Response times to items assessing momentary emotional experience have previously been used as an indirect measure of momentary emotional clarity. The rationale is that faster affect ratings reflect greater emotional clarity. When preparing response time data, researchers have to make a number of decisions. Such decisions involve the identification of outliers, the transformation of data to reduce skewness, and the inclusion or exclusion of the first item, among others. The aim of this study was to examine the influence of these decisions on the convergent (within-person) correlation of the indirect (response-time-based) measure of momentary emotional clarity with a direct (self-report) measure of momentary emotional clarity. Thus, we conducted a 3-week ambulatory assessment study (N = 419) and applied a multiverse analysis in which all reasonable decisions regarding the preparation of response time data were included. The results demonstrated that the most influential decision was the decision about whether and how to transform the response time data. Furthermore, we recommend that outliers be replaced by the cutoff (rather than by the mean) or that they be removed. The influences of the remaining multiverse parameters were rather small, indicating that the findings are robust to some extent and largely independent of the majority of data preparation decisions.
Ottenstein et al. (Thu,) studied this question.