Abstract
This study sought to examine the association between child protection mediation and permanency outcomes for children in foster care. Propensity Score Matching was used to match 315 mediated cases with 315 non-mediated cases that were resolved through the traditional adversarial process (N = 630). Logistic regression was used to examine outcomes. The findings indicate that participation in mediation did not have a discernable effect on whether permanency was achieved. The results suggest that the phenomena of permanency may be better explained not by one or two specific factors, but rather a combination of child, family, agency, court, and community factors.
Notes
*p ≤ .05
**p ≤ .001.
† p ≤ .10
*p ≤ .05
**p ≤ .01
***p ≤ .001.
*p ≤ .05
**p ≤ .01
***p ≤ .001.
1. Although the evaluation of the mediation pilot project extended from 1999 to 2005, CPS administrators indicated that administrative data for children prior to FY 2002 for some variables was not consistent with data compiled by the agency after FY 2002. Therefore, the decision was made to exclude data prior to FY 2002 and limit the data requested to children removed between FY 2002 and FY 2005.
2. Chi-square and t tests revealed no significant differences between the 19,418 children in the final merged data set and the subsample of 10,483 children used to select a comparison group of children whose families did not utilize mediation.
3. The 317 mediated cases included in the 10,483 subsample of “random and only children” were compared with the 323 mediated cases that were not included in the subsample. Comparisons were made between the two groups of mediated cases (selected cases/cases not selected) to discern if there were any differences on observed characteristics to be used in the PSM matching process. However, no significant differences were noted on any of the following variables: child's age at removal, child's gender, child's minority status, reasons for the child's removal, the geographic location of the court of jurisdiction for the case, and the year the child entered the foster care system. In addition, no significant differences were noted with regard to placement stability or individual race/ethnicities.