Taking a step back: developing interventions within a mediating-variable framework.

Abstract: This article discusses the recently introduced mediating-variable framework and summarizes the literature that supports the use of this framework to design health education interventions. In addition, common mediating variables through which a number of often-cited behavioral theories are activated are identified. Finally, example intervention methodology to help researchers and educators think about using the mediating-variable framework as a basis for designing future interventions are presented.

**********

Health educators use the concepts, methods, and measures of behavioral theory as a framework within which they can better understand health behaviors. Moreover, behavioral theory has provided the conceptual and empirical knowledge base for the design of health-promotion programs (Baranowski, Anderson, & Carmack, 1998). However, a number of community- and school-based trial interventions for chronic disease prevention that were based on popular behavioral theories have resulted in null or small behavioral effects (Carleton, Lasater, Assaf, et al., 1995; Fortmann, Taylor, Flora, & Winkelby, 1993; Glasgow, Terborg, Hollis, Severson, & Boles, 1995; Lando, Pechacek, Pirie, et al., 1995; Luepker, Murray, Jacobs, et al., 1994; Leupker, Ferry, McKinlay, et al., 1996). In addition, out of 23 published physical-activity intervention studies whose designs relied on behavioral theory, a substantial number demonstrated little or no impact on the target behaviors (Estabrooks, Courneya, & Nigg, 1996; King, Taylor, Haskell, & Debusk, 1988; Luepker, Murray, Jacobs, et al., 1994; Resnicow, Davis, Smith, et al., 1998).

These findings do not necessarily mean that interventions based on behavioral theory cannot elicit positive changes in behavior. Instead they may be telling us that we need to simplify our intervention designs. By basing our interventions on a few key variables--known as mediating variables and drawn from behavioral theory--rather than on complicated theoretical models, we can better examine the impact of the interventions on these mediating variables and isolate the relationship between the variables and our desired outcomes.

In this article, we discuss the recently introduced mediating-variable framework and summarize the literature that supports the use of this framework to design health education interventions. In addition, we identify common mediating variables through which a number of often-cited behavioral theories are activated and present example intervention methodology to help researchers and educators think about using this framework as a basis for designing future interventions.

BEHAVIORAL THEORY IN HEALTH EDUCATION

Among the behavioral theories most commonly cited in the health education literature as having provided a framework or model for designing behavioral interventions are the health belief model (Rosenstock, 1974), protection motivation theory (Prentice-Dunn & Rogers, 1986), social learning theory (Bandura, 1986), the theory of reasoned action (Fishbein & Ajzen, 1975), and the theory of planned behavior (Ajzen, 1985). Rather than developing in isolation, these theories often grew out of or elaborated upon one another. For example, the health belief model and social learning theory provided the basis for the development of the protection motivation theory (Prentice-Dunn & Rogers, 1986), whereas the theory of planned behavior combines insights from the theory of reasoned action and the social learning theory (Ajzen, 1998). As a result, these commonly used behavioral theories share construct (mediating) variables for behavior prediction.

It is increasingly apparent that these shared mediating variables may provide stronger correlates to desired behavioral outcomes than do the theoretical models as a whole. For example, the health belief model does not specify a mechanism by which health beliefs are translated into action. Such a mechanism is proposed, however, as a part of the theory of reasoned action, which suggests that the formation of a behavioral intention is the immediate antecedent of action and mediates the influence of other variables (Wallston, 1994). Additionally, Ajzen's (1985) theory of planned behavior extends the theory of reasoned action by adding perceived behavioral control (i.e., self-efficacy) as a predictor of both intention formation and action. This extension is an acknowledgment of evidence suggesting that self-efficacy is a reliable predictor of action (Ajzen, 1998). In short, these examples of theoretical development suggest that the addition of key mediating variables improves the predictive capabilities of interventions based on these theoretical models.

THE MEDIATING-VARIABLE FRAMEWORK

The mediating-variable framework, first presented by Hansen and McNeal (1996) and further developed by Baranowski et al. (1997, 1998), is based on the arguments that (i) interventions work by means of mediating variables; (ii) current theoretical models from which mediating variables are obtained often do not account for the substantial variability in targeted outcomes; (iii) interventions have not been shown to substantially effect change in the mediating variables; and (iv) together, these factors impose limits on the effectiveness of the interventions.

