Considerations for Employing Intersectionality in Qualitative Health Research

Abstract

Intersectionality theory has recently emerged in the health sciences as a critical theoretical and methodical approach. Though some scholars have outlined explicit guidelines for applying intersectionality in research using quantitative methods, others have cited epistemological concerns and additive thinking to advocate for the analysis of intersectionality with qualitative methods. Thus, there remains a need for additional guidance and support for utilizing and applying intersectionality theory throughout the qualitative research process. With the goal of demystifying the process of utilizing intersectionality as a methodological approach in qualitative research in the health sciences, this paper provides researchers with recommendations, specific examples, and important considerations for incorporating intersectional approaches into study conceptualization, participant recruitment, data collection, and data analysis. Additionally, this paper reviews challenges that researchers may experience in conducting research using intersectional approaches and offers suggestions for overcoming challenges. This paper offers timely and relevant information that can be used to strengthen the theoretical and methodological rigor of qualitative health research, especially studies that seek to advance health equity.

Considerations for Employing Intersectionality in Qualitative Health Research Intersectionality is an analytic sensibility, a way of thinking about identity and its relationship to power. Originally articulated on behalf of black women, the term brought to light the invisibility of many constituents … — all face vulnerabilities that reflect the intersections of racism, sexism, class oppression, transphobia, able-ism and more… Intersectionality has been the banner under which many demands for inclusion have been made, but a term can do no more than those who use it have the power to demand

(Crenshaw, 2015, para.5).

Illuminating socio-historical forces of marginalization and thereby better contextualizing phenomena under study, intersectionality theory can serve as a useful complement to qualitative health research. Intersectionality theory values and can guide the implementation of research methods that capture the lived and multifaceted experience of individuals at the crossroads of oppressed identities and social positions/locations. Numerous scholars have discussed the utility of intersectional approaches in empirical research (Bauer, 2014; Berger & Guidroz, 2009; Cole, 2009; Else-Quest & Hyde, 2016a; Few-Demo, 2014; Hancock, 2007; McCall, 2005; Warner & Shields, 2013). For example, explicit discussion and guidelines have been proposed for applying intersectionality in research using quantitative methods (e.g., sampling, measurement, and statistical analyses; Bauer, 2014; Bowleg, 2008; Cole, 2009; Else-Quest & Hyde, 2016b; Hancock, 2007). By contrast, qualitative researchers have critiqued the use of quantitative methods with an intersectional approach, emphasizing epistemological concerns around power and the pitfalls of additive, single-axis thinking (e.g., Birdsong, 2016; Bowleg, 2008; Shields, 2008). Nevertheless, there remains a need for more explicit guidance and support to utilize and apply intersectionality theory at each step of the qualitative research process. In particular, as intersectionality continues to be named but not deeply engaged (Else-Quest & Hyde, in press), clear guidelines are needed to gird the research of scholars who are new to intersectionality or who want to develop their intersectional approach. Our broad aim is to build upon existing scholarship on the utility and analysis of intersectionality with qualitative methods in order to demystify intersectionality and maximize the contributions that can be made within qualitative health research. To that end, we offer recommendations for “doing intersectionality” at various stages of the qualitative research process with key considerations for study methods and related logistics in the health sciences.

Intersectionality

Though the concept of intersectionality is embedded in Black feminism, the explicit theorizing of intersectionality is a more recent development (Alexander-Floyd, 2012; Cho, Crenshaw, & McCall, 2013; Hancock, 2016). Hancock (2016) describes the development of “intersectionality-like thought” throughout the 19th and 20th centuries, identifying the diverse origins of intersectionality, as in Sojourner Truth’s “Ain’t I a Woman” speech and the work of Anna Julia Cooper (1892). Decades after Cooper’s writings on the marginalized and ambiguous status of Black women in the U. S. during Reconstruction, Frances Beale (1970) described the “double jeopardy” of being a Black woman and experiencing both sexist and racist marginalization and discrimination. Echoing these ideas, the Combahee River Collective (1982), writing from the standpoint of Black feminist lesbians, asserted that systems of oppression such as sexism and racism are “interlocking.” Analyzing the interconnectedness of sexism and racism within the justice system, legal scholar Kimberlé Crenshaw (1989, 1991) introduced the term intersectionality to acknowledge, characterize, and explicitly examine how structures of oppression marginalize Black women.

Likewise, Collins described intersectionality as a framework for understanding the unique experiences of multiply-marginalized individuals within a “matrix of domination characterized by intersecting oppressions” (Collins, 2000, p. 23). And, building upon this imagery, May proposed that intersectional approaches require “matrix” thinking, rather than “single-axis” thinking, and that they are open-ended, dynamic, and “biased toward realizing collective justice” (May, 2015, p. 251). Other examples of Black feminist scholarship invoke consonant themes regarding the simultaneous membership in multiple social categories and the linked systems of power and inequality (e.g., Alexander-Floyd, 2012; Berger & Guidroz 2009; Carastathis, 2016). These writings have provided a diverse and generative foundation for the deployment of intersectionality as a critical theory across disciplines and multiple socially-constructed identities.

As a result of this rich history of critical scholarship, intersectionality offers researchers a robust analytical approach to understanding and examining the interconnectedness of numerous socially-constructed identities (e.g., race, gender, sexual orientation, class, etc.) as they collectively shape the lived experiences of individuals and groups. As intersectionality has evolved and distinguished itself across multiple disciplines, common elements of the theory have become clear: 1) the assumption that all individuals have multiple identities that converge; 2) within each identity is a dimension of power or oppression; and 3) identities, though possessed by individuals, are also created by socio-cultural context and are thus, mutable (Else-Quest & Hyde, 2016a).

