Maintaining attendance at a fitness center: an application of the decision balance sheet

Maintaining attendance at a fitness center: an application of the decision balance sheet

Claudio R. Nigg

The effect of a decision balance sheet intervention on attendance at a university-fitness facility was examined. Facility members were randomly assigned to control, placebo, and experimental conditions. The control condition received no intervention, whereas the placebo and experimental conditions were called by telephone and asked to complete either an irrelevant (smoking) or relevant (exercising at the fitness facility) decision balance sheet. Attendance was monitored surreptitiously for 4 weeks baseline and 8 weeks post intervention. Statistical analyses indicated that the control and placebo conditions showed a significant decrease in attendance from baseline to intervention, whereas those in the experimental condition maintained attendance levels. Discussion focused on broadening the application of the decision balance sheet, determining its theoretical boundaries, and the necessity and appropriateness of decision alternatives for the decision balance sheet in the exercise domain.

Index Terms: decision balance sheet, decision-making theory, exercise adherence, fitness center, intervention

Promoting regular exercise has become an important priority for public health interventions because of the documented physiological and psychological health benefits of exercise.[1] Despite the known benefits of regular exercise, overall participation rates are low. Only about 10% of the North American population exercises regularly.[2] Moreover, those who initiate an exercise program have trouble maintaining it. Dishman[3] estimated that approximately 50% of individuals who begin a structured exercise program drop out within the first 6 months. Results have been similar for children, college students, middle-aged and elderly persons, and in primary and secondary prevention and worksite settings.[4]

Given the documented health benefits and low participation and adherence rates associated with exercise, we believe that interventions to increase and maintain regular exercise must be crafted.[5] Various interventions have been applied to exercise behavior, including reinforcement, stimulus control, behavioral contracting, self-monitoring, health-risk appraisal, and goal setting.[6] The results have been generally positive, although modest, with standardized effect sizes (rs) typically in the .15 to .20 range.[7] One promising intervention for the exercise domain is the decision balance sheet (DBS) developed by Janis and Mann.[8]

The DBS is based on expectancy theory, which maintains that a person’s course of action changes as a function of increases or decreases in the relative strength of the anticipated gains and losses of taking the action. The more relevant the information considered before making a decision, the stronger the commitment to that decision and the better the long-term adherence to the decision.[8] A decision is usually made on the relative weights of the gains and losses associated with each decision alternative. The DBS is especially valuable for analyzing the degree to which a decision maker does a thorough and accurate job of exploring the full range of available alternatives and considering the consequences of each alternative. Decisional conflicts are thought to be overcome by systematically analyzing the information gathered on the gains and losses of each alternative and identifying the salient factors upon which the decision is to be made.[8]

Janis and Mann[8] identified two processes that may mediate the success or failure of the DBS. One process, referred to as self-persuasion, occurs when an individual systematically considers the positive outcomes of a decision alternative and becomes persuaded by these potential gains. This process serves to highlight and solidify the anticipated positive benefits of the decision alternative. A second process, termed emotional inoculation, occurs when an individual systematically considers the negative outcomes of a decision alternative. The person becomes emotionally inoculated against these negative outcomes and is better able to tolerate them because they were previously anticipated. This process is especially important for decisions that require short-term losses to realize long-term gains.

The DBS itself involves six steps consisting of (a) an open-ended interview that elicits the most salient alternatives and their consequences; (b) introduction of the DBS grid to stimulate thought concerning nonsalient considerations; (c) provision of a list of pertinent considerations to explore neglected pros and cons; (d) identification, through ranking, of the most important considerations; (e) exploration of alternatives not elicited in the initial interview; and (f) a second ranking of alternatives to encourage comprehensive consideration of all consequences.[8]

The pros and cons of each alternative correspond to a multidimensional set of values that make up the DBS grid. The DBS grid requires the person to write down anticipated outcomes in four broad categories under the headings of (a) gains and losses to self, (b) gains and losses to important others, (c) self-approval and disapproval, and (d) approval and disapproval of important others. The first category–gains and losses to self–covers all instrumental effects expected to result from the decision with regard to personal utilitarian objectives. The second or losses to important others–targets individuals or groups with whom the individual is affiliated. The third category–self-approval or disapproval–includes the individual’s basic moral and value systems. The final category–approval or disapproval of important others–targets the basic moral and value systems of individuals or groups with whom the individual is affiliated.

