Introduction
Open science is a broad and inclusive framework for making scientific knowledge openly available, widely accessible, and reusable while fostering collaboration, information sharing, and greater openness across research practices (UNESCO, 2021). This movement has gained increasing attention across diverse research fields, including second and foreign language (hereafter L2) research (e.g., Al-Hoorie et al., Reference Al-Hoorie, Cinaglia, Hiver, Huensch, Isbell, Leung and Sudina2024; Kremmel & Isbell, Reference Kremmel and Isbell2024; Liu, Reference Liu2023; Marsden et al., Reference Marsden, Morgan-Short, Thompson and Abugaber2018; Plonsky, Reference Plonsky2024). These open science practices enhance the transparency of original studies by allowing other researchers to access data and materials used in prior research. This accessibility enables researchers to reuse instruments or data to evaluate or replicate original findings (Plonsky, Reference Plonsky2024). Consequently, open science may offer advantages in reducing the risks of questionable research practices (QRPs), such as p-hacking, selectively reporting only statistically significant results, and HARKing (i.e., hypothesizing after the results are known) (see Larsson et al., Reference Larsson, Plonsky, Sterling, Kytö, Yaw and Wood2023, for these QRPs).
A key advantage of open science is its potential to facilitate various forms of secondary research. Based on previous studies (e.g., Cheong et al., Reference Cheong, Lyons, Houghton and Majumdar2023; Chong & Plonsky, Reference Chong and Plonsky2024; Heaton, Reference Heaton2004, Reference Heaton2008), we define secondary research as scholarly work that does not involve the collection of new primary data but instead reviews, summarizes, synthesizes, or reanalyzes previously published studies or previously collected datasets for secondary purposes. These secondary purposes may include addressing new research questions (e.g., Heaton, Reference Heaton2008; Ruggiano & Perry, Reference Ruggiano and Perry2019), examining existing data through different analytical approaches (e.g., Nicklin et al., Reference Nicklin, McLean and Vitta2026), and synthesizing findings across multiple studies (e.g., Norris & Ortega, Reference Norris and Ortega2000; Plonsky, Reference Plonsky2011). By contrast, primary research involves the collection of data directly from research participants to address research gaps identified in previous studies. Such research may aim to answer research questions (e.g., Chen et al., Reference Chen, Lin, Chen and Fu2023), test research hypotheses (e.g., McEown & Sugita-McEown, Reference McEown and Sugita-McEown2019), or develop and validate research instruments (e.g., C. Li et al., Reference Li, Jiang and Dewaele2018; Yashima & Nishida, Reference Yashima and Nishida2024). The data collected or generated through primary research can also serve as the foundation for secondary research.
When discussing secondary research, it is important to consider how data are collected through primary research and how they may be used in secondary analyses. This is because the nature of the original research context shapes the possibilities and limitations of secondary research. For example, in interview-based qualitative research, data and interpretations are co-constructed through interactions between participants and researchers (e.g., Kvale & Brinkmann, Reference Kvale and Brinkmann2015; Rapley, Reference Rapley2001). When such data are reused in secondary research, the original context and researcher-participant dynamics may influence not only the original data collection process but also the use and interpretation of the data in secondary qualitative data analysis. Therefore, when conducting secondary qualitative data analysis, researchers need to critically examine the original context in which data were generated or collected, as this context influences how existing data can be appropriately interpreted and reused.
To our knowledge, secondary research typically encompasses three major categories: (a) synthesis-based data analysis (e.g., systematic reviews and meta-analyses), (b) narrative reviews, and (c) reanalysis-based data analysis. Figure 1 illustrates the categorization of secondary research used in this article. The first major category, (a) synthesis-based data analysis, can be further divided into three levels. The first level consists of systematic reviews, which synthesize trends and general findings across previous studies, regardless of whether the original data or findings are quantitative or qualitative (e.g., Al-Hoorie et al., Reference Al-Hoorie, Oga-Baldwin, Hiver and Vitta2025; Hiver et al., Reference Hiver, Al-Hoorie and Evans2022, Reference Hiver, Al-Hoorie, Vitta and Wu2024). The second level focuses on synthesis-based quantitative data analysis, such as meta-analyses (e.g., Al-Hoorie, Reference Al-Hoorie2018; Boulton & Cobb, Reference Boulton and Cobb2017; S. Li, Reference Li2010; Norris & Ortega, Reference Norris and Ortega2000; Plonsky, Reference Plonsky2011; Webb et al., Reference Webb, Uchihara and Yanagisawa2023; Yanagisawa & Webb, Reference Yanagisawa and Webb2021).Footnote 1 The third level involves synthesis-based qualitative data analysis, as demonstrated in previous work on qualitative evidence synthesis (e.g., Chong & Plonsky, Reference Chong and Plonsky2021).
Classification of secondary research.
Figure 1. Long description
A flowchart with three levels of hierarchy.
At the top level is a central box labeled Secondary research.
Three arrows point down from this root to the second level:
1. To the left, Synthesis-based data analysis.
2. In the center, Narrative review.
3. To the right, Reanalysis-based data analysis.
The third level branches out from the left and right categories:
From Synthesis-based data analysis, three arrows point to:
- Systematic review.
- Synthesis-based quantitative data analysis e.g., meta-analysis.
