JOURNAL OF APPLIED BEHAVIOR ANALYSIS

NATURALISTIC TREATMENT OF AN AUTISTIC CHILD'VEY MICHAEL NORDQUIST AND ROBERT G. WAHLER

UNIVERSITY OF TENNESSEE

The present research experimentally evaluated a "naturalistic" treatment program foran autistic child administered by the parents over a 2-yr period. Operant reinforcementtechniques previously developed and tested in laboratory settings were initially assessedin a clinic and eventually in the family's home. Experimental manipulations were per-formed in both settings on rituals, crying and whining, compliance, non-verbal imita-tion, and verbal imitation. The results clearly indicated that parents can effectively treatautistic behaviors provided that they receive adequate training and supervision in operantreinforcement therapy, and provided that sufficiently potent reinforcers are available tomaintain behavior.

Operant reinforcement methods applied inlaboratory settings have been remarkably effec-tive in modifying a wide variety of "autistic" be-haviors. For example, self-destructive behaviors(Tate and Baroff, 1966; Simmons and Lovaas,1969; Lovaas, Freitag, Gold, and Kassorla,1965); tantrums (Lovaas, Schaeffer, and Sim-mons, 1965), avoidance and non-attentive be-haviors (Schell, Stark, and Giddon, 1967; Mc-Connell, 1967; Simmons and Lovaas, 1969;Lovaas, Schaeffer, and Simmons, 1965), ritualsand stereotyped behaviors (Lovaas, et al., 1965)and self-stimulatory behaviors (Lovaas, Litrow-nik, and Mann, 1971) have been successfullytreated in laboratory settings. Reinforcementtherapy has also been employed in laboratorysettings to produce appropriate social, imitative,and language skills (e.g., Lovaas, et al., 1965;Lovaas, Freitag, and Whalen, 1967; Lovaas,Freitag, Kinder, Rubenstein, Schaeffer, andSimmons, 1966; Simmons and Lovaas, 1969;Metz, 1965; Schell, Stark, and Giddon, 1965;Nelson and Evans, 1968; Hingtgen and Church-ill, 1971; Hingtgen, Coulter, and Churchill,1967; Hingtgen, Sanders, and DeMyer, 1965;Hingtgen and Trost, 1966).

Although the research cited above is im-

'Reprints may be obtained from either author, De-partment of Child Development, University of Ten-nessee, Knoxville, Tennessee 37916.

pressive, applications are limited by one ormore of the following conditions: (1) focus hasbeen on behaviors produced in contrived set-tings; (2) unnatural stimuli such as shock,buzzers, lights, and specially built apparatusnot readily utilized in non-laboratory settings, orconsummatory stimuli tied to biological cycles(e.g., food and liquids) have been employed tomanipulate autistic behaviors; (3) treatment hasbeen administered by a limited number of highlytrained professional personnel, and (4) accuraterecords of autistic behavior outside of thelaboratory setting have not been kept.

Unfortunately, those few studies that haveexamined reinforcement techniques in non-laboratory settings have not provided experi-mental evidence documenting the success oftheir treatment programs (e.g., Brown, Pace, andBecker, 1969; Hewett, 1965; Wetzel, Baker,Roney, and Martin, 1966; Wolf, Risley, andMees, 1964). Although Risley and Wolf (1966)did provide data that suggest that parents can betrained effectively to modify some autistic be-haviors, their conclusions are based on informa-tion obtained in a laboratory setting. Only Risley(1968) reported experimental evidence thatindicates that parents can successfully apply re-inforcement methods to autistic behaviors innatural settings. He trained the mother of a6-yr-old autistic girl to suppress dangerous

79

1973, 6., 79-87 NUMBER 1 (SPRING 1973)

VEY MICHAEL NORDQUIST and ROBERT G. WAHLER

climbing behaviors, opening the refrigerator,pulling clothes off of closet hangers, and throw-ing pots and pans out of the kitchen cupboards.The mother also attempted to teach her childnon-verbal imitative skills, but the programwas terminated before Risley (1968) couldadequately evaluate this phase of the research.

