Section 2: Risk factors for violence against women

By Hope Hutchins

• [Highlights] [Full article in PDF]

• Police-reported violence against women
• Self-reported victimization of women
• Summary
• Detailed data tables
• References
• Notes

While violence against women can cross all socio-demographic boundaries, data suggest that some groups of women and girls are more at risk. Previous research has suggested that particular socio-demographic and community factors are associated with a higher risk of self-reported violent victimization (Perreault and Brennan 2010). While these risk factors cannot be considered direct causes of victimization, they can individually or together contribute to an increased likelihood of victimization. Identifying risk factors, therefore, can help inform the maintenance and development of preventive strategies and responses to violence against women (Johnson 2006).

This section examines risk factors for violence against women aged 15 years and over using two types of data: (i) police-reported data from both the Incident-based Uniform Crime Reporting (UCR) Survey and the Homicide Survey and (ii) self-reported victimization data from the General Social Survey (GSS) on victimization.^{Note 1} Results are primarily based on descriptive analysis, which examines the prevalence of violent victimization for each individual "socio-demographic", "lifestyle" or "community-level" factor.^{Note 2} For the analysis of self-reported victimization, this discussion also indicates whether factors remain significant predictors of violent victimization after controlling for other known risk factors.^{Note 3}

Due to differences in the type and scope of information collected by police-reported and self-reported surveys, the analysis of risk factors varies by data source and should not be compared. Risk factors for police-reported violent crime are identified for spousal, dating, and non-intimate partner violence, while risk factors for self-reported victimization are described for spousal and non-spousal violent victimization.

Police-reported violence against women

Females aged 15 to 24 most at risk of police-reported violence

Consistent with patterns for violence overall, being young was a risk factor for all forms of police-reported violence against women, both within and outside the context of intimate partner relationships. In 2011, females aged 15 to 24 generally experienced the highest rates of violence, with rates subsequently decreasing with increasing age (Chart 2.1). The one exception was dating violence against women, where women aged 25 to 34 were most at risk (Chart 2.2). Even so, females aged 15 to 24 continued to be most vulnerable to certain types of dating violence, namely sexual violations.

Chart 2.1
Victims of police-reported violent crime, by sex and age group of victim, Canada, 2011

Description for chart 2.1

Note: Includes the population aged 15 years and over. Population estimates based upon July 1st estimates from Statistics Canada, Demography Division. Excludes incidents where the age and/or sex of the victim was unknown.      
Source: Statistics Canada, Canadian Centre for Justice Statistics, Incident-based Uniform Crime Reporting Survey.   

Chart 2.2
Victims of police-reported dating violence, by sex and age group of victim, Canada, 2011

Description for chart 2.2

Note: Includes the population aged 15 years and over. Population estimates based upon July 1st estimates from Statistics Canada, Demography Division. Excludes incidents where the age, sex, and/or the relationship to the accused of the victim was unknown.         
Source: Statistics Canada, Canadian Centre for Justice Statistics, Incident-based Uniform Crime Reporting Survey.   

Similar age-based risk was present when violence against women escalated to homicide. Based on data over the last decade, females aged 15 to 24 were most at risk. This was particularly evident for spousal homicide, where there were 18.8 spousal homicides per million females aged 15 to 24 (Table 2.1). This rate drops down to 10.3 homicides per million women aged 25 to 34, and 7.5 homicides per million women aged 35 to 44. Regardless of age, spousal homicide rates were consistently higher for women than men.

As with dating violence in general, the risk of dating homicide peaks at a later age. The rate of dating homicides for women was highest among those aged 35 to 44. Women's rates were consistently higher than the rates for men until the age of 55, after which men's rates were more elevated that those for women.

The rate of non-intimate partner homicide against women, while lower than the rate against men for every age group, was highest among females aged 15 to 24 (Table 2.1). The rate generally decreases with age, though rates increase for women aged 55 and older.

