Abstract
Most prior research on violence among tertiary students has been restricted to dating and sexual violence and neglected cross-cultural variation. To provide more comprehensive and intersectional understandings, this study examined the prevalence and cultural differences in intimate partner and family violence among women tertiary students in Australia. Data were collected from 1,845 women studying at post-secondary institutions and weighted by age and country of birth to reflect the population composition. Eighty-seven percent (86.7%) of participants reported having experienced violence (physical, sexual, psychological, and/or financial) from an intimate partner and/or family member during adulthood. Students who identified as Anglo were more likely to report victimization by intimate partners, while those who identified as non-Anglo and multicultural were more likely to report victimization by family members. The findings highlight the need for tertiary education institutions to prioritize preventing and responding to intimate partner and family violence in their culturally diverse student bodies.
Keywords
Introduction
Violence against women is a pervasive human rights and public health problem in Australia and around the world (Australian Institute of Health and Welfare [AIHW], 2019; World Health Organization [WHO], 2021). This violence takes many forms including physical, sexual, psychological, and financial, and most often occurs in intimate and family relationships (AIHW, 2019; WHO, 2021). Throughout this article, the term “intimate partner violence” (IPV) is used to refer to violence between current and former intimate partners, and “family violence” (FV) is used to refer to violence between non-intimate family members or relatives. These are collectively referred to as “intimate partner and family violence” (IPFV). This broader conceptualization reflects the co-occurrence and commonalities between IPV and FV, including in their etiology, risk factors, dynamics, and outcomes (Smith Slep & Heyman, 2001; Tolan et al., 2006). Globally, one in four (26%) ever-partnered women have experienced physical or sexual IPV since the age of 15 (WHO, 2021). In Australia, one in six (17%) women have experienced physical or sexual violence from a cohabiting partner since the age of 15 (Australian Bureau of Statistics [ABS], 2017), and one in 10 (10%) have experienced physical or sexual violence from a male relative since the age of 16 (Mouzos & Makkai, 2004). While such figures provide an indication of the scale of the problem, IPFV is not a unitary phenomenon, and some sociocultural groups face increased vulnerability (The Equality Institute, 2017).
Over the past three decades, violence against women in tertiary education has garnered substantial public and academic attention (Guberman et al., 2018; Rennison & Addington, 2014). This group is significant not only because of their unique risks, challenges, and impacts (Duval et al., 2020; Kaukinen, 2014; Molstad et al., 2023), but because tertiary institutions, as places of community, learning, and knowledge creation, are uniquely positioned to prevent gendered violence and lead change in wider society (Our Watch, 2021b, 2021c). Yet, little research in this area has been conducted in Australia, and most international research has been restricted to dating and sexual violence (Rennison & Addington, 2014). The full extent of violence in students’ relationships is unclear, hindering the development and expansion of policies and programs to address gendered violence in this cohort more comprehensively. Furthermore, despite the immense cultural diversity of the student subpopulation and population at large, few Australian studies have examined IPFV across cultural groups. This is needed to inform prevention and intervention efforts for culturally diverse victim-survivors in the tertiary education sector and beyond. The current study therefore sought to estimate the prevalence of IPFV victimization among women tertiary students in Australia and to explore cross-cultural differences therein.
Literature Review
Sexual Violence Among Tertiary Students
The primary focus of violence reduction in the tertiary education sector is sexual violence, particularly sexual assaults occurring on campuses (Guberman et al., 2018; Rennison & Addington, 2014). Research has consistently demonstrated the pervasiveness of this problem, with the most reliable estimates suggesting that one in five (20%) female college students in the United States (US) are sexually assaulted while in college (Muehlenhard et al., 2017). Systematic review evidence suggests that the most common forms of campus sexual assault are unwanted sexual contact (reported by 1.8% to 34% of college women in the US) and sexual coercion (1.7% to 32%), followed by incapacitated rape (1.8% to 14%) and forcible rape (0.5% to 8.4%; Fedina et al., 2018). The increased visibility of the issue has sparked global conversations and initiatives to address sexual violence in tertiary education in the US and beyond (Towl & Walker, 2019).
In 2016, Universities Australia launched the Respect. Now. Always. initiative to prevent and improve responses to sexual violence in university communities. As part of this initiative, the Australian Human Rights Commission (AHRC, 2017) conducted the first National University Student Survey on Sexual Assault and Sexual Harassment to establish the scale of the problem. Results were published in the landmark Change the Course report, revealing that almost two-thirds (63%) of female university students had been sexually harassed in 2016 and one in 10 (10%) had been sexually assaulted in 2015–2016 (AHRC, 2017). Universities Australia has since implemented a 10-point action plan to address the issue (Universities Australia, 2017).
While the attention to sexual violence in the tertiary education sector is critically important, it does not reflect the full extent of violence in students’ lives (Guberman et al., 2018). On the contrary, sexual assault may represent the “tip of the iceberg,” with data from both general population and student samples indicating that women are more commonly affected by non-sexual violence, and most often by their intimates (AIHW, 2019; Oswalt et al., 2018; WHO, 2021). Focusing exclusively on sexual violence may limit students’ awareness and help-seeking for other forms of violence and forgo opportunities to expand violence prevention and support programs. A broader focus on violence in students’ relationships may therefore be beneficial.
IPFV Among Tertiary Students
Though smaller than the literature on sexual violence, there is a developing body of research on IPV among tertiary students. Prevalence estimates vary widely based on definitions, reference periods, and samples, though collectively demonstrate that the issue is widespread (Rennison & Addington, 2014). Data from nationally representative samples of US college students have reported 12-month victimization rates of 16% to 34% for physical IPV, 10% to 36% for sexual IPV, and 17% to 76% for psychological IPV (Scherer et al., 2016). The largest international prevalence study in this area to date is the International Dating Violence Study (IDVS), which collected data from university students in all major world regions (Straus, 2011). Across nearly 16,000 students in 21 countries, a median of one in four (26%) reported physical assault, and one in four (24%) reported sexual coercion, by a dating partner in the previous 12 months (Chan et al., 2008). Among women students in the US, 28% reported physical assault, 28% reported sexual coercion, and 34% reported psychological aggression, with 55% reporting any dating violence victimization in the previous 12 months (Sabina & Straus, 2008).
