Investigating and Improving the Accuracy of US Citizens' Beliefs about the COVID-19 Pandemic: Longitudinal Survey Study

Aart Van Stekelenburg*, Gabi Schaap, Harm Veling, Moniek Buijzen

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review

24 Citations (Scopus)
26 Downloads (Pure)



The COVID-19 infodemic, a surge of information and misinformation, has sparked worry about the public's perception of the coronavirus pandemic. Excessive information and misinformation can lead to belief in false information as well as reduce the accurate interpretation of true information. Such incorrect beliefs about the COVID-19 pandemic might lead to behavior that puts people at risk of both contracting and spreading the virus. 


The objective of this study was two-fold. First, we attempted to gain insight into public beliefs about the novel coronavirus and COVID-19 in one of the worst hit countries: The United States. Second, we aimed to test whether a short intervention could improve people's belief accuracy by empowering them to consider scientific consensus when evaluating claims related to the pandemic. 


We conducted a 4-week longitudinal study among US citizens, starting on April 27, 2020, just after daily COVID-19 deaths in the United States had peaked. Each week, we measured participants' belief accuracy related to the coronavirus and COVID-19 by asking them to indicate to what extent they believed a number of true and false statements (split 50/50). Furthermore, each new survey wave included both the original statements and four new statements: Two false and two true statements. Half of the participants were exposed to an intervention aimed at increasing belief accuracy. The intervention consisted of a short infographic that set out three steps to verify information by searching for and verifying a scientific consensus. 


A total of 1202 US citizens, balanced regarding age, gender, and ethnicity to approximate the US general public, completed the baseline (T0) wave survey. Retention rate for the follow-up waves-first follow-up wave (T1), second follow-up wave (T2), and final wave (T3)-was high (≥85%). Mean scores of belief accuracy were high for all waves, with scores reflecting low belief in false statements and high belief in true statements; the belief accuracy scale ranged from-1, indicating completely inaccurate beliefs, to 1, indicating completely accurate beliefs (T0 mean 0.75, T1 mean 0.78, T2 mean 0.77, and T3 mean 0.75). Accurate beliefs were correlated with self-reported behavior aimed at preventing the coronavirus from spreading (eg, social distancing) (r at all waves was between 0.26 and 0.29 and all P values were less than .001) and were associated with trust in scientists (ie, higher trust was associated with more accurate beliefs), political orientation (ie, liberal, Democratic participants held more accurate beliefs than conservative, Republican participants), and the primary news source (ie, participants reporting CNN or Fox News as the main news source held less accurate beliefs than others). The intervention did not significantly improve belief accuracy. 


The supposed infodemic was not reflected in US citizens' beliefs about the COVID-19 pandemic. Most people were quite able to figure out the facts in these relatively early days of the crisis, calling into question the prevalence of misinformation and the public's susceptibility to misinformation.

Original languageEnglish
Article numbere24069
Number of pages15
JournalJournal of Medical Internet Research
Issue number1
Publication statusPublished - 12 Jan 2021

Bibliographical note

Publisher Copyright:
© 2021 Journal of Medical Internet Research. All rights reserved.

©Aart van Stekelenburg, Gabi Schaap, Harm Veling, Moniek Buijzen. Originally published in the Journal of Medical Internet Research (, 12.01.2021.

Research programs



Dive into the research topics of 'Investigating and Improving the Accuracy of US Citizens' Beliefs about the COVID-19 Pandemic: Longitudinal Survey Study'. Together they form a unique fingerprint.

Cite this