- Email: [email protected]
The effect of target behavior monitoring on weight loss and completion rate in a behavior modification program for weight reduction
AddictiveBehaviors,Vol. 11, pp. 337-340, 1986
0306-4603186 $3.00 + .GU Copyright e 1986 Pergamon Journals Ltd
Printed in the USA. All rights reserved.
BRIEF REPORT THE EFFECT OF TARGET BEHAVIOR MONITORING ON WEIGHT LOSS AND COMPLETION RATE IN A BEHAVIOR MODIFICATION PROGRAM FOR WEIGHT REDUCTION WILLIAM A. SPERDUTO Hofstra University
HARLEY S. THOMPSON Hicksville Public Schools, New York
RICHARD M. O’BRIEN B-MOD Associates Huntington,
New York
Abstract-Although behavior modification of obesity is usually descibed as a behavior change procedure, measurement is most often limited to the outcome variable of weight loss. The present investigation employed detailed behavior monitoring forms in half of 16 obesity groups (n = 173) matched across four different therapists. The percentage of compliance for nine specific treatment behaviors was charted from these monitoring forms. At the end of treatment, the eight groups that had the behavior monitoring treatment averaged over 6 pounds more weight lost than the matched control groups as well as showing significantly better attendance. The between group weight loss differences were still evident 3 months after treatment.
The major assumption in behavioral treatments of obesity is that being overweight is due to excess food consumption resulting from faulty eating habits (Wooley, Wooley, & Dyrenforth, 1979). Unfortunately, a failure to find strong relationships between selfreported behavior change and weight change despite significant changes in eating habits (Adams, Ferguson, Stunkard, & Agras, 1978; Krassner, Brownell, & Stunkard, 1979) has been noted in the literature. Two possible explanations have been advanced to account for such a discrepancy: changes in eating behaviors may not be linearly related to changes in weight, or positive relationships may have been obscurred by errors and biases in retrospective self-reports (Abramson, 1983; Jeffrey, Vender, & Wing, 1978). Attempts to directly work with eating behaviors in individual programs (Coates & Thoresen, 1981; Straw dz Terre, 1983) have produced more encouraging results; however, the maintenance of weight loss in individual programs has generally been less consistent than in group programs (Kingsley & Wilson, 1977; Straw & Terre, 1981). It would seem that further research is needed to systematically assess the effects of self-monitoring specific changes in behavior on weight loss in a group treatment program. Feedback through self-monitoring and charting has been shown to be an effective performance improvement technique in industry (Crawley, Adler, O’Brien, & Duffy, 1982), mental health settings (Kreitner, Reif, & Morris, 1977), and residential applicaPortions of this paper were presented at the annual meeting of the Association for the Advancement of Behavior Therapy, Toronto, November, 1981. This research was conducted at the Hofstra Center for Fitness and Health, Old Westbury, NY. Reprint requests should be addressed to Dr. Richard M. O’Brien, B-MOD Associates, 356 New York Avenue, Huntington, NY 11743. 337
338
WILLIAM A. SPERDUTO, HENRY S. THOMPSON and RICHARD M. O'BRIEN
tions (Jason & Klich, 1982; McDonald & Budd, 1983). Gilbert (1978) has emphasized that when one is trying to increase the performance of well-learned skills, it is best to monitor outcomes or results which in the case of obesity would be weight loss. But he has also suggested that monitoring behavior is most effective when new skills are being learned. Despite the emphasis on learning new eating and exercise behaviors in behavioral obesity programs, the measurement has focused almost exclusively on the results (weight loss). The present study was designed to examine the effect of actually monitoring and charting targeted eating and exercise behaviors on drop-out rate and number of pounds lost in weight reduction groups utilizing behavior modification. Based on the available literature, it was hypothesized that the weight loss of individuals utilizing behavior monitoring forms would be significantly greater than the weight loss of individuals treated without such forms. In addition, this difference should be maintained at 3-month follow-up. It was also hypothesized that the completion rate in groups utilizing the monitoring forms would be significantly greater than the completion rate in groups treated without such forms. METHOD
