Emotion and emotion regulation roBert w. Levenson
Emotion dynamics in couples interactions In typical laboratory studies of emotion, stimuli are standardized and the timing of stimulus and response are the same for all participants. In our new approach to studying emotion in couples, conversations were unrehearsed and idiosyncratic. Accordingly, emotional moments were different in kind and timing for each couple. Because of this, conventional data averaging techniques, which are useful for identifying common signals amidst random noise in classic experiments, are not very helpful. Figure 3 illustrates the problem. The top panel depicts a husband’s second-by-second rating of the valence (1 = very negative, 5 = neutral, 9 = very positive) of his emotional experience during a five minute period of sitting silently across from his wife, followed by a 15 minute discussion of a marital problem. This panel illustrates the ebbs and flows of emotion that are typical of these kinds of data. The bottom panel depicts the result of applying second-by-second averaging across 151 of these ratings obtained from husbands in a study of long-term marriages (Levenson, Carstensen, and Gottman, 1994). The resultant average is fairly featureless, reflecting the fact that there are few features in common across participants when these kinds of data are obtained in this manner.