Women’s Perceptions of Their Community’s Social Norms Towards Assisting Women Who Have Experienced Intimate Partner Violence

Abstract

The role of social norms has played an often unrecognized role in the perception of and action to assist low-income urban women who are in violent relationships. Two forms of social norms will be assessed, including descriptive norms—what people typically do to assist women in a violent relationship—and injunctive norms—defined as what people should do to assist women. This study will present our initial findings into the development of measures to assess women’s perception of their community’s social norms toward assisting women who have experienced intimate partner violence (IPV) and how these norms are related to women’s perception of the community, reasons for community assistance toward women experiencing IPV, and women’s own experience of IPV. Systematic measurement development processes were applied to reliably and validly develop the social norms measures. A three-phase approach was used to develop eight paired items measuring descriptive and injunctive norms. A total of 176 low-income urban women were interviewed and the scale responses were compared to length of time at the residence, perceptions of their neighborhood, perceived reasons for community involvement and non-involvement in assisting women experiencing IPV, and IPV experienced as an adult. The two developed social norms scales were found to have high internal consistency alpha coefficients of 0.84 for descriptive norms and 0.93 for injunctive norms. Paired t tests were statistically significant, denoting higher injunctive than descriptive social norms. Lowered descriptive norms were found among younger women, women who reported that they did not think their neighborhood was a good place to live, women who had ever experienced intimate partner violence as an adult, and perceived lower reasons for neighbor involvement and higher reasons for neighbor non-involvement toward assisting women experiencing IPV. Higher levels of injunctive social norms were statistically associated with living in a good place and increased perceived reasons for neighbor involvement toward assisting women experiencing IPV. Significant differences between descriptive and injunctive norms suggest that women, especially those who are currently experiencing IPV, would prefer greater support from community neighbors than they are currently providing. The descriptive and injunctive social norms scales demonstrated a high level of internal reliability and significantly associated with other influencing factors thought to be associated with social norms. Overall, the performance of the injunctive and descriptive norms scales support their use as a tool to investigate social norms toward neighbors taking action to assist women experiencing IPV.

Introduction

Although definitions of intimate partner violence (IPV) differ among individuals, cultures, countries, government agencies, and international health and human rights organizations,^1–3 the public health significance of IPV is undeniable. Recent results from the Behavioral Risk Factor Surveillance System, a USA telephone survey of a nationally representative sample, showed that nearly one quarter (23.6%) of women and 11.5% of men had been physically or sexually assaulted in their lifetime.⁴ An estimated 1.5 million women in the USA are raped or physically assaulted by an intimate partner each year, and studies have found that annual rates of IPV range from 5% to 25% and adult lifetime rates between 21% and 55%.^4–7

The real and perceived availability of resources to assist women who have experienced IPV is of paramount importance in efforts to decrease the deleterious physical and affective effects of IPV. Our work and that of others has shown that women who are abused are most likely to turn to those who are nearest to them, in geographic and psychological proximity.^8–11 Therefore, while social norms toward IPV in general have been shown to dramatically affect the rates and tolerance of IPV in communities,^12–20 investigations into the perceived availability of resources and response to IPV are in need of assessment. Low-income women living in an urban area represent a noteworthy sample to assess because many may lack financial resources to leave and relocate to a different environmental context, but have the availability of neighbors in relative close proximity to potentially depend upon for assistance.

As public health takes on a broader role in the field of IPV, there is a great need to begin the process by which we incorporate a social norms approach toward the prevention and “treatment” of IPV. There have been increasing demonstrations on the relative effect of social norms toward IPV and the role general social norms toward IPV can play in the prevention of IPV. However, there needs to be attention paid to the measurement of the perceived social norms regarding helping women involved in a violent relationships. The incidence and prevalence of IPV are significant and those women directly affected are in need of community responses that work with them to assist and effectively attenuate the effects of IPV.