In articulating the rationale for intersectional approaches in research, scholars have raised several critical concerns in how researchers conceptualize and apply intersectionality in their work. In describing these concerns, we aim to provide a clearer conceptualization of intersectionality and how mis-conceptualizations have limited its utility in research, with the ultimate goal of assisting researchers in avoiding these missteps. Notably, intersectionality scholars have been critical of additive approaches to understanding experiences associated with identity—that is, approaches that consider social categories such as race or gender as entirely independent, distinct, and mutually exclusive (Else-Quest & Hyde, 2016a; Spelman, 1988). Recent theoretical work focusing on how interconnected identities create unique experiences of marginalization have shown that treating identities as additive fails to capture the full delineations of oppression (Bowleg, 2008, 2012). In other words, the additive approach is counter to the conceptualization of intersectionality because discrete categories of identities such as race or gender or sexual orientation fail to account for the singular multidimensional lived experience of individuals experiencing multiple simultaneous forms of marginalization and this approach ultimately masks the true contours of oppression in the lives of such individuals.

An additional concern raised is that most research utilizing intersectionality as a guiding framework has focused on the experiences of marginalized groups, such as women of color, leading some to refer to the theory as a “content specialization” in the experiences of multiple oppressions (Hancock, 2007). Yet, insofar as intersectionality contends that all people are members of multiple social categories or groups, which contain a dimension of power or inequality, this framework is applicable to all groups. Thus, it can also be utilized to investigate the experiences of individuals with multiple privileged identities or a combination of privileged and oppressed identities (Christensen & Jensen, 2012; Yuval-Davis, 2011). Moreover, scholars have critiqued the “flattening of intersectionality”, described by Rosenthal (2016) as an over emphasis on research that seeks to understand the experiences of individuals with multiple marginalized identities rather than conceptualizing and framing the social inequities that shape their experiences and identifying potential solutions that address such inequities.

A final concern involves the limitation of applying intersectionality to only methodology (Bowleg, 2017). In qualitative research (particularly research that has material implications for individual wellbeing, like health equity research), it is important to also examine the epistemology (i.e., the study of knowledge) behind one’s approach. In other words, methodology based on intersectionality is incomplete without epistemology that also considers meta-theory congruent with an intersectional perspective (Bowleg, 2017). In the case of qualitative health research, intersectional epistemology involves methodological matters of reflexivity and interpretation; which is, in turn, directly related to the applicability of findings to contexts of health equity.

Intersectionality’s Potential for Impact in Qualitative Health Research

Qualitative research is well-suited for the utilization of intersectionality theory (Bowleg, 2008, 2017; Shields, 2008; Syed, 2010). Applying an intersectional lens in qualitative research can enable us to “see” what is outside the scope of most quantitative and qualitative health science studies. Through intersectionality, we can illuminate and dissect the complexities of minds and bodies as sites of intersectional oppression and generate new knowledge and more holistic representations of marginalized experiences and the forces that create those experiences to facilitate greater understandings of health as well as more comprehensive solutions. Further, the theory allows for the advancement of social justice via critical study of health disparities, the illness experience, constructions of cultural illness narratives, health behaviors and practices, experiences of caregivers, systematic influences on health outcomes, and health services at individual, interpersonal, community, structural, and a combination of the aforenoted levels. Importantly, intersectionality theory is useful to health researchers in the way it can be utilized to address underlying power structures that inform health inequities (Bowleg, 2017; Choo & Ferree, 2010).

Numerous scholars have examined health related topics while implicitly attending to components of intersectionality and such studies have unveiled the experience of multiple forms of marginalization in healthcare settings (Agénor et al., 2015; Jaiswal et al., 2019; Johnson et al., 2016), highlighted little or previously unknown influential sociocultural factors that can influence health behavior or outcomes (Bond & Gunn, 2016; Jaiswal et al., 2019; Johnson et al., 2016; Opara et al., 2019), and utilized the voices of those experiencing multiple forms of marginalization to highlight participant or patient recommendations for intervention or policy change (McLemore et al., 2018). However, in many qualitative health studies, the central tenets of intersectionality remain largely unacknowledged and no guidelines exist to assist researchers with incorporating the theory in their work. Thus, there is significant opportunity to utilize intersectionality as a unifying, explanatory, and analytical framework for guiding the conceptualization, conduct, and analysis of qualitative health research.

Doing Intersectionality in Qualitative Health Research

After identifying a research question or general area of study, the qualitative research process involves selecting an epistemological framework, theoretical lens(es), conceptual framework(s), approach (e.g., phenomenology, grounded theory, ethnography, etc.), methodology (e.g., observation, action research, case study, narrative analysis, etc.) and data collection techniques (e.g., interviews, focus groups, field notes, photographs, diaries, etc.), and format for presentation and writing (e.g., quotes, matrices, diagrams, etc.). When utilizing intersectionality as a theoretical lens and/or methodological framework, researchers can consider the timing and the “how” of incorporating intersectionality into their work. For example, researchers may utilize intersectionality as a theory to guide the entire research process or may wait to incorporate the framework into the data analysis stage. Although most qualitative approaches favor incorporation of theory early in the research process, there are some approaches (e.g., phenomenological) that encourage researchers to hold off on incorporation of theory so as not to bias the lens of data collection. Additionally, grounded theory approaches to research design task the researcher with the use of inductive and iterative aspects of methodology and coding, which, similar to phenomenology, do not require a priori identification of a guiding theory. Depending on the goals of the research, either approach (i.e., early or late incorporation of intersectionality) can be useful.

However, we recommend that intersectionality be considered during study conceptualization because even if the theory is not used to guide the design of the research study, the theory can still generate important considerations for interactions with participants during recruitment, data collection, and dissemination; we explore and address these considerations in depth in subsequent sections. Utilizing intersectionality as a theoretical lens early on can be particularly beneficial as it allows researchers to consider and account for the influence of sociohistorical forces of marginalization and understand participant identities as multidimensional and interdependent at each stage of the research process. Doing this will allow researchers to select approaches, methods, and data collection and analysis strategies that are more sensitive to the lived realities of participants, which may generate more robust and nuanced findings.

