Search reports, ask a question, or jump to a page.
The problem
Intervention: Child welfare service referral and interventions provided to children at risk of maltreatment
Who it was tested with: 1,708,565 children born in England 1995-1998, with analyses by sex (male and female subgroups).
Experience from people who deliver programmes: stories, case studies and lessons learned. Valuable for understanding how and why, and for spotting what to test next, but not a measure of effect.
"Multilevel logistic regression models for the whole population, and male and female subgroups, predict the odds of receiving any criminal caution or conviction by 2020 across all levels of child welfare service involvement, while accounting for other covariates." The study uses administrative data to examine associations between service involvement and later criminal outcomes, but the observational design means it functions as a longitudinal study documenting differential outcomes rather than a causal evaluation.
Observational design cannot establish causation; cannot distinguish whether elevated criminal contact is due to welfare service involvement, underlying vulnerabilities, or confounding factors; does not examine variation in intervention quality or type
Across countries and designs
1 study · 1 country
