Entertainment-education fiction narratives have proven to be a powerful tool for social change, dialogues and character design could being key elements for the effect of these narratives. Thus, our project will analyse an element related to the configuration of protagonists in EE fiction narratives designed to reduce prejudice: the type of evidence (arguments) used to criticise a rumour against a stigmatised group. For this purpose, three online experiments will be carried out with a 2 (type of evidence or arguments) x 3 (type of rumour) factorial design with independent measures. The main independent variable will be the type of evidence used by the protagonist (a character in transition) to deactivate the rumour. The second independent variable will be the type of rumour (using three different rumours per experiment). Finally, we will use the participant’s involvement with the topic as a moderating variable. In doing so, we attempt to contrast and extend the theoretical model proposed by the TRANCHAR project by applying it to three stigmatised groups (immigrants, LGTBIQ + people and people with disabilities). Therefore, our project aims to advance knowledge on the mechanisms that explain the impact of EE productions and the facilitating conditions, in the specific field of prejudice reduction.