Construct validity of dynamic risk factors: a multi-framework analysis

Dynamic risk factors (DRFs) occupy a central position in modern forensic practice, serving simultaneously as predictors of recidivism, explanations of offending and targets for intervention. Each of these functions makes fundamentally different evidential demands, and the field has not always maintained clear distinctions between them. This paper evaluates the construct validity of DRFs for sexual offending through three complementary theoretical frameworks: Cronbach and Meehl′s nomological network framework, Messick′s unified validity framework and Haig, B. D. ′s theory validation framework. Each framework reveals a distinct layer of the validity problem: the nomological architecture surrounding DRFs is theoretically underspecified; the validity argument is substantially stronger for prediction than for the clinical uses to which DRFs are also applied and DRFs largely operate as classificatory or dispositional theories rather than the mechanistic accounts their applied use presupposes. A research agenda is proposed to address these limitations.PRACTICE IMPACT STATEMENT Dynamic risk factors for sexual offending are routinely used to guide treatment targeting and clinical formulation, yet the construct validity evidence suggests that their causal status as intervention targets has not been adequately established. Practitioners and policy makers should exercise caution in treating DRF scores as direct indicators of the causes of offending, and should instead recognise them as theoretically incomplete but practically useful markers of recidivism risk.