Problem Statement (v2), combined and refined

• We want to: estimate (1) baseline racial disproportionality in child welfare referrals and acceptances and in active cases; (2) racial disparity in post-referral decision points, both baseline disparity and modeling decision outcomes as a function of race and other characteristics; (3) racial disparity in foster care outcomes, both baseline disparity and modeling outcomes as a function of race and other characteristics;
• Using: administrative data for Charlottesville from the Department of Social Services on (1) referrals for last three years and ongoing cases; (2) referrals for last three years (and ongoing cases?); (3) foster care cases and placement history for (last 17 years?);
• So that: we can determine the presence and/or extent of racial disproportionality and disparity across multiple outcomes and estimate the significance of race as a factor in shaping these outcomes;
• So that: the DSS can understand where disparities occur in order to better address them and the Charlottesville community can make progress in ensuring all families and children are treated fairly.

First Draft of Problem Statements, three takes

Based on the DataKind model

Take 1

• We want to: compare referrals, process, and outcomes for children of differnet races;
• Using: statistical analysis, visualization, decision point parameters and population parameters;
• So that: we can determine whether there is racial disproportionality and/or disparity in the Charlottesville Child Welfare System, and report that information to inform policy;
• So that: all children are treated fairly irrespective of race.

Take 2

• We want to: visualize and subset data;
• Using: foster care data;
• So that: we can determine the most significant factors that lead to children being placed in (or exiting) foster care;
• So that: the Charlottesville community can address specific groups or needs.

Take 3

• We want to: statistically examine (1) if the proportion of referrals called in, screened in, and founded is the same for white and non-white children; and (2) if there is a difference in length of stay and placement between white and non-white children;
• Using: the DSS data;
• So that: we can examine if there is disproportionality;
• So that: disproportionality/disparity can be addressed with changes in implementation to the process.