Operationalizing undifferentiated affect: Validity and utility in clinical samples

Abstract

Emotion differentiation is conceptualized as the process of categorizing one’s general affective experiences into discrete emotions. The experience of undifferentiated affect or the inability to distinguish the particular emotion or combination of emotions that one is experiencing is often considered a hallmark of emotion dysregulation. Some past research has attempted to operationalize the general tendency to experience undifferentiated affect at the trait level using explicit questionnaire measures. More recently, indirect measures using intraclass correlation coefficients (ICCs) to estimate the consistency between simultaneous measures of different in-the-moment emotional experiences have become the favored method of quantifying undifferentiated affect. While the ICC method constitutes an advancement in estimating undifferentiated affect, which is theorized to be a dynamic process that occurs at a very granular level, prior investigations have used aggregate ICC measures or momentary ICC derivations that ignore multiple sources of dynamic variability to make inferences about in-the-moment experiences. We introduce a new, flexible method of calculating ICC measures of undifferentiated affect at different levels of experience that takes full advantage of time-intensive data measurement and more closely maps onto the theorized process. This method provides more refined estimates of undifferentiated affect and its associations with various behavioral outcomes, as well as uncovers more nuanced associations regarding the temporal process of emotional differentiation. It also elucidates potential conceptual issues in mapping empirical estimates of emotion undifferentiation onto their underlying theoretical interpretations.