Women most at risk from an intimate partner

As discussed in section 1, women are most at risk of violence from an intimate partner^{Note 4} than any other type of perpetrator, though not all women experience the same level of risk from an intimate partner. In 2011, the rate of police-reported dating violence against women was 1.6 times higher than the spousal violence rate (631 versus 395 per 100,000 women).^{Note 5}

Also, there were notable differences in the proportion of female victims who were in an ongoing or previous relationship with the accused intimate partner. According to police-reported data, about one-third (34%) of female intimate partner victims of violence were currently married (legally or common-law) to the perpetrator, followed by those currently dating (31%), previously dating (20%) and previously married (14%) (Chart 2.3).

Chart 2.3
Victims of police-reported intimate partner violence, by sex and relationship of accused to victim, Canada, 2011

Description for chart 2.3

1. 'Other intimate' partners are defined in the Incident-based Uniform Crime Reporting Survey as "a person with whom the victim had a sexual relationship or a mutual sexual attraction but to which none of the other relationship options apply".     
Note: Includes those aged 15 years and over. Excludes incidents where the age, sex, and/or the relationship to the accused of the victim was unknown.
Source: Statistics Canada, Canadian Centre for Justice Statistics, Incident-based Uniform Crime Reporting Survey.   

While it is not possible to calculate rates for current or previous partners based on the Incident-based Survey,^{Note 6} data from the Homicide Survey suggest that once differences in the size of the married population are considered, the risk of spousal homicide was highest among legally separated women. In particular, over the previous five years, women were six times more likely to be killed by an ex-spouse than by a current legally married spouse (18.0 homicides per million versus 3.1 per million).^{Note 7}

Aboriginal women disproportionally represented as homicide victims

Aboriginal people, including women, are often overrepresented as victims of violent crime, including lethal forms of violence (Perreault 2011, Brennan 2011). While there are limitations in examining Aboriginal identity of victims using the Incident-based UCR Survey,^{Note 8} analysis is possible using the Homicide Survey. An important caveat is the high level of homicide incidents where the Aboriginal identity was unknown.^{Note 9}

Despite Aboriginal women representing about 4% of the population in 2006,^{Note 10} between 2001 and 2011, Aboriginal women accounted for at least 11% of dating homicide victims and at least 10% of non-intimate partner homicide victims^{Note 11} (Table 2.2). The proportion of Aboriginal women killed by a spouse (4%) was similar to their representation in the total population.

Rates of police-reported violence against women higher in non-CMAs than within CMAs

For both intimate partner violence and non-intimate partner violence overall, police-reported rates of violence against women were higher in non-census metropolitan areas (CMAs),^{Note 12} including small cities, towns and rural areas, than within CMAs (Chart 2.4).^{Note 13} This was also true for nearly all types of intimate partner and non-intimate partner violence, with the one exception of robbery.

Chart 2.4
Victims of police-reported intimate partner and non-intimate partner violence, by victim's place of residence, Canada, 2011

Description for chart 2.4

1. A census metropolitan area (CMA) consists of one or more neighbouring municipalities situated around a major urban core. A CMA must have a total population of 100,000 of which 50,000 or more live in the urban core. To be included in the CMA, other adjacent municipalities must have a high degree of integration with the central urban area, as measured by commuting flows derived from census data. A CMA typically comprises more than one police service. CMA populations have been adjusted to follow policing boundaries.       
Note: Includes the population aged 15 years and over. Population estimates based upon July 1st estimates from Statistics Canada, Demography Division. Excludes incidents where the age, sex, and/or the relationship to the accused of the victim was unknown.
Source: Statistics Canada, Canadian Centre for Justice Statistics, Incident-based Uniform Crime Reporting Survey.   

Self-reported victimization of women

Young women most at risk of self-reported violence

There are many socio-demographic factors associated with the self-reported violent victimization of women, many of which are characteristics that an individual cannot change, such as age. As with police-reported findings, being young was a significant risk factor for women's self-reported spousal and non-spousal violent victimization.

According to the 2009 GSS, women aged 15 to 34^{Note 14} with a current or former spouse^{Note 15} were about two to three times as likely as their older counterparts to report experiencing spousal victimization in the last 12 months (Table 2.3). Similar findings were evident for non-spousal violent victimization, as the rate for women aged 15 to 34 was over three times higher than the rate for those aged 35 and over (171 versus 50 incidents per 1,000 population) (Table 2.4).