As a result of the emergence of this framework, many researchers claim that priority should be placed on research that enhances our understanding of the relation between mediating variables and targeted outcomes and that measure directly the impact of interventions on these mediating variables (Baranowski, Lin, Wetter, Resnicow, & Hearn, 1997). According to Baron and Kenny (1986), a mediating variable could be said to account for the outcome effect of an intervention if a positive relationship between the intervention and outcome were rendered nonsignificant after statistically controlling for that variable. Using measurements such as effect size, variance percentage, and relative contribution via regression analyses to examine the individual predictive power of variables thus assist in determining what factors are impacting the targeted behavior. Examples of studies that control for individual factors to determine which are, in fact, the mediating variables of the intervention (i.e., those affecting behavior) are presented in the following sections. In addition to describing the statistical relationship for the mediating variable framework, it is also possible to graphically depict the mediating flow-through process (see figure 1).

[FIGURE 1 OMITTED]

Simply stated, behavioral theories contain common mediating variables. These variables may be used as the mechanisms through which behavioral interventions affect outcome behavior. By designing interventions to produce change in mediating variables that, in turn, affect outcome behaviors, researchers can more strongly link these outcome behaviors back to behavioral theory.

This framework can also be thought of in terms of mechanisms of action (MOA), a concept defined in the field of pharmacy as the process by which a given agent ffects a given outcome. The "outcome" can be thought of as the prediction of a behavior and the agent as the given behavioral theory. In other words, the mediating variable is the MOA that allows a given theory (agent) to result in the affecting the outcome of behavior.

The need for this type of research can be demonstrated with a look at the literature. For instance, in a meta-analysis used to quantify the relationship between 44 psychosocial variables (most of which were drawn from the theory of reasoned action) and heterosexual condom use, only nine of the variables had effect sizes that could be characterized as medium to strong. Only three of those had been drawn from the theory of reasoned action (Sheeran, Abraham, & Orbell, 1999). Additionally, an extensive review article revealed that the theory of planned behavior substantially predicted behavioral intention but was less predictive of behavior (Godin & Kok, 1996). These examples support the contention put forward by the developers of the mediating-variable framework that existing theoretical research accounts for a relatively small percentage of the variance in target behaviors. Taking a step back from complicated interventions whose designs attempt to incorporate all facets of these behavioral theories may provide insight into basic behavioral relationships. By focusing instead on the ability of interventions to effect changes in mediating variables, researchers are able to place another limit on the impact of an intervention on the outcome behavior (Baranowski, Lin, Wetter, Resnicow, & Hearn, 1997).

It is important to note that the mediating-variable framework does not question the value of behavioral theory for designing effective interventions. On the contrary, because the mediating variables for behavioral interventions are drawn directly from behavioral theories, the framework actually highlights the importance of theory for understanding intervention results.

COMMON MEDIATING VARIABLES DRAWN FROM BEHAVIORAL THEORY

Before an intervention based on the mediating-variable framework can be developed, possible mediating variables for a given behavior must be identified and examined individually. Some of the most common mediating variables that can be drawn from the above-mentioned behavioral theories include self-efficacy, past behavior or previous experience, social support, and behavioral intention.

SELF-EFFICACY

Self-efficacy is a person's belief that he/she is capable of successfully performing a behavior in order to achieve a desired outcome. Though it appears that no single variable solely determines adherence to either prescribed or self-initiated exercise and other health improvement programs, self-efficacy has been consistently identified as playing an important role in health and exercise behaviors (Bandura, 1986). This construct appears in several behavioral theories and practice models including the protection motivation theory (Prentice-Dunn & Rogers, 1986), the health action process approac (HAPA;Schwarzer, 1992), O'Donnell's model of health promotion behavior (first appeared in Wallston, 1994), and has even been added to the health belief model (Rosenstock, 1990; Rosenstock, Strecher, & Becker, 1988). Furthermore, there has been a growing consensus that self-efficacy is among the most important and modifiable predictors of physical activity behavior (Pate et al., 1995; Sallis et al., 1992; USDHHS, 1996).

Self-efficacy theory (Bandura, 1986) has generated an enormous amount of literature in multiple domains of behavioral functioning. It is hardly surprising that self-regulatory skills have been consistently implicated in successful adoption and maintenance of complex behaviors such as physical activity, which appears particularly difficult for many individuals. In these cases, focusing on methods for boosting self-efficacy as a mediating variable for the desired behavorial outcome can minimize the negative effects of discouragement, feelings of displeasure and incompetence, and proclivity to give up in the face of real or perceived adversity or challenge.

