Invoking identity to promote behavior change: a randomized controlled trial and three experimental tests

Abstract

Background: Although role identity predicts behavioral intentions in correlational studies, there have been relatively few experimental tests of the impact of identity invocation on health-related intentions or behavior. A randomized controlled trial tested whether survey items that invoked a relevant identity (“being an organ donorâ€) increases organ donor registration relative to items concerning the focal behavior (“organ donationâ€), or demographic items only. Methods: A nationally representative UK sample (N = 1,508) was randomly assigned to the three conditions, and completed 10 items concerning the identity or the behavior. Whether or not participants visited the national donor registration website was recorded. Findings: Contrary to expectations, participants generally had more favorable views of organ donation than being an organ donor. There was no reliable effect of condition on behavior (website visit rates = 2.5% to 3.8%). Three subsequent experiments concerning dietary behavior, physical activity, and alcohol consumption (Ns = 279, 162, and 335, respectively) also observed null effects of identity invocation. Discussion: Invoking a relevant identity does not appear to promote health-related intentions or action, at least for the behaviors examined here.