Effects of cognitive dissonance on intentions to change diet and physical activity among college students.

Abstract: The purpose of this study was to test the hypothesis that when college students are made to feel cognitive dissonance about their diet and exercise behaviors, they will be more likely to adopt healthier diet and exercise habits--particularly when the dissonance is tied to appearance rather than health concerns. One hundred twenty-six college students reported a number of diet and exercise behaviors after writing about why high-quality diet and exercise promotes health (dissonance-health), or physical appearance (dissonance-appearance); or they wrote about an unrelated topic (control). Risk perceptions related to negative health and appearance consequences emanating from diet and physical activity levels were then assessed. Following this, participants were instructed to indicate whether they intended to change their diet or exercise behaviors, for the better, anytime during the next 6 months. It was found that dissonance did not effect absolute levels of risk perceptions or intentions, but did influence the relationship between risk perceptions and intentions. In particular, there was no correlation in the control group, a negative correlation in the dissonance-appearance group, and a positive correlation in the dissonance-health group. Implications of these findings are discussed.

**********

Obesity has become an increasingly important public health problem in the United States (Centers for Disease Control and Prevention, 2006). Over the last few decades, the percentage of Americans who are clinically overweight has been steadily increasing (National Health Interview Survey, 2004). This trend has filtered into college-aged student populations. According to the American College Health Association's National College Health Assessment (2005), almost 30% of college students are classified as overweight. The high prevalence of overweight among college students can be attributed, in part, to inadequate health behaviors, particularly improper eating and lack of physical activity. Schuette and colleagues (1996) found that only 4% of college students reported eating less than 30% of calories from fat, and 25% of college students participate in no vigorous exercise for at least 20 minutes per week (National College Health Assessment, 2005). Due to a lack of healthy eating and regular exercise, a significant proportion of college students are becoming increasingly susceptible to overweight and obesity (Lowry et al., 2000).

As college students begin to manage their own lives, they are at liberty to regulate their involvement in various health behaviors. Health promotion programs targeted at college students must emphasize the importance of integrating healthy behaviors as routine during these seminal years (Martinelli, 1999). By encouraging college students to adopt healthy diet and physical activity behaviors, public health educators can help to extend the number of years of enhanced quality of life for individuals who might otherwise become predisposed to chronic disease (Fries, 2004). This philosophy becomes all the more practical as medical research indicates that sedentary individuals who eat poorly are much more likely to develop health problems such as diabetes, cardiovascular disease, cancer, and stroke (National Institutes of Health, 2005).

What then is the best way to elicit healthy diet and physical activity behaviors from college students? Most students are not meeting dietary and physical activity guidelines, which suggests a need for prevention interventions and increased understanding of overweight college students (Huang et al., 2003). While balanced diets and innovative exercise regimens provide the means to reduce obesity, these programs generally do not have lasting effects among college student populations (Nicklas et al., 1995; Engstrom, Tobelmann, & Albertson, 1997; American College of Sports Medicine, 2000; Centers for Disease Control and Prevention, 2004). Most interventions of this type focus on endorsing actual behaviors, but few focus on the cognitive barriers preventing healthy diet and physical activity. Attitudes constitute a considerable cognitive barrier limiting adherence to healthy behaviors (Ziebland et al., 1998).

Cognitive dissonance (Festinger, 1957) has been shown to affect attitudes and behavior by creating inconsistent cognitions within individuals (Draycott & Dabbs, 1998). Cognitive inconsistencies have been shown to stimulate individuals to actively attempt altering their behaviors in hopes of attaining consonance between attitude and behavior (Elliot & Devine, 1994). By causing college students to feel cognitive dissonance between their attitudes about diet and physical activity and their own self-reported diet and physical activity behaviors, they may be more likely to reassess their intentions to engage in these health behaviors. Once individuals feel dissonance between their attitudes and behaviors, the arousal they feel may serve as a catalyst for deciding to engage in healthy behaviors (Stone et al., 1994). After encountering cognitive dissonance, individuals may display increased levels of risk and worry regarding their negative health behaviors, which, in turn, could help to influence positive behavioral intentions.

