JOURNAL OF APPLIED BEHAVIOR ANALYSIS

EFFECTS OF A VARIABLE-RATIO REINFORCEMENT SCHEDULE WITHCHANGING CRITERIA ON EXERCISE IN OBESE AND NONOBESE BOYS

RAYLEEN V. DE LucA AND STEPHEN W. HOLBORNUNIVERSITY OF MANITOBA

The effects of a variable-ratio schedule of reinforcement on pedaling a stationary exercise bicyclewere examined. Three obese and three nonobese 11-year-old boys were individually tested five timesweekly for approximately 12 weeks. A changing-criterion design was used in which each successivecriterion was increased over mean performance rate in the previous phase by approximately 15%.The contingencies of the successive criteria resulted in systematic increases in rate of exercise for allchildren. Final variable-ratio rates were higher than those under fixed ratios found in previousresearch, with rates for 2 of the 3 obese boys approximating those of the nonobese.DESCRIPTORS: obesity, exercise, changing criterion design, variable-ratio schedule, children

Childhood obesity is a prevalent problem (Ar-istimuno, Foster, Voors, Srinivasan, & Berenson,1984) that has been demonstrated to cause phys-iological (e.g., Mann, 1974; Mayer, 1970; Mobbs,1970) as well as psychological problems (e.g., Israel& Stolmaker, 1980). Although behavioral treat-ment approaches for childhood obesity have beenexplored recently (e.g., Epstein, Wing, Koeske, &Valoski, 1987), Epstein and Wing (1987) assertedthat "additional research on exercise's role is need-ed" (p. 336). Because research has indicated thatobesity is often associated with physical inactivity(e.g., Bullen, Reed, & Mayer, 1964; Cohen, Gel-fand, Dodd, Jensen, & Turner, 1980; Johnson,Burke, & Mayer, 1956; Mayer, 1968; Rony, 1940;Stefanik, Heald, & Mayer, 1959), exercise appearsto be an important focus for research.The benefits of behavioral principles of rein-

forcement to increase physical activity with obesechildren have been demonstrated in recent studiesconducted by De Luca and Holborn (1985, 1990).Applying a fixed-interval (FI) schedule of rein-forcement to exercising in obese and nonobese boys,De Luca and Holborn (1985) demonstrated thatreinforcement schedules can be a powerful tool toincrease exercise duration. In an attempt to increase

The authors express appreciation to Gloria Eldridge andRandy Atkinson for assistance with the equipment as wellas for their suggestions and support throughout the study.This research was supported by the Manitoba Mental HealthResearch Foundation. Reprints may be obtained from Ray-leen V. De Luca, Department of Psychology, University ofManitoba, Winnipeg, Manitoba R3T 2N2, Canada.

response rate, De Luca and Holborn (1990) ex-amined the effects of a fixed-ratio (FR) schedule,matched to the reinforcement frequency of an F11-min schedule, with obese and nonobese childrenpedaling on a stationary exercise bicyde. After theinitial F1 phase, when the schedule was changed toFR, all children showed an increased rate of exer-cising, and the duration of exercising was main-tained at maximum during all contingency phases.The present research was designed to examinewhether variable-ratio (VR) schedules would in-crease and maintain even higher rates of exercisingthan FR and F1 schedules do.A single-case changing-criterion design (Hart-

mann & Hall, 1976; Kazdin, 1982) was used. Thechanging-criterion design has been demonstrated tobe effective in shaping behavior (Hartmann & Hall,1976). Brownell (1984) recommended that "onekey to exercise instruction is to begin at the patient'slevel" (p. 412). Using a changing-criterion design,the initial criterion is based on the subject's ownbaseline performance and is increased via small,successive increments. In this manner, high ratesof exercise can be approached gradually and sys-tematically.

