Risk-Need-Responsivity (RNR) classification models: Still evolving

Abstract

The RNR (Risk-Need-Responsivity) Framework identifies principles to facilitate assigning individuals to appropriate programs by triaging program placement based on static and dynamic risk factors that influence offending behavior (dynamic risk factors) through programming. This well-accepted framework has resulted in justice agencies adopting risk and need assessment (RNA) tools, but the tools themselves present barriers to a classification system. RNA tools suffer from methodological weaknesses that affect the valuation of the information gleaned from the tool. Six classification schemes are based on the RNR principles, and they vary on consistency with the small body of empirical studies conducted to identify statistically-based risk-need profiles. Furthermore, the models are weakly coupled with the criteria for good classification systems. The question that emerges is whether there is a theoretically sound classification system or if the field would be better off creating levels of care scaffolding based on the risk-need profile clusters. Addressing the methodological issues associated with the RNA tools will remove stumbling blocks to a meaningful classification system.