Individuals have been shown to feel the need to reduce any dissonance that they feel between their health attitudes and behaviors, by actively changing their health habits (Leary, Tchividjian, & Kraxberger, 1994). The dissonance that one feels can have a varying impact on behavioral intentions, depending on the circumstances associated with the dissonance. When individuals feel cognitive dissonance between their attitudes about health behaviors and their actual health behaviors, it is proposed that their feelings of risk regarding their physical appearance and health will become elevated. Risk perceptions are thought to be a key predictor of health behaviors in many health behavior models, such as the Health Belief Model (Janz & Becker, 1984), Protection Motivation Theory (Rogers & Prentice-Dunn, 1997), and the Precaution Adoption Process Model (Weinstein, 1998); however, the actual association between these two constructs varies considerably. For example, risk perceptions for HIV are only moderately correlated with intentions to have unprotected sex (Gerrard, Gibbons, & Bushman, 1996).

The association between risk perception and behavioral intentions may be influenced by a variety of situational constraints. For instance, sexual decision-making among young adults is often made under the influence of alcohol and in the context of increased sexual arousal, thereby reducing the salience of risk perceptions. Similarly, we were interested in whether the risk and worry posed by cognitive dissonance might influence the risk perception--intentions relationship. In related work, we found that giving people feedback that their diets were less healthy than that of their peers reduced the correlation between dietary variables (e.g., red meat consumption) and perception of risk for heart disease (Klein, Blier, & Janze, 2001). Although this study did not measure intentions, it shows that associations among health-related cognitions may be influenced by threats.

Studies have shown that interventions designed to threaten individuals' views of how they are perceived are more effective than traditional, information-based health interventions (Leary, Tchividjian, & Kraxberger, 1994; Feingold, 1992; Mahler et al., 2003). College students have self-reported reducing UV exposure after being subjected to a novel, photo-aging (premature wrinkling and age spots) intervention making the negative-appearance consequences of UV exposure more salient (Mahler et al., 2003). Research has shown that people will change their health behaviors in order to improve and/or maintain their outward physical appearance so that others will see them as desirable (Leary, Tchividjian, & Kraxberger, 1994). Those who possess a more attractive appearance are perceived as being more sociable, warm, intelligent, and socially skilled. Conversely, those who are viewed as less attractive draw inferences of laziness, self-indulgence, and lack of self-control (Feingold, 1992). Because of these perceptions, people may regulate their weight due to concerns about physical appearance (Leary, Tchividjian, & Kraxberger, 1994).

Since cognitive dissonance occurs when it is salient to people that they hold two inconsistent cognitions, any contextual factor that increases cognitive dissonance by definition is making cognitions more accessible. The strength of this relationship may depend, however, on the nature of the dissonance treatment. Appearance-based dissonance might reduce the association between health-related risk perceptions and intentions, whereas health-based dissonance might reduce the association between appearance-related risk perceptions and intentions. We were interested not only in effects of dissonance on absolute levels of risk perceptions and intentions, but also in the relationships between risk and intention.

In sum, we had two purposes in the study. The first was to test the hypothesis that when college students are made to feel cognitive dissonance about their diet and physical activity behaviors, they will be more likely to intend to adopt healthier diet and physical activity habits. It is hypothesized that dissonance would be more apparent when college students were encouraged to think about how their health habits influenced their appearance rather than their health. Dissonance participants were also expected to exhibit increased worry and perceived risk, which was expected to mediate the anticipated effect on intentions. Finally, we considered whether dissonance influences the important relationship between risk perceptions and intentions.

