Poster #214 - Psychometric Properties of the Children's Emotion Management Scales: Structure across Strategy and Emotion

• In Event: 1-169 - Poster Session 04
  In Poster Session: PS 04 Section - Social, Emotional, Personality

Integrative Statement

Regulating one’s emotions is a vital skill for effective youth adjustment. The Children’s Emotion Management Scales (CEMS) are popular measures of emotion regulation used in over sixty studies. A particular strength is that these scales assess regulation for three discrete negative emotions: sadness, anger, and worry. In separate scale-development papers, factor analyses suggested a three-factor structure (with subscales labeled inhibition, dysregulation, and coping) for regulation of each of the emotions (Zeman et al., 2001; Zeman et al., 2002; Zeman et al., 2010). To date, no study has replicated the original factor analyses, a limitation given that any factor analysis may capitalize on chance within a sample. Further, the overarching structure of regulation across these three emotions remains unknown. This, despite the fact that subscales within emotion and across emotions have been combined in various ways for analyses in numerous studies. It is not clear if combining subscales is empirically supported. Thus, our first aim was to replicate the three-factor structure for regulation of each of the discrete emotions. Given strong reliabilities found in previous studies using the CEMS, we hypothesized that we would replicate the structure for each CEMS scale. Our second aim was to investigate the overarching structure of all CEMS items across emotions. Without past literature to guide us, we did not form a particular hypothesis. Rather we tested competing plausible models, including 9-factor, hierarchical, and overarching strategy- and emotion-factor models.

Participants were 490 youth in two grade cohorts (213 in grades 6 & 7; 277grades 9-12; overall 60% girls; 73% White; Mage=13.96, SDage=2.48). As part of a larger study, adolescents completed the Children’s Emotion Management Scales, including the 11-item Anger Management Scale, the 12-item Sadness Management Scale, and the 10-item Worry Management Scale. For each of the CEMS, inhibition items generally assess hiding or suppressing emotions, dysregulation items assess outward-focused responses like crying and carrying on, and coping items assess calm, in-control, and problem-solving responses to emotions.

Confirmatory factor analyses used the WLSMV estimator with item responses considered ordered categorical variables. All fit statistics from models calculated are in Table 1. For aim 1, the 3-factor structure fit the data well for the anger scale. Neither the 3-factor structure for sadness nor worry fit the data. Allowing one item to crossload on the sadness scale and removing one item on the worry scale each led to excellent fit for their respective scales. For aim 2, only one model had acceptable fit. Specifically, results suggested that items could be modeled as being influenced by higher-order emotion factors (e.g., all worry items loading on a Worry factor) and by higher-order strategy factors (e.g., all inhibition items loading on an Inhibition factor). This suggests that to some extent adolescents engage in similar regulation strategies across negative emotions, but also that regulation is partly influenced by the type of emotion being regulated. Additional analyses will examine invariance by gender and grade cohort. Results will be discussed within the context of developmental theories of the structure of emotion regulation.

Authors

• Aaron Luebbe, Miami University
  Presenting Author

• Feven Asresehei Ogbaselase, Miami University
  Non-Presenting Author