Personality and Social Psychology Bulletin

Consistent with a mini-theory of motivational factors in unreciprocated love, (a) perceived potential value of a relationship with another; perceived probability of such a relationship; and desirability of the state of being in love with this other, even if it is unreciprocated, were each significantly and independently predictive of reported intensity of unreciprocated love; (b) incidence of unreciprocated love was greatest for those whose self-reported attachment style was anxious/ambivalent; and (c) perceived potential value of the relationship was most predictive of intensity for anxious/ambivalents and perceived desirability of the state was most predictive of intensity for avoidants. Findings are based on structural equation modeling analyses of questionnaire responses from a moderately large sample of U.S. undergraduates and were not qualified by gender.

Participants were either informed that contextual influences bias their judgment and asked to correct for the unspecified influence (blatant warning) or they were instructed that they should correct for the unspecified influence if they felt that there may be biasing influences (conditional warning). Whereas blatantly warned participants corrected under all conditions (Study 2), conditionally warned participants corrected their judgments when the source of bias was salient but not when the source was subtle (Studies 1 to 3). Implications for models of theory-driven correction are discussed.

This article examined the effects of group variability and processing goals on the ease of making judgments about social groups. Participants were presented with a heterogeneous or homogeneous category and were instructed either to form an impression of the group (impression set) or to attend to the similarities among, and the differences between, group members (integration set). Under impression sets, we predicted that participants would find it harder to judge heterogeneous groups as compared to homogeneous groups. Under integration sets, however, judgments were predicted to be relatively easy, regardless of group variability. These predictions were supported, regardless of whether judgment ease was operationalized via response latencies (Experiment 1) or through subjective reports (Experiment 2). Experiment 2 suggested that these effects were driven by the effects of group variability and instructional set on the importance of subtypes formed at the subordinate level. Implications of these results for research on the consequences of group variability are discussed.

This study investigated whether behavioral decisions (e.g., persistence, help seeking) are influenced by covariation information and if this influence is mediated by attributions. In Experiment 1, participants belonging to different age groups received covariation information, and causal attributions as well as behavioral-reaction decisions were assessed. Experiment 1 replicated the often-documented influence of covariation information on attributions. Furthermore, it revealed that covariation information affects behavioral decisions more strongly, and developmentally earlier, than do attributions. In Experiment 2, participants were provided with conflicting covariation information and attributions. Covariation information had a stronger impact on reaction decisions than attributions. Experiment 3 again provided participants with covariation information and assessed attributions and behavioral choices, as well as reaction times. Participants provided with covariation information needed less time to choose behavioral reactions than to choose attributions. We conclude that behavioral decisions can be directly determined by covariation information without attributional mediation.

Most bipolar models of affective processing in social psychology assume that positive and negative valent processes are represented along a single continuum that rangesfrom very positive to very negative. Recent research has raised the possibility, however, that the motivational systems for positive/approach and negative/defensive valent processing (positivity and negativity, respectively) are separable. In this article, the authors use unipolar positivity, negativity, and ambivalence ratings and bipolar valence, dominance, and arousal ratings of 472 slides from the International Affective Picture System to examine several aspects of the bivariate model of evaluative space. Analysis confirmed a positivity offset and negativity bias in the activation functions of the valent systems as wel as multiple modes of evaluative activation (e.g., reciprocal, uncoupled positivity, uncoupled negativity). Together, these data suggest that the bipolar structure of affective processes should be tested rather than assumed.

In two studies, the authors examined the accuracy of dysphoric and nondysphoric individuals' predictions about their future behavior. Participants predicted the occurrence of a variety of everyday events and reported on their occurrence over a period of either 4 (Study 1) or 8 (Study 2) weeks. As expected, dysphoria was unrelated to overall accuracy, but nondysphorics tended to be more accurate in making optimistic predictions and dysphorics tended to be more accurate in making pessimistic predictions. These differences were related to differences between the two dysphoric groups in base rates of reported outcome occurrence and certainty of judgments. The findings did not support depressive realism, the negative biasing effect of dysphoria on future predictions, or the contention that dysphorics are less accurate because they predict either more atypical or overly optimistic outcomes.

Two multi method studies probed the structure and correlates of self-reports of personality when participants were permitted to admit that they did not know assessment words. The first study (1,187 college students) examined the structure of self-evaluation using Goldberg's markers for his five-factor approach. Students reported greatest familiarity with Agreeableness markers and least familiarity with Emotional Stability. The second study (317 young adolescents) used two data sources: (a) individualized, self-paced, computer-based self-evaluations and (b) teacher ratings and evaluations of adjustment. The assessment (in either English or Spanish) allowed participants to obtain definitions. Adolescents requested the least help with agreeableness markers and the most help with emotional stability. Even with unknown words, and despite differences in methodology, similar five-factor structures emerged in both studies. Computer assessment methods, in combination with a five-factor approach, may be useful for probing processes that underlie the structure and development of personality and their links to adjustment.