Abstract
When predicting the future, people tend to overestimate the intensity of their emotions, a phenomenon known as the impact bias. Design: In two studies, we examined the impact bias for health outcomes. In Study 1, participants were randomized to think about a negative health outcome in the future or one in the past. In Study 2, participants came to the laboratory and were asked to predict and report their emotions surrounding an actual health outcome (consuming an unhealthy food). Results: In both studies, an impact bias emerged. In Study 1, participants thinking about an outcome in the future estimated more negative emotion than those thinking about an outcome in the past. In Study 2, when facing an actual health outcome, participants anticipated more negative and positive emotion than they experienced. Impact biases were also associated with behavioral motivation – desire to change the outcome (Study 1) and increased preventive intentions (Study 2). Additional analyses revealed that regret was a particularly important emotion. Conclusion: Although research has highlighted an impact bias for severe health outcomes like disease, these studies provide evidence of an impact bias for health outcomes generally. They also suggest that the bias may have implications for behavior intentions.
Acknowledgements
We thank Matthew Fallon and Adrian Hernandez for their assistance with data collection. This research was funded, in part, by the Grand Valley State University McNair Scholars Program.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
1 Of these 14 participants, nearly one-half (n = 6) of them wrote about both eating poorly and not exercising, so we combined these into one category.
2 We also conducted analyses (ANOVAs) controlling for the health behavior outcome participants reported. The condition effects on negative emotion and desire to change are similarly significant when entering this covariate.
3 We conducted one-way ANOVAs to test if there were differences for the three foods with respect to anticipated and experienced negative and positive emotions. These analyses revealed no differences among the three foods for any of these emotions, all Fs < 1. Additionally, we examined if the three foods differed for consumption, and similarly found no significant differences, F(2,75) = 1.31, p = .275.
4 Although the groups had unequal sample sizes, the Levene’s test revealed no violation of homogeneity for either food consumed or intentions.