BEHAVIORAL MANAGEMENT OF EXERCISE: CONTRACTING FOR AEROBIC POINTS
JOURNAL OF APPLIED BEHAVIOR ANALYSIS
BEHAVIORAL MANAGEMENT OF EXERCISE: CONTRACTING FOR AEROBIC POINTS
TIM WYSOCKI, GENAE HALL, BRIAN IWATA,1 AND MARY RIORDAN
WESTERN MICHIGAN UNIVERSITY, THE JOHN F. KENNEDY INSTITUTE, AND THE JOHNS HOPKINS SCHOOL OF MEDICINE
Behavioral contracting was used to encourage physical exercise among college students in a multiple-baseline design. Subjects deposited items of personal value with the experi- menters, which they could earn back on fulfillment of two types of contract contingencies. Subjects selected weekly aerobic point criteria, which they could fulfill by exercising in the presence of other subjects. In addition, subjects contracted to observe and record the exercise of other subjects and to perform an independent reliability observation once each week, with both of these activities monitored by the experimenters. Results indicated that the contract contingencies produced increases in the number of aerobic points earned per week for seven of eight subjects, that the aerobic point system possesses several ad- vantages as a dependent variable for behavioral research on exercise, and that inexperi- enced observers could be quickly trained to observe exercise behavior and to translate those observations into their aerobic point equivalents. Finally, in a followup question- naire completed 12 months after the end of the study, seven of the eight subjects reported that they were earning more aerobic points per week than had been the case during the baseline condition of this experiment. DESCRIPTORS: aerobic points, behavioral medicine, contracting, exercise, adults
A prevalent behavioral characteristic of our population is a general lack of regular physical exercise. In an extensive review of data com- paring the fitness levels of American versus Western European children, Cureton (1964) concluded that American children are inferior along several basic dimensions of physical fit- ness. A continued lack of exercise into adult- hood has been suggested as a precipitating fac- tor in the etiology of a wide variety of diseases (Krauss and Rabb, 1961; Fox, 1973). Further, researchers have demonstrated that many clinical conditions improve substantially in response to exercise as a treatment component (Frick, Katila, and Sjorgen, 1971; Morehouse and Miller, 1976). Finally, several large-scale longitudinal studies have associated physically active life- styles with both less frequent and less severe episodes of cardiovascular pathology (Morris,
‘Reprints may be obtained from Brian A. Iwata, Division of Behavioral Psychology, The John F. Ken- nedy Institute, 707 North Broadway, Baltimore, Mary- land 21205.
Chave, Adam, Sirey, Epstein, and Sheehan, 1973; Paffenbarger and Hale, 1975).
Lack of exercise was viewed as a problem of sufficient magnitude to induce the federal government to devote substantial expenditures to investigate potential solutions (President’s Council on Youth Fitness, 1961; President’s Council on Physical Fitness, 1962, 1965). Rec- ommendations of these commissions took two general forms: the dissemination of more in- formation to the public concerning the value of exercise, and the provision of more exercise opportunities (e.g., organized recreation pro- grams). These approaches can be viewed as attempts to encourage exercise through the manipulation of antecedent or setting events, the assumption being that deferred conse- quences such as weight loss or improved health would be sufficient to maintain changes in exer- cise behavior. However, exercise, like many behaviors, is probably associated with reinforc- ing consequences on a long-term basis, but punishing consequences on an immediate basis.
55
1979, 12, 55-64 NUMBER I (SPRING 1979)
TIM WYSOCKI et al.
Moreover, traditional programs aimed at en- couraging exercise have not paid serious atten- tion to the fact that behavior is often more strongly controlled by immediate rather than delayed consequences. Thus, one possible so- lution to the problem of lack of exercise would be to arrange contingencies such that reinforcers are delivered more closely in time following the exercise behavior than is usually the case.