The study involved 16 groups of adult subjects (n = 173, M = 43, F = 130) who were treated at a fee-for-service ($185 per person) clinic for obesity. In addition to the program fee, attendance deposits ($25) were collected from all subjects entering the program to be refunded upon completion of at least 12 treatment sessions. Previous research in obesity and smoking (Hagen, Foreyt, & Durham, 1976; Sperduto & O'Brien, 1983) has shown refundable deposits to be effective in maintaining a low attrition rate. Recruitment of subjects was primarily through newspaper advertisement directed at the same geographic location; therefore, the demographic characteristics of subjects (sex, race, SES, age) remained essentially similar throughout the study. Treatment was conducted in a group format (approximately 11 clients per group) and consisted of 15 weekly 1-hour meetings. Ferguson's book Learning to Eat (1975) was the manual for all groups. One half of the groups were given additional target behavior monitoring forms which listed a total of nine behaviors related to eating and activity levels. These forms were adapted from Ferguson's daily behavior checklist (1975) with certain modifications including conversion from a Likert-type scale to a frequency count. The number of times a desired eating behavior was engaged in was compared to the total eating episodes each day. This formed a ratio which the participant transformed each week into a percentage of compliance. The participant then recorded this percentage on an on-going graph which was provided for each behavior. Thus, the client received weekly visual feedback on their performance on each of the required new behaviors. The therapist then praised their effort and set goals for the next week. Similar to other obesity programs (Craighead, Stunkard, & O'Brien, 1981), individual weights were recorded on a weekly basis. Four different therapists had pairs of their groups randomly assigned to experimental or control treatments so that all therapists ran groups in both conditions. The therapists were advanced graduate students (3 males, 1 female) in a doctoral psychology program who had been specifically trained in behavior modification for weight control. Telephone contacts were made 3 months following treatment in order to gather maintenance data. This was deemed appropriate since the literature has indicated that self-reports of weight are extremely accurate (Schacter, 1982; Stunkard & Albaum, 1981).
Behavior monitoring for weight reduction
339
RESULTS
Weight loss Combining all eight groups that utilized target behavior monitoring forms as part of the behavior modification program showed a mean weight loss of 15.90 pounds (SD = 3.86) at the end of treatment compared to a mean loss of 9.67 pounds (SD = 1.9) for the eight nonbehavior monitoring groups. The difference of 6.2 pounds at the end of treatment was significant on a one-tailed test (t14 = 4.05, p < .005). At 3 month follow-up, this difference (6.4 lbs.) was maintained (t14 = 3.72, p < .005). The behavior monitoring effects might be best illustrated by a measure of association, such as the point biserial correlation. The association between treatment and outcome for weight loss at the end of treatment was rpb = .74 which means that 55% of the variance was explained by the monitoring forms. Similarly, the association between treatment and outcome at the 3 month follow-up was r,b = .71 demonstrating that 50% of the variance was accounted for by the monitoring forms. According to Cohen and Cohen (1975) both correlations represent a large effect size. Completion rate Those groups utilizing target behavior forms had a completion rate of 74.4% as compared to 56.8% for the nonbehavior monitoring groups. A one-tailed test revealed that this difference was statistically significant (t~, = 2.24, p < .05). The association between treatment and completion rate was rpb = .51 indicating that 26% of the variance was explained by the monitoring forms. Once again this represents a large effect size (Cohen & Cohen, 1975) produced by the monitoring forms. DISCUSSION The present results suggest that the problems observed in the behavior modification treatment of obesity may be due to the fact that people have been treating obesity and not eating behaviors. Although behavioral weight loss programs are phrased in terms of changes in performance (eating behaviors), feedback has traditionally been limited to outcome (weight change) and global self-reports on the practice of new behaviors. In any instance of learning, however, it is necessary for the organism to monitor its performance in order to self-correct. In the present study, the subjects systematically selfmonitored and charted specific eating behaviors while the therapists shaped compliance. This performance was reinforced in relation to increasingly difficult behavioral goals through praise and visual performance feedback. For example, 20O7o compliance with a new eating behavior, such as putting the fork down between bites or leaving food on the plate may have been acceptable during week 4 but by week 14, 80O7o was the standard. It should be noted that, as in any shaping procedure, these goals were individualized for each participant. Gilbert (1978) has criticized behavior modifiers for ignoring outcome variables, in this case weight loss. However, in learning new behaviors, even Gilbert recognizes the need to monitor behavior directly. It would be absurd to run an obesity program without measuring weight loss, but it is also crucial to monitor eating behaviors for learning to take place. The participants must receive adequate on-going feedback on their performance and the therapist must have accurate measures of eating behaviors, not just weight loss, so he or she can contingently reinforce the participant's performance and identify problem areas. Utilizing self-monitoring and charting allows the participant to self-reinforce and set realistic goals while providing the therapist with the opportunity to directly reinforce behavior change without having to wait for weight loss. Actually,
