Virginia Eubanks (“A Child Abuse Prediction Model Fails Poor Families” argues that the AFST (Allegheny Family Screening Tool) disproportionately targets poor families. Identify a specific feature of the algorithm that causes this to be the case. Describe it in 2-3 sentences, including a quotation from the Eubanks article. Finally, state whether you think these flaws in the algorithm could be corrected with better programming, or if you think residents of Allegheny County would be better served by relying on human workers (1-2 sentences).
AFST is a tool being used locally by the Allegheny County Department of Human Services to determine a child’s risk of being abused, and defend or protest its implementation.
"We definitely oversample the poor,” says Erin Dalton, Director of Allegheny County’s Office of Data Analysis, Research and Evaluation. “All of the data systems we have are biased. We still think this data can be helpful in protecting kids.” (Eubanks article)
When an allegation of maltreatment is received and the call screener enters details into the child welfare case management system (KIDS), a click will automatically generate the AFST score.
Pros
Limitations
In order to increase the accuracy of the tool , the call screeners and their supervisors need to focus on the immediate allegation, not the child’s future risk a year or two down the line. They should change their mindset to screen out the lowest-risk cases and screen in the highest-risk cases.
Virginia Eubanks (“A Child Abuse Prediction Model Fails Poor Families” argues that the AFST (Allegheny Family...