PAST BEHAVIOR (PRIOR EXPERIENCE)

In a meta-analysis of studies that examined the direct and indirect effects of past behavior on future responses, Oulette and Wood (1998) found strong support for the idea that past behavior has a significant impact on present and future behavior. According to the authors, past behavior affects future responses through one of two mechanisms. Well-practiced behaviors performed in familiar or constant contexts recur because the processing that initiates and controls their performance becomes automatic, affecting future performance. Alternately, when behaviors are not well learned or when they are performed in unstable or difficult contexts, conscious decision making is likely to be necessary to initiate and carry out the behavior. In this case, it is thought that past behavior contributes to intentions, and the current or future behavior is then affected by those intentions.

These mechanisms may explain the role that past health behavior has played in predicting future and actual health behavior in recent studies. Bozionelos and Bennett (1999) hypothesized that past behavior, personal normative beliefs, role beliefs, level of self-monitoring, and sex role identity would mediate the influence of variables drawn from the theory of planned behavior on intention to exercise and actual exercise behavior. Their findings revealed that past behavior was the most predictive variable of both intention to exercise and actual exercise behavior. This research supports the usage of past behavior as a mediating variable around which exercise and other health promotion interventions could be designed.

Additional studies support the consideration of past behavior as an important mediating variable for health behaviors. Conner et al. (1999) found past drinking behavior to be more predictive of intended drinking than attitudes, subjective norms, and perceived behavioral control. Past drinking behavior was also consistently more predictive of actual drinking behavior than were intention to drink and perceived behavioral control. Baker et al. (1996) found that considering prior use of condoms increased the ability m predict future condom use for both men and women. These authors recommend that the relationship between sexual partners and the individual's prior experience with condom use should be incorporated into attempts to understand the complex, dyadic behavior of condom use. Similarly, O'Callaghan et al. (1999) questioned high school students about their attitudes, subjective norms, perceived behavioral control, past behavior, intentions and actual behavior relating to the use of cigarettes. Results indicated that attitudes toward smoking, past behavior in relation to smoking, and perceptions of the attitudes of others were significant predictors of the intention to smoke. Intention, together with past behavior, predicted actual smoking behavior.

SOCIAL SUPPORT

Social support can be defined as support that one individual receives from another for engaging in a particular health behavior. It has been included in several behavioral theories, such as the HAPA (Schwarzer, 1992), the health promotion model (Pender, 1982) and O'Donnell's model of health promotion behavior (Wallston, 1994). Courneya and McAuley (1995) examined cognitive constructs as mediators of the relationship between selected social influence constructs and adherence to structured exercise classes. They found that social support appeared to be a mediating variable for perceived behavioral control and intention to exercise. In a meta-analysis of social influence and exercise, Carron et al. (1996) supported an overall conclusion that social influence has a positive impact on exercise behavior (both adherence and compliance), cognitions about exercise involvement (both intention to exercise and self-efficacy for exercise), and attitudes associated with the exercise experience. Another finding of this meta-analysis was that family did not represent the strongest source of social influence for adherence behavior, rather the influence of important others had a stronger impact. These researchers suggest that if important others provide support without exerting control over behavior, the individual can retain a perception of self-determination, and adherence should be enhanced. Duncan et al. (1993) endorsed the utility of a multidimensional view of social support and found that specific provisions of support may play a significant role in an individual's decision to adhere to a prescribed exercise regimen. Moreover, sources of social support can play a pivotal role in the adoption of other health behaviors as well.

BEHAVIORAL INTENTION

According to the theory of reasoned action, attitudes and subjective norms are the psychological prerequisites of intention formation. Attitudes are thought to be based on beliefs about, and evaluations of, the consequences of action, including beliefs about action effectiveness and barriers specified by the health belief model. Subjective norms are derived from an individual's beliefs about what other people think he or she should do and from motivation to comply with those views (Fishbein & Azjzen, 1975). Typically, the theory of reasoned action does a better job of accounting for a person's behavioral intentions than the health belief model does of predicting behavior (Schwarzer, 1992). The theory of planned behavior seems to account for behavioral intention better than all other theories, but it still fails to explain all the variability in behavioral intention and explains even less of the variability among individuals in their actual behavior (Wallston, 1994). Hence, there is a need to test behavioral intention as a mediating variable.

