Stuck in a Moment: Overcoming the Confound of Means and Variability for Measuring Everyday Emotion Dynamics

Introduction: Studying emotion dynamics in daily life is of central interest to many areas of psychology. However, substantive questions are hampered by a well-known empirical association between the mean and variance. Efforts to address this overlap in emotion dynamics has led to counterintuitive assertions, like neuroticism is not associated with variability in negative emotions. We tested whether shifting to using the median or mode of reported emotions for summarizing emotion patterns and studying their associations with other variables resolves problems with using the mean. Methods: We meta-analyzed 15 ecological momentary assessment (EMA) studies (N=2,088 participants;146,823 observations) to illustrate conceptual and statistical shortcomings of adjusting for the mean and to evaluate whether using the median or mode instead offers a better solution. Results: Results suggest that adjusting variability for the mean overcorrects for their association. We show the mode is less confounded with variability and remains associated with neuroticism over and above mean levels of negative emotion. Conclusion: Adjusting for the mode strengthens evidence for the role of emotion variability in neuroticism. Based on our results, we recommend using the mode instead of or in addition to the mean when measuring average levels of negative emotion in EMA studies.