The Relative Persuasiveness of Gain-Framed Loss-Framed Messages for Encouraging Disease Prevention Behaviors: A Meta-Analytic Review

Abstract

A meta-analytic review of 93 studies (N = 21,656) finds that in disease prevention messages, gain-framed appeals, which emphasize the advantages of compliance with the communicator's recommendation, are statistically significantly more persuasive than loss-framed appeals, which emphasize the disadvantages of noncompliance. This difference is quite small (corresponding to r = .03), however, and appears attributable to a relatively large (and statistically significant) effect for messages advocating dental hygiene behaviors. Despite very good statistical power, the analysis finds no statistically significant differences in persuasiveness between gain- and loss-framed messages concerning other preventive actions such as safer-sex behaviors, skin cancer prevention behaviors, or diet and nutrition behaviors.

Thanks to Lisa Benz Scott, Debbie Chang, Eamonn Ferguson, Cynthia Hoffner, Rama Jayanti, Amy Latimer, Deanna Lawatsch, Angela Lee, Norman Mundorf, Robin Nabi, Sherri Robertson, Rob Ruiter, Peter Salovey, Lijiang Shen, Patricia van Assema, Judith Weiner, and Holley Wilkin for supplying primary-research information.

Notes

^a The coding judgments, in order, follow: specific prevention behavior (1 = diet/nutrition, 2 = skin cancer prevention, 3 = safer-sex behavior, 4 = dental hygiene behavior, 5 = exercise behavior, 6 = smoking, 7 = inoculation/vaccination, 8 = other prevention behaviors); gain kernel-state language (1 = desirable states, 2 = undesirable states, 3 = both desirable and undesirable states, 4 = indeterminate); loss kernel-state language (1 = undesirable states, 2 = desirable states, 3 = both desirable and undesirable states, 4 = indeterminate).

∗p < .05

∗∗p < .01

∗∗∗p < .001.

^a These are power figures for detecting a population effect size of r = .10, assuming large heterogeneity, with a random-effects analysis, .05 alpha, and a two-tailed test (Hedges & Pigott, 2001)

References marked with an asterisk indicate studies included in the meta-analysis.