DESIGNING INTERVENTIONS WITHIN THE MEDIATING-VARIABLE FRAMEWORK

The goal, then, is for reseachers to develop behavioral interventions that effect change in mediating variables at some defined acceptable level (Baranowski, Anderson, & Carmack, 1998). The knowledge gained from such studies would then allow researchers and practitioners to identify more precisely the means by which behavioral theories may be translated into desired behavioral outcomes.

Outlined in Table 1 are example methodologies for a hypothetical exercise promotion intervention designed to effect change in common mediating variables for popular behavioral theories. These are provided in order to assist health educators and researchers to develop interventions in their own areas using the mediating-variable framework.

Methodologies designed around the mediating-variable framework can be adapted for either individual or community interventions. All of the intervention methods and health educator roles mentioned in Table 1 can be employed either with groups of individuals or in one-on-one sessions. And, of course, the intervention methods presented could be applied to other desired health behaviors, such as nutrition promotion, condom use, smoking cessation, or weight management.

Since it is thought that these common mediating variables are the primary mechanisms underlying the predictive capabilities of behavioral theories, interventions designed for the sole purpose of producing changes in these key variables should more directly influence behavior than those that attempt to incorporate all facets of a theory.

Several studies have used the above-mentioned mediating variables to successfully predict outcome behaviors, such as exercise adherence. Clark and Dodge (1999), in examining self-efficacy as a predictor for prescribed medicine use, getting adequate exercise, managing stress, and following a recommended diet among a sample of women, found that the variable was a significant predictor for the diet and exercise models (at both 4 and 12 months). Specifically, those women who at baseline had confidence in their own ability to stick to their diets or get adequate exercise were more likely at follow-up to believe the diet or getting exercise was beneficial to their health (Clark & Dodge, 1999).

Other studies have successfully used self-efficacy as a mediating variable for exercise behavior among sedentary adults. Calfas et al. (1997) demonstrated that change in efficacy for "making time for exercise," significantly predicted the change in a residualized activity score. Dunn et al. (1997) demonstrated that change in self-efficacy was significantly associated with achieving the CDC/ACSM criteria for accumulated bouts of physical activity on most days of the week.

In using social support and self-efficacy as mediating variables for participation in physical activity, Cousins (1996) found that these variables were significant predictors for physical activity participation among elderly Canadian women. In addition, Muto et al. (1996) found that colleague support and self-efficacy were significant predictors of leisure-time physical or sports activity among Japanese workers.

Finally, in using behavioral intention as a mediating variable for reported physical activity, Terry and O'Leary (1995) found the variable to be a significant predictor of self-reported adherence to prescribed exercise among Australian college students. Biddle et al. (1994) found behavioral intention to be a significant predictor of self-reported frequency of strenuous physical activity among British university employees. Fuchs (1996) found behavioral intention to be a predictive variable for both sedentary behavior and physical activity behavior among German adults.

IMPLICATIONS FOR BEHAVIORAL THEORY

Interventions designed for the purpose of producing changes in mediating variables can help to establish stronger links between the behavioral theories and the desired behaviors. Baranowski et al. (1997) suggest a progressive research process that would begin with testing the extent to which an intervention affects a single target mediating variable. This is an important process since mediating variables are drawn from behavioral theories hypothesized to affect the target behavior. Eventually, mediating variables would be added to the testing, leading ultimately to the building of a powerful intervention that incorporates many variables.

In the same way that mediating variables may be drawn from behavioral theory to build powerful interventions, the relationship established through the intervention between the mediating variables and the outcome measures may be used to strengthen behavioral theory. The final step of the flow-through process (see Figure 1) involves using the effect of the mediating variables on the targeted outcome behavior to redesign or modify the existing behavioral theory.

It is evident from the previously cited studies that not all components of behavioral theories necessarily assist in explaining behavior (i.e., not all of them prove to be mediating variables). It is further evident that certain mediating variables more closely predict behavior than others. Therefore, researchers need to use the knowledge gained from those interventions to redesign existing behavioral theory. Such revised theories will be better suited to frame future effective interventions.

