First,Do No Harm: From Diversity and Inclusion to Equity and Anti-racism in Interpersonal Violence Research and Scholarship

Journal of Interpersonal Violence (JIV) Diversity Statement

The JIV diversity statement, established in 2016 calls on IPV researchers to address diversity with regard to the “implications of the study questions, underlying research literature, methodology, analysis, and results.” Bent-Goodley (this issue) addresses each of these areas, outlining questions which researchers can consider at each stage of their research and offering examples to illustrate the value of cultural relevance. Bent-Goodley focuses on addressing diversity concerns in relation to race and ethnicity in the conceptualization, implementation, and interpretation of IPV research. In particular, she guides researchers through the framing of research questions, underscores the importance of a review of the literature from a cultural worldview, and walks researchers through measurement issues with particular attention to cultural validity of measures and data collection mechanisms. Her paper also supports careful consideration of sampling, including outreach and participant recruitment. She emphasizes participatory methods and offers examples of Cultural Adaptation, Community-Based Participatory Research (CBPR), and Community-Centered Evidence-Based Practice models. Bent-Goodley’s paper helps researchers to approach their work with cultural humility, emphasizing the importance of community representation and input from the community in interpreting findings. The present article aims to complement that content, avoid redundancy, and expand the range and types of factors researchers might consider in their research and scholarship to strengthen the cultural responsiveness and inclusivity of their work. The present article describes and explains factors for researchers to consider in the areas of researcher perspective and positionality, conceptualization, and measurement, discusses the research process with a focus on diversity, equity, and inclusion, and outlines factors reviewers might have in mind as they review manuscripts and assess the extent to which they address diversity and inclusion.

IPV takes many forms. Research and scholarship on IPV may focus on intimate partner/domestic violence (ip/dv), sexual violence, child abuse, youth violence, dating violence, elder abuse, and other forms and contexts of violence. This article aims to help researchers to better understand steps they might take to respond to diversity and inclusion concerns relevant to research in these areas of study. Understanding of diversity and inclusion in research is continually evolving. JIV’s diversity statement describes diversity as including “all aspects of human differences such as socioeconomic status, race, ethnicity, language, nationality, sex, gender identity, sexual orientation, religion, geography, ability, age, and culture.” It is beyond the scope of this article to describe inclusive and anti-oppressive research in every one of these areas. Instead, this article focuses on diversity with regard to race and ethnicity to illustrate how this might be approached, and touches upon each of these other forms of human diversity, as well as others. This article offers a model for considering diversity, equity, and inclusion in relation to all forms of social identity and JIV’s broad diversity statement invites researchers to apply this model to diversity in its many and varied forms.

Framing the Research Question

In research with racial, ethnic, or other marginalized communities, involving community members may help yield more relevant and useful research questions and build engagement with the research; participatory processes can shape actionable research questions to meet community needs and engender community buy-in in the research process and implementation. Bent-Goodley (this issue) describes a range of factors to consider with regard to the research question stage of research and offers recommendations to researchers for community-engaged research, emphasizing shared power, collaborative decision making at every stage, building upon community assets, seeking buy-in, and involving community members in interpreting and understanding the data. General research methods textbooks have begun to regularly address strategies for increasing the cultural relevance of studies, though they may stop short of identifying equity and anti-racism as the goal (Rubin & Babbie, 2017). Texts emphasize the need to develop research questions in consultation with community members, colleagues, and experts, including individuals with lived experience (Farmer & Farmer, 2020; Rubin & Babbie, 2017). For example, Strolin-Goltzman et al. (2010) engaged youth with lived experience in the foster care system to explore the topic of child welfare workforce turnover and retention. In focus groups, the youth shared their experiences, including the perception that the more caseworker turnovers they had, the more placements they experienced. The researchers examined this hypothesis within the department, finding a statistically significant positive relationship. The researchers urge continued engagement of youth in future research into the relationship between workforce retention and other indicators of child and youth well-being, such as attachment, educational outcomes, or permanency. Strolin-Goltzman et al. (2010) conclude, “child welfare agencies and partnering researchers should be challenged to harness the knowledge and experience of foster care consumers by collaborating with them on the development of interventions and innovations that aim to decrease turnover and ultimately seek to revolutionize systems of child welfare care” (p. 52). Youth have called for such participatory processes, echoing the self-determination call that emerged from the international disability rights movement: “Nothing about us without us” (Charlton, 2000; Day et al., 2012; Mockingbird Society, 2019). Engaging individuals with lived experience in developing research questions can increase the relevance of the research and advance research as action, although in doing so, researchers also need to critically consider whose voice is included and who is excluded.

Generalizability, Sampling, and Recruitment

Much research seeks knowledge that has “external validity,” which is generalizable to other people, places, and times. Research that seeks to inform broad public policy will likely aim for diverse representation to ensure that policy recommendations are applicable to a broad population. To be generalizable to a given population, the sample should be representative of that population on key social identity dimensions. However, not all research seeks generalizable knowledge. For example, qualitative interviews may aim to for deep understanding, may seek to give voice to a little heard perspective, valued not for its generalizability, but rather for its uniqueness, genuineness, and profoundness. As Farmer and Farmer (2020) describe it, “Authenticity is the degree to which the researcher fairly and completely shows a range of the lived experiences or realities of the participants” (e-book, location 3493).

Although some studies, especially large quantitative ones, have broad generalizability as a goal, other studies may aim for depth of understanding of within group experiences, or may seek to generalize to that subgroup, particularly when their aim is to understand experiences to inform prevention and intervention services to targeted groups. In the area of elder abuse, for example, Dong (2012) argues the need for more studies that enhance our understanding of cultural diversity and elder abuse. In particular, culturally relevant research could help examine barriers to reporting and help-seeking, prevalence rates, risk, and protective factors, as well as cultural norms, expectations, perceptions, determinants, and the impact of elder abuse within different racial and ethnic communities. Lee et al. (2014) conducted focus groups to explore and compare sociocultural and contextual factors related to elder abuse among Chinese and Korean immigrants. Their study found elder maltreatment to be a significant problem in these communities, despite the shared cultural value of venerating elders. They found sociocultural similarities, yet also differences in values and beliefs related to help-seeking. For example, the Korean participants reported greater spousal elder abuse, and voiced reluctance to disclose elder abuse in order to maintain the perception of family harmony and avoid public shame. Both cultural groups expressed some degree of belief in fate and endurance and a reluctance to complain about abuse, which posed a barrier to help-seeking. Although the study participants were not a representative sample, this qualitative research offered some insight into these specific cultural groups to help guide community-based interventions, with the limitation that the findings are generalizable only to those with social identities and experiences represented in the study sample.