This higher prevalence of non-spousal violence among young women was seen for all three types of violent victimization, namely physical assault, sexual assault and robbery. It was, however, most pronounced for sexual assault, where women under the age of 35 years had victimization rates that were five times higher than those of their older counterparts (73 versus 14^E incidents per 1,000 population).

Being young remained a strong predictor of violent victimization against women even after controlling for other factors associated with victimization. In other words, young age had a unique effect on women's risk of self-reported victimization, whereby young women, regardless of their other personal characteristics, their lifestyle behaviours or community, were particularly vulnerable to violent crime by either a spouse or another perpetrator.

Single women most at risk of violent victimization

The risk of self-reported violent victimization can vary by marital status, such as whether individuals are married or in a common-law relationship. For women, however, the prevalence of victimization by a current spouse^{Note 16} in the last 12 months did not differ between those in a legally married or common-law union. About 1.2% of legally married women were victims of spousal violence, similar to the proportion (1.7%^E) of women in common-law relationships. This was unlike the pattern for men, where the risk of spousal violence was about five times higher for men in common-law unions, compared to those who were legally married (3.8%^E versus 0.8%^E).

The rate of non-spousal victimization for single women was approximately four times higher than the rate for women in legally married or common-law unions (199 versus 50 incidents per 1,000 population) and three times higher than for women who were separated, divorced, or widowed (65 incidents per 1,000 population). This heightened risk among single women persisted even after accounting for other factors associated with victimization.

Prevalence of spousal violence elevated among lesbian and bisexual women

The 2009 GSS indicates that women who self-identified as lesbian or bisexual were significantly more likely than heterosexual women to report violence by a current or previous spouse in the previous five years (20.8%^E versus 6.1%)^{Note 17} (Table 2.5). It should be noted that the sex of the abusive spouse was not asked; therefore, the prevalence rates for lesbian or bisexual women could include some opposite-sex spouses.

Although estimates were not releasable for women who identified as lesbian or bisexual for non-spousal violence, overall non-spousal violence rates in the past 12 months of those who self-identified as gay or lesbian or bisexual were nearly five times the rate for those who self-identified as heterosexual (394^E versus 81 incidents per 1,000 population) (Table 2.4).

Women who participated in many evening activities at higher risk of spousal and non-spousal violence

Women who participated in a greater number of evening activities^{Note 18} per month had a higher prevalence of both spousal and non-spousal victimization. At particular risk were women who engaged in 30 or more evening activities per month. These women had more than double the risk of spousal violence compared to women who participated in 20 to 29 evening activities per month (3.1%^E versus 1.5%^E) (Table 2.3).

Similar risks were found for victimization outside a spousal relationship. The rate of victimization was highest among women who engaged in 30 or more evening activities a month. This rate was nearly double that of women who engaged in 20 to 29 activities, nearly three times higher than those who engaged in 10 to 19 activities, and more than five times higher than those who engaged in the fewest number of evening activities (less than 10 per month) (Table 2.4).

Even when other risk factors were taken into account, participating in a greater number of evening activities per month remained a significant predictor of women's risk of non-spousal victimization. However, engaging in 30 or more evening activities did not independently increase women's risk of spousal violence, when all other analyzed factors were considered.

Educational attainment and income not linked to women's risk

Overall, educational attainment had no bearing on women's risk of either spousal or non-spousal violence. This was also the case for income, a factor often influenced by levels of education. That is, income was not related to women's risk of either spousal or non-spousal violence.

Retired women experience lowest risk of non-spousal violent victimization

Time spent in particular domains, such as the household, paid employment, school, and leisure, may protect or expose individuals to potential perpetrators (de Léséleuc 2007). There is some evidence to suggest that this was the case for retired women's risk of non-spousal violence. According to the 2009 GSS, retired women were less likely than those working a paid job to be victimized by someone other than a spouse, when other factors were taken into account.

While the prevalence of spousal and non-spousal violence was lower among those whose main activity was household work, this could be attributed to other risk factors (Table 2.3, Table 2.4). Similarly, while students had more than double the rates of non-spousal violence compared to women in paid employment (204 versus 98 per 1,000 population), this elevated rate among students could be explained by the finding that students tend to be young, single and participate in more evening activities. Being a student on its own did not increase women's risk of violence.