There is a dearth of research on IPV among tertiary students in Australia, though Australian data from the IDVS provide some insight into its extent. Among female participants in Australia, 20% reported being physically assaulted, and 22% reported being sexually coerced, by a dating partner in the previous 12 months (Chan et al., 2008). In a more detailed analysis of IDVS lifetime data from female students in Australia and New Zealand, Cale et al. (2017) found that 27% had ever experienced minor physical assault, and 11% had ever experienced severe physical assault, by a dating partner. With regard to sexual coercion, 32% reported sex pursuant to insisting, 3.4% reported sex pursuant to threats, and 2.4% reported sex pursuant to force (Cale et al., 2017). Most commonly, these students reported psychological dating violence, with 74% reporting minor psychological aggression and 24% reporting severe psychological aggression (Cale et al., 2017). Evidently, the available data suggest that IPV is a pervasive problem for tertiary students.
In addition to violence in their intimate relationships, students may be vulnerable to violence in their family relationships, though this issue has received very little attention. This was highlighted by Guberman et al. (2018) reporting on Build. Act. Change., a gender-based violence program implemented at a university in Canada. Over the course of the two-year project, data were collected from multiple sources (e.g., surveys, focus groups, and reports to counselors) regarding the scope and impacts of violence against women students. The authors found that “what affected students most deeply was their experience of harm in the home by family members—fathers, brothers, mothers, and extended family” (Guberman et al., 2018, p. 79), which took physical, sexual, emotional, and financial forms. However, students described a lack of understanding and support for this issue at university (Guberman et al., 2018). Indeed, FV remains underrecognized and understudied in both the student and general population, and prevalence estimates for students are lacking. Given growing recognition of the overlap between IPV and FV, more integrative research examining IPFV is warranted (Tolan et al., 2006).
Theoretical Frameworks
Two key theoretical frameworks that underpin global and national approaches to gender-based violence prevention are intersectional feminism and social-ecological models (Our Watch, 2021a; UN Women, 2015; WHO, 2010). Traditional feminist theories posit that violence against women is driven by unequal gender power relations (Dobash & Dobash, 1979). Consistent with this theorizing, research has demonstrated a strong and consistent association between gender inequality and the prevalence of violence against women (Our Watch, 2021a). Furthermore, research suggests that four specific expressions of gender inequality consistently predict or “drive” this violence: the condoning of violence against women, rigid gender roles, gendered patterns of autonomy and decision-making, and male peer relations that emphasize dominance and aggression (Our Watch, 2021a). Yet, gender alone is not sufficient to explain the differential vulnerability of women across sociocultural groups (Vaughan et al., 2015). Intersectional feminism, first introduced by Crenshaw (1991), therefore suggests that women's experiences of violence are shaped by multiple social identities, including but not limited to gender, race, ethnicity, class, sexuality, and ability. An intersectional analysis considers how gender inequality and the gendered drivers of violence intersect with other forms of oppression and discrimination such as racism, colonialism, classism, heterosexism, cissexism, and ableism (Our Watch, 2021a).
Social-ecological models suggest that violence against women is the outcome of an interplay of factors at the individual, relationship, community/organizational, system/institutional, and societal levels (Our Watch, 2021a; UN Women, 2015). This framework departs from historically simplistic and individualistic explanations of violence, such as those focused on the perpetrator's psychology, past experiences, or personal circumstances (Our Watch, 2021a). In contrast, social-ecological models suggest that individual behavior is influenced by the norms, structures, and practices at each level of one's environment (Our Watch, 2021a; UN Women, 2015). Combining an intersectional and social-ecological lens allows us to understand how multiple social identities shape women's experiences of IPFV through factors operating at multiple levels (The Equality Institute, 2017). One such dimension of identity gaining increasing recognition is culture.
Examining Culture in IPFV Research
The terms “race,” “ethnicity,” and “culture” are often used interchangeably and imprecisely in the IPFV literature. While no consensus definitions exist, “race” is typically associated with shared physical characteristics, and “ethnicity” is more broadly associated with shared racial, national, linguistic, or cultural origins (Malley-Morrison & Hines, 2007). “Culture” is broader still, encompassing the shared beliefs, values, norms, and practices of a group, including but not limited to racial and ethnic groups (Ung, 2015). Attending to culture in IPFV research is critical to advancing intersectional understandings and approaches to addressing the issue. However, this work also presents serious complexities and risks, particularly related to biased interpretations and their consequences for marginalized communities. Indeed, there is a tendency to attribute violence in minority groups to cultural factors, while overlooking within-group variability and sociodemographic and structural influences, which can function to blame cultures and perpetuate harmful stereotypes (Kasturirangan et al., 2004; Malley-Morrison & Hines, 2007; Sokoloff & Dupont, 2005). When considering the impact of cultural minority status on IPFV, it is crucial to acknowledge both the immense diversity of minorities and their shared influences, including experiences of oppression both within their communities and from the dominant society (Kasturirangan et al., 2004). If cross-cultural differences in IPFV are contextualized and portrayed sensitively, this may inform better support for marginalized victim-survivors whose experiences and needs may differ from the majority (Malley-Morrison & Hines, 2007).
