The Level of Service Inventory–Revised Profile of English Prisoners

• Author: Clive R. Hollin,Emma J. Palmer
• DOI: 10.1177/0093854806286195
• Published date: 01 June 2006
• Date: 01 June 2006
• Subject Matter: Articles

THE LEVEL OF SERVICE

INVENTORY–REVISED PROFILE

OF ENGLISH PRISONERS

Risk and Reconviction Analysis

CLIVE R. HOLLIN

EMMA J. PALMER

University of Leicester

This reconviction study follows a previous needs analysis, using the Level of Supervision

Inventory–Revised (LSI-R), with English prisoners. LSI-R scores at prison discharge and

reconviction data are available for a sample of the original population. The reconvicted offend-

ers have a higher LSI-R total score, are in higher risk bands, and have higher levels of need on

several subscales. With statistical control of key variables, the difference in total score

remains, but differences in subscale scores largely disappear. Multivariate analysis reveals a

robust effect of LSI-R total score, but not subscale and risk band scores, in contributing to

models of both reconviction and time to reconviction. The study provides support for risk

assessment with the LSI-R in the English prison service.

Keywords: Level of Service Inventory; prediction; prisoners; reconviction

347

CRIMINAL JUSTICE AND BEHAVIOR, Vol. 33 No. 3, June 2006 347-366

DOI: 10.1177/0093854806286195

© 2006 American Association for Correctional and Forensic Psychology

AUTHORS’ NOTE: This study was supported by a grant from the Research

Planning Group, HM Prison Service. Any views or opinions expressed here are

those of the authors and not HM Prison Service. We wish to thank Danny Clark,

Research Planning Group (now Probation Unit, Home Office), for his help in data

collection at this stage of the project. We are also grateful to four anonymous

reviewers for their constructive and helpful comments on an earlier version of this

article. Correspondence concerning this article should be addressed to Clive

Hollin, Clinical Division of Psychiatry, Department of Health Sciences, University

of Leicester, Leicester General Hospital, Gwendolen Road, Leicester LE5 4PW,

UK; e-mail: crh9@leicester.ac.uk.

CJB286195.qxd 4/5/2006 4:54 PM Page 347

The Level of Supervision Inventory was described by Andrews

(1982) as a quantitative measure to inform assessment of risk

and need with offenders. Subsequently retitled the Level of Service

Inventory, a revised version (LSI-R) was published by Andrews and

Bonta (1995). Completed through file review and interview, the LSI-R

provides a structured assessment of 10 domains relevant to risk and need

assessment with offender populations. These criminogenic domains

are represented by the subscales of criminal history, education/

employment, finance, family/marital, accommodation, leisure/recreation,

companions, alcohol/drug problems, emotional/personal problems,

and attitudes/orientation. An offender’s scores on the individual sub-

scales identify areas of need for potential service delivery, and the

total score can be translated into a risk band to determine risk of

reoffending (Andrews & Bonta, 1995). A recent overview of the

extant evidence concluded that psychometrically the LSI-R is both a

valid and reliable instrument (Hollin, Palmer, & Clark, 2003).

The LSI-R has been used to inform risk and need assessment, in

both custodial and community settings, with adult women offenders

(e.g., Coulson, Ilacqua, Nutbrown, Giulekas, & Cudjoe, 1996), adult

male offenders (Loza & Simourd, 1994), and male and female young

offenders (Catchpole & Gretton, 2003). The LSI-R has also been

used to assess risk and need with different types of offenders, includ-

ing sex offenders (Simourd & Malcolm, 1998) and violent offenders

(Hollin & Palmer, 2003), and as an outcome measure in an evaluation

of an intervention with offenders (Bonta, Wallace-Capretta, &

Rooney, 2000). Indeed, the increasingly widespread use of the LSI-R

and similar instruments on a national basis in England and Wales

has prompted professional debate on the large-scale implementation

of standardized assessment within the criminal justice system

(Robinson, 2003).

Notwithstanding recent exchanges regarding the utility of differ-

ent measures employed in risk assessment (Gendreau, Goggin, &

Smith, 2002, 2003; Hemphill & Hare, 2004), it is broadly accepted

that the LSI-R functions effectively as a predictor of risk of offend-

ing (Kroner & Mills, 2001). The majority of studies of the utility of

the LSI-R in predicting reoffending have used measures of associa-

tion, either point-biserial correlations or phi coefficients, between

LSI-R total scores (i.e., risk) and reoffending. As Hemphill and Hare

348 CRIMINAL JUSTICE AND BEHAVIOR

CJB286195.qxd 4/5/2006 4:54 PM Page 348