METHOD

Subjects and SettingSix male 11-year-old children, whose parents

consented to their participation in a study involvingexercise on a stationary bicycle, were chosen assubjects. Half of the subjects were obese and half

671

1992)25,671-679 NUMBER 3 (FAu 1992)

62RYLEEN V. DE LUCA and STEPHEN W. HOLBORN

were of normal weight (as defined by age, height,and weight norms, Demirjian, 1980, and skin-foldthickness, Seltzer & Mayer, 1965). To qualify, obesechildren had to be at least 20% above average bodyweight for their age, sex, and height (Demirjian,1980), had no medical problems that contraindi-cated exercise, were not taking medication thatwould affect body weight, and were not involvedin a formal weight control program. The obese boyswere at least at the 97th percentile for weight andranged from 32% to 161.4% over the acceptablerange for percentage body fat. All obese subjectswere classified as "overfat" (Canadian DepartmentofHealth and Welfare [CDHW], 1984). The non-obese boys were at, or slightly below, the 50thpercentile for weight. Two nonobese boys werewithin the acceptable range for percentage body fatfor boys, and 1 was 5.5% over the acceptable rangefor percentage body fat and was dassified as "un-desirable" (CDHW, 1984). The heights of all boyswere within the normal range. All testing was con-ducted in the nurse's room of an elementary schoolin Winnipeg, Manitoba.

ApparatusA CCM stationary exercise bicyde was pro-

grammed to ring a bell and to illuminate a redlight simultaneously after each variable number(VR) of responses during the contingency phases.Lehigh Valley and BRS electromechanical moduleswere used for programming stimulus events andfor recording responses and reinforcements. The belland light were programmed via electromechanicaltimers and relay circuitry. Each wheel revolutionconstituted a response and was recorded on mag-netic counters. Folding 183-cm partitions were usedto obscure the equipment panels from the subject'sview. A 7-W red light was mounted on the frontmiddle panel of the bicyde in the subject's line ofvision. Throughout the study, the tension of thebicyde was set at moderately low resistance (2.27kg) for all subjects.

Experimental DesignA single-case changing-criterion design was em-

ployed (see Hartmann & Hall, 1976; Kazdin,

1982). Baseline data were taken for 3 obese and3 nonobese boys for eight sessions until behaviorappeared stable to visual inspection, after whichthe initial VR schedule was introduced. Eight ses-sions later, when responding was stable, the secondVR subphase was introduced. Following anothereight sessions, when stability was reached, the thirdVR subphase ensued for eight sessions. To exhibitfuiher control, a three-session return-to-baselinephase was followed by a final VR subphase at theprevious criterion, which lasted for five sessions.

ProcedureThroughout the study, subjects were individually

tested daily from Monday to Friday over approx-imately 12 weeks. The instructions given at thebeginning of each session were identical for all sub-jects: "Exercise as long as you like." No furtherencouragement was given. Recording began whenthe subject began pedaling the bicyde. The exper-imenter was seated behind the equipment panelsout of the subject's view. Sessions terminated whenthe subject dismounted from the bicyde or when30 min had elapsed, whichever occurred first.

Baseline. Prior to the first session, each childwas asked if he would like to exercise five timesweekly on the stationary bicyde. During the firstsession, following the consent of each subject toparticipate in the exercise program, the instructions"exercise as long as you like" were given. No stim-ulus changes occurred, nor was reinforcement ad-ministered during baseline. After eight sessions, thepoint system was introduced.

Reinforcement survey. Reinforcement surveys(Cautela, 1977) were individually administered toeach child during the first week of baseline. Tenitems from the preferred categories were selected asback-up reinforcers. Subjects rated on a scale of 1to 10 how much they liked each item. Higher costswere then assigned to the items with higher pref-erence ratings.VR (first subphase). The VR schedule of re-

inforcement was implemented during the ninth ses-sion, after a stable baseline had been achieved. Themean number of revolutions per minute during thebaseline was calculated for each subject. A separate

672

VR REINFORCEMENT SCHEDULE

ratio value was assigned to each subject during theinitial VR phase on the basis of approximately a

15% increase over mean responding during base-line. For example, the obese subject who pedaleda mean of 60 revolutions per minute during base-line received a VR 70 schedule; the schedule forthe nonobese subject, who pedaled a mean of 70revolutions per minute during baseline, was VR

80. This manipulation ensured that the shifts indensity of reinforcement would be similar for sub-jects and that the first criterion would not be overlystringent. Prior to beginning the first session of thisphase, subjects were allowed to examine the 10back-up reinforcers, tagged with the number ofpoints necessary for purchase. Reinforcers indudeda hand-held battery-operated game, kite, bicydebell, flashlight, model car, model plane, puzzle,adventure book, and comic books. Subjects were

advised that they could earn points by exercisingon the bicyde to "buy" the items that they likedbest. They were told that each time the bell rang

and the light went on, they would earn a point.At the end of each session, the subject received a

tally of the number of points he had earned duringthe session and the cumulative total to date.VR (second subphase). The second VR schedule

of reinforcement was implemented during the 17thsession, after stability had been achieved in the firstVR subphase. A separate ratio value was assignedto each subject on the basis of a 15% increase over

mean responding during the previous subphase. Inother words, a different criterion for performancewas specified for each subject, based on his per-

formance in the previous subphase.VR (third subphase). The third VR schedule

of reinforcement was implemented when stabilitywas achieved for the second VR subphase. Again,a separate ratio value was assigned to each subjectduring the third subphase on the basis of a 15%increase over mean responding during the previoussubphase.

Return to baseline. To provide further evidenceof experimental control, the return-to-baseline phasewas implemented after eight sessions on the thirdVR subphase, when stability in responding hadoccurred, and was continued for three sessions.

Return to VR third subphase. The subjects werereturned to their individual third subphase sched-ules for the last five sessions, which coincided withthe last 5 days of school prior to summer vacation.

Social validation. Each subject, as well as par-ents and homeroom and physical education teach-ers, completed a social validation questionnaire atthe end of the program. Questionnaires employeda 7-point rating scale with higher numbers reflect-ing increased social validation. Questionnaires as-certained consumer satisfaction with the treatmentprogram as well as perceived improvements inphysical activity and appearance of the children.

RESULTS

The dependent measures were (a) the overallrate of responding per session (which was calculatedby dividing the total number of revolutions persession by the total number of minutes spent ex-ercising) and (b) the total time spent exercising persession.

Figure 1 illustrates the response rates for allsubjects. During baseline there was some fluctua-tion in response rate, with a gradual decline for 2of the nonobese boys, Shawn and Steve. The re-sponse rate for the remaining nonobese boy, Scott,was relatively stable throughout the baseline phase.The nonobese boys responded at a mean of 71.9revolutions per minute during baseline. For theobese boys, there was an initial increase in responserate from the first to the second session, followedby generally stable responding for Peter and Paulduring baseline.

The performance of the remaining obese subject,Perry, continued to accelerate until the third session,when there was a sharp decrement followed bystable responding. The obese boys responded at amean of 59.2 revolutions per minute during base-line, which was 12.8 revolutions per minute slowerthan their slimmer counterparts. Upon implemen-tation of the initial VR schedules, response ratesfor all subjects increased. Performance was consis-tently above the criterion for all boys (criterion wasset at approximately 15% over each individual sub-ject's baseline performance). The mean rate of re-

673

674 RAYLEEN V. DE LUCA and STEPHEN W. HOLBORN

160BASELINE VR 80 VR 115 VR 130 BL VR 130

140120i-1001-

6080 .-2.....

