In 2004, the United Nations adopted the resolution Elimination of Domestic Violence against Women in recognition of the threat that violence in the private sphere posed for women’s equality and full participation in the public sphere (United Nations General Assembly 2004). Empirical evidence has documented that this threat is often a reality. Research has found an association between intimate partner violence (IPV) and increased absenteeism in both education (Vivolo-Kantor et al. 2016) and employment (Swanberg and Logan 2005), employment instability (Adams et al. 2013a, b), decreased workplace productivity (Ajala 2008) and lower levels of educational attainment and lifetime earnings for victimsFootnote 1 (Adams et al. 2013a, b). Despite widespread global admonishment of IPV, IPV remains a pervasive and costly social justice, public health, and legal issue. In Canada, the total costs of IPV have been estimated to be as high as $7.4 billion dollars annually, including costs to employers, counselling and social services for victims and perpetrators, and criminal and civil justice costs (Zhang et al. 2012). As such, the public has not only a vested moral but also an economic interest in reducing the incidence of IPV and supporting the safety of victims in violent relationships.

One tactic that IPV victims commonly use to promote their safety is soliciting the help of others. Research with survivors of IPV has documented an association between help seeking and violence severity; both male (Ansara and Hindin 2010; Machado et al. 2016) and female (Barrett and St. Pierre 2011; Ford-Gilboe et al. 2015) survivors of IPV have been found to be most likely to seek help from others when the violence they experience is more frequent and/or severe in nature. As prior research has identified severity of prior violence as a risk factor for intimate partner homicide (Block 2004; Sabri et al. 2014), soliciting effective help from others can potentially mean the difference between life and death.

While a vast body of literature exists examining socio-demographic and victimization characteristics as correlates of help-seeking in the aftermath of IPV for female victims, markedly fewer studies have addressed these factors in samples with male victims. The studies that have explored these issues with men tend to rely on relatively small samples (e.g. Cook 2009), use convenience sampling methods (e.g. Machado et al. 2016) or samples of men using specialized services/resources for men (e.g. Douglas & Hines 2011, 2015, Hines et al. 2007), or rely on perceptions of hypothetical case scenarios rather than the lived experiences of male survivors (e.g. Arnacky and Valliancourt 2014). As such, there are a dearth of large scale, community-based studies exploring men’s experiences with help-seeking in the aftermath of IPV. Although it is commonly believed that the help seeking of men differs from that of women, failure to include both men and women in the same study using the same methods, sampling procedures, and measures impede meaningful comparisons of gendered experiences of help-seeking.

Sex, Help Seeking, and Intimate Partner Violence

Sex and Help Seeking

Research has documented sex differences in the help seeking behaviors of males and females emerging in early childhood (Benenson and Koulazarian 2008) and continuing into adolescence (Ando et al. 2018) and adulthood (Moller-Leimkuhler 2002), with females requesting help more readily than males. In a systematic review of one hundred and forty four studies on help-seeking for mental health concerns, Clement et al. (2015) found that perceived stigma was a more salient deterrent in seeking help for males than for females. Research has found that males are significantly more likely than females to identify more systemic barriers to help seeking for psychological issues and to report that traditional masculine gender role traits are a deterrent to help seeking (Liddon et al. 2018). Other work has shown that the relationship between masculine gender role conflict and willingness to seek help for mental health problems is mediated in part by self-stigma and self-disclosure tendencies. Men with higher rates of gender role conflict report higher levels of self-stigma and lower levels of self-disclosure, which in turn contribute to higher levels of negativity towards counselling and lowered willingness to utilize counselling services (Pederson and Vogel 2007).

Other research suggests that the role between gender and help seeking is complicated, and men’s willingness to seek help varies based on the type of problem they are experiencing, the characteristics of potential helpers, and the available opportunities for seeking help (Lane and Addis 2005). For example, research on help seeking experiences of patients with a cancer diagnosis found that stage of diagnosis was a more salient predictor of help seeking than patient sex, with no significant sex differences found in the number of sources or type of help sought by male and female patients (Nagler et al. 2010). Research has also highlighted the salience of applying an intersectional framework to understanding men’s help seeking experiences, as men’s help seeking and service utilization experiences have been shown to vary by factors such as race, age, income, and geographic region (Cahn et al. 2017; Nam et al. 2010). Collectively, this research suggests that although sex differences have been found in the area of help seeking, men’s help seeking is highly influenced by context and the type of issue for which they are seeking help.

