Addressing COVID-19 Misinformation on Social Media Preemptively and Responsively

Emily K. Vraga; Leticia Bode

Disclosures

Emerging Infectious Diseases. 2021;27(2):396-403. 

In This Article

Results

Wave 1

First, we tested the effects of correction on misperceptions related to the effects of a hot bath on body temperature and COVID-19 prevention for wave 1. We limited these regression analyses to the 1,543 persons who passed a premanipulation attention check (Appendix 4, https://wwwnc.cdc.gov/EID/article/27/2/20-3139-App4.pdf). Exposure to the WHO graphic in any condition reduced misperceptions that a hot bath will raise body temperature as compared with the control, but had no effects on misperceptions that a hot bath will prevent COVID-19 infection (Table 1). When comparing the types of correction to each other, we found no differences by either source or placement, nor by the interaction between the 2 categories (Table 2). In other words, corrections were equally effective for body temperature misperceptions (and ineffective for COVID-19 prevention misperceptions) whether they came from a user or from WHO and when they were preemptive as well as responsive.

Wave 2

We replicated these analyses with the 1,110 participants who completed the follow-up survey and passed the attention check for wave 2 (12 participants failed the attention check in wave 2), controlling for the amount of time between taking the 2 waves of the survey. We found that exposure to the WHO preemptive, WHO responsive, or user responsive corrections all produced lower misperceptions than the control condition at wave 2 for body temperature misperceptions (Table 3). We also found that those exposed to the WHO responsive correction had significantly lower COVID-19 prevention misperceptions 1 week later than those in the control condition; results showed an average decline of 11% in COVID-19 prevention misperceptions from the control to the WHO responsive correction. However, the overall model predicting COVID-19 misperceptions was not significant, meaning that there were no differences in means averaged across the 6 experimental conditions even though there was a significant difference in directly comparing the WHO responsive correction to control condition, so this result must be interpreted with caution. We again found no significant differences in either type of misperceptions based on the source of the graphic (WHO versus Facebook user) or whether it was offered preemptively or responsively (Table 4).

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....