- Synthesis-based qualitative data analysis, which is highlighted with a thick red border.
From Reanalysis-based data analysis, two arrows point to:
- Reanalysis-based quantitative data analysis.
- Reanalysis-based qualitative data analysis, which is highlighted with a thick blue border.
At the bottom of the diagram, a box with a thick green border is labeled Secondary qualitative data analysis. Two arrows originate from this green box and point upward to the red-bordered Synthesis-based qualitative data analysis and the blue-bordered Reanalysis-based qualitative data analysis boxes respectively.
Narrative reviews summarize the literature on a specific topic to identify strengths and weaknesses and to critically evaluate findings in order to propose directions for future research (e.g., DeKeyser & Prieto Botana, Reference DeKeyser and Prieto Botana2015; Thomson & Derwing, Reference Thomson and Derwing2015). They are often used “to provide an authoritative argument based on published primary evidence” and “to enhance understanding of a topic and theory development” (Furley & Goldschmied, Reference Furley and Goldschmied2021, p. 2). The category of narrative reviews can include various types of reviews, such as conceptual reviews (e.g., Thomas & Rose, Reference Thomas and Rose2019) and critical reviews (e.g., Teng & Zhang, Reference Teng and Zhang2022), indicating that different forms of narrative reviews serve different purposes (Furley & Goldschmied, Reference Furley and Goldschmied2021). Notably, while systematic reviews follow more rigorous and standardized procedures, such as adhering to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelinesFootnote 2 to evaluate data collection methods and inclusion criteria, narrative reviews typically do not follow such formalized processes (Chong & Plonsky, Reference Chong and Plonsky2024; Ferrari, Reference Ferrari2015). Consequently, narrative reviews may be more subjective and potentially more vulnerable to arbitrariness than systematic reviews, a tendency often referred to as cherry picking (Greenhalgh et al., Reference Greenhalgh, Thorne and Malterud2018), suggesting that the personal perspectives of researchers may play a greater role in the synthesis. However, as Greenhalgh et al. (Reference Greenhalgh, Thorne and Malterud2018) also argue, systematic reviews are not free from subjectivity. Decisions about how to frame a topic, which questions to prioritize, and what constitutes relevant evidence also involve interpretive judgment. In this regard, they emphasize the importance of transparency in the review process so that the perspectives, choices, and interpretive positions of researchers are made explicit.
Reanalysis-based data analysis can be divided into two types: reanalysis-based quantitative data analysis and reanalysis-based qualitative data analysis. This type of research involves reusing original quantitative or qualitative datasets for secondary purposes, such as replicating findings or examining the data through different theoretical frameworks, research questions, or analytical methods (e.g., Heaton, Reference Heaton2008; Irwin, Reference Irwin2013; Koesten et al., Reference Koesten, Vougiouklis, Simperl and Groth2020; Nicklin et al., Reference Nicklin, McLean and Vitta2026). Unlike review or synthesis research, reanalysis-based data analysis directly reuses the original datasets generated in primary studies, and these datasets may be reanalyzed either by the original researchers or by other researchers (Heaton, Reference Heaton2008).
It should be noted that this article distinguishes between two broad types of secondary qualitative data analysis: synthesis-based secondary qualitative data analysis (i.e., the synthesis of qualitative findings or datasets) and reanalysis-based secondary qualitative data analysis (i.e., the reuse and reanalysis of a limited qualitative research dataset). This categorization is based on the assumption that synthesis-based secondary qualitative data analysis encompasses relatively large amounts of qualitative findings (or datasets) from multiple studies or sources and synthesizes them to address new research questions (e.g., Chong & Plonsky, Reference Chong and Plonsky2021). In contrast, the reanalysis-based secondary qualitative data analysis focuses narrowly on a limited dataset to investigate research questions in greater depth by applying different research frameworks and theoretical or analytical lenses (e.g., Dufour & Richard, Reference Dufour and Richard2019; Tate & Happ, Reference Tate and Happ2018). As these two types of analysis involve substantially different ranges of datasets and analytical approaches, they cannot be readily subsumed under a single category of secondary qualitative data analysis; therefore, we treat them as analytically distinct in this article (see Figure 1). Additionally, we assume that the reanalysis-based secondary qualitative data analysis generally requires access to the complete dataset, as opposed to the partial use of findings that may be sufficient for synthesis-based secondary qualitative data analysis. Accordingly, this article establishes a clear distinction between synthesis-based secondary qualitative data analysis and reanalysis-based secondary qualitative data analysis based on the analytical approach (i.e., synthesis vs. reanalysis), the number of datasets involved (i.e., a large number vs. a limited number), and the type of data used (i.e., partial findings or datasets vs. full datasets). Crucially, however, these two categories should not be understood as mutually exclusive. For instance, some studies seem to incorporate elements of both synthesis and reanalysis (e.g., Cheong et al., Reference Cheong, Lyons, Houghton and Majumdar2023; Heaton, Reference Heaton2015), indicating that a simple binary classification of secondary qualitative data analysis types would be overly reductive. Rather, the classification proposed here can serve as an organizing framework that helps researchers understand the broader landscape of secondary research and situate their own studies within it. Although individual studies may incorporate elements of both approaches, the classification provides a basis for identifying the primary methodological orientation of a study and does not imply that approaches can be combined arbitrarily. The selection and combination of approaches should instead be determined by the research purpose and research questions.