METHOD

PARTICIPANTS

One hundred twenty-six full-time college students (RR = 59.4%) from a school in the Southwestern United States participated in the study. The ages of the participants ranged from 18 to 23 years (mean = 20.61, SD = 1.25). Participants were enrolled as college students during the spring semester of 2005. The participants were students in a personal health course offered by the university.

DESIGN

The study was introduced as an attempt to examine how attitudes impact diet and exercise behaviors among college students and informed consent was obtained. Approval from the university's Institutional Review Board was sought and granted. Each individual was assigned randomly into one of three groups (control, dissonance-health, dissonance-appearance). The distribution of participants was relatively even in nature (Table 1). A written statement preceding the questionnaire informed the participants as to the intention of the research.

PROCEDURES

Participants were first asked to write a brief essay. The participants in the "control" condition wrote an essay about their favorite movie. Those subjected to the "dissonance-health" condition wrote the essay about why healthy diet and physical activity is integral to maintaining one's health. Participants in the "dissonance-appearance" group wrote the essay conveying why diet and exercise is necessary to maintaining an attractive physical appearance. Following completion of the essay, the participants completed questionnaires assessing 1) their present physical activity behaviors, 2) their dietary habits, 3) risk/worry they felt about their diet and exercise habits, and 4) diet/exercise intentions for the future.

INSTRUMENTS

The physical activity questionnaire assessed the number of days that the participants participated in a variety of physical activities during a normal week. It was based on a typical 7-day week, and looked for activity which lasted for 30-60 minutes. Activities included on the physical activity inventory were: intense cardiovascular exercise, weight training, moderate-intensity activity, stretching/flexibility, and active living (e.g., walking, jogging, bicycling to/from class). Questions were also adapted from The Multidimensional Health Behavior Inventory to evaluate how often participants engaged in behaviors such as aerobic activity, recreational physical activity, and stretching (Kulbok et al., 1999). The observed reliability of the data derived from these questions was acceptable ([alpha]=.72 ).

The dietary questionnaire asked the participants to indicate the number of days that they ate the recommended amount of servings from categories delineated by the new Food Guide Pyramid (U.S. Department of Health and Human Services, 2005). A copy of the pyramid was provided for the participants' reference. Participants were then asked whether they attempted to limit fat, sugar, and calories from their diet using questions adapted from The Multidimensional Health Behavior Inventory (Kulbok et al., 1999). The observed reliability of the data derived from these questions was also determined ([alpha]=.84).

Another questionnaire followed the diet and physical activity inventory assessing risk and worry as they related to health and appearance consequences of diet and exercise behaviors. Eight questions were devised to evaluate participants' risk and worry. All of the questions were rated on a 7-point scale (1 = not at all, 7 = to a great extent). Four questions evaluated risk, with two assessing risk for health problems ([alpha]=.75) and two assessing risk for diminished appearance as a result of diet and exercise behaviors ([alpha]=.85). Another four questions evaluated worry, with two assessing worry for health problems ([alpha]=.76) and two assessing worry for diminished appearance ([alpha]=.90) as a result of diet and physical activity behaviors.

After assessing risk and worry, participants were instructed to indicate whether they intended to change their diet or exercise behaviors, for the better, anytime during the next 6 months. Participants rated their intentions to change their diet and/or physical activity behaviors ([alpha]=.73) on separate 5-point scales (1 = no, not at all; 2 = probably not; 3 = maybe; 4 = likely; 5 = definitely).

DATA ANALYSIS

Upon completion of the questionnaires, comparisons were made among the control, dissonance-health, and dissonance-appearance groups on the risk, worry, and future intention items. The analysis involved three steps. In the first step, ANOVA was conducted to determine whether different treatments were associated with different levels of intention to engage in diet and physical activity behaviors. The second step used MANOVA calculations to examine whether different treatments were associated with different levels of risk perceptions of health (RiskH) and appearance (RiskA) outcomes. In the third step, multi-group comparisons, using Structural Equation Modeling, were performed to determine whether the effect of risk perceptions regarding health and appearance (RiskH, RiskA) on intention to engage in diet and physical activity behaviors (Intent) was influenced by the treatment strategies.

