Drug Control of Hyperactive Children

JOURNAL OF APPLJED BEHAVIOR ANALYSIS 1975, 8, 137-146 NUMBER

A BEHAVIORAL-EDUCATIONAL ALTERNATIVE TODRUG CONTROL OF HYPERACTIVE CHILDREN

TEODORO AYLLON,1 DALE LAYMAN, AND HENRY J. KANDEL

GEORGIA STATE UNIVERSITY

AND

UNIVERSITY OF ILLINOIS AT CHICAGO CIRCLE

A behavioral procedure for controlling hyperactivity without inhibiting academic per-formance is described. Using a time-sample observational method, the hyperactivitydisplayed by three school children was recorded during math and reading classes. Con-currently, math and reading performances were measured. The study consisted of twobaselines, one while the children were on medication and the second while they were offmedication. A multiple-baseline design across the two academic subject matters wasused to assess the behavioral intervention, which consisted of token reinforcement forcorrect academic responses in math and subsequently math and reading. Discontinuationof medication resulted in a gross increase in hyperactivity from 20% to about 80%,and a slight increase in math and reading performance. Introduction of a behavioralprogram for academic performance, during no medication, controlled the children'shyperactivity at a level comparable to that when they were on drugs (about 20%).At the same time, math and reading performance for the group jumped from about12%, during baseline to a level of over 85% correct. Each child performed behaviorallyand academically in an optimal manner without medication. Contingency managementtechniques provided a feasible alternative to medication for controlling hyperactivity inthe classroom while enabling the children to grow academically.DESCRIPTORS: drug therapy, hyperactivity, classroom behavior, academic behavior,

emotionally disturbed, multiple baseline, token economy

Hyperactivity or hyperkinesis in the class-

room is a clinical condition characterized byexcessive movement, unpredictable behaviors,unawareness of consequences, inability to focuson and concentrate on a particular task, andpoor academic performance (Stewart, Pitts,Craig, and Dieruf, 1966). It is estimated thatabout 200,000 children in the United States are

currently receiving amphetamines to controltheir hyperactivity (Krippner, Silverman, Cav-allo, and Healy, 1973).

Drugs such as methylphenidate (Ritalin) andchlorpromazine have been shown to controlhyperactivity in the laboratory and applied set-

1The cooperative spirit of the parents and teachersof the children in this study is gratefully acknowl-edged. Special thanks go to Dr. E. Ensminger for hisunflagging interest and encouragement. Reprints may

be obtained from T. Ayllon, Psychology Department,Georgia State University, University Plaza, Atlanta,Georgia 30303.

tings. The evidence from the laboratory is basedon recording devices actuated by the child'smovements (Hollis and St. Omer, 1972;Sprague, Barnes, and Werry, 1970; Sykes,Douglas, Weiss, and Minde, 1971). In the class-room, children have been rated by their teachersalong various dimensions to determine theeffectiveness of stimulants on their behavior.Comly (1971) found that of 40 hyperactivechildren, whose behavior was rated twice weeklyby teachers, those children receiving stimulantswere rated as having better listening ability, lessexcitability, less forgetfulness, and better peer

relationships. In a similar study, Denhoff, Davis,and Hawkins (1971) showed that teachers ratedhyperactive children on dextro-amphetamine(Dexedrine) as improved on measures of hyper-activity, short attention span, and impulsivity.In addition, global ratings by parents, teachers,and clinicians have shown that drugs such as

methylphenidate (Ritalin) and dextro-ampheta-137

2 (SUMMER 1975)

TEODORO AYLLON, DALE LAYMAN, and HENRY J. KANDEL

mine decreased children's hyperactivity in schooland at home (Conners, 1971).

While there is still some conflicting evidenceon drug effectiveness (Krippner et al., 1973),as well as a growing ethical concern for themorality and wisdom implied in administeringmedication to children, (Fish, 1971; Hentoff,1970; Koegh, 1971; Ladd, 1970) drugs arecommonly used to control hyperactivity in theclassroom.