Any given study should justify its target sample given its research question and the purpose of the research. The sample ought to be appropriate to the goal of the study, such as whether the aim is generalizability or specificity. Not all studies need to have samples that are diverse with regard to race or ethnicity, but having a diverse sample has benefits such as allowing for disaggregation by race/ethnicity and comparative analysis of any differences or disparities. Some studies may focus on a single racial, ethnic, or cultural group—in these cases, the study should offer a rationale for this focus, such as the goal of informing outreach efforts or the design of targeted interventions. For example, Yoshihama et al. (2014) investigated the relationship between enculturation, attitudes supporting intimate partner violence, and gender role beliefs among Gujaratis in the United States. They offer the following justification regarding their sample and comment on the implications for generalizability: “The strength of this study is that it focused on a previously under-studied population group and avoided a commonly done aggregation (e.g., studying Asian, South Asians, or Asian Indians) by studying a single ethnic group, Gujaratis. Yet, it is this study design of focusing on a single ethnic group that limits the generalizability of the study findings” (p. 258).

It is important that studies are clear about who is represented in the sample and who is excluded. To which populations, places, and times are the findings generalizable? Bent-Goodley discusses sampling in some detail, including describing strategies for recruiting diverse samples. For hard-to-reach groups, researchers should consider the barriers to recruitment of those who might be otherwise excluded. For example, it may be important to do targeted outreach, respond to any concerns, address transportation and childcare obstacles, offer financial or other incentives, and to include personalized touches (Rubin & Babbie, 2017).

When writing up the study, in addition to transparency about characteristics of the sample, researchers should describe the target population for the study and how representative the sample is of the population of interest. The representativeness of the sample may depend on recruitment methods, retention efforts, and data collection methods. In a study of Latinx participants and their experiences in protection order court, consider who is represented in the sample and who is not. Are English-language learners (ELL) under-represented? How might this affect the data and findings? Where population parameters are known and sample characteristics have been measured, the representativeness of the sample can be reported on, such as reporting the rate of ELL among Latinx populations in the study’s region compared to the sample. Researchers should report on the potential for sample bias on any characteristics of relevance to the study.

Research textbooks may discuss sample recruitment processes, including strategies to reach culturally diverse participants. Strategies include hiring bilingual and bicultural interviewers, training of interviewers, careful consideration of the location of interviews, and compensation considerations (Farmer & Farmer, 2020; Rubin & Babbie, 2017). In quantitative research, over-sampling participants who are more likely to represent a particular social identity of interest can help build a large enough sample for statistical analysis when a group is not represented at a high enough rate in the population. In nonprobability sampling, techniques such as stratified purposive sampling can be helpful to reach particular social identity groups (Farmer & Farmer, 2020). Outreach to marginalized, over-researched, and mistreated communities calls for participatory methods, such as CBPR as discussed below, in which affected communities are partners in the research and the researchers ensure that the research is designed to benefit the participants (Farmer & Farmer, 2020; Smith, 2012). It may help to advance inclusive research practices if researchers would routinely describe the processes by which they sought to recruit diverse participants, their successes and missteps, how they reached communities and engendered trust, and whether their final sample actually reflects the community of interest.

Ethical Considerations

The history and legacy of egregious treatment of racial and ethnic minority and indigenous peoples is well documented (Datta, 2018; Farmer & Farmer, 2020; Gamble, 1997; Smith, 2012; Zuberi & Bonilla-Silva, 2008). Researchers aiming to conduct research involving BIPOC and other marginalized groups must attend to ethical imperatives and earn people’s confidence to overcome a legacy of abuse of power by researchers. As Deloria (1991) has said, researchers “bear the burden of past mistakes” of other researchers (p. 467). In “Decolonizing Methodologies,” Smith (2012) details historical and current day experiences of indigenous peoples with research. She offers the following questions to consider about potential research projects: “Whose research is it? Who owns it? Whose interests does it serve? Who will benefit from it? Who has designed its questions and framed its scope? Who will carry it out? Who will write it up? How will its results be disseminated?” (pp. 43–44). Researchers would do well to ask these questions of themselves before approaching communities.

Ethical considerations in IPV research with diverse groups take many forms. Safety and confidentiality are foremost in ip/dv research. For example, surveys on gender-based violence that are conducted in the home require safety protocols, such as interviewing women in private, perhaps framing the survey as a health survey. Informed consent language and processes should be culturally appropriate (Farmer & Farmer, 2020) and should explain the precise nature of the research and procedures (WHO Report, 2001). A 2001 World Health Organization report describing international surveys recommends that “violence questions should only be incorporated into surveys designed for other purposes when ethical and methodological requirements can be met” (p. 11). Interviewers trained to refer victims for help and support must be informed about referral information relevant to different racial and ethnic groups and relevant to other marginalized identities represented in the sample, such as participants with vulnerable immigrant status, who have disabilities, or who are sexual or gender minorities. Researchers conducting international surveys of violence against women must be careful to minimize underreporting, as the implications of undercounts are significant, not only for perceptions of the extent of the problem, but also for funding allocations (WHO Report, 2001). Training native language interviewers or contracted interpreters on confidentiality is also critical, especially when studying sensitive topics and when interviewers and participants are from the same community.

