- October 2017
- Vol. 18, No. 7
Safeguards Against Misuse of Predictive Analytics
A new report urges caution when using child welfare data to predict future trends and offers suggestions for safeguarding against its misuse.
The 2017 report from the Kirwan Institute for the Study of Race and Ethnicity at Ohio State University acknowledges the potential benefits from using predictive analytics in child welfare (e.g., revealing patterns of social disparities and determining the best use of limited public and private resources). The report also offers reasons to be wary of this practice, including the risk of perpetuating certain cognitive and structural biases.
To guard against the misuse of predictive analytic tools, the report recommends the following:
- Develop a code of ethics to help guide how the field uses these tools and other applications to analyze large sets of data.
- Increase accountability by explaining algorithms and giving families recourse to dispute data-driven decisions to ensure greater transparency in decision-making.
- Gauge the equity impact of these tools through a thorough evaluation of predictive analytic models to address the disparities in child welfare.
- Broaden the scope beyond individual-level risk predictions to include neighborhood and citywide mapping to identify systemic risk factors such as poverty and lack of opportunity.
Foretelling the Future: A Critical Perspective on the Use of Predictive Analytics in Child Welfare is available at http://kirwaninstitute.osu.edu/wp-content/uploads/2017/05/ki-predictive-analytics.pdf (209 KB).