Migraine screen questionnaire: Further psychometric evidence from categorical data methods

Abstract

Background: Psychometric investigations of tools used in the screening of migraine including the migraine screen questionnaire (MS-Q), using an adequate statistical approach is needed. We assessed the psychometric properties of the migraine screen questionnaire (MS-Q) using categorical data methods. Material and methods: A total of 343 students at Mizan-Tepi University, Ethiopia, age range = 18-35 years were selected by a simple random sampling method to participate in a cross-sectional study. The respondents completed the MS-Q, a semi-structured socio-demographic questionnaire, and a visual analog scale for attention (VAS-A). Results: The cumulative variance rule (> 40%), the Kaiser's criteria (Eigenvalue> 1), the Scree test and, the parallel analysis (minimum rank) identified a 1-factor model for the MS-Q with the factor loadings in the range of 0.78 to 0.84. Fit indices favored a 1-factor model of the MS-Q as indicated by comparative fit index (0.993), weighted root mean square residual (0.048), root mean square error of approximation (0.067), the goodness of fit index (1.00), and non-normed fit index (0.987). The values of the Factor Determinacy Index (0.953), marginal reliability (0.909), H-latent (0.909), H-observed (0.727), explained common variance (0.906) and the mean item residual absolute loadings (0.225) further complimented finding of the 1-Factor model. McDonald's Omega (0.903) suggested adequate internal consistency. Discriminative validity was supported by significantly higher scores for the total and all the MS-Q items except one among those with complaints of attention. Conclusion: The categorical methods support the psychometric validity of the MS-Q in the study population.