Factor Structure and Rasch Analysis of the COVID Stress Scale: Evaluating Competing Factor Models

In response to the COVID-19 pandemic, specific assessments such as the COVID Stress Scales (CSS) were developed to measure pandemic-specific distress. The present study aimed to further validate the psychometric properties of the CSS and explore relationships between the CSS and measures of well-being. Adults in the U.S. ( N = 1,388, 63.3% female, 36.1% male, 58.4% Caucasian) completed the CSS, Satisfaction with Life Scale (SWLS), Patient Health Questionnaire-4 (PHQ-4), Brief Resilient Coping Scale (BRCS), and Cognitive and Affective Mindfulness Scale-Revised (CAMS-R) in June and July of 2020. Confirmatory factor analyses on the CSS were used to examine multiple model solutions. The six-factor solution provided the best fit (CFI = 0.98, TLI = 0.98, RMSEA = 0.06), outperforming the five-factor solution, which showed higher RMSEA (.07) and inadequate item separation according to the Rasch analysis. Rasch analysis found no misfitting items and the data fit the Rasch model. Results of the Differential Item Functioning (DIF) analysis supported gender invariance as only negligible DIF was observed across items. As expected, the CSS was positively correlated with measures of anxiety and depression (PHQ-4) and negatively correlated with life satisfaction (SWLS), resilience (BRCS), and trait mindfulness (CAMS-R). Our results partially replicate the factor structure of the CSS found in different adult samples.