Introduction

This article is motivated by a call to action for international business (IB) research to encompass a greater diversity of IB-related phenomena in the world so that we can better understand and address contemporary global issues (e.g., Buckley, Doh, & Benischke, 2017; Doh, 2015; Kolk, 2016). IB scholars have argued that understanding an empirical setting that has been under-researched can contribute to richer, more diversified theory development (e.g., Barnard, 2020; Barnard, Cuervo-Cazurra, & Manning, 2017; Bruton, Zara, Van de Ven & Hitt, 2022 ; Jack, Zhu, Barney, Brannen, Prichard, Singh, & Whetten, 2013; Meyer & Peng, 2005). In IB, the newness of an empirical setting is often expressed in geographic terms, but newness can equally encompass novel or emerging phenomena in familiar locations; for instance, the practices of multinational enterprises (MNEs) with respect to the UN Social Development Goals. When attention is directed to new empirical phenomena, there can be missed opportunities for theory development if they are forced into familiar theoretical categories, such as when refugees are conceptualized as expatriates (Szkudlarek, Nardon, Osland, Adler & Lee, 2021). One well-known discussion on this topic focuses on China and the extent to which Chinese management should be theorized from the “outside in,” starting with the lens of familiar theories to see whether and how they apply to the “new” setting of China, or from the “inside out,” starting with Chinese phenomena to identify issues of salience in that setting as a basis for theory development (Barney & Zhang, 2009; Child, 2009; Tsui, 2006; Tung, 2008).

This call to action emphasizes the desirability of generating theoretical insights into empirical settings – and the phenomena and research participants they encompass – that are under-represented in the scholarly IB literature. Doing so is consistent with the strengths of qualitative research methods. Qualitative research is a sense-making activity, where “researchers study things in their natural settings, attempting to make sense of, or interpret, phenomena in terms of the meanings people bring to them” (Denzin & Lincoln, 2000: 4). It is therefore well-suited to providing descriptions, conceptualizations, and interpretations of novel empirical settings (Birkinshaw, Brannen, & Tung, 2011; Johns, 2006).

However, as Boyacigiller and Adler noted 30 years ago when considering US influence over management research, even when theories produced by US studies are tested in other settings, “researchers usually select methods that are most acceptable according to American norms, thereby rendering results that are just as culturally conditioned” (Boyacigiller & Adler, 1991: 72). This conditioning is reinforced when researchers draw on established methodological authorities so that reviewers can judge the quality of a paper’s methods by the extent to which it draws on the precedents of prior research (Corley, Bansal & Yu, 2021). While qualitative researchers are continually working at improving their craft, there are sets of best practices and conventions that are commonly used (e.g., Gehman, Glaser, Eisenhardt, Gioia, Langley, & Corley, 2018; Harley & Cornelissen, 2022). These have emerged largely from a fairly limited range of empirical settings, based on data collected from participants in North America, Europe, and Australia/New Zealand, and it can be challenging to transfer those methods to other settings (e.g., Pelzang & Hutchinson, 2018; Stening & Zhang, 2007; Tsui, 2004). Qualitative researchers may encounter a “messiness” they did not anticipate when the methods they use have been developed in better-researched and thus more familiar settings (see, e.g., Michailova, 2004; Tsang, 1998). Data gathering precepts may not fit the characteristics of the setting, and data analysis precepts may constrain the generation of new theory. This can be confusing for researchers who hope to make a novel theoretical contribution in a top journal, as well as for reviewers who are struggling to assess the potential of a submission.

The purpose of this article is to describe challenges faced by qualitative researchers studying under-researched empirical settings and to suggest practices to manage them. We draw on published accounts of research by others, and on our own experiences in the field. Rather than specifying rules or templates for how data collection and analysis methods should be adapted for novel empirical settings, we take the position that qualitative IB researchers are bricoleurs who assemble methods and techniques as needed to fit the specifics of their research (Denzin & Lincoln, 2000; Pratt, Sonenshein & Feldman, 2022). Thus, we see the practices we identify as constituting a methodological toolkit to help IB researchers, editors, and reviewers navigate quality through transparency rather than a set of rules (see Jarzabkowski, Langley, & Nigam, 2021). To set bounds on the toolkit, we have opted to exclude issues associated with language and translation.1

The format of this article differs somewhat from most articles. To illustrate the messiness that can be found in practice and the issues that researchers can face in studying novel empirical settings, we draw on our own research. We provide descriptions of quandaries faced in our own research projects through four vignettes (see Figures 2, 3, 4, and 5), and refer to them in the discussion. We recommend that they be read before the article itself because the issues they raise provide essential background for the discussion. The Barnard vignette (Figure 2) relates to research on the executives of multinational enterprises in four African countries, carried out by South African researchers Helena Barnard and Anastacia Mamabolo. The Sasaki vignette (Figure 3) relates to research on long-living Japanese companies in the traditional craft sector, carried out by Innan Sasaki, who is half-Japanese and half-Finnish, and spent her childhood in Japan. The Couper vignette (Figure 4) relates to research on international partnerships carried out in the UK and China by Carole Couper, who is Caucasian, fluent in Mandarin, and worked professionally in China prior to her doctoral studies. The Alkhaled vignette (Figure 5) relates to research on Syrian women refugee-entrepreneurs forcibly displaced and temporarily (re)settled in refugee camps, neighbouring Arab countries and Europe, carried out by Sophie Alkhaled, a bilingual researcher of British and Syrian descent, who spent her childhood in the Middle East. All researchers were familiar with, and knowledgeable about, the empirical setting they studied.