RESULTS

DESCRIPTIVE ANALYSIS

The means and standard deviations for all the continuous variables in each group are listed in Table 2. The corrected overall variance/covariance matrix and the correlation matrix are reported in Table 3. In order to minimize the impact of measurement error on analysis results, the variance of each continuous variable was corrected by multiplying it by the Cronbach's alpha reliability coefficient (Bollen, 1989; Goldstein, 1979).

The correlation between risk for health problems and worry for health problems was extremely high. The correlation between risk for diminished appearance (RiskA) and worry for diminished appearance (WorryA) was also extremely high, and after correction, the correlation coefficient was greater than 1.06. This suggested that, with regard to both health and appearance, risk perception and worry were linearly dependent. This linear dependence caused the sample covariance matrix and correlation matrix to be non-positive definite, which prevented further analysis on these matrices. In addition, considering the strong theoretical connection between risk and worry, the investigators only included risk perception in subsequent analyses.

EFFECT OF TREATMENT ON INTENTION TO ENGAGE IN HEALTH BEHAVIORS

One-way ANOVA analyses were completed to test the difference among the mean levels of Intent among the three treatment groups. No statistically significant effect was found (p=0.864), and treatment only explained 2% of the variability in the intention scores.

EFFECT OF TREATMENT ON RISK PERCEPTION OF HEALTH AND RISK PERCEPTION OF APPEARANCE

In order to consider the correlation between RiskH and RiskA, Multivariate Analysis of Variance (MANOVA) was used to test the effect of treatment on both variables simultaneously. The effect of treatment was found to be neither statistically, nor practically significant (Wilk's lamda=0.969, p=0.427).

EFFECT OF TREATMENT ON THE RELATIONSHIP BETWEEN RISK PERCEPTION AND INTENTION

In the first two steps, it was found that the treatment strategies did not influence participant Intent to engage in healthy behaviors, and, furthermore, did not influence their RiskH and RiskA. Further analyses examined whether the treatments influenced the relationship between risk perceptions (RiskH & RiskA) and intentions (Intent) to engage in healthy behaviors. In order to determine if there was a relationship between these variables, path analysis was used to explore the overall effect of RiskH and RiskA on Intent. Once this analysis was completed, multi-group comparisons investigated whether the effects of risk perceptions on intention were different across the three treatments. Figure 1 represents the path model that was used. The path analysis and multi-group comparisons were conducted with corrected covariance matrices in AMOS 5.0 (SPSS, Inc., 2005).

[FIGURE 1 OMITTED]

When participants were mixed ignoring their group membership, RiskH had no significant effect on Intent ([beta]=-0.062, p=0.513); however, RiskA had a significant positive effect on Intent ([beta]=0.454, p<0.001). The result indicated that, overall, one standard unit increase in RiskA resulted in a 0.45 standard unit increase in Intent. The increase in RiskH, however, did not result in an increase in Intent.

Multi-group comparisons were used to investigate whether the influence of risk perceptions on intention were different across the three treatment strategies. The corrected covariance matrices for each group, which are listed in Table 3, were extrapolated. Two steps were conducted in this multi-group comparison. First, the investigated path coefficients were freed to be estimated independently in each group. Second, the investigated path coefficients were restricted to be equal across the three groups. The difference of the chi-square values and the associated differences of the degrees of freedom between the two steps were used to test whether the equal restriction was valid (SPSS, Inc., 2005).

The null hypothesis that the influence of RiskH on Intent was the same for all the three groups was rejected at the 0.001 significance level (chi-square=22.09, p<0.001). For the control group, one standard unit change in RiskH resulted in a 0.21 standard unit decrease in Intent (p=0.046). For the dissonance-appearance group, one standard unit increase in RiskH resulted in a 0.25 standard unit decrease in Intent (p=0.003). In contrast, for the dissonance-health group, one standard unit increase in RiskH resulted in 0.30 standard unit increase in Intent (p<0.001).

