Current theories argue that hyper-sensitisation of social threat perception is central to paranoia. Affected people often also report misperceptions of group cohesion (conspiracy) but little is known about the cognitive mechanisms underpinning this conspiracy thinking in live interactions. In a pre-registered experimental study, we used a large-scale game theory approach (N > 1,000) to test whether the social cohesion of an opposing group affects paranoid attributions in a mixed online and lab-based sample. Participants spanning the full population distribution of paranoia played as proposers in a modified Trust Game: they were allocated a bonus and chose how much money to send to a pair of responders which was quadrupled before reaching these responders. Responders decided how much to return to the proposers through the same process. Participants played in one of two conditions: against a cohesive group who communicated and arrived at a joint decision, or a non-cohesive group who made independent decisions. After the exchange, proposers rated the extent to which the responders' decisions were driven by (i) self-interest and (ii) intent to harm. Although the true motives are ambiguous, cohesive responders were reliably rated by participants as being more strongly motivated by intent to harm, indicating that group cohesion affects social threat perception. Highly paranoid participants attributed harmful intent more strongly overall but were equally reactive to social cohesion as other participants. This suggests that paranoia involves a generally lowered threshold for social threat detection but with an intact sensitivity for cohesion-related group characteristics.
Keywords: Conspiracy; Group cohesion; Paranoia; Trust game.