Extreme outcome expectations and affect intensity

Abstract

Two studies tested the hypothesis that extreme outcome expectations are associated with affect intensity. Study 1 (N=104) measured extreme outcome expectations in response to one's idiosyncratic goals, and Study 2 (N=93) measured extreme outcome expectations in response to common life events. Higher levels of affect intensity were associated with higher levels of extreme outcome expectations in both studies. The association between affect intensity and extreme outcomes expectations held even after controlling for shared variance with other affective variables (i.e., trait pleasant affect, trait unpleasant affect, affect variability) and other variables that might overlap with extreme outcome expectations (i.e., optimism, pessimism).

Notes

¹Schimmack and Diener (1997) proposed separating the frequency and intensity of affect when computing affect intensity scores. They suggested computing an average intensity score for each emotion only for the number of times that emotion was endorsed. This averaging method excludes from the averaging procedure the times in which participants indicated not experiencing an emotion. Thus, the overall mean affect (M) was decomposed to frequency (F) and intensity (I) components according to the following formula: M(affect) = F(affect)×I(affect)/N, with N being the number of ratings, F being the number of nonzero ratings, and I being the intensity score, computed as the average across all nonzero ratings.

²The results were the same regardless of whether AIM total score, AI total score, or the composite AIM/AI score was examined. Therefore, only the results for the composite score are presented.

³To avoid visual clutter, these correlations are not included in Figure 1. The correlations among the emotion variables were as follows: AI×PA: .29; AI×NA: .25; PA×NA: −.22. The correlation between extreme outcome expectations and goal achievement likelihood was .08.

⁴The positive extreme attribution and the negative extreme attribution scores were highly correlated with each other (r=.54). When they were entered separately in the same model, the associations between the predictors and the dependent variables were attenuated due to multicollinearity. When these predictors were added individually to the model, they still both predicted affect intensity. When positive outcome expectations were entered instead of the composite, the betas were as follows: positive outcome expectations with AI: .32; achievement likelihood with AI: .20; achievement likelihood with PA: .35; achievement likelihood with NA: −.20. When negative outcome expectations were entered in the model instead of the composite, the betas were: negative outcome expectations with AI: .44; achievement likelihood with AI: .33; achievement likelihood with PA: .32; negative outcome expectations with NA: .45.

⁵To avoid visual clutter, these correlations are not included in Figure 2. The correlations among the emotion variables were as follows: AI×PA: .35; AI×NA: .28; AVAR×PA: −.03, AVAR×NA: .47; AI×AVAR: .27; PA×NA: −.09. The correlation between extreme outcome expectations and optimism was −.18, and the correlation between extreme outcome expectations and pessimism was .21.

⁶Similar to Study 1, the positive extreme attribution and the negative extreme attribution scores were highly correlated with each other (r=.56). When they were entered separately in the same model, the associations between the predictors and the dependent variables were attenuated due to multicollinearity. When these predictors were added individually to the model, they still both predicted affect intensity. When positive outcome expectations were entered instead of the composite, the betas were as follows: positive outcome expectations with AI: .42; positive outcome expectations with AVAR: .33; positive outcome expectations with PA: .20; positive outcome expectations with NA: .20; optimism with AVAR: −.29; optimism with PA: .27. When negative outcome expectations were entered in the model instead of the composite, the betas were: negative outcome expectations with AI: .27; negative outcome expectations with AVAR: .25; negative outcome expectations with NA: .20; optimism with AVAR: −.37; optimism with PA: .27; optimism with NA: −.29.