Handling missing data in complex surveys: A case study of children’s functional disorders and their correlates

Studies based on complex survey designs often overlook the importance of rigorous statistical analysis, particularly with regard to sampling design and the handling of missing data. This study employs a two-stage statistical approach to investigate factors associated with functional difficulties among children aged 5–17 years. In the first stage, the missing data mechanism is identified, and an appropriate statistical method is applied to address the missing values. In the second stage, a survey logistic regression model is employed to identify key factors associated with children's functional disorders. The analysis utilizes data from the 2018–2019 Multiple Indicator Cluster Survey conducted in the Central African Republic. The results reveal significant associations with the child’s age, educational attainment, mother’s functional disorders, household wealth, and geographic region. Specifically, younger children, those with limited education, those whose mothers experience functional difficulties, and children from lower-income households are at greater risk. This study underscores the necessity of applying robust statistical methods to complex survey data and provides evidence to guide inclusive policies and targeted interventions aimed at supporting vulnerable children and youth with disabilities.