Cross-Cultural Differences in IPFV
The prevalence of IPFV among women appears to vary across regions and countries (e.g., WHO, 2021) and racial/ethnic groups within countries (e.g., Cho, 2012). In the US, data from the National Intimate Partner and Sexual Violence Survey 2010–2012 indicate that the lifetime prevalence of IPV (including physical violence, sexual violence, and stalking) is highest among multiracial women (57%), followed by American Indian/Alaska Native women (48%), Black women (45%), White women (37%), and Hispanic women (34%), and lowest among Asian or Pacific Islander women (18%; Smith et al., 2017). Data from other US national surveys show similar patterns across racial/ethnic groups; however, differences are less pronounced when other relevant variables are controlled (Breiding et al., 2008; Cho, 2012; Rennison & Planty, 2003). In a review of national survey data, Field and Caetano (2004) found that higher rates of IPV among ethnic minorities were reduced after adjusting for socioeconomic characteristics and alcohol use, though Black couples remained at higher risk for IPV than White and Hispanic couples. More recent analyses have produced similar and different results. For example, Breiding et al. (2008) found that when adjusting for age, income, and education, IPV prevalence was higher among multiracial women and lower among Asian and Hispanic women, but did not significantly differ for Black, American Indian/Alaska Native, or Native Hawaiian/Pacific Islander women, compared to White women. Cho (2012) found that when adjusting for age, education, employment, financial security, and social network, IPV prevalence was lower among Asian women, but did not significantly differ for Black or Latina women, compared to White women. In sum, this evidence suggests that IPV prevalence may differ across cultural groups, though crude differences may be at least partially explained by sociodemographic and other risk factors.
In Australia, IPFV appears to occur at higher rates in Indigenous than non-Indigenous communities, with an estimated three in five (65%) Aboriginal and Torres Strait Islander women having experienced physical or sexual IPV since the age of 15 (Webster, 2016). However, data on IPFV among other cultural minorities are scarce and inconclusive. In a supplementary analysis of Personal Safety Survey 2012 data, Cox (2016) found higher 12-month prevalence of physical or sexual violence by a male partner among women born in Australia (2.3%) than those born overseas (1.7%), and among women born in non-main English-speaking countries (1.9%) than those born in main English-speaking countries (1.2%). A more recent Australian study that collected data from 2012–2015 found no significant differences in the unadjusted or adjusted rates of IPV (including physical, psychological, and financial abuse) between migrant and non-migrant women (Satyen et al., 2021). Beyond these broad comparisons based on country of birth, there has been little exploration of cultural differences in IPFV in the Australian context.
IPFV in the tertiary student subpopulation may also vary cross-culturally. Data from the IDVS indicate that the 12-month prevalence of physical dating violence among female students ranged from 12% to 41%, and sexual dating violence ranged from 9.2% to 42%, across countries (Chan et al., 2008). Few studies have analyzed racial/ethnic differences in IPFV among students, particularly outside the US. Data from the National College Health Assessment (NCHA) 2011–2013 indicate that the 12-month prevalence of physically, sexually, and emotionally abusive intimate relationships was highest among Black students (3.4%, 2.1%, and 11%, respectively) and “Other” race/ethnicity students (2.8%, 2.3%, and 12%, respectively), and lowest among Asian students (1.7%, 1.2%, and 6.7%, respectively; Whitfield et al., 2021). When controlling for age, year in school, and marital status, “Other” race/ethnicity students had higher odds of all three types of IPV, Black students had higher odds of physical and emotional IPV, and Hispanic students had higher odds of physical IPV, than White students, while Asian students had lower odds of sexual and emotional IPV (Whitfield et al.). An analysis of NCHA 2011 data controlling for year in school, gender, and sexual orientation produced similar findings for Black, Hispanic, and Asian students (Oswalt et al., 2018). Additionally, this study found higher odds of physical and emotional IPV among multiracial students, and higher odds of physical IPV among Native American students, compared to White students (Oswalt et al.). These patterns broadly resemble those in the general population and suggest that cultural minority students may face increased risk of at least some forms of IPV.
In Australia, data on cross-cultural differences in violence among tertiary students are limited to sexual violence. The National University Student Survey on Sexual Assault and Sexual Harassment found that Aboriginal and Torres Strait Islander students were more likely to have been sexually assaulted (10%) and sexually harassed (62%) than non-Indigenous students (7.0% and 51%, respectively; AHRC, 2017). Conversely, international students were less likely to have been sexually assaulted (5.1%) and sexually harassed at university (22%) than domestic students (7.4% and 27%, respectively), and among domestic students, those who mainly spoke a language other than English at home were less likely to have been sexually harassed (45%) than those who mainly spoke English at home (53%; AHRC, 2017). To date, there have been no such comparisons of IPFV across cultural groups of students in Australia.
The Current Study
Evidently, the existing literature does not provide a comprehensive, cross-cultural picture of IPFV among women tertiary students in Australia. National and international data are largely restricted to sexual and dating violence, with little known about other forms of abuse and other intimate and familial relationships. Few studies have examined cross-cultural variation in IPFV in the student population, and research in the general population has yielded mixed results. These studies have varied in the degree to which they control other variables in cultural comparisons, and few have measured and adjusted for a comprehensive set of sociodemographics. Furthermore, most studies have defined culture in terms of race, ethnicity, or country of birth, yet these do not necessarily signify affiliation with a cultural group; participants’ self-identified culture may provide a more meaningful metric. Finally, most prior work on IPFV among tertiary students has focused on young university students, despite the increasing variety of post-secondary education options available (including both higher education and vocational education) and growing numbers of “non-traditional” or “mature-age” students. As tertiary students at large may face similar risks and challenges with respect to IPFV, more inclusive sampling is warranted.
To address the key gaps in the literature, the current study had two primary objectives: (a) to estimate the prevalence of IPFV victimization among women tertiary students in Australia and (b) to explore cross-cultural differences in victimization rates before and after adjusting for other sociodemographic factors. Additionally, to address some of the key limitations of prior research, we sought to: (a) examine violence from intimate partners and family members, considered separately and together (i.e., IPV, FV, and IPFV); (b) examine physical, sexual, psychological, and financial forms of abuse; (c) examine cross-cultural differences based on personal cultural identification; and (d) sample from the full population of adult women tertiary students in Australia, without restrictions based on age or institution type.