Although the importance of support for women experiencing IPV is clearly indicated in the literature, women are oftentimes reluctant to seek out support, and a number of studies have documented the relative dearth of support available for women experiencing IPV.^10,21–23 Sources of assistance for women seeking help for IPV have been categorized as formal sources (including law enforcement, shelters, health care providers, and counselors) and informal sources (including family, friends, and neighbors). Surveys have found that the most common forms of formal support for women experiencing IPV were health professionals and law enforcement.⁹ The utilization of formal sources of support was found to greatly lag behind the seeking of support from informal sources.^6,8,9 The vast majority of women who seek help with IPV report looking to friends, neighbors, and family—65% of women from O’Campo et al.⁸ and 75% of women from Coker et al.,⁶ with significantly less than half of women seeking help from formal sources of support. Identifying the challenges and barriers to seeking assistance from neighbors would help in the determination of effective strategies for abused women. Women’s perceptions of how and when their community would assist women involved in an abusive relationship will ultimately affect victims’ help-seeking behaviors, because abused women seeking help are most likely to turn to support systems within their social networks and communities.^8,11,15 These norms can have significant influence on how support for one who has experienced IPV is perceived by the victim, the perpetrator, and the community.^16–18,24 A woman’s decision to remain in a violent relationship or to flee is complex, controversial, and centers on the woman’s perception of the availability of resources.²⁵ The dynamics of abusive relationships are often misunderstood, and can lead to victim blaming when women choose to stay with their abusers. Low-income women lack sufficient monetary resources to move to a safer community; furthermore, there is the fear that leaving their partner will subject them to further aggression, as violence often increases during the weeks after a woman flees an abusive relationship. The perceived lack of acceptance by her neighbors, or willingness to offer assistance, may unintentionally shame a woman into remaining with her partner despite the violence, and for the woman to perceive that an offer of a safe haven is not available as an option. Consequently, as found by Koepsell et al.,²⁶ seeking out but not receiving support was negatively associated with leave taking, thereby making many women feel they have no other option but to remain with their abusers.

This study will present our initial findings into the development of measures to assess women’s perception of their community’s social norms toward assisting women who have experienced IPV and how these norms are related to women’s perception of the community, reasons for community assistance toward women experiencing IPV, and their own experience of IPV. It addresses the perceived influence of social norms around the issue of assisting low-income urban women in a violent relationship. Two forms of perceived norms will be assessed, including descriptive norms—what people typically do to assist women in a violent relationship—and injunctive norms—defined as what would be acceptable to others, that is, what people should do to assist women in violent relationships.^27,28 In addition to these two types of perceived norms, we examined those factors that may influence perceived norms, including perceptions of community neighbors reactions toward assisting women in violent relationships and women’s own experience of IPV.

Methods

In this study, we developed and tested three new measurement tools looking at descriptive norms, injunctive norms, and the perception of one’s community’s interaction with domestic violence. The study presented here is part of a larger Centers for Disease Control–funded effort (R49/CCR318515) entitled Community Pathways to Interpersonal Violence. Community Pathways to Interpersonal Violence was configured as a follow-up study of the Project WAVE cohort of primarily African American women living in low-income neighborhoods in Baltimore City, Maryland. The overall goal of the study was to seek a greater understanding of the community and neighborhood contextual factors related to violence, with specific attention paid to intimate partner violence.

Systematic measurement development processes were applied to psychometrically develop these measures. In the first phase, an extensive literature search and formative interviews were conducted to gather a list of possible items to be included. To generate a list of items related to the extent to which women perceive that their community shares common values that are supportive of helping abused women, we first conducted formative interviews with 23 women who had experienced recent IPV. We asked women to talk about their interactions with their neighbors and their experiences of abuse. The resulting themes were used to develop the initial formulation of the social norms scales.

In the second step, in-depth cognitive interviews were conducted to assess the inclusiveness and comprehensiveness of the measurement items. Lastly, in the third step, the final scales were assessed in the study sample of 176 women. These women were recruited from a previous project (Project WAVE). Women in the original study were recruited from low-income urban neighborhoods in Baltimore City, with active recruitment from outpatient and community clinics and neighborhood Healthy Start Services. To be eligible, the women needed to be over the age of 18 and consent to participate in the private interview session.

