UC Berkeley

Recent theoretical and empirical contributions to the emotion regulation and close relationships literatures have begun to elucidate the dynamic complexities and functional consequences of interpersonal emotion regulation (IER) processes. The present study sought to evaluate a naturalistic method for capturing IER processes in the daily lives of committed romantic couples by using ambulatory acoustic recording in tandem with ecological momentary self-report to measure enacted behavior and subjective experience. A coding manual was developed for this purpose. Then, a community sample of cohabiting romantic couples was recruited (final N = 59 couples, 118 participants) to consider the feasibility, reliability, and validity of this method. Couples completed a thorough baseline assessment, followed by eight days of ecological momentary assessment and ambulatory acoustic recording with the Electronically Activated Recorder. A team of undergraduate and post-baccalaureate coders was trained to identify and code the content of acoustically observable IER-relevant interactions that occurred within the couples. Participants’ self-reported evaluations of wearing the audio recorder, total number of recordings produced, the characteristics of the IER-relevant interactions that coders identified, inter-rater reliability among the coders, the correlational structure of the coder ratings, and correspondences between coder ratings and participant self-report were analyzed. Consistent with past research, participants generally reported that wearing the audio recorder was only modestly obtrusive for themselves and others. In total, participants produced nearly 2200 hours of audio. Using a periodic sampling strategy, coders identified 836 unique IER-relevant interactions. Many of these interactions appeared to have been “invisible” to the participants in that they reported not having engaged in IER during those epochs in ecological momentary self-report surveys. Inter-rater reliability ranged from tolerable to excellent for the coder ratings, albeit some codes were rarely endorsed in this sample. There were systematic correspondences between coder ratings and participant self-report, although the magnitude of these correspondences was generally quite modest. Overall, data from this study suggest that this approach to recording and coding spontaneously occurring IER interactions is feasible, reliable, and valid.