Methods
Study Design and Participants
A cross-sectional survey design was employed. Inclusion criteria were: identifying as a woman, being aged 18 years or older, currently studying at a tertiary education institution (e.g., university, college, or technical and further education [TAFE] institute) in Australia, and having sufficient English language proficiency to complete the survey. A total of 1,938 individuals participated in the study. Forty-two were excluded based on the eligibility criteria: 21 who did not identify as women or did not specify their gender, 3 who did not specify their age, and 18 who were not current students or did not specify their student status. A further 34 were excluded due to missing data for cultural identity, and 17 were excluded due to >50% missing data on the IPFV questionnaire. The final sample thus comprised 1,845 adult women tertiary students.
The sociodemographic characteristics of the sample are presented in Table 1. Participants were aged 18 to 70 years (M = 26.8, SD = 8.4). Seventy percent (69.7%) identified with Anglo cultures, 24.4% with non-Anglo cultures, and 5.2% with a mix of both. The sample generally reflected the population of adult female tertiary students in Australia on key demographic characteristics (i.e., age, state, area, country of birth, Indigenous status; see ABS, 2016). However, some groups were underrepresented, including ≥45-year-olds (4.2% sample vs. 10.9% population), New South Wales residents (12.3% sample vs. 32.5% population), urban residents (73.8% sample vs. 81.0% population), and migrants (27.6% sample vs. 36.1% population), while Victorian residents were overrepresented (50.7% sample vs. 26.6% population).
Sociodemographic Characteristics of Sample (N = 1,845).
Note. TAFE = technical and further education.
Measures
Sociodemographic Characteristics
Participants were asked to provide sociodemographic data including their gender identity, age, state, area type, employment status, tertiary institution type, domestic student status, country of birth, and cultural identity. Participants were also asked whether they had ever been in an intimate relationship since the age of 18 years.
Our measurement of cultural identity was informed by the Victorian Family Violence Data Collection Framework (Department of Premier and Cabinet, 2019). An open-ended question asked participants to specify the culture(s) that they most closely identified with. While cultural identity is multidimensional and cannot be fully captured by a single item or variable, this approach was intended to allow for a more subjective assessment that could reflect any combination of cultural factors (e.g., race, ethnicity, nationality, country of birth, language). Responses were categorized by Anglo-Saxon status and geographic region. Participants were categorized as Anglo if they identified with the majority culture(s) of Australia, Canada, New Zealand, the United Kingdom, and/or the United States. These countries represent the core “Anglosphere” as main English-speaking nations with close historical, cultural, and political ties (Legrand, 2016). Participants were categorized as non-Anglo if they identified with the culture(s) of any country outside the Anglosphere or as Indigenous (e.g., Aboriginal or Torres Strait Islander, Māori). Thus, Anglo status here accounts for both country and Indigeneity. Participants who identified with Anglo and non-Anglo cultures were categorized as “multicultural.” Cultural identity was also categorized as African, American, Asian, European, or Oceanian according to the United Nations geoscheme (United Nations Statistics Division, n.d.). Participants who identified with cultures of multiple regions were categorized as “multicultural.”
IPFV
Participants were asked whether they had experienced a range of violent behaviors from an intimate partner and/or family member since the age of 18 years. These included physically, sexually, psychologically, and financially abusive behaviors. Intimate partners were defined as any person that the participant had been in an intimate personal relationship with, including current and former spouses, boyfriends/girlfriends, dating partners, and sexual partners. Family members were defined as any person in the participant's immediate or extended family, including parents, grandparents, children, siblings, uncles/aunts, cousins, and in-laws.
Given the lack of existing instruments to measure IPFV comprehensively, items were adapted from multiple sources. Items measuring physical, sexual, and psychological abuse were adapted from the Family and Relationship Experiences and Attitudes Among College Students Survey (Gover et al., 2008). This survey is based on the Revised Conflict Tactics Scale, a widely used measure of IPV with cross-cultural reliability and validity among university students (Straus, 2016). Items measuring financial abuse and sexual harassment (the latter included in the sexual abuse subscale) were adapted from the Personal Safety Survey 2016 (ABS, 2017), the most comprehensive and robust survey of interpersonal violence in Australia. Example items include “Thrown something at you that could hurt” (physical abuse), “Used threats to make you have sex” (sexual abuse), “Prevented you from seeing family or friends” (psychological abuse), and “Controlled or tried to control your income or assets” (financial abuse).
For each item, participants could select “Partner” and/or “Family member” or “Neither.” On the basis of their responses, participants received a score of 1 (Yes) or 0 (No) on seven scales/subscales: IPFV, IPV, FV, physical abuse, sexual abuse, psychological abuse, and financial abuse. For IPFV, participants scored 1 if they selected “Partner” and/or “Family member” for any item and 0 if they selected “Neither” for all items. For IPV, participants scored 1 if they selected “Partner” for any item and 0 if they selected “Family member” or “Neither” for all items; FV scores were calculated analogously. For each abuse subscale, participants scored 1 if they selected “Partner” and/or “Family member” for any item within that subscale and 0 if they selected “Neither” for all items within that subscale. Other combinations were coded as missing (e.g., “Neither” selected for some items but all other items left blank). All scales demonstrated good to excellent reliability: IPFV α = 0.94; IPV α = 0.96; FV α = 0.91; physical abuse α = 0.90; sexual abuse α = 0.86; psychological abuse α = 0.87; financial abuse α = 0.85.
Procedure
Ethics approval was obtained from the Deakin University Human Research Ethics Committee. Participants were recruited through posts in over 400 Facebook groups for students in Australia. These posts included a brief overview of the study and specified that participants did not need to have experienced any violence to take part. The survey was open from February to September 2020. During the latter half of the recruitment period, Facebook groups for ethnic minority and international students were targeted to increase the cultural diversity of the sample.