Two errors, referred to as errors of commission and errors of omission, may occur in completing the DBS.[8] Errors of commission may occur if the person is overoptimistic about the expected gains of a decision and includes gains that cannot be reasonably expected. Errors of omission may occur when obvious and knowable losses are overlooked. These errors affect the completeness and accuracy of the DBS and may undermine the stability of a decision.[8]

Despite its intuitive appeal and strong theoretical basis, the DBS as an intervention tool in the exercise domain has been examined only three times.[9-11] All three studies have administered the DBS by telephone and have found promising results. Hoyt and Janis[9] examined the effect of a DBS intervention on attendance at an early-morning university exercise class that met three times per week for 7 weeks. The sample comprised 50 women (mostly faculty wives) who were randomly assigned to control, relevant (de, exercise), or irrelevant (de, smoking) DBS conditions. Results indicated a main effect for balance-sheet condition, with the relevant DBS group exhibiting more than twice the attendance rate for the 21 classes (11.9 times) of the irrelevant DBS (5.8 times) or control (5.6 times) groups.

The only researcher to follow-up on these results in the exercise domain was Wankel.[10, 11] Wankel and Thompson[10] examined the effect of a DBS intervention on attendance at a private fitness club consisting of 100 female members who had not attended the club in the previous month. The 100 participants were randomly assigned to (a) a control condition, (b) a standard telephone-call condition, (c) a full DBS condition (de, both pros and cons), or (d) a positive-only DBS condition. Attendance was monitored for I month after the intervention. Wankel and Thompson found that the women in the full DBS and the positive-only DBS conditions had significantly higher attendance rates than those in the control condition. The positive-only DBS group also had significantly higher attendance than those in the standard telephone-call condition. However, overall attendance of individuals in the sample in the 1-month study period was very low, on average: the positive-only DBS participants attended the fitness club 1.5 times; the full DBS group, 1.3 times; those who received the standard telephone call, 0.7 times; and the control group 0.2 times.

Finally, Wankel et al[11] investigated the effects of self-motivation and the DBS procedure on attendance at a community-based introductory fitness program that met once per week for 3 weeks. Fifty-two women were paired according to their level of self-motivation and were randomly assigned to DBS and control conditions. The findings indicated a significant main effect, with the group in the DBS condition exhibiting higher attendance (2.1 times) than the control group (1.5 times).

In summary, the limited research using a DBS intervention in the exercise domain has shown promising results. However, further research is necessary to replicate and extend these findings in terms of context, sample size, diversity, and length of postintervention monitoring. Previous research has examined female participants exclusively, used total sample sizes of fewer than 100, and monitored exercise following the DBS intervention for only 3, 4, and 7 weeks. Also needed is research to provide further evidence of the relative importance of the processes proposed to mediate the success of the DBS procedure.

Reports from previous research[9, 10] have suggested that the most likely process mediating the success of the DBS intervention is self-persuasion because the DBS had an immediate effect on exercise behavior,9 and a positive-only DBS had a stronger influence on exercise adoption than a balanced DBS.[10]

We examined the effectiveness of a DBS intervention on attendance at a university fitness center, using 153 male and female fitness club members who were randomly assigned to relevant DBS (de, exercise), irrelevant DBS (de, smoking), and control conditions; we followed the participants for 8 weeks after the intervention. On the basis of expectancy theory and previous DBS research, we hypothesized that only the relevant DBS condition would result in a positive effect on attendance at the fitness center.

Our secondary purpose was to examine the importance of the number of pros and cons listed by each individual in the relevant DBS condition. An implied hypothesis in the DBS procedure is that the greater the differential between pros and cons, the greater the commitment and adherence to the decision. This issue has not been addressed in previous exercise research, and we expected to provide evidence about the tenability of the two mediating processes of self- persuasion (from listing pros) and emotional inoculation (from listing cons). We also hypothesized, based on previous DBS research, that the number of pros would have a significant positive correlation with attendance, whereas no such relationship would be found for the number of cons.

METHOD

Participants

Participants in the study were 153 paying members of a university fitness club drawn from four categories of membership: alumni, support staff, academic staff, and the general public. Ages ranged from 16 to 66 years (M = 36.33; SD = 11.00). Seventy-seven percent of the participants were men.

Instruments

We used the mainframe computer from the recreation department to measure attendance. Each time a participant used the fitness facility, a staff member passed a computer wand over the person’s membership card, and the computer recorded the visit. We calculated pros and cons in terms of the number of pros or cons that the individual listed on the DBS and calculated the pro-con difference by subtracting the number of cons from the number of pros.