The hypothesis stating an equal effect of RiskA on Intent for the three groups was also rejected at 0.05 significance level (chi-square=8.70, p=0.013). For the control group, one standard unit increase in RiskA resulted in a 0.29 standard unit increase in Intent (p=0.006). For the dissonance-health group, one standard unit increase in RiskA resulted in 0.40 standard unit increase in Intent (p<0.001). In the dissonance-appearance group, one standard unit increase in RiskA resulted in a 0.63 standard unit increase in Intent (p<0.001).

DISCUSSION

SUMMARY OF FINDINGS

The study found that the treatment strategies did not influence the intention of the participants, and did not influence their risk perceptions in terms of health and appearance respectively; however, the treatment strategies did influence the relationship between risk perceptions and intention. For the dissonance-health group, RiskH had a positive relationship, but for the dissonance-appearance and control groups, RiskH had a negative relationship with Intent. Moreover, RiskA had the largest effect on Intent in the dissonance-appearance group.

INTERPRETATION OF FINDINGS

Even though the presence of dissonance had little direct effect on intention, the effect of self-reported risk on intention was found to be statistically and practically significant. Further analysis using structural equation modeling techniques revealed two especially interesting findings related to the effect of perceived risk on intention across the three treatment groups.

The relationship between risk for diminished appearance and intention to engage in healthier behaviors was different for each treatment group, with the dissonance-appearance group having the highest correlation between risk for diminished appearance and intention to engage in healthier behaviors. In this study, this particular finding lends credence to the idea that young people are more likely to question their health behaviors when they are made to feel at risk for a diminished physical appearance because of their unhealthy behaviors. Perhaps diet and physical activity interventions targeted at young people should be construed in the same way as the photo-aging UV intervention discussed earlier (Mahler et al., 2003). Just as Mahler and colleagues made individuals feel at risk for skin deterioration due to prolonged sun exposure, so to should health promotion specialists make young people aware of the detrimental physical appearance consequences of improper eating patterns and sedentary behavior. By making young people aware of the negative appearance repercussions that will ultimately result from unhealthy eating and inactivity, perceptions of self-risk may become elevated and intentions to change health behaviors may become greater.

The relationship between risk for health problems resulting from health behaviors and intention to engage in healthier behaviors was different for each treatment group, with the control group having no statistically significant correlation, the dissonance-appearance group having a statistically significant negative correlation, and the dissonance-health group having a statistically significant positive correlation. While the three treatment groups displayed the same mean level of risk, the effect risk had on intention was different for the three groups. The effect of risk on intention to become healthier was very treatment specific. In regards to health, an effect of this magnitude may only be replicated when individuals are specifically made to feel dissonant about their health, excluding any other aspect of their life.

The finding related to health risk was exclusive to participants in the dissonance-health group, with the opposite effect occurring among those in the dissonance-appearance group. It is possible that when young people are not asked to consider the health implications of their health behaviors, they are less likely to view risk for health problems as relevant to their own life schematic. While individuals in the dissonance-appearance group may have felt a high risk for health problems, they were not made to feel dissonant about their health behaviors, so they did not report greater intentions to engage in healthier behaviors. Even though health risk levels may have been high for those subjected to the dissonance-appearance condition, the participants in this group were not challenged to consider the disjoint between attitude and behavior related to their health, which may have made any association between risk and intention negative.