Because the often-implied objective behindthe use of drugs for the hyperactive child is thatof enabling him to profit academically, it issurprising that few data directly support thisbelief. Most studies have measured the effect ofmedication on component skills of learning,e.g., attention, concentration, and discrimina-tion. For example, Conners and Rothschild(1968), Epstein, Lasagna, Conners, Rodriquez(1968), Knights and Hinton (1968) tested drugeffects on general intelligence test performance.Sprague et al. (1970) studied children's re-sponses of "same" or "different" to pairs ofvisual stimuli presented on a screen. Conners,Eisenberg, and Sharpe (1964) studied the effectsof methylphenidate (Ritalin) on paired-asso-ciate learning and Porteus Maze performance inchildren with hyperactive symptoms. Others(Conners, Eisenberg, and Barcai, 1967; Spragueand Toppe, 1966), concentrated their efforts onthe effects of drugs on the attention of hyper-active children to various tasks. These labora-tory studies investigated the effects of drugs oncomponent skills related to learning, but theydid not measure academic performance per se(e.g., math and reading) in the classroom.

Sulzbacher (1972) experimentally analyzedthe effects of drugs on academic behaviors ofhyperactive children in the classroom. Measuresof correct solutions and error rates were takenin arithmetic, writing, and reading in threehyperactive children. In addition, measures weretaken of the children's rates of talk-outs inclass and their rates of out-of-seat behavior dur-ing class. The children were successively givena placebo, then 5 mg of dextro-amphetamine

(Dexedrine), and finally 10 mg of dextro-am-phetamine. The results showed that medicationof 5 mg improved the children's academic re-sponses; however, there was wide variance inacademic performance when the children wereadministered 10 mg. The results for social be-havior also varied. Of two children, one showedless hyperactive classroom behavior (talk-outsand out-of-seat behavior) at a dosage level dif-ferent than the second child. However, the pla-cebo had more effect on controlling the thirdchild's behavior than did medication. The au-thor's conclusion was that stimulant drugs "caneffectively modify disruptive behaviors withoutadversely affecting academic performance in theclassroom". Drug effects on academic perform-ance, however, were highly variable.

Since Sulzbacher's major interest was in de-termining the role of drugs on hyperactivityand academic performance, he did not pursuebehavioral alternatives to the control of hyper-activity. Yet, there is at present, a body of estab-lished findings indicating that such alternativesmay be available. For example, O'Leary andBecker (1967) found that when children wererewarded for sitting, making eye contact withthe teacher, and engaging in academically re-lated activities, their misbehavior was virtuallyeliminated. Ayllon, Layman, and Burke (1972)showed that misbehavior may be also reduced,not by rewarding the child for good conduct,but by imposing academic structure in the class-room. This structure involved giving academicassignments with a short time limit for theircompletion. Ayllon and Roberts (1974) foundthat another behavioral technique to eliminateclassroom misbehavior is to reward children foracademic performance only. These findings sug-gest that disruptive behavior can be weakenedby reinforcing incompatible academic perform-ance. Using this method, the child performswell both academically and socially withouttreating the disruptive behavior directly.The children in the above studies were dis-

ruptive, not hyperactive. Although the topog-raphy of the response is similar, hyperactivity

138

AN ALTERNATIVE TO DRUG CONTROL

differs from disruption in its magnitude, dura-tion, and frequency. Illustrations of this differ-ence are well documented, indicating that hy-peractive children are in constant motion, fidgetexcessively, frequently enter and leave the class-room, move from one class activity to anotherand rarely complete their projects or stay withone particular game or activity. Their academicperformance is typically poor (Campbell, Doug-las, Morgenstern, 1971; Freibergs and Douglas,1969; Stewart, Pitts, Craig, and Dieruf, 1966;Sykes, Douglas, Weiss, and Minde, 1971).Two questions arise:Can behavioral techniques used to decrease

disruptive behavior be at least as effective asdrugs in controlling an extreme form of class-room misbehavior such as hyperactivity? At thesame time, can such techniques help the hyper-active child to grow educationally? The presentstudy attempted to answer these questions.

