A Randomized Trial of Immunization Messaging on Parental Intentions to Immunize Children Against Influenza

Background: Recent outbreaks of vaccine-preventable illnesses and an increase in vaccination exemption rates highlight the importance of designing effective immunization campaigns. This experimental study examines how pro-vaccine flu messages, guided by the Extended Parallel Process Model, affect cognitive processing surrounding the decision to vaccinate children.

Methods: Parents of children six months to five years old in the United States (N = 975) were randomly exposed to one of four high-threat/high-efficacy messages (narrative, statistical, combined, or control). After reading the message, parents completed a follow-up survey assessing perceptions of authority, character, relevance, emotion, attention, manipulation, intention to seek more information, and intention to vaccinate their child.

Results: Multiple linear regression detected several variables that significantly (p < .05) predicted parental intentions. Relevance positively predicted, and manipulation negatively predicted, both intention to seek information and intention to vaccinate. Attention and emotion positively predicted information seeking, while character positively predicted intention to vaccinate. ANOVA and post-hoc Tukey tests detected significant (p < .05) differences between the four message conditions; the statistical message showed lower ratings of emotion and attention, while the narrative message showed lower ratings of authority and character and higher manipulation ratings. No significant difference was found for relevance (p = .06), intention to seek information (p = .14), or intention to vaccinate (p = .89).

Conclusion: There is evidence that presenting messages containing narratives and statistics together will improve cognitive processing of health messages. The results of this study have important implications for the design of health communication campaigns.