The ability to attribute intentions to others is a hallmark of human social cognition but is altered in paranoia. Paranoia is the most common positive symptom of psychosis but is also present to varying degrees in the general population. Epidemiological models suggest that psychosis risk is associated with low social rank and minority status, but the causal effects of status and group affiliation on paranoid thinking remain unclear. We examined whether relative social status and perceived group affiliation, respectively, affect live paranoid thinking using two large-N (N = 2030), pre-registered experiments. Interacting with someone from a higher social rank or a political out-group led to an increase in paranoid attributions of harmful intent for ambiguous actions. Pre-existing paranoia predicted a general increase in harmful intent attribution, but there was no interaction with either type of social threat: highly paranoid people showed the same magnitude of increase as non-paranoid people, although from a higher baseline. We conclude social threat in the form of low social status and out-group status affects paranoid attributions, but ongoing paranoia represents a lowered threshold for detecting social threat rather than an impaired reactivity to it.

Fear of others intending harm is central in paranoia but it is unclear to what extent social context alters paranoid attributions, and how this interacts with pre-existing paranoia. We examined social interactions via game theory paradigms across social rank (Experiment 1) and political group affiliation (Experiment 2) as a function of pre-existing paranoia using two pre-registered Dictator Games (N=2,030) for real money. Interacting with someone from a higher social rank or a political out-group led to an increase in paranoid attributions of harmful intent for ambiguous actions. Pre-existing paranoia predicted a general increase in harmful intent attribution but there was no interaction with social situation: highly paranoid people showed the same magnitude of increase to non-paranoid people, although from a higher baseline. We conclude that social context affects paranoid attributions but ongoing paranoia represents a lowered threshold for detecting social threat rather than an impaired reactivity to it.

No abstract