About | HeinOnline Law Journal Library | HeinOnline Law Journal Library | HeinOnline
LOG IN
LOG IN

35 J. Fam. Violence 1 (2020)

handle is hein.journals/jfamv35 and id is 1 raw text is: Journal of Family Violence (2020) 35:1-13 https://doi.org/10.1007/si0896-019-00074-y ORIGINAL ARTICLE Check fo'r Preventing Infant Maltreatment with Predictive Analytics: Applying Ethical Principles to Evidence-Based Child Welfare Policy Paul Lanier 1'23 - Maria Rodriguez4 - Sarah Verbiest2,s - Katherine Bryant2,5 - Ting Guan' - Adam Zolotor3,6 Published online: 7 June 2019 C Springer Science+Business Media, LLC, part of Springer Nature 2019 Abstract Infant maltreatment is a devastating social and public health problem. Birth Match is an innovative policy solution to prevent infant maltreatment that leverages existing data systems to rapidly predict future risk through linkage of birth certificate and child welfare data then initiate a child protection response. Birth Match is one example of child welfare policy that capitalizes on recent advances in computing technology, predictive analytics, and algorithmic decision making. We apply frameworks from business and computer science as a case study in ethical decision-making in child welfare policy. Current Birth Match policy applications appear to lack key aspects of transparency and accountability identified in the frameworks. Although technology holds promise to help solve intractable social problems such as fatal infant maltreatment, the decision to deploy such policy innovations must consider ethical questions and tradeoffs. Technological advances hold great promise for prevention of fatal infant maltreatment, but numerous ethical considerations are lacking in current implementation and should be considered in future applications. Keywords Family violence - Child welfare - Infants, decision making Algorithms and the data that drive them are designed and created by people - There is always a human ulti- mately responsible for decisions made or informed by an algorithm. The algorithm did it is not an accept- able excuse if algorithmic systems make mistakes or have undesired consequences, including from machine-learning processes. W Paul Lanier planier@unc.edu School of Social Work, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA 2 Jordan Institute for Families, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA 3 Injury Prevention Research Center, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA 4 Silberman School of Social Work, Hunter College, New York City, NY, USA s Center for Maternal and Infant Health, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA 6 School of Medicine, Department of Family Medicine, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA -Fairness, Accountability, and Transparency in Machine Learning (FAT/ML) ethical premise Introduction Similar to most sectors of government in the United States, the public child welfare system (CWS) is investing in the use of new technology and shifting toward data-driven, computer- powered policy and practice. Over the past decade, greater availability of big data and high-speed machine learning and computing has fostered the perception that technology can help solve some of the wicked child welfare problems (i.e., persistent problems that defy ordinary solutions; Chouldechova et al. 2018; de Haan and Connolly 2014; Kulkarni et al. 2016; Russell 2015). Increased use of data and computing technology signals at least two major shifts or innovations in child welfare policy and practice. First, the value of clinical prediction based on individual experience and training is replaced by an increasing availability of mathemat- ical actuarial prediction and judgement (Brauneis and Goodman 2018, p. 111). In other words, the numerous human decision points involved in child welfare services are increas- ingly informed by, and in some cases determined by, computer t Springer

What Is HeinOnline?

HeinOnline is a subscription-based resource containing thousands of academic and legal journals from inception; complete coverage of government documents such as U.S. Statutes at Large, U.S. Code, Federal Register, Code of Federal Regulations, U.S. Reports, and much more. Documents are image-based, fully searchable PDFs with the authority of print combined with the accessibility of a user-friendly and powerful database. For more information, request a quote or trial for your organization below.



Short-term subscription options include 24 hours, 48 hours, or 1 week to HeinOnline.

Purchase Access

Contact us for annual subscription options:

Contact Us

Already a HeinOnline Subscriber?

Log In

profiles profiles most