Non-spousal victimization higher among women who used alcohol or drugs

The 2009 GSS gathered information on how often individuals, regardless of whether or not they were a victim, drank alcoholic beverages or used drugs in the previous month. The prevalence of spousal violence was similar between women who drank heavily, defined as consuming five or more alcoholic drinks in one sitting, and those who did not (2.8%^E and 1.7%). It is important to note that the GSS does not indicate whether this alcohol or drug use began before or after the violent victimization.^{Note 19}

In contrast, women who drank heavily were more often a victim of non-spousal violence than those who did not drink heavily (185 incidents per 1,000 population versus 68 incidents per 1,000 population). However, drinking heavily was not an independent contributor to risk, but rather it was associated with other analyzed risk factors, such as being young, single and participating in evening activities.

The prevalence of spousal and non-spousal violence was consistently higher among women who reported using drugs.^{Note 20} In 2009, women who used drugs were more than three times as likely to self-report being a victim of spousal violence in the previous 12 months (5.4%^E versus 1.6%) (Table 2.3). Similarly, the prevalence rate of non-spousal victimization was over four times higher among women who used drugs compared to other women (285 versus 68 incidents per 1,000 population) (Table 2.4).

The higher prevalence of spousal violence among women who used drugs may be explained by the presence of other risk factors for violent victimization, as drug use on its own did not increase risk of spousal violence.^{Note 21} For non-spousal victimization, however, the risk of victimization among women who used drugs remained significantly higher, even after controlling for other risk factors.

Women with an activity limitation more likely to experience spousal violence

Activity limitation,^{Note 22} such as a physical or mental condition or health problem that restricts a person's activities, has been found to be associated with a significantly greater risk of violent victimization (Perreault 2009). Results from the 2009 GSS confirm that the prevalence rates of spousal and non-spousal violence were often higher among women with an activity limitation.

In particular, women with activity limitations experienced rates of spousal violence in the past five years that were nearly double those without limitations (9.3% versus 5.0%) (Table 2.5). This differs from men with some type of activity limitation, who experienced no increased risk of spousal violence.

The impact of activity limitations on women's risk of non-spousal violent victimization varied by type of violence. While women with activity limitations were not at greater risk of non-spousal violence overall,^{Note 23} they were over 2.5 times more likely to be a victim of robbery compared to women without any activity limitations (18^E versus 7^E incidents per 1,000 population).

When controlling for other socio-demographic factors, as well as lifestyle and community characteristics, women with an activity limitation continued to have an increased risk of spousal and non-spousal victimization. The heightened risk of violence among women with activity limitations may be related to their greater vulnerability and dependence on others (Brownridge 2006).

Spousal violence less prevalent among immigrant women than Canadian-born women

In Canada, there can be considerable overlap in the population of immigrant and visible minority women. Over three-quarters (76%) of recent immigrants in Canada were visible minorities in 2006, and more than two-thirds (68%) of visible minorities were immigrants (Chui and Maheux 2011, Chui 2011). The socio-demographic characteristics of these groups, while vastly heterogeneous, can differ from other groups of women, notably in the languages spoken (Chui 2011). As a result, it is important to note that since the GSS is conducted in English and French, some immigrants and visible minorities may be unable to participate due to language barriers^{Note 24} and may be under-represented among spousal and non-spousal violence victims (Sinha 2012, Johnson 2006).

According to the 2009 GSS, immigrant women had a lower risk of spousal violence compared to Canadian-born women. Specifically, 4.9% of immigrant women self-reported being a victim of spousal violence in the previous five years, compared to 6.8% of non-immigrant women (Table 2.5). This lower risk of spousal violence did not persist when other factors were taken into account. Although findings were not releasable by gender for non-spousal violence, immigrants were generally much less likely than non-immigrants to be victimized outside a spousal relationship in the previous 12 months (48^E versus 110 incidents per 1,000 population) (Table 2.4).