Help Seeking and Intimate Partner Violence

A constellation of barriers to help-seeking for female survivors of IPV have been well documented. Victims report a reluctance to discuss violence with others due to fear of retaliation, shame, or the possibility of blame or judgement (Cattaneo et al. 2008; Meyer 2016; Morgan et al. 2016), and when victims do discuss violence with others they tend to minimize their victimization experiences (Dunham and Senn 2000). Other research has found that some victims may be ambivalent in their goals for relationship continuity or termination while they are still in a relationship with their abuser and may not seek help until after they have made a decision to terminate the relationship, thus complicating their help seeking efforts (Evans and Feder 2014). Ambiguity surrounding definitions of violence and perpetrators attempts to minimize violence or shift blame to the victim have also been found to deter help-seeking efforts (Towns and Adams 2016). Variations in help-seeking experiences based on type of victimization experiences have also been reported. Women who experienced intimate terrorism were more likely to seek help than those who experienced situational couple violence (Leone et al. 2014). However, women who experienced intimate terrorism and who did not seek help were more likely to not do so because of fear, whereas women who experienced situational couple violence were most likely to not seek help because they did not believe they needed it (Leone et al. 2014).

Prior research and theorizing has suggested that it is not survivors’ lack of desire for help that impedes their abilities to terminate or exit violent relationships (if that is their desired outcome), but, rather, it is the barriers they experience when attempting to access assistance (Fugate et al. 2005; Liang et al. 2005; Moe 2007). These barriers may be more pronounced for victims experiencing oppression or marginalization based on other aspects of their identity that intersect with sex, such as immigrant women (e.g. Moe 2007), women of color (e.g. Hollenshead et al. 2006), women with disabilities (e.g. Macy et al. 2005), and low income women (e.g. Cattaneo et al. 2008), and may result in their under-utilization of formal sources of support.

Although there may be shared barriers that victims of IPV face, research has suggested differences in help-seeking. Using latent class analysis with a large sample of female survivors of violent crime (including IPV) in a national Canadian sample, Kaukinen (2002) found support for three classes of help-seekers: (1) minimal help seekers (who seek support from no or few sources); (2) family/friend help seekers (who seek help primarily from informal sources of support and eschew formal service providers); and (3) substantial help-seekers (who engage in a high volume of help seeking from multiple informal and formal sources of support). Consistent with other research, severity of violence (hurt/injury; fear; and use of weapon) was the strongest predictors of the classes, with substantial help-seekers experiencing the most severe violence.

Compared to what we know about women’s help-seeking, less is known about the frequency and correlates of help seeking among male victims of IPV. In a sample of eighty-nine male victims of female perpetrated IPV, Machado et al. (2016) found that the majority of men did not seek help; however, those who did were most likely to seek help from family and friends. Despite the common belief that men’s help seeking is likely to be met with negativity and disbelief, most men who sought help rated the help “a lot/very helpful” (Machado et al. 2016). Among men who did not seek help, the most commonly reported reason was that they did not notice that they were victims (64.7%), and they felt shame (30.9%), and distrust of the support system (19.1%). Unlike women, fear of not being believed (10.3%) was not as commonly reported as a barrier to seeking support. Similarly, in a survey of 302 men who self-identified as victims of IPV, the majority (84.9%) had sought help from friends, family members, neighbors, or other forms of informal support, and approximately 90% of those who did reported that the resource was “somewhat/very helpful” (Douglas & Hines 2011).

In contrast to men’s relatively positive experiences with informal sources of support, research has shown variations in men’s use of and satisfaction with formal service providers. Male survivors report highest levels of satisfaction when seeking help from mental health providers or counsellors and generally report dissatisfaction with their experiences with the criminal justice and domestic violence sectors (Douglas & Hines 2011; Machado et al. 2016, 2017). Research with domestic violence agencies in the United States found that agency workers attributed men’s lowered use of formal services to lack of awareness of services, shame and embarrassment about being a victim, denial in belief that one was a victim, stigmatization, and fear of reprisal from the perpetrator (Tsui et al. 2010). When men do use services, they are more likely to rate those services favorably when they validated men’s experiences, provided useful resources, and directly addressed how to get help for IPV (Douglas et al. 2012).

Only two studies comparing men’s and women’s help seeking were located (Ansara and Hindin 2010; Choi et al. 2018). In an analysis of case records of 157 male and 823 female survivors of IPV who utilized emergency room services, Choi and colleagues (Choi et al. 2018) found that males were significantly more likely than females to refuse on-site counselling services. The second study, which utilized data from the 2004 General Social Survey in Canada, found that severe violence was significantly related to help-seeking for males and females (Ansara and Hindin 2010). However, compared to female victims of IPV, males were less likely to use formal (32.2% v. 63.8%) and informal sources of support (57.1% v. 85.5%). While these studies provide preliminary evidence of sex differences in help seeking, further research is needed to understand the factors associated with help seeking in the aftermath of IPV. One possible factor that has yet to be explored in the literature is social belonging.