Procedure

The study was a fully randomized design consisting of control, placebo (de, irrelevant DBS), and experimental (de, relevant DBS) conditions. We selected individuals sequentially from the previously mentioned alphabetically listed membership categories and surreptitiously gathered 4 weeks of baseline attendance records before randomly assigning names to conditions. Our procedure was to assign names to the control, placebo, and experimental conditions alternately until each group consisted of 51 individuals.

The 51 participants who were randomly assigned to the control condition received no intervention. We telephoned the remaining persons and placed them alternately in the relevant (ie, exercise) and irrelevant (de, smoking) DBS conditions. We contacted 110 members over 3 consecutive days to achieve the 102 participants in the two conditions (8 fitness club members refused to participate, resulting in a 93% participation rate). The date on which we made the telephone call marked the end of the baseline period and the beginning of the intervention period.

The participants in the DBS conditions received a telephone call during which they were asked to get a pen and paper and to think systematically of and record the expected gains and losses of either exercising at the fitness center (relevant) or not smoking (irrelevant). The headings we gave to the participants were gains and losses to self, self-approval and disapproval, approval and disapproval from others, and other items. When the participants had completed the list, we asked them to think about their items and then read the list aloud to the interviewer. The interviewer recorded the pros and cons for each participant in each DBS condition. We then surreptitiously monitored attendance of all participants for 8 weeks.

RESULTS

The pattern of attendance for each condition across the 4 baseline and 8 intervention weeks is shown in Figure 1. Attendance generally declined over time, except for week 8 of the intervention, when we saw a noticeable increase. Week 8 coincided with the end of exams and that may have increased the time available for exercise for persons in our sample, which included both faculty and staff members. For analysis, we collapsed the attendance data into baseline (average of the 4 weeks of baseline attendance) and intervention (average of the total 8 weeks of intervention attendance). Internal consistencies for baseline and intervention attendance for successive weeks were in the .80s and .90s across the three conditions. For the means and standard deviations for baseline and intervention attendance across the three conditions, see Table 1. Pros, cons, and pro-con difference are also provided for the relevant DBS condition.

[Figure 1. ILLUSTRATION OMITTED]

TABLE 1 Means and Standard Deviations for Attendance Across Each Condition and for Pros, Cons, and Pro-Con Difference in the Relevant DBS Condition

Control Irrelevant DBS Relevant DBS

Condition M SD M SD M SD

Baseline

(4 weeks) 1.34 1.29 1.88 1.43 1.45 1.24

Intervention

(8 weeks) 1.11 1.16 1.42 1.30 1.41 1.25

Pros 4.51 1.69

Cons 1.88 1.62

Pro-con

difference 2.63 1.96

Future exercise research should examine the DBS in diverse contexts and across varied populations and should test the limits of its postintervention effectiveness. One important context is home-based exercise programs. Researchers have examined exercise only at fitness facilities. In terms of populations sampled, the research has been limited to normal, healthy adults. The DBS should be extended to other populations, including those using exercise as rehabilitation therapy (eg, cardiac or cancer patients). For those individuals, the decision to exercise is critical, and the DBS may be an effective tool to increase their commitment to the decision to exercise. Future research might also examine long-term exercise programs and determine if and when multiple administrations of the DBS would be effective in reducing dropouts and promoting program maintenance.

One interesting finding of our study was that the DBS intervention was effective in maintaining current exercise patterns, whereas the control and placebo groups experienced a significant decline in attendance over the 8-week period. This finding is in contrast to that of Hoyt and Janis,[9] who found that a DBS intervention resulted in an initial increase in attendance at fitness class, but that the rate of decline in attendance was consistent across the experimental, placebo, and control groups.

A perfunctory explanation is the obvious differences in contexts and samples. A more insightful explanation, perhaps, is the timing of the interventions. The Hoyt and Janis[9] intervention was conducted near the beginning of a structured fitness class when attendance patterns were being established and presumably were open to influence from an early intervention.

The DBS intervention in our current study was delivered to members of an ongoing fitness club who had probably established an attendance pattern and were undergoing a decline in attendance. It seems logical that a DBS intervention at this time would help prevent the decline in attendance but would not necessarily increase attendance. A distinction between using the DBS in the adoption versus maintenance of exercise has been noted earlier.[10] Future DBS research should consider the timing of the intervention in determining its expected results.