Ethical considerations also guide methodological approaches in IPV research. For example, because of participant safety issues and the potential for adverse events including the risk of emotional distress in participants, gender-based violence research rarely uses experimental design, randomized clinical trial procedures, or survey methods that involve longitudinal follow-up of survivors. Cross-sectional designs in which data are collected from survivors at one point in time without follow-up are safer and less risky. In a survey of intimate partner/domestic violence researchers to explore their use of various human subject protections, Hellmuth and Leonard (2013) reported that the most commonly utilized participant protections were anonymity; in person or telephone interviews; preventing participants’ partners from being aware of their participation or of their answers; certificates of confidentiality; using same-sex interviewers; training research staff in safety planning; and providing referral lists of community resources (Hellmuth & Leonard, 2013, p. 492). Because of the potential risks to participant safety and well-being in gender-based violence research, qualitative methods such as in-depth interviews are apt research designs which give voice to individuals with lived experience, enable participants to share from the perspective of their intersectional social identity, with skilled interviewers trained to engage participants to collect thick data and support a positive interview experience. Further, for research with marginalized racial and ethnic groups, qualitative approaches may be consistent with the epistemology and practices of the community. As Kanuha (2000) noted, “…the use of narratives, life histories, and intensive interviews is related to the reliance of many marginalized populations on oral communication, cooperative and mutual relationships, and narrative traditions that may be antithetical to many classical research methods” (p. 442).

Conceptualizing and measuring race and other social identities

Although race is a “biological fiction, it is nonetheless a social fact” (James, 2008, p. 38). JIV’s diversity statement describes diversity as including identities such as race, sex, gender identity, sexual orientation, and ability. From a constructionist perspective, each of these social identities is a social construct—there is no universal understanding of what they represent, therefore researchers need to be clear about how they have conceptualized and operationalized any social identity constructs of relevance to their research. In IPV and gender-based violence research and scholarship, race is especially relevant because of racism and the politicization of race in the United States. Structural and institutional racism borne of colonialism and White supremacy culture has led to race and ethnicity being associated with particular risk and protective factors, and with differential response on the part of the criminal and juvenile legal systems, child welfare system, and health and social service sectors. The lived experiences of IPV vary by race and ethnicity and other social identities, and it is important that interventions be informed by contextualized understandings of the experiences of diverse populations and that social responses address structural change in policy and practice.

In social research, race may be alternatively used as a grouping variable, control variable, dynamic process, or as a moderator (Rodgers-Farmer & Potocky-Tripodi, 2001). When race is a variable in research, how is it conceptualized and operationalized? How is race used in the study? In analyses? Is variation within race or ethnicity considered? There is no accepted convention for how race and ethnicity should be measured. It is most important to be clear about how and why the study used a given conceptualization and operationalization. To illustrate the extent of differences in the measurement of race, in the United States, the child welfare reporting data system AFCARS (Adoption and Foster Care Analysis and Reporting System) measures children’s race using the following categories: Black or African American; White; AI or AN; Asian; Native Hawaiian or Other Pacific Islander; and Unable to Determine. Ethnicity is measured as “Hispanic or Latino ethnicity,” separate from race. In contrast, in New Zealand, the Oranga Tamariki—Ministry for Children reports children’s race in the following categories: Maori; Maori and Pacific; Pacific; and New Zealand European and Other (which includes New Zealand European, European, Asian, Middle Eastern, Latin American, African, and other). These two very dissimilar measures of race reflect how two countries conceptualize race based on their particular population demographics, histories, geographic location, and social contexts.

If a study does not measure or report race/ethnicity, it should note why. Studies that do not include exhaustive categories of race and ethnicity should explain how they have conceptualized the construct and explain any categories they have excluded. Similar questions might be asked of any social identity characteristics of relevance to the study. For example, how is gender conceptualized and operationalized? With evolving conceptualizations of gender and how it is understood and with increasing recognition of gender fluidity, binary measures of “male” and “female” are not exhaustive and this should be addressed or identified as a study limitation.

Measurement in cultural context

Measurement issues are key to valid understandings and to making meaningful comparisons across groups. A sizeable literature details issues and techniques for assessing the conceptual equivalence of instruments in cross-cultural research. For example, Thanh et al. (2018) offer a step-by-step guide to measurement in cross-cultural research, whether adopting an instrument, adapting it, or developing a new instrument. They underscore the importance of having cultural experts on the team, to offer knowledge of language, history, and customs. When translating instruments, it is important to aim for conceptual equivalence, not just linguistic equivalence (Thanh et al., 2018). “Linguistic equivalence is the extent to which a measure is compatible with the language and culture to which it will be adapted” (Farmer & Farmer, 2020, e-book location 2268/8760). Going beyond language, “conceptual equivalence is the extent to which the concepts represented by the items making up the measure are understood to have the same meaning across cultural groups” (Farmer & Farmer, 2020, e-book location 2268/8760). For example, the concept of depression is complex and understood differently across cultures as people present with differing symptoms, which must be considered in sociocultural context (Chentsova-Dutton et al., 2014). Measures of depression based on the conceptualization in the American Psychiatric Association’s Diagnostic and Statistical Manual of Mental Disorders (DSM) emphasize intrapersonal dimensions of depression, whereas in some cultures, the concept of depression is understood as more interpersonal; researchers should consider measures that align with the construct as it is understood in the cultural context (Chentsova-Dutton et al., 2014). Best practice for achieving conceptual equivalence is through consultation with cultural experts or an advisory committee, and through the use of translation and back-translation of specific measures (Rubin & Babbie, 2017; Thanh et al., 2018). The metric or scalar equivalence is also important, as that assesses whether the measure’s response options are understood in the same way across groups (Farmer & Farmer, 2020). For cross-cultural or cross-national intervention studies, it is important to consider conceptual equivalence in the target problem, treatment elements, and in outcome measures across the samples (Thanh et al., 2018). To assess the psychometric properties of culturally developed or adapted instruments, appropriate analyses include computing reliability indices, item intercorrelations, and assessment of construct validity such as through factor analysis (Rodgers-Farmer & Potocky-Tripodi, 2001).