As can be seen from the vignettes, and the discussions of them in this article, we believe it is important for qualitative researchers to be highly reflexive about their personal role in the research process. Such reflexivity has always been important in the use of qualitative methods, given the centrality of the researcher in collecting and interpreting data (Van Maanen, 1979), and takes on heightened importance when researchers conduct personally-relevant research. As the range of home locations among IB scholars broadens, interest in personally relevant research is likely to continue to increase the study of more diverse empirical settings. For example, Khanna and Lakhani’s family experience and stories of the 1947 Partition of British India motivated their study of this topic (Khanna, Lakhani, Bhadada, Khan, Davé, Alam, & Hewett, 2021), and Alkhaled’s Syrian background motivated her study of Syrian refugees (Alkhaled & Sasaki, 2021).

Historically, professional distance and personal involvement have been seen as opposite ends of a continuum, based on the belief that personal involvement reduces professional distance and therefore research quality. However, there is a growing recognition that personal involvement with research projects need not be detrimental, and may instead create opportunities for more diversified theory development (Anteby, 2013; Jones & Bartunek, 2021). Rather than hiding personal connections or avoiding research with personal connections, researchers are increasingly being encouraged to be reflexive and transparent about personal involvements, and to navigate them explicitly (Jones & Bartunek, 2021; Langley & Klag, 2019). In doing so, it is essential for researchers to understand their own taken-for-granted assumptions and position vis-à-vis the research participants from whom they collect data (e.g., Alkhaled, 2021; Anteby, 2013; Couper, 2019; Hibbert, Sillince, Diefenbach, & Cunliffe, 2014).

The outline of the article is as follows. In the following section, we discuss three concepts essential to the discussion: setting, context, and contextualization. The discussion is summarized in Figure 1. In the subsequent sections, we draw on this foundation and present in turn three challenges associated with the collection and analysis of qualitative data when IB researchers are studying under-researched empirical setting. These issues exist to some extent in most, if not all, qualitative research projects. However, we believe they are intensified when scholars are studying empirical settings (phenomena and participants) that are unfamiliar to established theory and methodological norms, because of limited guidance from prior literature. They are also intensified when scholars hold multiple identities in an empirical setting, because this can affect their engagement with participants. We discuss practices that have been found to be useful in overcoming the challenges, and show how they can become opportunities for theory development.

Figure 1
figure 1

Empirical setting, research context, and contextualization.

Full size image

Empirical setting, research context and contextualization

Context has long been an important concept in IB research. It has been central to Decade Award-winning papers in JIBS over the past decade (Reuber & Fischer, 2022), and it continues to be an important part of current conversations in the journal (e.g., Beugelsdijk, Kostova, & Roth, 2017; Welch, Paavilainen-Mäntymäki, Piekkari & Plakoyiannaki, 2022). However, the term has had varied uses when applied to a research project. Sometimes, it simply refers to the empirical setting of the research. Often, it refers to a cluster of characteristics of the empirical setting that affect outcomes; see Johns (2006). For example, in the IB literature, “country-level context” and “cultural context” usually refer to a set of country and cultural characteristics that vary across countries and cultures and explain an outcome (Beugelsdijk et al., 2017). Sometimes, the use of “context” goes beyond the relationship between variable conditions and outcomes to include the explanatory mechanism linking them; for example, Welch, Piekkari, Plakoyiannaki and Paavilainen-Mäntymäki (2011: 741) define context as “the contingent conditions that, in combination with a causal mechanism, produce an outcome.” We think these conceptualizations are useful, but can be confusing for qualitative scholars who are asked by reviewers to contextualize their research. In particular, these conceptualizations do not distinguish whether the research context is on the empirical plane or the theoretical plane, and where the researcher fits in.

When writing papers and reviewing papers written by other authors, we find it useful to distinguish between the empirical research setting, the theoretical research context, and contextualization. We depict this diagrammatically in Figure 1. In thinking through the differences, it is helpful to start with simple dictionary definitions. First, the Cambridge Dictionary (2022) defines setting as the time, place, and materiality where something happens. We use the term “empirical setting” to reinforce that this encompasses the empirical. As shown in Figure 1, the empirical setting includes the phenomena being studied and the participants, or informants, who are providing data about it.

Second, the Cambridge Dictionary (2022) defines context as the situation within which something exists or happens, which helps to explain its meaning. Consistent with this, we believe that a research context goes beyond the physicality and temporality of the observed empirical setting, to have theoretical meaning associated with it. We view this added layer of explanation and meaning as being on the theoretical plane. Accordingly, we view the research context as identifying the theoretical category of which the empirical setting is an instance. For example, the empirical setting might be a company with a digital platform to manage medical data, and the theoretical category might be companies that are managing diverse regulatory regimes. When authors report their findings, they specify the context of their research using this theoretical category.

Contextualization therefore involves categorization, or the linking of observations from the empirical setting to theoretical categories (e.g., Cornelissen, Höllerer, & Seidl, 2021; Grodal, Anteby, & Holm, 2021). This is normally carried out by comparing the empirical data (something new) with known theoretical constructs and relationships (something borrowed), to uncover where they are not aligned (e.g., Becker, 1998; Vaughan, 1992). Good inductive and abductive qualitative research requires that researchers pay attention to the data rather than anticipate findings beforehand (e.g., Pratt, Kaplan, & Whittington, 2020). However, as depicted in Figure 1, the researcher’s understanding of the prior literature – their theoretical and methodological repertoire – is relevant to how they contextualize data from the empirical setting, and there is normally iteration among the data, prior literature, and emerging theoretical categories (e.g., Strauss & Corbin, 2008).

