The resilience of attitude toward vaccination: A web-based randomized controlled trial on the processing of online (mis)information.

Abstract

Before the COVID-19 pandemic, it was already recognized that internet-based misinformation and disinformation could influence individuals to refuse or delay vaccination for themselves, their families, or their child(ren). Reinformation, which refers to hyperpartisan and ideologically biased content, can propagate polarizing messages on vaccines, thereby contributing to vaccine hesitancy even if it is not outright disinformation. This study aimed to evaluate the impact of online reinformation on vaccine hesitancy. Specifically, our goal was to investigate how misinformation presented in the style and layout of a news article could influence the perceived tentativeness (credibility) of COVID-19 vaccine information and confidence in COVID-19 vaccination. We conducted a web-based randomized controlled trial (RCT) by recruiting English-speaking Canadians aged 18 years and above from across Canada through the Qualtrics paid opt-in panel system. Participants were randomly assigned to one of four distinct versions of a news article on COVID-19 vaccines, each featuring variations in writing style and presentation layout. After reading the news article, participants self-assessed the tentativeness of the information provided, their confidence in COVID-19 vaccines, and their attitude toward vaccination in general. The survey included 537 participants, with 12 excluded for not meeting the task completion time. The final sample comprised 525 participants distributed about equally across the four news article versions. Chi-square (χ2) analyses revealed a statistically significant association between general attitude toward vaccination and the perceived tentativeness of the information about COVID-19 vaccines included in the news article (χ21 = 37.79, P<.0001). The effect size was small to moderate, with Cramer's V = 0.27. An interaction was found between vaccine attitude and writing style (χ²₁ = 6.17, P= 0.01), with a small effect size, Cramer's V = 0.11. Additionally, a Pearson correlation revealed a significant moderate-to-strong correlation between perceived tentativeness and confidence in COVID-19 vaccination, r(523) = .48, P<.0001. The coefficient of determination (r²) was 0.23, indicating that 23% of the variance in perceived tentativeness was explained by confidence in COVID-19 vaccines. In comparing participants exposed to a journalistic-style news article with those exposed to an ideologically biased article, Cohen's d was calculated to be 0.38, indicating a small to medium effect size for the difference in the perceived tentativeness between these groups. Exposure to an online news article conveying misinformation may not be sufficient to change an individual's level of vaccine hesitancy. The study reveals that the predominant factor in shaping individuals' perception of COVID-19 vaccines is their attitude toward vaccination in general. This attitude also moderates the influence of writing style on perceived tentativeness; the stronger one's opposition to vaccines, the less pronounced the impact of writing style on perceived tentativeness. The study was part of a broader initiative with a registered protocol. The protocol was recorded on October 17, 2022, under the International Registered Report Identifier (IRRID): DERR1-10.2196/41012. RR2-10.2196/41012.