Are emotional clarity and emotion differentiation related?

Abstract

Distinct literatures have developed regarding the constructs of emotional clarity (people's meta-knowledge of their affective experience) and emotion differentiation (people's ability to differentiate affective experience into discrete categories, e.g., anger vs. fear). Conceptually, emotion differentiation processes might be expected to contribute to increased emotional clarity. However, the relation between emotional clarity and emotion differentiation has not been directly investigated. In two studies with independent, undergraduate student samples, we measured emotional clarity using a self-report measure and derived emotion differentiation scores from scenario-based (Study 1) and event-sampling-based (Study 2) measures of affect. We found that emotional clarity and emotion differentiation are: (i) associated to a very small and statistically insignificant degree; and (ii) differentially associated with trait and scenario-based/event-sampling-based measures of affect intensity and variability. These results suggest that emotional clarity and differentiation are distinct constructs with unique relations to various facets of affective experience.

Keywords:

• Emotional clarity
• Emotion differentiation
• Emotion granularity
• Emotional awareness

Acknowledgments

This article is based on data collected for Mügé Dizén's doctoral dissertation at the University of Illinois at Urbana-Champaign.

Notes

¹The LEAS (Lane et al., 1990), a performance-based measure of emotion differentiation, requires all participants to provide answers to a common set of questions for which the correct answer is the same for all individuals and judged by a third party. The LEAS might adequately measure one's understanding of other people's emotions (i.e., affect recognition) since there will likely be a high degree of consensus regarding the “correct” answer. However, the LEAS is not an adequate way to assess emotion differentiation or one's understanding of one's own emotions more broadly because the validity of an observer's inferences regarding another's emotions will necessarily be limited. For example, different individuals respond to the same event with different emotions. Therefore, we have chosen not to include the LEAS in the current research.

²One participant received a negative intraclass correlation coefficient, indicating measurement error. Based on recommendations of Cohen, Cohen, West, and Aiken (2003), we changed this score to zero and proceeded with analyses. Results were near identical to those reported below when removing this participant from further analyses.

³To avoid visual clutter, the following correlations are not included in Figure 1: positive emotion differentiation×negative emotion differentiation=.46; positive emotion differentiation×emotional clarity = −.01; negative emotion differentiation×emotional clarity=.12. Correlations between affect intensity and affect variability residuals were as follows: Scenario-response task (SRT) affect intensity×trait affect intensity=.26; SRT affect intensity×trait affect variability=.17; trait affect intensity×trait affect variability=.39.

⁴Shared between the current study and Study 2 in Disén and Berenbaum (2011) are participants, the ESM-based sampling procedure, and the ESM-based affect intensity and affect variability measures. Therefore, descriptions of these aspects of this study are similar to those presented in Disén and Berenbaum (2011). Additionally, the correlation coefficients between ESM-based measures of affect intensity and affect variability are reported in both articles. All other methods and results reported here are unique to this study and reported for the first time here.

⁵Similar to the event sampling study by Demiralp and colleagues (2012), we measured emotion differentiation by calculating the average Pearson correlation between pairs of affect items. Differentiation scores were highly related across the two computational methods for positive (r=.89, p<.001) and negative affect (r=.84, p<.001).

⁶A sample size of 99 was deemed appropriate for conducting a path analysis with three predictors.

⁷To avoid visual clutter, the following correlations are not included in Figure 2: positive emotion differentiation×negative emotion differentiation=.35; positive emotion differentiation×emotional clarity=.03; negative emotion differentiation×emotional clarity=.07. Correlations between affect intensity and affect variability residuals were as follows: Experience sampling method (ESM) affect intensity – positive affect (PA)×ESM affect variability (PA) = −.00; ESM affect variability (PA)×trait affect variability=.01; trait affect variability×trait affect intensity=.29; trait affect intensity×ESM affect variability – negative affect (NA)=.37; ESM affect variability (NA)×ESM affect intensity (NA)=.65; ESM affect intensity (PA)×trait affect variability = −.05; ESM affect intensity (PA)×Trait affect intensity=.19; ESM affect intensity (PA)×ESM affect variability (NA) = −.04; ESM affect intensity (PA)×ESM affect intensity (NA)=.09; PA affect variability (PA)×trait affect intensity=.12; ESM affect variability (PA)×ESM affect variability (NA)=.57; ESM affect variability (PA)×ESM affect intensity (NA)=.14; trait affect variability×ESM affect variability (NA)=.18; trait affect variability×ESM affect intensity (NA)=.29; trait affect intensity×ESM affect intensity (NA)=.26.

⁸We additionally tested whether affect intensity predicted emotional clarity and emotion differentiation using path analysis. In the first path analysis, using data from Study 1, we found that (i) trait affect intensity significantly predicted positive emotion differentiation (β = − 0.13, p=.05), but not negative emotion differentiation or emotional clarity (ps>.08); and (ii) SRT-based affect intensity significantly predicted positive (β = − 0.23, p<.001) and negative emotion differentiation (β = − 0.21, p=.002), but not emotional clarity (p=.51). In the second path analysis, using data from Study 2, we found that: (i) trait affect intensity significantly predicted positive (β = − 0.21, p=.03) and negative emotion differentiation (β = − 0.21, p=.05), but not emotional clarity (p=.83); (ii) ESM-based positive affect intensity significantly predicted positive emotion differentiation (β = − 0.31, p<.001) and emotional clarity (β = 0.25, p=.009), but not negative emotion differentiation (p=.91); and (iii) ESM-based negative affect intensity significantly predicted negative emotion differentiation (β = − 0.24, p=.02) and emotional clarity (β = − 0.30, p=.003), but not positive emotion differentiation (p=.23).