340
WILLIAM A. SPERDUTO, HENRY S. THOMPSON and RICHARD M. O'BRIEN
monitoring behavior avoids the possibility of inadvertently reinforcing potentially dangerous behaviors (fasts, crash diets, and/or taking protein supplements) to achieve weekly weight loss goals. Such monitoring also allows the therapist to reinforce all members of the group at their own rate of behavior change regardless of weekly weight loss. Therefore those who lose weight more slowly are less likely to be ignored and have their new behaviors placed on extinction. The fact that the attrition rate was lower in groups utilizing the self-monitoring and charting system would suggest that the participants found the experience more rewarding. The present results demonstrate that the addition of systematic self-monitoring and charting of eating behaviors to standard behavioral obesity programs increases treatment efficacy. Unfortunately, the self-monitoring and charting program in the present study was instituted as an independent rather than a dependent variable; therefore, these data were not collected from the participants at the conclusion of treatment. The direction of future research should be to compile the self-reports of compliance on each of the 11 eating behaviors to determine which of these behaviors correlates most highly with the final outcome goal of weight loss. REFERENCES Abramson, E. (1983). Behavioral treatment of obesity: Some good news, some bad news, and a few suggestions. The Behavior Therapist, 6, 103-106. Adams, N., Ferguson, J., Stunkard, A.J., & Agras, W.S. (1978). The eating behavior of obese and nonobese women. Behavior Research and Therapy, 16, 225-232. Coates, T.J., & Thoresen, C.E. (1981). Behavior and weight changes in three obese adolescents. Behavior Therapy, 12, 383-399. Cohen, J., & Cohen, P. (1975). Applied multiple regression~correlation analysisfor the behavioral sciences. Hillsdale, NJ: Erlbaum. Craighcad, L.W., Stunkard, A.J., & O'Brien, R.M. (1981). Behavior therapy and pharmocotherapy for obesity. Archives of General Psychiatry, 38, 763-768. Crawley, W.J., Adler, B.S., O'Brien, R.M., & Du ffy, E.M. (1982). Making salesmen: Behavioral assessment and intervention. In R.M. O'Brien, A.M. Dickinson, & M.P. Rosow (Eds), Industrial behavior modification: A management handbook (pp. 184-199). Elmsford, NY: Pergamon. Ferguson, J. (1975). Learning to eat: Behavior mod(fication for weight control. Palo Alto, CA: Bull. Gilbert, T.F. (1978). Human competence: Engineering worthy performance. New York: McGraw-Hill. Hagen, R.L., Foreyt, J.P., & Durham, T.W. (1976). The dropout problem: Reducing attrition in obesity research. Behavior Therapy, 7, 463-471. Jason, L., & Klich, M. (1982). Use of feedback in reducing television watching. Psychological Reports, 51, 812-814. Jeffrey, R.W., Vendor, M., & Wing, R.R. (1978). Weight loss and behavior change one year after behavioral treatment for obesity. Journal of Consulting and Clinical Psychology, 46, 368-396. Kingsley, R.G., & Wilson, G.T. (1977). Behavior therapy for obesity: A comparative investigation of longterm efficacy. Journal of Consulting and Clinical Psychology, 45, 288-298. Krassner, H., Brownell, K., & Stunkard, A. (1979). Case histories and shorter communications. Behavior Research and Therapy, 17, 155-156. Kreitner, R., Reif, W.E., & Morris, M. (1977). Measuring the impact of feedback on the performance of mental health technicians. Journal of Organizational Behavior Management, 1, 105-109. McDonald, M.R., & Budd, K.S. (1983). "Booster Shots" following didactic parent training: Effects of follow-up using graphic feedback and instructions. Behavior ModoCication, 7, 211-223. Schachter, S. (1982). Recidivism, and self-cure of smoking and obesity. American Psychologist, 37, 436-444. Sperduto, W.A., & O'Brien, R.M. (1983). Effects of cash deposits on attendance and weight loss in a largescale clinical program for obesity. Psychological Reports, 52, 261-262. Straw, M.K., & Terre, L. (1983). An evaluation of individualized behavioral obesity treatment and maintenance strategies. Behavior Therapy, 14, 255-266. Stunkard, A.J., & Albaum, J.M. (1981). The accuracy of self-reported weights. American Journal of Clincal Nutrition, 34, 1593-1599. Wooley, S., Wooley, O., & Dyrenforth, S. (1979). Theoretical, practical, and social issues in behavioral treatments of obesity. Journal of Applied Behavior Analysis, 12, 3-25.
