Our purpose in the present meta-analysis was to examine the extent to which discrete emotions elicit changes in cognition, judgment, experience, behavior, and physiology; whether these changes are correlated as would be expected if emotions organize responses across these systems; and which factors moderate the magnitude of these effects. Studies (687; 4,946 effects, 49,473 participants) were included that elicited the discrete emotions of happiness, sadness, anger, and anxiety as independent variables with adults. Consistent with discrete emotion theory, there were (a) moderate differences among discrete emotions; (b) differences among discrete negative emotions; and (c) correlated changes in behavior, experience, and physiology (cognition and judgment were mostly not correlated with other changes). Valence, valence-arousal, and approach-avoidance models of emotion were not as clearly supported. There was evidence that these factors are likely important components of emotion but that they could not fully account for the pattern of results. Most emotion elicitations were effective, although the efficacy varied with the emotions being compared. Picture presentations were overall the most effective elicitor of discrete emotions. Stronger effects of emotion elicitations were associated with happiness versus negative emotions, self-reported experience, a greater proportion of women (for elicitations of happiness and sadness), omission of a cover story, and participants alone versus in groups. Conclusions are limited by the inclusion of only some discrete emotions, exclusion of studies that did not elicit discrete emotions, few available effect sizes for some contrasts and moderators, and the methodological rigor of included studies.

In the United States, men are more likely to pursue math-intense STEM courses and careers than women. This investigation explored whether positivity bias in the degree to which people overestimate their past performance contributes to this gender gap. To find out, two studies were conducted with undergraduate college students in the Southern United States. In Study 1, participants (n=122) completed a math test and estimated the percent they had solved. They then were given feedback and completed a second math test and estimation. Men overestimated their performance more than women, judging they had done better on the test than they actually had. This gender difference was not present after feedback. Further, women, but not men, who reported a more positive previous experience with math were more likely to overestimate their performance. In Study 2, participants (n=184) completed a math test and estimated the percent they had solved. They also reported their interest in pursuing math courses and careers. Again, men overestimated their performance more than women. This greater overestimation of performance in men accounted for their greater intent to pursue math fields compared to women. The findings suggest that gender gaps in STEM fields are not necessarily the result of women underestimating their abilities, but rather may be due to men overestimating their abilities.

Lindquist, Siegel, Quigley, and Barrett (2013) critiqued our recent meta-analysis that reported the effects of discrete emotions on outcomes, including cognition, judgment, physiology, behavior, and experience (Lench, Flores, & Bench, 2011). Lindquist et al. offered 2 major criticisms-we address both and consider the nature of emotion and scientific debate. Their 1st criticism, that the meta-analysis did not demonstrate emotion-consistent and emotion-specific changes in outcomes, appears to have been based on a misunderstanding of the method that we employed. Changes in outcomes were coded according to predictions derived from a functional discrete emotion account. Their 2nd criticism, that the findings are consistent with a psychological constructionist approach to emotion, is partially supported by the data and our statements in Lench et al. (2011). However, only 1 meta-analytic finding is relevant to this hypothesis, and it does not offer unequivocal evidence. Further, we contend that no modern discrete emotion theories would make the claims described by Lindquist et al. as representing a "natural kind" perspective and that viewing a scientific debate as a war has negative implications for the ability to evaluate evidence.

No abstract

A risk-as-feelings approach suggests that factors irrelevant to the potential risk can influence risk perception. This investigation focused on the speed of events as one such factor. Negative events that occur relatively quickly were judged as more likely to occur than events that occur more slowly. Speed influenced risk perception when it was salient and differences in risk perception were reduced when it was not salient. Further, the likelihood of a negative outcome was judged to be more likely when the same event was described as occurring relatively quickly compared to slowly. Even when only the speed at which information was presented changed, faster events were judged to be riskier than slower events. Theoretically, these findings suggest that speed of an event contributes to risk judgements and suggest speed may be the reason people fear fast but low incidence events and fail to fear slower but higher incidence events.