|
Citation
|
Drake B, Barth RP, Berrick JD, Garcia A, Greeson JKP, Jonson-Reid M, Kim H, Kohl PL, Putnam-Hornstein E. Child Abuse Negl. 2023; 146: e106442. |
|
Abstract
|
Measurement
The FFCWS consists of a purposive 3:1 oversample of urban, largely unmarried mothers, yielding a disproportionately Black, Hispanic, and low-income sample (Princeton University, n.d.). Evangelist et al. cited prior analyses of the cumulative risks of CPS reports among all U.S. children (Kim et al., 2017). The Kim et al. study estimated CPS investigations by age 9 at 26.5 % for all U.S. children and 37.8 % for Black children. Meanwhile, Evangelist et al.'s caregiver-reported CPS contact rate was...
Causal inference
The purpose of the study was to determine if CPS contact has a unique toxic effect separate from any toxic effects caused by child maltreatment. This is no easy task. Identifying the "consequences of CPS contact" (p. 4) requires fully parsing the well-known and powerful negative impacts of child maltreatment, which can substantially increase negative outcomes. Even a very modest failure could account for the generally small effects found. Evangelist et al. used an advanced form of propensity...
Conclusion
Although many studies suffer from some of the limitations we described, the Evangelist et al. study must reach an unusually high bar, given that the authors position their study as advancing knowledge of "the consequences of child welfare involvement" (p. 3) and argue that their model "mirrors the intuition" of randomized controlled trials (p. 5). It is generally understood that correlational studies have imperfect controls, cannot make causal claims, and are fundamentally distinct from...
M. Evangelist et al.
Child protective services contact and youth outcomes
Child Abuse & Neglect
(2023)
Language: en |