Methodology

Analysis in this document focuses on acts of violence against women that have been quantified using statistical survey techniques. This provides indicators on many aspects of violence against women, services provided to victims, and women’s use of services. The analysis focuses on behaviours that could trigger a criminal justice response and addresses acts of violence against women that qualify as offences under the Criminal Code almost exclusively. The primary statistical data sources used to measure violence against women by Statistics Canada are victimization surveys, and data collected by police agencies, adult courts, emergency shelters for women and their children, and other service agencies providing assistance to crime victims.

Victimization surveys
Police statistics
Adult Criminal Court Survey
UCR2-ACCS linked database
Transition Home Survey
Victim Services Survey

Victimization surveys

In 1993, Health Canada commissioned Statistics Canada to conduct a large-scale survey dedicated to women’s experiences of violence perpetrated against them by male aggressors. The Violence Against Women Survey (VAWS) involved telephone interviews with a random sample of 12,300 women about their adult experiences of sexual and physical assault by men, including male partners, friends, acquaintances or strangers. The survey also included questions related to non-criminal forms of sexual harassment and to women’s fears of violence in public places.

Although the VAWS has not been repeated, some of the same or similar questions concerning spousal violence were included on the victimization cycle of the General Social Survey (GSS). A total of 14,269 women and 11,607 men were interviewed in the 1999 GSS. The GSS victimization survey was designed to be repeated every five years, thereby providing a reliable standard to measure experiences of violence over time. A third time point is provided by the 2004 GSS, in which 13,162 women and 10,604 men were interviewed. In 2004, respondents in the territories were also interviewed.

The GSS interviewed random samples of individuals 15 years of age and older about their experiences with crime in the previous year and their opinions concerning the justice system. Households were selected using random-digit dialling techniques. Once a household was chosen, an individual 15 years of age or older was randomly selected to respond to the survey. Households were excluded from the survey when they had no telephone or when the chosen respondent could not speak English or French. Also excluded were individuals living in institutions.

The methodologies of the 1993 VAWS and the 1999 GSS differ in important ways. As the VAWS was a dedicated survey, it focused exclusively on matters relating to violence against women and employed only female interviewers. The 1999 and 2004 GSS, on the other hand, are general victimization surveys with a special module of questions related to spousal violence that is based on the VAWS. The GSS employs both male and female interviewers, although respondents are offered the opportunity to switch to an interviewer of the other sex if they are uncomfortable responding to sensitive questions during the interview. As a result of these methodological differences, comparisons between the two surveys must be made with caution.

Although improvements have been made in the methodology for interviewing women about violence, victimization surveys still have limitations. The fact that surveys are conducted only in Canada’s two official languages presents a significant barrier for the full inclusion of Aboriginal and immigrant women in Canadian statistics. According to the 2001 Census, 2.6 million women in Canada were not fluent in either English or French. In many northern communities in Canada, especially those in the territories, older Aboriginal women have maintained their culture and language and many would not be able to participate in a telephone survey. Because interviewers did not use Aboriginal and Inuit languages, many Northern women were excluded.

The methodology of victimization surveys also does not allow for the inclusion of households without telephones or households with cellular telephones only. While telephone surveys are significantly better at generating responses than a standard mail or e-mail survey, this method automatically excludes households without telephones. These represent a small percentage of all households—about 4%—but excluding those without telephone access may under-represent certain groups in the population, such as lower income groups, northern people living in traditional communities, rural people, or women living in shelters or on the street as a direct result of violent victimization. Excluding those with only cell phones may underestimate young, single adults and transient populations.

