Information Integration in Child Welfare Cases: An Introduction to Statistical Decision Making

Abstract

Workers in the field of child maltreatment are required to make many complex and far-reaching decisions every week. In this article, two general methods for formulating decision-making policies are presented, along with a discussion of the considerable research literature demonstrating the superior predictive validity of statistical decision models over clinical prediction. A series of illustrative contrasts between the two approaches highlights the desirable mathematical properties of statistical equations as well as the cognitive biases and limitations inherent in unaided human judgment. Reasons for practitioners' adherence to the clinical approach are explored, with specific reference to child welfare decision making. Finally, recommendations are provided for enhancing the efficiency, validity, and ethical defensibility of decision making that seriously impacts the lives of children and their families.

Decision making in child protection: An international comparative study on maltreatment substantiation, risk assessment and interventions recommendations, and the role of professionals’ child welfare attitudes

“…The present study indicates that the impact of attitudes on risk assessments may be a universal phenomenon and should be addressed by training and policy. For instance, the use of structured decision making instruments may help reduce inter-worker inconsistencies (see Ruscio, 1998) that reflect personal preferences, rather than valid professional knowledge. It should be noted, however, that there is also evidence to indicate that "over-structuring'" the assessment process and eliminating professional secretion and individual differences may also have negative consequences (Munro, 2011).…”

Contribution of child protection workers’ attitudes to their risk assessments and intervention recommendations: a study in Israel

“…They lend credence to the widespread concern that intervention decisions may be biased by workers’ personal ideology and values and highlight the danger that families that pose similar risk to their children will be treated differently by workers with different attitudes. Such inter‐worker variability was seen by Ruscio (1998) as indicating low interpersonal reliability, which should be addressed by training and supervision to raise workers’ awareness of their underlying attitudes and of the role that those attitudes play in their intervention decisions.…”

“…Yet decision‐making in child welfare is characterised by low reliability, with different workers making different intervention decisions when presented with the same data (e.g. Lindsey 1992, Rose & Meezan 1996, Banach 1998, Ruscio 1998, Rossi et al. 1999, Schuerman et al.…”

“…1999), the question of what leads workers having the same information to arrive at different decisions is of paramount importance. Scholars concerned with this question have posited that social workers’ personal features, such as their biases, personality and temperament, enter into their decisions (Ruscio 1998, Coohey 2003, Gambrill 2005, Shlonsky & Wagner 2005). The present study investigates the possibility that workers’ decisions regarding children at risk for abuse and neglect may vary with their attitudes towards matters central to child protection (e.g.…”

“…The intervention recommendation concerns the best available way of handling the estimated risk. Both are the end points of numerous judgements and choices made under conditions of ambiguity and uncertainty (Ruscio 1998, Gambrill & Shlonsky 2000, Coohey 2003, Shlonsky & Wagner 2005).…”