CONCLUSIONS

Health educators and behavioral researchers need to take a step back from designing interventions based on complicated theoretical models that are overburdened with many mediating variables. Instead, they should first design interventions that focus on one or two mediating variables. In addition, as stated by Baranowski et al. (1998), different mediating variables must be tested among different subgroups, health behaviors, and environments to determine the conditions under which each variable has influence. Using these simplified interventions to produce change in these basic mediating variables would validate these variables, which could then be combined or recombined in more complicated interventions. As health education researchers and educators increase their knowledge and understanding of the mediating-variable framework and of the conditions under which certain variables are most effective, they should be able to design more effective interventions that have a significant impact on outcome behaviors. In addition, by applying their results back to the behavioral theories from which the variables were drawn, researchers may also significantly refine the theoretical framework available for designing future interventions.

Table 1. Using the Mediating-Variable Framework
to Design an Exercise Promotion Intervention.

Common
mediating         Method of          Health
variable          intervention       educator's role    Desired change

Self-efficacy     Use Bandura's      Serve as a role      To increase
                  methods of         model for           self-efficacy
                  increasing         exercise,           for exercise
                  self-efficacy:     assist in
                  1) Authentic       making
                     mastery         participants
                     experience      feel successful
                  2) Social          in exercise,
                     modeling--      be a source of
                     peer role       social support,
                     model           and educate
                  3) Modifying       participants
                     physiological   to better
                     reactions       understand
                                     bodily
                                     responses to
                                     exercise

Past behavior     1) Inquire about   Promote            To incorporate
or Past              PE with         positive past       any positive
experience (PE)      exercise--      experience with    past experience
                     both positive   exercise to         with exercise
                     and negative    encourage             into the
                  2) At program      participation          current
                     start, (may     and strategize        exercise
                     ask in face-    with                   program
                     to-face         participant to
                     sessions)       change
                     focus on how    perception of
                     to change       negative past
                     perception of   experiences
                     negative PE
                     and use
                     positive PE

Social Support    1) Identify        1) Promote         To increase the
                     those who          exercising         amount of
                     support            with a          social support
                     participant's      "buddy" or       received for
                     exercising         group              exercise
                  2) Elicit          2) Have               adherence
                     support from       participant
                     significant        and
                     others for         supporters
                     participant's      sign an
                     involvement        exercise
                     in the             contract
                     program            developed by
                  3) Have               participant
                     participant        & HE
                     invite
                     significant
                     others to
                     join exercise
                     program

Behavioral        1) Exercise        1) Educate           To increase
intention            attitudes:         participant     the behavioral
                     assess             regarding        intention to
                     beliefs about      positive           exercise
                     what positive      outcomes of        regularly
                     or negative        exercise and
                     outcome will       prepare
                     result from        participant
                     exercise           for any
                  2) Exercise           negative
                     social norms:      experiences
                     assess the         (soreness,
                     extent to          fatigue,
                     which              etc.)
                     exercise is     2a) Determine
                     motivated by        what
                     the wishes of       "others"
                     salient             motivate
                     others              participant.
                                         2b) Invite
                                         salient
                                         others to
                                         participate
                                         in the
                                         program
                                         and/or
                                         simply be
                                         included
                                         in the
                                         educational
                                         facet of
                                         the program

REFERENCES

Ajzen, I. (1985). From intentions to actions: A theory of planned behavior. In J. Kuhl & J. Beckmann (Eds.), Action-control: From cognition to behavior (pp. 11-39). Heidelberg: Springer.

Ajzen, I. (1988). Attitudes, personality, and behavior. Milton Keynes: Open University Press.

Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Process. 50, 179-211.

Ajzen, I. (1998). Models of human socal behaviour and their application to health psychology. Psychology and Health, 13, 735-740.

American Heart Association/American College of Sports Medicine (1998). AHA/ACSM Scientific Statement: Recommendations for cardiovascular screening, staffing, and emergency policies at health/fitness facilities. Circulation, 97, 2283-2293.

Baker, S. A., Morrison, D. M., Carter, W. B., & Verdon, M. S. (1996). Using the theory of reasoned action (TRA) to understand the decision to use condoms in an STD clinic population. Health Education Quarterly, 23, 528-542.

Bandura, A. J. (1986). Social foundations of thought and action. Englewood Cliffs, NJ: Prentice Hall.

Baranowski, T., Lin, L. S., Wetter, D. W., Resnicow, K. & Hearn, M. D. (1997). Theory as mediating variables: Why aren't community interventions working as desired? Annals of Epidemiology, 7, S89-S95.

