Proceedings of the Annual Meeting of the Cognitive Science Society

Previous studies have reported contradictory findings, showing that emotion can both enhance and interfere with working memory (WM). We propose that these inconsistencies partly arise from differences in how emotional effects are conceptually defined, rather than from emotional properties alone. Using a smartphone-based experience sampling method (ESM), we repeatedly assessed multidimensional emotional states and performance on two WM tasks over approximately three months in daily life. We conducted two analyses on the same dataset: one defining emotional effects based on emotional states at the measurement occasion, and another defining emotional effects as within-person deviations from each individual's typical emotional level using linear mixed-effects models. Analyses based on emotional states yielded mixed enhancement and interference effects depending on tasks and emotional dimensions. In contrast, emotional deviations from typical level were more consistently associated with interference effects on WM performance. Together, these findings demonstrate that how emotional effects are defined critically shapes their interpretation and consistency in emotion–WM research.