Replication

One of the most contentious debates regarding open qualitative research revolves around the question of whether replication is meaningful and feasible for qualitative research. Many researchers argue that the philosophical underpinnings of qualitative research render replication, or at least exact replication, problematic. For example, rooted in a social constructivist stance for feminist psychological research, Pownall (Reference Pownall2024) emphasizes that the subjectivity and positionality of the researcher are not biases to be removed but an inherent and intentional part of the research design and, therefore, expecting two researchers to produce an identical interpretive trajectory is neither feasible nor desirable. Taking a transcendental realism stance that views the world as composed of open, fluid, and complex systems, Pratt et al. (Reference Pratt, Kaplan and Whittington2020) further contend that the goal of qualitative research, which is often theory building and in-depth understanding of context-specific phenomena, directly clashes with the goal of replication in quantitative research seeking to test and confirm existing theories. This view favors a “horizontal notion of knowledge cumulation” (Pratt et al., Reference Pratt, Kaplan and Whittington2020, p. 5) where the goal is to build a wide repertoire of tentative theories for unique situations, rather than a vertical cumulation that builds on previously confirmed findings. In this line of view, researchers often refer to existing hallmarks and methods such as trustworthiness, triangulation, and member checking as better alternatives to replication, which is viewed as enforcing a positivist, quantitative, and experimental logic (Pownall, Reference Pownall2024; Pratt et al., Reference Pratt, Kaplan and Whittington2020).

In sharp contrast, other scholars advocate for the value of replication for qualitative research. Challenging the notion that replication is largely opposed by qualitative researchers, Makel et al. (Reference Makel, Meyer, Simonsen, Roberts and Plucker2022) cite evidence from a survey of 1,500 educational researchers to demonstrate that categorical opposition to replication is an infrequent stance, with less than 10% of qualitative researchers actually expressing the belief that it should be completely avoided. Building on this premise, they assert that replication can be applicable to qualitative inquiry in three principal areas. Specifically, replication reinforces methodological values like transparency and intentionality, provides a means to assess the transferability of findings, and can be used to evaluate the influence of researcher reflexivity on study results. Taking a similar stance against the assumption that replication is exclusively for quantitative research, Becker (Reference Becker2019) posits that qualitative replication not only can be done but should be encouraged as a powerful pedagogical tool because it helps novice researchers demystify interpretive decision-making and practice systematic evidence evaluation. Essentially, qualitative replication projects can serve as “training wheels” for novice researchers via a “learning-by-doing” approach (Becker, Reference Becker2019, p. 3).

Moving beyond this for-or-against dichotomy, a third strand of the debate proposes a reconceptualization of replication itself to enhance its compatibility with, and utility for, qualitative research. Porte and Richards (Reference Porte and Richards2012) argue that methodological and technological developments (e.g., stricter analytic standards, online data archives) mitigate earlier objections to replication and render it both meaningful and feasible for qualitative L2 research. Acknowledging that certain branches of qualitative research such as what Nespor (Reference Nespor2006) label as militant particularism (i.e., privileging contextual and local knowledge over generalized knowledge) are “inherently antithetic to replication,” they speak of a replication logic for realist-oriented qualitative research where each case is preserved as an analytic whole embedded in rich contexts yet intentionally sampled to extend or challenge emergent theory (Porte and Richards, Reference Porte and Richards2012, pp. 288–289). Taking a similar stance but situated in ethnomethodological conversation analysis, Markee (Reference Markee2017) proposes comparative reproduction as a modified form of replication, a practice that traces how the same social action is organized across languages or settings, and aims to establish the generality or prototypicality, not universality, of qualitative phenomena. Interrogating replication from a complex dynamic systems perspective, Al-Hoorie et al. (Reference Al-Hoorie, Hiver, Larsen-Freeman and Lowie2021) argue that because complex systems are open and defined by variability, the idea of an exact or direct replication is untenable. Instead of dismissing replication entirely, they propose a broadening of the scope of replication toward substantiation. This substantiation framework reframes the goal away from simple or methodological replication and adopts a more nuanced approach that includes extension, reproduction, and exploration, depending on the maturity of the theory under investigation.