0306-4603186 $3.00 + .GU Copyright e 1986 Pergamon Journals Ltd
Printed in the USA. All rights reserved.
BRIEF REPORT THE EFFECT OF TARGET BEHAVIOR MONITORING ON WEIGHT LOSS AND COMPLETION RATE IN A BEHAVIOR MODIFICATION PROGRAM FOR WEIGHT REDUCTION WILLIAM A. SPERDUTO Hofstra University
HARLEY S. THOMPSON Hicksville Public Schools, New York
RICHARD M. O’BRIEN B-MOD Associates Huntington,
New York
Abstract-Although behavior modification of obesity is usually descibed as a behavior change procedure, measurement is most often limited to the outcome variable of weight loss. The present investigation employed detailed behavior monitoring forms in half of 16 obesity groups (n = 173) matched across four different therapists. The percentage of compliance for nine specific treatment behaviors was charted from these monitoring forms. At the end of treatment, the eight groups that had the behavior monitoring treatment averaged over 6 pounds more weight lost than the matched control groups as well as showing significantly better attendance. The between group weight loss differences were still evident 3 months after treatment.
The major assumption in behavioral treatments of obesity is that being overweight is due to excess food consumption resulting from faulty eating habits (Wooley, Wooley, & Dyrenforth, 1979). Unfortunately, a failure to find strong relationships between selfreported behavior change and weight change despite significant changes in eating habits (Adams, Ferguson, Stunkard, & Agras, 1978; Krassner, Brownell, & Stunkard, 1979) has been noted in the literature. Two possible explanations have been advanced to account for such a discrepancy: changes in eating behaviors may not be linearly related to changes in weight, or positive relationships may have been obscurred by errors and biases in retrospective self-reports (Abramson, 1983; Jeffrey, Vender, & Wing, 1978). Attempts to directly work with eating behaviors in individual programs (Coates & Thoresen, 1981; Straw dz Terre, 1983) have produced more encouraging results; however, the maintenance of weight loss in individual programs has generally been less consistent than in group programs (Kingsley & Wilson, 1977; Straw & Terre, 1981). It would seem that further research is needed to systematically assess the effects of self-monitoring specific changes in behavior on weight loss in a group treatment program. Feedback through self-monitoring and charting has been shown to be an effective performance improvement technique in industry (Crawley, Adler, O’Brien, & Duffy, 1982), mental health settings (Kreitner, Reif, & Morris, 1977), and residential applicaPortions of this paper were presented at the annual meeting of the Association for the Advancement of Behavior Therapy, Toronto, November, 1981. This research was conducted at the Hofstra Center for Fitness and Health, Old Westbury, NY. Reprint requests should be addressed to Dr. Richard M. O’Brien, B-MOD Associates, 356 New York Avenue, Huntington, NY 11743. 337
338
WILLIAM A. SPERDUTO, HENRY S. THOMPSON and RICHARD M. O'BRIEN
tions (Jason & Klich, 1982; McDonald & Budd, 1983). Gilbert (1978) has emphasized that when one is trying to increase the performance of well-learned skills, it is best to monitor outcomes or results which in the case of obesity would be weight loss. But he has also suggested that monitoring behavior is most effective when new skills are being learned. Despite the emphasis on learning new eating and exercise behaviors in behavioral obesity programs, the measurement has focused almost exclusively on the results (weight loss). The present study was designed to examine the effect of actually monitoring and charting targeted eating and exercise behaviors on drop-out rate and number of pounds lost in weight reduction groups utilizing behavior modification. Based on the available literature, it was hypothesized that the weight loss of individuals utilizing behavior monitoring forms would be significantly greater than the weight loss of individuals treated without such forms. In addition, this difference should be maintained at 3-month follow-up. It was also hypothesized that the completion rate in groups utilizing the monitoring forms would be significantly greater than the completion rate in groups treated without such forms. METHOD
