DECODING SUBJECTIVE EVALUATIONS - HOW STEREOTYPES PROVIDE SHIFTING STANDARDS

Three studies examined how subjective evaluations relevant to stereoty pes are translated into open-ended descriptions (Study 1) and into obj ective judgments and Likert-type ratings (Studies 2 and 3). We expecte d that stereotypes would create an implicit context or standard that i ndividuals would use to ''decode'' subjective evaluations and against which targets would be judged. Although Study 1 found much similarity in descriptions of mothers and fathers evaluated as ''good'' or ''bad, '' this similarity was more apparent than real. As predicted by the sh ifting standards model, Study 2 demonstrated that subjective evaluatio ns were decoded using gender stereotypes as standards, with a woman de scribed as either a ''very good'' parent or an ''alright'' parent judg ed objectively to perform significantly more parenting behaviors than a similarly described man. Likert-type ratings failed to reveal this d ifference. In fact, evaluatively dissimilar targets (an ''alright'' mo ther and a ''very good'' father) were rated to be objectively the same on some dimensions but overall subjectively different. Study 3 extend ed these findings using race stereotypes and judgments of math ability . Thus, the ''same'' traits were deceptively not quite the same at all : they had different behavioral expectations associated with them, dep ending on the target's group membership. (C) 1997 Academic Press.