Baranowski, T., Anderson, C., & Carmack, C. (1998). Mediating variable framework in physical activity interventions: How are we doing? How might we be doing better? American Journal of Preventative Medicine, 15(4), 266-297.

Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173-1182.

Biddle, S., Goudas, M., & Page, A. (1994). Social psychological predictors of self-reported actual and intended physical activity in a university workforce sample. British Journal of Sports Medine, 23(3), 160-163.

Bozionelos, G., & Bennett, P. (1999). The theory of planned behavior as predictor of exercise: The moderating influence of beliefs and personality variables. Journal of Health Psychology, 4, 517-529.

Calfas, K. J., Sallis, J. F., Oldenburg, B., & French, M. (1997). Mediators of change in physical activity following an intervention in primary care: PACE. Preventive Medicine, 26, 297-304.

Carleton, R. A., Lasater, T. M., Assaf, A. R., et al. (1995). The Pawtucket Heart Health Program: Community changes in cardiovascular risk factors and projected disease risk. American Journal of public Health, 85, 777-785.

Carron, A. V., Hausenblas, H. A., & Mack, D. (1996). Social influence and exercise: A meta-analysis. Journal of Sport and Exercise Psychology, 18, 1-16.

Clark, N. M., & Dodge, J. A. (1999). Exploring of self-efficacy as a predictor of disease management. Health Education and Behavior, 26(1), 72-89.

Conner, M., Warren, R., Close, S., & Sparks, P. (1999). Alcohol consumption and the theory of planned behavior: An examination of the cognitive mediation of past behavior. Journal of Applied Social Psychology, 29, 1676-1704.

Courneya, K. S., & McAuley, E. (1995). Cognitive mediators of the social influence-exercise adherence relationship: A test of the theory of planned behavior. Journal of Behavioral Medicine, 18, 499-515.

Cousins, S. O. (1996). Exercise cognition among elderly women. Journal of Applied Sport Psychology, 8, 131-145.

Cox, M. H. (1984). Fitness and lifestyle programs for business and industry: Problems in recruitment and retention. Journal of Cardiac Rehabilitation, 4, 136-142.

Duncan, T. E., McAuley, E., Stoolmiller, M., & Duncan, S. C. (1993). Serial fluctuations in exercise behavior as a function of social support and efficacy cognitions. Journal of Applied Social Psychology, 23, 1498-1522.

Dunn, A. L., Marcus, B. H., Kampert, J. B., Garcia, M. E., Kohl, H. W. III, Blair, S. N. (1997). Reduction in cardiovascular disease risk factor: Sixth-month results from Project Active. Preventive Medicine, 26, 883-892.

Estabrooks, P. A., Courneya, K. S., Nigg, C. R. (1996). Effect of a stimulus control intervention on attendance at a university fitness center. Behavior Modification, 20, 206-213.

Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior: An introduction to theory and research. Reading, MA: Addison-Wesley.

Fletcher, G. F., Balady, G. J., Blair, S. N., Blumenthal, J., Casperson, C., Chaitman, B., Epstein, S., Froelicher, E. S., Froelicher, V., Pina, I., & Pollock, M. (1996). Statement of exercise: Benefits and recommendations for physical activity programs for all Americans. Circulation, 94, 857-862.

Fortmann, S. P., Taylor, C. B., Flora, J. A., & Winkelby, M.A. (1993). Effect of community health education on plasma cholesterol levels and diet: The Stanford Five City Project. American Journal of Epidemiology, 137, 1039-1055.

Fuchs, R. (1996). Causal models of physical exercise participation: Testing the predictive power of the construct "pressure to change." Journal of Applied Social Psychology, 26(21), 1931-1960.

Glasgow, R. E., Terborg, J. E., Hollis, J. F., Severson, H. H., & Boles, S. M. (1995). Take Heart: Results from the initial phase of a worksite wellness program. American Journal of Public Health, 85, 209-216.

Godin, G., & Kok, G. (1996). The theory of planned behavior: A review of its applications to health related behaviors. American Journal of Health Promotion, 11, 87-89.

Hansen, W. B., & McNeal, R. B. Jr. (1996). The law of maximum expected potential effect: Constraints placed on program effectiveness by mediator relationships. Health Education Research Theory and Practice, 11, 501-507.

King, A. C., Taylor, C. B., Haskell, W. L., & DeBusk, R. F. (1988). Strategies for increasing early adherence to and long term maintenance of home-based exercise training in healthy middle aged men and women. American Journal of Cardiology, 61, 628-632.

