Theoretical and empirical work on the processes by which we attribute dispositional characteristics to others has focused almost exclusively on how such processes proceed once the perceiver has been motivated to initiate them. The problem of identifying the factors which prompt the perceiver to engage in an attributional analysis in the first place has been relatively ignored, even though the influence of such factors may extend beyond the initiation of the causal analysis to affect the manner in which it unfolds and, ultimately, the form and substance of its conclusion. From the assumption that the function of an attributional analysis is effective control of the social environment, it was hypothesized that high outcome dependency upon another, under conditions of high unfamiliarity, is associated with the initiation of an attributional analysis as evidenced by increased attention to the other, better memory of the other's characteristics and behavior, more extreme and confidently given evaluations of the other on a variety of dispositional trait dimensions, and increased attraction to the other. These hypotheses were tested within the context of a study of heterosexual dating relationships in which men and women volunteers anticipated varying degrees of dependence upon another for their dating outcomes. The findings support the view that the data processing operations of the social perceiver-from attention to memory to attribution-are part of a unified whole and may be viewed as manifestations of an underlying motivation to predict and control the social environment.

No abstract

Two studies investigated differences in the situational control of behavior as a result of variation in attentiveness to situational and interpersonal guides to social appropriateness. In Study 1, subjects participated in a group discussion that made salient either of two reference groups. The effects of this manipulation on social conformity were examined as a function of self-monitoring and neuroticism measures. Social conformity of high self-monitoring and low neuroticism subjects differed reliably between discussion contexts. Low self-monitoring and high neuroticism subjects were unaffected by discussion context. In Study 2, raters judged their generosity, honesty, and hostility in nine relevant situations. A variance measure of perceived situational variability indicated that high self-monitoring subjects reported more situational variability than did low self-monitoring subjects. Implications for the interaction of situational and dispositional determinants of behavior are discussed.Research and theory have Idng attempted to weigh the relative influence of situational and dispositional determinants of human behavior. Inevitably, participants in this endeavor have taken sides. Classical personality formulations have emphasized traits and types and searched for cross-situational consistencies in behavior (Alker, 1972; Cattell, 1950; Guilford, 19S9). Unfortunately for this position, as Mischel (1968) has documented, cross-situational correlation coefficients in any domain of social behavior rarely exceed .30. Social learning theorists (e.g., Mischel, 1968Mischel, , 1973 have interpreted this as evidence for the situational control of behavior. Nevertheless, whenever comparisons of the amount of variance in self-report, self-rating, or actual behaviors attributable to person variables and situation variables are made, rarely do either account for sizable amounts of variance. Bowers (1973) has reported that across 18 comparisons, the average variance due to persons

The attempt to find high correlations between trait measures and behavioral measures has often been unsuccessful. Furthermore, some personality theorists have suggested that the only time one might expect a substantial correlation between a trait measure and a behavioral measure is when the behavioral measure is comprised of a multiple-act criterion. The present research suggests that it would be premature to abandon the attempt to predict single-act criteria. The results of two studies indicate that traits can be used to predict behavior in some situations but that traits cannot be used to predict behavior in other situations. As a consequence, the relative efficacy of multiple-act criteria may be due in part to the fact that multiple-act criteria have a higher probability of including at least one predictable situation. Thus, a reasonable alternative strategy to employing multiple-act criteria is to attempt to specify those conditions in which single-act criteria are highly correlated with trait measures. The data indicate that one of the most promising means of specifying when traits can and cannot predict behavior is by the assessment of the strength of the situational pressures present in the setting in which the behavior is measured.

This study examined whether social cognitions that have been assumed to influence aggression actually forecast change in aggressive habits over time. Participants were 189 3rd- through 7th-grade boys and girls; data on social cognitions and social behaviors were collected in the fall and spring of the school year. Aggression-encouraging cognitions assessed in the fall indeed promoted aggression over the school year, but such developments hinged critically on child sex and on initial (fall) levels of aggression and victimization. Results illustrate the principle that cognitions affect behavioral development mainly when the child's transactions with the social environment support the use of the cognitions as guides for behavior.