Study Conceptualization

The first stage of qualitative research, and the first point at which intersectionality can be considered, is study conceptualization. When conceptualizing a study, it is important to consider the identities of individuals designing the study, recruiting participants, and collecting, analyzing, and disseminating data. Even in the context of participatory designs, differences and similarities in the cultural backgrounds of researchers and participants can significantly influence the process of qualitative inquiry. For example, researcher characteristics may contribute to important oversights in the design of research, difficulties with participant recruitment, and/or misinterpretations of data—all of which can threaten the quality of the research (American Psychological Association, 2003).

The “insider-outsider” paradigm has been used by many qualitative researchers to explore and understand how differences and similarities in identity between researchers and participants influence researcher biases as well as what information is sought and communicated, how it is communicated, and how it is interpreted by participants and researchers (Maxwell, Zunga, Abrams, & Mosavel, 2016; Mullings, 1999). Designing studies from an “insider” perspective is, among some, an imperative and recommended methodological practice since it centers the voices of individuals from marginalized groups (McLemore & Choo, 2019) and can promote empathy and rapport-building (Ross, 2017). If the experiences of marginalized groups are captured and told from an insider perspective, this approach can allow researchers to operate with pre-existing knowledge that may better facilitate access to and interactions with participants (Greene, 2014), allowing narratives rendered to be more “authentic” and reflective of those under study while also providing potential emotional benefits for participants (Ross, 2017).

On the contrary, there are several advantages associated with “outside” or “partial” group membership in the research context. For example, with regard to increasing neutrality, “outsider” researchers are believed to have an advantage. Although some scholars argue that complete neutrality is impossible in research, unfamiliarity with a research context is believed by others to maximize impartiality and minimize the distortion of meaning (Asselin, 2003; Greene, 2014) as outsiders may be more likely to ask clarifying questions, which could reveal information that may have remained submerged with a more familiar researcher (Corbin Dwyer & Buckle, 2009). However, it also argued that an “outsider-as-expert” approach serves to further marginalize individuals from these populations due to its tendency to legitimize approaches to research that focus on deficit-based models of healthcare and reduce representation of group members within academic institutions (McLemore & Choo, 2019).

While the ‘insider-outsider’ paradigm brings to light important considerations for the incorporation of intersectionality into study conceptualization and planning, the framework is restrictive via polarization that inadequately accounts for researchers who occupy the ‘space in-between’ insider and outsider statuses by having both shared and different identities with participants (Corbin Dwyer & Buckle, 2009). Now more often conceptualized as a continuum, the insider-outsider paradigm remains restrictive when considering the notion of the intersectionality in social identity and position. For example, despite shared gender identity and ethnicity, a White cisgender Latina interviewer may struggle to establish trust and rapport with a research participant who is a Black transgender Latina because of disparate experiences relating in part to birth-assigned gender, cisgenderism, racism, and cultural norms associated with the intersections of their identities.

Identity is a profoundly complex construct that becomes even more obfuscated when considering researchers and participants simultaneously occupy more than one identity. Because identities are properties of individuals, but also shaped by socio-cultural context, their meaning and salience are dynamic and shifting. Further, identities ascribed to researchers by participants (or to participants by researchers) are not always concordant due to differing perspectives and contexts between researchers and participants (Simon & Mosavel, 2011). And, in the context of qualitative research with human participants, identity salience and concordance can shift almost continuously based on environmental changes, questions asked, information shared, verbal and nonverbal responses, sociopolitical contexts, institutional affiliations, and the juxtaposition of research partners. These psychosocial processes stand to impact the overarching research question as well as the amount and quality of data a researcher is able to gather. Thus, it is important to recognize and attend to the ever-shifting tide and complexity of identity during the planning stage of the research process to account for and attempt to minimize its influence via composing diverse research teams (that include insiders and outsiders), engaging in critical self-reflection, and engaging with communities’ participants are part of before the research process begins.

Further, as all researchers have additional positionality and symbolic capital due to institutional affiliation and academic background, power imbalances are still inherent in researcher-community interactions (Simon & Mosavel, 2011; Sprague et al., 2019), even for “insider” researchers who share backgrounds or experiences with participants (Ross, 2017). If not addressed, such imbalances can serve to place communities in subordinate positions in the research process, reinforcing problematic hierarchies that can serve to further disenfranchise and potentially harm communities under study (Muhammed et al, 2015; Simon & Mosavel, 2011; Sprague et al., 2019).

When possible, training community members as study team members (including interviewers) can also enhance team diversity and promote active collaboration and co-leadership with communities (Simon & Mosavel, 2011; Muhammed et al, 2015). Additionally, practices such as reflexive journaling (i.e., written systematic self-awareness) and bracketing (i.e., identification and suspension of researcher biases) have been useful in helping researchers understand the influence of their personal identities and associated biases on the research process, including study conceptualizing, participant recruitment, data collection, and data analysis (Gearing, 2004; Ortlipp, 2008). For researchers who do share backgrounds or experiences with participants, employing “caring reflexivity” by opening dialogue with participants about power imbalances and concerns during the study can enhance trust and build rapport (Ross, 2017).

Additionally, addressing power imbalances, expectations, potential benefits, and concerns in researcher-community relationships before the study protocol is implemented is critical, and can often be addressed through collaboration with community stakeholders and/or use of community advisory boards that are involved in decision-making processes surrounding research methodology and dissemination before any research is conducted (Simon & Mosavel, 2011; Sprague et al., 2019). Involving communities as co-leaders in study conceptualization can also facilitate a co-learning environment that gives communities shared agency in the creation of knowledge based on their experiences, while further building trust and rapport that aids the research process (Muhammed et al, 2015). Formulating community-researcher collaborations through methodologies such as community-based participatory research (CBPR) allows individuals’ voices to be heard and amplified early in the research process.

