Remembering affect between moments: assessing peak-end effects in continuous affect measures

In experience sampling (ESM) studies, participants are measured repeatedly to track momentary affect changes. The data is collected in specific time intervals, leaving gaps between assessments. To capture affective changes between assessments, we developed a measure where participants retrospectively draw their affect in a continuous drawing. This approach, while more complete, may be influenced by retrospective biases, such as the peak-end effect, where intense or recent moments are remembered disproportionately. In an ESM study, 115 participants rated their momentary positive and negative affect six times daily, drew affect changes between assessments, reported events, and completed one daily measurement burst with up to six momentary ratings. We examined the divergence of the retrospective continuous affect drawings from the momentary affect and event ratings, respectively. For positive affect, intensity and recency were associated with lower divergence; the association was nonlinear. Divergence was greatest at low intensity and reduced at moderate to high levels. For negative affect, intensity was associated with higher divergence and recency with lower divergence. The effect of intensity diminished over time, indicating that divergence is mostly determined by recency. Findings were similar for events, except that for negative events, there was no recency effect. Future research should consider the effect of time and intensity, especially the differential effect of positive and negative intensity, on short retrospective ESM measures.