Purpose: Adolescent school dropout remains a major concern in the United States, accounting for 5.3% of the entire high school students in 2022. This study aims to investigate the relationships between socioeconomic status (SES), the availability of mental health and substance use treatment facilities, and school dropout rates across thirteen states.
Method: We used geospatial and machine learning techniques to integrate datasets and impute missing values. Regression analyses were used to examine the relationships among SES, service availability, and dropout rates.
Results: Our findings suggest that computational techniques can help conduct multi-state analysis of educational data. The availability of services negatively moderates the relationship between SES and dropout.
Discussion: This study demonstrates the potential of geospatial and machine learning techniques for multi-level analysis. The findings provide implications for social work, public policy, and education, indicating that community-level resources may serve as protective factors in supporting youth educational continuity.
* This article has been accepted to the Special Issue on Emerging Computational Approaches for Social Work Research and Practice.
* Myeong Lee is the corresponding author.