Preregistration

Similar to replication, preregistration, as one of the core practices in mainstream open scholarship, is another source of contention. Critics argue that preregistration originated as a means to enhance hypothesis testing, a logic foreign to and incompatible with the exploratory and theory-building nature of qualitative research. Haven and Van Grootel (Reference Haven and Van Grootel2019) identify three interconnected objections to qualitative preregistration: the absence of testable hypotheses, the iterative nature of the research design, and the subjectivity of the researcher. Because qualitative researchers often refine sampling, adjust data collection strategies, and pursue emergent lines of inquiry, registering a protocol would seem counterintuitive or even counterproductive to skeptics (Haven & Van Grootel, Reference Haven and Van Grootel2019). Bennett (Reference Bennett2021) extends this logic from a feminist standpoint, warning that current open science practices risk further marginalizing methods that privilege listening, reflexivity, and social change over statistical inference. In this view, preregistration is not just unnecessary; it is conceptually misaligned with qualitative paradigms.

Against this backdrop, a growing number of scholars contend that preregistration can be meaningful and feasible insofar as it is tailored to qualitative epistemologies and foregrounds flexibility and reflexivity. Haven and Van Grootel (Reference Haven and Van Grootel2019) argue that making study aims, analytic lenses, and anticipated procedures publicly visible enables wider scrutiny regardless of whether specific confirmable hypotheses exist. The emphasis thus shifts from pre-specifying predictions to documenting theoretical commitments and methodological trajectories. Building on this premise, Haven et al. (Reference Haven, Errington, Gleditsch, van Grootel, Jacobs, Kern, Piñeiro, Rosenblatt and Mokkink2020) conceptualize preregistration as a systematic starting point that should never become a straitjacket. Researchers are free to amend their plans as needed while leaving an auditable trail of decisions. Branney et al. (Reference Branney, Brooks, Kilby, Newman, Norris, Pownall, Talbot, Treharne and Whitaker2023) likewise highlight the capacity of qualitative preregistration templates to render the often invisible “reflexive labor” visible (p. 6). Importantly, infrastructure is catching up with these conceptual arguments: Several journals, including BMJ Open Science, F1000Research, and PeerJ, accept qualitative Registered Reports, and the Peer Community In initiative has begun recommending such submissions, albeit in modest numbers (Branney et al., Reference Branney, Brooks, Kilby, Newman, Norris, Pownall, Talbot, Treharne and Whitaker2023). Parallel efforts to refine templates and guidelines, such as the Qualitative Preregistration template (Haven et al., Reference Haven, Errington, Gleditsch, van Grootel, Jacobs, Kern, Piñeiro, Rosenblatt and Mokkink2020), further enhance the utility of preregistration for qualitative research.

Data sharing

Whether and how qualitative data can and should be shared is another focal point of discussion. Before reviewing this debate, it is helpful to distinguish between data (e.g., transcripts, recordings) and research materials (e.g., interview protocols). The sharing of materials is a crucial step for making the research process transparent and often faces fewer ethical barriers than sharing data. L2 researchers have been sharing materials for quite some time (e.g., the establishment of IRIS, an open repository for research into languages, in 2011), including qualitative research materials. For instance, there are 156 entries of interview protocol/schedule indexed on IRIS (as of January 20, 2026), which evidences existing engagement of qualitative research materials sharing.

In contrast to the uptake in materials sharing, the conversation regarding qualitative data sharing takes place against the backdrop of a growing push from the top down. While large text corpora have long existed as a form of open data in L2 research, and projects such as Languages and Social Networks Abroad Project (Mitchell et al., Reference Mitchell, Tracy-Ventura and McManus2017) have shared qualitative data such as oral interviews, there is an increasing push from funding agencies, publishers, and governments mandating open access to research data to enhance research transparency and maximize return on scientific investment (Bishop & Kuula-Luumi, Reference Bishop and Kuula-Luumi2017; Irwin, Reference Irwin2013). The United Kingdom illustrates how policy and infrastructure can work in tandem. The UK National Data Strategy (2009–2012), for example, expedited the expansion of the UK Data Archive and related specialized services such as the QualiBank repository. The Economic and Social Research Council, a major funder of social research in the UK, has been mandating funded researchers to deposit data into the Qualitative Data Archival Resource (Moore, Reference Moore2007).

