Fig. 4: Loneliness is associated with perceptions that others are more volatile and with perceived fragility of positive states for the self.
From: Loneliness is associated with unstable and distorted emotion transition predictions
A Lines represent the estimated simple slopes from a meta-analysis of mixed-effects models predicting emotion transition ratings from loneliness separately for each valence transition category. Higher loneliness is associated with greater expectations that others will transition between positive and negative states (positive to negative in dashed green line, negative to positive in dashed yellow line) and lower expectations that others will remain in the same valence state (solid lines; positive to positive in pink, negative to negative in blue). In other words, lonely individuals expect more volatility in others’ emotion transitions. Shaded bands indicate 95% confidence intervals (N = 6 studies). Asterisks indicate significance levels of the slope differing from 0: p < 0.05 (*), p < 0.001 (***). B The meta-analytic fixed-effect estimates for the relationship between loneliness and emotion transition ratings, separately for each valence transition category, are visualized. Each point represents the estimated standardized beta coefficient, with horizontal error bars denoting the 95% confidence intervals (NStudy1 = 113; NStudy2 = 185; NStudy3 = 376; NStudy4 = 91; NStudy5 = 68; NStudy6 = 41). C Lines represent the estimated simple slopes from a meta-analysis of mixed-effects models predicting emotion transition ratings from loneliness separately for each valence transition category. Higher loneliness was associated with greater expectations that the self will transition into a negative state from a positive state (in green) and lower expectations that the self will transition into a positive state from a positive state (in pink). In other words, lonely individuals perceive positive states as less stable for themselves. Shaded bands indicate 95% confidence intervals (N = 7 studies). Asterisks indicate significance levels of the slope differing from 0: p < 0.001 (***). D The meta-analytic fixed-effect estimates for the relationship between loneliness and emotion transition ratings separately for each valence transition category are visualized. Each point represents the estimated standardized beta coefficient with horizontal error bars denoting the 95% confidence intervals (NStudy1 = 113; NStudy2 = 185; NStudy3 = 376; NStudy4 = 91; NStudy5 = 68; NStudy6 = 41; NStudy7 = 856).