BASELINE VR 8S VR 11S VR 125 BL VR 125150.1 I_ _

130.I10~~~~~~~~~~

70

.I t1!aI II III I tm SHAWN

rm~~~

15l0 r BASELINE VR8S VR11S VR125 BL VR12S130t _

40 '

co0

BASELINE VR 70 VR 95 VR 100 BL VR 1001401-

o 120 '- * j

> 100so~ ~~~N

BASELINE VR 80 VR 105 VR 120 BL VR 120140

120 k, i_ __

so

140 BASELINE VR 70 VR 90 VR 110 BL VRI 110

120t

100I . -

so~ ~ ~ ~ I

40_,t'.O"1 s RO 20 2Y 10 L YR 10

SESSIONSFigure 1. Mean revolutions pedaled per minute during baseline, VR 1 (VR range, 70 to 85), VR 2 (VR range, 90 to

115), VR 3 (VR range, 100 to 130), return to baseline, and return to VR 3 phases for obese and nonobese subjects.

VR REINFORCEMENT SCHEDULE

sponding during the first VR subphase was 98.89revolutions per minute for the nonobese boys and85.51 revolutions per minute for the obese boys.

The change in criterion in the second VR sub-phase produced an increase in response rate for allsubjects. The mean rate of responding during thesecond VR subphase for nonobese boys was 114.2,whereas the obese boys' rate was somewhat slowerat 101.2 revolutions per minute. However, unlikeperformance during the first VR subphase, eachboy's response rate was initially slightly slower thanthe set criterion. Of course, falling below criterionstill produced substantial reinforcement on the VRschedule, due to variability in the ratios requiredfor reinforcement (i.e., some of the ratios fell below80 responses per minute).

There was an initial increase in performance forall boys upon implementation of the third VRsubphase. One nonobese and 2 obese boys per-formed at or above the criterion throughout thissubphase. The remaining subjects performed at ap-proximately the criterion rate. The mean responserate for the nonobese subjects was 130.0 revolutionsper minute, whereas the obese boys pedaled a meanof 117.0 revolutions per minute.

Implementation of the return-to-baseline phaseproduced a reduction in response rate for all boys.The mean response rate of 95.3 for nonobese boyswas higher than the obese boys' mean rate of 83.6revolutions per minute.

Reintroduction of the token economy producedsubstantial increases in all subjects' response rates.Both nonobese and obese subjects achieved theirhighest response rates during this subphase. Themean rate of responding during the final VR sub-phase was 138.7 revolutions per minute for non-obese boys, whereas the obese boys performed atthe mean rate of 123.6 revolutions per minute.During the final subphase of the program, 3 non-obese and 2 obese boys consistently performed wellabove the criterion, and upward trends were evi-dent. The response rate for the remaining obesechild, Perry, was relatively stable (M = 98.9)throughout the phase, and he performed slightlyabove the criterion during the last session.

Figure 2 shows the total time spent riding theexercise bicyde by all subjects. (Data points areabsent for Perry for eight sessions because he wasabsent from school.) It can be seen that duringbaseline there was a general decline over sessionsin the duration ofexercise for all subjects. Nonobeseboys exercised a mean of 15.2 min per session,whereas obese boys spent slightly less time pedaling(M = 12.9 min). The implementation of a rein-forcement program produced a substantial increasein the amount of time spent exercising by all sub-jects. Although 1 obese subject, Paul, did not ex-ercise for the entire allotted period of time duringthe first session, performance of all subjects stabi-lized at the maximum of 30 min. All boys exercisedthe full 30 min during all contingency phases. In-troduction ofthe return-to-baseline phase produceddecreases in the time spent exercising for all non-obese and obese boys. Nonobese boys exercised amean of 11.2 min per session during the return-to-baseline phase, whereas the obese children ped-aled a mean of 13.3 min per session. The reintro-duction ofreinforcement increased performance onceagain, and all subjects exercised- the maximum of30 min per session for the duration of the program.

Responses to the social validation questionnairefrom the children, parents, and teachers were uni-formly positive. With respect to satisfaction withthe treatment procedures, the modal rating for allgroups was 7 (the highest possible score on the7-point rating scale). Ratings of improvements inphysical activity and appearance were also extreme-ly high (mode = 7).