The study involved 16 groups of adult subjects (n = 173, M = 43, F = 130) who were treated at a fee-for-service ($185 per person) clinic for obesity. In addition to the program fee, attendance deposits ($25) were collected from all subjects entering the program to be refunded upon completion of at least 12 treatment sessions. Previous research in obesity and smoking (Hagen, Foreyt, & Durham, 1976; Sperduto & O'Brien, 1983) has shown refundable deposits to be effective in maintaining a low attrition rate. Recruitment of subjects was primarily through newspaper advertisement directed at the same geographic location; therefore, the demographic characteristics of subjects (sex, race, SES, age) remained essentially similar throughout the study. Treatment was conducted in a group format (approximately 11 clients per group) and consisted of 15 weekly 1-hour meetings. Ferguson's book Learning to Eat (1975) was the manual for all groups. One half of the groups were given additional target behavior monitoring forms which listed a total of nine behaviors related to eating and activity levels. These forms were adapted from Ferguson's daily behavior checklist (1975) with certain modifications including conversion from a Likert-type scale to a frequency count. The number of times a desired eating behavior was engaged in was compared to the total eating episodes each day. This formed a ratio which the participant transformed each week into a percentage of compliance. The participant then recorded this percentage on an on-going graph which was provided for each behavior. Thus, the client received weekly visual feedback on their performance on each of the required new behaviors. The therapist then praised their effort and set goals for the next week. Similar to other obesity programs (Craighead, Stunkard, & O'Brien, 1981), individual weights were recorded on a weekly basis. Four different therapists had pairs of their groups randomly assigned to experimental or control treatments so that all therapists ran groups in both conditions. The therapists were advanced graduate students (3 males, 1 female) in a doctoral psychology program who had been specifically trained in behavior modification for weight control. Telephone contacts were made 3 months following treatment in order to gather maintenance data. This was deemed appropriate since the literature has indicated that self-reports of weight are extremely accurate (Schacter, 1982; Stunkard & Albaum, 1981).
Behavior monitoring for weight reduction
339
RESULTS
Weight loss Combining all eight groups that utilized target behavior monitoring forms as part of the behavior modification program showed a mean weight loss of 15.90 pounds (SD = 3.86) at the end of treatment compared to a mean loss of 9.67 pounds (SD = 1.9) for the eight nonbehavior monitoring groups. The difference of 6.2 pounds at the end of treatment was significant on a one-tailed test (t14 = 4.05, p < .005). At 3 month follow-up, this difference (6.4 lbs.) was maintained (t14 = 3.72, p < .005). The behavior monitoring effects might be best illustrated by a measure of association, such as the point biserial correlation. The association between treatment and outcome for weight loss at the end of treatment was rpb = .74 which means that 55% of the variance was explained by the monitoring forms. Similarly, the association between treatment and outcome at the 3 month follow-up was r,b = .71 demonstrating that 50% of the variance was accounted for by the monitoring forms. According to Cohen and Cohen (1975) both correlations represent a large effect size. Completion rate Those groups utilizing target behavior forms had a completion rate of 74.4% as compared to 56.8% for the nonbehavior monitoring groups. A one-tailed test revealed that this difference was statistically significant (t~, = 2.24, p < .05). The association between treatment and completion rate was rpb = .51 indicating that 26% of the variance was explained by the monitoring forms. Once again this represents a large effect size (Cohen & Cohen, 1975) produced by the monitoring forms. DISCUSSION The present results suggest that the problems observed in the behavior modification treatment of obesity may be due to the fact that people have been treating obesity and not eating behaviors. Although behavioral weight loss programs are phrased in terms of changes in performance (eating behaviors), feedback has traditionally been limited to outcome (weight change) and global self-reports on the practice of new behaviors. In any instance of learning, however, it is necessary for the organism to monitor its performance in order to self-correct. In the present study, the subjects systematically selfmonitored and charted specific eating behaviors while the therapists shaped compliance. This performance was reinforced in relation to increasingly difficult behavioral goals through praise and visual performance feedback. For example, 20O7o compliance with a new eating behavior, such as putting the fork down between bites or leaving food on the plate may have been acceptable during week 4 but by week 14, 80O7o was the standard. It should be noted that, as in any shaping procedure, these goals were individualized for each participant. Gilbert (1978) has criticized behavior modifiers for ignoring outcome variables, in this case weight loss. However, in learning new behaviors, even Gilbert recognizes the need to monitor behavior directly. It would be absurd to run an obesity program without measuring weight loss, but it is also crucial to monitor eating behaviors for learning to take place. The participants must receive adequate on-going feedback on their performance and the therapist must have accurate measures of eating behaviors, not just weight loss, so he or she can contingently reinforce the participant's performance and identify problem areas. Utilizing self-monitoring and charting allows the participant to self-reinforce and set realistic goals while providing the therapist with the opportunity to directly reinforce behavior change without having to wait for weight loss. Actually,