Further, it is important to acknowledge both that the empirical setting may change over the course of a research project and that participants are more than simple suppliers of data. They have their own reasons for participating and beliefs about what constitutes success when doing so, and can gain different understandings of the research over time. As the researcher and research participants react and interact, they can influence each other. Thus, the qualitative IB researcher may not only be outside the empirical setting they are framing as the research context but they may also simultaneously be an integral part of it, as depicted in Figure 1. This reciprocity is why it is essential for researchers to be reflexive and transparent in reporting their research, and to think carefully about how they should best characterize their involvement in the empirical setting (see Langley & Klag, 2019).

The Role of Prior Literature in Studying Under-Researched Empirical Settings

All empirical settings are characterized by many diverse attributes; for example, culture, history, economic conditions, temporality, and so on (Askegaard & Linnet, 2011). In contextualizing their research and aligning their research question with an empirical setting, researchers need to select which of these attributes are relevant to the theoretical contribution they can make. Certain attributes are likely to be more salient to a researcher because they are familiar with them from prior research. Since qualitative research findings are dependent on the perceptions and interpretive framework of the researcher who collects and analyzes the data (Van Maanen, 1979), a researcher’s theoretical understanding of prior literature is consequential. It is relevant to their contextualization of a project before it even starts because it influences their perceptions of why particular phenomena and/or participants are likely to be theoretically interesting to study. It continues to be relevant throughout the project, and to data collection and analysis because prior research is used to frame a paper’s new theoretical contributions (Ragin, 1992). Moreover, qualitative researchers often use an existing theory as a lens and conceptual vocabulary to guide their attention (Dolbec, Fischer, & Canniford, 2021). Thus, a researcher’s knowledge of prior literature suggests which conditions, explanations, and outcomes may be theoretically interesting, and which may not be.

The central role of prior literature in contextualization heightens the complexity of categorizing new-to-the-literature empirical settings in two ways. First, researchers may need to “unlearn,” or at least suspend, the theoretical ideas with which they are familiar, so that they are not biased towards well-known theoretical categories and thus miss the unique qualities of a novel empirical setting (Vaughan, 1992). The Barnard vignette (Figure 2) shows the difficulty that she and her co-author had in recognizing the importance of religion in their data. They initially dismissed participants’ comments about religion to the point that they did not even register the many mentions of religion during interviews. However, a side comment by a participant serendipitously linked religion to a theoretical category from institutional economics with which they were familiar. This triggered a chain of thoughts alerting the researchers to the importance of God in business in Africa. They started to pay attention to the hitherto ignored mentions of God in the workplace in their transcripts, and adjusted their interview guide going forward to prompt for it.

Figure 2
figure 2

Barnard vignette: Linking empirical insights with theoretical conversations.

Full size image

Such adjustments are not uncommon in qualitative data analysis. Through analyzing their data, researchers may come to see a misalignment between their data and their preconceived theoretical frameworks, and this facilitates the development of new theory (Locke, Golden-Biddle, & Feldman, 2008; Vaughan, 1992). An example of this in the IB literature is the “discovery” of international new ventures: firms that Oviatt and McDougall (1994) noticed were selling in foreign markets much earlier than existing theory at the time would have predicted.

Second, in order to make a contribution to an established body of literature, researchers need to categorize new constructs and relationships in a way that recognizes their uniqueness, while also associating them with known categories in extant literature. For example, the Japanese concept of shinise was novel in the literature to which Sasaki and her colleagues (Figure 3) were contributing, and they associated them with long-lasting, high-status traditional family firms (see Sasaki, Ravasi, & Micelotta, 2019). Similarly, Barnard and Mamabolo (Figure 2) theorized religion as a unique category associated with a familiar category from institutional theory (see Barnard & Mamabolo, 2022). In this case, the authors benefited from learning how scholars from another field, the sociology of religion, theorized the phenomenon. The challenge – and also the opportunity – for researchers is to construct novel theoretical insights that resonate with established literature, while retaining the authenticity of the new empirical setting and the voices of participants.

Figure 3
figure 3

Sasaki vignette: Contextualizing while retaining the authenticity of the setting.

Full size image

Prior literature provides not only theoretical but also methodological direction. Prior literature on research methods directs attention as to how high-caliber, rigorous qualitative research should be carried out and identifies appropriate practices that can be used. At the same time, other important actors in the research ecosystem – such as funding agencies and institutional review boards – provide guidelines for the collection and analysis of qualitative data. Over time, these prescriptions change (Reuber & Fischer, 2022), but they tend to change slowly and at any given time it is not difficult for researchers to understand methodological norms and expectations. Often more difficult is applying these norms and expectations in empirical settings where they may not fit well. For example, the Couper vignette (Figure 4) describes the chaos of the research setting for an interview with a busy Chinese executive, and the Alkhaled vignette (Figure 5) shows the difficulty in following established ethics procedures when interviewing refugees. The current emphasis on methodological transparency rather than rules in qualitative methods (e.g., Jarzabkowski et al., 2021; Pratt et al., 2020) supports variation in practices across empirical settings, which encourages discussion of these issues in the Methods section of papers. Although some qualitative IB researchers do disclose explicitly the variations to conventional methods-related practices that were necessary to study a particular empirical setting (e.g., Couper, 2019; Seriki, Hoegl, & Parboteeah, 2010; Yagi & Kleinberg, 2011), many do not and we think this avoidance can reduce the richness and transparency of the narrative.