Although limited in scope, Risley's (1968)results suggested that parents of autistic childrencan be trained to administer reinforcementtherapy in a home setting. If parents could infact be trained to modify a wide variety ofautistic behaviors, a "home based" treatmentapproach would have several distinct advantagesover a laboratory (or clinic) approach: (1) costto parents would be appreciably reduced becausethe parents, not professionals, would administerthe treatment package; (2) treatment conductedin the home would have to rely on reinforcersindigenous to that setting and consequently,(3) concern for generalization from the labora-tory (or clinic) to the home would not benecessary. This last advantage cannot be takenlightly because others have either suggested(e.g., Breger, 1965) or demonstrated (e.g., Ris-ley, 1968) that successful treatment in a labora-tory setting is no guarantee that treatment willtransfer to the child's natural environment.

Thus, the primary objective of the presentresearch was to assess the therapeutic impact ofa "naturalistic" treatment program administeredby the parents and designed to modify a widevariety of autistic behaviors. Because laboratoryevidence (cited above) had consistently indi-cated that several months were required toadminister such a program, another objective ofthis research was to assess the impact of thetreatment over an extended time period.

METHOD

Subject and SettingsJoey was referred by his parents to the Psycho-

logical Clinic, University of Tennessee, at theage of four. During the preceding 2-yr periodhe had been seen by a number of professionals

and was variously diagnosed as mentally re-tarded, schizophrenic, brain damaged, aphasic,and autistic. The parents reported that theywere usually advised to place the boy in aninstitution.When first seen by the second author, Joey

exhibited a number of deviant behaviors andbehavioral deficits characteristically associatedwith childhood autism. Language skills wereabsent; the parents reported that the boy hadnever spoken words. Social skills were also ab-sent, eye contact was very poor and close physi-cal contact with other people, including theparents, usually elicited struggling, whining,and sometimes intense tantrum behavior. Mostof Joey's behavior could have been classified asritualistic, self-stimulatory, or self-destructive.He was often observed playing with one toy,"sighting" on his fingers as one might sight ona gun barrel, or pouring sand, dirt, or sometimesa liquid on his head. Arm biting was also a prob-lem and occasionally he would bang his headagainst a wall or a chair. Screaming and a loud,high-pitched sound, something like a factorywhistle, occurred quite often. However, theparents were most concerned about attacks di-rected toward a younger, 2-yr-old sister. Al-though the attacks had just begun to occur atthe time of his referral and were more vocalthan physical, the parents expressed the fearthat Joey would seriously harm his sister.

Because of the need for immediate inter-vention, efforts to obtain baseline estimates ofthe boy's behavior were not instituted. Instead,the second author described to the parents atreatment program based on some of the labora-tory research cited earlier and expressed aninterest in training them to administer the pro-gram in their own home. They were informedof the experimental nature of the research andalso told that their home would eventually be-come a laboratory base for conducting the re-search. That is, the parents were told that theywould be trained to administer essentially thesame therapeutic techniques that had beensuccessfully administered by professional per-

80

NATURALISTIC TREATMENT OF AN AUTISTIC CHILD

sonnel in laboratory settings. They were alsotold that several sessions would be required to

teach them these skills and because of this, theywould be asked to make 10 hourly visits to theclinic for the first two months. After two

months, visits would be scheduled on a weeklybasis whenever possible. It was also explainedto them that both their behavior and their child'sbehavior would be monitored closely by ob-servers and that it would be necessary at timesto discontinue treatment briefly for the purpose

of evaluation.