Broadly, study conceptualization rooted in intersectionality can examine the complex system of feedback loops and interactions between macrosystems of power (e.g., institutions) and microstructures (e.g., identity) (Choo & Feree, 2010). For example, highlighting the value of examining multilevel and intersectional influences, Gonazelez-Lopez and Vidal-Ortiz (2008) examined the intersection of sexual health behaviors, Latin cultural values, and sexual orientation in relation to HIV/AIDS, stigma and identity. Researchers may also apply such principles in examining how extant intersectionality paradigms may have overlooked contributions from nontraditional sources of scholarship, such as contributions from Chicanas/Latinas (Baca Zinn & Zambrana, 2019).

Participant Recruitment

Qualitative health researchers can benefit from employing an intersectional approach when determining “who” to include in their study. Cole (2009) suggests that researchers begin by asking themselves: 1) Who are the individuals in the category of interest?, 2) What is the role of inequality in their lives?, and 3) What commonalities exist across the multiple identities of participants? After identifying participants of interests and their intersecting identities researchers should then critically examine the role of marginalization and the social forces that drive inequities as it relates to the phenomena under study. For example, while considering the topic of interest also consider how racism, sexism, classism, cisgenderism, homophobia, disease status, or any relevant combination of those factors impact illness experience, experience in a healthcare system and/or with healthcare providers, perceptions and internalization of stigma, cultural mistrust, or access to prevention resources, healthcare, and treatment. Though the research question may not be centered on inequality, it is important that researchers recognize the pervasiveness of marginalization and consider mechanisms of influences as well as the ways such experiences may influence the health or healthcare experiences of potential participants - such considerations may provide key insights for participant recruitment and engagement.

Next, researchers can focus on identifying commonalities across intersecting identities and social positionality. Researchers may discover that intersectionality not only centers around traditional categories of race and gender but is ideally person-centered (Else-Quest & Hyde, 2016a, 2016b). In other words, that there are commonalities across intersectional locations and focusing only on comparisons or differences in categories or identities might obfuscate those commonalities. For example, in an effort to identify similarities in experiences across those with varied identities, researchers using an intersectional approach may frame a health issue under study (e.g. breast cancer) as one shared commonality across categories of race, gender, sexual orientation, and so on. Other commonalities can include health-promoting or health-compromising behaviors, shared illness experiences, and/or barriers or facilitators to health resources.

Alternatively, identifying how commonalities differ among certain intersectional identities that share a common axis (e.g., gender, sexual orientation) can aid researchers in exploring how barriers or facilitators are differentially efficacious among minority subgroups. For example, a common facilitator of cervical cancer screening, sexual and reproductive health services use, is associated with increased likelihood of screening among White sexual minority women but not among Black or Hispanic sexual minority women (Agénor, Krieger, Austin, Haneuse, & Gottlieb, 2014). Determining shared categories of identities and experiences can assist researchers in determining how to frame research questions using an intersectional lens and may also assist researchers with identifying optimal locations for recruitment. Additionally, use of CBPR can aid researchers in determining how to frame research questions that guide research design by bringing in the perspectives, experiences, and priorities of communities under study (Muhammed et al., 2015). Further, CBPR that takes a co-leadership approach can enhance trust within communities, thereby reducing barriers to participant recruitment via active collaboration (Sprague et al., 2019).

Where to recruit.

As marginalized groups often face unique barriers to utilizing healthcare, including structural, interpersonal, and cultural factors related to help-seeking (Alegría et al., 2002; Rodríguez, Valentine, Son, & Muhammad, 2009), focusing recruitment entirely at traditional health sites (e.g., hospitals, healthcare institutions, academic research centers) may overlook those with intersectional marginalized identities as these social structures presume a certain level of access and self-identification. Expanding recruitment efforts may mean an emphasis on smaller intimate communities (Simon & Mosavel, 2011). Given that the effects of inequality are manifested and confronted in various social spaces (Cole, 2009), additional recruitment locations may include advocacy organizations, religious centers, empowerment groups, community centers, hair and nail salons, restaurants, sororities/fraternities and other public service organizations, and web-based locations such as social media, chatrooms, blogs, and support groups. Similarly, those locations themselves become an important component of an intersectional approach, insofar as power and inequality are conferred by social, political, historical, and geographical context.

How to recruit.

Of the recruitment strategies available to qualitative health researchers, purposive, quota, and snowball sampling may be the most useful for those employing an intersectional approach (Else-Quest & Hyde, 2016b; Mack, Woodsong, MacQueen, Guest, & Namey, 2005; Suzuki, Ahluwalia, Arora, & Mattis, 2007). Purposive and quota sampling are similar as both strategies facilitate the identification of participants based on preselected criteria relevant to the research question. One key difference, however, is that quota sampling in qualitative research delineates participant categories and a minimum number of participants required for each category (Robinson, 2014). For example, if race, socioeconomic status, and gender are of interest in a study about experiences with seeking cardiac care, a quota sample might aim to recruit a minimum of at least 10 individuals at the intersection of each possible identity category. This strategy is often used when researchers are interested in making comparisons between or within groups and helps facilitate the representation of perspectives from groups of interest (Robinson, 2014).

Another useful strategy is respondent-driven sampling, also known as participant referral or snowball sampling. This method involves asking and/or incentivizing participants to recruit additional participants (Meyer & Wilson, 2009), which can be particularly useful for targeting those with stigmatized or hidden identities (Bostwick & Hequembourg, 2013; Heckathorn, 1997). Respondent-driven sampling can also be buttressed by the vocal support of a community leader, as they may provide guidance on effective recruitment techniques, help researchers establish trust with potential participants, and assist researchers in preemptively addressing concerns community members may have about the research project (Berg, 1999, Milburn et al., 1991, Valente & Pumpuang, 2007). Though respondent-driven sampling can be useful in increasing access to stigmatized or hidden populations, the benefits of this approach must be considered in tandem with potential drawbacks. Given its nature, respondent-driven sampling is inherently biased toward inclusion of individuals with interrelations, which can potentially limit diversity of the sample and contribute to a greater likelihood of missing individuals who are not connected to the accessed social network. Such individuals may have contrary experiences and viewpoints that could be valuable to achieving a given study’s aims.