Figure 4
figure 4

Couper vignette: Navigating unexpected data collection conditions.

Full size image
Figure 5
figure 5

Alkhaled vignette: Managing multiple and complex researcher identities.

Full size image

The Role of the Researcher in Studying Under-Researched Empirical Settings

In addition to being the orchestrator of research activities, a qualitative IB researcher serves as an important instrument to collect interview and observational data (McCracken, 1988; Pratt, Kaplan & Whittinton, 2020; Van Maanen, 1979). They need to be open to opportunities that arise during interviews and observation sessions in order to follow promising avenues of inquiry, and this may require on-the-spot modification of protocols. As Pratt et al., (2020: 7) point out, researchers “would not be doing good research if they robotically ran through an interview protocol.” Thus, the interaction between researcher and participant are consequential for data collection and interpretation. This interaction is likely to be affected by a participant’s perceptions of a researcher’s personal characteristics, such as age, nationality, and status, and by the relationship between the two individuals, which may evolve over time (see, e.g., Michailova, 2004; Tsang, 1998).

One often-discussed dimension on which researchers vary is whether they are an insider or an outsider with respect to the empirical setting being studied. Prior literature has identified advantages and disadvantages of both insidership and outsidership. Insiders understand the etiquette of social situations, and have the cultural competence to establish trust and rapport with participants and interpret their responses accurately (Karra & Phillips, 2008; Pelzang & Hutchinson, 2018). They can understand complex local concepts and recognize the political and cultural sensitivity associated with certain lines of research (Yang & Lê, 2008). For example, the Sasaki vignette (Figure 3) emphasizes the importance of insidership in understanding key concepts relevant to long-lived Japanese craft firms. And the Alkhaled vignette (Figure 5) shows that insidership can build trust between the researcher and participants from vulnerable groups, such as women refugees, who are afraid of sharing their stories and identities with “outsiders,” when they are seeking political asylum in their host communities.

However, insiders may have less critical distance from the empirical phenomenon and participants than do outsiders. As we discuss later, this may result in role conflicts with participants, and/or pressure to conform with shared cultural norms. Moreover, familiarity may render a phenomenon mundane and taken for granted (Karra & Phillips, 2008). This was the case in the Barnard vignette (Figure 2), where both researchers were well aware of the importance of religion in Africa and, as a result, overlooked references to it in their data until their attention was triggered by a chance association.

Being perceived as an outsider by participants can also be beneficial. In the Sasaki vignette (Figure 3) and the Couper vignette (Figure 4), participants provided detailed explanations of their behavior so that an outsider researcher could better understand it, resulting in rich data. The Couper vignette also shows that knowledgeable outsiders can benefit from their outsider status when they exceed participants’ low expectations of outsiders. Couper was able to build trust quickly with Chinese participants because she exceeded their stereotyped expectations of Caucasian researchers.

Being an insider or an outsider is not a dichotomy; researchers are rarely at one extreme or the other. Moreover, although the labels are useful as a broad framework to think about the benefits and drawbacks of insidership and outsidership, even a continuum is too simplistic to capture the multiple, complex and changing identities that an IB researcher has with respect to an empirical setting. For example, the Sasaki and Alkhaled vignettes (Figures 3 and 5, respectively) illustrate the benefits of a bicultural, or hybrid, background (see also Tung, 2008), and the Couper vignette (Figure 4) highlights the benefits of being a “knowledgeable outsider.” As the Couper and Alkhaled vignettes show, a researcher can hold multiple identities in the eyes of participants that are overlapping and even conflicting. It is therefore necessary for the researcher to be reflexive in managing their identities during the research process.

To summarize, contextualizing qualitative IB research is an evolving and dynamic process involving the empirical setting, the researcher, and the researcher’s understanding of prior literature. This is always true of qualitative research, but it can be particularly challenging when a researcher is studying an empirical setting that is under-represented in prior IB research. In the following sections of the article, we highlight three distinct challenges related to the contextualization process: managing researcher identities, navigating unfamiliar data gathering conditions, and theorizing the uniqueness of novel empirical settings. While the first two challenges are mainly applicable to data collection, and the third one to data analysis and interpretation, they are inter-related because qualitative data collection and analysis are comprised of iterative and intertwined processes. In the remainder of the paper, we suggest practices that have been found useful in managing these challenges. This “toolkit” for authors, editors, and reviewers of qualitative IB research is summarized in Table 1.

Table 1 Toolkit of practices for qualitative researchers studying novel empirical settings
Full size table

CHALLENGES OF MANAGING RESEARCHER IDENTITIES

As emphasized in the previous section, the researcher is the actor who contextualizes a research project, interpreting both the empirical phenomenon under study and the prior literature that may be relevant to it. At the same time, the researcher is also the primary instrument for the collection and analysis of qualitative data, and, to be effective, needs to manage the insider and outsider aspects of multiple identities reflexively and relationally. Since researchers and participants influence each other, this requires researchers to be sensitive to the way they should position themselves vis-à-vis participants and to the dynamics of researcher-participant interaction.