Behavior Classification,Observers, and Reliability

Records of Joey's behavior were obtainedthrough a behavior checklist similar to thatdescribed by Hawkins, Peterson, Schweid, andBijou (1966). The method required an ob-server to make coded checks for the occurrence

of a behavior class within successive 10-secintervals over a 30-min observation period; any

occurrence of a class, regardless of its duration,was scored as a single unit. Thus, no more than180 units could be scored for a behavior classduring an observation period.

Four observers were used in the research.Each observer was sophisticated in the use ofoperant principles and natural science methodsof observation. Only one observer was aware ofthe experimental nature of the research. Thisobserver collected approximately one third ofthe data. Reliability checks were made at peri-odic intervals on 33% of the data.

Several of the initial clinic visits were usedto adapt the boy and his parents to the ob-server's presence and to obtain written recordsof parent-child interactions. During these ses-

sions, the parents and the child were observed ina playroom that contained several toys, a sandbox, a chalk board, punching bag, and a smalltable with chairs. The parents were told to playwith the boy and periodically request that hedo something for them. Occasionally, theserequests were intended to elicit imitative re-

sponses. For example, one of the parents would

take a peg board and mallet, demonstrate itsuse to Joey, then give the toy to him and say:"Here. Now you try it. Do it like mommy".

After the fifth session, the written recordswere analyzed in the following way: Joey'sbehavior was grouped into classes on the basisof physical or functional similarities among theseparate responses. For example, screaming, cry-ing, and whining were considered members ofthe same response class because of their similarphysical characteristics. Compliant responses,although physically dissimilar, were consideredtogether because of their relationship to com-mon stimulus events, namely, parental com-mands. Similarly, imitative responses were alsoconsidered members of the same response classbecause of their relationship to parent modelcues. Thus, five separate child response cate-gories were identified and defined as follows:Rituals-"sighting" on fingers, flapping hands,twirling objects or unusual gestures. Crying andWhining-crying, whining, screaming, and de-structive behavior such as face slapping, biting,or hitting that sometimes occurred along withcrying and whining. Compliance-following aparental command within the interval the com-mand was delivered or within the interval im-mediately following the command. Non-VerbalImitation-imitation of parental non-verbal be-havior in the interval the parental cue was de-livered or in the interval immediately followingthe presentation of the non-verbal cue. VerbalImitation-the same scoring criteria were ap-plied to this category that were applied to thenon-verbal imitative category.

General ProcedureTreatment: Phase I. The parents were first

instructed to use procedures intended to reducethe frequency of rituals, crying, and disruptivebehaviors and increase the frequency of imitativebehaviors. A timeout procedure was used topunish inappropriate behaviors. Whenever thesebehaviors occurred in the home setting, theparents were asked to remove the child im-mediately to his room and leave him there for

81

VEY MICHAEL NORDQUIST and ROBERT G. WAHLER

at least 10 min. If a tantrum occurred while hewas in his room, he was not released until thetantrum had subsided for 2 or 3 min. Timeoutwas used to punish inappropriate behaviors fortwo reasons: first, earlier research (e.g., Wolf,Risley, and Mees, 1964) indicated it might suc-

cessfully suppress the frequency of these be-haviors. Second, Wahler (1969) demonstratedthat the timeout procedure is sometimes associ-ated with increases in parental reinforcementvalue. Therefore, there was reason to believethat reductions in inappropriate behaviors mightbe associated with an increase in the reinforce-ment value of the parents. Consequently, theparents were also instructed to attend to com-

pliant behavior as often as possible.To teach Joey how to produce verbal re-

sponses, an imitative training procedure2 was

implemented, similar to those described byLovaas et al., (1967), Nelson and Evans (1968),and Risley and Wolf (1967). Because Joey hadno verbal imitative repertoire to speak of, it wasnecessary first to teach him to imitate simplenon-verbal tasks. A number of researchers (citedabove) have found that the acquisition of verbalimitative skills is facilitated when a child is firsttaught to imitate non-verbal behaviors. Usually,