In ip/dv research, the Revised Conflict Tactics Scale (CTS-2) (Straus et al., 1996) is widely used and measures physical and psychological assault and sexual coercion within intimate partner relationships. The CTS-2 has been translated into over 15 languages, used in over 20 countries (Chapman & Gillespie, 2019) and is reported as being reliable and valid across different populations and cultures. Gender-based violence researchers may augment the CTS-2 with other assessment tools to address the limitations of the measure. For example, the CTS-2 is a behavioral measure which does not focus on the situational context of the violence, experiences of coercion and control, or assess relationship dynamics over time (Lindhorst & Tajima, 2008). The Women’s Experiences with Battering scale (WEB) is a measure that is widely used in gender-based violence research and which focuses on assessing coercive control and the meaning of the abuse, including psychological and emotional victimization and fear. The WEB is a 10-item scale that has been normed among women. The Composite Abuse scale—Revised, Short Form (CAS-R SF) is a brief, 15-item instrument that has been tested reliable and valid among a sample of women in Canada representing diverse races, including indigenous participants (Ford-Gilboe et al., 2016). The authors urge additional norming in other countries and with other identity groups. In particular, the CAS-R-SF has been normed among cisgender women, therefore additional psychometric testing is needed with people of all genders, particularly trans and nonbinary populations, who “face unique societal, political and historical challenges that affect their experiences of violence within intimate relationships” (Ford-Gilboe et al., 2016, p. 10).

Culturally developed or adapted instruments

Appropriate consideration of diverse populations in IPV research requires tools that are reliable, valid, and culturally appropriate. The intimate partner/domestic violence field offers valuable examples of tools that have been developed or adapted for particular groups. Haddad et al. (2011) validated an Arabic version of an instrument to measure ip/dv among Arabic and Middle Eastern women and reported on prevalence and patterns of abuse in that cultural context. Specifically, the researchers developed an Arabic version of the NorVold Abuse Questionnaire (NorAQ), a tool originally developed by the Nordic research network NorVold and normed with Scandinavian samples, which measures gender-based emotional, physical, and sexual abuse, and abusive experiences in the health care system (Swahnberg & Wijma, 2003). Translated into Arabic, with back-translation, Haddad et al. assessed the scale’s content validity, construct validity (using factor analysis), and reported reliability (via Cronbach’s α), offering evidence of both reliability and validity with their sample (Haddad et al., 2011). The adapted version of the NorAQ for Arabic women excluded those items from the original instrument which measure the construct “Abuse in the health care system” and focused exclusively on abuse in relationships. This study of the psychometric properties of the NorAQ with Jordanian Arabic women represents the first validated tool with this population. In addition to items measuring mild, moderate, and severe emotional, physical, and sexual abuse, the tool includes questions about the perpetrator. In the Jordanian cultural context, the researchers found that fully 30% of the reported abuse was perpetrated by the women’s brothers before marriage (Haddad et al., 2011). Studies of domestic violence often focus exclusively on violence by intimate partners—cross cultural research reminds us that to understand women’s experience of domestic violence and abuse, we may need to expand measurement to consider other perpetrators such as brothers (Haddad et al., 2011) or mothers-in-law (Yoshihama, 2005). Intimate partner violence measurement tools may need to be adapted in order to be accurate (valid) measures of the construct of domestic violence in that cultural context.

Another example of a culturally adapted instrument is Messing et al’s (2013) adapted danger assessment (DA) instrument (developed by Campbell) for immigrant women. The DA-I is a 26-item scale, designed for use in safety planning with women with immigrant status. Immigrant women are more vulnerable to ip/dv and intimate partner femicide for several reasons, including because they are often socially isolated, financially dependent, without work visas, and other factors (Messing et al., 2013). The DA was adapted to include new items, including some from the Women’s Experience of Battering (WEB, Smith et al., 1995). Examples of the new items are as follows: “Partner is not foreign born,” “Prevents you from going to school, getting job training, or learning English,” “Threatens to report you to child protective services, immigration, or other authorities,” “Victim is not employed” and from the WEB: “I hide the truth from others because I am afraid.” The researchers found that the adapted “DA-I predicts any and severe intimate partner violence significantly better than the original DA” (Messing et al., 2013, p. 270). Notably, several items in the original DA did not predict risk among immigrant women participants: “partner owns a firearm,” “partner uses drugs,” and “partner is unemployed,” illustrating the importance of adapting and testing instruments with target sample groups so that only predictive items are assessed.

In a study of gender-based violence among an Asian sample of women, Yoshihama (2001) found that adding even just three socioculturally relevant items augmented disclosure of physical violence and sexual violence by approximately 5% and 10%, respectively. The items added are as follows: “Had dining table overturned” and “doused with liquid” (added to the physical violence items) and “forced sexual intercourse when concerned with people nearby” (added to sexual abuse items). Using culturally relevant instruments not only increases validity, it also means that the tool has allowed women to “voice their concerns about how violated they feel within a cultural framework that is meaningful to them” (Sokoloff & Dupont, 2005, p. 42).

Contextualization

Bent-Goodley notes the need to contextualize measures of violence so they have cultural relevance and cultural validity. Similarly, Lindhorst and Tajima (2008) argue for greater contextualization of quantitative measures of intimate partner/domestic violence against women, to go beyond the measurement of behavioral indicators of violence to measure contextual dimensions such as the cultural context, the context of systemic oppression, and of historical trauma. “Culture matters to the conceptualization of intimate partner violence in that it plays a role in defining what behaviors are deemed abusive, in what particular situational contexts, and it shapes the personal meaning which victims attach to their experiences” (Lindhorst & Tajima, 2008, p. 374). Elder abuse too takes many forms; studies have delineated a range physical, psychological, financial, sexual, and neglect dimensions. However, there may be additional culturally specific dimensions, or culturally specific forms of abuse. For example, in a workshop on elder abuse and prevention, one researcher noted that some Native American elders talked of spiritual abuse such as being excluded from ceremonies or being denied the support of a traditional healer (National Center for Biotechnical Information [NCBI] Workshop proceedings, 2014). Another researcher reported that, among Chinese elders, psychological abuse, including disrespect and ignoring was perceived as more severe than physical abuse. The workshop attendees concluded that “conceptualizations and definitions of elder abuse need to be culturally relevant” (NCBI proceedings, 2014). Contextualizing measures in IPV research means careful consideration of the cultural context in selecting, developing, or adapting measures. Measures of IPV victimization should also consider cultural perceptions of the victimization experience, including attention to the perceived severity and impact of the experience (Yoshihama, 2005).