In addition, although the sample sizes used in these surveys are relatively large, analysis of smaller groups is not always possible. Without disaggregated information the indicators cannot show in detail the unique experiences of Aboriginal women, immigrant and refugee women, visible minority women, women with disabilities, teenaged and younger girls, older women, women living in low income situations, women in rural and remote communities, and lesbian and bisexual women. (See Table Number of women and men interviewed for the 1993 Violence Against Women Survey (VAWS) and the 1999 and 2004 General Social Survey (GSS) on Victimization)

How violence is measured using victimization surveys

Violence by a spouse or common-law partner is measured in the General Social Survey on Victimization GSS and the Violence Against Women Survey VAWS by a module of 10 questions. This approach consists of asking respondents about specific actions instead of simply asking about “violence” or “assaults”, in order to minimize differing interpretations of what constitutes violent behaviour. The module of questions with introductory statement follows:

It is important to hear from people themselves if we are to understand the serious problem of violence in the home. I’m going to ask you 10 short questions and I’d like you to tell me whether, in the past five years, your spouse/partner has done any of the following to you. Your responses are important whether or not you have had any of these experiences. Remember that all information provided is strictly confidential.

During the past five years, has your partner:

1.  threatened you with his/her fist or anything else that could have hurt you?
2.  thrown anything at you that could have hurt you?
3.  pushed, grabbed or shoved you in a way that could have hurt you?
4.  slapped you?
5.  kicked, bitten, or hit you with his/her fist?
6.  hit you with something that could have hurt you?
7.  beaten you?
8.  choked you?
9.  used or threatened to use a gun or knife on you?
10. forced you into any unwanted sexual activity by threatening you, holding you down, or hurting you in some way?

In the VAWS, these 10 questions were asked in order. In the GSS, the first two questions were given in sequence to all respondents and the remaining eight questions were asked in random order.

How criminal harassment is measured

The series of questions designed to measure the prevalence of criminal harassment or stalking conforms to the Criminal Code definition and includes:

In the past five years, have you been the subject of repeated and unwanted attention that caused you to fear for your safety or the safety of someone known to you? By that I mean has anyone:

1. phoned you repeatedly or made silent or obscene phone calls?
2. followed you or spied on you?
3. waited outside your home?
4. waited outside your place of work or school or other places you were, when they had no business being there?
5. sent you unwanted e-mail messages?
6. sent you unwanted gifts, letters or cards?
7. persistently asked you for a date and refused to take no for an answer?
8. tried to communicate with you against your will in any other way?

If the respondent stated that they had experienced at least one of these acts they were then asked, “Did you fear for your safety or the safety of someone known to you?” If the respondent stated “yes”, they then were considered to be a stalking victim.

In addition to these questions, respondents were asked two additional questions that did not require them to state that they felt fear because threats were explicit in the questions. Respondents who responded yes to either of these questions were also considered to be victims of stalking.

9. In the past five years, has anyone attempted to intimidate or threaten you by threatening or intimidating someone else?
10. In the past five years, has anyone attempted to intimidate or threaten you by hurting your pet(s) or damaging your property?

Top of Page

Police statistics

Uniform Crime Reporting Survey

The Canadian Centre for Justice Statistics, a division of Statistics Canada, developed the Uniform Crime Reporting (UCR) Survey with the co-operation and assistance of the Canadian Association of Chiefs of Police. The aggregate UCR Survey, which became operational in 1962, collects crime and traffic statistics reported by all police agencies in Canada. UCR survey data reflect reported crimes that have been substantiated (or founded) through police investigation.

More detailed incident-based crime statistics are collected through the Incident-based Uniform Crime Reporting Survey (UCR2 Survey). This survey allows detailed examination of accused and victim characteristics, as well as characteristics of the incident itself. Collection began in 1988. By 2004, 166 police agencies in nine provinces, representing 53% of the national volume of reported crime, were responding to the UCR2 Survey. The sample of police forces is not nationally representative. The largest percentage of cases originates in Ontario and Quebec.

The UCR2 Trend Database contains historical data, which permits the analysis of trends in the characteristics of incidents, accused and victims, such as weapon use and victim/accused relationship. This database currently includes 68 police services that have reported to the UCR2 Survey consistently since 1998. These accounted for 37% of the national volume of crime in 2004.