The survey was administered online using Qualtrics software and took approximately 10–15 minutes to complete. Participants were required to read a plain language statement explaining the nature and purpose of the study, eligibility criteria, potential risks and benefits, and confidentiality and data storage procedures, before providing their consent. Participants were assured of the anonymity of their responses, with no identifying information collected. They were advised to only complete the survey if it was safe for them to do so, and that they could withdraw their participation at any time before submitting their responses. Contact details of relevant national helplines were provided before and after the survey, and a custom information sheet about IPFV services and resources in Australia was made available to view and download.
Data Analysis
Data analyses were conducted using IBM SPSS Statistics version 26. The proportion of missing data on outcome variables ranged from 0.2% to 4.0%. Missing data were included in descriptive statistics but excluded from bivariate and multivariate analyses. Participants who had never been in an intimate relationship were excluded from IPV analyses only. The prevalence of IPFV, IPV, and FV was calculated overall and as a function of abuse type. These estimates were weighted to the 2016 Australian Census data for adult female tertiary education students (ABS, 2016) to reflect the population distributions for age (18–24, 25–34, 35–44, ≥45 years) and country of birth (Australia, overseas). The term “prevalence” is used here when referring to weighted data for the full sample and the term “rate” is used when referring to unweighted data for subgroups.
Crude cultural differences in victimization rates were analyzed using a series of bivariate logistic regressions with one predictor (cultural identity by Anglo status or by region) and one outcome (IPFV, IPV, FV, physical abuse, sexual abuse, psychological abuse, or financial abuse) at a time. The cultural majority group (Anglo or Oceanian) served as the reference category. Adjusted cultural differences were analyzed using a series of multivariate logistic regressions including cultural identity (by Anglo status and by region) and other sociodemographic variables (gender identity, age group, area type, employment status, tertiary institution type, domestic student status, country of birth, and Indigenous status) as predictors for each outcome.
All assumptions for logistic regression were checked prior to analyses. Regression diagnostics were used to screen for outliers, high leverage points, and influential observations (according to Field, 2018). Most analyses contained small proportions of outliers and/or high leverage points, but these were not influential. However, in the IPFV analyses, one observation in the African subgroup was influential, and results must therefore be interpreted with caution. The beta coefficients in each model were screened for overly small or large standard errors, confirming the absence of overdispersion and complete separation, respectively (Field, 2018). Finally, collinearity diagnostics were used to screen for multicollinearity in the multivariate models. Variance inflation factors and variance proportions were within acceptable ranges, indicating a lack of multicollinearity (Belsley et al., 1980; Field, 2018).
Results
Prevalence
The weighted prevalence estimates of IPFV are presented in Table 2. Eighty-seven percent (86.7%) of participants reported having experienced violence from an intimate partner and/or family member during their adult lifetime. This most commonly took the form of psychological abuse (experienced by 81.3% of participants), followed by physical (56.0%), sexual (52.8%), and financial (48.3%) abuse. Three in four (73.5%) participants who had ever been in an intimate relationship reported having experienced violence from an intimate partner, and three in five (60.4%) participants reported having experienced violence from a family member. Polyvictimization was common, with 38.1% of participants (44.0% of those who reported any IPFV) reporting violence from both an intimate partner and family member.
Weighted Prevalence Estimates of Intimate Partner Violence (IPV), Family Violence (FV), and Intimate Partner and/or Family Violence (IPFV) Victimization.
Excludes participants who had never been in an intimate relationship.
Item frequencies indicated that psychological abuse was most often verbal in nature, including shouting/yelling (experienced by 72.8% of participants), insulting/swearing (63.6%), and name-calling (56.6%). It also frequently involved controlling behaviors (e.g., insisting on knowing the participant's whereabouts all the time; 47.2%) and threats of physical harm (39.5%). Physical abuse most often involved pushing/shoving (41.6%), grabbing (39.7%), and throwing things at the participant (34.2%). More severe physical violence was also common, including punching/hitting (27.5%), kicking (18.2%), and choking (16.5%). Sexual abuse most often involved insisting on sex when the participant did not want to (39.1%), making inappropriate comments about the participant's body or sex life (34.7%), and non-consensual sexual touching (30.1%). Substantial numbers of participants reported being threatened (16.4%) or forced (9.5%) into sex. Financial abuse most often involved controlling the participant's knowledge of or access to household money (29.7%), damaging/stealing property (27.2%), and controlling income/assets (23.9%) or study (23.9%).
Cross-Cultural Differences
Bivariate Analyses
Results of the bivariate analyses of associations between cultural identity and IPFV are presented in Table 3. Students who identified with non-Anglo cultures were significantly less likely to report any IPFV victimization than those who identified with Anglo cultures. Considered separately, Anglo students reported significantly higher rates of IPV than non-Anglo students, while non-Anglo and multicultural students reported significantly higher rates of FV than Anglo students. With regard to regional groups, students who identified with Asian cultures were significantly less likely to report any IPFV victimization than those who identified with Oceanian cultures. Considered separately, Asian students reported significantly lower rates of IPV, while multicultural students reported significantly higher rates of FV, than Oceanian students. Differences between other minority groups and the majority were not significant.
Bivariate Logistic Regression Analyses of Cultural Identity Predicting Intimate Partner Violence (IPV), Family Violence (FV), and Intimate Partner and/or Family Violence (IPFV) Victimization.
Note. CI = confidence interval; OR = odds ratio.
Excludes participants who had never been in an intimate relationship.
*p < .05; **p < .01; ***p < .001.
Results of the bivariate analyses of associations between cultural identity and violence/abuse types are presented in Table 4. Students who identified with non-Anglo cultures were significantly less likely to report all four types of abuse, and those who identified as multicultural were significantly less likely to report sexual abuse, than those who identified with Anglo cultures. With regard to regional groups, Asian-identifying students reported significantly lower rates of physical, sexual, and psychological abuse, and multicultural students reported significantly lower rates of sexual abuse, than Oceanian-identifying students. African-identifying students reported significantly higher rates of physical abuse than Oceanian-identifying students.