Taking these different types of secondary qualitative data analysis into account, this article primarily focuses on the reuse and reanalysis of a limited primary qualitative dataset for secondary analytical purposes (i.e., the reanalysis-based secondary qualitative data analysis), an area that has received relatively limited attention in L2 research.
What data can be used for secondary qualitative data analysis?
Secondary qualitative data analysis has gained increasing recognition and interest in social research since the mid-1990s. This type of research involves reusing qualitative data collected in prior studies (Heaton, Reference Heaton2008). Recent work by Cheong et al. (Reference Cheong, Lyons, Houghton and Majumdar2023) has expanded this scope by including existing research data from primary studies, as well as other resources not originally intended for research purposes. A growing body of literature has discussed its methodological foundations, analytic procedures, ethical considerations, and issues of rigor and validity (e.g., Camfield & Palmer-Jones, Reference Camfield and Palmer-Jones2013; Chatfield, Reference Chatfield2020; Chauvette et al., Reference Chauvette, Schick-Makaroff and Molzahn2019; Corti, Reference Corti2007; Dunn et al., Reference Dunn, Arslanian-Engoren, DeKoekkoek, Jadack and Scott2015; Heaton, Reference Heaton2004, Reference Heaton2008; Long-Sutehall et al., Reference Long-Sutehall, Sque and Addington-Hall2011; Thome, Reference Thome1998). Empirical and methodological applications of secondary qualitative data analysis have also been reported across diverse fields such as health research, development studies, and sociology (e.g., Cheong et al., Reference Cheong, Lyons, Houghton and Majumdar2023; Dufour & Richard, Reference Dufour and Richard2019; Heaton, Reference Heaton2015; Szabo & Strang, Reference Szabo and Strang1997; Whiteside et al., Reference Whiteside, Mills and McCalman2012), demonstrating its potential for extending the value of existing qualitative datasets. However, despite this expanding literature, the application of secondary qualitative data analysis remains rare in L2 research, and systematic discussions of how such analysis can be operationalized in L2 research are largely lacking. This gap suggests the need to begin by clarifying what types of data can be used for secondary qualitative data analysis in L2 research and how such data can be analyzed and interpreted in light of existing studies, rather than emphasizing originality at this early stage of inquiry. This article mainly adopts Heaton’s (Reference Heaton2008) definition of secondary qualitative data analysis; however, we also embrace more flexible and innovative approaches to data collection adopted by Cheong et al. (Reference Cheong, Lyons, Houghton and Majumdar2023). These approaches may enable researchers to conduct secondary qualitative data analysis using a broader range of data, including data collected for both research and non-research purposes. This is based on the recognition that non-research data, such as social media content and interviews that were not originally conducted for research purposes, can be valuable sources for secondary analysis (e.g., Cheong et al., Reference Cheong, Lyons, Houghton and Majumdar2023; Selvi, Reference Selvi2026; Shahri & Karizmeh, Reference Shahri and Karizmeh2025).
Qualitative data, such as interviews, diaries, field notes, and responses to open-ended questions, often contain rich and detailed information about participants, which may raise legal and ethical concerns owing to the identifiable nature of the data. Consequently, the sharing and reuse of qualitative data for secondary research (i.e., secondary qualitative data analysis) have become a critical issue and have been the subject of ongoing debate among researchers (Broom et al., Reference Broom, Cheshire and Emmison2009; Chatfield, Reference Chatfield2020; Chauvette et al., Reference Chauvette, Schick-Makaroff and Molzahn2019; Irwin, Reference Irwin2013; Ruggiano & Perry, Reference Ruggiano and Perry2019). Possible solutions and the potential benefits of conducting secondary qualitative data analysis have also been discussed (e.g., Cheong et al., Reference Cheong, Lyons, Houghton and Majumdar2023; Heaton, Reference Heaton2008; Ruggiano & Perry, Reference Ruggiano and Perry2019). These discussions may be in line with broader global initiatives supporting open science and data reuse, which have gained increasing momentum, particularly since the mid-1990s, when the Organization for Economic Cooperation and Development advocated open access to publicly funded research data (Chauvette et al., Reference Chauvette, Schick-Makaroff and Molzahn2019). A notable example is the establishment of QualiData in 1994 by the UK Economic and Social Research Council, which is now part of the UK Data Service, as the world’s first archive of qualitative data (Chauvette et al., Reference Chauvette, Schick-Makaroff and Molzahn2019). Extending such foundational efforts, countries such as Canada, the United States, and Australia have been actively developing the infrastructure and frameworks necessary to facilitate the sharing and secondary use of qualitative data (Chauvette et al., Reference Chauvette, Schick-Makaroff and Molzahn2019). The following sections introduce a more detailed discussion of the processes, risks, and benefits associated with sharing qualitative data.
Given the growing interest in and global support for sharing qualitative data, a critical next step is considering the appropriate types of qualitative data for secondary analyses. This requires careful conceptual clarification, especially in distinguishing secondary analysis from related practices, such as documentary analysis. According to Heaton (Reference Heaton2008), in her book “Reworking Qualitative Data,” published in 2004, the “focus on non-naturalistic qualitative data distinguishes secondary analysis from documentary analysis (Plummer, Reference Plummer1983, Reference Plummer2001), which involves working with naturalistic or ‘found’ materials, such as auto-biographies, personal diaries and photographs” (p. 34).
