JMIR Publications

ABSTRACT

Background:

In an effort to signal the authenticity of user accounts, social networking sites (SNS) such as Facebook and Twitter utilize visual heuristics (blue checkmarks) to signify whether accounts are verified. While verification markers are generally well recognized (and often coveted) by SNS users, relatively little is known about how they affect users’ perceptions of accuracy or their likelihood of engaging with online information. This is particularly true in the case of information posted by medical experts and healthcare professionals.

Objective:

This study employs an experimental survey design to assess the affect of verification markers on SNS users’ assessments accuracy, as well as their proclivity to recirculate health information or “follow” verified medical experts in their social network.

Methods:

A survey experiment, utilizing random assignment, was conducted on a representative sample of 534 adult SNS users in the State of Florida. Two separate experimental scenarios exposed users to vaccine-related posts from verified medical experts on Twitter. In each case, the original post contained a platform issued verification marker (treatment group) which was subsequently edited out of the image as an experimental control. For each scenario, respondents were randomly assigned to either the treatment or control group, and responses to three follow-up questions were assessed through a series of chi-square analyses, as well as two logit regression models. Responses were fielded using a stratified, quota sampling approach to ensure representativeness of the state’s population based on age, gender, race, ethnicity, and political affiliation.

Results:

Users assessments of information accuracy were not significantly impacted by the presence or absence of verification markers, and users exposed to the experimental treatment (verification marker) were not significantly more likely to repost the message or “follow” the author. While verification makers did not influence user’ assessments or subsequent behaviors, reliance on social media for health-related information and political affiliation were significant predictors of accuracy assessments in both experimental scenarios. In Scenario 1 – which included a post addressing COVID-19 vaccine efficacy – users who relied on social media “a great deal” for health information were 2 times more likely to assess the post as accurate (OR = 2.033, 95% CI = 1.129 – 3.661, P = .018). In Scenario 2 – which included a post about measles vaccines – registered Republicans were nearly 6 times less likely to assess the post as accurate (OR = 0.171, 95% CI = 0.097 – 0.299, P ≤ .001).

Conclusions:

For health professionals and medical experts wishing to leverage social networks to combat misinformation and spread reliable health-related content, verification markers appear to offer little by way of added value. Based on prior research, other heuristics and communication strategies are likely to yield better results. Clinical Trial: n/a