Abstract
In this article, we address the question of how scholars innovate research methods, using the example of qualitative research in strategy and management. We define innovation as “the intentional introduction and application of ideas, processes, products or procedures, new to the relevant unit of adoption, designed to significantly benefit” (building on West & Farr, 1990, p. 9). This definition explicitly acknowledges the application of new ideas, rather than just the creative process itself (Acar, Tarakci, & van Knippenberg, 2019). Our focus is specifically on innovation in research methods, in other words, the introduction and application of (parts of) research methods that are new or seldom used in a research field. Critically, such innovation does not simply entail changing the content of a method, that is, applying an existing method to a new empirical or theoretical context. Rather, in the process of so doing, it also transforms the research tool and how it is used.
It is critically important to understand innovation in research methods for various reasons. First, methodological innovation is often credited for generating new theoretical insight (Bansal & Corley, 2011; Bansal, Smith, & Vaara, 2018; Ketchen, Boyd, & Bergh, 2008; Lê & Schmid, 2019; Locke, Feldman, & Golden-Biddle, 2016; Reay, Zafar, Monteiro, & Glaser, 2019). Indeed, method innovation is central to creating the “generative capacity” (Gergen, 1978), “disciplined imagination” (Weick, 1989), or “discovery” (Locke et al., 2016) needed to make significant theoretical contributions. Thus, it is perhaps not surprising that editorials and reviews on qualitative methods highlight new or unconventional qualitative methods as an effective strategy to achieving high-impact publications (Bartunek, Rynes, & Ireland, 2006; Bluhm, Harman, Lee, & Mitchell, 2011; Elsbach & Kramer, 2016; Lê & Schmid, 2019).
Second, institutional pressures can discourage methodological innovation. For example, there is currently a trend toward convergence of qualitative research through the emergence of templates in management research (editorial, this issue; Langley & Abdallah, 2011). Indeed, leading scholars are lamenting the fact that repeated application of similar ways of conducting and presenting research has led the field down a path of standardization in which certain elements, models, or templates for qualitative research dominate, reducing methodological pluralism (Bansal & Corley, 2011; Bansal et al., 2018; Langley & Abdallah, 2011; Reay et al., 2019). Yet, it is precisely this innovativeness that has emerged some of the key contributions in the management field. Indeed, scholars have always emphasized the diversity and flexibility of qualitative methods as key strengths of this type of method (e.g., Glaser & Strauss, 1967; Langley, 1999; Pratt, 2009; Suddaby, 2006; Wiles, Crow, & Pain, 2011), actively arguing against homogeneity of methods: “Constructing a single boilerplate is not something qualitative researchers strive to achieve” (Pratt, 2009, p. 857). For instance, Gersick (2016, p. 317) encourages “trial-and-error tinkering”: “Get all the help you can from existing methods, but be willing to try making up something new.”
Third, despite the value of innovation, we actually know relatively little about the actual practice of research method innovation. Existing work presents exemplars of innovative methods along the research process from research setting to design, forms of data, data collection, and analysis (cf. Elsbach & Kramer, 2016). Other work (Bansal & Corley, 2011) calls for innovating methods via new research questions, nontraditional data sources (e.g., narratives, photographs, organizational artifacts, nonverbal interaction), new and varied data analyses (e.g., longitudinal and multilevel process research, including innovative approaches to mixed methods that give equal weight to qualitative methods), and new or unconventional data display and research presentation. While this provides some useful pointers with regard to areas of innovation, it says little about the actual process of innovation. Yet, it is precisely this process that we need to understand if we want to explain and encourage innovation.
This study therefore explores how scholars innovate methods, focusing in particular on qualitative research in strategy and management. As outlined above, creative use and innovation of methods are central to qualitative methods. Qualitative research also reflects common themes regarding methodological innovation. One such recurring theme is the tension between converging around standards and diversity in methods. This is particularly true for the strategic management field, which has been both criticized for a lack of qualitative research (Bettis, 1991) and praised for the emergence of novel approaches to qualitative research (e.g., Langley & Abdallah, 2011).
Drawing on a systematic 10-year review of qualitative strategy research, we identify and closely examine innovative papers, that is, papers that demonstrate novelty in every part of the research process. We work through these papers in detail to demonstrate their innovativeness, highlighting concrete ways in which scholars have innovated the interconnected parts of the research process: data generation, data analysis, and presentation. Based on this in-depth review of exemplar studies, we develop a two-layered “iceberg” model of method innovation that identifies innovative research practices and emerges the underlying assumptions that underpin these practices.
Building on the analogy of an iceberg, where only the tip of the iceberg is visible and much of the mass remains concealed under water, our empirically grounded framework deconstructs the process of innovating research methods into two layers. Specifically, we found that, in our innovation exemplars, researchers made use of new or seldom applied research tools as part of their study. These research tools and how they were used were explicitly communicated in the published papers. This overtly visible innovative use of research tools (Layer 1), however, was guided by implicit principles of innovating methods (Layer 2) that—though central to the actual research practice—were often not made explicit by the authors. While a summary of innovating practices is helpful, successful innovation depends also on understanding the underlying principles of innovating methods, which act to provide “the rules” by which tools can be effectively innovated. This is thus our core contribution. Specifically, we explicate four core innovation principles as part of our model. This enables us to more clearly delineate the process of innovation in research methods, thereby providing guidance and encouragement for a more creative use of methods that moves beyond templates.
In the remainder of this article, we first explain our approach to this study. We then showcase concrete examples of methodological innovation along the entire research process, highlighting specific innovative research practices. Finally, we explain our innovation iceberg, outlining the core principles that guided overtly visible method innovation in published research papers.
Our Approach
This article was inspired by something we noted while we were conducting a review of qualitative research methods: Qualitative research methods are diverse and innovative! Our initial review was directed at assessing the state-of-science in the use of qualitative methods within a field of management research (for further detail on our approach and findings in this review, please refer to Lê & Schmid, 2019). The main result of this review was a set of commonly used research approaches, that is, various ways in which qualitative work was conducted and presented in top tier journals. As part of this review, we noted both conservatism in the use of methods, that is, many scholars used a few well-established research approaches, and remarkable creativity in how scholars used research methods. Specifically, as we looked across our sample of papers, we noted a lot of innovation; certainly much more innovation than we were expecting based on the debate around the convergence of qualitative research methods and cautions against the overreliance on templates (Bansal & Corley, 2011; Bansal et al., 2018; Langley & Abdallah, 2011; Pratt, 2009; Reay et al., 2019). It was this observation that inspired our current paper: We wanted to better understand and showcase innovation in research methods.