One framework that may be helpful in examining the timing of a DBS intervention is Prochaska’s transtheoretical model (TM).[12] The TM identifies five stages of change (ie, precontemplation, contemplation, preparation, action, and maintenance) and integrates them with the constructs of pros and cons from the Janis and Mann decision-making theory.[8] Some evidence exists to suggest that the perceived pros and cons of a behavior are very important to transition through the early stages. Cons always outweigh the pros during the precontemplation stage, whereas the opposite is true in the action and maintenance stages.[13] The crossover of pros and cons takes place in either the contemplation or preparation stage, depending on individual behavior. The effectiveness of a DBS intervention, therefore, is likely to depend on a person’s stage of change.

It seems clear that a DBS intervention will be effective only for persons whose pros for regular exercise outweigh their cons. That criterion was met in previous research and among the participants in the present study because one can assume that fitness center members would be in the preparation stage at least (only 5 persons had a negative pro con difference index). In terms of the TM, it is likely that a DBS intervention will not be effective for precontemplators and may actually have a reverse effect. A DBS intervention might be helpful, however, for those contemplating a decision to exercise, those preparing to increase their exercise, and those in the action and maintenance stages in terms of exercise. An examination of the effectiveness of the DBS intervention across the stages of change would be an interesting future study.

The pro-con results of our present study provide information to evaluate the relative importance of the two processes proposed to mediate the success of a DBS intervention –self-persuasion and emotional inoculation. Previous research[9, 10] has suggested that the most likely process mediating the success of the DBS intervention in the exercise domain is self-persuasion, wherein the person is persuaded by the number of pros that the behavior has positive benefits and should be performed.

The self-persuasion hypothesis was endorsed in previous studies because the DBS intervention had an immediate effect on exercise behavior,[9] and a positive-only DBS condition had a stronger influence on exercise adoption than a balanced-DBS condition.[10] The research we have described in this article appears to support both the self-persuasion and emotional inoculation hypotheses. The number of pros listed by the relevant DBS group had a significant positive correlation with attendance during the 1st post-intervention week, indicating that completing the DBS might have had an immediate self-persuasion effect. It was not until the 5th and 6th weeks after the intervention, however, that the number of cons listed by this group had a significant positive correlation with attendance.

A positive correlation between the number of cons and attendance is what would be expected, according to the emotional inoculation hypothesis. The more cons that are listed, the more the negatives of the decision are anticipated and the better the adherence to the decision. The delayed effect would also be expected in that, as the length of time engaged in the behavior increases, so does the number of negatives encountered.

The relative independence of the self-persuasion and emotional inoculation hypotheses also seems supported in the present study because the pro-con difference measure did not correlate with attendance at any week post intervention. This finding seems to be an indication that people consider either the pros or cons of a behavior following a decision to perform, but do not consider both pros and cons at the same time. It is important to note that change scores tend to be less reliable, and this may also explain the lack of relationship between the pro-con index and attendance. These results are modest and preliminary, however, and future researchers examining the issue might consider using a more complex weighting scheme for the pros and cons in contrast to the simple counting procedure we used. Janis and Mann[8] do not give any guidelines on the quantification of the pros and cons in a DBS intervention, so it is not clear if or how they should be weighted and combined.

In the present study, the magnitude of the effect of the DBS intervention was small to moderate, which is consistent with previous DBS research and exercise-intervention research in general.[7] A modest effect size seems acceptable, however, given the need for experimental control and standardized interventions, which necessarily reduce the intensity of the treatment intervention.[11] Participants completed the DBS during a telephone call of only 10 to 15 minutes. Its minimal requirements make the DBS a simple and cost-effective procedure for initiating and maintaining exercise behavior in a wide variety of situations. Future research might examine a more intense DBS treatment by using a longer interview involving face-to-face contact that is also administered at many points in an exercise program. This approach should be compared with a single brief telephone administration of the DBS, which has been the method of choice in all previous studies.

Another important issue is the presence and appropriateness of alternatives considered in a DBS intervention. Previous exercise research, including our present study, has asked participants to consider the pros and cons of only one alternative–exercising. By definition, at least two alternatives exist in any decision. The most common alternative is not doing the behavior (eg, not exercising or remaining sedentary), and this approach has been recommended for the exercise domain.[14] A more precise approach may be to have the person consider the pros and cons of the specific behavior against which exercise is competing. For example, Hoyt and Janis[9] examined early-morning exercise classes. At that time of day, it is likely that the alternative behavior is staying in bed. A thorough assessment of alternatives in this example would be to compare the pros and cons of sleeping in with the pros and cons of exercising. This approach would provide a clearer and more direct contrast among the choices available to the individual and should result in a more informed and committed decision.

