Life course epidemiology: Modeling educational attainment with administrative data

Area of Science:

• Life course epidemiology
• Developmental psychology
• Public health

Background:

• Understanding childhood and adolescent factors influencing later life inequalities requires longitudinal data.
• Linkable administrative data can track individuals over time to study pathways of inequity.
• Research using large administrative databases has both advantages and disadvantages.

Purpose of the Study:

• To examine predictors of educational attainment across the life course.
• To identify critical and sensitive periods for social and health influences on high school graduation.
• To utilize administrative data for a comprehensive analysis of developmental trajectories.

Main Methods:

• Longitudinal study of 89,763 children born in Manitoba (1982-1995) followed to age 19.
• Logistic regression used to model five time-varying predictors (3 social, 2 health) of high school graduation.
• Analysis of six temporal patterns, including full, accumulation, sensitive, and critical period models.

Main Results:

• Predictors in early adolescence showed the highest odds ratios for high school graduation.
• Low-income neighborhood residence was a significant predictor of not graduating.
• Risk transmission across developmental periods demonstrated significant implications for later life stages.

Conclusions:

• Life course epidemiology can be advanced by using administrative data to link social, cognitive, and health measures.
• Analyzing temporal patterns is valuable for understanding outcomes like educational achievement and workforce participation.
• Adolescence is a critical period for interventions aimed at improving long-term educational success.