In order to do so, we embarked on another review of qualitative research methods. Again, we began by restricting the scope of our study. First, as contributions are relative to specific research fields, and we wanted to ensure comparability across studies, we zoomed in on a specific subfield of management: strategic management. Next, we restricted the time horizon. We chose the last 10 years (2008–2017), as recent reviews have reported sizable progress in qualitative management research during this period (Bluhm et al., 2011; Lê & Schmid, 2019) and because we wanted to include the most recent methods developments (e.g., multimodality; see Zamparini & Lurati, 2017). We also, in an effort to impose some quality control, sampled papers only from journals that are consistently rated top strategy and management journals in national and international journal rankings:
To explore in detail how scholars innovated methods, we probed our dataset for evidence of innovative methods. We began by closely reading each empirical paper, carefully studying it for evidence of innovation. We coded the papers based on seven domains (Glaser & Strauss, 1967; see also Strauss & Corbin, 1990): (a) contribution, (b) research perspective, (c) research approach, (d) research setting and case selection, (e) data sources and major stages/steps in data collection, (f) analytic strategies and tools, and (g) the presentation of findings. Having mapped dominant research approaches in a previous review (Lê & Schmid, 2019), we were now able to detect when scholars diverged from these designs through innovation. Our first step was to systematically record the innovation we observed in each paper, discussing this innovation within the research team. We then sought similarities and differences, grouping innovation into categories based on these.
An important insight from this analysis was that scholars generally did not engage in methodological innovation for its own sake, but rather did so in an intentional and targeted way that would emerge unique findings and answer novel research questions. Another observation was that papers differed in the degree of innovation. Building on key categories of innovating (see, e.g., Crossan & Apadyin, 2010; also Gopalakrishnan & Damanpour, 1997; Henderson & Clark, 1990), we noted that many papers engaged in incremental innovation, that is, minor changes within dominant research approaches, such as making more extensive use of archival data within a well-established comparative case study approach (e.g., Santos & Eisenhardt, 2009). We, however, also detected some studies that more significantly departed from established research practice. These latter studies included radical innovation, that is, they introduced and/or elaborated a research approach that was new or seldom used in the field of strategic management research. They also included architectural innovation, that is, featuring novel ways of combining new and well-established research approaches. Given our aspiration to explore and encourage methodological innovation, we focused on exemplars that featured radical innovating, at times combined with architectural innovating. Focusing on these “extreme” cases allowed us to make important features of how researchers innovated research methods more “transparently observable” (following a sampling strategy introduced by Pettigrew, 1987; see also Eisenhardt, 1989; Miles, Huberman, & Saldana, 2014; Yin, 2014). In making these choices, we acknowledge that the distinction between incremental and radical innovation is not a true dichotomy, but rather a gradual sliding scale.
We engaged in a detailed review of exemplar studies to further elaborate our understanding of the practice of innovating methods. In this highly iterative process, we refined our sample of studies and moved toward a mode of theory elaboration based on relating our emerging findings to established theory (cf. Locke, Golden-Biddle, & Feldman, 2008). As part of the process, we carefully reread and reviewed each exemplar paper and iterated our analytic categories several times, ultimately excluding categories that focused on theoretical innovation (such as, substantive contributions of the study) and settling on three highly visible areas of methods innovation: (a) data generation, including research setting and data sources, (b) data analysis, covering analytic process and strategies, and (c) display and presentation of findings.
We also returned to the literature on methodological innovation (e.g., Bansal & Corley, 2011; Cassell, Buehring, Symon, & Johnson, 2006; Elsbach & Kramer, 2016; Gersick, 2016; Wiles et al., 2011). While we noted that the value of focusing on innovation along the stages of the research process had previously been recorded (Elsbach & Kramer, 2016), our findings offered a more comprehensive representation of innovation. This provided an initial confirmation of our findings and helped us elaborate our own contribution. In particular, we used recent descriptions of methodological innovation to extend our sample of exemplar papers. For example, drawing on recent reviews of new methods in strategy and management research that highlighted and elaborated a historic turn in strategy research (e.g., Vaara & Lamberg, 2016), we included an exemplar of historical methods in our analysis to explore the historical embeddedness of strategic processes and practices.
Finally, we engaged practice theory to make sense of two additional findings that emerged during our in-depth review of exemplar studies. First, methodological innovation entailed specific ways to make creative use of novel or uncommon research methods. Drawing on the notion of tools-in-use (Jarzabkowski & Kaplan, 2015), we differentiated relatively generic research tools (e.g., video ethnography) from the specific ways in which researchers made use of these tools in their research practice (e.g., using handheld devices and engaging in multimodal analysis using partial transcriptions of videos and move/sequence analysis). As we aimed to provide hands-on guidance to researchers by deciphering the worked practice of innovating, we sought detailed evidence of the latter, that is, on specific, creative ways of making use of novel or uncommon research tools within exemplar papers. Delving deep into the papers confirmed the value of analyzing exemplar papers along the parts of the research process, as it helped flesh out the practice of innovating and add empirical richness to our explanations.
Second, we realized that visible innovation in research tools was not random. Rather, it was informed and guided by core innovation principles. For example, we noted that exemplar studies produced methodological innovation that was holistic in nature, that is, researchers introduced not just single changes in established research practice (e.g., using an uncommon data source), but multiple, consistent changes that cascaded across the entire research process. We thus sought to systematically detect key principles in our exemplar papers that were often not stated explicitly, yet manifest in how methodological innovation was implemented. We conceive these guiding principles as core intuitions that inform researchers on how they can reproduce and/or change methodological practices (building here on the elaboration of guiding principles in practice theory by Feldman & Orlikowski, 2011; Feldman & Worline, 2016). Critically, our emergence of these key principles of method innovation from a detailed review of the practice of innovating methods using exemplar studies, more closely aligns with the notion of guiding principles than related terms, for example, best practice (on this distinction, see Feldman & Orlikowski, 2011). Indeed, we did not seek to identify “best practice” or “ideal types.” Rather, we see value in showcasing the various innovation practices and principles within exemplars as a way of understanding some of the ways of innovating methods currently represented in the literature. We acknowledge that there are likely other ways to innovate and see our paper as a starting point to understanding the multiple ways of innovating that are possible. We use the remainder of this article to outline key exemplars of how researchers innovate.
Exemplars of Innovating Research Methods
In this section, we closely examine seven published research papers as exemplars of method innovation. As described above, these papers were selected based on a comprehensive, 10-year review of qualitative studies in strategy research. Since we were interested in how researchers significantly depart from established research practice, we focus here on exemplars that showcase radical innovation in research methods, that is, the researchers introduce or elaborate a new or uncommon research approach and/or novel ways to relate existing research approaches.
In what follows, we work through each of our primary exemplar papers in detail to demonstrate concrete ways in which scholars have innovated three interconnected parts of the research process: (a) data generation, including research setting and data sources, (b) data analysis, covering analytic process and strategies, and (c) display and presentation of findings. Zooming in on these studies enables us to showcase concrete examples of method innovation as a means of inspiring the creative use of methods. Table 1 provides summaries of the exemplar papers and the innovative research tools used in each study. It offers an overview so that the reader can decide on which exemplars to study in greater detail.
Seven Exemplars of Innovating Methods.
Our exemplars, while all innovating across the entire research process, showcase different ways of innovating. The first two exemplars illustrate novel ways for investigating activities and practice in organizations. Exemplar 1, Smets, Jarzabkowski, Burke, and Spee (2015), employs a global, team-based ethnography to investigate the microdynamics of balancing multiple institutional logics across multiple sites and over an extended period of time. Exemplar 2, LeBaron, Christianson, Garrett, and Ilan (2016) entails an ethnomethodological study harnessing the full communicative spectrum, including speech, gestures and body positioning, via a powerful multimodal analysis of video recorded routines.