Among the female visible minority population,^{Note 25} the rate of spousal victimization was equal to the rate for non-visible minorities. This is in contrast to visible minority men who were less likely to state that they had been a victim of spousal violence compared to non-visible minority men (3.2%^E versus 6.4%). Due to small counts, it was not possible to examine visible minority women's risk of non-spousal violence, though overall, visible minorities had lower rates of non-spousal victimization than non-visible minorities (63^E versus 103 incidents per 1,000 population).

Prevalence of victimization higher among Aboriginal women

As previously mentioned, Aboriginal persons are more often the victim of violent victimization compared to non-Aboriginal persons (Brennan 2011, Brzozowski et al. 2006). The higher risk for violence against Aboriginal women may be partially explained by factors associated with violent victimization, such as age, marital status and participation in evening activities (Perreault 2011). Indeed, Aboriginal women's 2.5 times higher rate of spousal violence (Table 2.5) can be explained by the presence of these other analyzed risk factors, as identifying as an Aboriginal woman was not independently related to spousal violence risk.

While Aboriginal women also experienced much higher rates of non-spousal violence compared to non-Aboriginal women (Chart 2.5), other factors could not fully explain this heightened risk.

Chart 2.5
Self-reported non-spousal violence victimization rates in the past 12 months, by sex of victim and victim's Aboriginal identity, Canada, 2009

Description for chart 2.5

^E use with caution
† reference category
* significantly different from reference category (p < 0.05)
Note: Rates are calculated per 1,000 population aged 15 years and older. Responses of "Don't know" and "Not stated" are not listed, therefore, the sum of percentages may not add up to 100%. Data from the Northwest Territories, Yukon and Nunavut were collected using a different methodology and are therefore excluded.          
Source: Statistics Canada, General Social Survey, 2009.  

Some possible reasons for the higher prevalence of violent victimization among the Aboriginal population may be attributed to factors specific to Aboriginal persons. The Standing Committee on the Status of Women (2011) identified the following possible larger socio-economic factors as factors related to violence:

• economic and social inequalities,
• discrimination against Aboriginal peoples,
• possible loss of understanding of their history and culture, and
• residential schools and the intergenerational cycle of violence.^{Note 26}

Spousal violence more common among women who experienced emotional and financial abuse

Some risk factors are specific to spousal violence, such as a spouse's psychologically abusive behaviour. These behaviours, which can be precursors to the commission of physical or sexual assault, can include limiting contact with others, put-downs and name-calling, jealousy, harming or threatening to harm someone close, demanding to know whereabouts, damaging property, and preventing access to income. In 2009, 18% of women reported experiencing emotional or financial abuse by a current or previous partner at some point during their relationship. This was not statistically different to the proportion for men (17%).

Victimization data indicate that women and men who had experienced emotional or financial abuse by a spouse were much more likely to be physically and/or sexually abused by their current or previous spouse in the previous five years. In 2009, 19% of women who experienced emotional or financial abuse by a current spouse reported being a victim of physical or sexual assault by this same spouse. This compares to 2% of women who did not experience emotional or financial abuse. The heightened risk was also evident when the violence involved previous spouses (32% versus 4%^E).

Some forms of psychologically abusive behaviour were more often related to violent acts than others. The 2009 GSS found that the risk of physical or sexual abuse was highest for women who reported that their spouse damaged their personal property (Chart 2.6). Overall, emotional and financial abuse remained correlated to spousal violence, even when other risk factors were taken into consideration.

Chart 2.6
Percentage of women that have ever been emotionally or financially abused, who were physically or sexually assaulted by their spouses in the last 5 years, by type of psychological abuse and relationship status, Canada, 2009

Description for chart 2.6

F too unreliable to be published
† reference category
* significantly different from reference category (p < 0.05)
Note: Emotional or financial abuse is measured over the respondent's lifetime. 'Spouse' refers to legally married and common-law spouses, including same-sex spouses. Responses of 'Don't know' and 'Not stated' are not listed, therefore, the sum of percentages may not add up to 100%. Data from the Northwest Territories, Yukon and Nunavut were collected using a different methodology and are therefore excluded.       
Source: Statistics Canada, General Social Survey, 2009.  

