Evidence for bidirectional causation between trait and mental state inferences

To navigate the social world, people must understand each other’s momentary thoughts and feelings (mental states) and enduring personalities (traits). How do people make trait and mental state inferences in the real world? Prior research has artificially separated these two processes, primarily studying each topic on its own. However, in real life, these two processes constantly co-occur and rely on partially overlapping information. It is likely that people inform one type of inferences with the other. Here we investigate this possibility using naturalistic paradigms, statistical learning, and stimulus optimization techniques. We first demonstrated the correlation between trait and mental state inferences of targets in naturalistic videos (Study 1) and familiar others in real life (Study 2). Targets perceived to have similar traits were judged to experience various mental states with similar frequencies. We showed that this association was causal in two experiments. Learning that two people experienced similar mental states across a range of situations caused participants to attribute similar traits to them (Study 3). Conversely, observing two people had similar traits caused participants to attribute similar mental states to them across a range of situations (Study 4). Together, these four preregistered studies (total N = 762) reveal that trait and mental state inferences continually run in parallel and that people rely on others’ enduring traits to predict their momentary states, and vice versa. These findings highlight that biases in trait impressions may distort understanding of others’ situation-specific states, and others’ uncharacteristic states may influence judgments of their enduring traits.