The next three exemplars showcase innovative ways to engage research participants and the research field. Exemplar 3, Lüscher and Lewis (2008), employs action research to leverage collaboration in the research process, actively sparring with participants in order to encourage reflection and generate data. Exemplar 4, Bernstein (2012), utilizes deception in his study, placing confederates on the shop floor to work with and observe unsuspecting “subjects.” Exemplar 5, Michel (2007), draws on “insider” experience working at a Wall Street bank in order to increase empathy with and facilitate access to informants in the banking sector.
The last two exemplar studies center on novel ways of relating research traditions from different disciplines. Exemplar 6, Sonenshein (2010), combines multiple traditions of linguistic analysis, drawing together narrative and content analysis. Exemplar 7, Lubinski and Wadhwani (2020), introduces historical research methods to emerge a richer, contextualized understanding of organizational and management phenomena.
Exemplar 1—Harnessing the “Mundane” Through a Global, Team-Based Ethnography
A first exemplar of methodological innovation is the study by Smets et al. (2015) on the dynamic balancing of incompatible logics of action in reinsurance trading at Lloyds of London. Their institutional and practice-based research approach sees them combine the study of relatively standard macro-level phenomena (organizational and market outcomes) with the study of emerging micro-level phenomena (everyday mundane work).
Drawing on institutional studies (e.g., Scott, 2001; Thornton, Ocasio, & Lounsbury, 2015), the authors adopt an empirical focus on organizational responses to institutional complexity (Smets et al., 2015, p. 932). Creatively, they also draw on practice-based studies to emerge a simultaneous empirical focus on actions, as “practices are understood to be the primary building blocks of social reality” (Feldman & Orlikowski, 2011, p. 1241; also Schatzki, Knorr-Cetina, & Von Savigny, 2001). While practice-based studies are by no means new, and scholars often achieve their contributions through shared methodological preferences (Langley & Abdallah, 2011, pp. 220–221; also Feldman & Orlikowski, 2011), Smets et al. (2015) are particularly innovative in their approach. Specifically, the authors utilize new methods of data generation (global team-based ethnography of frontline workers), as well as novel analytic techniques (multimodal analyses involving materials and tools) and ways to display findings (the composite narrative). They thus introduce novelty across the entire research process.
Smets et al. (2015) utilized an impressive, large scale design. Specifically, they invented the “global team-based ethnography,” an internationally collaborative research design, which allowed them to study reinsurance practices across multiple international sites over the period of one year, spending a total of 180 days emerged in the field (Jarzabkowski, Bednarek, & Cabantous, 2015). This approach maximized their spread of data and enabled the research team to also undertake comparative national and international analysis.
The author team complemented this with other innovative data collection strategies, also delving “into the mundane everyday practices by which reinsurance underwriters assess risks, place capital, and, in doing so, balance the seemingly irreconcilable demands of the financial market and their Lloyd’s community” (Smets et al., 2015, p. 933); “we coded all the mundane micropractices we observed in underwriters’ daily work, from ‘putting on tie and coat’ to ‘generating loss curves,’ ‘calculating financial returns,’ ‘helping brokers out,’ ‘sharing gossip,’ or ‘lunching with peers’” (p. 939). This emphasis on the mundane brought about innovation. First, it stimulated a focus on different type of actors, specifically moving the authors to “re-direct attention from organizational leaders to a previously neglected class of actors: people on the frontline” (p. 939). Second, it enabled the authors to work with a different type of data. As part of their global ethnography of Lloyds, Smets et al. (2015) did not simply rely on the standard triad of observations, interviews and documents (Langley, 1999). Rather, they also recorded and video-recorded many, specific trading transactions (see Smets, Burke, Jarzabkowski, & Spee, 2014). The detailed recordings of particular events allowed them to capture the multifaceted and interconnected nature of practices. This critical research practice enabled them to observe not just participants’ discursive practices, but also the material, bodily, emotive, interactive, and spatial practices in reinsurance trading (see Smets et al., 2015, p. 938): “In our notes, we documented the material context of work, the information shared, and the emotional reactions sparked, such as people gesticulating, laughing, or raising their voices.”
Smets et al. (2015) also engaged in novel analytic techniques, specifically multimodal analyses involving materials and tools. While the authors identified a number of material practices, for example, “Jim opens the spreadsheet on his PC” (p. 949), they also identified “the locales where practices were being performed (e.g., office, trading floor, meeting)” (p. 940). They could thus match practices to locales, for example, “analytic micro-practices (e.g., calculating financial returns, generating loss curves, modeling) always occurred in the office” (p. 940) while “community-oriented practices (e.g., social talk with brokers at the box) only occurred at Lloyd’s” (p. 940).
The paper is equally innovative in the way it presents key findings. In particular, Smets et al. (2015) feature “a thick description of a ‘day in the life’ of a reinsurance underwriter” (p. 939). This “composite narrative” (Jarzabkowski, Bednarek, & Lê, 2014) diverges from standard research practice to capture narratives, as it is composed of representative elements of analysis, from multiple actors and across multiple days, that are subsequently woven into a single flowing narrative and presented as one coherent story in order to illustrate the core dynamics: “To illustrate the everyday enactment of these logics in practice, this section presents a composite narrative that exemplifies a ‘day in the life’ of Lloyd’s underwriter ‘Tim’ at SafeCo…this composite narrative presents the full breadth and depth of our data within a single evocative story intended to ‘render the actual—and to do so persuasively’ (Van Maanen, 2011: 232)” (p. 944).
This rich descriptive way to analyze and present data allows access to the micropractices of reinsurance trading, including insight into nuances like scheduling, core activities, and dress code: “At 11:15 a.m., Tim gets ready to go to Lloyd’s. He picks a tie from his drawer and puts it on, slips on his coat, which had been hanging over his chair, and checks his black shoes. He grimaces and quickly runs a tissue over a scuffmark. Finally, he gathers his files and, adjusting the knot on his tie, heads to the lift with the other underwriters” (p. 952). Such very rich data enabled the researchers to drill into the microdetail of practice, showing that something as slight as putting on a tie had major implications for interactive dynamics and, ultimately, organizational responses.
Exemplar 2—Harnessing the Interactive and Multimodal Nature of Communication in an Ethnomethodological Study
Another powerful exemplar of methodological innovation is the study by LeBaron et al. (2016). The authors investigate handover routines of intensive care unit (ICU) physicians by drawing on “two complementary research traditions: routine dynamics and ethnomethodology” (p. 514). While routine dynamics has recently seen a resurgence in the literature (cf. Feldman, Pentland, D’Adderio, & Lazaric, 2016), ethnomethodology remains a relatively rare approach. Ethnomethodology “foregrounds the role of action in creating shared understanding…to study coordinating flexible performance, we must take seriously the obstacles that participants encounter during performance. Not being able to read each other’s minds, people do not have a way of knowing what other people are thinking, except through the actions that they make available to each other during performance—that is, what they say and what they do” (p. 515). This study moves beyond existing work in all areas of the research process. It innovates in terms of data collection (videography of handoff routines and reflective interviews), employs analytic techniques rarely used in management (conversation analysis and multimodality), and uses novel ways to present the findings (video stills and Jeffersonian transcript, explained below).