Despite this policy push, many qualitative researchers caution against the view that sharing data is unequivocally beneficial. A commonly cited concern is the risk of breaching participant confidentiality despite anonymization (Irwin, Reference Irwin2013; Mauthner et al., Reference Mauthner, Parry and Backett-Milburn1998). De-identification of qualitative data can be labor-intensive, and at times problematic, especially for research on small or marginalized groups where seemingly innocuous details may reveal identity (Prosser et al., Reference Prosser, Hamshaw, Meyer, Bagnall, Blackwood, Huysamen, Jordan, Vasileiou and Walter2023). Relatedly, some researchers perceive sharing data as a potential ethical violation that could undermine the trust between the researcher and the participant (Irwin, Reference Irwin2013). Obtaining informed consent from participants is also deemed tricky as future use of the data may be unknown (Broom et al., Reference Broom, Cheshire and Emmison2009). Opposing stances become strongest when it comes to issues of power imbalances. Couldry and Mejias (Reference Couldry and Mejias2019) warn that publicly hosted data may be mined by outsiders for purposes never intended by the original contributors. Feminist decolonial authors such as Simpson (Reference Simpson2007) and Tuck and Yang (Reference Tuck, Yang, Paris and Winn2013) assert that silence, or the act of withholding information, is a legitimate ethical stance when speaking might expose marginalized groups to harm. Legal and commercial issues also complicate the matter when data are collected in partnership with private organizations that claim ownership or fear competitive disadvantage from data sharing (Zuiderwijk et al., Reference Zuiderwijk, Janssen, van de Kaa and Poulis2016). Additionally, there are also barriers to data sharing on a psychological level for researchers. The fear of premature scrutiny may have unintended consequences that lead to over-sanitized or linear reporting and suppress the messy but creative side of qualitative work (Mauthner et al., Reference Mauthner, Parry and Backett-Milburn1998).

As the debate continues, rebuttals and measures to address these concerns are also mounting in the literature. Countering the anxiety about scrutiny over one’s own research, some researchers reframe openness as a hallmark of credibility rather than a threat. As Elman and Kapiszewski (Reference Elman and Kapiszewski2014) put it, a more thoughtful and comprehensive application of data access and research transparency allows qualitative scholars to better showcase the robustness and rigor of their work and enable a broader audience to understand and interpret their findings. The availability of a detailed account of method and data serves to benefit multiple groups: Experienced researchers can verify or audit the research (DuBois et al., Reference DuBois, Strait and Walsh2017), students can retrace the process for learning purposes (Becker, Reference Becker2019), and the public can gain more trust in research through enhanced transparency (DuBois et al., Reference DuBois, Strait and Walsh2017). The risk of confidentiality breaches can be mitigated as national legislation and archive procedures provide layered protection for sensitive materials. Archives such as the UK Data Archive all maintain “comprehensive systems for protecting data” that include secure servers, access controls, and enforceable end-user licenses, which can also be used for protecting proprietary interests while preserving research access (Bishop, Reference Bishop2009, p. 261). In regard to rebalancing power and preventing abuse, Bishop (Reference Bishop2009) reminds us that researchers hold reciprocal duties not only to protect participants but also to benefit society, duties difficult to fulfill without data access. Importantly, when participants are asked, many consent to data sharing (Kuula, Reference Kuula2011) and endorse this broader vision to advance science (Bishop, Reference Bishop2009). In this sense, data archiving need not reinforce extractive power differentials and instead can honor participants’ wish to inform future scholarship (DuBois et al., Reference DuBois, Strait and Walsh2017).