Lando, H. A., Pechacek, T. F., Pirie, F. L., et al. (1995). Changes in adult cigarette smoking in the Minnesota Heart Health Program. American Journal of Public Health, 85, 201-208.

Luepker, R. V., Ferry, C. L., McKinlay, S. M., et al. (1996). Outcomes of a field trial to improve children's dietary patterns and physical activity. The Child and Adolescent Trial for Cardiovascular Health (CATCH). JAMA, 275, 768-776.

Luepker, R. V., Murray, D. M., Jacobs, D. R., et al. (1994). Community education for cardiovascular disease prevention: Risk factor changes in the Minnesota Heart Health Program. American Journal of Public Health, 84, 1383-1389.

Muto, T., Saito, T., & Sakurai, H. (1996). Factors associated with male workers' participation in regular physical activity. Indonesian Health, 34, 307-321.

O'Callaghan, F. V., Callan, V. J., & Baglioni, A. (1999). Cigarette use by adolescents: Attitude-behavior relationships. Substance Use & Misuse, 34, 455-468.

Oulette, J. A., & Wood, W. (1998). Habit and intention in everyday life: The multiple processes by which past behavior affects future behavior. Psychological Bulletin, 124, 54-74.

Pate, R. R., Pratt, M., Blair, S. N., Haskell, W. L., Macera, C. A., Bouchard, C., Buchner, D., Ettinger, W., Heath, G. W., King, A. C., Kriska, A., Leon, A. S., Marcus, B. H., Morris, J., Paffenbarger, R. S., Patrick, K., Pollock, M. L., Rippe, J. M., Sallis, J., & Wilmore, J. H. (1995). Physical activity and public health: A recommendation from the Centers for Disease Control and Prevention and the American College of Sports Medicine. JAMA, 273, 402-407.

Pender, N.J. (1982). Health promotion in nursing practice. New York: Appleton-Century-Crofts.

Prentice-Dunn, S., & Rogers, R. W. (1986). Protection motivation theory and preventative health: Beyond the health belief model. Health Education Research, Theory, and Practice, 1, 153-161.

Resnicow, K., Davis, M., Smith, M., et al. (1998). Results of the Teach Well Worksite Wellness Program. American Journal of Public Health, 88, 250-257.

Rosenstock, I. M. (1974). Historical origins of the health belief model. Health Education Monographs, 2, 328-335.

Rosenstock, I. M. (1990). The health belief model: Explaining health behavior through expectancies. In Health behavior and health education: Theory, research, and practice, edited by K. Glanz, F. M. Lewis, and B. K. Rimer. San Francisco: Jossey-Bass.

Rosenstock, I. M., Strecher, V. J., & Becker, M. H. (1988). Social learning theory and the health belief model. Health Education Quarterly, 15, 175-183.

Sallis, J. F., Hovell, M. F., Hofstetter, R., & Barrington, E. (1992). Explanation of vigorous physical activity during two years using social learning variables. Social Science Medicine, 34, 25-32.

Schwarzer, R. (1992). Adaptation and maintenance of health behaviors: A critical review of theoretical approaches. In Self-efficacy: Thought control of action. Ed. R. Schwarzer. New York: Hemisphere.

Sheeran, P., Abraham, C., & Orbell, S. (1999). Psychosocial correlates of heterosexual condom use: A meta-analysis. Psychological Bulletin, 125, 90-132.

Terry, J. T. & O'Leary, J. B. (1995). The theory of planned behavior: The effects of perceived behavioral control and self-efficacy. British Journal of Social Psychology, 34, 199-220.

USDHHS. (1996). Physical activity and health. Report of the Surgeon General. Pittsburgh, PA.

Wallston, K. A. (1994). Theoretically based strategies for health behavior change. In M. P. O'Donnell & J. S. Harris (Eds.), Health promotion in the workplace (2nd ed., pp. 185-203). Albany, NY: Delmar.

Kristine S. Calderon, Ph.D., CHES, is a Research Assistant Professor and Jill W. Varnes, Ed.D., CHES, is a professor at the University of Florida. Address all correspondence to Dr. Calderon at the Office of Research Support; College of Nursing; Health Science Center; P.O. Box 100187; University of Florida; Gainesville, FL, 32601-0187; Phone: [phone], FAX: [phone]; e-mail: [e-mail].