2The imitation training procedures were directedto the establishment of non-verbal imitative skills.Non-verbal behaviors (see Footnote 3) were selectedthat were known to be within the boy's behavioralrepertoire. After instructing Joey to "Watch me. Dothis," the mother (or father) would present the cue,

then give the materials to the boy and say "Ok. Nowyou do it." During the first several sessions, any

gross approximation to the parents behavior was rein-forced with social approval. Gradually, the parentswere instructed to withhold approval until betterapproximations were produced. This "shaping" pro-cedure was consistently used for all new modelstimuli throughout the study. If no behavior occurredafter a cue presentation, the parents were instructedto take the boy physically through the imitation andthen reinforce. Gradually, the prompts were fadeduntil Joey produced the response on his own. To test

for the effect of non-verbal training on the acquisitionof verbal imitative skills, verbal cues were alternatedwith non-verbal cues. Non-verbal trials were dis-continued after Session 45 because they no longerappeared to facilitate the acquisition of verbal imita-tion (see Figure 1).

several non-verbal behaviors can be identifiedand presented to the child by an adult model. Ifthe child does not imitate the model, prompting,shaping, and fading procedures (see Risley andWolf, 1967) may be used to elicit, shape, andmaintain the behavior. Once the child haslearned reliably to imitate several non-verbaltasks, he is usually ready to begin verbal imita-tion training. Thus, the purpose of using non-verbal imitation training in the present researchwas to facilitate the acquisition of verbal imita-tion skills.

Assessments of Phase I began after thehabituation and behavior classification sessions.Assessments were made only in the clinic during30-min observation periods, although the par-ents did implement the same training programin the home on the days they did not visit theclinic. In order to obtain continuous records ofJoey's behavior, timeout was not administeredduring the recording sessions. Instead, theparents were asked to begin presenting non-verbal and verbal imitative cues to the boy atregular intervals and attempt to control inap-propriate behavior by delivering firm com-mands. Each session proceeded in the follow-ing manner: the parents, an observer(s), andJoey entered the laboratory room describedabove; all of the materials remained in theroom. Joey was seated by his parents at a smalltable and one of the parents (usually themother) sat opposite the boy and began to

administer the imitative cues.3 Non-verbal and

3Both non-verbal and verbal cues were selected onthe basis of familiarity and probability that the boycould produce the behavior. Thus, the non-verbalcues consisted of lifting a cup to the mouth, using afork and plate to eat "imaginative food", poundinga peg board, and stirring a spoon in a cup. Theverbal cues (also presented visually on cards) initiallyconsisted of sounds associated with the letters a, e, i,o, u, b, p, t, k, d, and s. By Session 20, it was apparentthat Joey could imitate all of these sounds fairly well.Therefore, very short one or two syllable words suchas cat, dog, eat, me, top, baby, etc., were introduced tothe boy. All of the words used were selected essen-tially on the basis of length, difficulty, and theirassociation to common objects or events found in the

82

NATURALISTIC TREATMENT OF AN AUTISTIC CHILD

verbal cues were alternated so that the sameper cent of verbal and non-verbal cues werepresented each session. Initially, any responsewas reinforced with social approval and the boywas allowed about 1 min to play in the roomafter each trial. If no response was observed (onnon-verbal trials only), the parents were in-structed to prompt the response and slowly fadethe prompt until Joey produced the responsewithout help from the parents. Gradually, thenumber of trials preceding play were increaseduntil a trials:play ratio of 3:1 was attained.Access to play was discontinued after Session 54.As later data will show, the per cent of ritualsand crying and whining observed across all ses-sions of the Phase I treatment period graduallydeclined. This change was accompanied by in-creases in the per cent of compliance, non-verbalimitation, and verbal imitation. Thus, a reversalprocedure was briefly implemented to test forcause-effect relationships. During this period,the parents were instructed to discontinuetimeout and social reinforcement for compliantbehaviors. Once causality was demonstrated, theparents were asked to reinstate the treatmentprocedures.