In order to stay true to the ethnomethodological approach, LeBaron et al. (2016, p. 518) engage in creative data generation practices. In particular, the research team video-recorded the handover routines of physicians, by placing a camera in the corner of the handover room with the aim of capturing “emergent action processes as they unfold in time and space, through orchestrations of discourse, bodies, and things (Streeck et al. 2011).” Video recordings also enable researchers to purposefully return to the data and “repeatedly observe the empirical details of select events and activities within organizations” (p. 518). This rich multimodal dataset was complemented with reflective interviews that could contextualize and give deeper meaning to the captured interactions. In introducing ethnomethodology, the authors emphasize three defining features of ethnomethodology and explain how these emerge powerful insights. These features are (a) the centrality of situated actions, that is, what people say or do in a given situation, (b) the importance of surfacing expectations and setting them against actions, and (c) the role of action sequences in recognizing that “actions acquire their meaning in relation to prior and subsequent actions” (p. 517).
These data supported sophisticated and innovative—conversational and multimodal—data analysis. In combining these two analytic traditions, the focus was on “talk-in-interaction” as “talk is always situated within a material environment that includes objects, artifacts, and tools; and talk is always embodied by those who look, point, and orient their bodies in ways that constitute social action (Streeck et al. 2011)” (p. 518). Conversational analysis involves detailed transcriptions of talk, including microdetail like pauses, stutters, and elongation of words. These are generally facilitated by Jeffersonian transcription. The Jeffersonian transcript is a specific way of transforming speech into text, composed by Gail Jefferson as she transcribed Harvey Sacks interviews, which captures fine details of language use and interaction (see Jefferson, 1984). It commonly includes notations for laughter, pauses, and positioning. This enables researchers to capture subtle linguistic nuances that influence conversational dynamics and signal particular meanings. Multimodal analysis enables the author team to capture additional complexity in the interaction, including gaze and positioning of body within a space. Such detail is meaningful in terms of enabling greater understanding of interactional dynamics and their consequentiality.
Unusually for research in the field of management, the authors did not define a research question at the start of the research project, but rather engaged in the process of data generation and analysis “until a research question eventually emerged and enabled our theoretical contribution (Sacks 1992)” (p. 518). Hence, the authors sought to “identify, articulate, and then answer” their research question through their analysis of handoffs (p. 518). The analytic process cascaded through three main phases: (a) coding of action patterns, (b) identifying themes in interviews, and (c) conducting additional analysis of handovers using interviews to identify contextual cues. The authors used “medical informants” to understand the medical terminology and support the analytic process; in fact, two of the authors are trained medical practitioners (Christianson and Ilan). Also including a scholar extensively trained in communication studies and anthropology (LeBaron), the author team brings together a diverse knowledge and skillsets.
Some of the most innovative features of this article are the visual displays that are used in the presentation of findings. LeBaron et al. (2016) use several video stills from recorded handovers in order to decipher and present sequences. In presenting these, the authors draw attention to various gazes, gestures and body positioning to support their interpretation. Additionally, excerpts from their Jeffersonian transcripts are presented in the body of the text and analyzed line by line. Together these different forms of data bring to life the handover routines, making the various sequences much more visceral and accessible, and consequently present a more holistic and convincing sequence of events that cast light on the multimodal nature of coordinating.
Exemplar 3—Harnessing Collaborative Inquiry through Action Research
Our third exemplar of methodological innovation is the Lüscher and Lewis (2008) action research study on change and managerial sensemaking at the Danish Lego company. It represents the first of three exemplars that illustrate innovate ways for engaging with research participants and the research field. “Although largely absent from mainstream journals,” the authors develop a highly collaborative action research approach because it “offers exceptional access to and support of organizational sensemaking” (p. 221). Despite the importance of managerial sensemaking during organizational change, related studies are rare, in part because of research challenges. Such investigations require exceptionally intimate, real-time, and longitudinal research access…examining their sensemaking requires a highly interactive method (Balogun & Johnson, 2004)…. Argyris (1993)…called for more collaborative methods, stressing the potential for action research to support sensemaking and enable induction…intervention may help actors surface more subconscious anxieties, cope with defenses, and alter their cognitive frames. (p. 223) Lewin (1946) recommended that subjects and researchers be jointly responsible for developing and evaluating theory to ensure that the results of inquiry (1) reflect the knowledge created through the participative process and (2) help improve the social situation of the subjects. Researcher and subject engagement are critical to ensuring relevance. According to Argyris (1993), its participatory nature makes action research ideal for exploring latent dynamics of organization life. In contrast, more detached approaches miss discrepancies between “espoused theories” and “theories-in-use,” as researchers may only have access to what actors can, will, and/or are allowed to express. (pp. 223–227)
The study departs from the status quo in terms of data collection and analysis practices in several ways. Specifically, the interactive, engaged research process enacted through action research involves exploring possible patterns and emerging themes with the participants; leveraging outsiders such as internal consultants and other researchers to critique the research process; and researchers encouraging subjects’ ongoing experimentation and reflection to assess the validity and value of findings. In the study by Lüscher and Lewis, interventions by the research team served as triggers for managers to articulate and challenge their own understanding, and apply new understanding in subsequent practice in order to reflect on and challenge that understanding. In short, the purpose was to iteratively challenge and grow understanding in order to facilitate change. This required a “highly flexible” and adaptable research design, emulating the “most open” approach within the spectrum of action research. As the authors state, “Our approach requires a degree of flexibility and involvement foreign—and potentially anxiety-provoking—to many trained researchers” (Lüscher & Lewis, 2008, p. 238).
The project unfolded over three iterative and cyclical phases: groundwork, intervention, and theory building. While “groundwork” built on data familiar to most qualitative researchers—semistructured interviews and archival data—and theory building involved processes similar to those found in other forms of theorizing inductively from the data (e.g., triangulation, outsider review, member checks), the intervention was quite novel. Specifically, the authors used serial individual and team sparring sessions to encourage reflection and practice adaptation. These sparring sessions were supported by “interventive” questions (Tomm, 1987) that differ from common, open-ended questions used in qualitative interviewing (see the following quote).
The authors also develop unique ways for elaborating and presenting their research findings. Rather than separating out method and findings, as is often the case in research papers, they stay true to their action research approach by including in the findings section both their collaborative research process and their resultant substantive findings. In the discussion, the authors then elaborate an integrated methodological and theoretical contribution. They work their research process into a new process methodology for paradoxical inquiry (“Sparring: A Collaborative Process of Working through Paradox,” p. 227). The authors explicate three aspects of organizational change—paradoxes of performing, belonging and organizing—and illuminate them with detailed sensemaking data, contributing to “understandings of interconnections among paradoxes” (p. 236).