With respect to insidership and outsidership, it has long been recognized that the benefits of both can be gained by the establishment of research teams composed of insiders and outsiders (e.g., Boyacigiller & Adler, 1991; Peterson, 2001). For example, the process of theorizing described in the Sasaki vignette (Figure 3) was facilitated by communication and negotiation with a collaborator who was an outsider. Researchers are better equipped to find collaborators who complement them when they understand their own position and how participants perceive them.

Recognizing that the researcher–participant relationship is dynamic and situational can help researchers manage their identities over time. For example, outsidership can put constraints on access to participants, awareness of cultural sensitivities, and interpretation of the data. Even though Sasaki was half-Japanese, she felt all of these constraints (see Figure 3). She managed the outsidership aspects of her identity in several ways. Although she had organized her first round of data collection by sending e-mails to case companies without introductions, and relying on contacts of family friends, for the second round of data collection she obtained a visiting researcher position at a top local university. This gave her local legitimacy when trying to set up interviews. She received advice from an insider on how to conduct interviews, including how she should introduce herself, how to exchange business cards, and how to manage her body language. The insider also ensured that her e-mails were written in the right tone. Further, she had a script ready to explain any deficiencies that Japanese participants might pick up on, which helped to set the expectations of participants. Sasaki started every interview with the following sentence: “Before starting the interview, I would like to apologize in advance, as being half-Japanese, I may make some mistakes during the interview and not fully manage the polite form of speech.” Because she had never worked for a Japanese organization, her language was only at the high-school level, and she had not had much opportunity to learn the respectful form of speaking in formal settings. Starting an interview with this disclaimer allowed her to make mistakes and turn otherwise awkward situations into moments of curiosity. Asking “stupid” or “too open” or even “rude” questions became an asset, because participants tried to respond in a way that a “half-foreigner” could understand. Therefore, being seen as a half-outsider became an asset, because it facilitated access to meanings that were local to the empirical setting.

This same vignette shows how a researcher’s identity can change over time as their relationships with participants change, and as their familiarity with the phenomenon under investigation changes. The project involved three data-gathering visits to Japan over eight months. Over subsequent rounds of data collection, Sasaki was increasingly invited to socialize with participants. She got to know them informally and even developed friendships with some. Over time, she became more of an insider to the empirical setting.

Another challenge arises when participants perceive a researcher as having multiple identities, potentially with conflicts between them. One example of this is from the Couper vignette (Figure 4), where a Chinese participant unexpectedly drew Couper into his interview responses, by comparing a business partner unfavorably with her, and in the presence of other participants. Given her knowledgeable outsider status in China, Couper was regularly asked for advice on conducting business between the UK and China, and even for help in identifying potential partners in the UK. To manage the multiple roles of researcher and advisor, she made sure to offer advice or help after data collection, so that it did not influence participants' responses during interviews. As her research progressed, Couper realised that Chinese participants saw the relationship as a “two-way street,” where they helped with her research and expected some help from her in return. Accordingly, after the research ended, she arranged a debriefing session to share her results. For one dyad, she was able to highlight particular aspects of the relationship she had observed where UK and Chinese partners had misunderstood each other's situation or behavior, and this had led to erosion of trust. By explaining why the misunderstandings had happened, she was able to help them restore some level of trust. This also showed participants that their voices had been heard.

A second example of conflict between researcher identities is from Alkhaled’s vignette (Figure 5). The Syrian refugees whom Alkhaled studied wanted her to champion them and publicize their plight. They also asked her to share their digital platforms where they were (anonymously) advertising and selling their craftwork. Obligated to follow her university’s code of ethical conduct and protect the anonymity of participants, Alkhaled was hesitant to publicize their work. However, her experience in the field made her appreciate their personal, political, and financial desperation. Like Couper, she felt that the relationship between the researcher and participants could not be a “one way-street,” where the researcher takes information with little return to the participant. Given the refugee women’s desperate state, Alkhaled was compelled to give back in return for them sharing their stories. She did so by writing up their stories in media articles, which she then shared with them as proof that she had stuck to her promise to “let the world hear their voices.” She also shared their anonymous social media sites, as well as paper-based marketing material, with her local network in Amman. This resulted in several of the women gaining catering contracts with local restaurants, and others being put in touch with supportive charities and refugee networks. Alkhaled’s actions deepened the trust of participants, which consequently provided a platform for longitudinal ethnographic work.

The Alkhaled vignette points out that it is not only researchers who recognize conflicting identities, but that participants may do so as well. For example, the vignette shows that Alkhaled carefully cultivated her identity as a “Syrian girl,” to be viewed as an insider like her participants, and that this was valuable in building trust with them. However, this trust was threatened by her outsider identity as a Western researcher when she asked them to sign consent forms. To manage this identity conflict, she had to find a more suitable way to obtain consent (see Figure 5).

In these instances, participants valued the multiple roles of the researcher. They perceived the researchers to have more information and access to a wider forum and array of actors than they did. This may be more likely to happen in under-researched empirical settings, in which participants have little exposure to researchers and may have an inflated perception of researchers’ clout and capabilities. In managing their multiple identities, it is important for the researcher to reflect on their position of power and greater access to resources from which participants may benefit. If researchers can build reciprocal give-and-take relationships with participants, they may be able to develop deeper and more trusting relationships, which benefit their research agenda as well as being personally meaningful.