Finally, venue-based sampling or time-location sampling (TLS) may also be utilized to recruit hard-to-reach populations. As an example, Medina-Perucha, et al. (2019) conducted a qualitative study on intersectional stigma, sexual health, and substance use to examine how health inequities are driven by stigma. Participants were recruited from drug service and service for sex workers sites, with staff at those sites identifying eligible participants (Medina-Perucha, et al., 2019). Similarly, Muhib et al. (2001), utilized venue-based sampling at nightclubs to recruit self-identified LGBTQ, young adults who engaged in club drugs, such as ecstasy. Venue-based sampling or TLS allows researchers to intercept hard-to-reach populations in places and times where they might gather.

Researchers may consider combining sampling approaches to minimize selection bias and related threats to the trustworthiness of data. Combining strategies (e.g., employing quota and snowball sampling online and in community-based settings) may better capture participants who are considered “hard to reach,” especially if the identities of the population of interest are hidden or associated with illegal activity (e.g., illicit drug use or sex work). For example, researchers can network with social service providers, community organizations, and advocacy groups, offline and online, that empower these groups (Mehra, Merkel, & Bishop, 2004). Advertising in community forums, newsletters or agency listservs, online social groups hosted by platforms such as www.meetup.com, www.craigslist.com, and social media may be other routes researchers can pursue for recruitment (Babbit, 2013). It is also important to note that, “matched” researchers (i.e., those similar in race, age, gender, and/or culture to potential participants) have also been shown to enhance recruitment efforts (Berg, 1999; Milburn, Gary, Booth, & Brown, 1991).

Innovative recruitment strategies may also be necessary to connect with individuals at the intersection of multiple oppressions (Williams & Fredrick, 2015). A relatively new innovation for research recruitment and data collection, Amazon’s Mechanical Turk, has been shown to yield participants with greater demographic diversity, in terms of socioeconomic status and ethnicity, than traditional recruitment strategies research studies (Casler, Bickel, & Hackett, 2013). In addition, recruitment through social media (e.g., Facebook and Instagram) and dating applications (e.g., Grindr and Tinder), which allows researchers to target potential participants based on location, demographics, and interests, may help extend the reach of traditional recruitment strategies to broader and diverse samples (e.g., Arcia 2014, Fenner et al., 2012, Zickuhr & Smith, 2012). However, using web-based recruitment techniques raise issues regarding ensuring privacy, confidentiality, and informed consent as well as concerns on establishing research legitimacy online and verifying participant information (Berry, 2004; Koo & Skinner, 2005). The Harvard Catalyst Regulatory Foundations, Ethics, & Law Program has prepared an extensive guide - complete with ethical, cultural, and legal considerations - to assist researchers with using online platforms such as social media as a recruitment tool (Gelinas, Pierce, Winkler, Cohen, Lynch, & Bierer, 2017).

In addition to dating apps and social media platforms, web-based survey platforms and research-specific recruitment apps also exist to help facilitate participant recruitment. For example, Survey Monkey and Qualtrics have participant pools from which researchers can recruit and each platform allows researchers to specify the demographic characteristics of desired participants, which can facilitate direct and targeted access to intersectional populations. Furthermore, apps such as Research Unlimited and My Clinical Study Buddy as well as the web-based company Research Match were designed specifically for the purpose of connecting researchers to potential participants. These companies advertise studies to app users and also utilize technology to remind participants of appointments. Although the “digital divide” may restrict access or utilization of app or web-based platforms for some populations, such as older adults or homeless persons, as well as people in less developed regions, web-based recruitment efforts have been shown to net more diverse samples than traditional techniques (Gosling, Vazire, Srivastava, & John, 2004).

Data Collection

There are several types of data collection techniques available to qualitative health researchers, among these are: interviewing, individually or in groups (i.e., pairs or focus groups, or the World Cafe Method of rotating group dialogues), participant observation (field notes, videos, photographs, social media), and analyzing pre-existing documents including, but not limited to, diaries, records, or memos (Padgett, 2016; Suzuki et al., 2007) and newer technologies for CBPR like Photovoice. Here we discuss strategies for incorporating intersectionality theory into data collection via focus groups and interviews, two particularly popular methods in qualitative health research.

Considering the tenets of intersectionality during the thoughtful process of developing effective interview guides can shape the quality of data collected and influence results in unanticipated ways. Numerous scholars have provided recommendations for developing quality interview guides and highlight the utility of refining questions throughout the research process as a reflection of increased understanding of phenomena that progressively occurs with each data collection experience (e.g., Charmaz, 2006; Creswell, 2007). Questions can also be refined based on a researcher’s observations about their biases and perspectives (Agee, 2009). In addition to developing and refining questions related to how the intersections of participant identities may influence or be influenced by the phenomena under study, we advocate for broadening the foci of the study to make meaning around and/or gauge the influence and relevance of intersections of societal structures and systems related to the phenomena under study with a keen eye toward identifying and explicating the influence of multiple levels of social injustice. In other words, the influence of structural stigma (i.e., the sociopolitical and institutional processes that drive inequality; Metzel & Hansen, 2014) should be accounted for when developing routes of inquiry, as qualitative research has the potential to reinforce structural stigma when formulating interview guides from a deficits-based approach of health (Muhammed et al., 2015). Collaborating with communities in the development of interview guides can reduce power imbalances in the research process, and thereby help reduce risk of reinforcing structural stigma and perpetuating social injustice.

