COVID-19 Anxiety Measurement invariance Differential item functioning Rasch analysis Item response theory


March 15, 2024


Introduction: The global COVID-19 pandemic has caused increased levels of anxiety and stress. A recent study reported that 99% of its sample experienced anxiety, stress, stigma, and concerns about potential social isolation. Aims: This study aims to address the lack of a specific COVID-19 anxiety scale in Oman. Methods: The online scale was distributed via Google Form to 431 students, representing 3% of the total student population at Sultan Qaboos University (SQU). The students were enrolled in different colleges of Sultan Qaboos University. Exploratory factor analysis revealed a unidimensional factorial structure consisting of four items. Item loadings ranged from 0.617 (item 4) to 0.847 (item 3). Results: The results indicated that the validity fit indices were excellent. The RMSEA and SRMR values were greater than 0.05 indicating a perfect fit. The X2/DF value was 0.854, with a p-value of less than 0.653, while the GFI, AGFI, IFI, NFI, TLI, CFI, RMR, and RMSEA values were 0.999, 0.995, 1.003, 0.998, 0.959, 1.00, 0.015, 0.000, respectively (90% CI [0.000; 0.074]). The COBAS-4 showed convergent validity through its significant association with similar scales. Discriminant validity was established via its insignificant correlation to distinct constructs. McDonald’s omega and Cronbach’s alpha showed adequate reliability at 0.810 each. The results of multiple group CFA indicated configural, metric, and scalar invariance between male and female students (ΔCFI ≤ 0.01, ΔRMSEA ≤ 0.015). No significant differences were found between the two sexes. The scale was calibrated using the Rasch rating scale, which provided robust results. The Mantel-Haenszel test showed no differential item functioning (DIF) across sexes. Conclusion: The COBAS-4 is a valid, reliable, simple, and easy-to administer self-report instrument for assessing COVID-19-related anxiety.