CHALLENGES OF UNFAMILIAR DATA GATHERING CONDITIONS

The research methods and practices for data gathering that are familiar to qualitative IB researchers have been developed for use within relatively homogeneous empirical settings, and, therefore, reflect the norms and expectations of what best research practice in these settings should involve. Challenges in following these accepted but often tacit norms may become evident to researchers only while they are collecting data in settings with different characteristics (e.g., Michailova, 2004). They can impact the contextualization of the findings by constraining the quality and quantity of data collected. In this section, we describe challenges that we have encountered and provide suggestions for how researchers can navigate them.

A dominant theme underlying the literature on data collection practices is the importance of planning and deliberate design. There is often mention made of the possibility of an unexpected event – for example, if a participant opens up a novel line of discussion – but the underlying assumption is that the data collection process is scripted and controlled by the researcher. For example, McCracken (1988:41) emphasizes the scripted nature of a pre-set interview protocol:

"The interview itself will open with a carefully contrived section in which respondent anxieties are laid to rest. The grand-tour questions and prompting strategies are then set in train and the interviewer must labor to identify key terms, minimize respondent distortion, choose the most promising avenues of inquiry, and listen for material that is indexed by respondent testimony but not made explicit in it. All of this activity must be set in a generous time-frame in order to let respondents tell their own story in their own terms."

Methodological authorities also emphasize that it is important for researchers to design data collection around specific participants. Researchers are encouraged to learn about participants and their situations ahead of time, to have the background knowledge to understand their vantage point, and to tailor questions to them (e.g., Rubin & Rubin, 2012: 60–61). Further, researchers may plan to collect data from multiple participants simultaneously for specific purposes. For instance, focus groups can provide data on multiple perspectives (e.g., Yin, 2018: 120), and observing a group conversation, such as a Board meeting, can provide data on group-level effects on outcomes (e.g., Kyprianou, Graebner & Rindova, 2015). Overall, the expectations are that interviews will take place in participants’ places of work, private meeting rooms, or online, according to pre-agreed interview schedules and named participants, and that there will be uninterrupted time for the “respondents to tell their own story in their own terms,” as McCracken (1988) describes in the quote above.

One challenge to such precepts is that data collection in some empirical settings can be much more fragmented and chaotic than norms and guidelines suggest, and less controllable by the researcher. When participants do not have an online profile, it is difficult to pre-plan interviews. Barnard and Mamabolo (Figure 2) did an initial set of site visits specifically to gather data to help them select participants for the project and to develop relationships to aid in setting up formal interviews. The refugees in the Alkhaled vignette (Figure 5) were not comfortable with sharing identifying information with the researcher until they met her face-to-face and trusted that she was a “fellow Syrian”, and therefore did not pose a threat to their informal (unregistered/illegal) businesses or precarious immigration status. Alkhaled had a local insider as a research assistant who was already on the ground and could introduce the researcher and participants to each other and help to establish trust between them.

The Couper vignette (Figure 4) shows the difficulty of following a pre-scripted interview plan when participants require interviews to take place in unexpected venues, with unanticipated individuals present, and with frequent interruptions (see also Couper, 2019). This example is consistent with a polychronic approach to time characteristic of China, but such a situation can be stressful to new qualitative researchers, because it does not conform to textbook models of how interviews should unfold (see also Tsang, 1998). However, unexpected benefits arose from the chaotic nature of Couper’s interviews. It prompted negotiation of future access for repeat interviews of key individuals, in case the quality of data from a single interview was insufficient. Also, spending long periods of time physically trailing participants around provided the opportunity to observe them in the conduct of their daily business and social activities. Couper’s interviews of British participants aligned with common interview norms, and did not generate similar opportunities for observational data collection and a deeper understanding of participants’ work environment. Even when interviews with Chinese participants were conducted in more formal venues, such as offices, the participants would often suggest that the formal interview be followed by a more informal and social get-together. Such get-togethers, often at a restaurant, yielded additional and at times more valuable data. Since Couper was unable to record them, she needed to be diligent in recording detailed and comprehensive field notes after they occurred.

This example suggests that data-gathering challenges can become opportunities. In particular, when conducting interviews in new empirical settings, researchers should be open to serendipitous opportunities for observation to acquire richer data. As another example, while doing site visits in Zimbabwe, Barnard and Mamabolo (Figure 2) found that observational data about the sites, such as whether the gardens were tended or the walls freshly painted, provided insights into the financial well-being of the firm. This contrasted with visits in the other, relatively more affluent, African countries, and provided rich evidence about the lived experience of institutional dysfunction.

A second difficulty in following data collection norms relates to the tone of the interview. Again, established guidelines assume control by the researcher. Rubin and Rubin (2012: 31–34) distinguish between cultural interviews, which focus on norms, values, and taken-for-granted rules, and topical interviews, which focus more narrowly on particular issues or events, although they recognize overlap between them. Researchers are advised to be more relaxed in cultural interviews, letting participants tell their own stories, and more directive in topic interviews, eliciting detail about the topic under investigation. In our experience, all interviews are to a large extent cultural interviews, in that participants’ responses are based on their norms, values, and taken-for-granted rules, and so these should underpin the tone of an interview. Understanding and interpreting responses can be more difficult to the extent that the researcher is an outsider, because it can be difficult to interpret non-verbal responses, such as gestures and body language, and to understand the meaning of the interview’s content from a participant’s perspective. Piloting the data collection protocols with individuals who have characteristics in common with the participants can suggest improvements and improve the quality of the data collected.

