Do people sincerely believe conspiracy theories that they endorse?

Surveys are regularly used to estimate the prevalence of belief in conspiracy theories and test hypotheses about causes and consequences of these beliefs. In such research, it is typically assumed that belief reports are sincere. However, evidence for sincerity is rarely provided. In the present study we examine the issue of participant sincerity in a sample of 1,044 Australians demographically matched to the voting population. We find endorsement of six preexisting conspiracy theories to be widespread, with 33.4% of participants (n = 349) endorsing at least one and 5.0% (n = 52) endorsing all of them. However, we identified evidence that not all participants who endorsed these conspiracy theories sincerely believe them. Strikingly, we found that of the 19.6% of participants (n = 205) who failed at least one of two sincerity checks, 63.9% endorsed at least one conspiracy theory (n = 131) and 26.0% (n = 51) endorsed all six, while of the 80.4% of participants (n = 839) who passed both sincerity checks, only 26.0% endorsed at least one conspiracy theory (n = 218) and only 0.1% endorsed all six (n = 1). Overall, our results suggest that a substantial minority of participants answering surveys endorse conspiracy theories that they do not sincerely believe.