However, Heaton (Reference Heaton2008) noted that “some types of qualitative material, such as life stories and diaries, may be subject to either secondary analysis or documentary analysis, depending on to what extent the material was solicited and shaped by researchers’ involvement in collecting the material” (p. 34). Therefore, clearly defining and assessing the nature of the possible qualitative data is important, as not all such data may be suitable for secondary analysis. Table 1 summarizes the types of qualitative data in L2 research that may be ideal for either secondary or documentary analysis, based on Heaton’s (Reference Heaton2004, Reference Heaton2008) descriptions. Data appropriate for secondary analysis typically include materials collected through the direct involvement of researchers, such as interview transcripts, responses to open-ended survey questions, classroom observation notes, and teacher or learner logs.
Classification of possible qualitative data in L2 research: secondary vs. documentary analysis
Table 1. Long description
The table consists of three columns: Type of analysis, Type of data, and Examples of data.
* The first row describes Secondary analysis. The data type is Non-naturalistic data involving researchers. Examples include interview transcripts, responses to open-ended survey questions, classroom observation notes, teacher or learner logs, stimulated recall data, focus group data, and researcher field notes.
* The second row describes Documentary analysis. The data type is Naturalistic data with no researcher involvement. Examples include autobiographies, personal diaries, photographs, blog posts, forum discussions, social media posts, learner essays written independently, and teaching materials created without research intent.
* The third row describes Either analysis depending on the context. The data type includes both non-naturalistic and naturalistic data depending on researcher involvement. Examples include classroom video or audio recordings, reflective journals, lesson plans, institutional reports, and student portfolios.
By contrast, documentary analysis involves naturally occurring data not produced for research purposes, including autobiographies, personal diaries, photographs, blog entries, social media posts, learners’ essays, and teaching materials developed independently by practitioners. Certain types of data, such as classroom videos or audio recordings, reflective journals, lesson plans, and student portfolios, may be included in either category depending on the context in which they are produced. The classification depends on the degree of researcher involvement and the original purpose of the data collection. Even seemingly naturalistic data can be suitable for secondary analysis if they were initially generated through researcher-initiated processes.
Recent technological developments have enabled access to a wider range of online sources or databases (e.g., Carmichael, Reference Carmichael, Fielding, Lee and Blank2017; Chatfield, Reference Chatfield2020; Cheong et al., Reference Cheong, Lyons, Houghton and Majumdar2023). Such resources allow researchers to record and archive various forms of data, regardless of whether the data were originally collected for research purposes. Although Cheong et al. (Reference Cheong, Lyons, Houghton and Majumdar2023) did not explicitly clarify their position regarding the use of qualitative data that can be used for secondary qualitative data analysis (e.g., data originally collected for research purposes vs. data not collected for research purposes) as discussed by Heaton (Reference Heaton2008), they adopted a more flexible approach. As demonstrated in their study, establishing clear criteria for evaluating data reliability through the PRISMA flowchart allows researchers to determine whether target data, including data not originally collected for research purposes, can be used for secondary qualitative data analysis. This growing inclusion of data not originally collected for research purposes may reflect the impact of technological progress, which enables the use of diverse data resources, provided that appropriate guidelines are in place. It is important to note, however, that Cheong et al.’s secondary qualitative data analysis is closer to synthesis-based analysis than to reanalysis-based analysis as suggested by Heaton (Reference Heaton2008). This implies that data criteria, such as the inclusion of non-research data in secondary qualitative data analysis, may differ depending on the type of secondary analysis (e.g., synthesis-based vs. reanalysis-based). Therefore, the selection of qualitative data for secondary qualitative data analysis should ultimately be based on research goals and questions (e.g., exploring common patterns across multiple L2 qualitative studies vs. examining a specific L2 qualitative dataset in depth), theoretical frameworks (e.g., examining learners’ cognitive learning vs. investigating socially situated practices), and the intended type of analysis (i.e., synthesis vs. reanalysis).
How to conduct secondary qualitative data analysis
Heaton (Reference Heaton2008) introduces three main modes of secondary data analysis. Table 2 summarizes these three modes and provides potential examples from L2 research. The first mode is “formal data sharing” (Heaton, Reference Heaton2008, p. 35), in which researchers obtain archived datasets from public or institutional repositories and examine them for research purposes that differ from those of the original study. Because such datasets are generally prepared for broader use, they are often accompanied by relevant documentation and shared only after meeting necessary ethical and legal requirements. Heaton noted that, although secondary researchers may communicate with the original researchers when clarification is needed, the original researchers typically do not participate as collaborators in subsequent publications. To the best of our knowledge, the practice of openly sharing qualitative data has not yet evolved sufficiently, and therefore is not widely adopted in the field of L2 research. However, certain leading journals, such as Studies in Second Language Acquisition and Language Learning, recommend that researchers share their data. In addition, repositories for research data and instruments, such as the Open Science Framework (OSF: https://osf.io/),Footnote 3 Figshare (https://figshare.com/), and Instruments and Data for Research in Language Studies (IRIS: https://www.iris-database.org/), allow researchers to store and archive research materials, thereby creating opportunities for secondary qualitative data analysis. Through the policies of such academic journals and the preservation of data in repositories, qualitative research datasets may be archived, allowing other researchers to access them for future secondary qualitative data analysis. For instance, a recent study by Sunardi et al. (Reference Sunardi, Rahayu, Rusmawaty, Hermagustiana and Sunggingwati2025) shared their interview data through a journal platform and Figshare, allowing other researchers to reuse these data for secondary qualitative data analysis. This represents an example of formal data sharing.
