Predictive analytics tools are an emerging strategy to making data-driven decisions and enhancing decision-making about child welfare services and child protection interventions. These tools are being employed at a variety of locations across the country. Use the resources below to learn about the use of predictive analytics tools, their implementation, and results. 

Developing Predictive Risk Models to Support Child Maltreatment Hotline Screening Decisions (PDF - 5,970 KB)
Allegheny County Department of Human Services (2019) 
Provides an overview of the Allegheny Family Screening Tool, a predictive risk-modeling tool being used in Pennsylvania to improve child welfare screening decisions. This publication includes eight documents that review the tool.

The Foster Care System Turns to Big Data: Promising or Profiling? 
Loudenback (2022)
The Imprint
Describes the use of predictive analytics over time in Los Angeles County. 

Implementing Executive Order 50 (2019): Summary of Agency Compliance Reporting (PDF - 566 KB) 
Thamkittikasem (2020)
City of New York Office of the Mayor
Provides a directory of algorithmic tools being used in New York City, including the tool being used by the Quality Assurance Unit within the Division of Child Protection to review child welfare investigations and support the selection of cases for review. 

Implementing the Hello Baby Prevention Program in Allegheny County: Methodology Report (PDF - 1,024 KB) 
Vaithianathan, Putnam-Hornstein, & Benavides-Prado (2020)
Auckland University of Technology, Centre for Social Data Analytics
Discusses a predictive risk-modeling tool, Hello Baby, that uses proactive outreach and matches family’s needs to existing support services. The Hello Baby program allows Allegheny County (PA) to overcome the limitations of traditional referral pathways, which rely on families or providers for referrals to services.

Oregon DHS Safety at Screening Tool — Development and Execution (PDF - 430 KB)
Oregon Department of Human Services (2019)
Discusses a predictive risk-modeling tool being used in Oregon to support child welfare decision-making. 

What Is Predictive Risk Modeling and How Can It Be Applied to Child Welfare?
Casey Family Programs (2018)
Presents a series of videos and information from a child welfare convention during which leaders and experts shared innovative strategies being used to improve child safety and reduce child abuse and neglect fatalities. The videos highlight the predictive risk model being used in Allegheny County (PA) to support effective hotline decision-making.