DISCUSSION

The results of the present study indicated thatthe rate of exercise, as measured by the speed ofbicyde pedaling, can be increased using a VR sched-ule of reinforcement. As expected, the introductionof the initial VR subphase of the changing-criteriondesign produced marked increases in the rate ofexercise for all subjects. The outcome is an extensionof De Luca and Holbom's (1990) findings thathigh rates of exercise were established during a ratio

675

RAYLEEN V. DE LUCA and STEPHEN W. HOLBORN

50

403025201510S

50

4030252015

10S

50

4030.

25201510

5

50

4030252015

105

501

4030252015

105

50

403025201510S

BASELINE VR 80 VR 115 VR 130 BL VR 130

4-1 II

,~~~~~~

BASELINE VR 85 VR 115 VR 12S B. VR 125

FBASELINE *VR 8S VR1llS VRI12S BL VR 125

;*

wS

BASELINE VR 80 VR 105 VR 120 BL VR 120

tBASELINE VR 70 VR 90 VR 110 III VR 110I *. I412 2 I L . a 2 1 1- II

_*S * . .

_, S'' 4' 4~

. .. ..4... . . . .. . .. . . . . . . . ........................BASELINE V 70 VR 95 VR 100 BL VR 100I-. 4 0_##8v_ I ,

* 5 0o 15 20 50 35 40

SESSIONS

Figure 2. Total number of minutes spent exercising during baseline, VR I (VR range, 70 to 85), VR 2 (VR range,

90 to 115), VR 3 (VR range, 100 to 130), return to baseline, and return to VR 3 phases for obese and nonobese subjects.

676

Ii

I

VR REINFORCEMENT SCHEDULE

(FR) schedule of reinforcement. It is noteworthythat the highest rates achieved under VR by bothobese and nonobese boys in the present study weregenerally greater than the most accelerated ratesunder the FR in the earlier De Luca and Holbom(1990) research. High rates of responding withinfrequent pausing are typically produced with hu-mans and animals on VR schedules (e.g., Lowe,1979; Orlando& Bijou, 1960). In addition, furthersupport is provided for the view that VR schedulescan attenuate past history differences (Baron & Ga-lizio, 1983; Kaufman, Baron, & Kopp, 1966). Inparticular, the obese subjects' histories of reducedrates of responding were eventually overcome bysuccessively increasing levels of the VR schedule.For 2 of the obese subjects (Peter and Paul), thefinal levels of exercise approximated those of thenonobese boys. Perry's final VR performance waslower, but he was absent for much of the training.

Another contributing factor to the higher ratesmay have been use of a changing-criterion design.In the present study, the rate of exercise generallyimproved in increments to match or exceed thecriterion specified. An exception to this occurredwith 1 obese and 1 nonobese boy who did notmeet the criterion during the third subphase. Inmany studies, not meeting the criterion would havenegated reinforcement (Kazdin, 1982); however,in the present study, because of the variable natureof the schedule it resulted only in reduced amountsof reinforcement. The 2 children's behavior cameunder control ofthe criterion in the following phase,and their exercise behavior accelerated as they at-tempted to "catch up" in points to their peers.

The results also corroborate De Luca and Hol-bom's (1985, 1990) findings that exercise durationcan be increased using behavioral principles of re-inforcement with children. The introduction of atoken economy produced increases in duration ofexercise to the maximum 30 min. Control wasevident in the children's duration of bicyde ped-aling. When the children received reinforcement forbiking, they continued for the maximum allottedtime; when they did not receive reinforcement, theirpedaling rapidly decreased. When reinforcement

was withdrawn and the bicyde's bell and light nolonger signaled points, the boys reduced their timespent pedaling and complained about no longerbeing able to collect points.