Three modes of secondary qualitative data analysis based on Heaton (Reference Heaton2008)
Table 2. Long description
The table is organized into three columns: Type of mode, Datasets, and Examples in L2 research.
* Row 1: Formal data sharing. Datasets are described as archived from public or institutional repositories with documentation, shared after ethical and legal requirements are met. Examples include L 2 qualitative datasets like interview transcripts, survey responses, and observation notes shared via OSF, Figshare, and IRIS.
* Row 2: Informal data sharing. Datasets are directly exchanged among researchers through independent sharing, collaborative secondary analysis, or integration of multiple datasets. Examples include L 2 qualitative datasets shared among researchers for secondary analyses.
* Row 3: Self-collected data. Datasets collected by original researchers are reused for subsequent research or integrated by original researchers for collaborative analyses. Examples include L 2 qualitative datasets reused by original researchers, original teams, or collaborative teams involved in the original studies.
The second mode is referred to as “informal data sharing” (Heaton, Reference Heaton2008, p. 35). Unlike formal data sharing, this approach does not necessarily involve depositing datasets in public or institutional archives. Data are exchanged directly among researchers through various collaborative arrangements. For example, original researchers may share their datasets with other researchers without participating in subsequent analyses, collaborate with new researchers as members of a secondary analysis team, or combine independently collected datasets for collaborative secondary analyses. The involvement of original researchers can, therefore, influence the nature of secondary analyses because they may contribute contextual knowledge, including information about the participants, gained during the initial data collection process. In L2 research, for example, researchers may contact original researchers to obtain L2 qualitative datasets and reanalyze these datasets using different theoretical frameworks or research questions. Alternatively, original researchers who have independently collected L2 qualitative data may form a collaborative team to combine their respective datasets and conduct further analyses, allowing them to obtain a more comprehensive understanding of specific L2 learning phenomena.
The third mode concerns the reuse of “self -collected data” (Heaton, Reference Heaton2008, p. 35). In this mode, researchers return to qualitative datasets from their own previous studies and analyze them from new perspectives. Such reanalysis may serve different purposes, including extending the scope of the original research, exploring previously unexamined questions, or reconsidering earlier interpretations through further analysis. In L2 research, this approach can involve either collaborative efforts among researchers who contributed to previous studies or the independent reuse of existing datasets by original researchers. In the former case, researchers may reanalyze datasets from multiple L2 studies. Although this resembles informal data sharing, participation is limited to researchers who were involved in the original studies. In the latter case, individual researchers or original research teams reuse their own datasets without involving external researchers (Heaton, Reference Heaton2008).
We now turn to how primary data can be used for secondary qualitative data analysis in several ways. For example, if researchers aim to examine new research questions and hypotheses using data originally collected by others, they may adopt the second mode described above. In such cases, obtaining consent to reuse the data from the original researchers and the participants involved in data collection is advisable from ethical and legal standpoints. This is because such secondary analysis may require access to the entire dataset, which could potentially reveal detailed information about the participants in the original studies. This type of secondary qualitative data analysis also requires careful consideration of how findings are generated by secondary analysts who were not involved in the original data collection. Because qualitative research often emphasizes the sociocultural context in which researchers and participants co-construct meanings and interpretations (e.g., Flick, Reference Flick2023; Kvale & Brinkmann, Reference Kvale and Brinkmann2015; Rapley, Reference Rapley2001), data and findings are inherently shaped by these interactions. Therefore, failing to consider the epistemological stance, methodological frameworks, analytical procedures, research context, and interview questions of the original study may lead secondary researchers to overlook how these factors shaped the data collected and the accounts elicited from participants. This may result in interpretations that are insufficiently grounded in the original research context.
The three modes discussed above typically involve original data collected for research purposes, either by the same researchers or others; however, another important category should be considered. This involves original data not initially collected for research purposes. In a recent study by Cheong et al. (Reference Cheong, Lyons, Houghton and Majumdar2023), they demonstrated that such data could be valuable for secondary analyses. They employed extensive search techniques to collect a wide range of qualitative data, including data generated for purposes other than research (i.e., “publicly available interviews or anecdotes from around the world” [p. 6]). In contrast to Heaton’s (Reference Heaton2004, Reference Heaton2008) more restrictive view, this broader approach would be feasible if researchers clearly define the criteria for evaluating the reliability and relevance of such data. Cheong et al. (Reference Cheong, Lyons, Houghton and Majumdar2023) support this view by showing that non-research data can offer meaningful insights when carefully assessed. In addition, recent technological developments have expanded access to various types of datasets, such as social media content, which may enable researchers to conduct various types of secondary data analyses or related studies (e.g., Selvi, Reference Selvi2026; Shahri & Karizmeh, Reference Shahri and Karizmeh2025). This shift suggests that more flexible methods, such as those proposed by Cheong et al., may offer broader advantages than approaches based on narrower definitions of eligible data. However, the criteria for data selection, including whether non-research data can be incorporated, may vary depending on the type of secondary qualitative data analysis (i.e., synthesis-based vs. reanalysis-based).
