Forecasting Turbulence: Evidence Of Affective Projection Biases In Momentary Predictive Fluctuations Using Dynamic Structural Equation Modelling

Document Type

Article

Publication Date

4-1-2026

School

Psychology

Abstract

Background Belief formation and behavior are informed by predictions about future emotional outcomes (affective forecasting). Hopelessness is a potent suicide risk, but temporal relationships with subsequent negative affect (NA) may be biased toward systematic over- or under-estimation of future affect. We examined relationships between these using ecological momentary assessment (EMA) and Dynamic Structural Equation Modelling (DSEM). Methods Undergraduate psychology students ( N = 175) completed baseline questionnaires and 10 days of EMA (five prompts/day) assessing hopelessness, five NA indicators (anger, stress, sadness, worry, anxiety), and momentary SI. We (1) modeled NA trajectories using linear mixed models (LMM); (2) validated the latent NA construct using two-level confirmatory factor analysis (CFA); (3) estimated DSEM parameters; and (4) explored associations between hopelessness and SI using generalized linear mixed models. Results (1) Quadratic time trends best fit NA. (2) CFA fit was acceptable. (3) DSEM showed strong concurrent associations (βHtoN = 0.46, p < .001), modest NA (AR[1]NA = 0.25, p < .001) and hopelessness (AR[1]HPL = 0.09, p = .007) inertia (autoregressive carry-over, AR[1], of a process over adjacent occurrences), small NA to future hopelessness cross-lag (Lag[1]NtoH = 0.08, p = .016), but no significant hopelessness to NA cross-lag ( p = .379). (4) Trait hopelessness predicted baseline suicide risk (β = 0.997, p < .001). State hopelessness predicted momentary SI (β = 0.231, p = .001). Discussion Concurrent but not lagged effects are consistent with projection bias. Robust associations between trait and state hopelessness and SI suggest clinical relevance. Limitations Single hopelessness item, nonclinical sample, small sample of EMA-endorsed SI. Conclusion Momentary hopelessness reflects concurrent and prior NA but poorly predicts future NA. Findings highlight intervention targets focused on affective reactivity and variability.

Publication Title

Journal of Affective Disorders

Volume

398

Link to Full Text

Share

COinS