DIAG-Party-Newspaper

Recent research has demonstrated that people infer ideological categories such as conservative, leftist, racist, feminist, environmentalist or party-categories like Democrat or Republican from short behavioral descriptions or expressions of opinion. The current study investigates whether these inferences can be reduced. To the extent that ideological inferences serve to predict a person's political opinions and behaviors they should be sensitive to perceived cue diagnosticity, that is, the extent to which a cue (e.g., an observed behavior) predicts membership in category A and non-membership in category B. In the current research we manipulate cue diagnosticity via counter-stereotypical exemplars. For instance, we present participants with a Republican or Democrat who is either pro-life or pro-choice and test whether the stereotype-disconfirming exemplar (i.e., a pro-choice Republican or a pro-life Democrat) compared to the stereotype-confirming exemplar (i.e., a pro-life Republican or a pro-choice Democrat) leads to reduced inferences of partisanship on the basis of opinions about abortion.