Community factors had little impact on women's risk of victimization

The 2009 GSS captured information related to respondents' communities, including the individual's place of residence, their levels of social ties and interactions, and their perceptions of social disorder in their neighbourhood, such as noisy neighbours, vandalism, people using or dealing drugs, and prostitution.^{Note 27} Some research has suggested the absence of social ties can reduce an overall sense of community well-being, resulting in negative consequences, such as increased levels of crime (Gannon et al. 2005). Furthermore, signs of physical and social disorder may indicate that a community lacks surveillance and is uncared for by its residents (Keown 2008).

According to victimization data, the effect of community factors^{Note 28} on women's risk of victimization was minimal. There was no impact on victimization for women living in a census metropolitan area compared to those residing in small cities, towns and rural areas. Similarly, the social ties within a community rarely contributed to women's risk. For instance, women who reported knowing few or no people in their neighbourhood had a similar risk of victimization, both spousal and non-spousal, as women with many contacts (Table 2.6).

While women living in a neighbourhood with any indicators of social disorder did not have an elevated risk of spousal violence compared to other women,^{Note 29} they were nearly three times as likely to be a victim of non-spousal violent victimization (169 incidents per 1,000 population versus 60 per 1,000). In addition, social disorder was independently associated with this higher risk of non-spousal victimization.

Summary

The likelihood of being violently victimized is not equal among all groups of women, but rather differs based on socio-demographic, lifestyle behaviours, and community characteristics, as well as other life events (experiences of emotional and financial spousal abuse). Using both police-reported and self-reported victimization data, this section identified several factors that raise a woman's risk of violent victimization.

Police-reported findings indicate that young women are the most-at-risk group of women, a finding consistent with earlier research. The only exception to this pattern was women in dating relationships, where risk was highest among those in their late 20s and early 30s. Rates of police-reported violence were also elevated among those living in non-census metropolitan areas, while homicide data indicate that Aboriginal women were disproportionally victims of homicide.

Some of these factors, namely age and Aboriginal identity, were also related to risk of self-reported violent victimization. In addition, for spousal victimization overall, risk was heightened for women who had an activity limitation and were emotionally and/or financially abused by a spouse. These two factors, along with age, were independently associated with risk of spousal violence.

A number of factors elevated women's risk of non-spousal violence, including young age, being single, participating in many evening activities, using drugs, being an Aboriginal person, and living in a community characterized by social disorder.

Detailed data tables

Table 2.1 Victims of homicide, by age group and sex of victim and accused-victim relationship, Canada, 2001 to 2011

Table 2.2 Victims of homicide, by sex of the victim, accused relationship to victim, and Aboriginal identity, Canada, 2001 to 2011

Table 2.3 Self-reported spousal violence victimization in the past 12 months, by sex of victim and victim socio-demographic characteristics, Canada, 2009

Table 2.4 Self-reported non-spousal violence victimization in the past 12 months, by sex of victim, victim socio-demographic and lifestyle characteristics, Canada, 2009

Table 2.5 Self-reported spousal violence victimization in the past 5 years, by sex of victim and selected victim socio-demographic characteristics, Canada, 2009

Table 2.6 Self-reported non-spousal violence victimization in the past 12 months, by sex of victim and community factors, Canada, 2009

References

Brennan, S. 2011. "Violent victimization of Aboriginal women in the Canadian provinces, 2009." Juristat. Statistics Canada Catalogue no. 85-002. (accessed September 6, 2012).

Brownridge, D. A. 2006. "Partner violence against women with disabilities: Prevalence, risk and explanations." Violence Against Women. Vol. 12, no. 9. p. 805-822.

Brzozowski, J., A. Taylor-Butts and S. Johnson. 2006. "Victimization and offending among the Aboriginal population in Canada." Juristat. Vol. 30, no. 3. Statistics Canada Catalogue no.85-002. (accessed September 6, 2012).

Chui, T. 2011. "Immigrant women." Women in Canada: A Gender-based Statistical Report, Sixth Edition. Statistics Canada Catalogue no. 89-503. (accessed August 29, 2012).

