Algorithmically Based Decision Support Tools: Skeptical Thinking about the Inclusion of Previous Involvement

Abstract

Decision making in social work, especially in child protection services, is increasingly being assisted by algorithmically based decision support tools. These tools assign a quantified level of risk of child maltreatment, or what is considered a proxy for child maltreatment, and have been used to assist decisions about which referrals to screen in for further investigation. These tools may have been trained using vast amounts of data held in public services databases and perhaps dozens of characteristics of both parents and children. Characteristics, or predictors, are given a weighting in terms of their correlation with child maltreatment and one characteristic that has been given a high weighting is that of previous involvement of a child or parent with child protection services. In this article, through a process of ‘skeptical thinking’, the different meanings of previous involvement are explored in relation to parents and children and child protection agencies. The aim is to further debate about the development and use of algorithmically based decision tools and highlight the role social workers can play in this process.

Keywords:

• decision support tools
• algorithm
• previous involvement
• child protection

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This research is supported by the Australian Research Council through the award of a Future Fellowship (FT170100080).

Notes on contributors

Philip Gillingham

Dr Philip Gillingham Correspondence to: Australian Research Council Future Fellow, School of Nursing, Midwifery and Social Work, University of Queensland. Email: [email protected]