In other cases, researchers may consider integrating selected qualitative findings from primary studies using such diverse sources as free descriptions, narratives, case studies, or classroom observations. This may be more accurately characterized as a synthesis of qualitative findings (i.e., synthesis-based qualitative data analysis), as introduced by Chong and Plonsky (Reference Chong and Plonsky2021), rather than reanalysis-based secondary qualitative data analysis. In such instances, rather than reanalyzing the entire dataset for a specific purpose, researchers may choose to synthesize key patterns or particularly important findings from the original studies. In this regard, this approach may not require the direct involvement of the original researchers because secondary researchers typically do not need access to the full dataset and collect and extract relevant findings from published research papers or reports. Chong and Plonsky (Reference Chong and Plonsky2021) introduced a procedure for this type of secondary qualitative research in an L2 learning and teaching context. They explained how qualitative data or findings can be integrated into secondary research, how data can be collected for such studies, and how those data can subsequently be analyzed. In addition, they indicated the limited applicability of qualitative secondary research in applied linguistics. Nevertheless, L2 researchers can draw on various modes of qualitative secondary research, including synthesis-based and reanalysis-based qualitative data analysis. These approaches may enable them to integrate findings across qualitative studies, reanalyze existing datasets without collecting new data, and generate new insights by applying different theoretical or methodological frameworks.
Benefits and risks of secondary research using qualitative data
Although secondary qualitative data analysis is prevalent in other research fields, such as medical and social research, this practice remains limited in L2 studies. This may be due, in part, to the fact that primary qualitative data collected by other researchers are often not available for secondary analysis. Nevertheless, previous studies in other academic fields have discussed various advantages of sharing qualitative data and conducting secondary qualitative data analysis (e.g., Ruggiano & Perry, Reference Ruggiano and Perry2019; Sherif, Reference Sherif2018; Tate & Happ, Reference Tate and Happ2018). A potential benefit of this approach is that it allows researchers to reuse existing data to examine new research questions and hypotheses (Ruggiano & Perry, Reference Ruggiano and Perry2019), thereby generating new insights and contributing to knowledge advancement. In addition to enabling the investigation of new perspectives, secondary qualitative data analysis can be used to reexamine the original findings using the same data. This makes it particularly valuable for identifying alternative interpretations or for extending prior conclusions from different theoretical or contextual perspectives. Another potential advantage of this method is that it can improve time efficiency (Sherif, Reference Sherif2018). Qualitative data collection typically requires substantial time from both researchers and participants. As such, reusing existing qualitative data may reduce the burden on both parties. Therefore, data sharing and secondary analysis offer the potential to generate valuable new findings across various research fields while reducing the burden of data collection.
Despite its potential benefits, qualitative data sharing for secondary qualitative data analysis has several limitations and challenges (e.g., Campbell et al., Reference Campbell, Javorka, Engleton, Fishwick, Gregory and Goodman-Williams2023; Chauvette et al., Reference Chauvette, Schick-Makaroff and Molzahn2019; Cheong et al., Reference Cheong, Lyons, Houghton and Majumdar2023; Feldman & Shaw, Reference Feldman and Shaw2019; Mozersky et al., Reference Mozersky, Parsons, Walsh, Baldwin, McIntosh and DuBois2020; Ruggiano & Perry, Reference Ruggiano and Perry2019; Tamminen et al., Reference Tamminen, Bundon, Smith, McDonough, Poucher and Atkinson2021). One key challenge concerns how qualitative data can be shared and reused for secondary purposes. As mentioned above, Heaton (Reference Heaton2008) explains three different modes of secondary qualitative research and analysis that involve decisions about who collects and shares data and how the analysis is conducted. As Heaton noted, the extent to which primary researchers are involved in secondary studies can greatly affect how secondary research is conducted. For instance, if primary researchers are not involved, secondary analysts may face limitations in accessing or interpreting data owing to ethical and legal constraints. This may hinder their ability to analyze the data as originally intended. While this issue may be resolved if primary researchers are fully involved in the secondary project or if they reuse their own data for a new purpose, such involvement is not always feasible.
Another challenge closely related to this issue is the extent to which qualitative data can be shared, both ethically and legally. As Chauvette et al. (Reference Chauvette, Schick-Makaroff and Molzahn2019) indicated, some organizations have begun to establish guidelines for sharing and archiving qualitative data for secondary use. However, such guidance, particularly when adapted to qualitative data, has not been well established in the L2 field. Consequently, concerns remain regarding the potential ethical and legal violations associated with data sharing. Therefore, developing clear criteria to evaluate the sensitivity of personal data may be beneficial. Figure 2 illustrates the different levels of ethical concerns regarding the secondary use of qualitative data.
Levels of ethical concerns in secondary qualitative data analysis.
Figure 2. Long description
The diagram consists of three vertical rectangular panels positioned above a thick blue double-headed arrow labeled Ethical concerns, with Lower on the left and Higher on the right.