In addition to mitigating concerns and articulating the benefits, a sub-strand of the pro-sharing literature is dedicated to practical solutions that translate principles into workflows. Joyce et al. (Reference Joyce, Douglass, Benwell, Rhys, Parry, Simmons and Kerrison2023) posit that the methodological traditions of conversation analysis provide a compelling case for the value of sharing qualitative data and offer practical models that other qualitative approaches could adopt. Grounded in ethnomethodology, conversation analysis treats data not as a static record but as a phenomenon constituted for the first time in any specific analysis and thereby dissolves the conventional distinction between primary and secondary data that often underpin contentions about data sharing (e.g., issues of context) (Joyce et al., Reference Joyce, Douglass, Benwell, Rhys, Parry, Simmons and Kerrison2023, p. 649). In addition to the more widely adopted practice of corpus sharing (i.e., archiving the complete dataset on an open repository), Joyce et al. (Reference Joyce, Douglass, Benwell, Rhys, Parry, Simmons and Kerrison2023) recommend a distinctive integrated research practice known as data sessions where recordings and transcripts are scrutinized by researchers outside the immediate project team to enhance the rigor and transparency of the analysis (p. 650). Recognizing the practical and non-trivial challenge of sharing contextually embedded data ethically and responsibly, Abramson and Dohan (Reference Abramson and Dohan2015) propose a more visual solution. They introduce a format called ethnoarray, which arranges coded ethnographic data in an interactive matrix so that secondary readers can trace how specific observations underwrite broader theoretical claims. Both proposals accept that contextualized materials require careful curation, but they maintain that thoughtful design choices can balance openness with ethical considerations. Along similar lines, Thomas and Bowen (this issue) propose participant narratives (i.e., co-constructed accounts developed collaboratively with participants) as an alternative to sharing raw transcripts that can better represent what participants intended to convey while reducing privacy risks.

Data reuse

Closely related to and partly overlapping with the debate on data sharing and replication is data reuse. While various typologies have been proposed to define data reuse (e.g., Heaton, Reference Heaton2004), the use cases can be thought of along two broad categories: verifying the original research findings, and generating new knowledge through reanalysis, the latter of which can be done solely on the original data or in conjunction with other primary or secondary data.

A central concern surfacing in this debate on data reuse (and the debate on data sharing) is the epistemological compatibility of data reuse and qualitative inquiry. As Heaton (Reference Heaton2004, p. 54) succinctly summarized, the crux of the tension lies in three central problems: “the problem of data ‘fit’” (i.e., whether data can be repurposed for new inquiries), “the problem of not having ‘been there’” (i.e., the impact of the secondary analyst’s detachment from the data generation process), and “the problem of verification” (i.e., whether data can be effectively used to confirm or dispute initial conclusions). Building on these concerns, Walters (Reference Walters2009) argues that reanalyzing historical qualitative data is prone to misinterpretation because the information is intrinsically linked to the social, cultural, and political realities of its time. Taking a step further, Mauthner et al. (Reference Mauthner, Parry and Backett-Milburn1998) assert that the notion of data reuse rests on a “naively realist” assumption that interview transcripts and fieldnotes exist as objective entities that can be detached from their contexts (p. 743). Similar reservations can be found in the discussion of participatory action research, where the goal is to empower communities by embedding inquiry within local social relations (Irwin, Reference Irwin2013). In Irwin’s (Reference Irwin2013) words, “it is hard to see what secondary analysis would add to such approaches, nor is data recorded and preserved in such a way that would make secondary analysis possible” (p. 298). Beck (Reference Beck2023) adds to the conversation from a nursing research perspective and warns that bias inherent in the primary dataset may be amplified by a secondary analyst who lacks the insider knowledge and relationship with the original researchers and participants. On the specific issue of whether data reuse can accomplish the task of verification, researchers have also expressed doubts. In Hammersley’s (Reference Hammersley1997) view, the concept of auditing research via archived data is flawed due to its incorrect epistemological foundation, which he refers to as “epistemological foundationalism,” which wrongly assumes that research is a simple deduction from self-evident data (p. 133).