Exemplar 4—Harnessing the Power of “Embedded Participation” via “Undercover Researchers”
The exemplar we showcase next is the unique mixed methods study by Bernstein (2012), which combines an embedded participation-based study with a field experiment. The paper investigates the implications of transparent organizational design on organization performance and worker productivity by means of two studies conducted in a mega mobile phone factory in China over an 18-month period: “Study 1 was a month-long inductive qualitative study with data gathered by three embedded researchers who were simultaneously operators on the factory lines and participant-observers for the study. Study 2 was a field experiment using an intervention and, again, embedded researchers to collect data on the factory lines, this time for the first five weeks of a five-month study” (p. 185). A central element of both studies was covert embedded participation. This method adaptation emerged from practical considerations: “this research design was born out of necessity from my inability to fit in” (p. 187). In short, the author, a Western foreigner, would not have blended in on a Chinese shop floor.
As a result of this need to adapt the research design to the practical circumstances, the study features innovation in all parts of the research process. In particular, it uses innovative data collection techniques (deployment of “confederates” onto the shop floor and debriefing into recording devices), utilizes unusual analytic techniques (grounded-analysis typical of observation-based study and statistical analysis characteristic of field experiment), and uses interesting ways to present the findings (rich illustrative quotes alongside objective frequency graphs and statistical reporting). The data generation process in this research project is particularly innovative. As Bernstein (2012, p. 185) himself notes, I adopted a special methodology to avoid contaminating the environment and the behaviors I was attempting to observe. My research team included three undergraduate students (two females and one male), all three of whom had been born and raised in China until at least the age of ten. The three students were inconspicuously placed on the factory lines as ordinary employees—only the GM, head of HR, and head of operations of the 14,000-person facility knew their true identities, which were carefully guarded.
The analytic process was tweaked to reflect the research design and the nature of the data. While in Study 1 a grounded theory approach was adopted (Glaser & Strauss, 1967; Miles & Huberman, 1994), Study 2 used descriptive statics and a difference-in-differences estimation model, which “consists of first identifying a specific treatment and then comparing the difference in outcomes, intervention, for groups affected by the intervention to the unaffected” (p. 198). While neither of these approaches is particularly unique, the combination is novel. Particularly rare is the relatively equal emphasis on qualitative and quantitative methods.
The novelty filtered into the presentation of findings, where we see an interpretive qualitative study followed immediately by a more positivistic, quantitative study. Additionally, unusual for interpretive work, the studies are presented within a positivistic framework, where each study is described, methods outlined, findings displayed, and discussion presented in turn. Detailed descriptive findings were presented in Study 1: I have a feeling that he tempers with the numbers, because only 2 days ago, I was one station away from him. I don’t think he’s actually counting them one-by-one. I think he controls the numbers throughout the day—sometimes they produce more, sometimes they produce less. Our blue hat always writes down 240. And he never…I actually pay attention to what he does…he never actually looks at what it says on the Precision IT screen. He just notices that it’s about the right time—let’s say it’s 9:46—and he will write down for the block from 9-10 that we made 240.
Exemplar 5—Harnessing the “Insider” Experience of the Researcher
The two-year ethnography of investment bankers by Michel (2007) showcases the innovative contributions of studies in which researchers adopt an “insider” perspective. Specifically, using an immersive ethnographic design and drawing on her own experience of working at a Wall Street bank, Michel (2007, p. 507) is able to develop a multilevel grounded theory that “depicts the mutual constitution of individual and organizational cognition, including the relation of these levels over time and how the same individual cognitive process operates differently in a different organizational cognitive context.” Michel (2007, pp. 514–515) describes her research role as a “marginal native,” experienced and empathetic, but nonparticipatory: Prior to this research, I worked at a Wall Street investment bank for four years, first as an analyst and then as an associate in the mergers and acquisition department, which provided me with the kind of knowledge Resnick, Pontecorvo, and Säljö (1997) deemed essential to studying distributed cognitive processes. I also worked for one year in the bank’s training department, where I conducted a survey of the industry’s best practices in professional development. The personal connections I gained during this process helped me gain access for the present investigation. My background in banking enabled me to attend client meetings in which bankers justified my presence as a form of quality control…. My personal experience increased my empathy with informants and positioned me as an in-group member, such that bankers included me in work and sometimes non-work activities and trusted me with private information. Both empathy and social inclusion were crucial for investigating cognitive change processes. As elaborated below, I chose the observer (versus the participant) as my primary research role to maintain the position of the “marginal native” (Freilich, 1970; Hammersley and Atkinson, 1997) to balance deep familiarity with the detachment necessary for intellectualizing the experience.
Michel’s (2007) data collection is novel in terms of the quality of access her insider position affords her and the large datasets she can produce based on this high quality access. For instance, her preliminary study consists of 48 days of observation and 84 interviews. Her main dataset is even more extensive, consisting of “four overlapping data sources: overt participant and non-participant observation (about 7,000 hours); 136 formal, semi-structured interviews; informal interviews with 120 informants; and analysis of company materials” (p. 515). Structuring her data collection around two “high contrast” cases in investment banking, she immerses herself deeply in the organizations, shadowing bankers and adopting their working hours: “My most intense participation was during the first year, when I observed between five and seven days a week (80–120 hours), mirroring the bankers’ working week…. To monitor or ‘shadow’ informants continuously, I moved into the cubicles of traveling bankers and took notes on what the banker next to me said and did” (p. 515). This involved detailed charting of data to individual bankers to ensure balanced observations.
Michel’s (2007) data analysis also incorporated features facilitated by her “insider status.” While her grounded theory approach incorporates many of the standard techniques of the method, she also incorporates unique features, for instance, treating each banker as an individual case and involving bankers in coding in order to assess themes by asking “friendlies” to code data and explain their ratings. Given the grueling schedule of bankers, this commitment to the research project and willingness to engage with mundane tasks like coding is impressive.
In terms of presentation of findings, Michel (2007) offers an in-depth, almost personal account that is lively, emotive and authentic. For instance, she quotes one associate in the following way, drawing on the interpersonal and collegial nature of learning: “Jeff is my big buddy. On the first day, we sat down and put together a list of all the things I should be learning during the next six months. When I feel that I know how to do something by myself, I cross it off the list. If I am not staffed on deals where I can learn these tasks, Jeff will sit down with me and teach me.” Big buddies let associates watch them doing tasks and let associates do increasingly more of the task (p. 529).
Exemplar 6—Harnessing Multiple Linguistic Traditions
Sonenshein (2010) showcases another way of innovating research methods. Specifically, the author harnesses synergies between two rarely combined traditions—discursive narrative analysis and content analysis. Sonenshein clearly positions this combination as a distinguishing feature of his study: “In a single-site case study, I used narrative analysis (Riessman, 1993) and content analysis (Berelson, 1952) as two related approaches to examining the discourses individuals used to construct meaning (Gephart, 1993; Pentland, 1999)” (Sonenshein, 2010, p. 480). Using these two techniques in conjunction enables Sonenshein (2010) to produce a unique case-study based research approach, which unveils that “managers tell strategically ambiguous, interwoven narratives about how an organization changes and how it remains the same, thereby attempting to both unfreeze and freeze the existing meanings employees attribute to the organization” (p. 477). The work is innovative, as the author used unusual data (retrospective and real-time data; surveys and observations), rare combinations of analysis (narrative and content analysis), and unique findings presentation (tabling of contradictory views and competing narratives, use of composite narrative, supported by frequency counts).
