Developing Predictive Risk Models to Support Child Maltreatment Hotline Screening Decisions: Allegheny County Methodology and Implementation
Centre for Social Data Analytics, Auckland University of Technology
Ref: SRC-001-USA-002
Accessed: 10/30/2025
Summary
Primary methodology report documenting the AFST's development, data sources, variable selection, model specification (probit and boosted probit), model performance (AUR 0.77 placement, 0.73-0.74 re-referral), external validation against hospital data, implementation details, and discussion of race as a predictor variable. Describes the full sample of 76,964 referrals used for model training.
View Harvard reference
Vaithianathan, R., Putnam-Hornstein, E., Jiang, N., Nand, P. and Maloney, T. (2017) Developing Predictive Risk Models to Support Child Maltreatment Hotline Screening Decisions: Allegheny County Methodology and Implementation. Auckland: Centre for Social Data Analytics, Auckland University of Technology. Available at: https://www.alleghenycountyanalytics.us/wp-content/uploads/2019/05/Methodology-V1-from-16-ACDHS-26_PredictiveRisk_Package_050119_FINAL.pdf (Accessed: 30 October 2025).