In parallel and counter to the above narrative, some researchers view secondary analysis as a pragmatic and ethically responsible research strategy. An often-cited benefit is the economic advantage gained through reusing and reanalyzing data. Hinds et al. (Reference Hinds, Vogel and Clarke-Steffen1997) lamented that even though quantitative secondary data analysis was a “respected, common, and cost-effective” strategy for maximizing data utility, it was underutilized in qualitative research (p. 408). Subsequent bibliometric evidence suggests the trend may be turning: for instance, the number of qualitative secondary analyses in nursing journals alone has more than doubled each decade since the 1990s (Beck, Reference Beck2023), and a synthesis of 71 studies published between 2006 and 2016 confirms the steady normalization of qualitative secondary analysis across disciplines (Ruggiano & Perry, Reference Ruggiano and Perry2019). Drawing on collaborative, big-team research, Brower et al. (Reference Brower, Jones, Osborne-Lampkin, Hu and Park-Gaghan2019) demonstrate that pooling social media data or open-ended survey responses extends qualitative inquiry to questions that would not be possible to address with fresh fieldwork alone. Additionally, as Mauthner et al. (Reference Mauthner, Parry and Backett-Milburn1998) and Beck (Reference Beck2023) pointed out several decades apart, archived data also have high value as the raw material for historical and documentary analysis, which enables researchers to track emergent social trends by placing past narratives in dialogue with present concerns. Reusing archived interviews can also reduce participant burden and facilitate studies with vulnerable or hard-to-reach populations, an ethical gain when repeated data collection might exacerbate fatigue or risk (Beck, Reference Beck2023).

Notably, counterarguments to the “not having been there” problem (Heaton, Reference Heaton2004) also emerge to reframe a secondary analyst’s remove from the original data as an analytical affordance rather than loss. This essentially shifts the dialogue from “should we reuse qualitative data?” to “how can we do so meaningfully?” Mason (Reference Mason2007) advocates for an “investigative epistemology” that treats archived datasets as springboards for new questions rather than repositories of fixed facts. Echoing the same sentiment, Moore (Reference Moore2007) directly challenges Mauthner and colleagues’ (e.g., 1998) claims by arguing that they are “almost as realist as those they critique, in their implication that because it is not possible to archive everything, including the reflexive production of data, that the possibilities of meaning-making from secondary data are limited” (p. 6). In line with Mason (Reference Mason2002), Moore (Reference Moore2007) argues that anti-reuse views often rest on the mistaken belief that primary researchers and participants hold an exclusive claim to knowledge derived from the study. Instead of viewing reanalysis as an analysis of pre-existing data, it should be understood as an active process of “recontextualizing and reconstructing data” within the current/new project, the value of which lies not in trying to reproduce the past but in attending to the reflexive producing of meaning and context at present (Moore, Reference Moore2007). More recently, Hughes et al. (Reference Hughes, Hughes and Tarrant2022) operationalized this vision by proposing a typology of continuous, collective, and configurative modes of secondary qualitative analysis. These approaches showcase that varying degrees of temporal and epistemic distance from data are not deficits but analytic leverages. Tarrant and Hughes (Reference Tarrant, Hughes, Kara and Khoo2021) extend this logic by reframing reanalysis as an ethical imperative whereby preserving and revisiting data becomes a collective responsibility to safeguard marginalized voices across historical contexts.

Publication policy

A parallel strand in the conversation about open qualitative research asks how journals and professional bodies should provide guidelines for openness. At its center is a governance question: who gets to write the rules, and whose epistemology do those rules serve? Lorenz and Holland (Reference Lorenz and Holland2020) set the tone by warning that any attempt at systemic reform will risk privileging certain paradigms over others and reproducing methodological and epistemological inequity. This concern is echoed in political science, where early criticism of the transparency movement cautioned that imposing a single template on diverse research traditions would ignore the multiple logics of research that characterize qualitative inquiry (Jacobs et al., Reference Jacobs, Büthe, Arjona, Arriola, Bellin, Bennett, Björkman, Bleich, Elkins, Fairfield, Gaikwad, Greitens, Hawkesworth, Herrera, Herrera, Johnson, Karakoç, Koivu, Kreuzer and Yashar2021). While not targeting open science issues per se, Lazaraton (Reference Lazaraton2003) articulated similar concerns regarding L2 journal guidelines that prioritize specific research methodologies. This critique was accompanied by a call for researchers to move beyond simple adherence to methodological checklists and instead critically examine the fundamental principles and evaluative standards that underpin their work (Lazaraton, Reference Lazaraton2003).

