4.1.2.1. Baseline session (equal probabilities)

Non-social intentions: As previously shown in Chambon et al. (xxxx), healthy participants’ RTs decreased and hits improved as the amount of visual information increased. A similar effect was found in schizophrenic patients (all participants, both motor and Superord. tasks: F(2,96)>251.1, p<.001). Moreover, both groups showed no significant differences in correct responses between all types of intentions presented, indicating that prior to biasing participants, there wasn’t any a priori preference for one intention over the other ones (both tasks: F(1,48)>.01, p>.85). Finally, we didn’t find any interaction between the amount of visual information and the bias in both comparison and schizophrenic participants, showing that improved participants’ performance for higher amounts of information was independent of the type of intention presented (both tasks: F(2,96)>2.3, p>.12).

Social intentions: As for non-social intentions, a significant main effect of the amount of visual information was obtained on participants’ hits and RTs, in both comparison and schizophrenic participants (both social motor and social superord. tasks: F(2,94-96)>159.9, p<.001). However, we found groups’ differences in the main effect of the factor ‘Bias’: while healthy participants showed an early preference (i.e. prior to being biased) for ‘tit-for-tat’ mode of reciprocity, patients did not. In the social superordinate task, a significant effect of the interaction Group*Bias was thus obtained for hits rate (F(1,48)=6.3, p=.01), with patients showing no significant preference for intentions congruent with a ‘tit-for-tat’ strategy compared to the other ones (e.g. ‘always cooperate’, ‘always defect’ strategies). In the social motor task, this interaction did not reach significance however (p=.09).