Participants rated schematic faces from two categories, gnomes and leprechauns, on feature widths and pleasantness of facial configuration. Three target faces shared critical facial features across categories while two contextual faces extended the range for that category to include either very wide or very narrow features. In Experiment 1a, results indicated contextual effects on judgments of target faces when they were rated in separate categorical blocks [Wedell, D. H., & Pettibone, J. C. (1999). Preference and the contextual basis of ideals in judgment and choice. Journal of Experimental Psychology: General, 128, 346–361], but not when faces from both categories were rated together in Experiment 1b. Two additional experiments explored this failure to use categorical information. Categorical context was found to produce contrast effects on descriptive ratings of feature width when participants were forced to rely on name cues rather than actual faces in Experiment 2. In Experiment 3, both contrast effects on descriptive ratings and assimilation effects on ideals for pleasantness were found when the names for the faces were learned separately for each category. These results identify constraints on the nature of category-based stereotyping effects on judgment while isolating the influence of recent and categorical context.
Two important influences on social judgments of persons, behaviors or events are the recent set of persons, behaviors or events one has encountered and the information represented in categories relevant to the target being judged. The first of these influences has been documented in the literature on context effects in judgment (for reviews, see Eiser, 1990, Mussweiler, 2003, Schwarz and Bless, 1992 and Wedell and Parducci, 2000) and the second has been documented in the literature on stereotyping (for reviews, see Hilton and von Hippel, 1996 and Macrae and Bodenhausen, 2000). These two types of judgment phenomena have been studied largely in isolation of each other, although some researchers have attempted to examine the relationship between them (e.g., Manis, Biernat, & Nelson, 1991). Others have pointed out how these influences are difficult to separate within commonly used social judgment paradigms (Wedell & Pettibone, 1999). The purpose of the research reported here was to determine how these two sources of evaluative influence might operate on two different types of judgment, dominance judgments and ideal-point judgments (Coombs, 1964), and to understand the conditions under which each may be used.
Dominance judgments are perhaps the more common type of evaluation examined in the social literature and reflect the situation in which values on the response scale are monotonically related to values on the underlying dimension. Thus, for example, ratings of a person’s aggressiveness increase with increase in the number of aggressive behaviors, or ratings of a person’s likableness increase with number of positive behaviors. Ideal-point judgments, on the other hand, typically follow a nonmonotonic relationship between responses and the underlying stimulus attribute. The specific form of ideal-point judgments tends to be a single-peaked function. These types of judgments are most often found in attitudinal endorsements (Eagly & Chaiken, 1993). For example, a politically moderate individual will disagree with attitudinal positions at the liberal or conservative ends of the spectrum and endorse those in the middle. The stimulus value corresponding to the peak of the endorsement function is referred to as the ideal point and represents the individual’s own attitude (Thurstone & Chave, 1929). Item characteristic curves for Likert-type scales typically follow a single peaked function and thereby are consistent with an ideal-point judgment process (Roberts and Laughlin, 1996 and Roberts et al., 1999). Ideal-point functions are also characteristic of many preference domains, in which there can be too much or too little of an attribute, such as the preferred amount of sugar in one’s tea or coffee.