Treatment: Phase II. The treatment pro-cedures used during Phase I were only partially

boy's home. Thus, the words food, water, bathroom,outside, door, car, etc. were frequently presented be-cause it was intended that the parents begin elicitingthese words "naturally" as soon as the boy had learnedto produce them. By Session 81, Joey had learned toimitate virtually any word presented to him. There-fore, the sentences "I want the .", "This is a

.", and "Let's go to (the) ." were intro-duced. Each word was presented separately at first,but reinforcement was not delivered until the boy hadproduced the last word of the sentence and pointedto the object, until the parent handed him the ob-ject requested, or until the parent had given him apicture of the place he was to go. Gradually, theparent faded the verbal prompts until the procedurewent as follows: the parent, holding up an object (orpicture of a place) would say "Joey, what is this?(What do you want: Where do you want to go?)".Joey, after producing the correct response, wouldthen write the response on paper, repeating verballyeach word as he wrote it. After 1 min of "free-time"to write, a new stimulus was then presented.

successful in increasing the per cent of com-pliant and imitative behavior and decreasingrituals and crying and whining. The variabilityin the boy's performance from one session to thenext suggested that parental attention, althoughreinforcing, was not a potent enough reinforcerto produce behavioral stability. Consequently, asearch of the boy's home environment wasundertaken in an effort to identify a reinforcerpowerful enough to maintain the treatmentprogram. From home observations and theparents' reports, it was learned that Joey spenta great deal of time watching television; com-mercials, soap operas, and cartoons were viewedwith equal interest. Thus, the treatment pro-gram was modified so that imitation of parentalcues was reinforced by a token that Joey couldimmediately insert in a slot and earn 15 sec oftelevision viewing. At the end of this phase, thefamily and research team temporarily discon-tinued observations and terminated the clinicvisits to allow for a well-earned vacation break.Although on vacation, the parents continued touse the timeout and differential attention pro-cedures. Because the parents had become quiteskilled in the administration of treatment, planswere formulated for implementing observationand training sessions in the family's home.

Treatment: Phase III. It did not take long todetermine that television was not going to pro-duce response stability. Crying and whining,compliance, and verbal imitation per cents con-tinued to fluctuate markedly from session tosession. Thus, after briefly implementing asecond reversal period, the parents once againinstituted imitation training and continued topunish inappropriate behaviors with timeout.However, appropriate imitative behaviors werenow reinforced by short writing periods. Inaddition to watching television, it was learnedthat Joey spent a considerable amount of timedrawing pictures and copying letters, words, andeven sentences. Thus, the training procedure waschanged so that a correct verbal imitation wasfollowed by a brief period when Joey wasgiven paper and pencil and asked to copy a

83

VEY MICHAEL NORDQUIST and ROBERT G. WAHLER

written representation of the word (or sentence)he had just verbalized.

Joey was very skilled with paper and pencil,usually reproducing the written cue in a fewseconds. It soon became apparent that less than10% of the training sessions were actually spentwriting; more than 90% of the time Joey wasrequired to produce imitative behaviors. It is notsurprising, therefore, that his performance con-tinued to fluctuate across sessions. To deal withthis problem, the parents were instructed (Ses-sion 79) to increase the "free-writing" periodsto a full 1 min. After Joey had reproduced thewritten stimulus, he was free to write or drawwhatever he wished. A brief reversal of the finalprocedure was put into effect toward the end ofthe phase.

RESULTS

ReliabilityInter-observer agreement per cents were com-

puted separately for each response class by add-ing the total number of interval agreements tothe total number of interval disagreements anddividing the sum into total interval agreements.

The mean agreement per cents for each responseclass were as follows: Rituals, 0.87 (range: 0.78-1.00); Crying and Whining, 0.80 (range: 0.71-0.94); Compliance, 0.81 (range: 0.76-1.00);Non-Verbal Imitation, 0.96 (range: 0.82-1.00);Verbal Imitation, 0.87 (range: 0.69-1.00). Nodifferences were observed between per centsbased on clinic sessions and per cents based onhome sessions.