The items included in the final scales were scored on a Likert scale with a score of 5 = strongly agree with the statement, 4 = agree, 3 = neither agree or disagree, 2 = disagree, and 1 = strongly disagree. The data have been factor analyzed within their conceptualized subscale using a principal components analysis with a varimax rotation. The rotation was used in the data reduction analysis to obtain a one-factor solution for each scale construct. We describe the final scale constructions below; the psychometric testing of the scales is described in “Results.” For each of the subscales, a mean item score was calculated to enable comparisons to be made between subscales with differing item number compositions.

The final descriptive and injunctive social norms scales each included eight items that were comparative in their construction. The descriptive social norm items (except for one) began “In my neighborhood, neighbors will…,” whereas the injunctive social norms (except for one) began “I think neighbors should…” followed by the social norm. The final descriptive social norms scale included eight items such as “In my neighborhood, neighbors will call the police when they hear or see a couple yelling or screaming,” and “In my neighborhood neighbors will go out of their way to help a woman who is being abused.” The final scale had a mean item score of 3.08 (SD = 1.04) with internal reliability alpha equal to 0.92, and factor analyses indicated that the single scale explained 68% of the variance, with item- to- total factor loading ranging from 0.75 to 0.87. The injunctive social norm scale included eight items that assessed what women believed should happen in their neighborhood, including items such as “I think neighbors should try to break up a fight between a couple,” and “I think neighbors should go out of their way to try and help a woman who is being abused.” This single factor scale had a mean score of 3.41 (SD = 1.08) with an internal consistency equal of 0.92, and the items explained 70% of the total variance, with item- to- total factor loading ranging from 0.64 to 0.83. Difference scores were computed by subtracting the descriptive norms from the injunctive norms to ascertain the association between the relative difference in these two forms of social norms and potential influencing factors listed below.

Potential Influencing Factors We assessed a number of covariates that may influence a woman’s perception of social norms, including her own experience of intimate partner violence as an adult, time lived in her current neighborhood, and overall perception of her neighborhood.The perceived reasons for neighbor involvement and non-involvement toward assisting women experiencing intimate partner violence encompassed two subscales examining (1) reasons for community involvement and (2) reasons for community non-involvement. In the first scale, items measuring reasons why neighbors would be actively involved in neighborhood domestic violence included five items, with an item mean = 3.24(SD = 1.05; α = 0.83). In the second subscale, reasons for non-involvement included eight items such as “neighbors think it is a woman’s own fault if she is being abused,” “neighbors believe that couples fighting is a private matter,” “neighbors think that there is nothing that can be done to deal with the problem of domestic violence,” and “neighbors don’t know how to help a woman who is being abused” (mean 3.22; SD = 0.95; α = 0.85).

Intimate Partner Violence Status The Conflict Tactics Scale²⁹ was used to assess if women had ever experienced physical or sexual abuse as an adult (over the age of 18) and whether the abuse was experienced in the last year (current abuse). Responses were categorized as never experiencing abuse as an adult, ever experiencing abuse as an adult but not in the last year, and experiencing abuse in the last year.

Time Lived in Neighborhood To anchor women’s conceptualization of the boundaries of their neighborhood, women were asked where, in what neighborhood, they currently resided. They were then asked how long they had resided in this neighborhood. Responses were dichotomized to less than or greater than 4 years.

Perception of Neighborhood Women were asked about the perception of their neighborhood and to indicate if they agreed or disagreed with the statement that their neighborhood is a “good place to live.”

Sociodemographics Sociodemographics assessed included the respondent’s racial/ethnic background, educational attainment, income, and marital status. Racial/ethnic background consisted of two categories: (1) African American and (2) Other. Over 95% of the women self-identified as African American, and therefore, this is not included in the final analyses. Education: Respondents reported their highest grade in school. This variable was then collapsed into two categories: (1) less than high school and (2) high school or equivalent. Income: Respondents selected one of eight income levels that represented their total family monthly income. A mean income level was computed for each category and divided by the number of household residents this income supported. The categories were collapsed into two levels per capita: (1) <$300.00 a month and (2) >$300.00 a month. Marital status categories were collapsed into two categories: (1) not currently married or living with someone and (2) currently married or living with someone.