Sex, Social Belonging, and IPV

Numerous studies have documented the important role social support plays in facilitating the resiliency of survivors and mitigating the negative mental health consequences of IPV (Beeble et al. 2009; Coker et al. 2002; Constantino et al. 2005; Howell et al. 2018). Due to the isolation that many survivors experience (El-Bassel et al. 2001), however, survivors often have diminished social networks and limited access to sources of support. Research has found that female survivors of IPV have fewer social support networks (Levondosky et al. 2004) and lower levels of perceived social support (Thompson et al. 2000) than non-abused women. Indeed, some research has found that compared to victims who do not use services, victims who use formal services have lower levels of social support, suggesting that women may need to resort to formal support when their personal support network is lacking (Macy et al. 2005).

Although social support has been extensively examined in the trauma literature, it is a broad construct that encapsulates multiple domains. Such domains include social embeddedness (one’s psychological sense of community or belonging), perceived social support (one’s perception of the availability and adequacy of various forms of instrumental or emotional support in their social milieu), and enacted support (the specific forms of support provided to a person by others). These domains constitute distinct but inter-related phenomenon; however, they are often conflated in the literature under the umbrella term social support (Barrera 1986). Social belonging has similarly been described under the rubric of social support, but it is distinct from other aspects of social support. Akin to Barrera’s conceptualization of social embeddedness, social belonging refers to one’s psychological sense of connectedness with their surroundings (Hagerty et al. 1992), whereas social support is commonly used to describe the perceived availability or enacted behavioral support available in one’s social network (Turner and McLaren 2011). While social support has been the subject of much inquiry in the field of IPV, social belonging has not.

Although we were not able to locate any studies which specifically explored the relationship between IPV, help seeking, and social belonging, it can be hypothesized that individuals who feel less connected to others in their environment may be less likely to seek out support, even if possible sources of support are available to them. Further, research has shown notable sex differences in the availability of possible sources of support for men and women, with men reporting a larger number of individuals in their networks but women reporting a larger number of individuals with whom they feel close to (Furher and Stansfeld 2002). Thus, it is possible that sex differences in help seeking may be associated with differences in their levels of connectedness with others in their environment. As such, it is important to examine the potential role that social belonging plays in the help-seeking experiences of both men and women.

Methods

The purpose of the current study was to explore predictors of help-seeking for survivors of IPV in Canada. The primary research questions of interest for this study were:

  1. (1)

    Are there sex differences in the overall and specific help seeking experiences of male and female survivors of IPV?

Consistent with the literature reviewed on sex differences in help-seeking, we hypothesized that in bi-variate group comparisons between male and female survivors of IPV, male survivors would be less likely than female survivors to seek help from all forms of informal and formal support.

  1. (2)

    Controlling for socio-demographic characteristics and violence severity, do IPV survivors with a strong sense of social belonging differ from IPV survivors who do not have a strong sense of social belonging in seeking help from any source?

Based on the literature reviewed, we hypothesized that victims of IPV who self-reported a strong sense of social belonging would be more likely to seek any form of help (informal or formal) than survivors who do not have a strong sense of social belonging.

  1. (3)

    Controlling for socio-demographic characteristics and violence severity, do IPV survivors with a strong sense of social belonging differ from IPV survivors who do not have a strong sense of social belonging in the specific sources of help that they seek?

Based on the literature reviewed, we hypothesized that victims of IPV who self-reported a strong sense of social belonging would be more likely to seek help from all informal sources (family members, friends or neighbors, co-workers, and religious/spiritual advisors) and formal sources (crisis centre/line, counsellor/psychologist, doctor/nurse, lawyer, police, community centre/family centre).

Data Source

This study involved a secondary analysis of data from the 2009 Canadian General Social Survey (GSS) Cycle 23-Victimization main file. The GSS is a cross-sectional nationally representative survey conducted by Statistics Canada to measure indicators of social well-being among Canadians. The theme of the GSS rotates annually, with a focus every 5 years on victimization, including spousal abuse. A detailed description of the methods employed in Cycle 23 of the GSS has been published (see Statistics Canada 2010) and are briefly outlined below.

Data Collection & Sampling

The target population for the 2009 GSS was Canadian residents 15 years of age or older living in the ten provinces of Canada. Full time residents of institutions and those residing in households without landline telephones were excluded from the sampling frame. Although individuals who solely used cell phones were not included in the GSS, it was estimated that only 8% of Canadians solely possessed cell phones in 2008, the year immediately preceding Cycle 23 of the GSS (Statistics Canada 2010).

The GSS employed a geographically stratified sample. Random digit dialing (RDD) was then used to identify target households within each geographic stratum using simple random sampling procedures. Within each household, one resident aged 15 years or older was then randomly selected to be invited to participate in a telephone interview. Interviews were conducted in either English or French (the language of the respondent’s preference) using computer assisted telephone interviewing (CATI). Beginning with the 1999 GSS victimization survey (cycle 13), interviewers received specialized training by a psychologist because of the sensitive nature of the survey focus. The target sample was 23,500, and a final sample of 19,422 was achieved, with slight under-representation of respondents from low income households (Statistics Canada 2010). Under-representation of low income persons was attributed to lower rates of telephone ownership among the lowest income households (lower than $10,000CAN), with 86% of households in this income bracket owning telephones compared to 96% or higher for all other income groups (Statistics Canada 2011). However, as only 0.9% of the target population for the GSS did not own telephones, survey weights were constructed by Statistics Canada to represent all persons in the target population, inclusive of households with no telephones (Statistics Canada 2011).

