A test of conversational and testimonial messages versus didactic presentations of nutrition information

Objective: Conversational and testimonial approaches to presenting messages relevant to fruit and vegetable consumption were compared with a traditional didactic message. Design: A Greco-Latin square, within-subjects experimental design was employed with 3 formats and 3 topics. Setting: Community centers. Participants: Thirty-one ethnically diverse adults. Intervention: Presented conversational, testimonial, and didactic nutrition messages. Variables Measured: Believability, clarity, and perceived usefulness of messages presented; identification with sources in message; and self-efficacy with respect to produce consumption. Analysis: Analysis of variance. Results: Conversational formats were perceived as more believable than the control newsletter article. No differences were found with respect to clarity, usefulness, or self-efficacy. Identification with characters did significantly interact with message format in predicting the message assessment variables. However, the conversation and testimonial versions did not increase identification with persons in the message over the didactic presentation. Conclusions and Implications: Audience response to narrative messages is contingent upon identification with persons portrayed in the message. The results underscore the need for careful pretesting of involvement with the story and identification with persons portrayed to maximize the effectiveness of narrative messages in health interventions.