TreatmentFigure 1 presents changes in the per cent of

rituals, crying and whining, compliance, verbalimitation, and non-verbal imitation over a 2-yrtreatment period. One is immediately impressedby the marked variability in Joey's behavior.However, in spite of the variability, the percents of ritualistic behavior and non-verbal imi-tation eventually stabilized. That is to say, bySession 46, non-verbal imitation clearly stabil-ized near 100%; by Session 64, the per cent ofritualistic behavior decreased to zero. Thus, datafor both categories terminated at these points.

Although rituals and non-verbal imitationeventually stabilized at acceptable treatmentlevels, crying and whining, compliance and

PHASE I PHASE xsoe" R laamp9S1a.U-tSIt RubeM11 "s_* AdiTVR*Mf1ewm_I

PHASE x

50 l V V \J1Vs1 1 111ISessions I 5 l 0 0o 20 30 40 45 4 0 55 s0 5 7 o 0to 90o 90

J.E'69 ' JULY AUGUST k N|M(ER I0ECA9' JANer 900 rEA IMAM. I L | MAY AUG. SETC I4-aI| PE.lUFE N MN MY71

Fig. 1. Per cent of Rituals, Crying and Whining, Compliance, Non-Verbal Imitation and Verbal Imitationrecorded during 30-min treatment sessions conducted initially in a clinic (Sessions 1 to 65) and later in thehome (Sessions 66 to 90). Reinforcement contingencies are temporarily discontinued during the Test Probesessions. Changes in verbal model cues are also indicated; i.e., letters to words (Session 20) and words to sen-tences (Session 82).

84

NATURALISTIC TREATMENT OF AN AUTISTIC CHILD

verbal imitation per cents varied considerablyuntil Session 79. It will be recalled that theparents instituted a full 1-min period of "free-writing" reinforcement during this session.Thereafter, Figure 1 depicts sharp increases inthe per cents of verbal imitation and compliance,as well as concomitant decrease in the per centof crying and whining. These per cents werestable over the remaining treatment sessions.

The data in Figure 1 clearly suggest that thevariability in Joey's performance was predomi-nantly a function of weak reinforcers. Most ofthe boy's crying and whining was due to the factthat he did not want to "work" very hard forparental approval or 15-sec of television view-ing. Crying and whining and compliance percents decreased and increased accordingly onlywhen a powerful reinforcer was made contin-gent on verbal imitation (Session 79). Ritualisticbehavior did not appear to be associated with(produced by) the treatment procedure. Thus,rituals were eventually suppressed. Non-verbalimitative behaviors are not incompatible withcrying and whining. Joey could easily imitatenon-verbal cues and still cry or whine, as he oftendid. However, verbal imitation is incompatiblewith crying and whining. Thus, it is not sur-prising to see such a close correspondence be-tween an increase in verbal imitation on the onehand and a decrease in crying and whining onthe other, especially when 1-min free-writingperiods were introduced at Session 79.

Figure 1 provides additional sources of infor-mation. First, the acquisition of verbal imitativeskills appeared only after the acquisition of non-verbal imitation. Therefore, the data suggestthat verbal imitative learning is facilitated whenthe autistic child can reliably produce non-verbalimitative behaviors. Second, shifts in presenta-tion of the model cues (e.g., letters to words andwords to sentences) produced only mild and verybrief performance decrements, indicating thatcue changes should be made when a subject canimitate approximately 75 to 80% of the model'sbehavior. Finally, Figure 1 presents evidenceindicating that changes in the boy's behavior

were produced by the treatment operations.When the parents were instructed to discontinuetreatment (see the Test Probes), per cents forall five response categories reversed direction.