Before the interview, respondents were informed by the interviewer of the confidential nature of the survey with the following statement: “This study is being conducted under the authority of the Statistics Act, and your answers will be kept strictly confidential and used only for statistical purposes.” The assurance of confidentiality was repeated again prior to the administration of questions regarding IPV. To further protect respondents’ confidentiality, variables deemed to be of a sensitive nature were suppressed in the public use microdata file for the GSS. We were granted access to the full dataset, including suppressed variables, through the Statistics Canada Research Data Centre (RDC) program. All data analysis was conducted on-site at a secure RDC facility and all output was vetted by Statistics Canada to ensure compliance with all protocols for use with this dataset. Research Ethics Board (REB) approval was obtained by the Principle Investigator (first author) of this study from the university where all authors of this paper were affiliated as faculty members or students.

Measures

Intimate Partner Violence Victimization

GSS respondents who reported a current or former spouse or common-law partner with whom the respondent had been in contact within the previous 5 years were administered a Spousal Abuse module. Because this module was limited to marital and common law relationships, it did not capture other forms of intimate partner violence (such as violence in casual, dating, or non-co-habiting relationships). As same sex unions are recognized in Canada, the Spousal Abuse model was administered to those in both heterosexual and same sex marital or common-law relationships. However, due to RDC minimum cell counts for data release, we were unable to separately assess the experiences of gay, lesbian, and bisexual Canadians in the present analysis due to the low percentage of respondents (5.2%) who identified as gay, lesbian, or bi-sexual.

Items regarding physical abuse perpetrated by current or former partners in the last 5 year were assessed in separate modules (a current partner module and an ex-partner module). Respondents were asked to indicate whether their partners had engaged in any of the following behaviors (yes/no): (1) threatened to hit you with his or her fists or anything that could 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 you, bit you, or hit you with his or 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; and (10) forced you into any unwanted sexual activity by threatening you, holding you down, or hurting you in some way. Responses were binary coded (yes =1, no = 0), with those who reported any one of these items coded as a survivor of physical spousal abuse.

The present analysis included both male and female respondents who reported physical abuse perpetrated by a current or former partner within the last 5 years (n = 900 respondents [385 males and 515 females]). Table 1 provides a description of this sample. Although spousal violence is the term employed by Statistics Canada, we have elected to use the more common nomenclature of intimate partner violence (IPV) in our reporting of results for consistency with the broader body of literature in this field.

Table 1 Description of participants in General Social Survey who experienced Physical/Sexual Intimate Partner Violence Victimization
Full size table

Socio-Demographic Characteristics

Socio-demographic factors assessed in this study include: age [measured as a continuous variable]; sex [female = 1, male = 0]; AboriginalFootnote 2 identity [Aboriginal = 1, not Aboriginal = 0]; visible minority identityFootnote 3 [visible minority = 1, white = 0]; country of origin [born outside of Canada =1, born in Canada = 0]; education [did not graduate high school =1, high school graduate or higher = 0]; physical or mental limitationsFootnote 4 [physical and/or mental limitations = 1, no reported limitations =0]; incomeFootnote 5 [less than $30,000CAN = 1, greater than $30,000CAN =0], rural or urban residence [rural =1, urban =0], and participation in religious activities [participated in religious activities at least once in the last month =1; no participation in religious activities = 0].

Severity of Abuse

Severity of abuse was assessed in the present study with three measures: (1) number of incidents of abuse; (2) physical injury; and (3) fearing for one’s life. Respondents who indicated yes to at least one form of physical abuse were asked a series of follow up questions to assess the chronicity and severity of the abuse: “You said yes to at least one of the previous 10 questions concerning physical and sexual violence. During the past 5 years, has he or she been violent to you on more than one occasion?” If a respondent answered affirmatively, a subsequent question was asked: “How many times did these things happen during the past 5 years?” Responses were coded as zero if the respondent indicated the abuse occurred one time, and one if the respondent indicated more than one incident of violence. Physical injuries were assessed with the question: “You said that during the past 5 years your spouse/partner was violent. During this/any of these incidents were you ever physically injured in any way, for example, bruises, cuts, broken bones, etc.?” with responses dichotomously coded (yes =1, no =0). Finally, respondents were asked if they had ever feared for their lives: “During the past 5 years, did you ever fear that your life was in danger because of your spouse/partner’s violent or threatening behavior?” with responses dichotomously coded (yes =1, no = 0).

Sense of Social Belonging

Social belonging was assessed with the question “How would you describe your sense of social belonging to your local community?”[response options = very strong, somewhat strong, somewhat weak, very weak]. Respondents were coded as having a sense of social belonging if they responded “very strong” or “somewhat strong” to this question.