International and cross-national research

All researchers should be familiar with the historical context of the country in which they are doing research, its experience with colonization, and how racism, sexism, and classism are manifested. For researchers doing research in international settings or involving cross-national comparisons, planning should involve partnership with in-country researchers and community partners. For researchers who are nationals, consideration of diversity and inclusion may mean reaching out to indigenous and underrepresented researchers and communities. In global or cross-national research, researchers and reviewers should not perpetuate White supremacy thinking by expecting a Western or White comparison group and should offer a rationale to justify comparison groups across different countries. Interventions developed for Western groups may not be effective in other cultural contexts (Shokane et al., 2018). For example, scholars have underscored that research conducted in countries in Africa should approach from an Afrocentric worldview and adopt Afrocentric research methods (Asante, 1987; Mazama, 2003). Given the diversity within the multicultural African continent, researchers should consider intervention strategies appropriate to the cultural context of particular subregions and consider whether the cultural assumptions underlying a proposed intervention align with the specific cultural context.

Acculturation and Enculturation

Acculturation and enculturation are constructs that may be included in IPV research with racial and ethnic groups and communities. Acculturation is understood to be a multidimensional process of behavioral and cognitive adaptation occurring through exposure to different cultural groups, particularly racial and ethnic groups exposed to numerical majority cultures (Cote & Bornstein, 2003; Phinney & Flores, 2002; Tajima & Harachi, 2010). Enculturation describes a process by which an individual is socialized into their own cultural heritage through participation in and through the retaining or relearning of practices and beliefs (Kim et al., 2009; Wolsko et al., 2007; Zimmerman et al., 1996). Measures of acculturation are especially relevant to research involving immigrant and refugee communities; measures of enculturation are relevant to minoritized racial and ethnic groups and indigenous tribes and nations (Zimmerman et al., 1996). Assessment of acculturation or enculturation is appropriate when the study topic may be theoretically associated with cultural attachment and participation.

Although acculturation is sometimes conceptualized as a single continuum, evidence indicates that it is better understood as orthogonal (LaFromboise et al., 1993; Phinney & Flores, 2002). That is, acculturation to a different culture is not necessarily associated with reduced attachment or enculturation within one’s culture of origin. People may be bi- or multi-cultural. Researchers may elect to include both acculturation and enculturation scales (Do et al., 2019; Kim et al., 2009; Yoshihama et al., 2014) to assess the cultural values, beliefs, behaviors, and community participation.

In a study of cultural identity, stress, coping and psychological well-being among Yup’ik communities in the Yukon-Kuskokwim Delta region, Wolsko et al. (2007) utilized measures of acculturation and enculturation. The researchers found that among study participants, enculturation to the traditional Yup’ik way of life was associated with significant mental and behavioral health benefits. In contrast, acculturation to the (White) majority culture was associated with increased stress and increased substance use as a coping mechanism, offering evidence of the “positive health benefits of greater self-determination and of affirming the intrinsic worth of indigenous worldviews” (p. 52).

In IPV research involving immigrant groups, thought should be given to whether and how to measure these constructs, and whether acculturation or enculturation might mediate or moderate outcomes or be otherwise relevant to the research. Research with immigrants and refugees should also consider accounting for their migration experience and the social reception they encountered upon entry to the host culture, as these can vary significantly between and within groups (Tajima & Harachi, 2010). In addition to measuring these aspects of acculturation, enculturation, and social reception, IPV studies involving immigrant households might also consider the possible role of factors such as acculturation conflict, which may occur between parents and second-generation children, or the protective dimension of enculturation. An example of the latter, Do et al. (2019) examined peer victimization and the perpetual foreigner stereotype among Sikh American youth, with a focus on the role of behavioral enculturation. Their study reported protective mental health effects among male adolescents who maintained Sikh cultural and religious practices such as speaking Punjabi or attending Sikh activities. Specifically, they found that greater behavioral enculturation was predictive of lower depression in Sikh American boys (Do et al., 2019). The study also reported that behavioral enculturation moderated the relationship between feeling like a perpetual foreigner and symptoms of depression. Specifically, the perpetual foreigner stereotype related to higher depressive symptoms for Sikh boys who reported less behavioral enculturation, suggesting that feeling socially excluded may be experienced more deeply by those youth who regard themselves as more assimilated to U.S. culture.

Methodology and Participatory Research Processes

Bent-Goodley has described participatory methods and offered examples of each, so I will not add great detail here but I will underscore that attending to diversity, equity, and inclusion calls for inclusive research processes, for “working alongside and in relationship” (Hess, 2018, p. 586) and for critical thought as to who is included and excluded. CBPR is a collaborative research process appropriate to both qualitative and quantitative studies that engages community partners in conceptualizing the research question, designing the study, and analyzing, interpreting, and disseminating findings. As Spencer (2015) notes, CBPR “focuses on social, structural, and/or environmental inequalities through active involvement of community members, organizational representatives, and providers in all aspects of the research process” (p. 45). Central to CBPR is the recognition of community strengths, a commitment to co-learning, community capacity building, power sharing and to long-term commitment (Israel et al., 2012).

Informed by Lorde’s oft-cited “the master’s tools will never dismantle the master’s house” (1984), researchers have pointed to the colonialist bedrock of scientific research methods and call for new methodologies and ways of approaching our work that challenge notions of expertise and scientific objectivity (Ziai et al., 2020). Datta (2018), an indigenous researcher, points to the settler colonial roots of his own research training and describes his personal process of unlearning and re-learning decolonizing methodologies. From his perspective, “decolonization is a continuous process of anti-colonial struggle that honors indigenous approaches to knowing the world, recognizing indigenous land, indigenous peoples, and indigenous sovereignty—including sovereignty over the decolonization process” (p. 2). In her book, “Decolonizing Methodologies,” Smith (2012) notes, “in all community approaches, process—that is methodology and method—is highly important. In many projects the process is far more important than the outcome. Processes are expected to be respectful, to enable people, to heal and to educate. They are expected to lead one step further toward self-determination” (pp. 218–219).

For many researchers trained in positivist and nonparticipatory processes, centering diversity, inclusion, equity, and anti-racism will stretch them, and they may need to unlearn and relearn. I hold though that this offers researchers the opportunity to consider who they are as researchers and who they want to be. In IPV research, to support anti-racist and anti-oppressive approaches to IPV prevention and intervention, researchers might pursue community-driven and community-based research from a restorative justice approach rather than research that further reinforces a punitive or retributive approach. DeCuir-Gunby et al. (2019) call for dismantling and reappropriating methods as scholar-activists who ask the question, “How can we help?” (p. 224).

