Combining Self-Affirmation With the Extended Parallel Process Model: The Consequences for Motivation to Eat More Fruit and Vegetables

Abstract

There is potential for fruitful integration of research using the Extended Parallel Process Model (EPPM) with research using Self-affirmation Theory. However, to date no studies have attempted to do this. This article reports an experiment that tests whether (a) the effects of a self-affirmation manipulation add to those of EPPM variables in predicting intentions to improve a health behavior and (b) self-affirmation moderates the relationship between EPPM variables and intentions. Participants (N = 80) were randomized to either a self-affirmation or control condition prior to receiving personally relevant health information about the risks of not eating at least five portions of fruit and vegetables per day. A hierarchical regression model revealed that efficacy, threat × efficacy, self-affirmation, and self-affirmation × efficacy all uniquely contributed to the prediction of intentions to eat at least five portions per day. Self-affirmed participants and those with higher efficacy reported greater motivation to change. Threat predicted intentions at low levels of efficacy, but not at high levels. Efficacy had a stronger relationship with intentions in the nonaffirmed condition than in the self-affirmed condition. The findings indicate that self-affirmation processes can moderate the impact of variables in the EPPM and also add to the variance explained. We argue that there is potential for integration of the two traditions of research, to the benefit of both.

Notes

3. ¹Participants were also given the opportunity to take away with them two leaflets, one giving tips on how to eat more fruit and vegetables and one on portion sizes. Consistent with previous findings in which self-affirmation promoted such behaviors as leaflet taking, there was a significant main effect of self-affirmation on leaflet taking, with self-affirmed participants taking significantly more leaflets (M = 1.58, SD = .68) than did nonaffirmed participants (M = 1.01, SD = .90), F(1, 78) = 7.13, p = .009, η² = .08. Because EPPM variables did not predict this behavior, it is ancillary to the focus of the current article and not discussed further here.

4. ²Adding the self-affirmation × threat × efficacy term at a fifth step did not add to the variance explained nor change any of the significant effects. The three-way interaction did not approach significance, ß = −.02, p = .85.

5. ³This effect of self-affirmation on the efficacy−intention relationship has positive and negative potential for subsequent behavior: positive, because more people with lower efficacy may attempt to eat more fruit and vegetables than would have done without self-affirming; negative, because people may attempt what they cannot achieve and thus become demoralized.