METHOD

Subjects and SettingThree school children, (Crystal, Paul, and

Dudley) clinically diagnosed as chronically hy-peractive, were all receiving drugs to controltheir hyperactivity.

Crystal was an 8-yr-old girl. She was 47 in.(118 cm) tall and weighed 76 lb (34.2 kg).She had an I.Q. of 118 as measured on theWISC. She was enrolled in a learning-disabili-ties class because of the hyperactive behaviorshe displayed before taking medication andbecause of her poor academic work. She hadbeen on drugs since she was 5 yr old, when herdoctor felt that her behavior was so unpredict-able that he prescribed 5 mg of Methylpheni-date q.i.d. to calm her down.

Paul was a 9-yr-old boy. He was 53 in. (133cm) tall and weighed 65 lb (29.2 kg). He hadan I.Q. of 94 as measured on the WISC. He hadbeen enrolled in the learning-disabilities classfor 2 yr before the study and had been taking5 mg of methylphenidate b.i.d. for 1 yr to con-trol his hyperactive behavior.

Dudley was a 10-yr-old boy. He was 5 5 in.(138 cm) tall and weighed 76 lb (34.2 kg). Hehad an I.Q. of 103 as measured on the WISC.He was enrolled in a learning-disabilities classfor 2 yr before the study and on the advice ofhis doctor had been taking 5 mg of methyl-phenidate t.i.d. for 4 yr.

In addition to their drug treatment, Crystaland Dudley were under the care of a child psy-chiatrist and a pediatrician during the study.

The three children attended a private elemen-tary school. They were enrolled in a self-con-tained learning disability class of 10 childrenand one teacher. The children and the teacherremained together throughout the school dayin the same room. Other personnel during thestudy consisted of two observer-recorders: oneof the authors and an undergraduate student.

Response DefinitionHyperactivity and academic performance

across two academic periods, math and read-ing, were measured.

Math. Math was defined as addition of wholenumbers under 10. The teacher wrote 10 prob-lems on the board at the beginning of each class.The children were given 10 min to complete theproblems. Problems were taken from LaidlawSeries Workbooks, Levels P and 1.

Reading. Reading was defined as comprehen-sion and was measured by workbook responsesto previously read stories in a basal reader. Eachchild had 20 min to complete a 10-questionworkbook page per day. The books were Merril-Linguistic Readers - 3. In both math and read-ing, the written response served as a permanentproduct from which the percentage of correctanswers could be determined.The academic assignments in both math and

reading increased slightly in difficulty as thechild progressed through the work.

Hyperactivity. Since hyperactive behavior hasoverlapping topographical properties with otherdeviant behaviors, hyperactive behavior was de-fined using the same response definition as pre-sented by Becker, Madsen, Arnold, and Thomas

139

TEODORO AYLLON, DALE LAYMAN, and HENRY J. KANDEL

for deviant behavior in the classroom (1967).To define and record deviant behavior, Beckerand his colleagues used seven general categoriesof behavior incompatible with learning. Theseincluded gross motor behaviors, disruptive noisewith objects, disturbing others, orienting re-sponses, blurting out, talking, and other miscel-laneous behaviors incompatible with learning.In the present experiment, the behaviors of thehyperactive children most often fell into thefollowing four categories: gross motor behav-iors, disruptive noise, disturbing others, andblurting out. The most frequently recorded cate-gory for these hyperactive children was grossmotor behaviors, which included runningaround the room, rocking in chairs, and jump-ing on one or both feet. Disruptive noise withobjects included the constant turning of bookpages and the excessive flipping of notebookpaper. Disturbing others and blurting out in-cluded the constant movement of arms, result-ing in the destruction of objects and hittingothers, screaming, and high-pitched and rapidspeech. Categories that were not recorded withany consistency included orienting responses andtalking, as in a conversation with another per-son. Thus, although the response definition fordeviant behavior was used, the actual recordingwas heavily weighted on those behaviors de-scribed by Stewart et al. (1966) as being typicalof hyperactive children.