Secondary analyses. Researchers conducting secondary analysis of existing data should interrogate whether systematic bias has excluded certain groups. Which groups are not represented or are underrepresented in the data? Has systematic bias led to BIPOC people being overrepresented in the data? For example, with data from the criminal legal system in the United States, does the article acknowledge the institutionalized racism and systemic bias that leads to BIPOC being overrepresented in official data? At every stage, from police interrogation and arrest, through prosecution, sentencing, and incarceration, BIPOC communities are overrepresented in officially reported law enforcement data.

Just as participatory processes and methods such as CBPR can enhance the relevance of and community investment in primary research, collaborative processes may inform and strengthen secondary analysis of existing data. Collaborating with affected groups and communities as consultants and research partners can ensure that the “right” research questions are asked, that is, that the research yields useful and actionable findings. Further, partnership with communities and those with lived experience can help assess the appropriateness of existing measures of key constructs and help identify the limitations of the data. As Bent-Goodley (this issue) has pointed out, community involvement in interpreting the results of secondary analysis is important to obtaining accurate understandings.

Systematic reviews. In the case of systematic reviews, researchers do not have control over the original studies. There are nonetheless diversity and inclusion concerns to consider in the systematic review process and focus. For example, in their discussions of diversity and inclusion, authors of systematic reviews should describe efforts to locate, access, and include studies from non-Western countries and cultures. In developing search terms, it is important to capture different spelling conventions, such as American English as well as English used in most other countries (e.g., color and colour, self-defense and self-defence, or pediatric and paediatric). Consideration of terms used in different countries will also help avoid biasing searches toward certain countries.

Because the demographic composition of the sample relates to the representativeness of the findings for different populations, systematic reviews should report relevant sample characteristics such as race/ethnicity. If a reviewed study did not report the racial and ethnic characteristics of their sample, it would be appropriate to note this in the systematic review. This might be noted as a limitation of the individual study and, if a common omission across multiple studies, pointed to as a gap in the literature, which should be addressed in future research. Systematic reviews might help advance diversity, inclusion, and equity in research and scholarship if they would regularly describe and analyze the extent to which the reviewed studies attended to these factors in their sampling, measures, and analyses.

Meta-analysis is a quantitative form of systematic review designed to synthesize research findings across multiple sources (Wolf, 1986). Meta-analyses aggregate findings across individual studies and typically calculate an overall effect size. Such aggregated analyses may lead to erroneous conclusions regarding effects for particular demographic groups if they were underrepresented in the original studies reviewed, or if small effect sizes for those groups were aggregated with large effect sizes for other demographic groups. Researchers conducting meta-analyses could address this by calculating overall effect sizes for population subgroups, recognizing that doing so limits the total number of aggregated studies and therefore the power to detect overall effects.

Analyses. In studies involving race and ethnicity as a grouping variable, analyses should be disaggregated so that groups may be described separately and differences between groups can be examined, especially prior to combining any groups (Farmer & Farmer, 2020; Rubin & Babbie, 2017). Patterns and relationships may not be identical across groups, and effects may be reduced, obscured, or eliminated when groups are combined. With disaggregation, disparities are evident and it is possible to see the structural forces at play. If racial or ethnic groups have been combined for analyses, what are the implications of combining the groups? Studies should describe the rationale for combining racial, ethnic, or other social identity categories. It ought to be clear who the analytic sample is representative of and researchers’ thinking should be transparent.

Studies should also note when any racial groupings were dropped from analysis because of low N (sample size). Ong et al. (2020) has pointed to some of the ways that data can contribute to “erasure” of AI and AN peoples, such as the way that the U.S. census has historically undercounted AI/AN groups, and the fact that in many regions of the United States, AI/AN peoples represent such small numbers that their responses are often not analyzed or are aggregated with other groups. This practice ought to be resisted and not perpetuated. In studies with N’s too small for statistical analysis, such as with some secondary analysis of existing data, this limitation should be expressly noted. Researchers might reflect on the reasons for the exclusion and nonrepresentation of particular populations or social identity groups, and not from a participant deficit perspective, but rather what engagement, sampling, and recruitment steps were missed that might have reached these groups?

In statistical analyses such as logistic regression, where categorical groups are compared to reference groups, researchers ought to give careful consideration to which groups are the comparison group, to avoid reinforcing White supremacy, patriarchy, heteronormativity, ableism, and other oppressive systems. For example, in logistic regression analyses by race/ethnicity, rather than comparing all groups to White people, consider calculating separate models with a different referent group each time—doing so will allow the study to report results comparing each racial group to each of the other groups. By disaggregating data and by including a range of factors in statistical analyses, researchers can explore associations related to race, as well as discern when race is not related, such as when factors other than race are at play (Lindhorst & Tajima, 2008; Rubin & Babbie, 2017).

Interpretation and Discussion of Findings

The discussion section and interpretation of findings is an opportunity to contextualize the study results and their significance. It is a place to highlight the diversity, inclusion, and equity dimensions of the findings and note how they advance understanding of the research topic toward policy and practice change. It is also a place to reflect on the extent to which one’s research addresses the various points introduced in this article. For studies of particular cultural groups or settings, researchers might contextualize the discussion by linking back to literature reviewed regarding the sociocultural, historical, institutional, and political context. For research with anti-racist or anti-oppression aims, to what extent does the study contribute to understanding of disparities and the impact of racism and oppression? Does the discussion include implications for social justice? Does it offer implications for reform of racist or oppressive policy? Does the interpretation of findings and discussion focus at the individual level, blaming individual deficits without acknowledging the role of larger structures? Is the role played by racism called out explicitly? In “Health Affairs” Boyd et al. (July, 2020) describe standards for publications on racial health inequities. They call for researchers to define race and to name racism and its role in health (i.e., whether the racism is systemic, institutional, individual, or internalized). They underscore that discussion of alternative explanations should be evidence-based and especially not baseless genetic attributes. In the field of IPV, research findings should be understood as situated in the context of systems of oppression, including White supremacy, patriarchy, colonialism, xenophobia, imperialism, heterosexism, ableism, ageism, classism, and economic exploitation, as these all function to perpetuate violence, delimit options, and shape the consequences of IPV for victims (Lindhorst & Tajima, 2008).

