4.2. Bias effect

There was no main effect of the group, indicating that, regardless to the type of intention, the effect of the bias was as important in patients as in healthy participants (F(1,191)=2.1, p=.15). The main effect of the Amount of visual information was significant, with an effect of the bias progressively increasing as the amount of information decreased (F(2,382)=93.8, p<.001). We also observed a main effect of the Type of intention (F(3,191)=11.7, p<.001), accounting for a greater influence of the bias for inferring superordinate vs. motor intentions, whatever the type of target considered (social as non-social). Finally, the effect of interaction between the Type of intention and groups was significant (F(3,191)=11.06, p<.001). Interestingly, patients’ performance testified to a greater influence of the bias than controls in the non-social superordinate condition (LSD test: p=.003), while in social conditions the bias affected patients’ response to a lesser extent than in healthy participants (LSD tests: all p<.008) (figure 4). This interaction effect was not further modulated by the Amount of visual information (Group*Amount*Intention: F(6,382)=.69, p=.66).

Fig. 4. Bias effect (%) for all types of intention considered. The greater the bias effect, the more participants respond toward the preferred (i.e. biased) intention.