Factor structure and psychometric properties of the german version chronic uncertainty scale (CU-20)

Abstract

Abstract Background The experience of uncertainty is ubiquitous and universal across the globe. Many available tools measuring uncertainty are focused on one aspect of uncertainty, e.g., patients with life-threatening illnesses, hence a measure considering (chronic) uncertainty as an integral experience reflect ongoing uncertainties from a socio-cultural perspective is missing. Additionally, current tools do not account for an extended timeframe to measure chronic forms of uncertainty. The objective of this study is to validate a translated German version of the 20 item Chronic Uncertainty Scale (CU-20). Methods The full sample comprised N = 462 participants. Most of the participants were young German citizens and the sex distribution was relatively balanced (60% females; age in average: M = 24.56; SD = 4.78). Using equally split samples, an exploratory factor analysis (EFA) evaluated the CU-20 factor structure, followed by a confirmatory factor analysis (CFA) to test the established factor structure. Measurement invariance between male and female groups was evaluated. Internal consistency of the six-factor model was shown and scale discrimination was shown against chronic stress. Results The EFA results showed decent model fit for the five-factor structure, however based on the CFA results, the theoretically established six-factor model fits the data significantly better. Measurement invariance between male and female groups was shown to be clearly scalar invariant. Cronbach’s alpha, omega and lambda all support internal consistency and reliability of CU-20. Conclusions The CU-20 is a valid and reliable measure of one’s state of chronic uncertainty reflecting the individuals’ experiences of macrosocial forms of uncertainty, compared to the existing ones. This scale is especially useful in the context of migration, refugees or during global crises. Further psychometric testing is required in more diverse samples and a deeper look into measurement invariance is recommended.