Outcome Variables

Help Seeking: Informal Sources of Support

Respondents who reported at least one experience of physical abuse were then asked if they had talked to anyone about the incident. Participants were asked to indicate (yes/no) whether they had spoken about the abuse with: (1) family; (2) a friend or neighbor; (3) a co-worker; and (4) a priest, rabbi, imam, elder, or another spiritual advisor. In addition to assessing each source of informal support separately, we created two composite variables: (1) use of any informal source of support; and (2) total (mean) number of informal sources of support used by respondents.

Help Seeking: Use of Formal Services

To assess the use of formal services in response to abuse, respondents who had at least one experience of physical IPV were asked about specific services used within the last 5 years to help with the violence (yes/no). For the present study, we assessed formal sources of help as: (1) crisis center or crisis line; (2) another counsellor or psychologist; (3) doctor or nurse; (4) lawyer; (5) police; or (6) community centre or family centre. In addition to assessing each formal service separately, we created two composite variables: (1) use of any formal service; and (2) total (mean) number of formal services used. Sample descriptive statistics for use of informal and formal supports are outlined in Table 2.

Table 2 Description of help seeking by Physical/Sexual Intimate Partner Violence Victims
Full size table

Analysis

Research question 1 was assessed through a series of bi-variate chi-square analyses to test group sex differences (men/women) in seeking help (yes/no) from each source of formal and informal help. Research questions 2 and 3 were assessed through a series of hierarchical binary logistic regression equations predicting seeking help from any source (research question 2) and help seeking from each specific informal and formal source of help (research question 3). Control variables were conceptually grouped and entered into the equations sequentially in separate blocks (block 1: socio-demographic factors, including gender; block 2 severity of violence), with our primary predictor variable of interest, social belonging, entered in block 3. Hierarchical binary logistic regression was selected as the method of analysis as it enters grouped variables in a sequential manner to allow for testing hypotheses derived from theory (Petrocelli 2003). As research questions 2 and 3 examine theoretically based hypotheses that social belonging will significantly predict help seeking above and beyond the effects of socio-demographic variables and violence severity, hierarchical regression is an appropriate analytic tool for testing these associations. For ease of interpretation (Bograd 2016), marginal effects (ME) are reported. As per Statistics Canada requirements for public release of data output using this dataset, all analyses were conducted using normalized weights and the numbers of participants reported in descriptive statistics have been rounded to a base of 5. All findings reported reflect weighted analyses. Missing data was excluded from all analyses.

Results

Among survivors of IPV, female survivors reported more chronic and severe abuse than did male survivors. While approximately one-third of female survivors of physical IPV feared for their lives (32.74%), only approximately 5 % (5.34%) of male survivors feared their lives were in danger. Over 40% (42.28%) of female IPV survivors were injured as a result of the violence in comparison to 16.89% of male survivors. While over half (50.17%) of female survivors experienced multiple instances of abuse, less than one third (30.80%) of male survivors experienced ongoing violence. There were no significant sex differences between males and females regarding their sense of social belonging.

Research Question 1

As outlined in Table 2, most survivors of IPV engaged in some form of help seeking in response to violence. Over half of survivors (67.40%) sought help from at least one informal source of help, with a higher percentage of female survivors (79.64%) soliciting help from informal sources than male survivors (54.89%). Survivors sought help from an average of 1.35 (SD = 1.21) informal sources of help, with female survivors seeking help from a higher number of informal sources (M = 1.63, SD = 1.16) than male survivors (M = 1.07, SD = 1.07). In regard to service utilization, 40.55% of survivors sought assistance from at least one formal service provider. Again, a higher percentage of female survivors sought the assistance of a helping professional (53.25%) than did male survivors (27.56%). On average, survivors solicited help from 0.97 service providers, with female survivors (M = 1.41, SD = 1.84) using a higher number of formal services than male survivors (M = 0.54, SD = 0.54).

Chi-square analyses (Table 2) revealed significant sex differences in help seeking. Female survivors of IPV were more likely than male survivors to solicit help from each form of informal and formal support assessed in this study, supporting our hypothesis for research question 1.

Research Question 2

Our first hierarchical binomial logistic regression equation assessed the probability of using any source of help (yes/no) in response to IPV. In the final model, four variables emerged as significant predictors of seeking help from any source: age, fear for one’s life, number of incidents of violence, and injury. Every increase of 1 year in age was associated with an increase of 0.5 percentage points that one had used any form of help (p ≤ .001). Fearing for one’s life was associated with a 24.0 percentage point increase in the probability of help seeking (p ≤ .001), while experiencing multiple instances of violence (ME = .164) and experiencing injury (ME = .122) were both also associated with an increase in the probability of soliciting help (p ≤ .001). Social belonging was not associated with seeking help from any source at a statistically significant level. As such, our hypothesis for research question 2 was not supported.

