Determinants of educational attainment in Ghana: ordered probit versus generalized ordered logit approach

This study investigates the determinants of educational attainment in Ghana, with a focus on disparities across educational levels, using data from the Ghana Living Standards Survey (GLSS) Round 7. Employing ordered probit, generalized ordered logit, and multinomial logit models, the analysis examines how individual and household factors influence progression through four ordered educational categories: no formal education, primary/Junior High School (JHS), secondary/Senior High School (SHS), and tertiary education (including vocational/technical/commercial, undergraduate, and postgraduate levels). The results identify several significant predictors of educational attainment, including gender, marital status, disability, region, place of residence, parental education, labor force status, age, income, and household size. Specifically, males, individuals who have never married, persons without disabilities, urban residents, those with highly educated parents, individuals who are not in the labor force, persons from high-income households, and those from smaller households are more likely to attain higher levels of education compared to their respective counterparts. Additionally, the study finds a quadratic relationship between age and educational attainment, whereby increasing age is associated with a higher probability of achieving advanced education among younger individuals, but beyond a certain age threshold, this probability declines. The proportional odds assumption is rejected, supporting the generalized ordered logit as the preferred specification, while multinomial logit results confirm robustness. These findings underscore the need for equity-focused interventions to address structural and socio-economic barriers in Ghana’s education system, offering evidence for policy alignment with Sustainable Development Goal 4.