Further, the tone of an interview may take an unexpected turn. This can be beneficial for the research project when it adds richness to the data and the insights that can be gained. This happened when Mreji and Barnard (2021) encountered an unexpected level of emotion – anger and sadness – during their interviews of returnee entrepreneurs in Kenya. To help put respondents at ease, the first author – herself a returnee – shared some of the distressing events she had encountered. The intensity of the emotion in the data reinforced the authors’ conclusions about the liabilities that returnee entrepreneurs can face, and so was an interview challenge that proved to be valuable for framing the phenomenon theoretically.

When researchers anticipate uncomfortable situations, from pilot or early interviews, they can devise a method to navigate them. For example, as part of her pre-interview planning, Alkhaled (see Figure 5) prepared a “local Syrian girl” narrative to build a rapport with the women refugee-entrepreneurs she interviewed. However, she did not anticipate that some of their husbands would feel uncomfortable and emasculated in the presence of a Syrian woman researcher. They stated that they would have felt “less shame” about their lack of ability to provide for their families in front of a Western woman. She was able to draw on her mixed identity to sympathize with the pressures on men to be the bread winners in Arab society, and report that observing men in supportive roles of their wives’ businesses is very much the norm outside of the Middle East and is not associated with emasculation but family teamwork. In this situation, to manage the tone of the interview and put participants at ease, her role went beyond letting participants tell their story, to providing them with a Western interpretation of their story.

A third difficulty in following data collection norms relates to norms and practices around research ethics. Norms underpinning ethics approval processes are often inherited from scientific/medical research, where the risk of harm to patients is reduced through the protection of anonymity and extensive formal written consent. These practices have led to requirements that organizations and participants participating in IB research projects be anonymous, and sign written consent forms after reading documentation about the study. The Academy of International Business Journals Code of Ethics (2020) governing JIBS and Journal of International Business Policy states “Authors have a responsibility to preserve and protect the privacy, dignity, well-being and freedom of human subjects and research participants. Informed consent should be sought from all human subjects, and if confidentiality or anonymity is requested, it should be honored” (section 3.71) and then goes on to say that “Manuscripts involving human subjects (surveys, simulations, interviews) should comply with the relevant Human Subject Protocol requirements at the Author's (Authors’) university(ies)” (section 3.7.2). Therefore, the ethics review board at a researcher’s university is the relevant decision-making body when a researcher wishes to request adjustments to standard practices.

Conventional ethics norms and practices may be questioned by participants, and there is a growing recognition of the need for ongoing negotiation with participants and ethics review boards (see, e.g., Bell & Kothiyal, 2018). In some research settings, participants can be wary of signing written documents, so the researcher can record verbal consent, as Couper (Figure 4) did for some Chinese participants. In other settings, not all participants may be literate and thus able to read or sign written consent forms. When facing this dilemma, Alkhaled (Figure 5) read aloud an Arabic translation of the consent form and participants put an X on a paper version of it. Once she had provided an explanation of how and why the processes of obtaining informed consent needed to be adapted, her university’s ethics board approved the adaptations.

Researchers also need to be aware that participants may challenge the norm of anonymity. For example, there has been a long history of cultural appropriation of African ideas (e.g., Cabrita, 2020), and knowledge of this has led some of Barnard’s African participants to be offended by the suggestion that their names would be removed from research data and outputs. Researchers can initiate a discussion about the purpose of the research project and the purpose of anonymity to elicit participants’ concerns and preferences. The informed consent document will reflect the agreement between the researcher and the participant. When participants do not want anonymity, managing the ethics process can be similar to the ethics process for developing teaching cases where the name of the firms and the protagonists are known. Transcripts are shared with participants before they are analyzed to verify that the information is correct. If there are multiple participants and only some of them prefer anonymity, anonymity should be maintained for everyone. In this situation, the researcher needs to explain to participants why all participants need to be treated the same, and reiterate that they can withdrew from the project if they no longer wish to participate because their contribution cannot be attributed to them.

CHALLENGES OF THEORIZING THE UNIQUENESS OF NOVEL SETTINGS

Buckley and Chapman point out that participants in an empirical setting use their own categories to characterize it (Buckley & Chapman, 1997). They argue that researchers may lose what is important and unique about a setting when they  try to map these empirical categories into familiar theoretical categories. There is a risk of an overly general and simplistic set of concepts that obscure the richness of the setting being investigated. This reduces the potential for theory development that comes from studying more diverse phenomena. At the same time, the potential for theory development is reduced if the empirical setting is viewed as literally unique, a “unicorn,” and the research findings cannot be transferred to other empirical settings. Therefore, a contextualization challenge for qualitative IB scholars, especially those studying under-researched phenomena, is retaining the uniqueness and authenticity of an empirical setting that is unfamiliar to other researchers, while, at the same time, expanding a theoretical conversation in the literature that extends beyond that setting (see Walton, 1992).

In this section, we suggest four practices to help in navigating this core tension of contextualization. The first practice is setting up a collaboration in which researchers have distinct roles, as illustrated in the Sasaki vignette (Figure 3). One collaborator is an insider who is intimately familiar with the empirical setting and the data. A second collaborate is an outsider to the empirical setting but an insider with respect to the theoretical conversation they are hoping to join. In the early stages of data analysis, the insider researcher codes and categorizes the empirical data. This insider can keep the empirical setting – and participants’ voices – alive by labeling the emerging categories as close to the language of participants as possible. To do this, the researcher draws on insights from the data collected, insider knowledge of the empirical setting, and familiarity with prior literature. This is important because “without intimate knowledge of a target culture, the lack of conceptual or functional equivalencies may elude a researcher" (González & Lincoln, 2006: 3). In Sasaki’s case, the insider knowledge was gained not only by being half-Japanese and understanding the language and the culture but also by being embedded in the empirical setting of long-living Japanese firms for a prolonged period.

