaching use of gesture, alone or in combination withvocal behavior, children with autism acquire anotherskill that is a needed component of social interaction.More important, teaching gesture in the presence ofvarying nonverbal and verbal discriminative stimulishould help to provide the context in which gesturalcommunication should be used (Shah & Wing, 1986).With the addition of gestural behavior under the con-trol of appropriate social stimuli to the communica-tion repertoires of individuals with autism, hopefullythese individuals will appear more "typical" in theirsocial interactions and in turn become more reinforc-ing to their social partners (Koegel & Frea, 1993).

In the current study, children with autism wereable to expand their communicative repertoires bylearning gestures and verbalizations that could beused under a variety of conditions in which theteacher presented nonverbal and verbal discrimina-tive stimuli. Although these discriminative stimuliwere chosen to closely represent those that would oc-cur in a child's natural setting, future research effortsshould investigate the spontaneous use of gesturesand verbalizations (learned during discrete trialteaching) in the child's natural environment. Such ef-forts should lead to the development of more "natu-ral" teaching paradigms (e.g., incidental teaching) tofacilitate generalization of the gestural skills fromtraining situations to novel situations. It would alsobe interesting to determine whether early interven-tion efforts that include gestural communicationtraining would ultimately produce generalized imita-tion of gesture.

REFERENCES

American Psychiatric Association. (1987). Diagnostic and statisticalmanual of mental disorders (3rd ed.). Washington, DC:Author.

Attwood, A., Frith, U., & Hermelin, B. (1988). The understandingand use of interpersonal gestures by autistic and down's syn-drome children. Journal of Autism and Developmental Disor-ders, 18, 241-257.

Barten, S. S. (1979). Development of gesture. In N. R. Smith &M. B. Franklin (Eds.), Symbolic functioning in childhood (pp.139-151). Hillsdale, NJ: Erlbaum.

Bates, E., Camaioni, L., & Volterra, V (1975). The acquisition ofperformatives prior to speech. Merrill-Palmer Quarterly, 21,205-226.

Bijou, S. W. & Baer, D. M. (1961). Child development I: A system-atic and empirical theory. New York: Appleton-Century-Crofts.

Bijou, S. W, & Baer, D. M. (1965). Child development II: Universalstage of infancy. New York: Appleton-Century-Crofts.

Buitelaar, J. K., van Engeland, H., de Kogel, K. H., de Vries, H.,& van Hooff, J. A. (1991). Differences in the structure ofsocial behavior of autistic children and non-autistic retardedcontrols. Journal of Child Psychology and Child Psychiatry, 32,995-1015.

Carr, E. G., & Kemp, D. C. (1989). Functional equivalence ofautistic leading and communicative pointing: Analysis andtreatment. Journal of Autism and Developmental Disorders, 19,561-578.

Catania, H. C. (1992) Learning (3rd ed.). Englewood Cliffs, N.J.:Prentice-Hall.

Garfin, D. G., & Lord, C. (1986). Communication as a social prob-lem in autism. In E. Schopler & G. B. Mesibov (Eds.), Socialbehavior in autism (pp. 133-149). New York: Plenum Press.

Kasari, C., Sigman, M., Mundy, R, & Yirmiya, N. (1990). Affectivesharing in the context of joint attention interactions of nor-mal, autistic, and mentally retarded children. Journal ofAutism and Developmental Disorders, 20, 87-100.

Koegel, R. L, & Frea, W D. (1993). Treatment of social behaviorin autism through the modification of pivotal social skills.Journal of Applied Behavior Analysis, 26, 369-377.

Landry, S. H., & Loveland, K. A. (1988). Communication behav-iors in autism and developmental language delay. Journal ofChild Psychology and Psychiatry and Allied Disciplines, 29, 621-634.

Landry, S. H., & Loveland, K. A. (1989). The effect of social con-text on the functional communication skills of autistic chil-dren. Journal of Autism and Developmental Disorders, 19,283-299.