340
WILLIAM A. SPERDUTO, HENRY S. THOMPSON and RICHARD M. O'BRIEN
monitoring behavior avoids the possibility of inadvertently reinforcing potentially dangerous behaviors (fasts, crash diets, and/or taking protein supplements) to achieve weekly weight loss goals. Such monitoring also allows the therapist to reinforce all members of the group at their own rate of behavior change regardless of weekly weight loss. Therefore those who lose weight more slowly are less likely to be ignored and have their new behaviors placed on extinction. The fact that the attrition rate was lower in groups utilizing the self-monitoring and charting system would suggest that the participants found the experience more rewarding. The present results demonstrate that the addition of systematic self-monitoring and charting of eating behaviors to standard behavioral obesity programs increases treatment efficacy. Unfortunately, the self-monitoring and charting program in the present study was instituted as an independent rather than a dependent variable; therefore, these data were not collected from the participants at the conclusion of treatment. The direction of future research should be to compile the self-reports of compliance on each of the 11 eating behaviors to determine which of these behaviors correlates most highly with the final outcome goal of weight loss. REFERENCES Abramson, E. (1983). Behavioral treatment of obesity: Some good news, some bad news, and a few suggestions. The Behavior Therapist, 6, 103-106. Adams, N., Ferguson, J., Stunkard, A.J., & Agras, W.S. (1978). The eating behavior of obese and nonobese women. Behavior Research and Therapy, 16, 225-232. Coates, T.J., & Thoresen, C.E. (1981). Behavior and weight changes in three obese adolescents. Behavior Therapy, 12, 383-399. Cohen, J., & Cohen, P. (1975). Applied multiple regression~correlation analysisfor the behavioral sciences. Hillsdale, NJ: Erlbaum. Craighcad, L.W., Stunkard, A.J., & O'Brien, R.M. (1981). Behavior therapy and pharmocotherapy for obesity. Archives of General Psychiatry, 38, 763-768. Crawley, W.J., Adler, B.S., O'Brien, R.M., & Du ffy, E.M. (1982). Making salesmen: Behavioral assessment and intervention. In R.M. O'Brien, A.M. Dickinson, & M.P. Rosow (Eds), Industrial behavior modification: A management handbook (pp. 184-199). Elmsford, NY: Pergamon. Ferguson, J. (1975). Learning to eat: Behavior mod(fication for weight control. Palo Alto, CA: Bull. Gilbert, T.F. (1978). Human competence: Engineering worthy performance. New York: McGraw-Hill. Hagen, R.L., Foreyt, J.P., & Durham, T.W. (1976). The dropout problem: Reducing attrition in obesity research. Behavior Therapy, 7, 463-471. Jason, L., & Klich, M. (1982). Use of feedback in reducing television watching. Psychological Reports, 51, 812-814. Jeffrey, R.W., Vendor, M., & Wing, R.R. (1978). Weight loss and behavior change one year after behavioral treatment for obesity. Journal of Consulting and Clinical Psychology, 46, 368-396. Kingsley, R.G., & Wilson, G.T. (1977). Behavior therapy for obesity: A comparative investigation of longterm efficacy. Journal of Consulting and Clinical Psychology, 45, 288-298. Krassner, H., Brownell, K., & Stunkard, A. (1979). Case histories and shorter communications. Behavior Research and Therapy, 17, 155-156. Kreitner, R., Reif, W.E., & Morris, M. (1977). Measuring the impact of feedback on the performance of mental health technicians. Journal of Organizational Behavior Management, 1, 105-109. McDonald, M.R., & Budd, K.S. (1983). "Booster Shots" following didactic parent training: Effects of follow-up using graphic feedback and instructions. Behavior ModoCication, 7, 211-223. Schachter, S. (1982). Recidivism, and self-cure of smoking and obesity. American Psychologist, 37, 436-444. Sperduto, W.A., & O'Brien, R.M. (1983). Effects of cash deposits on attendance and weight loss in a largescale clinical program for obesity. Psychological Reports, 52, 261-262. Straw, M.K., & Terre, L. (1983). An evaluation of individualized behavioral obesity treatment and maintenance strategies. Behavior Therapy, 14, 255-266. Stunkard, A.J., & Albaum, J.M. (1981). The accuracy of self-reported weights. American Journal of Clincal Nutrition, 34, 1593-1599. Wooley, S., Wooley, O., & Dyrenforth, S. (1979). Theoretical, practical, and social issues in behavioral treatments of obesity. Journal of Applied Behavior Analysis, 12, 3-25.