Despite the important contributions of this study, we recognize that a number of limitations should be taken into account in interpreting the results and planning future research. One limitation is the short time of the intervention (8 weeks), despite its being the longest period to date for DBS exercise research. Studies of longer duration are needed to determine the limits of DBS’s effectiveness for exercise.

A second limitation is the exclusive focus on frequency (attendance) of exercise, with no attempt to assess the effects of the intervention on the intensity and duration of exercise that are necessary components for improving fitness and health. A third limitation is that we made no attempt to assess changes in fitness or health, although changes are the ultimate outcomes researchers pursue in studies of exercise adherence.

Our study has extended the effectiveness of the DBS by showing that it maintains exercise behavior in a new context, within a large mixed-sex sample, and for an extended period of postintervention monitoring. Future exercise research should continue to examine the DBS in diverse contexts, across varied populations, and should test the limits of its effectiveness. Suggestions for future research include an examination of the effectiveness of the DBS intervention across the stages of change; consideration of a more complex weighting scheme for the pros and cons; administration of a more intense DBS treatment, using a face-to-face interview; and including appropriate alternatives to the decision to exercise.

NOTE

For further information, please address correspondence and reprint requests to Kerry S. Courneya, PhD, Faculty of Physical Education, University of Alberta, E401 Van Vliet Center, Edmonton, Alberta T66 2H9 CANADA.

REFERENCES

[1.] Bouchard C, Shepard RJ, Stephens T. Physical Activity, Fitness, and Health: Consensus Statement. Champaign, IL: Human Kinetics; 1993.

[2.] Stephens T, Caspersen LJ. The demography of physical activity. In: Bouchard C, Shepard RJ, Stephens T, eds. Physical Activity, Fitness, and Health: Consensus Statement. Champaign, IL: Human Kinetics; 1993:203-213.

[3.] Dishman RK. Exercise adherence research: Future directions. Am J Health Promo. 1988;3:52-56.

[4.] Robison JI, Rogers MA. Adherence to exercise programs: Recommendations. Sports Med. 1994;17:39-52.

[5.] Dishman RK, Sallis JF. Determinants and interventions for physical activity and exercise. In: Bouchard C, Shephard RJ, Stephens T, eds. Physical Activity, Fitness, and Health: Consensus Statement. Champaign, IL: Human Kinetics; 1993: 214-238.

[6.] Dishman RK. Increasing and maintaining exercise and physical activity. Behav Ther. 1991;22:345-378.

[7.] Dishman RK. Introduction: Consensus, problems, and prospects. In: Dishman RK, ed. Advances in Exercise Adherence. Champaign, IL: Human Kinetics; 1994.

[8.] Janis IL, Mann L. The decision balance sheet. In: Janis IL, ed. Decision Making: A Psychological Analysis of Conflict, Choice, and Commitment. New York: Macmillan; 1977:135-169.

[9.] Hoyt MF, Janis IL. Increasing adherence to a stressful decision via a motivational balance sheet procedure: A field experiment. J Pers Soc Psychol. 1972;5:833-839.

[10.] Wankel LM, Thompson C. Motivating people to be physically active: Self persuasion vs. balanced decision making. J Appl Soc Psychol. 1977;7(4):332-340.

[11.] Wankel LM, Yardley JK, Graham J. The effects of motivational interventions upon the exercise adherence of high and low self-motivated adults. Can J Appl Sport Sci. 1985;10(3):147-156.

[12.] Prochaska JO, Marcus BH. The transtheoretical model: Applications to exercise. In: Dishman RK, ed. Advances in Exercise Adherence. Champaign, IL: Human Kinetics; 1994: 161-180.

[13.] Prochaska JO, Velicer WF, Rossi JS, et al. Stages of change and decisional balance for twelve problem behaviors. Health Psychol. 1994;13:39-46.

[14.] Knapp DN. Behavioral management techniques and exercise promotion. In: Dishman RK, ed. Exercise Adherence: Its Impact on Public Health. Champaign, IL: Human Kinetics; 1988: 203-237.

At the time this research was done, Mr Nigg, Dr Courneya, and Mr Estabrooks were with the Faculty of Kinesiology at the University of Calgary in Alberta, Canada. Dr Courneya is now with the Faculty of Physical Education at the University of Alberta in Edmonton.

COPYRIGHT 1997 Heldref Publications

COPYRIGHT 2004 Gale Group