Methodology

Data Sources

General Social Survey on Victimization
Data limitations
Incident-based Uniform Crime Reporting (UCR2) Survey
Homicide Survey

General Social Survey on Victimization

In 2004, Statistics Canada conducted the victimization cycle of the General Social Survey for the fourth time. Previous cycles were conducted in 1988, 1993 and 1999. The objectives of the survey are to provide estimates on the extent to which people experience incidences of eight offence types (physical assault, sexual assault, robbery, theft of personal property, break and enter, motor vehicle theft, theft of household property and vandalism), examine risk factors associated with victimization, examine reporting rates to police, and measure fear of crime and public perceptions of crime and the criminal justice system.

The target population for the GSS includes all non-institutionalized persons 15 years of age and older. The 2004 GSS had a sample size of 24,000 households across the provinces. Individual households were selected using Random Digit Dialing (RDD) yielding a response rate of 75%. The use of telephones for sample selection and data collection means that the 2004 GSS sample in the provinces only covers the 96% of the population that had telephone service.

Top of Page

Data limitations

The data that appear in this profile are based on estimates from a sample of the Canadian population and are therefore subject to sampling error. Sampling error refers to the difference between an estimate derived from the sample and the one that would have been obtained from collecting data from every person in the population.

This profile uses the coefficient of variation (CV) as a measure of the sampling error. Any estimate that has a high CV (over 33.3%) has not been published because the estimate is too unreliable. An estimate that has a CV between 16.6% and 33.3% should be used with caution and the symbol ‘E’ is used.

When comparing estimates for significant differences, we test the hypothesis that the difference between two estimates is zero. We construct a 95% confidence interval around this difference and if this interval contains zero, then we conclude that the difference is not significant. If however, this confidence interval does not contain zero, then we conclude that there is a significant difference between the two estimates.

In addition, non-sampling errors may have also been introduced. Types of non-sampling errors may include the refusal by a respondent to report, a respondents’ inability to remember or report events accurately, or errors in coding and processing of the data. In addition, individuals who could not speak English or French well enough to complete the survey were not included. For these reasons, the victimization data should be used with caution.

Top of Page

Incident-based Uniform Crime Reporting (UCR2) Survey

The Incident-based Uniform Crime Reporting (UCR2) survey collects information on criminal incidents reported to, or detected by police including detailed characteristics of victims, accused persons and the incidents. In 2005, detailed data were collected from 122 police services representing 71% of the population of Canada. Other than Ontario and Quebec, the data are primarily from urban police departments. As such, the reader is cautioned that these data are not geographically representative at the national or provincial levels.

The UCR2 Trend Database contains historical data that permits the analysis of trends in the characteristics of incidents, accused and victims. This database currently includes 62 police services that have reported to the UCR2 survey consistently since 1998. These respondents accounted for 51% of the population of Canada in 2005.

Top of Page

Homicide Survey

The Homicide survey began collecting police-reported data on homicide incidents, victims and accused persons in Canada in 1961. When a homicide becomes known to the police, the investigating police department completes a survey questionnaire which is then forwarded to the Canadian Centre for Justice Statistics (CCJS).

The homicide count for a particular year represents all homicides reported in that year, regardless of when the death actually occurred. The survey was revised in 1991 and again in 1997 to include additional variables such as previous conviction histories and employment status of both the accused and the victim, and whether the victim precipitated the incident with the use or threat of force.