Agent-based simulation and child protection systems: Rationale, implementation, and verification.

Area of Science:

• Computational Social Science
• Public Health Modeling
• Child Welfare Systems Research

Background:

• Simulation models are under-utilized in child protection research.
• Operational research and decision-making in healthcare and other fields benefit from simulation.
• Evidence-based management of child protection systems is increasingly critical.

Purpose of the Study:

• To describe the rationale and development of an agent-based simulation model for a US child protection system.
• To introduce the OSPEDALE prototype for simulating child welfare system interactions.
• To demonstrate the utility of simulation models as a source of research evidence in child protection.

Main Methods:

• Developed an agent-based simulation model (OSPEDALE) using historical data from 600,000 children in an urban child welfare system.
• Governing equations probabilistically simulate investigation onset and subsequent system interactions (recurrence, referrals, placement).
• Model incorporates child characteristics, risk levels, and prior system involvement to predict outcomes.

Main Results:

• OSPEDALE was tested by comparing simulated admission counts with empirical data.
• The simulation closely replicated the empirical count of new admissions.
• Results justify further investment in the OSPEDALE model for child protection system analysis.

Conclusions:

• Simulation models, like OSPEDALE, provide a valuable source of research evidence for public child protection systems.
• Effectively specified simulation models offer real-world relevance comparable to other research methods.
• The development of OSPEDALE supports the growing need for evidence-based decision-making in child welfare.