Most of the controversies regarding publication guidelines center on sharing data and materials. Prosser et al. (Reference Prosser, Hamshaw, Meyer, Bagnall, Blackwood, Huysamen, Jordan, Vasileiou and Walter2023) point out that many journals gesture toward open data but provide guidance specifically written for quantitative archives. This leaves qualitative authors without adequate support to navigate vague directives that can feel more punitive than enabling. Preparing context-rich datasets for public release is both time-consuming and resource intensive and the burden lands hardest on early-career scholars, colleagues at less well-funded institutions, as well as researchers studying sensitive topics or working in high-risk settings (Jacobs et al., Reference Jacobs, Büthe, Arjona, Arriola, Bellin, Bennett, Björkman, Bleich, Elkins, Fairfield, Gaikwad, Greitens, Hawkesworth, Herrera, Herrera, Johnson, Karakoç, Koivu, Kreuzer and Yashar2021). These considerably higher ethical and practical burdens of making qualitative research open deepen equity concerns over who can afford to publish in leading outlets and what kinds of topics get published.

Just as there is wide agreement that one-size-fits-all prescriptions are ill-suited for the philosophical plurality within qualitative research (Yadav, Reference Yadav2022), there is a general consensus against creating a single set of publication rules for openness. Drawing on an extensive consultative process, Jacobs et al. (Reference Jacobs, Büthe, Arjona, Arriola, Bellin, Bennett, Björkman, Bleich, Elkins, Fairfield, Gaikwad, Greitens, Hawkesworth, Herrera, Herrera, Johnson, Karakoç, Koivu, Kreuzer and Yashar2021) conclude that publication policy should rely on differentiated judgment rather than blanket expectations. Editors are urged to weigh four factors when considering data sharing requests: the intellectual gain to be had, the minimum volume of materials required, the ethical risk to participants, and the labor involved in preparing files for meaningful reuse. Instead of rigid mandates, editors and reviewers might defer to an author’s reasoned justification for their transparency choices (Jacobs et al., Reference Jacobs, Büthe, Arjona, Arriola, Bellin, Bennett, Björkman, Bleich, Elkins, Fairfield, Gaikwad, Greitens, Hawkesworth, Herrera, Herrera, Johnson, Karakoç, Koivu, Kreuzer and Yashar2021). Two very recent sets of journal guidelines exemplify how differentiated judgments are being codified within L2 research. TESOL Quarterly’s (n.d.) open science principles (based on Al-Hoorie et al., Reference Al-Hoorie, Cinaglia, Hiver, Huensch, Isbell, Leung and Sudina2024) outline tiered expectations covering transparent methodological description (required), open materials (required where possible), open data (encouraged), and preregistration (encouraged). Complementing this, The Modern Language Journal (Hiver et al., Reference Hiver, Al-Hoorie and Huensch2026) is rolling out Transparency in Research and Reporting Practices guidelines in 2026 and 2027 that promote open materials, code, and data for all research submitted to the journal. Together, these documents illustrate how publication policy is moving toward structured, yet flexible, expectations that can accommodate epistemic diversity while elevating transparency and rigor. Beyond our field too, many qualitative journals (e.g., Qualitative Inquiry, the International Journal of Qualitative Methods, Qualitative Research, and The Qualitative Report) have positioned themselves as active participants in open science reform by innovating and reframing what open means so that it fits with their particular epistemology, ethics, and practice. There is an increasing move in elite qualitative research journals toward sharing practices that are “as open as possible, as closed as necessary.”