What makes an Expert in Risk Assessment in Child Protection?

Objectives In child protection research it is still not known, what factors contribute to professional expertise in risk assessment. Moreover, it is controversial how to empirically identify good or correct risk judgements.... [ view full abstract ]

Objectives
In child protection research it is still not known, what factors contribute to professional expertise in risk assessment. Moreover, it is controversial how to empirically identify good or correct risk judgements. This preliminary study conducted with 1.200 social work students explores the relationship between personality, knowledge and skills and the quality of judgements in risk assessment in child protection.

Method
Based on fictional vignettes participants estimated the risk of child maltreatment and neglect. Vignettes were constructed in factorial design. Participants are free how to do the assessment (by intuition, by application of a risk assessment tool, etc.). Individual quality of judgements in risk assessment was calculated with a coherence measure developed by James Shanteau and David Weiss (CWS-Expertise-Index). Participants also completed a self-report questionnaire on personality, judgements styles and conducted a knowledge and risk literacy test.

Results
Findings show that there is a significant positive relationship between the coherence of risk judgements and formal education, competence in probabilistic reasoning, conscientiousness in using prediction strategies, not using some formal risk assessment tool and risk assessment format (frequency vs. logical concept of probability). Apart from impulsivity there was no relationship to other personality constructs like rational vs. intuitive judgement or analytic vs. prototype judgement style.

Conclusion
Results show that professional judgement can be trained, since it depends less on thinking styles rooted in personality but rather on general education and competence in probabilistic reasoning as well as on adequate representation of risk in the prediction task. There are limitations of the study. First, the sample so far only consists of students. To assess the influence of social work experience and of different prediction strategies professionals use in daily practice the study has to be conducted with professionals of different experience levels. Second, judgements are based on fictional vignettes, not on real cases. Therefore, only judgement coherence but not judgement accuracy could be measured. The authors plan a study addressing these issues.