Observational and recording procedure for hy-peractivity. Initially, six children were identifiedby the school director as being hyperactive andreceiving medication for it. These children wereobserved across two class periods: math andreading. The duration of each class period was45 min. Each child was observed in successiveorder on a time-sample of 25 sec. At the end ofeach 25-sec interval, the behavior of the childunder observation was coded as showing hyper-activity or its absence. At that time, the observermarked a single slash in the appropriate inter-val, on a recording sheet, if one or more hy-peractive behaviors occurred. If no hyperactivebehaviors were observed at that time, the ap-

propriate interval was marked with an "O". Thenumber of intervals of hyperactivity over thetotal number of intervals for each child gavethe observer the per cent of intervals in whicheach child was hyperactive. Each of the sixchildren was observed a total of 17 times per45-min class period. Using this recording pro-cedure, it was possible to determine, duringbaseline, that the most chronically hyperactivechildren were Crystal, Paul, and tDudley. Bydropping observations on the less-severely hy-peractive children it was possible to increase thenumber of observations for the chronically hy-peractive ones. Recording hyperactivity from onechild to the next was now sampled about every18 sec in the manner described above. Each childwas now observed approximatetly 50 times eachclass period throughout the remaining phases ofthe experiment.

Observer agreement on academic perform-ance and hyperactivity. The percentage of cor-rect math and reading problems was checkedby the teacher and one of the authors each dayand the obtained agreement score was 100%on each occasion for each child.

Reliability checks for hyperactivity weretaken by one of the authors and one of threeundergraduate students in Special Education.The student was given the list of deviant be-haviors described by Becker et al. (1967) oneday before the reliability check to become fa-miliar with the responses. The students werenot told of the purpose of the study or of thechanges in experimental conditions. Each ob-server during the reliability check used a watchwith a sweep second hand. In addition, a pre-pared sheet showed the observers the sequencein which the children were to be sampled andthe intervals at the end of which each observerwas to look at the subject and record whetheror not the behavior was occurring at that in-stant. Each observer sat on opposite sides ofthe room to ensure unbiased observations.The percentage of agreement for hyperactive

as well as nonhyperactive behavior was calcu-lated by comparing each interval and dividing

140

AN ALTERNATIVE TO DRUG CONTROL

agreements in each by the total number of ob-servations and multiplying by 100. Reliabilitychecks were taken to include the baseline pe-riod under medication (Blocks 2, 3, 5, and 6;in Figures 1, 2, and 3), the period when medi-cation was discontinued and no reinforcementwas available (Blocks 7 and 9), and the finalperiod when reinforcement was introduced inboth math and reading (Block 11). Reliabilityscores for hyperactivity for each child were al-ways more than 85%, with the scores rangingfrom a low of 87% to a high of 100%. Theaverage reliability score was 97%.

Check-point system and back-up reinforcers.A token reinforcement system similar to thatused by O'Leary and Becker (1967) in a class-room setting was used. Children were awardedchecks by the teacher on an index card. Onecheck was recorded for each correct academicresponse. The checks could be exchanged for alarge array of back-up reinforcers later in theday. The back-up reinforcers ranged in pricefrom one check to 75 checks, and included suchitems and activities as candy, school supplies,free time, lunch in the teacher's room, and pic-nics in the park.

ProcedureEach subject's daily level of hyperactivity and

academic achievement, on and off medication,were directly observed and recorded before thebehavioral program. In addition, using a mul-tiple-baseline design, the relative effectivenessof the motivational system on (a) hyperactivityand (b) academic performance, in math andreading was evaluated. This type of design al-lowed each child to serve as his own control,thereby minimizing the idiosyncratic drug-be-havior interactions that have the potential forconfounding the interpretations and even theresults when comparing one subject with an-other. This design is particularly useful in thestudy of the effects of discontinuing drugs onbehavior, since as Sprague et al. (1970) and Sulz-bacher (1972) have pointed out, the inherentproblem in assessing effects of medication lies

in the fact that each child reacts to the presenceor absence of medication on an individual basis.The design of the study included the follow-

ing four phases:Phase 1: on medication. Crystal, Paul, and

Dudley were observed for 17 days to evaluatehyperactive behavior when they were takingdrugs. Academic performance in math and read-ing was also measured.With the full cooperation of the children's

doctors and their parents, medication was dis-continued on the eighteenth day, a Saturday.An additional two days, Sunday and Monday (aschool holiday) allowed a three-day "wash-out"period for the effect of medication to disappear.It is known that these stimulant drugs are al-most completely metabolized within one day.No measures of hyperactivity or academic per-formance were obtained during this weekendperiod.