Analysis

Psychometric analysis of the final scales was determined by the use of a systematic process incorporating univariate examination of the subscale item distribution, the calculation of a mean item response for each subscale, factor analysis with a varimax rotation, and an internal consistency using Cronbach’s alpha utilized to determine scale properties. Paired t tests were used to assess participant perceptions of injunctive and descriptive norms items. Relations between the subscales of social norms and reasons for neighbor involvement or non-involvement were assessed using correlation coefficients. Mean score responses using an analysis of variance were utilized to assess the relationship between social norms and community interaction subscales, and categorical sociodemographic and IPV-related variables. For all analyses, an a priori p value of 0.05 was used as the criterion by which to assess statistical significance.

Results

A total of 176 low-income urban women are included in our analysis, and their sociodemographic information is displayed in Table 1. Over half of the women had a per capita monthly income of $300 or less, experienced physical or sexual abuse as an adult, lived in their neighborhood for over 4 years, and agreed that their neighborhood was a good place to live.

Table 1.

Characteristics of 176 women involved in the development of the social norms and community interaction toward intimate partner violence measurement tools

Characteristic	N	%
Income level
<$300 per capita/month	129	73.3
>$300 per capita/month	47	26.7
Education level
Less than high school	76	43.2
High school	71	40.3
More than high school	29	16.5
Married or living with someone
No	135	76.7
Yes	41	23.3
Time lived in neighborhood
<4 years	69	39.2
>4 years	107	60.8
Neighborhood good place to live
Not agree	70	39.8
Agree	106	60.2
Adult domestic violence
No abuse	73	41.5
Ever abused, not in last year	68	38.6
Current abuse	35	19.9

Injunctive and Descriptive Social Norms

The 16 items of the descriptive and injunctive social norms scales were developed to comprise eight paired items, each assessing the two different types of norms. The univariate and total scale descriptive statistics are listed in Table 2. Women were least likely to agree that their neighbors should try to break up a fight between a couple (mean = 2.97, SD = 1.46) or that women who are being abused should feel comfortable talking to their neighbors about the abuse (mean 3.07, SD = 1.43), and more likely to agree that neighbors should call the police when witnessing a couple physically fighting (mean = 3.85, SD = 1.15) or witnessing couples yelling and screaming (mean = 3.83, SD = 1.17). When asked how neighbors currently respond to intimate partner violence, women were least likely to agree that their neighbors try to break up a fight between a couple (mean = 2.62, SD = 1.25) or offer a woman who is being abused a place to stay (mean = 2.88, SD = 1.47). Neighbors were likely to call the police when witnessing a couple physically (mean = 3.37, SD = 1.32) or verbally (mean = 3.56, SD = 1.30) fighting. The psychometric properties of our two social norms scales showed high internal consistency. Paired t tests were conducted to assess individual participant perception differences between the two forms of social norms. The majority of the paired t tests indicate statistically significant discrepancies between the two forms of social norms, with participants reporting higher scores on the injunctive norms scale than the descriptive norms.

Table 2.

Intimate partner violence injunctive and descriptive social norms

Social norm item	Injunctive norm (“I think neighbors should…”)	Descriptive norm (“In my neighborhood, neighbors will…”)	Paired t test (df = 175)
	Mean (SD)	Mean (SD)
Try to break up a fight between a couple	2.97 (1.46)	2.62 (1.25)	3.03*
Try to convince a woman who is being abused that she should leave her partner	3.27 (1.45)	3.07 (1.32)	1.69
Offer a woman who is being abused a place to stay	3.32 (1.40)	2.88 (1.47)	3.77*
Try to convince a woman who is being abused that she and her partner should get counseling	3.47 (1.36)	3.05 (1.23)	3.83*
Go out of their way to try to help a woman who is being abused	3.50 (1.36)	3.09 (1.34)	3.69*
Call the police when they hear or see a couple physically fighting	3.85 (1.15)	3.34 (1.32)	4.46*
Call the police when they hear or see a couple yelling and screaming	3.83 (1.17)	3.56 (1.30)	2.42*
I think that a woman who is being abused should feel comfortable talking to her neighbors about ita	3.07 (1.43)	3.03 (1.28)	0.33
Mean item score	3.41 (1.08)	3.08 (1.04)	4.08*
	α = 0.93; 68% of total variance explained	α = 0.84; 60% of total variance explained

*p < 0.05

^aDescriptive norm wording for this item is: “In my neighborhood, a woman who is being abused feels comfortable talking to her neighbors about it.”

