Overview: Human services organizations have increasingly turned to data-driven surveillance and decision-making approaches in their work with individuals and families. Presenters will discuss theoretical frameworks and the results of a descriptive policy analysis to examine ways that educators can prepare practitioners to effectively, ethically, and equitably evaluate and use such tools.Proposal text: Data-driven approaches to surveillance and decision-making, such as the use of big data, predictive analytics, and algorithmic tools, have become increasingly-common strategies for child welfare agencies and other human services organizations as they seek to improve the comprehensiveness and consistency of their practice. On one hand, the use of data-driven decision-making is ubiquitous in industry and in many areas of government, and it has consistently been shown to produce more accurate and less variable decisions (Coulthard et al., 2020; Kahneman et al., 2021). On the other hand, scholars and family advocates have called for the implementation of such strategies to be guided by clear ethical, equitable, and evaluative frameworks that protect families’ human rights, due to the potential for historical patterns of human racial and socioeconomic bias to be replicated and even amplified in predictive models and tools (Gillingham, 2019; Lanier et al., 2020; Rodriguez et al., 2019). The evidence for promise and caution is further complicated by indications from the structured decision-making literature that suggest that child welfare practitioners often do not trust these more quantitative and rational approaches to their work and tend to override or ignore their intended influences (Bosk, 2018; Gillingham & Humphreys, 2010; Rea & Erasmus, 2017).
Accordingly, social work practitioners in child welfare and other human services fields are likely to find themselves entering into practice environments in which data-driven decision-making approaches of various types are rapidly becoming the norm, and they will need a robust understanding of the justifications, intended benefits, and possible risks related to these types of tools. Further, they will need to be equipped with the necessary competencies to advocate for effective and ethical tool design and to integrate the use of data-driven insights into their clinical judgment and practice. This proposed presentation will (1) summarize current data-driven initiatives in child welfare and human services fields, (2) provide attendees with the relevant literature pertaining to decision-making theories such as naturalistic decision-making and dual-process theory (Kahneman et al., 2021; Klein, 2015), (3) introduce emerging ethical and evaluative frameworks developed by data science (Diakopoulos et al., n.d.) and social work researchers (Drake et al., 2020) and (4) describe the results of the presenters’ recent descriptive policy analysis of a child welfare data-driven initiative known as Birth Match as an example of these principles and their implications. Special attention will be given to the ways in which this information can be incorporated into social work curricula and trainings so that practitioners may be equipped to produce positive outcomes for their clients while safeguarding their clients’ rights to fair treatment and privacy in this evolving technological landscape.
Learning Objectives:
Describe current initiatives to incorporate data into child welfare decision-making, including the use of algorithmic tools, predictive analytics, and advanced data visualization.
Use theoretical and ethical frameworks to analyze and evaluate the usefulness, impact, transparency, and fairness of data-driven decision-making strategies.
Prepare social work students interested in the human services field with the knowledge and skills required to effectively assess and use data-driven decision-making tools, with an emphasis on securing clients’ rights to well-being, fair treatment, and privacy.