DISCUSSION

One objective of the present research wasto evaluate experimentally the effectiveness of aparent-administered treatment program for anautistic child. The results clearly suggest thatparents can be trained successfully to treat awide variety of autistic behaviors. The resultsalso indicate that techniques essentially de-veloped in laboratory settings do not presentsetting limitations; i.e., the present treatmenttechniques, though similar to those used bylaboratory researchers, were easily learned andeffectively applied by parents in a clinical settingand at home. Thus, in some cases it may notbe necessary to institutionalize the autistic childor depend entirely upon professional personnelto administer the treatment. Once a parent haslearned the basic principles of operant rein-forcement therapy, and has had an opportunityto receive supervision in the administration ofoperant techniques, he can begin to assume fullresponsibility for administering the treatmentpackage in the home setting. The professionalcan then direct his attention to the design ofthe treatment package, making changes andadvising the parents as new problems arise.Such a "home based" program would consider-ably reduce treatment costs incurred by thefamily and appreciably diminish concern for thegeneralization of treatment from the laboratory(or clinic) to the child's natural environment.

Although a "naturalistic" treatment approachhas several distinct advantages, its success de-pends upon the ability of professionals andparents to satisfy the following conditions: first,the parents must be willing to make a prolongedcommitment to the program. They must spenda considerable amount of time familiarizingthemselves with operant reinforcement prin-ciples and subject themselves to several training

85

86 VEY MICHAEL NORDQUIST and ROBERT G. WAHLER

sessions before attempting to "go it on theirown". Most important, the parents must fullycomprehend the extent of their commitment.Many months of continuous therapy may be re-quired to attain treatment results. Second, thepresent research indicates quite clearly that anaturalistic approach will work only when po-tent reinforcers are accessible to the parents.Fortunately, a home based treatment programcan utilize most reinforcers employed in labora-tory settings; also, reinforcers indigenous to thenatural setting may be identified that otherwisemight have gone unnoticed had treatment beenconducted solely in a laboratory or clinic setting.

Thus, the implications of the present researchare clear. Naturalistic treatment of the autisticchild is one therapeutic alternative to therapyconducted in a laboratory or clinic setting. Con-ceivably, teachers as well as parents might betrained to treat autistic behaviors. Certainly, theextension of the treatment locus to include boththe school and the home offers exciting researchpossibilities.

REFERENCES

Breger, L. Comments on "Building social behaviorin autistic children by use of electric shock."Journal of Experimental Research in Personality,1965, 1, 110-113.

Brown, R. A., Pace, Z. S., and Becker, W. C. Treat-ment of extreme negativism and autistic behaviorin a 6-year-old boy. Exceptional Children, 1969,36, 115-122.

Hawkins, R. P., Peterson, R. F., Schwied, D., andBijou, S. W. Behavior therapy in the home:amelioration of problem parent-child relationswith the parent in a therapeutic role. Journal ofExperimental Child Psychology, 1966, 4, 99-107.

Hewett, F. M. Teaching speech to an autistic childthrough operant conditioning. Journal of Ortho-psychiatry, 1965, 35, 927-936.

Hingtgen, J. N. and Churchill, D. W. Differentialeffects of behavior modification in four muteautistic boys. In D. Churchill, G. Alpew, and M.DeMyer (Eds.), Infantile autism. Charles E.Thomas (Pub.): Illinois, 1971. Pp. 185-199.

Hingtgen, J. N., Coulter, S. K., and Churchill, D. W.Intensive reinforcement of imitative behaviorin mute autistic children. Archives of GeneralPsychiatry, 1967, 17, 36-43.

Hingtgen, J. N., Sanders, B. J., and DeMyer, M. K.Shaping cooperative responses in early childhoodschizophrenics. In R. Ullman and L. Krasner(Eds.), Case studies in behavior modification. NewYork: Holt, Rinehart & Winston, 1965. Pp. 130-137.

