In two studies we found that feelings of guilt provoke individuals to cooperate in repeated social bargaining games (a prisoner's dilemma in Study 1 and an ultimatum game in Study 2). Feelings of guilt were either experimentally manipulated (Study 1) or assessed via self-report (Study 2) after participants had played one round of a social bargaining game. As predicted, individuals who experienced feelings of guilt (compared to individuals who felt no guilt) after pursuing a non-cooperative strategy in the first round of play, displayed higher levels of cooperation in the subsequent round of play (even one week later). Results are discussed in terms of an "affect-as-information" model, which suggests that non-cooperating individuals who experience the negative affective state associated with guilt in a social bargaining game may be using this feeling state as "information" about the future costs of pursuing an uncooperative strategy. Because in guilt the focus is on the specific, individuals are capable of ridding themselves of this emotional state through action (Lewis, 1993, p. 570).

The average probability estimate of J > 1 judges is generally better than its components. Two studies test 3 predictions regarding averaging that follow from theorems based on a cognitive model of the judges and idealizations of the judgment situation. Prediction 1 is that the average of conditionally pairwise independent estimates will be highly diagnostic, and Prediction 2 is that the average of dependent estimates (differing only by independent error terms) may be well calibrated. Prediction 3 contrasts between-and within-subject averaging. Results demonstrate the predictions' robustness by showing the extent to which they hold as the information conditions depart from the ideal and as J increases. Practical consequences are that (a) substantial improvement can be obtained with as few as 2-6 judges and (b) the decision maker can estimate the nature of the expected improvement by considering the information conditions.

The self is defined and judged differently by people from face and dignity cultures (in this case, Hong Kong and the United States, respectively). Across 3 experiments, people from a face culture absorbed the judgments of other people into their private self-definitions. Particularly important for people from a face culture are public representations--knowledge that is shared and known to be shared about someone. In contrast, people from a dignity culture try to preserve the sovereign self by not letting others define them. In the 3 experiments, dignity culture participants showed a studied indifference to the judgments of their peers, ignoring peers' assessments--whether those assessments were public or private, were positive or negative, or were made by qualified peers or unqualified peers. Ways that the self is "knotted" up with social judgments and cultural imperatives are discussed.

SummaryThis study examined the mental health outcomes of job stress among Chinese teachers in Hong Kong. A total of 269 Chinese teachers participated in Study 1 which provided crosssectional data regarding the associations among stress resource factors, burnout, and negative mental health. Study 2 was a six-month longitudinal study which aimed to establish the direction of the associations among the hypothesized variables across two time points with a separate sample of 61 Chinese secondary school teachers. Results of the structural equation modelling analyses on the cross-sectional data at T1 showed that stress resource factors of self-ef®cacy and proactive attitude were negatively related to burnout, which in turn had a direct effect on negative mental health. Stress resource factors were also directly linked to mental health status of teachers. Results of similar analyses on the longitudinal data at T2 further indicated that burnout at T1 had a direct impact on burnout at T2, which in turn had a direct effect on negative mental health at T2. Findings and limitations of the study were discussed.

Six-person mock civil juries awarded significantly larger amounts for damages than did 12-person juries, and individuals preferred even larger average awards. A reversal of the "deep-pockets bias" observed earlier, an explanation involving temporal fluctuation in normative standards, during the time interval between the studies, was supported by independent data showing temporal trends in actual civil trial awards. A computational model of consensus that assumed a strong majority of those members with the most similar (closest) personal preferences decided on the median of their preferences accurately predicted award magnitude. Computer simulations explored the effects of critical faction size (majority, etc.) and location within the group, features that might in turn depend on task environment, cultural dynamics, and social context.