On Sanders, Trump, and rhinoceroses: Quantifying subjective construals helps predict political attitudes

People with different worldviews may have different interpretations for the same object. Freedom fighters for some are terrorists for others; immigrants can be seen as a threat or being threatened. To incorporate differences in interpretations into quantitative models, I propose the construct of value-instantiating beliefs, or perceived consequences of actions for basic human values. I hypothesize that these beliefs moderate the relationship between personal values and attitudes. In a preregistered mixed-methods study (N = 2038) I assess personal value priorities, attitudes, and perceived consequences of four political behaviours. Beliefs about consequences of the behaviours differed with political affiliation: for example, liberals saw voting for Trump in light of negative consequences for universalistic values, while conservatives stressed positive consequences for self-direction and security. The moderation hypothesis was confirmed. Including value-instantiating beliefs in the models significantly improved the predictions of attitudes towards and intentions to perform the behaviours.