Sonenshein (2010, p. 482) clearly emphasizes the unique features of his study in relation to data collection and generation: Unlike most research on change, which only uses retrospective data (Van de Ven, 1992), this study includes both retrospective and real-time data. In total, I spent 15 months collecting data during the testing and implementation of the strategy, and after the project was no longer a strategic priority. I used five data sources: interviews, documents, archival records, observations, and surveys.
Drawing on these data, Sonenshein combines divergent forms of data analysis in innovative ways. Rare for a study of meaning making, Sonenshein (2010, p. 480) combines narrative analysis (cf. Riessman, 1993)—one of the standard analytical techniques used in the field—with content analysis (cf. Berelson, 1952)—a technique seldom employed within this particular field, positioning them as complementary: “Narrative analysis focuses on the rich use of discourse embedded in context (Pentland, 1999), particularly emphasizing how protagonists interpret experiences (Bruner, 1990). Content analysis is a technique for understanding the factors that explain differences in interpretations (Langley, 1999).” He uses the techniques sequentially, constructing narratives to identify and outline broad issues, and then drilling into the specificity of these “big picture” issues with content analysis: “Narrative analysis affords an opportunity to examine managerial and employee meanings embedded in a rich context by such means as examining large fragments of discourse; content analysis provides an opportunity to hone in on specific meanings in those narratives to explain patterns (Langley, 1999)” (p. 484).
Thus, drawing on narrative analysis, Sonenshein (2010, p. 484) initially “used individuals’ discourse (managers or employees) to create composite narratives of group constructions of the change (Currie & Brown, 2003; Dunford & Jones, 2000; Eisenberg, Murphy, & Andrews, 1998)” (p. 483). Then, seeking to understand how the meanings constructed had evolved, he used content analysis and “inductively derived codes about meanings from the narrative analysis, focusing on emergent core themes,” even providing “percentages of mentions” (p. 484; see also Table 4, p. 497). Integrating the analytic techniques in this way allows Sonenshein to produce an innovative and value adding piece of work.
This innovative practice filters through to the unique presentation of findings. Particularly, staying true to the narrative tradition in a way seldom seen in the management field, Sonenshein (2010) includes contradictory views and competing narratives (see the extensive Table 2, pp. 487–494), for instance, showing the same manager simultaneously positioning the change as significant and insignificant in the same conversation: We as the MallCo chain have gone through a period of sales decreases…we had to do something in order to change that trend. And part of changing that trend was leveraging the BigBoxCo name, refreshing the stores.…Attracting that mall customer that might be in there getting them to walk into our BigBoxCo Light store just as they would walk into a BigBoxCo store, but not necessarily a MallCo store [change as significant]. It’s basically the same other than the paint and some of the fixtures and the sign out front [change as insignificant].
Exemplar 7—Harnessing History via Historical Research Methods
A final exemplar of methodological innovation is the historical case study by Lubinski and Wadhwani (2020) on multinational strategy formation in the context of rising economic nationalism. Following a recent “historical turn” in management and organization studies (e.g., Kipping & Üsdiken, 2014; Rowlinson et al., 2014; Vaara & Lamberg, 2016), such studies leverage history as a key component of theory and empirical analysis, and draw on historical research methods (Argyres et al., 2020). Lubinski and Wadhwani (2020) showcase methodological innovation in the data generation practices (using historical cases; documentary data from archives as primary sources) as well as data analysis and presentation techniques (narrative strategy).
Lubinski and Wadhwani innovate data generation methods in the following ways. While we would usually expect contemporary or “live” cases, the authors use historical cases. In so doing, their empirical treatment of time and temporality shifts in innovative ways (Rowlinson et al., 2014). Specifically, whereas organization scholars often focus on the chronological order of events, historical methods consider the historical context as central to the interpretation and understanding of events. The focus is explicitly on revealing the social, cultural, and institutional construction of a phenomena at a specific time and in a specific space. Lubinski and Wadhwani (2020) showcase this periodization of events when, based on various different sources, they carefully construct “the historical case of [two] German MNEs [multinational enterprises] in late colonial India and the context of interwar international business [period between World War I and II]” (p. 404).
Given the focus on retrospective cases, the study is primarily “based on research in corporate and government archives in Germany, India, Great Britain and the United States” (p. 405). This emphasis on archival data reflects a fundamental, disciplinary difference regarding research evidence (Rowlinson et al., 2014). Organization scholars generally prefer real-time data constructed from specified procedures (such as semistructured interviews conducted with the managers and employees of a firm) and tend to relegate archival sources to providing background information or validating retrospective accounts (see Jarzabkowski, 2008). By contrast, historical approaches cast retrospective interviews as notoriously unreliable “testimony,” especially when collected years later, and therefore prefer the eclectic, but verifiable documentary sources found in organizational archives (Rowlinson et al., 2014). Working with public archives has the added advantage of allowing Lubinski and Wadhwani to explicitly name the two companies they study—Siemens and Bayer—rather than anonymizing them like most organization research studies do. They also list the archives they drew their evidence from, thereby providing verifiable locations of their sources of evidence. Finally, they cite an extensive set of primary and secondary sources, including, for instance, company letters and governmental reports from the time of study. This is also seldom and unique.
Lubinski and Wadhwani (2020) further draw on historical research methods in their data analysis and presentation of findings. The key innovation here is their strong reliance on analytical narratives (see Ingram, Rao, & Silverman, 2012; Rowlinson et al., 2014). Historians traditionally rely on the literary narrative form (often without explicit theory and method) as a powerful means to represent and explain historical events. This entails greater reflexivity regarding the issue of representing the past through narrative construction (Rowlinson et al., 2014). For example, Lubinski and Wadhwani explicitly position themselves as historical realists and describe their strategy of an “analytically structured narrative (Ingram, Rao & Silverman, 2012; Rowlinson et al., 2014)” (p. 405). More specifically, the authors construct an historical account that was focused on the conceptualization of economic nationalism [and] the strategy formation at Siemens and Bayer…. These companies faced a highly uncertain environment in the wake of World War I, and their initial strategy vis-à-vis nationalism is best characterized as “emergent.”…Over the subsequent decades, the firms’ strategies vis-à-vis nationalism became increasingly deliberate, and the narrative traces this evolution from opportunism to intentional strategy. (p. 405)
Discussion
Based on an in-depth review of exemplar studies, we develop an understanding of how scholars innovate research methods. To do so, we use this section to elaborate an empirically grounded model of innovating methods. We call this model the “innovation iceberg,” paying tribute to the multilayered and often implicit process of innovating methods. In order to improve visibility of innovation in the research process, we break innovation into two interconnected layers: (a) the explicitly mentioned and visible elements of methods innovation, represented through new or seldom applied research tools as reported in a given study; and (b) the more implicit core principles that guide the innovative use of research methods. Next, we describe and elaborate these two layers, outline the implications of our model, and develop options for future research.