Research Question 3

Table 3 outlines the findings of our hierarchical binomial logistic regression equations predicting utilization of each form of informal and formal support.Footnote 6

Table 3 Hierarchical binomial logistic regression models predicting use of informal and formal support (Marginal Effects Reported)
Full size table

Informal Sources of Help

Family Members

Four variables were associated with the probability of soliciting help from family members in our final model. In Block 1, being female (ME = .085, p ≤ .05) and having an incomes of less than $30,000 (ME = .119, p ≤ .01) elevated the probability of seeking assistance from one’s family. In Block 2, fearing for one’s life (ME = .146, p ≤ .01) and experiencing two or more episodes of violence (ME = .208, p ≤ .01) were similarly associated with a heightened probability of seeking help from family members. Block 3 (sense of social belonging) was not significantly associated with seeking the help of family members.

Friends or Neighbors

In our final model, five variables were significant. In regard to socio-demographic factors, being female increased the probability of soliciting friends or neighbors for help by 11.6 percentage points (p ≤ .01) and being Aboriginal increased the probability of talking to friends or neighbors by 18.2 percentage points (p ≤ .05). In regard to severity of violence, fearing for one’s life increased the probability of soliciting friends or neighbors for help by 12.9 percentage points (p ≤ .05) and experiencing two of more instances of violence increased the probability of speaking to friends or neighbors by 13.6 percentage points (p ≤ .01). Sense of social belonging was associated with an 11.6 percentage point increase in the probability of soliciting help from friends or neighbors (p ≤ .01).

Co-Workers

Having an income of less than $30,000 (CAN) decreased the probability of seeking help from a co-worker by 9.1 percentage points (p ≤ .05). In contrast, fearing for one’s life elevated the probability of talking to co-workers by 12.8 percentage points (p ≤ .05) and experiencing two or more incidents of abuse increased the likelihood by 9.2 percentage points (p ≤ .05). Sense of social belonging was not significantly related to help-seeking in this model.

Formal Sources of Help

Counsellor or Psychologist

Having a mental or physical disability increased the probability soliciting help from a counsellor or psychologist by 7.7 percentage points (p ≤ .05). Severity of violence similarly increased the probability of seeking help from a counsellor or psychologist. Fearing for one’s life was associated with an 8.1 percentage point increase in using the services of a counsellor or psychologist points (p ≤ .05) while experiencing two or more instances of violence was associated with a 16.0 percentage point increase (p ≤ .01). Sense of social belonging was not significantly associated with use of this (or any other) formal service.

Doctor or Nurse

Seeking help from a doctor or nurse was significantly predicted by three socio-demographic variables and all three measures of violence severity. Being female (ME = .075, p ≤ .05), increased years of age (ME = .004, p ≤ .01), and having physical or mental limitations (ME = .080, p ≤ .01) were all associated with an elevated probability of soliciting help from a doctor or nurse. Fearing for one’s life (ME = .106, p ≤ .01), experiencing two or more instances of violence (ME = .083, p ≤ .05), and being injured (ME = .086, p ≤ .01) also increased the probability of talking about the abuse with a doctor or nurse.

Police

In our final model, socio-demographic factors did not significantly predict the probability of reporting violence to the police. In contrast, all three violence severity measures were associated with an elevated probability of soliciting police help. Fearing for one’s life increased the probability of seeking police assistance by 13,7 percentage points (p ≤ .01), experiencing two or more instances of violence increased the probability by 7.0 percentage points (p ≤ .05), and incurring injuries elevated the probability by 6.7 percentage points (p ≤ .05).

Lawyer

Socio-demographic factors were not significantly associated with the probability of contacting a lawyer in our final model; however, all measures of violence severity significantly heightened the probability of soliciting a lawyer’s assistance. Fearing for one’s life elevated the likelihood of contacting a lawyer by 14.7 percentage points (p ≤ .01), experiencing two or more incidents of violence increased the probability by 9.2 percentage points (p ≤ .01), and incurring injuries was associated with an 8.4 percentage point increase (p ≤ .01) in the probability of soliciting help from a lawyer.

In summary, social belonging was significantly associated with seeking help from a friend or neighbor but was not associated at a statistically significant level with any other informal or formal source of support. Thus, our hypothesis for research question 3 was only partially supported.

Discussion

This study assessed the relationship between social belonging and help seeking among IPV survivors in a nationally representative sample of Canadians. To our knowledge, this is the first study to assess the association between social belonging and help seeking in a community based sample comprised of both male and female survivors of IPV. As such, this study provides a novel and important contribution to the help-seeking literature. Our hypothesis that social belonging would have a significant association with help seeking over and above the effects of socio-demographic factors and violence severity was partially supported, in that a strong sense of social belonging was significantly associated with an increased probability of seeking help from friends and neighbors. This suggests that strengthening community bonds among neighborhood residents is a potentially powerful meso-level strategy for building systems of support for survivors of IPV.