Phase 2: off medication. Following the three-day "wash-out" period, a three-day baseline whenthe children were off medication was obtained.Time-sampling observations of hyperactivitywere continued, as well as measures of academicperformance. This phase served as the basisagainst which the effects of reinforcement onhyperactivity and academic performance couldlater be compared.

Phase 3: no medication; reinforcement ofmath. During this six-day period, the childrenremained off drugs while the teacher introduceda reinforcement system for math performanceonly. Observations of hyperactivity continuedand academic performance was measured.

Phase 4: no medication; reinforcement ofmath plus reading. During this six-day phase,the children remained off drugs while reinforce-ment was added for reading and reinforcementof math was maintained. Observations of hyper-activity and measures of academic performancewere continued.

RESULTSWhen Ritalin was discontinued, the level of

hyperactivity doubled or tripled its initial level.

141

TEODORO AYLLON, DALE LAYMAN, and HENRY J. KANDEL

However, when reinforcement was systemati-cally administered for academic performance,hyperactivity for all three children decreased toa level comparable to the initial period whenRitalin chemically controlled it.

Figure 1 shows that hyperactivity for Crystalduring the drug phase in math averaged about20%, while academic performance in math waszero. When Ritalin was discontinued, hyperac-tivity rose to an average of 87% and math per-formance remained low at an average of 8%.When math was reinforced, and Crystal con-tinued to stay off drugs,significantly from 879%

hyperactivity droppedto about 9%. Math

performance increased to 65 %. Hyperactivityin math was effectively controlled through re-inforcement of math performance. However, themultiple-baseline design shows that concurrentlyCrystal's hyperactivity during reading class re-mained at 90% before reinforcement was intro-duced for correct reading responses.

At the same time measures were taken in thearea of math, hyperactivity and academic per-formance were also measured in the area ofreading. Crystal's hyperactivity during readingclass averaged approximately 10% under medi-cation. Academic performance in reading waszero under medication. When Crystal was taken

MEDICATION NO MEDICATION

0

ID

*inm@peuftfumo 10oo

i 2 3 4 B C OF 9DAYS (IN MLOCKS OF 3)

dbw HYPERACTIVITY

-

CRYSTA __o n~r&NP.4.CRYSTIAL pACADEMIC PERFORMANCE

Fig. 1. Crystal. The percentage of intervals inwhich hyperactivity took place and the per cent ofcorrect math and reading performance. The first andsecond segments respectively show the effects of med-ication, and its subsequent withdrawal, on hyperac-tivity and academic performance. A multiple-baselineanalysis of the effects of reinforcement across mathand reading and concurrent hyperactivity, is shownstarting on the third top segment. The last segmentshows the effects of reinforcement on math plusreading and its concurrent effect on hyperactivity.(The asterisk indicates one data point averaged over

two rather than three days).

MEDICATION NO MEDICATIONI I 01 0

O 1

234 5 6 ? 8 9 10 II

DAYS (IN BLOCKS OF 3)PAUL *-.HYPERACTIVITYP 0 ACADEMIC PERFORMANCE

Fig. 2. Paul. The percentage of intervals in whichhyperactivity took place and the per cent of correctmath and reading performance. The first and secondsegments respectively show the effects of medication,and its subsequent withdrawal, on hyperactivity andacademic performance. A multiple-baseline analysisof the effects of reinforcement across math and read-ing and concurrent hyperactivity is shown starting onthe third top segment. The last segment shows theeffects of reinforcement on math plus reading and itsconcurrent effect on hyperactivity. (The asterisk indi-cates one data point averaged over two rather thanthree days).