It is also important to discuss the limitations of the research with regard to the diversity, inclusion and equity points raised in this article. How has the diversity of the sample or lack of diversity affected the findings and the questions which the study is able to answer? For studies focused on specialized cultural groups, to whom are the findings generalizable? To whom are they not generalizable?

Researchers might also include discussion of the perspectives of participants and communities with regard to the study findings. Bent-Goodley emphasizes the value of consultation with communities and affected groups at the stage of interpreting the data. The multicultural validity of findings is enhanced when participants contribute their wisdom to interpreting the data (Kirkhart, 2013). In qualitative research involving diverse groups, member checking “may allow the researcher to become aware of his or her biases, enhance the quality of the interpretation of the data, and help contextualize the findings” (Farmer & Farmer, 2020, location 6368).

Language and write-up. Language choice is something to be mindful of in writing and reviewing manuscripts. In shifting from “color-blind” to “race conscious,” it is important to avoid euphemism (Hess, 2018). The term “diversity” itself may be euphemistic and rhetorical. Using clear, race-specific language can help move the literature beyond neutral, “color-blind” language that serves to perpetuate the status quo of norming White people. For example, we might do a better job of contextualizing our samples. Said differently, it is typical in IPV research to specify when the sample involves people who are not White (whether in the title or abstract) but to be “color-blind” when the sample is White or predominantly White. This reinforces White as the norm. Contextualized samples would specify the sample on key demographic characteristics. For example, rather than the title “Dating violence among teenagers,” a more contextualized title might be: “Dating violence among a predominantly White, heterosexual sample of teenagers in the U.S. Northwest.” In using race-conscious language, the goal is to decenter White as the norm and increase transparency about who is represented. Specifying other social identity characteristics such as in the example above similarly helps to shift norms on other marginalized identities, such as by sexual orientation, region, and country. Correcting Eurocentrism or White norming in this way can help interrupt the problematic worldview of “the West and the rest” (Ziai et al., 2020, p. 4, citing Hall, 1992). It is important to consider language in our papers so as not to represent participants in ways that perpetuate harm, reinforce colonialist narratives, or that oppress. Representational ethics reminds us to be mindful of how we represent participants and communities in our research and that “the act of writing—selecting, paraphrasing, summarizing—needs to be considered as part of ethics in the research process” (Haarlammert et al., 2017, p. 420).

Strengths-based. Research that aims to be culturally responsive, inclusive, and anti-racist should not approach from a deficit perspective but rather be strengths- and asset-based. It has been observed that the research tradition of “problem definition” itself may perpetuate a deficit perspective when studies involve marginalized groups, by conceptualizing of the research topic or the research participants themselves as the problem (Farmer & Farmer, 2020). Deficit perspectives may be reflected in research questions, measures, language used to describe participants, or in how results are interpreted. This distinction is illustrated in the following two research questions, on essentially the same topic: (1) “Do gang members have higher High-School dropout rates than other youth?” and (2) “Do schools support the educational success of youth who are involved in gangs?” The two studies might use the same educational data, but the wording of the first study suggests individual failings, reports graduation data as “HS dropout” rates and might conclude that observed differences represent an “achievement gap.” In contrast, the research question in the second study appears to focus on the institutional responsibility to engage and support all students, might use measures such as HS retention and graduation rates, and might draw attention to the “opportunity gap” as a structural problem. Regardless of whether the focus is individual or structural, it is important to be strength-based with regard to language, measures, and with regard to how findings are interpreted.

In social research, using “person-first” language is generally recommended, to help ensure that people are described first and foremost as people, not as a condition they may have or a category to which they may belong. Examples of person-first language are “people with disabilities” rather than “disabled people,” or “the disabled”; “children in foster care” instead of “foster children”; and “youth involved in the juvenile justice system” instead of “delinquent youth” or “delinquents.” However, I also note that perspectives and language are evolving, and particularly in relation to disabilities, some advocate for the use of identity-first language and believe it to be more affirming of disability identity and culture (Dunn & Andrews, 2015).

In the field of IPV, it is particularly important to emphasize strengths-based perspectives and approaches and not perpetuate individual-level, deficit-based, victim-blaming research and scholarship. It is critical that social, legal, and health system responses to IPV victimization support those who have been victimized from a strengths-based, asset-based, and trauma-informed perspective. These response systems need to acknowledge the responsibility of organizations and institutions to improve their policies and practices to better support healing and restoration among those who are victims as well as those who have perpetrated violence.

Guiding principles

Researchers and scholars offer a range of suggestions and guiding principles relevant to the present discussion of diversity, inclusion, equity, and anti-racism in IPV research. This article offers three exemplars. I do not expand upon them, as they are included simply to illustrate what guiding principles researchers might follow with regard to cultural responsiveness, inclusion, equity, and anti-racism.

Child Trends researchers, Andrews et al. (2019) offer a guide to incorporating a racial and equity perspective in child welfare research, and include the following guiding principles:

Examine your own biases

In “Quants and Crits” (referring to quantitative researchers and critical race theorists), Crawford et al. (2018) outline five principles for race-conscious analysis (p. 25):

Centrality of racism

As researchers consider the cultural context, Afrocentric theorists and scholars have articulated an Afrocentric paradigm from which to consider Afrocentric research methods (Asante, 1987; Mazama, 2003). As one illustration of this, Mazama (2003) has summarized the following as criteria for an Afrocentric methodology (cited in Mabvurira & Makhubele, 2018, p.16):

African experience must guide and inform all inquiry

For manuscript reviewers

JIV expects manuscript reviewers to assess the extent to which the manuscript has addressed diversity-related concerns. Reviewers may find this article helpful in their assessment in that it outlines dimensions along which they might assess the manuscript, and might help guide reviewers in offering comments to researchers to support more culturally responsive, inclusive, equitable, and anti-racist research processes and goals. Mnguni (2020) notes, “Researchers, including editors and reviewers must ask themselves: To what extent does their research embrace social justice?” (Mnguni, p. 28). All reviewers can and should grapple with this question and use their position as a reviewer to guide authors regarding the social justice implications of their research and findings. Researchers, editors, and reviewers in the field of IPV all play a role in advancing equity, anti-racism, anti-oppression, and social justice in the literature. Editors should make active efforts to recruit and include reviewers representing diverse social identities. At the same time, I note that it is not the singular responsibility of reviewers who are BIPOC or who represent other marginalized communities to give feedback and guidance on diversity and equity.