LIMITATIONS

Evaluating the diet and exercise behaviors of students before presenting them with questions on risk and worry may have also been beneficial. The presence of too many students, within the study, adhering to proper diet and exercise recommendations could have caused biased results. Those participants feeling satisfied with their current diet and physical activity habits may have not felt the need to cognitively examine their behaviors. By having participants complete a pre-trial diet and exercise inventory, separate from the remainder of the questionnaire (regarding risk, worry and intention), the research design would be better able to identify unhealthy participants most vulnerable to the dissonance treatment. After the initial diet and physical activty inventory took place, then the questionnaires on risk, worry, and intention could have been distributed to at-risk individuals. Participants who exhibited sub-standard diet and physical activty behaviors would then have the greatest likelihood of being susceptible to the dissonance manipulation. The scope of the study would then be specifically targeting "unhealthy" college students.

In addition, the population of this study constituted a small convenience sample of college students which limits the generalizability of the results to all college students. The students in the study self-selected to enroll in the personal health course, which was a course required for students completing an undergraduate degree in health education. Perhaps these students were more inclined to monitor their health behaviors more closely, which could have had an impact on the results. Future studies should include a larger number of randomly selected participants. However, our study still provided some insight about the effect of cognitive dissonance on college students' intention to improve diet and physical activity behaviors. Also, there are the ethical considerations of making college students (especially female) cognizant of detrimental appearance consequences of health behavior. Females prone to eating disorders or body dismorphia may not respond positively to this type of appearance based treatment.

CONCLUSION

Utilizing the Theory of Cognitive Dissonance to elicit positive health behavior change in future interventions could be useful in improving the diet and physical activity behaviors of college students. More importantly, making college students feel greater levels of risk for a diminished physical appearance may positively effect their intentions to change their diet and physical activty behaviors. However, making college students feel greater levels of risk for health problems associated with diet and exercise behaviors may have little affect on their intentions to change health behaviors. Interventions designed to emphasize the effect of health behaviors on appearance may be appropriate for certain populations of college students.

ACKNOWLEDGEMENTS

The lead author wishes to thank members of the 2006 Society of Public Health Education Awards Committee for awarding this manuscript with the distinction of Graduate Student Paper of the Year.

REFERENCES

American College of Sports Medicine. (2000). ACSM Guidelines for Exercise Testing and Prescription. 6th ed. Baltimore, MD: Lippincott Williams & Wilkins; 2000.

Bollen, K. A. (1989). Structural equations with latent variables. New York: Wiley.

Centers for Disease Control and Prevention (2004). Behavioral Risk Factor Surveillance System. Atlanta, GA: US Department of Health and Human Services.

Centers for Disease Control and Prevention (2006). Overweight and Obesity: Home. Retrieved March 27, 2006, from http://www.cdc.gov/nccdphp/dnpa/obesity/

Draycott, S., & Dabbs A. (1998). Cognitive dissonance: An overview of the literature and its integration into theory and practice in clinical psychology. British Journal of Clinical Psychology, 37(3), 341-353.

Elliot, A. J., & Devine, P. G. (1994). On the motivational nature of cognitive dissonance: Dissonance as psychological discomfort. Journal of Personality and Social Psychology, 67(3), 382-394.

Engstrom, A., Tobelmann, R. C., & Albertson, A. M. (1997). Sodium intake trends and food choices. American Journal of Clinicial Nutrition, 65, 704S-707S.

Feingold, A. (1992). Good-looking people are not what we think. Psychological Bulletin, 111, 304-341.

Festinger, L. (1957). A theory of cognitive dissonance. Evanston, IL: Row, Peterson.

Fries, J. F. (2004). Aging Well: An Inside Look at the Compression of Morbidity. Omaha, NE: A Wellness Councils of America Interview.

Gerrard, M., Gibbons, F. X., & Bushman, B. J. (1996). The relationship between perceived vulnerability to HIV and precautionary sexual behavior. Psychological Bulletin, 119(3), 390-409.

Goldstein, H. (1979). The design and analysis of longitudinal studies: Their role in the measurement of change. New York: Academic Press.

Huang, T. K., Harris, K. J., Lee, R. E., Nazir, N., Born, W., & Kaur, H. (2003). Assessing overweight, obesity, diet, and physical activity in college students. Journal of American College Health, 52(2), 83-86.