Important in coding is understanding how “they” (the participants) want to express themselves, compared to how “we” (the research community) want to understand them. In later stages of analysis, the challenge is to retain the essence of these meanings while developing theoretical categories that are more abstract. The outsider collaborator takes on a larger role. Throughout the analytic process of iterating among the data, prior literature, and emerging theoretical categories (e.g., Strauss & Corbin, 2008), there is continuous negotiation between the two collaborators until there is a consensus that the theoretical model reflects the uniqueness of the empirical setting, while making a novel theoretical contribution to the extant literature. We encourage researchers to “doubt” the emerging theoretical model throughout this process, because “the living state of doubt drives and energizes us to generate possibilities, try them out, modify, transform, or abandon them, try again, and so on until new concepts or patterns are generated that productively satisfy our doubt” (Locke et al., 2008: 908). We also encourage them to discuss the final theoretical model with participants to ensure that it is truthful to their experience (González & Lincoln, 2006: 7; Nag, Corley, & Gioia, 2007: 829).

The second practice is retaining the language of the empirical setting by refraining from translating concepts that are not amenable to translation. It is not always possible to find accurate translations; for example, the Couper vignette (Figure 4) points out her struggle to communicate to an English co-author the meaning of two distinct Chinese terms that have only one meaning in English. In the Sasaki vignette (Figure 3), the insider researcher, who was bilingual in Japanese and English, could not find a suitable English word to use for certain Japanese concepts. When it proved to be impossible to fully explain the concept to her outsider collaborator, they decided to keep the Japanese words intact in the paper, rather than substituting an incorrect English word. The untranslated words were written in italics and definitions were provided. This avoided a loss of meaning when the research was written up for publication.

The third practice that can help in navigating the core tension of contextualization is looking for ideas to appropriate from outside the theoretical conversation to which the researchers are contributing. Sasaki found appropriable concepts in Selznick’s (1957) ideas (see Figure 3). Appropriable concepts are those that can be recast with new meaning because they are out-dated or ill-defined. They are attractive for abstracting theoretical insights from empirical settings that are under-represented in the literature because, while they represent existing theoretical categories, they also have room to be refurbished with fresh meaning. This is less likely to be the case with well-used and well-defined theoretical categories. Researchers can look outside the IB literature for appropriable concepts; for example, Barnard (Figure 2) drew on past research on the sociology of religion, and Sasaki (Figure 3) drew on Japanese literature about the industry she was studying. In looking for appropriable concepts in other literatures, we recommend that researchers not take them at face value, but carefully understand the similarities and differences between them and the empirical concepts they have identified.

The fourth practice is to establish clear boundary conditions, clarifying the scope of the theory generated. This is always crucial when theorizing from qualitative data (see Eisenhardt, 2021), but it is more challenging and more important when researchers are contextualizing an empirical setting that is unfamiliar in the literature. The intentional retention of setting-specific nuances and participants’ voices throughout data analysis has the potential to give the impression that the transferability of theoretical insights is limited. Reviewers may view the findings as unique to an unfamiliar empirical setting and assess the paper’s theoretical contribution unfavorably. It is therefore especially important for researchers to address the issue of transferability explicitly, by delineating the empirical and the theoretical boundary conditions of their findings. For example, Sasaki et al.’s (2019) study is not merely a case of how a set of long-living heritage crafts firms collectively maintain a high social status in Kyoto, Japan, but is also a case more broadly of how local communities maintain status hierarchies among firms. By explicitly delineating the boundary conditions of the study, it is clear their findings are applicable to other empirical settings, such as European heritage-based clusters.

SUMMARY AND CONCLUSIONS

International business scholars are hungry to understand new, and often personally-relevant, empirical phenomena. Recent conference themes, journal special issues, webinars, and shared interest groups signal the importance of studying new phenomena that have hitherto received limited attention from established scholarship. Somewhat surprisingly, there has been less acknowledgement that this may heighten challenges associated with how qualitative scholars can study under-researched phenomena in a rigorous way to make a novel theoretical contribution and therefore be welcome in top journals. We hope that practices summarized in the toolkit in Table 1 are helpful to qualitative IB researchers in addressing these challenges and turning them into opportunities for more diverse theorization.

In closing, we note that the Methods section is the “heart” of a paper, located in the middle of the manuscript and connecting prior literature (something borrowed) with the findings (something new). We argue that contextualization of qualitative research takes place in this “heart.” It is where the empirical setting is introduced, and where the practices used for data gathering and analysis are explained. In other words, the Methods section is where the theoretical world and the empirical world collide. Reviewers should be comfortable with unexpected collisions and authors should fully explain them. We hope that such reflexivity and transparency can pave the way for more diverse and contextualized qualitative international business research, while retaining the rigor and transferability required for publication.

Notes

  1. 1.

    We see issues related to language and translation as important to the overall objective of increasing the diversity of IB-related phenomena studied by qualitative IB researchers (see Brannen, Piekkari, & Tietze, 2014; Chidlow, Plakoyiannaki, & Welch, 2014; Tenzer, Terjesen, & Harzing, 2017), but separable from the issues discussed here. The researchers we draw on were able to speak fluently with participants in their preferred language and did not need to rely on translation by others.