On the Dominance of Moral Categories in Impression Formation.

A goal-oriented approach to person perception (Trzebinski, 1985) and personality description (Pervin, 1989) assumes that goals are central both to the scientific assessment and lay perception of personality. This approach distinguishes two general types of goal categories: one that refers to the actor's intended goals and another that concerns the probability of goal attainment. Identification of the actor's goal is frequently a prerequisite for drawing inferences about his or her traits (Read, Jones, & Miller, 1990), and the intended goal is paramount in deciding whether an action is moral or immoral (Shultz & Wright, 1985). On the other hand, efficiency of goal attainment is crucial in determining the actor's competence and abilities (Darley & Goethals, 1980). It may be expected, then, that morality (M) and competence (C) constitute two basic and relatively independent meanings of social behavior and personality traits.

This expectation found substantial support in a number of studies. Moral- and competence-related traits frequently appear in voters' open-ended commentaries on political candidates; Kinder and Sears (1985) claimed that M and C constitute two separate and basic clusters of traits in the perception of political leaders. Wojciszke (1994) asked his participants for recollections of episodes in which they had come to clear-cut evaluative conclusions on other people or themselves. Content analyses of more than 1,000 episodes showed that in three fourths of them, the evaluative impression was based on M- or C-related considerations. Finally, in classical studies on the structure of implicit personality theories, Rosenberg and his coworkers (Rosenberg & Sedlak, 1972) showed that co-occurrences of traits in person impressions were underlain by two relatively independent dimensions. Although Rosenberg dubbed the dimensions intellectually good-bad and socially good-bad, the terms competence and morality may be equally or even more appropriate. Numerous traits marking the intellectually good-bad dimension have more to do with competence in general than with intellect (e.g., persistent, industrious, wavering), whereas many of the traits defining the socially good-bad dimension clearly pertain to morality (e.g., sincere, helpful, dishonest).

Differences Between Moral- and Competence-Related Categories

At least two theories--the schematic model of attribution (Reeder, 1985) and the cue-diagnosticity model of impression formation (Skowronski & Carlston, 1987)--hypothesize differences in the processing of M and C information. Both theories assume that people infer personality traits from others' actions and that in the M domain, negative information is more decisive or diagnostic than positive information, whereas the opposite is true in the C domain. An important effect of this positive-negative asymmetry is that integration of incongruent information results in a negativity bias in the M domain, but it results in positivity bias in the C domain (Kubicka-Daab, 1989; Skowronski & Carlston, 1987; Wojciszke, Brycz, & Borkenau, 1993).

Another important difference between the two domains is that M judgments are more saturated with affect than C judgments. For example, using literally hundreds of behavior descriptions, Wojciszke, Pienkowski, Maroszek, Brycz, and Ratajczak (1993) found that behavioral acts elicited more extreme evaluations when they exemplified M traits rather than C traits. In a similar vein, Brycz and Wojciszke (1992) showed that in the M domain, lay predictions of a target person's future behavior were to some degree based on the perceiver's purely affective responses to the target, whereas in the C domain, the predictions were solely based on cool ascriptions of a relevant personality trait.

Why should morally relevant behavior instigate a stronger emotional response than behavior revealing (in)competence? One possible answer is that perceivers typically tend to construe incoming information in terms of its bearing on their own self-interest (unless they are driven by other specific goals) and that others' morality is typically more relevant for the perceiver's interests than is their competence. Usually, an individual's immoral behavior is harmful to other people, whereas his or her moral behavior is beneficial to them. Perceivers are involved as targets of those harms or benefits, either actually or potentially. On the other hand, C qualities of behavior are of only secondary importance to perceivers (with the exception of observers processing information under a specific, C-related goal, such as when making employment decisions). They are consequential to the extent that the actor's competence leads to a higher or lower efficiency in inflicting harm or furnishing benefits.(1)

Dominance of Moral Categories in Global Impression Formation

All of this suggests that M-related information should receive special treatment in the global appraisals of others. Such appraisals have been studied in countless experiments under the impression formation heading. The ease and willingness with which people make such impressions, even without a specific goal in mind and based on a very impoverished base, suggests that this is a natural task for them. Why do people engage in formation of evaluative impressions in natural settings? The most straightforward answer seems to be that such impressions reflect the location of others on the approach-avoidance dimension. All organisms have at least one mechanism for differentiating agreeable environments from adverse environments (Martin & Levey, 1978), and people have many such mechanisms, including evaluative processes, such as attitudinal responses (Cacioppo & Berntson, 1994) and global impressions of others.

If the main function of global evaluative impressions is to distinguish between persons who should be approached and persons who should be avoided, it is clear why M categories occupy a privileged position in impression formation. These categories are instrumental in locating others on the approach-avoidance dimension to a higher extent than any other concept (C traits included)--a decision about whether a person is moral amounts to a direct settlement of whether the person is beneficial rather than dangerous.(2) In contrast, information on his or her competence plays the role of a modifier. That is, it helps to decide how beneficial or how dangerous the person is, and it comes to play only after the basic approach-avoidance decision (i.e., moral judgment) has been made. Therefore, we expect M categories to play a dominant role at different stages of impression formation--from gathering information on which an impression is to be based to concluding what the final impression is.

Perceiver's Sex

In traditional sex-role stereotypes, caring for others' needs and well-being is considered the domain of women rather than men, whereas task orientation and striving for occupational achievement and excellence is considered the domain of men rather than women. Although in modern societies these sex-role stereotypes have decreased in their scope and intensity, they are still being built into people's self-identities. In the course of socialization, males learn more about the importance of C traits, whereas females learn more about the importance of M traits (Eagly, 1987). These differential socialization practices can result in females' tendency to base their impressions of people to a higher degree on M judgments than on C judgments; the opposite is true for male perceivers. In effect, the dominance of M categories over C categories postulated in impression formation should be more pronounced for female perceivers than for male perceivers.

Overview of the Studies and Hypotheses

In Study 1, we asked our participants to list those personality trait descriptors that they considered to be most important in other persons. These trait names were then rated for their M relatedness and C relatedness to test the hypothesis that the most (chronically) accessible descriptors of others are more related to morality than to competence. In Study 2, we examined the information gathering process and tested the hypothesis that when forming global impressions of others, people are more interested in their M traits than their C traits. In Study 3, we studied M traits and C traits ascribed to several target persons and tested hypotheses that global impressions of real persons are better predicted from M trait than C trait ascriptions. Finally, in Study 4, we investigated global impressions based on behavioral information concerning both M and C of fictitious target persons. We tested the hypothesis that the global impression of positivity-negativity was decided mainly by the M content of targets' behavior, whereas C information served only as a relatively weak modifier of impression intensity.

STUDY 1: CHRONIC ACCESSIBILITY OF M TRAITS AND C TRAITS

It is well known that people differ in the content of the chronically accessible constructs that they use in their perception of others, presumably due to frequent and consistent use of these categories in the past. Heightened chronic accessibility of a construct (e.g., such as the person descriptor of honesty) results in a greater likelihood of detection of information that is relevant to the construct (Bargh & Pratto, 1986) and greater sensitivity to the construct-relevant information in the person-perception process (Higgins, King, & Mavin, 1982). Chronic accessibility was understood in the cited studies as individual differences in heightened probability of use of single constructs (such as intelligent, honest, or conceited). However, the present theorizing implies chronic accessibility of a whole class of concepts (i.e., M-related traits). Despite possible individual differences, this heightened accessibility (compared to C-related …