Homicide Survey

The Homicide Survey has provided detailed police-reported data on homicide incidents since 1974. When police become aware of a homicide, a survey questionnaire is completed. The count for a particular year represents all homicides reported in that year, regardless of when the death actually occurred. The survey remained unchanged until 1991, at which time more detailed information was collected. A question regarding the history of domestic violence between the accused and victim, and more detailed victim-offender relationship categories were added to the survey in 1991.

Top of Page

Adult Criminal Court Survey

The Adult Criminal Court Survey (ACCS) provides a database of statistical information on the processing of cases through the adult criminal court system. The survey consists of a census of Criminal Code and other federal statute charges dealt with in adult criminal courts. The ACCS represents approximately 90% of the national adult criminal court caseload. Manitoba is not included in the survey for any year. Data from Nunavut were included as part of the Northwest Territories prior to April 1, 1999; however, Nunavut has not reported to the ACCS since the creation of the territory. Data from the Northwest Territories are not available for 1996/97, 2000/01, 2001/02, 2002/03, or 2003/04. New Brunswick and British Columbia began reporting to the ACCS in 2001/02. Also, information from Quebec’s municipal courts (which account for approximately one-quarter of Criminal Code charges in that province) is not yet collected. With the exception of Prince Edward Island, Nova Scotia, New Brunswick, Alberta, British Columbia and Yukon, no data are provided from superior courts.

The absence of data from all but six superior court jurisdictions may result in a slight underestimation of the severity of sentences imposed across Canada. The reason for this is that some of the most serious cases, which are likely to result in the most severe sanctions, are processed in superior courts.

Top of Page

UCR2-ACCS linked database

A pilot study linked data from the UCR2 Survey for the years 1997 through 2001 to cases with convictions for violent crimes from the ACCS for the years 1997/98 through 2001/02. The UCR2 variables that were linked to the court files include: relationship of victim to accused; sex of victim; age of victim; level of injury; and presence and type of weapon. The objective of this study was to examine court outcomes of spousal violence cases. As these variables are in the UCR2 but not the ACCS, analysis was made possible through this linkage process.

Coverage for the UCR2 and ACCS databases includes 18 urban areas in four provinces for the study period 1997/98 to 2001/02 (Newfoundland and Labrador, Ontario, Saskatchewan and Alberta). The 18 urban areas that are included in the pilot project are:

Because the study focuses on selected urban areas, it is not a representative sample, but rather a location-specific analysis of sentencing patterns. Data from these urban areas are rolled up to produce a total for all areas.

Top of Page

Transition Home Survey

The Transition Home Survey was developed under the federal government’s Family Violence Initiative in consultation with provincial and territorial governments and transition home associations. The objectives of the survey are to collect information on residential services for abused women and their children during the previous 12 months of operation, and to provide a one-day snapshot of the clients being served on a specific day. In 1991/92, Statistics Canada began to collect basic information on transition home services and clientele. The survey was repeated with some changes in 1992/93, 1994/95, 1997/98, 1999/00, 2001/02 and 2003/04.

The Transition Home Survey is a mail-out/mail-back census survey of all residential facilities providing services to abused women and their children. In 2003/04, of the 543 residential facilities providing services to abused women and their children, 473 returned their questionnaires for a response rate of 87%. Separate questionnaires were completed for facilities that had two or more residences under the same name or address.

Top of Page

Victim Services Survey

The Victim Services Survey was funded by Justice Canada’s Policy Centre for Victim Issues. The objectives are to provide a profile of victim service agencies, information on the types of services offered, and some insight into the clients who use them through a snapshot of clients serviced on October 22, 2003.

Victim services are defined as agencies that provide direct services to primary or secondary victims of crime and are funded in whole or in part by a ministry responsible for justice matters. The survey covers system-based, police-based, court-based and community-based agencies, sexual assault centres, criminal injuries compensation programs and other financial benefit programs. Corrections-based victim services are not included.