Unreliable conspiracies: survey results about COVID-19 conspiracy theories lack temporal stability

Introduction Conspiracy theories complicating public reaction to the COVID-19 pandemic inspired quantitative research on conspiracy theories, mostly using survey-based, correlational designs. Data from similar studies may, however, be unreliable due low temporal stability (Graham, 2021). Objectives We examine of a popular survey measure beliefs (CCBs). Methods CCBs were measured by set items developed in first months pandemic, addressing that was hoax (CCH) and it artificially created for evil purposes (CCC) (Imhoff & Lamberty, 2020), 179 students medicine. In March 2022, twice same questions presented once with numeric (N1 measure) Lickertian (L1 scale, filler between. The sample May 2022 (L2 measure). Results mean agreement did not differ between previous January 2021. expressed as correlation L1 L2 measurement poor (r = .57 CCC, r .67, CCH). difference positively correlated =.21, p < .01 CCH; .44; .001 CCC). Out 18 respondents reporting CCC 5 CCH, only 8 still reported 1 CCH May. Finally, participants split based their score into groups “mainstreamers”, “undecided”, “conspirators”. For there no recorded N1, L1, L2. versus N1=L2, suggesting random effects (regression mean). “conspirators”, scores equal, while lower May, is better explained situational factors. Conclusions Temporal survey-reported low, particularly among those CCBs: When respondent reports CCB survey, they are more likely disagree than agree two later. seems affected incorrect or answers, but also Implications: First, measures may inflate spread population. Second, correlations other variables surveys inflated via common method bias , distorting our understanding predictors consequences CCB. Disclosure Interest None Declared