After the three review sites have been determined, the Children's Bureau draws two random samples of cases to be reviewed (a total of 150 in-home services cases and approximately 150 foster care cases) from the respective universe of cases in the three sites to be reviewed. The sample of in-home services cases is selected by family, and the sample of foster care cases is selected by child. Before the Children’s Bureau sends the sample of 150 foster care cases to the State, it randomizes the records in the sample. This is designed to preclude any bias when the State selects the cases to be reviewed at each of the three sites.
For in-home services cases, the “universe” of cases is a State-provided list of in-home services cases that were open for services for at least 60 consecutive days during the sampling period and in which no children in the family were in foster care for 24 hours or longer during any portion of the review period. The State should provide this list of in-home services cases to the Children's Bureau because that information is not currently available through the NCANDS or other national data sources. The sampling period for in-home services cases extends 2 months beyond the sampling period for foster care cases, for a total of 8 months. This is because the CFSRs review in-home services cases that were open for at least 60 days.
For foster care cases, the “universe” (list) of cases is the State’s 6-month AFCARS submissions that correspond with the sampling period for the three review sites. To ensure that sites selected for the onsite review will have a sufficient number of the targeted foster care cases for review, the Children's Bureau will sort the AFCARS foster care file by the four categories and by jurisdiction within a State. A table is then generated for each State identifying the jurisdictions and the number of cases in each of the four categories. This assists in the site selection process after sites are proposed through the Statewide Assessment.
Local Site Coordinators then schedule the 65 cases for onsite reviews across the three sites. At each review site, approximately 15-35 cases are reviewed (for example, the Onsite Review Team typically reviews up to 35 cases in the largest metropolitan subdivision and no fewer than 15 in the other two sites), unless otherwise agreed upon by the Children's Bureau and the State. The Children's Bureau, however, will review no fewer than 15 cases at any review site.