A trifactor model for comparing retrospective and momentary assessments of emotions.

It has long been known that different processes underlie momentary and retrospective reports of the same emotions. When answering retrospective questions about emotions, respondents rely on episodic memory and thus engage a variety of heuristics, leading to a so-called "memory-experience gap". The current study uses ecological momentary analysis (EMA) to systematically model the gaps between three types of measures of positive and negative emotions: momentary, daily retrospective, and weekly retrospective. Participants (N = 223) responded to momentary, daily retrospective, and weekly retrospective versions of the modified Differential Emotions Scale (mDES) of positive and negative affect for five weeks. We used a novel application of the trifactor model (Bauer et al., 2013), originally developed to reconcile scores from multiple raters, to systematically model the differences between momentary, daily retrospective, and weekly retrospective ratings of emotions. The measurement properties of individual items -- i.e., factor loadings and item endorsement thresholds -- were largely invariant across the three types of ratings. However, there was a systematic upward bias for daily retrospective measures of both positive and negative affect, relative to momentary measures; this bias was even greater for weekly measures. Relations between these discrepancies and personality measures are also discussed.

https://psychology.wfu.edu/veronica-cole/