Janz, N. K., & Becker, M. H. (1984). The health belief model: a decade later. Health Education Quarterly, 11(1), 1-47.

Klein, W. M., Blier, H. K., & Janze, A. M. (2001). Maintaining positive self-evaluations: Reducing attention to diagnostic but unfavorable social comparison information when general disregard is salient. Motivation and Emotion, 25(1), 23-40.

Kulbok, P. A., Carter, K. F., Baldwin, J. H., Gilmartin, M. J., & Kirkwood, B. (1999). The multideminsional health behavior inventory. Journal of Nursing Measurement, 7(2), 177-195.

Leary, M. R., Tchividjian, L. R., & Kraxberger, B. E. (1994). Self-presentation can be hazardous to your health: Impression management and health risk. Health Psychology, 13(6), 461-470.

Lowry, R., Galuska, D. A., Fulton, J. E., Wechsler, H., Kann, L., Collins, J. L. (2000). Physical activity, food choice, and weight management goals and practices among US college students. American Journal of Preventative Medicine, 18(1), 18-27.

Mahler, H. I., Gibbons, F. X., Gerrard, M., Kulik, J. A., & Harrell, J. (2003). Effects of appearance-based interventions on sun protection intentions and self-reported behaviors. Health Psychology, 22(2), 199-209.

Martinelli, A. M. (1999). An explanatory model of variables influencing health promotion behaviors in smoking and nonsmoking college students. Public Health Nursing, 16(4), 263-269.

National College Health Assessment (2005). Reference Group Executive Summary. Baltimore, MD: American College Health Association.

National Center for Health Statistics. (2004). National Health Interview Survey. Hyattsville, MD: Center for Disease Control and Prevention.

National Institutes of Health. (n.d.). Do you know the health risks of being overweight? Retrieved August 13, 2005, from http://win.niddk.nih.gov/publications/health_risks.htm

Nicklas, T. A., Farris, R. P., Myers, L., & Berenson, G. S. (1995). Dietary fiber intake of children and young adults: The bogalusa heart study. Journal of the American Dietetics Association, 95, 209-214.

Rogers, R. W., & Prentice-Dunn, S. (1997). Protection motivation theory. In D. Gochman (Ed.), Handbook of Health Behavior Research: Vol. 1. Personal and Social Determinants (pp. 113-132). New York: Plenum.

Schuette, L. K., Song, W. O., & Hoer, S. L. (1996). Quantitative use of The Food Guide Pyramid to evaluate dietary intake of college students. Journal of American Dietetic Association, 96, 453-457.

SPSS Inc. (2005). AMOS 5.0 for Windows. SPSS Inc., Chicago IL.

Stone, J., Aronson, E., Crain, A. L., Winslow, M. P., & Fried, C. B. (1994). Inducing hypocrisy as a means of encouraging young adults to use condoms. Personality and Social Psychology Bulletin, 20(1), 116-128.

U.S. Department of Health and Human Services. (2005). The Food Guide Pyramid. Retrieved January 26, 2005, from http://www.mypyramid.gov/pyramid/index.html

Weinstein, N. D. (1988). The precaution adoption process. Health Psychology, 7(4), 355-386.

Ziebland, S., Thorogood, M., Yudkin, P., Jones, L., & Coulter, A. (1998). Lack of willpower or lack of wherewithal? "Internal" and "external" barriers to changing diet and exercise in a three year follow-up of participants in a health check. Social Science & Medicine, 46, 461-465.