Innovating Research Tools along the Research Process
As shown by our exemplar studies and in Figure 1, scholars can innovate research methods along all parts of the research process. They can introduce new or uncommon ways of data generation (e.g., doing a video ethnography in a research team at multiple sites, doing covert observation via “undercover researchers,” relying on organizational archives as primary source), they can innovate the process and strategies of data analysis (e.g., engaging in a multimodal analysis, doing data analysis collaboratively with research participants through focus group review sessions), and they can present their findings in unconventional, creative ways (e.g., write up their findings as a narrative, report both research process and products). Our focus on exemplar studies means that we review a select set of method innovation, specifically radical innovation that involves the introduction or elaboration of new or uncommon research approaches.
Innovating research methods as two-layered process (“innovation iceberg”).
We would thus like to remind readers that they may also find inspiration in work based on radical innovation not covered in this study (e.g., autoethnography as showcased by Empson, 2013) or based on more incremental innovation. Indeed, in our review, we identified several other examples that feature meaningful innovation. For instance, Huy (2011, p. 1393) draws on “cognitive appraisal theories of emotion” to code emotion in informal conversations. Stigliani and Ravasi (2012) study the material artifacts used in product design, zooming in on activities such as collecting objects and interacting with users, in order to better understand collective sensemaking. Lingo and O’Mahony (2010, p. 54) immersed themselves in the research context in a creative way in order to gain firsthand experience of the challenges research participants in the music industry faced, “The first author also co-produced a single song: working with a co-producer to write and select from a portfolio of songs to be recorded and co-managing the three-hour recording session with an engineer, drummer, lead guitarist, pianist, and bass guitar player.” Gail Whiteman (Whiteman & Cooper, 2011) brings to life the notion of ecological sensemaking through a vivid first-person account of falling into a roaring river during an ethnographic observation of Cree hunters in subarctic Canada. Another example is Danneels’s (e.g., 2011) use of the extended case method from critical sociology (Burawoy, 1998) to enable theory elaboration from case studies. These are only a few examples of the many other ways of innovating that we identified in our review.
Had we sampled more widely, including a greater range of years and journals, many additional ways of innovating would also be showcased. For instance, Plowman et al. (2007, p. 515) start their paper atheoretically, showcasing the rich case narrative: “The church faces scenic downtown Mission Park and is wedged between two historic hotels that regularly host elegant weddings and expensive executive education seminars.” Heracleous and Jacobs (2008) utilize workshops with organizational actors in which they use Lego and other materials to develop “embodied metaphors.” Balogun and Johnson’s (2004, p. 526) advance research on managerial cognition by using diaries, “a form of personal document…as the primary data collection mechanism.” There are many more examples like these to acknowledge and celebrate, but we cannot capture them all within the realm of a single paper. It is thus critical to understand the underlying principles that guide method innovation. We turn to this topic next.
Guiding Principles of Innovating Methods
As main contribution of this article, we reveal that the use of innovative research tools is guided by core principles of innovating methods. These principles are often not explicitly stated by the authors, yet manifest in the how the study is conducted. Based on our review of exemplar studies, we explicate four of these principles as a central part of our model (cf. Figure 1). In so doing we give particular focus to the way they link to established principles of “good” empirical research (“criteria of goodness”; see Denzin & Lincoln, 2005; Hammersley, 2007; see also Keleman & Bansal, 2002). The four principles we identified are (a) innovating methods holistically, (b) establishing the credibility, appropriateness, and value of the new method through excruciating clarity of method, (c) attending to the entwined nature of theory and method, and (d) remaining reflexive when innovating methods.
The holistic nature of method innovation
Method innovation may begin with single parts of the research process, such as access to or interest in an unconventional data generation strategy (e.g., working with visual and video data, or using self-report data). However, as the various parts of any research design are entwined rather than independent from each other, such decisions have effects that cascade through the entire research process in ways that beget innovation (cf. Figure 1). Indeed, powerful research designs are internally consistent, that is, the different choices regarding, for example, data generation, data analysis, and presentation of findings fit together and support each other (Gephart, 2004).
The emergent process of innovating methods in impactful ways is therefore often holistic in nature. This means that researchers engage in the process of innovating methods by progressively crafting and implementing innovation in multiple parts of the research design. While adaptations in select parts of the research process are also possible, these are more likely to be pragmatic adaptations to a particular research situation, rather than the type of shift in ideas that we see in more holistic innovation. Indeed, all of our exemplar studies demonstrate holistic methodological innovation. For instance, Lubinski and Wadhwani (2020) started their historical case study with negotiating access to corporate and governmental archives, and then innovated data analysis and presentation techniques in line with these data when constructing and writing up their findings in a historical narrative. Similarly, Bernstein’s (2012, p. 187) use of deception arose out of practical necessity—he was not able to blend in on the floor of a Chinese factory and thus had to find a creative way to access this site. Ultimately, he recruited Chinese-born students to work on the assembly line; these students generated data through observation and interaction with other assembly line workers, debriefing into recording devices at regular intervals. The data were later analyzed using techniques borrowed from grounded theory, and emerging theory was tested with a field experiment. Innovation thus cascaded through every part of the research process.
Establishing the credibility, appropriateness, and value of the new method through excruciating clarity
Method innovation entailed, in our exemplar studies, an explicit and detailed presentation of the new method used in the study. This innovation principle resonates with the well-established principle in scientific research that a detailed account of method and evidence is an important part of an empirical study (Gephart, 2004). For example, qualitative researchers commonly demonstrate the trustworthiness of their findings by establishing “audit trails” (e.g., Lincoln & Guba, 1985). In analogy to mathematical audits or audits of fiscal accounts, an audit trail chronologically and systematically documents what the researchers did, how they did it, and how they arrived at their interpretations. This information empowers the reader to independently assess the credibility of the study as well as its appropriateness and value.
The need to transmit the credibility, appropriateness and value of a method, however, is particularly pronounced when the researcher needs to communicate a method largely unfamiliar to a target audience. In such cases, a skilled presentation of the method is critical in order to convey the general usefulness of the method and demonstrate competence in the method to the audience. Our exemplar studies also illustrate this guiding principle of method innovation. For instance, Lüscher and Lewis (2008) use their method section to provide an extended introduction to action research in order to familiarize management scholars with this relatively rare method, repeatedly demonstrating the value and credibility of the method. Similarly, LeBaron et al. (2016) introduce the ethnomethodological perspective that inspires their multimodal approach to data generation and analysis before explaining principles of conversation analysis and sociomateriality that support multimodal work. These authors thus educate their audience in “bite-size chunks,” so they can take the reader with them on the journey of discovery.
The close interplay of theory and method development
Another widely established quality principle is the emergence of a meaningful theoretical contribution; that is, the study creates new theory, synthesizes previous theories, integrates multiple perspectives, or comes up with an important, potentially counterintuitive finding (e.g., Bartunek et al., 2006). Prior research has highlighted that method innovation facilitates focus on unresolved research questions and produces novel findings (e.g., Bansal et al., 2018).
