Reexamining Health Messages in the Digital Age: A Fresh Look at Source Credibility Effects

Abstract

This article investigates the influence of perceived source credibility on the effectiveness of health-related public service announcements (PSAs) and electronic word-of-mouth (eWOM) communications. Findings indicate that online commenters who are perceived to be credible are instrumental in influencing consumers’ responses to pro- versus antivaccination online PSAs. Results further suggest it is not the advertising message (i.e., the PSA-advocated position) alone that influences consumers’ responses (even when consumers perceive the PSA sponsor to be highly credible) but rather the commenters’ reactions to the claims presented in the PSA that also independently contribute to consumers’ vaccination attitudes and behavioral intentions. Finally, results also show that when the relevant expertise of online commenters is identified, the effectiveness of the PSA's advertising message is moderated by the interactive effect of the online comments and their associated perceived credibility.

Notes

1. Given the sensitive nature of our manipulations, which dealt with the ongoing national debate regarding the pros and cons of vaccinations, all participants were debriefed twice following data collection. First, a few days after participation, individuals received an e-mail directing them to three different impartial websites (e.g., http://www.vaccines.gov). Participants were strongly encouraged to visit these sites to obtain reliable information regarding vaccinations and were reminded to not base their own or their family's future vaccination decisions on any of the information that was contained in the experimental stimuli. Second, several weeks later, they received another e-mail providing additional vaccine-related information and were encouraged to contact the authors if they had any questions or concerns. This project was evaluated and approved by the institutional review board (IRB) of the authors’ university.

2. This reduction yielded cell sizes ranging from 47 to 54 participants. Analyses reported in Table 1 were also conducted with the total sample of participants (i.e., including respondents who failed either of the two manipulation checks). A similar pattern of results was obtained, whereby both interactive terms in each regression analysis were highly significant (p < .001) for all four dependent measures.

3. This reduction in sample size yielded an N = 174 for the medical doctor versus lobbyist comparison (with cell sizes ranging from 35 to 49) and N = 196 for the medical doctor versus student comparison (with cell sizes ranging from 44 to 54). Analyses reported in Tables 3A and 3B were also conducted with the total sample of participants (i.e., including respondents who failed any of the three manipulation checks); a similar pattern of results was obtained.