Triangulation, an approach that aligns well with intersectionality theory, is another strategy that can be used to enhance our understanding of phenomena while simultaneously enhancing the validity, quality, and trustworthiness of data (Padgett, 2016). With the goal of corroborating findings via cross verification, triangulation refers to combining methods, investigators/observers, theories, and data sources to generate more comprehensive knowledge related to the topic of study (Curry & Nunez-Smith, 2014; Padgett, 2016). This strategy could be particularly useful in simultaneously examining experiences of individuals with multiple marginalized identities while also examining the structures and systems that frame their experiences. For example, researchers studying the experiences of Black gay men living with HIV who are navigating the HIV care continuum can triangulate data sources by utilizing interviews with the population of interest, focus groups with healthcare providers, and archival materials (e.g., health records, social media posts, public health campaign materials, etc.). Such strategies can be useful in obtaining information at individual, relational, cultural, and structural levels.

Additionally, there are a few other well documented practices in qualitative research methods that may be especially useful for scholars interested in incorporating intersectionality in their work. For example, piloting interview guides with individuals from the population of interest can be beneficial to ensure that potential participants will understand questions and that questions will elicit information relevant to the overarching research question (Padgett, 2016). It can also be useful to actively collaborate with individuals from the community of interest about ideal data collection sites, time periods for data collection, and compensation for participation. In this regard, community-based participatory research approaches can be particularly useful for designing research that is sensitive to the needs and experiences of participants and can assist researchers with avoiding common pitfalls that can impede study progression or negatively influence data quality. Further, CBPR promotes an intersectional approach, where experiences can be understood within the context of community rather than as a comparison to the dominant norms (Weber & Parra-Medina, 2003). Such collaborations perform best when researchers employ shared decision-making processes with communities of study, thereby promoting co-leadership throughout the research process that facilitates both study progression and community agency (Sprague et al., 2019). Shared decision-making, a concept adapted from medical literature (Peek et al., 2016), should involve an ongoing process of information-sharing, open dialogue about benefits and harms, and decision-making about research methodology, epistemology (with respect to cultural norms, boundaries, and ownership), and dissemination of findings.

Ethical concerns.

There are several ethical considerations that need to be applied sensitively and with additional caution when working with populations with marginalized identities, among them are informed anonymity, privacy, confidentiality, and consent. Sharkey et al. (2011) argues that ethical concerns and risks may be multiplied when working with vulnerable populations and that participants themselves should be queried on how to address issues of confidentiality and anonymity. This is in line with an intersectional approach that values “inclusion [and] transcends representation, offering the possibility to repair misconceptions engendered by the erasure of minority groups and the marginal subgroups within them” (Cole, 2009, p. 172). Moreover, it accounts for the relationship between participant autonomy (including a right to disclose) and protecting confidentiality; for participants belonging to multiply-marginalized groups, confidentiality and anonymity agreements may be perceived as the negation of their right to be identified (Giordano, O’Reilly, Taylor, & Dogra, 2007). This practice may unintentionally silence the voices of participants, a practice at odds with the spirit of qualitative research. Thus, carefully navigating this tension and ethically offering opportunities for participants to forego anonymity if desired is essential for qualitative health researchers working with participants or communities who may have experienced a history of being silenced because of their intersectional locations.

Relatedly, informed consent may also require additional considerations when working with marginalized groups. Consider that some individuals with multiple marginalized identities may not want to be identified or known to be associated with a research study on the topic of their identity, this may be especially true for individuals with hidden or stigmatized identities and those associated with illegal activity (e.g., gay man living with HIV or transwoman engaged in illegal sex work). It is possible that for some participants, signing a consent form that documents their identity may serve as a deterrent for participation. To overcome this issue, many ethical review committees and institutional review boards around the world have offered researchers alternative methods for obtaining informed consent.

As such, researchers may find it useful to work with their ethics review committee to learn more about and utilize options for informed consent that facilitate greater privacy for participants such as a verbal consent process or a waiver of consent. In addition, utilization of participatory methods may be particularly useful as these approaches can allow individuals that represent the population of interest to work with researchers to ensure linguistic and cultural appropriateness of written or verbal consent documents by collaboratively drafting and/or reviewing and providing feedback on existing documents and scripts (Simon & Mosavel, 2011). Also, utilizing a pre-consent quiz, can help ensure that potential participants understand the research goals and procedures as well as their rights (Simon & Mosavel, 2011).

In light of these ethical considerations, cultural humility and structural competence training may help ensure that research team members are trained to uphold and be critical of ethical principles. Though the specific content of the training may vary based on the population of interest, at minimum team members should be made aware of the tenets of intersectionality, historical issues of health-related research, cultural mistrust, identifying and addressing stereotypes and relevant consequences, and effective communication skills (including appropriate verbal/written terminology and body language/gestures). Further, cultural humility also requires team members to be reflexive and aware of the boundaries of their own understanding, predicated on power differentials and privilege associated with their own identities as well as their positionalities as researchers (Muhammed et al., 2011). From this perspective, it is essential for researchers to be authentic about their own identities (Simon & Mosavel, 2011) and transparent about their own backgrounds in their interactions with participants in order to build trust and rapport (Simon & Mosavel, 2011; Ross, 2017). As research team members who share identities or positionalities with the population of interest can also face marginalization in the research process, care should be taken to democratize knowledge and processes within the research team (Muhammed et al., 2011) while accounting for the potential need for these interviewers to debrief with someone of a similar background (Ross, 2017). Equipping team members with essential knowledge and team-building practices may help improve their cultural sensitivity, consciousness, and ability to collaborate, thereby enhancing their skills to collect quality data.