Level 1. Low ethical sensitivity.
Ethical concerns. Minimal to none, provided that the data are properly anonymized and clear consent for future reuse is obtained during the original research process.
L2 research examples.
* Open-ended survey responses on learners’ preferred study methods and teachers’ instructional practices.
* Anonymized forum posts or public blog entries that have been publicly archived, as well as publicly available school curricula for L2 education.
* Data for which explicit consent has been obtained for secondary use in future educational or research contexts.
Level 2. Moderate ethical sensitivity.
Ethical concerns. Potential for contextual re-identification. Unclear participant expectations regarding secondary use. Personal experiences and perspectives that do not involve highly sensitive information.
L2 research examples.
* Interview transcripts, such as those of students’ language learning experiences, teachers’ professional experiences, parents’ experiences of supporting L2 learning, and other personal but non-sensitive aspects of L 2 learning.
* Reflective journals containing some personal details but no directly identifiable information.
* Data collected with consent forms that contain only general statements regarding possible future academic use, without explicitly addressing secondary use.
Level 3. High Ethical Sensitivity.
Ethical concerns. Risk of identifying individual research participants. Possible emotional harm resulting from the reuse of sensitive personal narratives. Lack of informed consent for secondary use.
L2 research examples.
* Audio or video recordings of classroom interactions in which participants are clearly identifiable through their voices and appearances.
* Interviews and related transcripts documenting deeply personal language learning struggles such as anxiety, trauma, or discrimination.
* Data collected without any mention of future reuse in the consent forms.
In this figure, we classify ethical concerns into three levels: lower, moderate, and higher. At the lower level, data are appropriately archived, and participants cannot be identified even when different researchers reuse the data for secondary analysis. Additionally, informed consent regarding the potential reuse of collected qualitative data in future research is obtained during the original research process. For example, open-ended responses that do not contain any identifying information about research participants, anonymized forum posts and public blog entries that have been publicly archived, and publicly available school curricula for L2 education can be categorized at this level.
At the moderate level, although data are properly archived, they may contain potentially re-identifiable information or non-sensitive personal information. In addition, participants may not fully understand the expectations regarding secondary use. For example, interview transcripts, such as those concerning students’ language learning experiences, teachers’ professional experiences, and parents’ experiences of supporting their children’s L2 learning, and other personal but non-sensitive aspects of L2 learning may be included in this level. Furthermore, reflective journals containing some personal details but no directly identifying information and data collected with consent forms that include general statements regarding possible future academic use can also be categorized at this level.
At the higher level, the data include clearly identifiable information about participants and may contain sensitive content that could cause emotional harm or raise religious or cultural concerns. Consent from participants may also not have been appropriately obtained for the secondary use of collected qualitative data. For example, audio or video recordings in which participants are clearly identifiable are included in this category. Additionally, interviews discussing deeply personal language learning struggles (e.g., anxiety, trauma, and discrimination) and data involving sensitive topics, such as gender identity, immigration experiences/status, and political views, can also be categorized at this level.
As this figure illustrates, establishing appropriate criteria for assessing the level of concern associated with sharing and reusing qualitative data is essential, particularly when such data are used by researchers who were not involved in the primary study. Importantly, although our examples represent only a limited range of qualitative data collected from participants, various types of qualitative data can be obtained from diverse groups in L2 contexts, such as learners, instructors, parents, migrants, and educational institutions. Therefore, applying sound judgment and careful evaluation is crucial when conducting secondary qualitative data analysis.
An additional issue arises from an epistemological perspective (e.g., Campbell et al., Reference Campbell, Javorka, Engleton, Fishwick, Gregory and Goodman-Williams2023; Chauvette et al., Reference Chauvette, Schick-Makaroff and Molzahn2019; Feldman & Shaw, Reference Feldman and Shaw2019). As discussed earlier, qualitative research often involves a co-constructed process between researchers and participants, in which findings emerge through their interaction (e.g., Flick, Reference Flick2023; Kvale & Brinkmann, Reference Kvale and Brinkmann2015; Rapley, Reference Rapley2001). In particular, the individuals involved in data collection are closely linked to the construction of the findings. This may be especially evident in interview-based data collection, as interactions, questioning, and responses may vary considerably during interviews depending on who is involved, even when researchers use the same topic. However, in secondary qualitative data analysis, the original researchers who collected the data are not always involved (Heaton, Reference Heaton2008). Accordingly, the epistemological grounding of the data may be weakened. In this regard, some consideration of the involvement of the original researchers may be necessary depending on the type of data. For instance, secondary analysis of interview data may be most appropriately conducted either by the original researchers or with their involvement because such data use may raise epistemological issues as described above. When a secondary researcher is required to conduct a qualitative secondary analysis without any involvement of the original researchers, it would be preferable to contact them and check the findings and interpretations to ensure that they do not diverge from the original research context. For data such as logs or responses to open-ended questions, which may be collected without the active involvement of the original researchers, secondary analysis conducted without the original researchers may be more acceptable.
