Multi-agent accumulator-based decision-making model of incivility (MADI)

Abstract

While behaviour can either be perceived as respectful or disrespectful, incivility reflects relatively minor violations of social norms within a group. In the present study, we used an accumulator-based model of decision-making, assuming that social agents attempt to classify behaviour as respectful or disrespectful based on available social cues and reciprocate toward other group members once a criterion amount of evidence is accumulated. Perceived incivility is derived from the model by taking the balance of evidence of the respectful and disrespectful social cues, reflecting uncertainty in decision-making. In multi-agent interactions, the model averages perceived incivility (i.e., uncertainty) over multiple trials. We demonstrate that this model can differentiate between attitudes and behavior in a single social agent as well as how incivility can arise within a group as a result of small differences in response threshold to disrespectful behaviour and biases in social cue identification accuracy.