Our main hypothesis is that speech type will affect source memory (better source memory in the indirect speech condition) but not content memory. So we expect to find an interaction effect. Based on our previous experiments, we expect people to be better at addressing statements to a protagonist in the indirect speech condition than in the direct speech condition. Like in Exp. 3, we included additional information with respect to the protagonists. Therefore, we expect to find an effect of speech type on the number of correctly addressed statements based on the name of the protagonist. This effect is thought to be smaller than that in Exp. 3 (d = .44) because of the addition of another block of questions that will be answered first by half of the participants. We do not expect to find an effect of speech type on the number of correctly addressed statements based on the textbase of the story. In other words, we do not expect to find an effect of speech type on content memory. NB. Although we are not primarely interested in the effect of speech type on source memory at the gender level in this study, we do expect to find this effect, like in Exp 1-3 (with better source memory in the indirect speech condition than in the direct speech condition). This effect is thought to be larger than in Exp 1 (d = .39) and Exp 2 (d = .37), and smaller than that in Exp. 3 (d = .56; because of the addition of another block of questions that will be answered first by half of the participants).