Bivariate Logistic Regression Analyses of Cultural Identity Predicting Types of Abuse from Intimate Partners and/or Family Members.
Note. CI = confidence interval; OR = odds ratio.
*p < .05. **p < .01. ***p < .001.
Multivariate Analyses
Results of the multivariate analyses of IPFV are presented in Table 5. After adjusting for other sociodemographic variables, cultural differences in IPFV, IPV, and FV were no longer significant, except that Asian-identifying students remained at significantly lower risk for IPV than Oceanian-identifying students. Intimate partner violence was otherwise predicted only by age group and country of birth, with older (25–34, 35–44, and ≥45-year-old) and Australian-born students significantly more likely to have experienced IPV than younger (18–24-year-old) and overseas-born students. Family violence was predicted by age group, gender identity, employment status, and tertiary institution type. In contrast to IPV, older (25–34 and 35–44-year-old) students were significantly less likely to have experienced FV than younger (18–24-year-old) students. Additionally, trans/non-binary, unemployed, and TAFE students were significantly more likely to have experienced FV than cisgender, full-time employed, and university/college students, respectively. Age group was the only significant predictor of IPFV combined, with 25–34 and 35–44-year-olds more likely to report any victimization than 18–24-year-olds.
Multivariate Logistic Regression Analyses of Sociodemographic Variables Predicting Intimate Partner Violence (IPV), Family Violence (FV), and Intimate Partner and/or Family Violence (IPFV) Victimization.
Note. CI = confidence interval; OR = odds ratio; TAFE = technical and further education.
Excludes participants who had never been in an intimate relationship.
*p < .05. **p < .01. ***p < .001.
Results of the multivariate analyses of violence/abuse types are presented in Table 6. As above, most cultural differences did not remain significant after adjusting for other sociodemographic variables. However, African-identifying students remained at significantly higher risk for physical abuse, and Asian-identifying students remained at significantly lower risk for sexual abuse, than Oceanian-identifying students. Abuse types were otherwise predicted only by age group and tertiary institution type. Older (25–34, 35–44, and ≥45-year-old) students were significantly more likely to have experienced all four types of abuse than younger (18–24-year-old) students (except ≥45-year-olds for psychological abuse), and TAFE students were significantly more likely to have experienced physical, psychological, and financial abuse than university/college students.
Multivariate Logistic Regression Analyses of Sociodemographic Variables Predicting Types of Abuse From Intimate Partners and/or Family Members.
Note. CI = confidence interval; OR = odds ratio; TAFE = technical and further education.
*p < .05. **p < .01. ***p < .001.
Discussion
This study provides the first estimates of the prevalence and cross-cultural differences in IPFV among women tertiary students in Australia. Over four in five participants reported having been victimized by an intimate partner and/or family member since the age of 18. Considered separately, three in four participants reported having been victimized by an intimate partner and three in five reported having been victimized by a family member. Students who identified with Anglo cultures were more likely to report any IPFV than those who identified with non-Anglo cultures. Considered separately, Anglo-identifying students were more likely to report violence from intimate partners, while non-Anglo and multicultural-identifying students were more likely to report violence from family members. However, these differences did not remain significant after adjusting for other sociodemographic variables. Findings for the full sample and cultural subgroups are discussed and integrated with the existing literature below.
Prevalence of IPFV
The prevalence estimates of 43%, 55%, and 62% for physical, sexual, and psychological IPV, respectively, from this study roughly correspond to those from the Australian and New Zealand component of the IDVS, which indicated lifetime prevalence rates of 27%, 32%, and 75% for minor physical assault, sexual coercion, and psychological aggression by a dating partner (Cale et al., 2017). These estimates are substantially higher than those among women in the general population in Australia, with the Personal Safety Survey 2016 indicating rates of 16%, 5%, and 23% for physical violence, sexual violence, and emotional abuse from a cohabiting partner since the age of 15 (ABS, 2017). Similarly, the prevalence estimates of 34% and 12% for physical and sexual FV, respectively, from this study are higher than those from the Australian component of the International Violence Against Women Survey, which indicated rates of 8% and 3% for physical and sexual violence from a male relative since the age of 16 (Mouzos & Makkai, 2004). There are no analogous prevalence estimates for financial abuse, though our results indicate that this less well-recognized form of IPFV is also highly common among students and therefore warrants further research.
Important methodological and contextual differences between the current and previous studies likely contribute to variability in prevalence estimates. We used broader definitions of intimate partners and family members (e.g., including cohabiting and non-cohabiting relationships) and violence/abuse (e.g., including sexual harassment as a form of sexual abuse). We also sampled from the broader population of tertiary students, including TAFE students, who reported higher rates of IPFV than university/college students. Another key consideration is that the data for this study were collected during the COVID-19 pandemic, which was associated with an increase in IPFV cases and media attention to the issue (Kourti et al., 2023). In Australia, a national survey of IPV during the initial stages of the pandemic found that 1.5% of women had experienced physical or sexual violence, and 2.8% had experienced emotional abuse or coercive control, by a cohabiting partner for the first time in the previous three months (Boxall et al., 2020). While lifetime prevalence estimates such as those from the current study will be less affected than period (e.g., 12-month) prevalence estimates, the pandemic may have nonetheless resulted in higher victimization rates than previously. Conversely, women who could not safely and privately complete the survey were discouraged from participating, which may have excluded many who were experiencing violence from cohabiting partners or family members during lockdowns, thereby leading to an underestimation of prevalence. Thus, the impacts of the COVID-19 pandemic on the findings are unclear and potentially complex.