Turning to indices of social validity, the ques-tionnaire data for the children, their parents, andtheir teachers uniformly indicated maximum or doseto maximum ratings of satisfaction with both treat-ment procedures and outcome. Also, there wereother anecdotal indicators of consumer satisfactionthat are equally compelling. For example, in re-lation to satisfaction with treatment procedures, ona number of occasions boys in the program askedif their friends could participate. Many times whenthe 30 min were over, the children asked if theycould bicyde for a longer time to "get more points."Parents of all of the boys reported that their sonshad spoken positively about the program, with oneparent of an obese boy stating "that's all he talkedabout for weeks." Each of the obese boys convincedtheir parents to purchase a new bicyde for himduring the course of the study. The boys daimedthat although each had ridden bicydes in the past,they really had not cared much for biking.

In relation to satisfaction with treatment out-come, the convergence of a variety of anecdotalobservations proved persuasive of positive effect.Every subject participated in the school's track andfield events. The nonobese boys participated in run-ning events, and the obese boys participated in ballthrowing, long jumping, and high jumping. Twoof the obese boys received ribbons of achievementfor the first time in these events. Although all ofthe nonobese boys received ribbons at the track andfield meet, 2 of them (Shawn and Steve) went onto compete in the 200-m and 400-m runs in thecitywide track and field meet. Shawn won an awardfor his division. Shawn's mother reported that shefelt her son's participation in the biking programimproved his stamina, enabling him to compete ina larger meet for the first time.

Reduced girth measurements were most notice-able for Peter. He was excited that he could nowwear his stepfather's pants (Size 32) rather thanthe Size 36 he had previously wom. The vice-

677

678 RAYLEEN V. DE LUCA and STEPHEN W. HOLBORN

principal ofthe school commented that "Peter lookslike a new boy. He looks terrific." Peter's motherreported that when he went for a physical exami-nation prior to going to camp, the doctor compli-mented Peter on his improved appearance and fit-ness. Playing soccer and swimming became frequentactivities for Peter.

All parents were very supportive of the programand felt their sons had benefited from it. The par-ents of 1 obese boy stated that they had beeninitially hesitant about the program, but were "verythrilled" with the outcome and their son's "changedattitude" about participating in physical activity.Peter's mother asked if her younger son could alsobe involved in the program. Teachers stated thatthe program was very beneficial for all students.However, changes in physical activity were noticedparticularly for the obese boys, who they felt hadparticularly benefited from the program.A number of interesting areas remain for future

research. Primary among these should be the designof experiments to assess and program for general-ization of trained exercise behaviors to the naturalenvironment. Because there are more overweightgirls than boys (Manitoba Department of Educa-tion, 1978), it will be important to indude obesegirls, as well as boys, in future research. Althoughthe results of the present study are promising, fur-ther research on curtailment of obesity at an earlyage is crucial, if enduring life-style changes are toprevent obese youngsters from becoming obeseadults.

REFERENCES

Aristimuno, G. G., Foster, T. A., Voors, A. W., Srinivasan,S. R., & Berenson, G. S. (1984). Influence of persistentobesity in children on cardiovascular risk factors: TheBogglusa Heart Study. Circulation, 69, 895-904.

Baron, A., & Galizio, M. (1983). Instructional control ofhuman operant behavior. Psychological Record, 33, 495-520.

Brownell, K. D. (1984). The psychology and physiologyof obesity: Implication for screening and treatment.Jour-nal of the American Dietetic Association, 4, 406-414.

Bullen, B. A., Reed, R. B., & Mayer, J. (1964). Physicalactivity of obese and non-obese and adolescent girls ap-praised by motion picture sampling. American Journalof Clinical Nutrition, 14, 211-233.

Canada. Department of Health and Welfare. (1984). Stan-dardized tests offitness. Ottawa, Ontario: Fitness andAmateur Sport.

Cautela,J. R. (1977). Behavior analysisformsfor clinicalintervention. Champaign, IL: Research Press.