Hingtgen, J. N. and Trost, F. C. Shaping coopera-tive responses in early childhood schizophrenics:reinforcement of mutual physical contact andvocal responses. In R. Ulrich, T. Stachnik, and J.Mabry (Eds.), Control of human behavior. Scott,Foresman & Co.: Glenview, Ill., 1966. Pp. 110-113.

Lovaas, 0. I., Freitag, G., Gold, V. J., and Kassorla,I. C. Experimental studies in childhood schizo-prenia: analysis of self-destructive behavior.journal of Experimental Child Psychology, 1965,2, 67-84.

Lovaas, I. I., Freitag, G., Kinder, M. I., Rubenstein,B. D., Schaeffer, B., and Simmons, J. Q. Estab-lishment of social reinforcers in two schizo-phrenic children on the basis of food. Journal ofExperimental Child Psychology, 1966, 4, 109-125.

Lovaas, 0. I., Freitag, K. N., and Whalen, C. Theestablishment of imitation and its use for thedevelopment of complex behavior in schizophrenicchildren, Behaviour Research and Therapy, 1967,5, 171-181.

Lovaas, 0. I., Litrownik, A., and Mann, R. Responselatencies to auditory stimuli in autistic children.Behaviour Research and Therapy, 1971, 9, 39-49.

Lovaas, 0. I., Schaeffer, B., and Simmons, J. Q.Building social behavior in autistic children byuse of electric shock. Journal of ExperimentalResearch in Personality, 1965, 1, 99-109.

Lovaas, 0. I. and Simmons, J. Q. Manipulation ofself-destruction in three retarded children. Journalof Applied Behavior Analysis, 1969, 2, 143-157.

McConnell, 0. L. Control of eye contact in anautistic child. Journal of Child Psychology andPsychiatry, 1967, 8, 249-255.

Metz, J. R. Conditioning generalized imitation inautistic children. Journal of Experimental ChildPsychology, 1965, 2, 389-399.

Nelson, R. 0. and Evans, I. M. The combination oflearning principles and speech therapy techniquesin the treatment of non-communication children.journal of Child Psychology and Psychiatry,1968, 9, 111-124.

Risley, T. The effects and side effects of punishingthe autistic behaviors of a deviant child. Journalof Applied Behavior Analysis, 1968, 1, 21-34.

Risley, T. and Wolf, M. M. Experimental manipula-tion of autistic behaviors and generalization intothe home. In R. Ulrich, T. Stachnick, and J.Mabry (Eds.), Control of human behavior. Scott,Foresman & Co.: Glenview, Ill., 1966. Pp. 187-198.

Risley, T. and Wolf, M. M. Establishing functional

NATURALISTIC TREATMENT OF AN AUTISTIC CHILD 87

speech in echolalic children. Behaviour Researchand Therapy, 1967, 5, 73-88.

Schell, R. E., Stark, J., and Giddan, J. J. Develop-ment of language behavior in an autistic child.Journal of Speech and Hearing Disorders, 1967,32, 51-64.

Simmons, J. Q. and Lovaas, 0. I. Use of pain andpunishment as treatment techniques with child-hood schizophrenics. American Journal of Psy-chotherapy, 1969, 23 (1), 23-36.

Tate, B. B. and Baroff, G. S. Aversive control ofself-injurious behavior in a psychotic boy. Be-haviour Research and Therapy, 1966, 4, 281-287.

Wahler, R. G. Oppositional children: a quest for

parental reinforcement control. Journal of Ap-plied Behavior Analysis, 1969, 2, 159-170.

Wetzel, R., Baker, J., Roney, M., and Martin, M.Outpatient treatment of autistic behavior. Be-haviour Research and Therapy, 1966, 4, 169-17 7.

Wolf, M., Risley, T., and Mees, H. Application ofoperant conditioning procedures to the behaviorproblems of an autistic child. Behaviour Researchand Therapy, 1964, 1, 305-312.

Received 10 April 1972.(Revision requested 14 June 1972.)(Final acceptance 10 August 1972.)