Chui, T. and H. Maheux 2011. "Visible Minority Women." Women in Canada: A Gender-based Statistical Report, Sixth Edition. Statistics Canada Catalogue no. 89-503. (accessed August 29, 2012).

de Léséleuc, S. 2007. "Criminal victimization in the workplace." Canadian Centre for Justice Statistics Profile Series. No. 13.Statistics Canada Catalogue no. 85F003M. (accessed September 12, 2012).

Gannon, M., K. Mihorean, K. Beattie, A. Taylor-Butts and R. Kong. 2005. "Criminal Justice Indicators, 2005." Statistics Canada Catalogue no. 85-227. (accessed September 12, 2012).

Johnson, H. 2006. Measuring Violence Against Women: Statistical Trends 2006. Statistics Canada Catalogue no. 85-570.

Keown, L. 2008. "Life in metropolitan areas: a profile of perceptions of incivility in the metropolitan landscape." Canadian Social Trends. Vol.86. Statistics Canada Catalogue no. 11-008. (accessed August 6, 2012).

Perreault, S. 2009. "Criminal victimization and health: A profile of victimization among persons with activity limitations or other health problems." Canadian Centre for Justice Statistics Profile Series. No. 21. Statistics Canada Catalogue no. 85F003M. (accessed September 12, 2012).

Perreault, S. 2011. "Violent victimization of Aboriginal people in the Canadian provinces, 2009." Juristat. Statistics Canada Catalogue no. 85-002. (accessed September 12, 2012).

Perreault, S and S. Brennan. 2010. "Criminal victimization in Canada, 2009." Juristat. Vol. 30, no. 3. Statistics Canada Catalogue no. 85-002. (accessed September 16, 2012).

Sinha, M. 2012. "Victimization among visible minority and immigrant populations." National Clearinghouse on Family Violence Bulletin. Public Health Agency of Canada. Ottawa. (accessed August 6, 2012).

Standing Committee on the Status of Women. 2011. Ending Violence Against Aboriginal Women and Girls: Empowerment – A New Beginning. House of Commons Canada. Ottawa. (accessed September 6, 2012).

Statistics Canada, 2006. Census of Population, Statistics Canada Catalogue no. 97-552-XCB20060007.