RESPONSIBILITIES AND COMPETENCIES OF HEALTH EDUCATION

Responsibility I--Assessing Individual and Community Needs for Health Education Competency A: Obtain health related data about social and cultural environments, growth and development factors, needs, and interests Competency D: Determine factors that influence learning and development

Michael Stellefson, MS, is a Doctoral Candidate in the Department of Health and Kinesiology at Texas A&M University. Zhongmiao Wang, PhD, is affiliated with the Department of Educational Psychology at Texas A&M University. William Klein, PhD, is an Associate Professor in the Department of Psychology at the University of Pittsburgh. Please address all correspondence to Michael Stellefson, MS, Department of Health and Kinesiology, Texas A&M University, MS 4243, 166 Read Building, College Station, TX 77843; PHONE: [phone]; FAX: [phone]; EMAIL: [e-mail].

Table 1. Distribution of Participants Across Conditions

Participant Condition Breakdown
Condition                 Number of Participants     Percentage (%)

Control                             41                    32.8
Dissonance-Health                   45                     36
Dissonance-Appearance               40                     32

Table 2. Means and Standard Deviations for all the Continuous
Variables in Each Group (n = 48 words)

                                                            Standard
   Group        Variable                           Mean     Deviation

Dissonance-     Risk perception of health          2.71       1.30
Health          Risk perception of appearance      4.14       1.82
                Worry about health                 3.01       1.42
                Worry about appearance             4.06       1.98
                Intention to exercise and          4.01       0.77
                eat well

Dissonance-     Risk perception of health          2.42       1.23
Appearance      Risk perception of appearance      3.42       1.92
                Worry about health                 2.74       1.45
                Worry about appearance             3.42       2.02
                Intention to exercise and          3.92       0.94
                eat well

Control         Risk perception of health          2.70       1.13
                Risk perception of appearance      3.80       1.53
                Worry about health                 3.16       1.24
                Worry about appearance             3.78       1.52
                Intention to exercise and          3.93       0.86
                eat well

Overall         Risk perception of health          2.62       1.22
                Risk perception of appearance      3.81       1.78
                Worry about health                 2.98       1.38
                Worry about appearance             3.77       1.86
                Intention to exercise and          3.96       0.85
                eat well

Table 3. Corrected Covariance Matrix and Correlation Matrix

                RiskH       RiskA       WorryH      WorryA     Intent

                 1.13
   RiskH        (1.00)
               0.91 **       2.67
   RiskA        (0.52)      (1.00)
               1.22 **     1.44 **       1.43
  WorryH        (0.96)      (0.74)      (1.00)
               0.83 **     3.08 **     1.30 **       3.13
  WorryA        (0.44)      (1.07)      (0.61)      (1.00)
                 0.13      0.50 **      0.25 *     0.51 **      0.52
 Intention      (0.17)      (0.42)      (0.29)      (0.39)     (1.00)

Note: 1. Numbers in parenthesis are correlation coefficients.

      2. ** p<0.01, * p<0.05

      3. RiskH = Risk for negative health outcomes
         RiskA = Risk for negative appearance outcomes
         WorryH = Worry for negative health outcomes
         WorryA = Worry for negative appearance outcomes

Table 4. Corrected Covariance Matrix for Each Group

             RiskH      RiskA     WorryH     WorryA     Intent

                           Dissonance-Health

 RiskH       1.26
 RiskA       0.94       2.84
 WorryH      1.40       1.48       1.65
 WorryA      0.78       3.27       1.15       3.63
 Intent      0.36       0.60       0.36       0.58       0.42

                          Dissonance-Appearance

 RiskH       0.98
 RiskA       0.84       2.90
 WorryH      1.17       1.57       1.54
 WorryA      0.80       3.77       1.55       3.53
 Intent      0.05       0.70       0.26       0.74       0.69

                                 Control

 RiskH       1.13
 RiskA       0.88       2.14
 WorryH      1.06       1.25       1.10
 WorryA      0.85       2.14       1.18       2.13
 Intent      -0.04      0.18       0.12       0.20       0.50

Note: 1. RiskH = Risk for negative health outcomes
         RiskA = Risk for negative appearance outcomes
         WorryH = Worry for negative health outcomes
         WorryA = Worry for negative appearance outcomes