Data Analysis

Several types of data analytic strategies exist; including but not limited to the popular grounded theory, content, and thematic analytic approaches. Though each approach varies in its purpose, all three often involve the analysis of text data to identify the occurrence of popular categories across data (Hsieh & Shannon, 2005). In addition, there exists an analytic strategy, known as disconfirming case analysis (also called negative or deviant case anaylsis), focused on identifying contradictory patterns in data (Mays & Pope, 2000; Yardley, 2000; 2008). Each of these analytic approaches serve different strengths depending on the research question. Further, these approaches utilize a variety of methods that differ in terms of coding schemes, coding origins, and susceptibility to bias (Hsieh & Shannon, 2005). However, data analysis utilizing either approach can be strengthened through intersectional framing.

In a discussion of intersectionality ontology and analysis Anthias (2013) suggested that intersectional framing must consider:

1. How are social categories (e.g., gender, ethnicity, class) or concrete social relations (i.e., hierarchy, inequality, and outcomes and processes of power dynamic) of interest in the framing of the research question? As Vardeman-Winter and colleagues (2013) suggest, a key aspect of intersectionality theory is the proposition that social constructions of identity are not independent and both privilege and oppression exist concomitantly. Thus, the scope of these social categories or concrete relations can be defined in terms of the categories of interest and their relations to power structures and care should be taken to not code for identities or related experiences in ways that assume an additive nature of social identities (Bowleg, 2008) and instead examines their discrete, intersecting construction (Anthias, 2013; Bowleg, 2008; Vardeman-Winter et al., 2013).

2. Which “societal arenas” (i.e., contexts for social categories and concrete social relations) are relevant to the research question(s)? The importance of context in producing social categories or shaping their significance is fundamental to an intersectional approach. For instance, analysis of context can involve organizational arenas (i.e., how population-level social categories are organized), representational (i.e., the flow of information in different institutional or organizational frameworks), intersubjective (i.e., intergroup relations or interactions between certain social categories or concrete social relations and societal institutions), and experiential arenas (i.e., affective narratives). Such distinctions are important, as certain identity intersections, social categories, or relations may be more salient in particular arenas (Vardeman-Winter et al., 2013). For example, Bridges’ (2011) ethnography of a women’s health clinic in a New York City hospital describes how race is socially constructed within that institutional space, highlighting the racial hierarchy of the white physicians overseeing first-generation immigrant women who provide nursing care to pregnant and poor women of color. Similarly, historical or time context are essential to an intersectional approach, as Anthias (2013) notes in their third point.

3. What historical outcomes or processes are of interest to the research questions in terms of time contexts (i.e., historicity)? From Bowleg’s (2008) perspective, historicity involves the analysis of implicit data, and requires the coding of narratives through the use of knowledge about historical contexts (such as an understanding of institutional heteronormativity in the healthcare system). Thus, depending on the type of coding performed, researchers may form meta-categories that are implicitly derived from social and historical contexts relating to oppression in participant narratives.

In terms of analysis, the type of coding methodology is often based on the types of framing used. For instance, if a health sciences researcher is interested in how sexual orientation and racial concordance in healthcare encounters (a concrete social relation) impacts patients’ breast cancer treatment narratives (an experiential arena) relative to the medicalization of minority identities (a historical outcome or context), they may choose an inductive approach to coding that seeks to analyze how individuals make meaning of their experiences (Starks & Brown Trinidad, 2007). In another example, a project that focuses on the representation of youth of color (a social category) in social media-based tobacco prevention messaging (representational arenas) in relation to cultural scripts of substance use risk behaviors (a historical outcome), deductive analysis using a priori codes and theory may be more suitable. However, it remains useful to give attention to both time and historical context of the selected theories’ critical formation and the communities that have been explored as part of its formation. As such, inductive analysis should be used when possible, as it allows for codes to be derived by existing data.

In addition to type of analysis, researchers can incorporate intersectional framing into the level of analysis. According to Braun and Clarke (2006), analysis often occurs on one level, the semantic level. Data analysis that occurs at the semantic level involves analyzing data at face value, only considering what participants have articulated or written. While this approach is valuable, there is added value in moving beyond semantic analytic strategies to discover what lies beneath spoken or written content. Analyzing data at the latent level, accomplished via interpretative analysis, allows researchers to undertake this task via identification of assumptions, beliefs, thought patterns, and conceptualizations that characterize semantic content (Braun and Clarke, 2006). Incorporating intersectional framing via analyzing data at both levels can assist researchers with achieving a more comprehensive analysis. Regardless of the level of analysis or approach, it is important to note that there will always be a bias resulting from the interaction between the data and the researchers’ backgrounds and care should be taken to avoid reproducing inequality within the coding and analytic processes (Hankivsky et al., 2010).

Ultimately, an intersectional framework of analysis requires health sciences researchers to assess topics of interest in terms of what aspect of identity and power dynamics they seek to study, the categorical arena these data are derived from, and the relevant historical context (processes or outcomes). It can be beneficial to select coding methods that take intersectional framing into account and, in subsequent analysis, be attentive to implicit themes of domination and exploitation that are suggested by temporal and spatial historical contexts (Bowleg, 2008; Hankivsky et al., 2010; Anthias, 2013). Finally, thorough practices of reflexivity occur throughout the coding process as well as subsequent interpretation and reporting.

Engaging in Reflexive Practices

The practice of reflexivity is a cultivated awareness of the influence of identity and power differentials embedded in research. Reflexivity lends itself well to utilization of intersectional approaches in qualitative health research as both approaches are rooted in the injunctive to challenge upward categories of oppression (Pillow, 2003). Reflexivity asks the researcher to create and maintain a subjective awareness of their multiple privileges, intentions, and identities, and further suggests that overlooking relational authority to position framing is a strong source of error (Bourdieu, 2004). Reflexivity can be applied as a personal practice and is strongly recommended as a methodological practice that can assist researchers in identifying and managing assumptions, sentiments, and beliefs and preventing them from unintentionally influencing data collection and analytic processes. Reflexivity asks the researcher to reflect on their motivations, intentions, and assumptions - without reflexivity a researcher may avoid self-censure and not deeply engage with the data collection and analysis processes (Berger, 2013).

