Development and Psychometric Validation of the Sudan Women’s Empowerment Index (SWEI-SD): Integrating Factor Analysis, Item Response Theory, and Measurement Invariance

Abstract

Background: Women’s empowerment is multidimensional and context-dependent, yet few psychometrically validated instruments are available for conflict-affected Sudan. This study developed and validated the Sudan Women’s Empowerment Index (SWEI-SD) as a culturally grounded measure of women’s agency. Methods: A cross-sectional psychometric study included 1,100 Sudanese women aged 18 years or older. An initial 42-item pool spanning seven conceptual domains was evaluated using exploratory factor analysis (EFA; n = 550), independent confirmatory factor analysis (CFA; n = 550), bifactor modelling, graded response item response theory, local-dependence diagnostics, differential item functioning (DIF), and measurement-invariance analyses. Item reduction integrated statistical performance and conceptual coverage. Results: EFA supported three factors rather than seven distinct empirical dimensions, while bifactor analysis demonstrated strong general-factor dominance (ECV = 0.830; omega = 0.979; omega hierarchical = 0.959). All 42 items showed satisfactory discrimination and ordered thresholds. Evidence from redundancy, local dependence, item fit, DIF, invariance, and substantive coverage supported reduction to 28 items, retaining four per conceptual domain. The SWEI-SD-28 achieved EAP reliability of 0.944, mean posterior SE of 0.220, and preserved the 42-item latent ranking almost perfectly (Pearson r = 0.990; Spearman rho = 0.987). No practically meaningful DIF remained after recalibration. Among 1,019 eligible respondents, the mean 0–100 score was 54.41 (SD = 21.54). Conclusion: The SWEI-SD-28 provides a concise, reliable, and contextually appropriate measure of women’s empowerment in Sudan. External replication, longitudinal validation, and full WLSMV invariance testing are warranted before implementation