Source Credibility and the Biasing Effect of Narrative Information on the Perception of Vaccination Risks

Abstract

Immunization rates are below the Global Immunization Vision and Strategy established by the World Health Organization. One reason for this are anti-vaccination activists, who use the Internet to disseminate their agenda, frequently by publishing narrative reports about alleged vaccine adverse events. In health communication, the use of narrative information has been shown to be effectively persuasive. Furthermore, persuasion research indicates that the credibility of an information source may serve as a cue to discount or augment the communicated message. Thus, the present study investigated the effect of source credibility on the biasing effect of narrative information regarding the perception of vaccination risks. 265 participants were provided with statistical information (20%) regarding the occurrence of vaccine adverse events after vaccination against a fictitious disease. This was followed by 20 personalized narratives from an online forum on vaccination experiences. The authors varied the relative frequency of narratives reporting vaccine adverse events (35% vs. 85%), narrative source credibility (anti-vaccination website vs. neutral health forum), and the credibility of the statistical information (reliable data vs. unreliable data vs. control) in a between-subjects design. Results showed a stable narrative bias on risk perception that was not affected by credibility cues. However, narratives from an anti-vaccination website generally led to lower perceptions of vaccination risks.

Acknowledgments

The authors are grateful to Karl-Philipp Henschelmann and Philipp Schmid for their help in conducting the study as well as Heather Fuchs for helpful comments on a previous draft of this article. The authors also gratefully acknowledge http://www.sparbaby.de, http://www.onmeda.de, and http://estudy.zpid.de/ for posting a link to the study on their websites.

Notes

¹The time required to complete the study was calculated as the time between having read the initial instructions and having answered the last of the manipulation checks (recall of forum name). The numeracy test was excluded from this calculation, as the seven items took approximately half as long to complete as the entire rest of the survey, with a very large standard deviation, that is, 6 min 9 s (4 min 20 s).

²In the laboratory sample, the mean (SD) age was 22.69 (3.77) years. 81 participants (96.4%) reported an Abitur grade with a sample mean (SD) of 2.15 (0.44). Grades on this exam range between 1.0 and 4.0, with 1.0 being the best possible grade. The mean time (SD) to complete the study was 10 min 24 s (2 min 16 s). In the online sample, the mean (SD) age was 31.07 (5.09) years. Thirty-three participants (18.2%) did not complete the Abitur, 46 (25.4%) completed the Abitur, 92 (50.8%) received a higher education degree, and 10 (5.5%) received a PhD. The mean time (SD) to complete the study was 8 min 51 s (4 min 9 s).

³We used the SPSS macro PROCESS provided by Andrew Hayes at http://afhayes.com.