We further elaborate the close entanglement of theory and method development by illustrating how new methods can generate powerful new theory. Our exemplar studies, for instance, consistently demonstrated that new methods emerge new findings and new theoretical insights. Additionally, we show how theory informs method innovation. In particular, in our exemplar studies, scholars either introduced a new theoretical perspective that spurred methodological innovation or introduced a new method and then used established theoretical perspectives to elaborate and justify the new method. For example, expanding practice theory into strategy research required strategy scholars to innovate methods in ways that would allow them to capture and reconstruct patterns of action linked to large scale strategy phenomena (e.g., Jarzabkowski, 2008). As we illustrate using the example of Smets et al. (2015), this can entail, for instance, global, team-based ethnographic research to capture the situated enactment of strategizing practices in specific incidents and across multiple sites. In a similar vein, Michel (2007) uses the theoretical concept of “cognitive uncertainty” from social psychology (Trope & Liberman, 1996), combining it with a practice perspective to demonstrate the “mutual constitution of individual and organizational cognition” that had previously been obscured in the literature (Michel, 2007, p. 507).
Reflexivity in innovating methods
Any research approach or method originates in particular research traditions and projects (Lê & Schmid, 2019). Their inventors develop them for specific research purposes and with particular assumptions about how to do research. Scholars introducing new methods from other fields need to stay true to these meta-theoretical assumptions. Bansal et al. (2018, p. 1192) make a similar point when they state that an effective use of established qualitative methods entails a reflexivity of underlying assumptions: “Know Your Epistemology: Specific genres [of qualitative research] reflect particular onto-epistemological assumptions that should be taken seriously through the research process and writing of the paper.” In our study, we show that method innovation requires similar reflexivity. Specifically, there is implied reflexivity in our exemplar studies. For example, Lüscher and Lewis (2008) transferred action research from the field of organizational development to broader management research, preserving the original, intervention-based participatory research tradition underlying the method. They also explicated the assumptions underlying the method and explained them in the light of the assumptions commonly held in the targeted research field. Additionally, they explicitly talk about engaging in reflexive activity, such as sparring and reflexive questions, with their research participants.
Taken together, the four principles we identified here demonstrate that scholars can orient their use of new methods by building on established rules and standards of high-quality research. This finding informs prior research that has raised conservative skepticism regarding the innovation of methods. For example, in their review of innovation in qualitative methods, Wiles et al. (2011) found that methodological innovation was often limited, as authors over-stated their innovation or imported methods from other fields, rather than creating entirely new methods. In a similar way, Travers (2009) recognized that authors using qualitative methods may feel forced to engage in methodological innovation for its own sake, losing sight of basic virtues and practices of qualitative inquiry, such as a deep and repeated engagement with empirical data (see also Lê, Smith, Crook, & Boyd, 2019). By contrast, our study suggests that an adherence to established principles of “good” research and the aspiration to continuously innovate methods are not mutually exclusive. They can, as our exemplar papers show, reinforce and support each other. Certainly, scholars may hold different, even contradictory, understandings of what constitutes “good” research (or quality criteria—see, e.g., Amis & Silk, 2008; Flick, 2006; Hammersley, 2007; Patton, 2002). At the same time, scholars often engage in similar quality assurance practices (e.g., detailed reporting of methods). Such shared practices can help produce method innovation that is both novel and high quality.
Implications for Teaching and Learning Research Methods
While method texts and a careful examination of exemplar studies highlight important ways to attain methodological knowhow and learn new methods, scholars will also learn about and innovate methods “by doing” empirical research (see also Bansal et al., 2018). As such formal and informal research methods training is central to the emergence of good research, we close with implications for learning and teaching (new) research methods drawn from our iceberg model of methods innovation.
A first implication of our model is that the process of innovating research methods entails advanced expert knowledge and skill. The guiding principles of innovating methods we identified in this study reflect this high level of methodological skill. According to our selective review of exemplar studies, researchers may need, for example, the ability to integrate multiple practices of methodological innovation, integrating them into a coherent research design. They also need the capacity to educate and convince other scholars who may be less familiar with the new method of the value it generates. Researchers may also need sophisticated skill in both theories and methods so that they can justify and elaborate a new method within theoretical perspectives and, in turn, use the affordances of the new method to advance these theories. Finally, researchers need to be highly reflexive about meta-theoretical assumptions and aspirations underlying the new method in order to make effective use of this method and be able to communicate it within and across established research traditions and perspectives.
Given this skill requirements, method innovation may require individual scholars to commit to particular methods, engaging in a career-long learning process that involves repeat publications. This constitutes a significant personal investment. In fact, our review made salient to us that authors associated with particular methods sometimes had to go through a relatively long process before they were able to publish the exemplar study in a leading journal. Future research may take up this focus on scholars innovating new methods and their individual learning and career trajectories.
Another important implication is that method innovation is often implemented by interdisciplinary research teams. In the exemplar studies reported here, the author (team) often had training and experience in areas outside of the field of organizational research. For instance, LeBaron et al. (2016) was written by a diverse team, including a scholar extensively trained in communication and anthropology (LeBaron), two physicians (Christianson and Ilan), one person trained in organizational behavior (Christianson), and an early career scholar (Garrett). This unique combination resulted in a paper based on innovative data collection practices (videography of handoff routines) and rarely used analytic techniques (conversation analysis and multimodality). We thus see working with a diverse set of collaborators, particularly people from outside of the field, as one potential source of methodological innovation. We suspect that working with such “diversity” and “outsider perspective” can push boundaries and challenge thinking in ways not previously conceived. Collaborations also allow us to bring key skills together in new, innovative and generative ways.
Finally, our exploration of method innovation has implications for the teaching of research methods. Our work suggests that we should place the teaching of proven methods alongside the development of practical skill in innovating methods. This is critical to ensuring that methods retain the fluidity and flexibility with which they were created (e.g., Gioia, 2019), rather than simply becoming tools that reproduce knowledge. Hence, rather than teaching well-established research approaches as fixed templates, method trainings might use these as initial frameworks to orient and support students in order to later make creative use of established approaches for their own study. We would also emphasize that the specific innovation exemplars we showcase are meant to provide a path away from templates; we thus strongly discourage scholars from using them as templates in their own right. In addition, it may be important to teach a broader spectrum of methods, including methods less well established in organizational research, in order to encourage scholars to import methods from other fields, create entirely new methods or combine methods in novel ways. Going further, the interplay of theory and method development in the process of innovating also suggests that integrated theory and method courses would be one valuable way to teach/learn how to innovate methods in theoretically meaningful ways. In short, our study suggests new ways of teaching methods that engage students in actual research practice as one way for encouraging innovation.
Conclusion
By critically examining the process of methodological innovation, we identified multiple ways that scholars introduce or combine methods in creative ways. While prior research has focused primarily on explicating specific new methods, our study provides an empirical account of the process of methodological innovation. Our study has focused on qualitative research within strategy and management research. Due to their diversity and flexibility, qualitative research methods can be seen as an “extreme case,” that is, methodological innovation may be particularly important and common. We thus need future studies investigating how scholars innovate other types of methods. Still, our study demonstrates the value of methodological innovation for theory development, and, we hope, encourages further innovation in research methods.
Footnotes
Declaration of Conflicting Interests
Funding
ORCID iD
Notes
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