Several recent studies suggest that such an approach would be successful. Libb and Clem- ents (1969) used a token reinforcement proce- dure to increase the rate of exercise on a sta- tionary bicycle by geriatric patients. Other studies have demonstrated that contingency management approaches can be effective in modifying aspects of various athletic perform- ances: swimming (McKenzie, Note 1; McKen- zie and Rushall, 1974; Rushall and Pettinger, 1969), basketball (Jones, Note 2), football (Komaki and Barnett, 1977), and baseball (Heward, 1978). However, although results from these studies are highly promising for the development of procedures for modifying exer- cise behavior in specific settings, each of the studies possesses limitations that may render their techniques somewhat unattractive as meth- ods for encouraging exercise in the natural en- vironment. First, each study involved direct observation of exercise by the experimenter. Implementation of this procedure on a large scale may be extremely costly and would place excessive demands on the trainer’s time and resources. Second, all of these studies required the subjects to engage in a single form of physi- cal activity. However, changes in exercise be- havior may be more likely to persist in the natural environment if subjects could engage in those activities possessing highest reinforcement value to a given individual.
The purpose of the present study was to provide a preliminary evaluation of a procedure that might prove effective in encouraging exer- cise in the natural environment. Three specific problems were addressed in an attempt to ac- complish this goal: (1) the derivation of opera-
tional definitions of exercise behavior, (2) the development of an observation system that provides maximal assurance that exercise has occurred, yet places minimal demands on the experimenter, and (3) the selection of a con- tingency likely to produce changes in exercise behavior.
Cooper (1970) described a category of ac- tivities called aerobic exercises, consisting of moderate, sustained activity that produced ele- vated levels of oxygen consumption over ex- tended periods of time. He studied the re- sponses of some 15,000 Air Force cadets to programs of aerobic conditioning. In that re- search, he measured the amount of oxygen consumed by the body during various combi- nations of intensity and duration for each of the aerobic exercises (running, walking, swim- ming, handball, cycling, stationary running, rope skipping, basketball, racquetball, squash, golf, badminton, fencing, football, hockey, soccer, skating, skiing, tennis, volleyball, and wrestling) and reduced the oxygen consumption data to a single quantitative dimension, which he termed the aerobic point. Each aerobic point is equal to oxygen consumption at the rate of 7.0 ml/kg/min. Thus, in its simplest terms, the aerobic point is a measure of how much oxygen the body consumes during a specific activity. Table 1 presents the aerobic point values for a number of common exercises at various combinations of intensity and duration. Note, for example, that running one mile in 13 min is worth two aerobic points, while covering the same distance in 9 min is worth four points.
In addition to Cooper’s research, others have provided evidence suggesting a functional re- lationship between the amount of oxygen con- sumed during exercise and the cardiovascular benefit of that exercise (Daniels and Oldridge, 1977; Pollock, Cureton, and Greninger, 1969). Several exercise physiologists (Morehouse and Miller, 1976; Sharkey, 1975) have further con- cluded that aerobic capacity is the best single measure of physical fitness currently available.
56
CONTRACTING AND EXERCISE
In view of its demonstrated validity and ability to reduce a wide variety of exercises into a common dimension, the aerobic point was se- lected as the dependent variable in the present research.
Several strategies are available in developing an observation system that ensures that exercise has occurred. Although personal observation of all exercise by the experimenter is likely to yield reliable data, it limits the subject’s choices as to how and when to exercise. An al- ternative procedure might be to recruit media- tors in the subject’s natural environment to assist in performing observations (Tharp and Wetzel, 1969), and a likely pool of such media- tors would be other subjects. This approach would allow the experimenter to distribute resources across a larger number of subjects and might also be expected to enhance maintenance of behavior change. Thus, subjects in the pres- ent study were required to observe and record the exercise behavior of other subjects and to perform occasional checks, with both activities being monitored by the experimenters. With respect to contingencies, behavioral
contracting appeared attractive for the purposes of the present research (Tighe and Elliot, 1968). The procedure involves negotiation of a con- tract between a client and therapeutic agent. The client deposits various items of worth (e.g., money, goods) with the agent; the agent returns the items contingent on contract ful- fillment (behavior change) on the part of the client. Behavior contracting has been effective in controlling numerous behaviors in the natu- ral environment: body weight (Mann, 1972), smoking (Tighe and Elliot, 1968), studying (Homme, 1969), alcoholic drinking (Miller, 1972), and delinquency (Stuart, 1971). Several recent studies have evaluated the use of be- havioral contracting in encouraging physical exercise (Turner, Pooly, and Sherman, 1976; Vance, 1976). However, in both of these studies no attempt was made to monitor directly the frequency or intensity of the subjects’ exercise behavior. Instead, the experimenters relied on
the subjects’ verbal reports of…