A significant body of literature in the field of IPV has identified community capacity building as an important strategy for the prevention of IPV and supporting survivors in the aftermath of violence (e.g. Mancini et al. 2006; Sabol et al. 2004). While providing community members with training to enhance their capacity to offer helpful assistance to survivors of abuse is a critical public endeavor (see Flanigan 2011 for an evaluation of one such effective model in Ontario, the Neighbours, Friends, and Families Program), findings from the present study indicates that survivors will be less likely to seek help from their friends and neighbors if they do not feel a sense of belonging in their social context. Given the ability of those in one’s natural networks to provide free and individualized support to address specific needs not only during an immediate period of crisis but also on an ongoing basis beyond the period of availability of traditional service providers (Budde and Schene 2004), there are multiple advantages to survivors in soliciting help from members of their neighborhood community. Strategies for bringing together community members in every day contexts (not solely in the aftermath of violence) may be important for enhancing survivors’ sense of belonging and increasing the likelihood that they will solicit help in the future if needed. Further, enhancing neighborhood cohesion and connection for all members of a community will benefit the survivors in that community without having to specifically “target” survivors as a population for intervention. As survivors are often an invisible and isolated population, broad based strategies for building community connections that are not predicated explicitly on violence prevention and intervention have the potential to be particularly beneficial for reaching survivors who may not openly identify as victims of violence.

Counter to our hypothesis, social belonging was not significantly associated with soliciting help from any other form of help in the present study in multivariate analysis. This unexpected finding may be due in part to the operationalization of social belonging through a single item indicator measured in the GSS as “social belonging to your local community.” It may be that sense of social belonging is domain specific and that an individual’s sense of social belonging with their friends and neighbors in their community may differ from that of their sense of social belonging in their familial or employment contexts specifically. Further, one’s sense of social belonging may potentially be connected to specific attitudes about the desirability of accessing support from certain formal service providers. For example, research has shown that perceptions of police fairness are associated with social identity, with individuals who feel that they belong to the group that police are intended to represent rating police as fairer than individuals who did not feel that they belonged to the groups that police represent (Bradford 2012). Thus, individuals who feel a strong sense of social belonging to a group that has developed a social norm of distrusting the police may be less likely to seek help from the police than other formal service providers (e.g. doctors or nurses). As such, the use of a single item which assessed social belonging solely in the domain of one’s community is a notable limitation of the present study. Further examination of domain-specific experiences of social belonging and its associations with survivors’ help seeking with sources of support within those specific domains is an important area of future inquiry to further develop our understanding of the relationship between social belonging and help seeking.

While social belonging was only found to be associated with help seeking from friends and neighbors, findings from the present study suggest that chronicity and/or severity of violence are among the most consistent predictors of help seeking. This finding is consistent with previous research using Canadian samples (e.g. Ansara and Hindin 2010; Barrett and St. Pierre 2011; Ford-Gilboe et al. 2015). Fearing for one’s life as a result of violence and experiencing a higher number of instances of violence were significantly associated with not only the probability of seeking any form of help but were also significantly associated with each specific form of help assessed. This continues to underscore that when survivors’ elect to seek help for abuse, the violence in their relationship is most likely to have already escalated in its level of danger. As survivors of violence are most likely to seek help from informal sources (most commonly family, friends, and neighbors), these findings highlight the importance of educating the public not only how to recognize the signs of abuse but also how to best support survivors’ strategies for ensuring their own safety.

The importance of community education to help members of the public provide survivor-centered, empowerment based responses to violence disclosures is essential given that not all reactions on the part of potential sources of help are necessarily viewed as helpful by survivors. While our research shows that survivors are most likely to solicit help from family members, previous research has documented that some family members may pressure survivors to accept the abuse and/or reconcile with abusive partners in the name of preserving the family (e.g. Anderson et al. 2003; Beaulaurier et al. 2007; Raj and Silverman 2007). Other research has shown that when survivors’ solicited help from their friends (the second most frequently utilized source of help identified in the present study) a common reaction they received was an admonishment that they should leave their partners (El-Bassel et al. 2001). For many survivors, this reaction was not perceived as helpful or consistent with their desires. Still other research (Trotter and Allen 2009) has found that some survivors felt blamed for their abuse when they disclosed it to friends and/or family members and that disclosing their abuse resulted in the person they disclosed to emotionally distancing themselves from the survivor. Because the present study involved an analysis of data in an archived data set that did not include measures of the specific reactions that survivors’ received when seeking help, we were not able to assess whether survivors’ help seeking efforts actually resulted in responses that were perceived as helpful by survivors (another notable limitation of our study). With that said, given our finding that both male and female survivors of IPV are more likely to reach out to informal rather than formal sources of assistance, the continued support of evidence-based public education initiatives is essential for equipping community members to serve as supportive and effective responders to disclosures of abuse.

