Predictive and prescriptive analytics, machine learning and child welfare risk assessment: The Broward County experience

Abstract

This paper presents the findings from a study designed to explore whether predictive analytics and machine learning could improve the accuracy and utility of the child welfare risk assessment instrument used in Broward County (Ft. Lauderdale, Florida). The findings from this study indicate that, indeed, predictive analytics and machine learning would significantly improve the accuracy and utility of the child welfare risk assessment instrument being used. If the predictive analytic and machine learning algorithms developed in this study would be deployed, there would be improved accuracy in identifying low, moderate and high risk cases, better matching between the needs of children and families and available services and improved child and family outcomes. This paper also identifies further areas of research and study.