Reasons for Community Non/Involvement Toward IPV The reasons for community interaction scale comprised two subscales: reasons for neighbors being involved and reasons for non-involvement (Table 3). Participants were likely to agree that their neighbors believe there is no excuse for a partner to beat up on a woman (mean = 3.52, SD = 1.37), and that neighbors feel pity or sorry for a woman who is being abused (mean = 3.52, SD = 1.31). Reasons for non-involvement included neighbors’ having too many problems of their own (mean = 3.39, SD = 1.42) and that couples’ fighting is a private matter (mean = 3.41, SD = 1.43). Neighbors were also perceived as not knowing how to help a woman who is being abused (mean = 3.24, SD = 1.25) and knowing very little about domestic violence (mean = 3.21, SD = 1.42). The psychometric properties of the subscales were found to be adequate for our sample of participants. However, to maintain consistency with the other potential covariates, the two scales were dichotomized by a median split (mean score = 3.00) with those participants with a mean score of ≤3.00 categorized as “low” and those with a mean score of more than 3.00 categorized as “high.” Reasons for neighbor involvement were categorized as 49% low and 51% high, and non-involvement as 46% low and 54% high.

Table 3.

Descriptive statistics for women’s perceptions of reasons for community assistance toward women experiencing IPV

Item	Mean	SD	Item total r
Reasons for neighbor involvement
Neighbors think that a woman who is being abused should feel comfortable talking about it with them	2.86	1.40	0.61
Neighbors think that you should offer a woman who is being abused a place to stay	2.99	1.38	0.65
Neighbors think people should call the police when they hear a couple physically fighting	3.32	1.30	0.68
Neighbors pity or feel sorry for a woman who is being abused	3.52	1.31	0.66
Neighbors believe that there is no excuse for a partner to beat up on a woman	3.52	1.37	0.55
Mean item score	3.24	1.05
α = 0.84; 60% of total variance explained
Reasons for neighbor non-involvement
Neighbors think it’s a woman’s own fault if she is being abused	3.14	1.55	0.51
Neighbors think that couples’ fighting is just a normal part of being together	3.36	1.34	0.56
Neighbors have too many problems of their own to deal with someone who is being abused	3.39	1.42	0.77
Neighbors believe that couples’ fighting is a private matter	3.41	1.43	0.74
Neighbors think there is nothing that can be done to deal with the problem of domestic violence	3.03	1.24	0.70
Neighbors think that domestic violence isn’t a problem in our community	3.04	1.39	0.54
Neighbors generally know very little about domestic violence	3.21	1.42	0.70
Neighbors don’t know how to help a woman who is being abused.	3.24	1.25	0.56
Mean item score	3.22	0.95
α = 0.85; 63% of total variance explained

Factors Associated with Perceived Injunctive and Descriptive IPV Social Norms

Statistically significant associations were found between the descriptive and injunctive social norms and reasons for community involvement in assisting women experiencing IPV; participants perception of their neighborhood; and experience of abuse as an adult (Table 4). Lowered descriptive norms were found among younger women; participants who reported that they did not think their neighborhood was a good place to live; those who had ever experienced intimate partner violence as an adult; those who perceived lower reasons for neighbor involvement; and those who perceived higher reasons for neighbor non-involvement toward assisting women experiencing IPV. Higher levels of injunctive social norms were statistically associated with living in a good place and increased perceived reasons for neighbor involvement toward assisting women experiencing IPV.

Table 4.