Many interacting factors at the individual, relationship, organizational, institutional, and societal levels contribute to IPFV (Our Watch, 2021a), and further research is needed to understand those most relevant to tertiary students. In the broader Australian context, there is an emerging focus on attitudes toward violence against women as a key factor in its etiology and thus a key target for primary prevention (Powell & Webster, 2018). Attitudes are understood to shape social norms and the underlying conditions in which violence against women occurs; at the same time, attitudes are themselves shaped by broader social norms, structures, and practices and thereby reflect a society's progress in addressing the problem (Powell & Webster, 2018; Webster et al., 2018). In the most recent National Community Attitudes toward Violence Against Women Survey, 26% of Australians were classified as having a low level of understanding of violence against women, and 25% were classified as having a high level of endorsement of attitudes supportive of violence against women (Webster et al., 2018). Only a handful of Australian studies have examined tertiary students’ attitudes toward IPFV, and these have focused on frontline healthcare professions (i.e., nursing, midwifery, and paramedicine) and the influence of education and training on knowledge of IPV. These studies have generally reported poor understanding of IPV and violence-supportive attitudes in these cohorts (Beccaria et al., 2013; Doran & Hutchinson, 2017; Sawyer et al., 2017). However, there have been no such studies of knowledge and attitudes toward IPFV in the student population more broadly. As per the general population, attitudinal data could be used in conjunction with prevalence data to develop better understandings of IPFV, identify targets for prevention and intervention, and monitor the sector's progress in addressing the issue (Webster et al., 2018).
Cross-Cultural Differences in IPFV
Given our novel examination of cultural identity rather than race, ethnicity, or country of birth, the cultural differences found in this study cannot be directly compared to past research; only broad parallels can be drawn. The finding that non-Anglo-identifying students reported lower crude rates of IPV than Anglo-identifying students is broadly consistent with general population data indicating that overseas-born women report lower rates of IPV than Australian-born women (Cox, 2016). Similarly, the finding that Asian-identifying students reported lower rates of IPV than Oceanian-identifying students is broadly consistent with US data indicating that Asian women report lower rates of IPV than White women in both the general population (Breiding et al., 2008; Cho, 2012; Smith et al., 2017) and student population (Oswalt et al., 2018; Whitfield et al., 2021). Patterns of different types of violence/abuse (i.e., physical, sexual, psychological, and financial) were largely similar across cultural groups in our sample. Indeed, the lower crude rates of IPFV among non-Anglo-identifying students compared to Anglo-identifying students, and Asian-identifying students compared to Oceanian-identifying students, were consistent across all four abuse types. Only African-identifying students reported higher rates of physical abuse than the majority. This contrasts with US data indicating higher rates of physical and emotional IPV among Black, Hispanic, Native American, and multiracial students compared to White students (Oswalt et al., 2018; Whitfield et al., 2021). To the best of our knowledge, this is the first study to examine cultural differences in FV specifically. Contrary to IPV, the results suggest that non-Anglo and multicultural-identifying students face greater risk of FV than the majority.
Further research is needed to explore the many factors that may contribute to heightened risk of IPV among cultural majority students and heightened risk of FV among cultural minority students. We must be careful to avoid selective cultural interpretations that construe violence within minority communities, but not majority communities, as arising from cultural norms, values, and practices (Sokoloff & Dupont, 2005). Rather, the disproportionate rates of FV among cultural minority students in the current study must be understood in the context of multiple forms of structural and systemic oppression and discrimination including sexism, racism, and ethnocentrism, which may exacerbate abusive family relationships (Malley-Morrison & Hines, 2007; Sokoloff & Dupont, 2005). Beyond prevalence, future studies should seek to unpack cultural differences in the dynamics and complexities of IPFV in students’ lives.
Importantly, most cross-cultural differences in victimization rates were not significant in the multivariate analyses, except that Asian-identifying students remained at lower risk for IPV and sexual abuse, and African-identifying students remained at higher risk for physical abuse, than Oceanian-identifying students. This is consistent with prior research indicating that racial/ethnic differences in IPV prevalence are reduced when other sociodemographic factors are controlled (Field & Caetano, 2004; Malley-Morrison & Hines, 2007; Sokoloff & Dupont, 2005). The multivariate models in the current study included gender identity, age, area type, employment status, tertiary institution type, domestic student status, and country of birth, and the results suggest that crude cultural differences may be partially attributable to these factors. Those variables which emerged as significant multivariate predictors may be particularly relevant; for example, trans or non-binary gender identity, younger age, unemployment, and TAFE (as opposed to university) study may be more prominent risk factors for FV than minority cultural identity. In conclusion, these and previous findings suggest that culture may better conceptualized as a risk marker rather than a risk factor for IPFV (Malley-Morrison & Hines, 2007).
Limitations
Several limitations of the current study must be considered. A convenience sample was utilized, with prevalence estimates weighted to the population distribution for age and country of birth to minimize demographic bias and improve generalizability. As with any non-probability sample, the results can only be interpreted as indicative of the true population prevalence. Data were collected during the COVID-19 pandemic which, beyond potentially impacting prevalence estimates as discussed above, may have led to the underrepresentation of international students due to travel restrictions. Additionally, the survey was only available in English, which may have excluded some linguistically diverse students. There were particularly small numbers of participants in the African (n = 24) and American (n = 30) cultural identity groups, which warrants cautious interpretation of their results; for the African-identifying group, this was reflected in large confidence intervals around estimates.
We attempted to take a more contemporary approach to the analysis of culture by assessing participants’ subjective identification with a culture(s) rather than using race, ethnicity, or country of birth as a proxy. Still, culture was categorized in country-based groups, and groups were combined into broader categories to facilitate reliable statistical comparisons. It is important to reiterate that culture is not restricted by geographic boundaries, and there is both immense diversity within these broad groups and shared elements of culture between them. Future cross-cultural research should seek to design and utilize increasingly sophisticated measures of cultural identity; for example, based on participants’ degree of identification with specific aspects of a culture (e.g., cultural norms and values). A further cultural limitation of this study is that all statistical comparisons used the majority group as the reference category, and differences among minorities were not analyzed. Future studies should seek to collect sufficient data from diverse groups to permit analysis of variability between and within minorities, thus providing a more nuanced cross-cultural picture of IPFV. Finally, while we sought to analyze the robustness of cultural differences after adjusting for other sociodemographic factors, some potentially relevant variables were not assessed in the survey, such as socioeconomic status, sexuality, and disability. Future cross-cultural research should seek to account for increasingly comprehensive sets of sociodemographic and other risk factors.
