Structured variation in daily life experience within and across individuals

Human experience varies across contexts and individuals. Yet, psychological studies typically seek to constrain rather than discover structured variation. Here we leverage a powerful alternative approach to examine variability in emotional experience by sampling multimodally from individuals across a variety of contexts throughout their daily life and employing unsupervised machine learning to discover reliable, person-specific, multimodal patterns of experience. Healthy adults (N = 97) completed 14 days of biologically triggered experience sampling. Participants wore mobile cardiac monitors for 8 hr/day and reported their current valence, arousal, primary activity (e.g., relaxing), social context (e.g., alone), and emotional experience (via free report) when prompted. Prompts were initiated whenever a sustained change in cardiac interbeat interval was detected in the absence of gross bodily movements (and twice randomly each day). From each event (10,755 total, M = 110.9 events/person), we extracted cardiovascular, postural, affective, and contextual features, which were submitted to integrative clustering to derive reliable patterns for each individual. Clustering identified 313 multimodal patterns (M = 3.2 patterns/person). Multimodal patterns varied across people in number and, to varying extents, in nature. The features that best distinguished patterns from one another also varied by person. Critically, all participants mapped many emotion labels to a single multimodal pattern and a single emotion label to many patterns (i.e., a many-to-many mapping). Simulated data sets validated our novel analytic strategy. This comprehensive approach has broad utility for discovering structured variation in emotional experience "in the wild" and provides further evidence that emotions are diverse populations of instances constructed by a perceiver in context. (PsycInfo Database Record (c) 2026 APA, all rights reserved).