Qualitative health researchers can apply reflexivity to their methodology in a variety of ways: 1) acknowledging, examining, and setting aside biases and assumptions (bracketing), 2) reflecting on assumptions in the course of research (epistemological reflexivity), and 3) examining socio-political conditions linked to research (critical theory standpoint; Rae & Green, 2016). Reflexivity can unearth dimensions of power differentials, such as those between the researcher and participant as well as those unique to the context of research (Carstensen-Egwuom, 2014). This can be especially illuminating if the researcher identifies with multiple categories and or shares commonalities with participants.

There is also utility in attending to often overlooked (e.g., ability and sexual orientation identities) and invisible identities (e.g., disease status and trauma history). Implementation of reflexivity allows researchers to examine their own identities and relationships to structures of privilege and oppression in relation to the research topic and participants. Awareness and reflection on commonalities and power differentials can bring greater clarity to the process of how these categories confer power and disadvantage at multiple social locations and how such processes influence systematic inquiry.

Implications

Akin to other theories that challenge inequitable social institutions and their deleterious influences (e.g., critical race theory and reproductive justice theory), at the core of intersectionality is promotion of social justice and equity (Else-Quest & Hyde, 2016a; May, 2015; Rosenthal, 2016). Thus, incorporation of intersectionality in qualitative health research can be particularly useful for work that seeks to elucidate experiences of and address health disparities. Per Kilbourne and colleagues (2006), health disparities research occurs in three phases: 1) defining and detecting disparity populations, 2) understanding the determinants of disparities, and 3) reducing disparities through intervention and subsequent evaluation research. During the first and second stage of health disparities research, utilization of intersectionality theory can alter the scope of research questions to allow for more nuanced investigations of illness narratives and the cultural constructions and contexts behind systems that influence illness and care seeking experiences. Further, utilization of the theory allows for investigators to capture nuance in the influence of multiple identity categories and hierarchical power dynamics in a manner that goes beyond the narrow focus of additive approaches. For example, investigators may seek to describe the healthcare experiences of individuals with cancer within categories of race and gender, instead of by race or gender.

Additionally, qualitative health research findings can provide important context for quantitative research that utilizes intersectionality within its analysis framework. This allows for a better understanding of how particular intersectional locations may be characterized by experiences of marginalization within healthcare systems (Hankivsky et al., 2010). As data collection is largely influenced by the defining of social identity categories and participant-researcher power dynamics (which may be particularly salient in research in clinical settings; Karnieli-Miller, Strier, & Passach, 2009), intersectionality framing facilitates transparency at this stage of the research process through its applications to reflexivity and data analysis.

In the second and third phase of health disparities research, many scholars rely on theory to guide their work in understanding and targeting determinants of disparities through interventions. However, many health science theories are informed by homogenous samples from “WEIRD” (Western, Educated, Industrialized, Rich, and Democratic) societies (Henrich, Heine, & Norenzayan, 2010), which limits the content validity and generalizability of such theories to diverse underrepresented populations. This reality highlights how theoretical approaches in health sciences research are socio-culturally deficient and reflective of social hierarchy and oppression. Qualitative research conducted with an intersectional framework, particularly if participatory methods are employed, can facilitate the undoing of these gross oversights through the generation of new knowledge, theory, and quantitative research tools that are informed by voices of marginalized groups, sensitive to their experiences, and keen on addressing relationships among systems of power, oppression, and identity salience under different socio-historical contexts and systems related to health.

In turn, intersectionality has important implications that extend beyond qualitative health research to the practice and policies that are grounded in research findings. That is, public health efforts and clinical guidelines may need revising or adapting in order to meet unique health needs or circumstances of multiply marginalized groups. For example, qualitative health research with lesbian, bisexual, and queer women and transgender men indicates that many members of these gender and sexual minority groups report both that they experience discrimination in health care settings because of their gender expression and that they receive irregular cervical cancer screenings (Agénor et al., 2019; Johnson et al., 2016).

Thus, intersectionality-informed public health campaigns may focus on outreach to members of those marginalized communities to promote cervical cancer screenings. Similarly, clinical practice guidelines for working with members of gender and sexual minority groups might involve creating more welcoming and inclusive healthcare settings. In these examples, the empirical research is leveraged to reduce health disparities in gender and sexual minority groups, whose needs may be marginalized or ignored in the context of cisnormative and heteronormative values and political motivations. Additionally, if communities are actively involved in intersectional work via CBPR methods, they have more immediate access to research findings and thus greater ability to advocate for their own interests as a direct result of the study (Muhammed et al., 2015).

Conclusion

This paper is to be used as a guide for researchers interested applying intersectionality theory to the qualitative health research process. Regardless of whether research questions explicitly incorporate intersectionality, the theory captures important considerations for researchers to consider during study conceptualization, recruitment, data collection, and data analysis. Additionally, as health sciences researchers are increasingly focused on mitigation and elimination of health disparities among marginalized populations, this paper offers timely and relevant information that can be used to strengthen the theoretical and methodological rigor of qualitative health research. Attention to interwoven and dynamic aspects of identity is essential for the illumination of definitions and experiences of wellness and disease, the identification and elimination of health disparities, and the accomplishment of health equity across diverse groups. In conclusion, qualitative health research is often utilized to make meaning of phenomena and listen to and amplify voices of the marginalized; in this regard, intersectionality theory is well-suited to empower participants, shift and expand vision and objects of focus in research, and hold researchers accountable for critical inquiry throughout the research process.

Research Highlights.

• Intersectionality holds researchers accountable for critical inquiry in the research process.

• Intersectionality has implications for study design, sampling, data collection, and analysis.

• Through intersectionality, we can generate new knowledge to advance health equity.