142

AN ALTERNATIVE TO DRUG CONTROL

off drugs, hyperactivity rose dramatically from10% to an average of 91%. Academic perform-ance remained low at approximately 10%. Onlywhen reinforcement was administered for read-ing was hyperactivity in this area reduced from91% to 20%. Reading performance increasedfrom 10% to an average of 69%.

Similar results were found for Paul and Dud-ley, as can be seen in Figures 2 and 3.

Figure 4 shows the pre and post measures ofhyperactivity and academic performance forDudley, Crystal, and Paul as a group. It canbe seen that when the children were takingdrugs, hyperactivity was well controlled and

100-

75-

z

25-

o-100-

75-

z( 50-a49

2P5-

MEDICATION NO

0t ^ rsv VVJ0

_} All~~~~~~~~~~~~~~~~~~

0 ° o O0

fl*I - I

12

MEDICATIONO 01 0 0

REINFO

MATH

12 34 56 789101011

* DAYS (IN BLOCKS OF 3)DUDLEY -HYPERACTIVITY

0 0 ACADEMIC PERFORMANCE

Fig. 3. Dudley. The percentage of intervals inwhich hyperactivity took place and the per cent ofcorrect math and reading performance. The first andsecond segments respectively show the effects of med-ication, and its subsequent withdrawal, on hyperac-tivity and academic performance. A multiple-baselineanalysis of the effects of reinforcement across mathand reading and concurrent hyperactivity is shownstarting on the third top segment. The last segmentshows the effects of reinforcement on math plus read-ing and its concurrent effect on hyperactivity. (Theasterisk indicates one data point averaged over tworather than three days).

II-IU

I-bN93

11 WYPRACMM

SACADO11 "ffOCEION OFF

MEDICATION MEDICATION

Fig. 4. Average per cent of hyperactivity and aca-demic performance in math and reading for threechildren. The first two bars summarize findings fromthe 17-day baseline under drug therapy. The last twobars show results for the final six-day period withoutdrug therapy but with a reinforcement program forboth math and reading performance.

averaged about 24% during math and read-ing. When medication was discontinued anda reinforcement program was established tostrengthen academic performance, the combinedlevel of hyperactivity was about 20% duringmath and reading for the three children. Thislevel (20%) of hyperactivity matched that ob-tained under medication (24%).

During the period when the children weretaking drugs, their per cent correct in math andreading combined, averaged 12%. When medi-cation was discontinued and a reinforcementprogram was established, their average per centcorrect in both.academic subjects increased from12% to 85%.

DISCUSSION

These findings show that reinforcement ofacademic performance suppresses hyperactivity,and they thus support and extend the findingsof Ayllon and Roberts (1974). Further, the aca-demic gains produced by the behavioral pro-gram contrast dramatically with the lack ofacademic progress shown by these children un-der medication.2

2For a systematic replication of this study see Lay-man, unpublished.

143

TEODORO AYLLON, DALE LAYMAN, and HENRY J. KANDEL

The multiple-baseline design demonstratesthat token reinforcement for academic achieve-ment was responsible for the concurrent sup-pression of hyperactivity. Indeed, while this con-trol was demonstrated during math periods, thechildren's concurrent hyperactivity during read-ing remained at a high level, so long as the re-inforcement procedure for reading was with-held. Only when reinforcement was introducedfor both math and reading performance did thehyperactivity for all three children drop to lev-els comparable to those controlled by the drug.

