A Practical Measure of Offence Seriousness: Sentence Severity , continued
The first step was to obtain a set of offenses on which to measure the dependent and independent variables. The list of 204 NSCS crime descriptions was examined and those which would not result in a prison sentence in this jurisdiction were excluded. This was necessary because the data available on sentencing were concerned solely with prison sentences and not other terms of penalty such as fines, probation, etc. The next step was to go through the list of NSCS crime descriptions again and fit each with the most appropriate offence category label from the Criminal Code of Canada (CCC). The resulting set of offenses and their crime descriptions were then submitted (a) to a psychologist expert in inmate classifications for an exercise of judgement on the validity of the offence/crime description matches and (b) to a group of six forensic psychologists for the identification of offenses for which the relationship between sentence severity and offence seriousness might be partially obscured due to varying perceptions of offence seriousness in the U.S.A. and Canada. There was strong agreement that drug offenses are likely jurisdictional variants (interjudge reliability coefficient of .80). In addition, on the basis of the differences between the two countries in gun control laws, weapons offenses were designated as jurisdictional variants. The next step was to generate the sentence severity and NSCS offence seriousness data associated with the list of offenses. The sentence severity measure posed little difficulty. Sentence severity (mean sentence length) was calculated for all offenses for which 24 or more counts of the offence were available and at least one count of the offence occurred during each of the three years of the observation period. For the NSCS offence seriousness data, however, the task was not quite so straight-forward since many of the offence categories were associated with a multiplicity of crime descriptions. The solution to the problem of choosing the value for NSCS offence seriousness to attach to each offence category was to conduct two separate studies of the relationship between sentence severity and offence seriousness.
The purpose of both studies was to determine the validity of sentence severity as a measure of offence seriousness. In each study, validity is demonstrated by a high correlation between sentence severity and the criterion measure of offence serious-ness obtained from the NSCS seriousness scores.