Another objective of the present study was to assess gender differences in the solicitation of help for IPV. Based on our review of the literature on gender differences in help seeking in general and in relation to IPV specifically, we hypothesized that males would be significantly less likely than females to solicit help from all informal and formal sources of support. This hypothesis was fully supported in the present study. With that said, our research found that many male survivors of IPV (54.89%) did solicit help from at least one informal source. This suggests that while male survivors may remain less likely than female survivors to seek help, approximately half of male survivors do solicit help from those in their informal network. Prior research has suggested that men’s help-seeking may be highly context specific and contingent on the specific type of situation for which men need help and the characteristics of potential helpers available to assist them (Lane and Addis 2005). Thus, it may be that although men are significantly less likely to seek help than women in general, they will seek help from individuals in their close networks whom they trust (family or friends) when they are in situations where they are facing physical violence that places them in potential danger (IPV). Further research which identifies the specific contexts and helper characteristics which reduce gender specific barriers to men’s help seeking may help us to better understand the complexities and nuances of men’s experiences soliciting support in the aftermath of IPV.

While some scholars (see Tsui et al. 2010) have argued that the proportionately lower help seeking rates among male survivors may be due to a constellation of social barriers that discourage men from seeking needed help, it may also be the case that male survivors are less likely to solicit help because they experience less severe forms of abuse than do females. In the present study, females were significantly more likely than males to experience chronic and severe violence, and chronicity and severity of violence significantly increased the probability of seeking help from all sources. This is not to dismiss the existence of social barriers grounded in patriarchal notions of masculinity which may serve to stigmatize male survivors (both within male-female as well as within male-male partnerships); however, it does suggest that the reasons for the disproportionately lower rates of help seeking among male survivors of IPV may also be due (at least in part) to the lower levels of severity of violence experienced by men. This also would suggest that the subset of men who do disclose experiences of IPV may be the ones who are experiencing an elevated risk of harm.

Further, survivors of IPV may have different needs and motives for soliciting help. For example, some survivors may be seeking emotional support when disclosing violence while others may be seeking instrumental support, such as material assistance (e.g. help with money, transportation, housing, or child care). Although the methods of the present study did not allow us to ascertain the specific types of assistance sought by survivors when seeking help, it may be the case that there are sex differences in the types of assistance sought by male and female survivors when seeking help. For example, due to the feminization of poverty and women’s assumed responsibility for the care of children, female survivors may be in need of more instrumental forms of assistance than males in the aftermath of violence (which may help account, in part, for their increased rates of help seeking when compared to male survivors). While an emerging body of literature on the help seeking experiences of male survivors exists (see Douglas & Hines 2011 and Parry and O’Neal 2015), a dearth of research exists which compare the help seeking experiences of men and women in the same study. Future research comparing not only the rates and sources of help seeking among survivors but also the relationship between help seeking and types of assistance is critical to identify sex differences in help seeking.

While there are numerous strengthens to the methods employed in the present study (such as the use of a large scale nationally representative sample, the inclusion of male and female survivors of abuse, and the examination of a range of sources of help), there are also limitations which must be considered when assessing this work. First, we were not able to identify the specific types of help desired and whether desired help was received (as well as the survivors’ levels of satisfaction with the help they were offered and/or received). As these sources of information are critical for a nuanced understanding of survivors’ help seeking, the omission of this information in the present study is a notable limitation of our work. Further, while we were able to include both male and female survivors of abuse, we were unable to assess the specific nature of violence experienced by survivors (specifically in regard to the directionality of the violence, e.g. whether respondents used violence in self-defense or retaliation as opposed to being the primary aggressors of violence). Thus, although we categorized any person (male or female) who reported any experience of physical abuse as a survivor of IPV, it is possible that some individuals who experienced abuse may have experienced violence enacted by a partner in self-defense or retaliation (thus bringing into question whether such individuals should be appropriately categorized as survivors).

Another important limitation of this work is the slight under-representation of low income respondents, largely due to low rates of telephone ownership among households in the lowest income bracket (lower than $10,000CAN). Research has documented that those in extreme poverty are at highest risk of IPV, with a current study of violence in England and Wales finding that women in households that earned less than £10,000 annually (approximately $16,700CAN) were more than four times as likely to experience IPV than those in higher income households (Office of National Statistics 2018). Other recent research has found that survivors living in the lowest income census tracks are most likely to experience first time reports to the police for IPV than those in higher income census tracks (Bonomi et al. 2014). Given the heightened risk of victimization among this population and their over-representation in official reports to the police, the under-representation of respondents in situations of extreme poverty in the present study must be considered when making extrapolations from our findings to survivors in the most vulnerable income brackets. Finally, while we sought to employ an intersectional analysis through the inclusion of multiple socio-demographic factors in our analysis, the dichotomous nature of coding for many of such variables (e.g. visible minority, Aboriginal, physical or mental limitations, etc.) meant that we were unable to fully assess within group differences among these diverse populations.