The Pandemic Stressor Scale: factorial validity and reliability of a measure of stressors during a pandemic

Abstract

Background

This study aimed to assess the factorial validity and reliability of the Pandemic Stressor Scale (PaSS), a new measure to assess the severity of distress for different stressors relevant during a pandemic or epidemic.

Methods

The PaSS was administered in N = 2760 German participants. Exploratory factor analysis was used to extract factors. The factor structure obtained in the German sample was examined in N = 1021 Austrian participants using confirmatory factor analysis. χ2, RMSEA, SRMR, CFI, TLI were assessed as global goodness of fit indices for two models (Model 1: nine-factor model; Model 2: nine-factor model combined with a second-order general factor). We additionally assessed factor loadings, communalities, factor reliability, discriminant validity as local fit indices. Internal consistency, item discrimination, and item difficulty were assessed as additional test quality criteria.

Results

The results of the exploratory factor analysis suggested a nine-factor solution with factor loadings accounting for 50.4% of the total variance (Factor 1 ‘Problems with Childcare’, Factor 2 ‘Work-related Problems’, Factor 3 ‘Restricted Face-to-Face Contact’, Factor 4 ‘Burden of Infection ‘, Factor 5 ‘Crisis Management and Communication’, Factor 6 ‘Difficult Housing Condition’, Factor 7 ‘Fear of Infection’, Factor 8 ‘Restricted Access to Resources’, Factor 9 ‘Restricted Activity’). The confirmatory factor analysis showed a sufficient global fit for both tested models (Model 1: χ2 (369, N = 1021) = 1443.28, p < .001, RMSEA = .053, SRMR = .055, CFI = .919, TLI = .904; Model 2: χ2 (396, N = 1021) = 1948.51, p < .001, RMSEA = .062, SRMR = .074, CFI = .883, TLI = .871). The results of the chi-square difference test indicated a significantly better model-fit of Model 1 compared to Model 2 (∆χ2 (27, N = 1021) = 505.23, p < .001). Local goodness of fit indices were comparable for both tested models. We found good factor reliabilities for all factors and moderate to large factor loadings of the items as indicators. In Model 2, four first-order factors showed small factor loadings on the second-order general factor.

Conclusion

The Pandemic Stressor Scale showed sufficient factorial validity for the nine measured domains of stressors during the current COVID-19 pandemic.