Notes

^E use with caution

1. For information about methodological differences between the GSS and the UCR surveys, see discussion in the Introduction section.
2. For both self-reported spousal and non-spousal violence, the analysis of risk factors is primarily based on a one-year snapshot, since many of these factors can change over time. The only exceptions are for the following socio-demographic characteristics and life events of spousal violence victims: sexual orientation, presence of an activity limitation, Aboriginal identity, visible minority status, immigrant status, and experiences of emotional and/or financial abuse. Sample counts were too low for these characteristics to be reliably presented for the previous 12 months.
3. Risk factors for victimization seldom exist in isolation. For example, being young is often associated with participating in evening activities, both of which can be risk factors for violent victimization. To evaluate the independent effect of each factor to the risk of victimization, logistic regression modeling techniques were used. By doing so, it was possible to identify factors that independently predict women's risk of self-reported violent victimization, even after controlling for the potential effects of other factors.
4. Intimate partners include legally married, separated, divorced, opposite and same sex common-law, dating partners (current and previous) and other intimate partners.
5. The higher risk of dating violence is despite the fact that the rate of dating violence is underestimated. Underestimation is a result of the inflated size of population used in the calculation of dating violence rates. That is, the calculation of a dating violence rate uses the entire population of unmarried persons, regardless of their dating relationship status.
6. This is because the UCR Survey relationship categories do not match population data.
7. Excludes homicides committed by same-sex spouses, ex-same-sex spouses or separated common-law spouses since there are no population data specifically for these groups.
8. Police services do not consistently report on the Aboriginal identity of victims or accused.
9. Aboriginal identity was known for 55% of spousal homicides, 55% of dating homicides, and 54% of non-intimate partner homicides against women.
10. Based on the 2006 Census, Statistics Canada.
11. The calculation of percentages includes those instances where the Aboriginal identity of the victim was unknown, since it may be more likely that unknown cases of Aboriginal identity may be biased.
12. A CMA (census metropolitan area) consists of one or more neighbouring municipalities situated around a major urban core. A CMA must have a total population of at least 100,000 of which 50,000 or more live in the urban core. To be included in the CMA, other adjacent municipalities must have a high degree of integration with the central urban area, as measured by commuting flows derived from census data. A CMA typically comprises more than one police service.
13. It is not possible to examine census metropolitan area (CMA) rates of spousal and dating violence, since population data for spouses and unmarried persons are not available at the CMA level.
14. More detailed age breakdowns were not possible, due to small samples for detailed age categories and the corresponding high likelihood of sampling error.
15. For spousal violence, prevalence is calculated as the percentage of individuals with a current and/or former spouse who were victimized, while for non-spousal violence, it is calculated as the rate of incidents per 1,000 total population.
16. It is not possible to calculate prevalence of violence by a previous spouse by marital status.
17. In order to release an analysis of the association between sexual orientation and violent victimization by gender with 2009 GSS data, the categories of self-identifying as gay or lesbian and self-identifying as bisexual were combined.
18. Evening activities can include working, attending classes, going to meetings/restaurants/movies/bars/pubs/casinos, participating in sports or recreational activities, shopping, and visiting with friends or family.
19. The GSS asked victims of spousal violence if they had used alcohol or drugs to cope with violence. Due to small counts, this information was not releasable.
20. In the GSS, "drugs" can include illicit drugs and the abuse of solvents and other hazardous substances, but excludes medication taken on a doctor's prescription or bought over-the-counter.
21. Based on results from multivariate analysis.
22. The GSS defines persons with activity limitations as those who reported difficulty in their daily lives or a physical or mental condition or health problem that limited the quantity or type of activities in which they could engage. This is based on the World Health Organization's (WHO) framework definition of disability (See Perreault 2009).
23. This differs from findings from multivariate analysis.
24. According to the 2006 Census, about 2% of women in Canada could not speak English or French. Although the General Social Survey does collect information on the cultural origins of respondents, the sample size is not large enough to calculate reliable estimates by race or cultural background. However, those who identified themselves as a visible minority or an immigrant were not found to be associated with increased levels of spousal violence. Similar findings have been found for victimization in general. It should be noted that despite improvements in the methodology used for interviewing women about violence, surveys are only conducted in Canada's two official languages which may present a barrier to the collection of data from immigrant women.
25. As per Census definitions, visible minorities refer to those who self-identify as belonging to one or more of the following racial or cultural groups: Chinese, South Asian, Black, Filipino, Latin American, Southeast Asian, Arab, West Asian, Korean, Japanese, or another group. Non-visible minorities include single origin White, single origin Aboriginal, and multiple origin White/Latin American and White/Arab-West Asian.
26. For more information on the interconnected issues affecting Aboriginal peoples, see the Royal Commission Report on Aboriginal Peoples (1996). (accessed November 26, 2012).
27. The 2009 GSS measured several indicators of disorder, including, noisy neighbours or loud parties, people hanging around on the streets, people sleeping in public places, garbage or litter lying around, vandalism, graffiti, and other deliberate property damage, people being attacked or harassed because of their skin colour, ethnic origin, or religion, people using or dealing drugs, people being drunk or rowdy in public places, and prostitution.
28. Community factors were only examined for spousal violence by a current spouse in the last 12 months given that women may move neighbourhoods after marital separation. For multivariate analysis, however, the model examining the effects of community factors included victimization by both a current and previous spouse to allow for larger sample size.
29. May contain some cases where respondents did not state whether social disorder was a problem in their neighbourhood.
30. For the territories, given small counts, the analysis of risk factors for spousal violence refers to victimization in the past five years. However, as with the provinces, data for non-spousal violence are only available for the past 12 months. Many findings for spousal and non-spousal victimization in the territories should be used with caution due to small counts. The results are based on descriptive analysis.
31. Unlike other differences, this was only significant at the (p < 0.10) level.
32. As with the provinces, community factors were only examined for spousal violence by a current spouse given that women may move neighbourhoods after marital separation.
33. May contain some cases where respondents did not state whether social disorder was a problem in their neighbourhood.