Reviewers assessing the extent to which a manuscript addresses diversity issues might think about the research and write-up specifically from a diversity, equity, and inclusion perspective, and might have a rubric in mind to assess the manuscript and from which to suggest revisions. In my graduate research class, students conduct a small-scale research project and write up their study. For this assignment, I ask them to reflect on the implications of their study for empowerment, co-learning, and/or social change and I ask them to reflect on their positionality and role as a researcher. To help assess the extent to which they have addressed these areas, I use a rubric to evaluate the extent to which they: present a thoughtful reflection on social justice and their role as a researcher; demonstrate strong awareness of positionality in relation to the study; discuss the role of the study participants and power dynamics involved; present thoughtful discussion of their role as an “insider” or “outsider”; and offer compelling ideas for further research to advance social justice in this area. Manuscript reviewers might consider assessing manuscripts on whatever dimensions of diversity, equity, and inclusion are relevant to that particular study.

Toward an Equity and Anti-racism Research Agenda in IPV

Is the research racist? Not racist? Or anti-racist? Angela Davis is quoted as saying, “In a racist society, it is not enough to be not racist, we must be anti-racist.” Ibram X. Kendi, author of “How to be an Anti-Racist” (2019) has underscored that the opposite of being racist isn’t being not racist, it’s being anti-racist. It’s active. Researchers would do well to ask themselves: Where would I situate my research? Is it racist, not racist, or anti-racist? Being not racist is arguably the minimum—Do no harm. However, the United States is at a point of reckoning with race where it is not enough to be not racist. To be simply not racist is to be complicit. Research on IPV that is simply “not racist” perpetuates the status quo.

Racism exists at the individual and institutional level and takes overt and covert forms (Stokely & Hamilton, 1967). Racist discussions may essentialize race, generalize from individuals to the population, or make assumptions about culture. A paper recently published in Society (July 21, 2020) entitled “Poverty and Culture” attributed low SES in the United States to the low aspirations and effort among Black and Latinx people, and argued that “the solution to poverty is for the poor themselves to adopt the more inner-driven individualist style.” Following petitions representing hundreds of signatories, the journal retracted the paper on July 31, 2020, with apologies, citing inadequate editorial oversight. This offers an example of contemporary social welfare scholarship that was widely regarded as racist.

Not all IPV research will have the express goal of dismantling racial inequities, but we ought to aspire to do this. The IPV field could benefit from a research agenda to advance racial equity and anti-racism in research, policy, and practice (Furtado & Banks, 2016). Anti-racist and anti-oppression research into the root causes of IPV and the structures that contribute to it can help expose and show their inner workings. Racial disproportionality in the child welfare system is well documented—less is known about what works in interrupting bias and structural racism in that system, how to address the impact of historical trauma or how to re-imagine that system. Research evaluating anti-racist initiatives and transformative practices would advance our understanding in this area; active efforts are needed to dismantle, undo, repair, and restore. Stewart and Sweetman (2018) remind us that Martin Luther King, Jr. called on social and behavioral scientists to “tell it like it is.”

Researchers might ask themselves whether their IPV research perpetuates rather than challenges militarized law enforcement or the prison industrial complex. Does it take (or even acknowledge) an institutional approach? Does it gather evidence regarding disproportionality, bias, and racism in institutions and systems such as the criminal legal system, police, courts, child welfare, and in immigration policy? For example, anti-racist research in the IPV field might aim to advance our understanding of the ethics of machine learning, its disproportionate burden on BIPOC, and its significance to racial equity. Algorithms in the criminal legal system such as COMPAS for bail and parole contribute to disproportionate incarceration (Huq, 2018). Although some may view algorithms as less biased than humans, they are programmed by humans and used within structurally racist institutions. If we are serious about advancing an anti-racist research agenda in IPV, we will also need to tackle the question of funding: Who will fund anti-racist research? Who will fund community-based studies that actively challenge institutions of the state that perpetuate those durable inequalities across race, gender, and class? (Tilly, 1998).

Even those studies that are anti-racist in their purpose or that challenge existing inequitable institutions should also describe their efforts toward diversity and inclusion in their research processes. For instance, a study of the effectiveness of a community-based diversion program implementing a restorative justice intervention with Latinx youth involved in gangs would still be expected to describe who their sample is representative of, who was excluded, describe efforts to ensure cultural validity of measures used, and share experiences of consultation with community members and stakeholders in conceptualizing the study and discussing implications of the findings.

We could benefit from a racial equity tool to help assess IPV research—not to define or enforce conformity in conceptualization and process, but to promote best practices and offer guidance, including from those who have a depth of experience with culturally relevant and anti-racist research and who can help lead the way. Research can expose, challenge, and help dismantle systemic racism and inequity or it can perpetuate it. To quote Mnguni, (2020) “#BLM, even in research: A call to researchers to take a knee.” We have a responsibility to tell it like it is.

Attention to diversity, equity, and inclusion can be demonstrated in various ways and will not be identical across studies. It may be demonstrated in the topic of research, such as research that aims to address violence against marginalized and oppressed groups, such as people who are BIPOC, immigrants, transgender, or people who are poor. It may mean including a statement of positionality, recognition of intersectionality, clarity about who is included and excluded in the sample; it may mean involving the community in planning and in interpreting results, or it may be critical discussion of the role of systemic racism in the IPV under study. There is no formula or complete checklist for researchers to describe how they have addressed diversity and inclusion in their research. Rather, it is an opportunity to be earnest in sharing the ways in which they have grappled with and sought to promote diversity, to be inclusive, and to seek equity. We can all improve, dig deeper. Doing so calls for self-reflection and reflexivity as a continuous practice, for intentional critical examination of our positionality, our thinking, and our methods, and for engaging and consulting with others to continually learn and grow. We can share what we have done so others may learn from our experiences. We can share the limitations of our work with regard to diversity and inclusion, equity, and anti-racism—this too may help others address these in their own research and in doing so, help the IPV field to pursue justice, dismantle oppressive systems and eradicate violence in its many forms.