Associations between intimate partner violence injunctive and descriptive social norms, sample characteristics, and perceived reasons for neighbor involvement/non-involvement toward assisting women experiencing IPV (results in bold indicate statistical significance at p < 0.05)

		Injunctive social norms, mean (SD)	Descriptive social norms, mean (SD)	Difference between descriptive and injunctive norms, mean (SD)
Age	<35 years	3.34 (1.14)	2.87 (1.13)	0.47 (1.06)
	>35 years	3.44 (1.01)	3.23 (0.92)	0.22 (1.04)
Current marital status	Not married	3.36 (1.10)	3.13 (1.03)	0.23 (1.06)
	Married	3.53 (0.91)	2.91 (1.00)	0.62 (0.98)
Education level	Less than high school	3.58 (0.94)	3.03 (1.01)	0.55 (1.12)
	High school or greater	3.26 (1.13)	3.12 (1.01)	0.14 (0.97)
Income level (per capita)	<$300/month	3.33 (1.03)	3.01 (0.98)	0.32 (1.07)
	>$300/month	3.59 (1.13)	3.28 (1.10)	0.31 (1.00)
Time lived in neighborhood	< 4 years	3.33 (0.89)	3.01 (0.87)	0.32 (1.06)
	> 4 years	3.47 (1.18)	3.12 (1.13)	0.35 (1.04)
Neighborhood good place to live	Not agree	3.17 (1.11)	2.86 (0.94)	0.31 (1.26)
	Agree	3.57 (1.03)	3.23 (1.07)	0.34 (0.89)
Experience intimate partner violence as an adult	Never	3.57 (1.11)	3.33 (1.04)	0.24 (0.96)
	Ever	3.20 (1.04)	2.89 (0.94)	0.31 (1.07)
	Current	3.50 (1.02)	2.94 (1.14)	0.56 (1.19)
Neighbor involvement	Low	2.97 (1.01)	2.30 (0.72)	0.58 (1.13)
	High	3.81 (0.95)	3.75 (0.80)	0.07 (0.90)
Neighbor non-involvement	Low	3.56 (1.00)	3.47 (0.92)	0.05 (0.93)
	High	3.26 (1.10)	2.75 (0.99)	0.53 (1.10)

Differences between the participants’ perceived level of descriptive norms and injunctive norms were calculated, and the means stratified by selected influencing factors are presented in Table 4. Statistically significant differences between descriptive (what is commonly done) and injunctive norms (what ought to be done) were found for marital status, education level, reasons for neighbor involvement, and reasons for neighbor non-involvement in assisting women.

Discussion

In this study, we utilized the social influencing factor of social norms to construct and test measurement tools looking at descriptive norms—that is, what people actually do in a given situation—and injunctive norms—perceptions of what people ought to do in a given situation. The low-income urban women who participated in this project convey that the current actions of their neighbors are insufficient to assist women experiencing IPV, as indicated by the significantly lowered levels of descriptive norms compared to injunctive norms. This relationship was also found for the majority of the individual paired items of the two norms scales, indicating that women, especially those who have experienced abuse, would like to see their neighbors offer more instrumental support than they are currently providing to women who are abused by their intimate partners. Significantly higher perceptions of neighbor descriptive norms were found for older women; those who perceived of the neighborhood as a good place to live; those who had never experienced IPV; and those who reported their neighbors having greater reasons to intervene and fewer reasons not to intervene. Lower levels of injunctive norms were found for women with a high school education or greater; those who reported the current neighborhood was not a good place to live; and neighbors having fewer reasons to intervene. Significant differences between reported descriptive and injunctive norms were found for marital status, with married women reporting a greater increased difference than unmarried women. Women with less than a high school education reported significantly greater differences between the two norms than women with higher educational attainment. In addition, participants reporting few reasons for neighbor involvement in helping women experiencing IPV showed significantly greater difference scores for social norms than those perceiving more reasons for neighbor involvement, demonstrating that those participants perceiving currently low reasons for neighbor involvement/high reasons for non-involvement do want neighbors to engage in assisting women living with IPV. Although approaching significance, one of the greatest differences between descriptive and injunctive norms was found for women who are currently experiencing IPV. The descriptive norm mean was calculated to be 2.94, with the injunctive norm calculated to be 3.50; a difference of 0.56, indicating women who are experiencing IPV reporting what is currently happening to assist women experiencing IPV to be less than the midpoint (below 3.00; neither agree or disagree). However, women currently experiencing IPV do report that neighbors ought to be doing much more to assist women such as themselves; these women represent a key stakeholder for this investigation.