Implications
The findings from this study indicate that the overwhelming majority of women tertiary students in Australia have experienced violence from intimate partners or family members in adulthood. In considering the implications for tertiary education providers, it must be acknowledged that these experiences may or may not have occurred during participants’ tertiary studies or within tertiary settings. Yet, regardless of when and where victimization has occurred, tertiary institutions are in a unique position to prevent and respond to IPFV. Indeed, tertiary education has been identified as one of five priority sectors for action to prevent violence against women in Australia, given its broad reach (including otherwise hard-to-reach audiences such as international students and recent migrants) and high potential for widespread impact on social and institutional norms, structures, and practices (Our Watch, 2021a). The sector is also well placed to support students who have experienced IPFV, including both recent and past experiences, to reduce the impacts of victimization and the risk of revictimization.
The findings from this and previous research suggest that sexual violence is but one dimension of violence in the lives of women students, and not the most pervasive (Guberman et al., 2018; Oswalt et al., 2018). Nevertheless, considerable policy and procedural work has been undertaken to address sexual violence in the tertiary education sector, and this can provide foundations and directions for addressing IPFV. The Tertiary Education Quality and Standards Agency (TEQSA, 2019) found that, by 2018, universities had largely implemented the recommendations of the AHRC's (2017) Change the Course report and were responding comprehensively to sexual assault and harassment, though TAFE and independent providers had not made such progress. For example, 93% of universities were found to have a policy on sexual assault and harassment and 93% had conducted a review of existing policies and response pathways, compared to 58% and 13% of TAFE and independent providers, respectively (TEQSA, 2019). This points to potential strengths and weaknesses in the sector's capacity to address other forms of gendered violence. As TAFE students reported higher rates of IPFV than university/college students in the current study, it may be important that TAFEs are provided additional support to respond to the issue.
While many tertiary institutions also have policies and procedures in place for IPFV, the focus of reform and guidance from peak bodies has been sexual assault and harassment. It would therefore be prudent for institutions to consider how current best practice guidelines for policy responses to sexual violence could be applied or adapted to IPFV. In Australia, these include Universities Australia's (2018) Guidelines for University Responses to Sexual Assault and Sexual Harassment and TEQSA's (2020) Good Practice Note: Preventing and Responding to Sexual Assault and Sexual Harassment in the Australian Higher Education Sector. Per the recommendations for sexual violence, tertiary institutions should consider developing standalone policies for IPFV that are guided by the principles of compassion, support and assistance, confidentiality and privacy, cultural competence, and natural justice (TEQSA, 2020; Universities Australia, 2018). These policies should include clear definitions and behavioral examples of IPFV, and information about victim-survivor's rights, disclosure and reporting options, and support mechanisms (TEQSA, 2020; Universities Australia, 2018).
It is critical that policies and procedures regarding IPFV, as per those regarding sexual assault and harassment, sit within a broader institution-wide approach to addressing violence against women (Universities Australia, 2018). Our Watch has recently launched a set of evidence-based tools and resources to guide universities and TAFEs in promoting gender inequality and preventing gendered violence (Our Watch, 2021b, 2021c). This framework uses an intersectional and social-ecological approach to challenge the gendered drivers of violence across all levels and areas of an institution. Still, given the prevailing focus on sexual violence in the sector, efforts should be made to specifically address IPFV within broader violence prevention and intervention initiatives, and awareness-raising may be necessary prior to change-oriented strategies (Our Watch, 2021b, 2021c). Comprehensive IPFV education and information materials should be readily accessible to students and regularly updated (TEQSA, 2020). These should address the broad spectrum of violence in intimate and family relationships, particularly less well-recognized forms such as psychological and financial abuse. Tertiary institutions should also seek to establish and strengthen IPFV response systems, support services, and referral pathways, with the understanding that awareness and education initiatives may lead to increased disclosure and help-seeking (Our Watch, 2021b, 2021c). A range of student services may be involved in responding to IPFV including health, counselling, academic, security, housing, and financial services. Staff in these services should therefore be equipped to respond to disclosures of IPFV in a trauma-informed way and to provide appropriate support and referrals.
While cross-cultural differences in IPFV prevalence appear to diminish after adjusting for other sociodemographic factors, this does not eliminate the need to address higher rates of victimization in some cultural groups. For example, non-Anglo and multicultural students may benefit from targeted prevention and intervention strategies for violence from family members. Per current best practice guidelines, such strategies should take an intersectional approach that considers the diversity of victim-survivors’ experiences and actively addresses the multiple systems of oppression and discrimination that drive violence against minority women (Our Watch, 2021a). Continued efforts are needed to improve the cultural competence and responsiveness of IPFV support services for students. Service providers should be equipped to assess the intersectional challenges and needs of cultural minority students and to provide referrals to culturally specific IPFV services as necessary.
Further research is needed to understand the factors that underpin cross-cultural variability in IPFV among tertiary students. A variety of statistical methods can be used to assess the influence of cultural identity on IPFV in the context of other relevant factors, including controlling for such other variables when analyzing cultural differences, examining associations and interaction effects between culture and other variables, and analyzing whether other variables mediate or moderate associations between culture and IPFV. Thoroughly interrogating and contextualizing cultural differences will not only protect against essentialist and pathologizing interpretations, but may facilitate more nuanced understandings of the role of culture in IPFV and provide more meaningful targets for prevention and intervention.
Footnotes
Acknowledgments
We would like to sincerely thank each participant in this study for their time and willingness to share their personal experiences of violence. We would also like to thank the two anonymous reviewers for their insightful and constructive feedback on the manuscript.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Laura Zark is supported by the Lina Ricciardelli Memorial Scholarship from the School of Psychology at Deakin University.
Author Biographies
).