The control over hyperactivity by the en-hancement of academic performance was quick,stable, and independent of the duration anddosage of the medication received by each childbefore the program. One child had been undermedication for as long as 4 yr, another childfor 1 yr. Despite this extreme difference in his-tory of medication, the behavioral effects werenot differential to that history.When medication was discontinued, hyperac-

tivity increasedimmediately and to a high levelin all three children. The effectiveness of medi-cation in controlling hyperactivity, evaluatedthrough direct observations of behavior, sup-ports the data of earlier studies using record-ings based on instrumentation (Hollis et al.,1972; Sprague et at., 1970; Sykes et al., 1971).During the few days of no medication, hy-

peractivity became so severe that the teacherand parents freely commented on the gross dif-ference in the children's behavior in school andat home. Their reports centered around suchdescriptions as "He's just like a whirlwind","She is climbing the walls, it's awful", "Justcan't do a thing with her . . ." "He's not attend-ing, doesn't listen to anything I tell him", andothers. It was only with a great deal of supportand counselling that the teacher and parentswere able to tolerate this stressful period. It wasthis high level of hyperactivity shown by allthree children that allowed the opportunity totest the effectiveness of a reinforcement pro-gram for academic performance in controllinghyperactivity.

Since both hyperactivity and academic per-formance increased concurrently, as soon asmedication was discontinued, it might be con-strued that these two dimensions are compati-ble. This may be an unwarranted conclusion,however, because the slight increments in aca-demic performance concurrent with incrementsin hyperactivity may only reflect the type of re-cording method used in this study. For example,measures of the behavior of the children showthat once they had finished their academic as-signments, they became hyperactive. Thus, aca-demic performance and hyperactivity could takeplace sequentially. When the time limit foracademic performance had expired (e.g., after10 or 20 min, depending on the subject matter)the child could engage in hyperactivity for therest of the class period.

It usually took only one session for each childto learn that academic performance was associ-ated with reinforcement while hyperactivity wasnot, suggesting that in the absence of medica-tion these children react to reinforcement asnormal children do. The classroom with rein-forcement procedures now set the occasion foracademic performance, rather than hyperactivity.The present results suggest that the continued

use of Ritalin and possibly other drugs to con-trol hyperactivity may result in compliant butacademically incompetent students. Surely, thegoal of school is not to make children into doc-ile robots either by behavioral techniques or bymedication. Rather, the goal should be one ofproviding children with the social and academictools required to become successful in their so-cial interactions and competent in their aca-demic performance. Judging from the reactionsand comments of both parents and teacher, thisgoal was achieved during the reinforcement pe-riod of the study. The parents were particularlyrelieved that their children, who had been de-pendent on Ritalin for years, could now func-tion normally in school without the drug. Simi-larly, the teacher was excited over the fact thatshe could now build the social and academicskills of the children because they were more

144

AN ALTERNATIVE TO DRUG CONTROL 145

attentive and responsive to her than when theywere under medication.On the basis of these findings, it would seem

appropriate to recommend that hyperactive chil-dren under medication periodically be given theopportunity to be drug-free, to minimize drugdependence and to facilitate change through al-ternative behavioral techniques. While this studyfocused on behavioral alternatives to Ritalin forthe control of hyperactivity, it is possible thatanother drug or a combination of medicationand a behavioral program may also be helpful.

This study offers a behavioral and education-ally justifiable alternative to the use of medica-tion for hyperactive children. The control ofhyperactivity by medication, while effective, maybe too costly to the child, in that it may retardhis academic and social growth, a human costthat schools and society can ill afford.

REFERENCES

Ayllon, T. and Kelly, K. Effects of reinforcementon standardized test performance. Journal of At-plied Behavior Analysis, 1972, 5, 477-484.

Ayllon, T., Layman, D., and Burke, S. Disruptivebehavior and reinforcement of academic perform-ance. Psychological Record, 1972, 22, 315-323.

Ayllon, T. and Roberts, M. Eliminating disciplineproblems by strengthening academic performance.Journal of Applied Behavior Analysis, 1974, 7,7 1-76.

Becker, W., Madsen, C., Arnold, C., and Thomas, D.The contingent use of teacher attention and praisein reducing classroom behavior problems. TheJournal of Special Education, 1967, 1, 287-307.

Campbell, S., Douglas, U., and Morgenstern, G.Cognitive styles in hyperactive children and theeffect of methylphenidate. Journal of Child Psy-chology and Psychiatry, 1971, 12, 55-67.