Loveland, K. A., & Landry S. H. (1986). Joint attention and lan-guage in autism and developmental language delay. Journalof Autism and Developmental Disorders, 16, 335-349.

Loveland, K. A., Landry, S. H., Hughes, S. O., Hall, S. K., &McEvoy, R. E. (1988). Speech acts and the pragmatic deficitsof autism. Journal of Speech and Hearing Research, 31, 593-604.

McHale, S. M., Simeonsson, R. J., Marcus, L. M., & Olley, J. G.(1980). The social and symbolic quality of autistic children'scommunication. Journal of Autism and Developmental Disor-ders, 10, 299-310.

Morford, M., & Goldin-Meadow (1992). Comprehension and pro-duction of gesture in combination with speech in one-wordspeakers. Journal of Child Language, 19, 559-580.

Mundy, R, Sigman, M., & Kasari, C. (1990). A longitudinal studyof joint attention and language development in autistic chil-dren. Journal of Autism and Developmental Disorders, 20, 115-128.

Mundy, P., Sigman, M., Ungerer, J., & Sherman, T (1986). Defin-ing the social deficits of autism: The contribution of non-ver-bal communication measures. Journal of Child Psychology andPsychiatry, 27, 657-669.

Mundy, R, Sigman, M., Ungerer, J., & Sherman, T (1987). Non-verbal communication and play correlates of language devel-opment in autistic children. Journal of Autism andDevelopmental Disorders, 17, 349-364.

Partington, J. W, Sundberg, M. L., Newhouse, L., & Spengler, S.M. (1994). Overcoming an autistic child's failure to acquire


a tact repertoire. Journal of Applied Behavior Analysis, 27,733-734.

Pierce, K., & Schreibman, L. (1995). Increasing complex social be-haviors in children with autism: Effects of peer-implementedpivotal response training. Journal of Applied Behavior Analysis,28, 285-295.

Prizant, B. M., & Wetherby, A. M. (1987). Communicative intent:A framework for understanding social-communicative behav-ior in autism. Journal of the American Academy of Child andAdolescent Psychiatry, 26, 472-479.

Ricks, D. M., & Wing, L. (1975). Language, communication, andthe use of symbols in normal and autistic children. Journal ofAutism and Childhood Schizophrenia, 5, 191-221.

Ricks, D. M., & Wing, L. (1976). Language, communication andthe use of symbols. In L. Wing (Ed.), Early childhood autism(pp. 93-134). New York: Pergamon.

Shah, A., & Wing, L. (1986). Cognitive impairments affecting so-cial behavior in autism. In E. Schopler & G. B. Mesibov

(Eds.), Social behavior in autism (pp. 153-167). New York: Ple-num Press.

Sigman, M., Mundy, P., Sherman, T., & Ungerer, J. (1986). Socialinteractions of autistic, mentally retarded and normal childrenand their caregivers. Journal of Child Psychology and Psychiatryand Allied Disciplines, 27, 647-656.

Sigman, M., & Ungerer, J. A. (1984). Cognitive and language skillsin autistic, mentally retarded, and normal children. Develop-mental Psychology, 20, 293-302.

Smith, I. M., & Bryson, S. E. (1994). Imitation and action inautism: A critical review. Psychological Bulletin, 116, 259-273.

Wetherby, A. M. (1986). Ontogeny of communicative functions inautism. Journal of Autism and Developmental Disorders, 16,295-315.

Wetherby, A. M., & Prutting, C. A. (1984). Profiles of communi-cative and cognitive-social abilities in autistic children. Journalof Speech and Hearing Research, 27, 364-377.

Zimmerman, I. L, Steiner, V.G., & Pond, R. E., (1979). PreschoolLanguage Scale. San Antonio, TX: The Psychological Corp.

Procedures for Teaching Appropriate Gestural Communication ... · Procedures for Teaching Appropriate Gestural Communication Skills to Children with Autism Dawn M. Buffington,1,2

Documents