Abstract
This study applied the control theory explanation of memorable message to H1N1 prevention. Participants (N = 204) described an H1N1 prevention memorable message and then evaluated it along the dimensions of argument, directness, and vulnerability; they also completed measures of positive and negative health self-concept from the Health-Related Self-Concept scale, general health behaviors, and flu-related health behaviors. Descriptively, campus and mass media were the most common sources of H1N1 messages, and wash hands, avoidance, and getting shot/vaccine were the most common contents. Correlation analyses revealed that positive health self-concept was positively related to the argument message dimension and flu-prevention behaviors; negative health self-concept was negatively related to the argument message dimension and hand washing. Further, the interaction of self-image-actual behavior discrepancy and message explicitness fell just short of conventional significance levels in predicting hand washing behavior. These results provide suggestive evidence supporting the control theory explanation of memorable messages.
Acknowledgments
The authors would like to thank Marquita Berry for assistance with data collection and Dr. Chris Carpenter for helpful feedback on an earlier draft of the article.
Notes
Note. N = 174 for point-biserial correlations.
*p < .05. **p < .01.
Note. N = 174 for correlations with message dimensions. HPM = Health Protection Motivation; HRH = Health-Risky Habits.
*p < .05. **p < .01.
*p < .05. **p < .01. † = .06.
Note. *p < .05. **p < .01.
In the time frame posited by the survey, the flu shot item should be a dichotomous variable. However, 18 participants marked 2 for this item, and 20 marked a 3 for the item. It is impossible to know if they misunderstood/overlooked the time frame referred to in the directions or were referring to different kinds of shots, or something else entirely. For exploratory purposes, the item has been left as reported by respondents.
In the interests of economy, statistical details of these preliminary analyses have been omitted. Results are available from the first author.
Follow-up analysis reveals that, if the entire sample were used, the interaction would be significant, beta = .20, p < .05.