Matt Williams, Mathew D. Marques, Mathew Ling, Stephen Hill, John Kerr, Edward J. R. Clarke, Ahnya Martin, Robert Ross
Abstract Recent research suggests that some survey respondents who endorse conspiracy theories are responding insincerely. However, it is unclear to what extent this problem may distort important empirical findings with respect to conspiracy theories. In this registered report, we recruited a demographically representative quota sample of the New Zealand population (n = 810), and probed sincerity by two methods: presenting a farcically bizarre conspiracy theory and explicitly asking respondents if they had been insincere in any of their responses. We found that respondents classified as insincere according to either of these methods displayed substantially higher levels of conspiracy theory endorsement. Respondents who displayed signs of insincerity were also much more likely to endorse contradictory conspiracy theories. Furthermore, we found that the relationships between endorsement of conspiracy theories and well-established predictors (paranoia, belief in a dangerous world, cognitive reflection, and trust, but not anxiety) had significant interactions with insincerity. While some uncertainty remains about the validity of the methods that we used to detect insincerity, our findings suggest that insincere responses may distort empirical findings relating to belief in conspiracy theories. Researchers should not assume that every survey respondent who endorses a conspiracy theory believes that theory.