Cohen, E. A., Gelfand, D. M., Dodd, D. K., Jensen, J., &Turner, C. (1980). Self-control practices associated withweight loss maintenance in children and adolescents. Be-havior Therapy, 11, 26-37.

De Luca, R. V., & Holborn, S. W. (1985). Effects of afixed-interval schedule of token reinforcement on exercisewith obese and non-obese boys. Psychological Record,35, 525-533.

De Luca, R. V., & Holborn, S. W. (1990). Effects of fixed-interval and fixed-ratio schedules of token reinforcementon exercise with obese and non-obese boys. PsychologicalRecord, 40, 67-82.

Demirjian, A. (1980). Anthropometry report: Height,weight and body dimensions. A report from NutritionCanada. Ottawa, Ontario: Department of Health andWelfare.

Epstein, L. H., & Wing, R. R. (1987). Behavioral treat-ment ofchildhood obesity. Psychological Bulletin, 101,331-342.

Epstein, L. N., Wing, R. R., Koeske, R., & Valoski, A.(1987). Long-term effects of family-based treatment ofchildhood obesity. Journal of Consulting and ClinicalPsychology, 55, 91-95.

Hartmann, D. D., & Hall, R. V. (1976). The changingcriterion design. Journal ofApplied Behavior Analysis,9, 527-532.

Israel, A. C., & Stolmaker, L. (1980). Behavioral treatmentof obesity in children and adolescents. In M. Hersen, R.M. Eisler, & P. M. Miller (Eds.), Progress in behaviormodification (Vol. 10, pp. 82-109). New York: Aca-demic Press.

Johnson, M. L., Burke, B. S., & Mayer, J. (1956). Prev-alence and incidence of obesity in a cross-section of ele-mentary and secondary school children. American Jour-nal of Clinical Nutrition, 4, 231-238.

Kaufman, A., Baron, A., & Kopp, R. E. (1966). Someeffects of instructions on human operant behavior. Psy-chonomic Monograph Supplements, 1, 243-250.

Kazdin, A. E. (1982). Single-case research designs: Meth-odsfor clinical and applied settings. New York: OxfordUniversity Press.

Lowe, C. F. (1979). Determinants of human operant be-havior. In M. D. Zeiler & P. Harzem (Eds.), Advancesin analysis of behavior: Reinforcement and the orga-nization ofbehavior (Vol. 1, pp. 159-192). Chichester,England: Wiley.

Manitoba Department of Education. (1978). Manitobaschoolsphysicalfitness survey, 1976-1977. Winnipeg,Manitoba: Author.

Mann, G. V. (1974). The influence of obesity on health.New England Journal of Medicine, 291, 178-185.

Mayer, J. (1968). Overweight: Causes, cost and control.Englewood Cliffs, NJ: Prentice-Hall.

Mayer, J. (1970). Some aspects of the problem of regu-

VR REINFORCEMENT SCHEDULE 679

lating food intake and obesity. International PsychiatryClinics, 7, 225-234.

Mobbs, J. (1970). Childhood obesity. InternationalJour-nal of Nursing Studies, 7, 3-18.

Orlando, R., & Bijou, S. W. (1960). Single and multipleschedules of reinforcement in developmentally retardedchildren. Journal of the Experimental Analysis of Be-havior, 3, 339-348.

Rony, H. R. (1940). Obesity and leaness. Philadelphia:Lea and Febiger.

Seltzer, C. C., & Mayer, J. (1965). A simple criterion ofobesity. Postgraduate Medicine, 38, 101-107.

Stefanik, P. A., Heald, F. P., & Mayer, J. (1959). Caloricintake in relation to energy output of obese and non-obese adolescent boys. American Journal of ClinicalNutrition, 7, 55-62.

ReceivedJuly 9, 1990Initial editorial decision September 13, 1990Revisions received December 13, 1990; April 29, 1991;

July 21, 1991Final acceptance August 23, 1991Action Editor, Charles R. Greenwood