Automating Risk Assessment Instruments and Reliability

• Published date: 01 February 2017
• Date: 01 February 2017
• DOI: http://doi.org/10.1111/1745-9133.12272
• Author: Thomas H. Cohen

POLICY ESSAY

RECIDIVISM RISK ASSESSMENT

Automating Risk Assessment Instruments

and Reliability

Examining an Important but Neglected Area in Risk

Assessment Research

Thomas H. Cohen

Administrative Office of the U.S. Courts

The utilization of risk assessment instruments is increasingly common in various

criminal justice contexts as practitioners use these instruments to predict re-

cidivism, identify addressable criminogenic needs, and direct resources to those

offenders at greatest risk of reoffending (Andrews and Bonta, 2010; Andrews, Bonta, and

Hoge, 1990; Cullen and Gendreau, 2000). The predictive efficacy of these instruments is

of paramount importance because of the sheer numbers of persons currently placed under

some form of correctional supervision and their proclivity to reoffend. The Bureau of Justice

Statistics (BJS), for example, has reported that in 2014, there were 6,851,000 adults under

correctional supervision in the United States (Kaeble, Glaze, Tsoutis, and Minton, 2016);

moreover, within 5 years of being released from prison, 77% of an estimated 400,000 state

prisoners eventually reoffended (Durose, Cooper, and Snyder, 2014).

The advent of actuarial risk assessment instruments has given rise to a rich literature

demonstrating the superiority of these tools at predicting recidivism over traditional ap-

proaches involving professional or clinical judgment (Andrews and Bonta, 2010; Andrews,

Bonta, and Wormith, 2006; Bonta, Law, and Hanson, 1998). As discussed by Grant Duwe

and Michael Rocque (2017, this issue), however,the aim of much of this literature has been

focused on issues of validity or on the predictive power of these instruments (Gendreau,

Goggin, and Smith, 2002; Gendreau, Little, and Goggin, 1996; Smith,Cullen, and Latessa,

2009). There have been fewer efforts to examine inter-rater reliability or the extent to which

The views and opinions expressed in this essay are solely the author’s. They do not represent the views of the

Administrative Office of the U.S. Courts. Direct correspondence to Thomas H. Cohen, Administrative Office of

the U.S. Courts, One Columbus Circle NE, Washington, DC 20544. (e-mail: Thomas_cohen@ao.uscourts.gov).

DOI:10.1111/1745-9133.12272 C2017 American Society of Criminology 271

Criminology & Public Policy rVolume 16 rIssue 1