Translating research into practice: Offender Risk Assessment

PDF (12KB)

Research summary
Vol. 7 No. 1
January 2002


What is involved when correctional systems implement offender risk scales that are developed by researchers?


During the past twenty years, there has been considerable research in the development of objective offender risk assessment instruments. These instruments measure the risk of an offender committing a new crime based on social and personal-demographic information. The ability to reliably differentiate higher risk offenders from lower risk offenders is tremendously important for public safety and effective programming.

The research-based offender risk instruments show significantly better predictive accuracy compared to the more traditional use of professional or clinical judgements in making decisions about offender risk. Throughout the world, more and more correctional systems are introducing these risk scales into their daily decisions on the management of offenders. How well these risk instruments work and what kinds of difficulties are encountered when the risk scales are introduced in "real world" settings is an issue with important implications.


A series of reviews of the literature on the implementation of risk instruments were undertaken. The reviews focused on implementation issues encountered when offender risk instruments were introduced at the national (Canada), provincial and state levels (U.S.). In addition, data relevant to the integrity of implementation was analyzed from one large correctional jurisdiction.


Research-based risk assessments can be implemented very well in the everyday setting provided the correctional organization is attentive to certain potential difficulties. The implementation literature reveals three main themes that appear to hold true across governmental levels (i.e., national, provincial, state). First, and most obviously, the implementation of offender risk scales requires a large investment in training. Even the simplest offender risk scale requires staff to be trained and usually, computerized offender information systems must be modified. It is simply not enough to give staff the scale and a scoring manual and hope that the instrument will be administered uniformly with few errors. With more sophisticated risk scales a number of days of intensive training is required.

Second, introducing a new risk scale will be met with staff resistance unless they are properly prepared. New assessment procedures mean abandoning old ways of doing things and for some this may be difficult. Staff need to be convinced of the merits of the new approach and fully supported in their learning of the new procedures.

Third, monitoring and retraining is very important. A training program may successfully train staff to a high skill level but the skills often deteriorate with time. In one study, hundreds of videotapes of offender risk assessments conducted by correctional officers found an average error rate of 13%. After booster training sessions, the percentage of errors decreased to one percent.

Policy implications

  1. Risk scales developed under highly controlled research conditions can be successfully implemented into real world, correctional settings. The research is also unequivocal in that research-based risk scales improve offender predictions of risk beyond traditional methods.
  2. There is now a body of literature surrounding the conditions favourable to the implementation of offender risk scales. This literature can guide correctional organizations by identifying what needs to be done and what must be avoided to ensure successful implementation of new risk scales.
  3. Three important factors to consider when implementing offender risk scales are:
    1. devoting sufficient financial resources to training,
    2. staff preparation for a new procedure, and
    3. monitoring implementation and retraining staff when necessary.


For further information

James Bonta, Ph.D.
Public Safety Canada
340 Laurier Avenue West
Ottawa, Ontario
K1A 0P8
Tel 613-991-2831
Fax 613-990-8295

Date modified: