Modeling the Scaling Up of Early Crime Prevention
Published online on May 22, 2017
Abstract
Research Summary
Although the amount of research evidence on the effectiveness of developmental crime prevention has grown considerably in recent decades, the translation of this scientific knowledge into policy and practice has lagged behind. In this article, we consider the challenges as well as the opportunities associated with scaling up evidence‐based programs and we offer an approach for considering the potential effects of deviations in implementation protocols during replications. We use results from the series of studies on the Nurse‐Family Partnership (NFP) to develop a computer simulation model. Based on a large number of simulations, we systematically adjusted key inputs (e.g., target population and fidelity) to mimic a range of possible implementation conditions and to observe the impacts on the estimated intervention effects. As the process progresses from the baseline condition, which reflects the initial implementation conditions specified in the NFP model, to alternative experimental scenarios reflecting problematic deviations in implementation, the number of arrests accumulated by treatment participants begins to increase. This indicates that these implementation challenges have a negative impact on program effects and that we can go some way toward predicting what might occur in implementation as they emerge and interact.
Policy Implications
The challenges facing program dissemination and their impact on program results have not been enumerated all that precisely in the literature, which has resulted in the assignment of arbitrary penalties when evaluating the prospects for taking programs to scale. Establishing efficacious interventions is only one part of the process of moving research to practice, and further consideration of the scaling‐up process is integral to translational criminology. Informed computer simulation may be one tool to help guide this program implementation process.