Comly, H. Cerebral stimulants for children withlearning disorders. Journal of Learning Disabili-ties, 1971, 4, 484-490.

Conners, C., Eisenberg, L., and Barcai, A. Effect ofdextro-amphetamine in children: studies on sub-jects with learning disabilities and school behav-ior problems. Archives of General Psychiatry,1967, 17, 478-485.

Conners, C., Eisenberg, L., and Sharpe, L. Effects ofmethylphenidate (Ritalin) on paired-associatelearning and Porteus performance in emotionallydisturbed children. Journal of Consulting Psy-chology, 1964, 28, 14-22.

Conners, C. and Rothschild, G. Drugs and learningin children. In G. Helmuch (Ed.), Learning dis-orders, Vol. 3. Seattle, Special Child Publications,1968.

Conners, K. Recent drug studies with hyperkineticchildren. Journal of Learning Disabilities, 1971,4, 476-483.

Denhoff, E., Davis, A., and Hawkins, A. Effects ofdextro-amphetamine on hyperkinetic children: acontrolled double blind study. Journal of Learn-ing Disabilities, 1971, 4, 491-499.

Epstein, L., Lasagna, L., Conners, K., and Rodriguez,A. Correlation of dextro-amphetamine excre-tion and drug response in hyperkinetic children.Journal of Nervous and Mental Disease, 1968,146, 136-146.

Fish, B. The "one child, one drug" myth of stimu-lants in hyperkinesis: importance of diagnosticcategories in evaluating treatment. Archives ofGeneral Psychiatry, 1971, 25, 193-203.

Freibergs, V. and Douglas, V. Concept learning inhyperactive and normal children. Journal of Ab-normal Psychology, 1969, 74, 388-395.

Hentoff, N. The drugged classroom. Evergreen Re-view, December, 1970, 6-11.

Hollis, J. and St. Omer, V. Direct measurement ofpsychopharmacologic response: effects of chlor-promazine on motor behavior of retarded chil-dren. American Journal of Mental Deficiency,1972, 76, 397-407.

Keogh, B. Hyperactivity and learning disorders: re-view and speculation. Exceptional Children, 1971,38, 101-109.

Knights, R. and Hinton, G. Minimal brain dys-function: clinical and psychological test charac-teristics. Academic Therapy, 1968, 4, 265-273.

Krippner, S., Silverman, R., Cavallo, M., and Healy,M. A study of hyperkinetic children receivingstimulant drugs. Academic Therapy, 1973, 8,261-269.

Ladd, E. Pills for classroom peace? Saturday Re-view, November, 1970, 66-83.

Layman, D. A behavioral investigation: the effectsof medication on disruptive classroom behaviorand academic performance. Unpublished doctoraldissertation, Georgia State University, 1974.

O'Leary, K. D. and Becker, W. C. Behavior modifi-cation of an adjustment class: a token reinforce-ment program. Exceptional Children, 1967, 33,637-642.

Sprague, R., Barnes, B., and Werry, J. Methylpheni-date and thoridazine: learning, reaction time, ac-tivity, and classroom behavior in disturbed chil-dren. American Journal of Orthopsychiatry, 1970,40, 615-628.

Sprague, R. and Toppe, L. Relationship betweenactivity level and delay of reinforcement. Experi-mental Child Psychology, 1966, 3, 390-397.

Stewart, M., Pitts, F., Craig, A., and Dieruf, W. The

146 TEODORO AYLLON, DALE LAYMAN, and HENRY J. KANDEL

hyperactive child syndrome. American JournalofOrthopsychiatry, 1966, 36, 861-867.

Sulzbacher, S. Behavior analysis of drug effects inthe classroom. In G. Semb (Ed.), Behavior analy-sis and education. University of Kansas, 1972.

Sykes, D., Douglas, V., Weiss, G., and Minde, K.Attention in hyperactive children and the effectof methylphenidate (Ritalin). Journal of ChildPsychology and Psychiatry, 1971, 12, 129-139.

Received 17 October 1974.(Final acceptance 6 January 1975.)