Overall, the performance of the injunctive and descriptive norms and reasons for neighbor involvement scales support its use as a tool to investigate social norms toward neighbors taking action to assist women experiencing IPV. The descriptive and injunctive social norms scales demonstrated a high level of internal reliability and were significantly associated with other influencing factors thought to be associated with social norms. The scales were shown not to be consistently influenced by women’s sociodemographics but were influenced by perception of their neighborhood and the IPV experienced as an adult. A future longitudinal assessment could be used to ascertain the directionality and strength of the relationship between these constructs and experiences and perceived norms and neighbor involvement toward assisting women experiencing IPV.

The need to increase the actual and perceived support to be afforded by informal sources of support is particularly important as these are the sources women experiencing IPV will typically turn to for help and assistance, and have been found to be a key protective factor associated with increased coping, health, and well-being.^{10,21–23,30} The utilization of formal sources of support (shelters, crisis hotlines, law enforcement, health care providers) were found to be positively related to the severity of the abuse experienced by women, with more women seeking this form of support as the severity of the violence increased.⁹ Therefore, to best meet the needs of abused women, especially in potentially earlier stages of IPV when informal sources of support are more widely used than formal support, it is important to examine methods to increase the capacity of community neighbors’ safe involvement in assisting women. Participants’ responses to the descriptive and injunctive norm items elicit possible areas of action the public health community can take to assist women experiencing IPV. These social norms indicate that intervention in an urban environment could focus on programs and interventions to increase the descriptive norms, what is currently being done to better match women’s perceptions of what ought to be done to assist women in need. For example, one area of development could be increasing neighborhood awareness of what could be done to assist women experiencing IPV. Women in the current study reported that they perceived that their neighbors did not know what to do if they thought a woman in their neighborhood was being abused by her partner. These results point to the need to engage urban community members to work collectively to increase knowledge, awareness, and efficacy in assisting women who are involved in violent relationships. Such norms must be addressed in our public health prevention efforts. These efforts can be enhanced with the integration of constructs of social influence to develop structurally based public health campaigns and programmatic methods to define social norms around assisting those experiencing IPV.^31–33

Intimate partner violence is a complex social and public health issue. If people intend to act in desired ways in regards to IPV but fail to carry out the actual behavior, a complex constraint (i.e., norm) limiting their action may exist. Previous research has shown that combined support provided by family, friends, formal authorities (i.e., police response), and the community have a profound effect on decisions to continue or leave an abusive relationship.^8,10,24,34 Real or perceived disapproval by each of these parties or spheres of society is also likely to serve as a social sanction for abusers that may be more poignant than the threat of legal punishment. Hence, the behaviors of persons coexisting with those directly involved in IPV are crucial in overall IPV reduction. Future interventions should focus on generating action by each of these groups, and the social norms that govern those actions, to ensure that timely and responsive support is available to women experiencing IPV.

It is important to note that social change, particularly shifts in descriptive and injunctive subjective norms, occurs gradually over time. This change transpires in response to a range of interventions that eventually result in a “tipping point” when large-scale behavioral change occurs.³⁵ Even community galvanization can be incremental particularly in communities where deep-rooted norms may not be conducive, in assisting women living with IPV and unintentionally providing the perception of tolerance of IPV.

The results of the current study do carry methodological limitations to their generalizability. The work presented is cross-sectional and represents the perceptions of the low-income urban women interviewed. The association of the norms and experiences of IPV are correlational and therefore directionality cannot be assigned. While we are unable to assess which came first, the experience of IPV or the perception of the injunctive or descriptive social norm, the results from the study indicate that effective and responsive programs and policies should be developed to change the perception that not enough is currently being done to address assistance for women experiencing IPV.

Conclusion

Intimate partner violence is a complex public health issue enveloped by social norms that have the ability to influence whether action or is not taken by the community to assist women experiencing IPV. An ecological macro-level approach is necessary to influence social norms surrounding collective action or inaction in regards to IPV. If society as a whole believed and acted in ways that expressed willingness to assist women experiencing IPV and proactively assisted those who experienced IPV, and social norms around IPV in turn clearly rejected such violence, the incidence and prevalence of IPV would in turn lessen. Public health must continue to educate communities regarding the complex social and health issues surrounding IPV and the directions people can take to assist women living with HIV, and in turn, over time, social norms will eventually become anti-IPV in character and practice.