On the other hand, reanalysis-based secondary qualitative data analysis can take various forms, including studies conducted with or without the involvement of original researchers (Heaton, Reference Heaton2008). However, when full datasets are reused, at least one researcher involved in the original study should preferably participate in the secondary research. The involvement of original researchers would help to mitigate the ethical and legal issues associated with data reuse. In addition, clarifying the roles of all contributors in the research and writing processes would be essential. This consideration is often discussed in relation to authorship ethics and research transparency (e.g., Astaneh et al., Reference Astaneh, Schwartz and Guyatt2021; De Costa et al., Reference De Costa, Sterling, Lee, Li and Rawal2021). As previously noted, qualitative findings can differ depending on the researcher and context involved. Therefore, explicitly reporting who analyzed the data during preregistration and in the manuscript may contribute to more accurate interpretation of the findings and enhance the overall transparency and credibility of the research.
Conducting reanalysis-based secondary qualitative data analysis
A body of qualitative research has accumulated in L2 research to date (Benson et al., Reference Benson, Chik, Gao, Huang and Wang2009; Riazi et al., Reference Riazi, Rezvani and Ghanbar2023; Richards, Reference Richards2009, for reviews), yet practices for openly sharing qualitative data remain limited (Liu et al., Reference Liu, Al-Hoorie and Hiver2026). In response to this gap, many scholars (e.g., Al-Hoorie et al., Reference Al-Hoorie, Cinaglia, Hiver, Huensch, Isbell, Leung and Sudina2024; Isbell & Kremmel, Reference Isbell and Kremmel2024; Plonsky, Reference Plonsky2024; Winke, Reference Winke2024) have advocated the adoption of open science practices in recent years. This advocacy has been accompanied by the emergence of platforms such as IRIS and Figshare, which provide opportunities for publicly sharing datasets. Such platforms may in turn facilitate qualitative data sharing (e.g., Siegel, Reference Siegel2026; Sunardi et al., Reference Sunardi, Rahayu, Rusmawaty, Hermagustiana and Sunggingwati2025) and, by extension, reanalysis-based secondary qualitative data analysis in L2 research.
Drawing on insights from the existing literature on secondary qualitative data analysis and qualitative data reuse (e.g., Chatfield, Reference Chatfield2020; Cheong et al., Reference Cheong, Lyons, Houghton and Majumdar2023; Heaton, Reference Heaton2004, Reference Heaton2008) and the potential risks and challenges associated with secondary qualitative data analysis (e.g., Branney et al., Reference Branney, Brooks, Kilby, Newman, Norris, Pownall, Talbot, Treharne and Whitaker2023; Campbell et al., Reference Campbell, Javorka, Engleton, Fishwick, Gregory and Goodman-Williams2023; Ruggiano & Perry, Reference Ruggiano and Perry2019; Tamminen et al., Reference Tamminen, Bundon, Smith, McDonough, Poucher and Atkinson2021), we propose a tentative framework to guide future secondary qualitative data analysis, with a particular focus on reanalysis-based secondary qualitative data analysis in L2 research (see Figure 3). The proposed framework comprises four steps. The framework is intended to outline the overall procedure by illustrating how reanalysis-based secondary qualitative data analysis can be conducted through these four steps. In the following sections, we explain how each step of the framework can be applied to reanalysis-based secondary qualitative data analysis in L2 research.
An example framework for conducting reanalysis-based secondary qualitative data analysis.
Figure 3. Long description
A cyclical flowchart with four numbered boxes connected by thick yellow arrows in a clockwise loop.
1. Clarifying the research goal and plan. This step involves creating a research plan. Key questions include identifying investigation aims, necessary data types like interviews or observations, data sources such as existing databases, ethical and legal data management, involvement of original researchers, and the epistemological stance.
An arrow points right to Step 2.
2. Pre-registering research if applicable. This involves submitting the research plan to a journal or registry. It requires presenting goals and contributorships, explaining data storage and ethical management, clarifying original researcher involvement or data reuse permission, and describing anticipated contributions and limitations.
An arrow points down to Step 3.
3. Conducting research. This step focuses on executing the research plan. Actions include collecting and analyzing data based on goals and epistemological stance, consulting original researchers, interpreting findings, ensuring data confidentiality, and evaluating trustworthiness through credibility, dependability, and transferability.
An arrow points left to Step 4.
4. Presenting and reflecting on the results. This involves reporting the findings. Tasks include deciding which findings to report, defining contributorships, indicating data usage and sharing potential, and reflecting on the process to revisit the research plan for future studies, with an arrow pointing back up to Step 1.
Readers may find it easier to understand the process if it is illustrated with a concrete example. Thus, we present an illustrative example of reanalysis-based secondary qualitative data analysis using the study by Sunardi et al. (Reference Sunardi, Rahayu, Rusmawaty, Hermagustiana and Sunggingwati2025), which investigated the emotional experiences of graduate preservice student teachers in Indonesia. We selected this study because the authors provided the interview transcripts as supplementary qualitative data, enabling us to illustrate how existing qualitative data may be reanalyzed for secondary analytical purposes.
In this example, we consider a hypothetical reanalysis-based secondary qualitative data analysis of the interview data provided by Sunardi et al. (Reference Sunardi, Rahayu, Rusmawaty, Hermagustiana and Sunggingwati2025). Their dataset was originally generated through interviews with 15 graduate preservice student teachers during their teaching practicums at the Teacher Training and Education Faculty of Mulawarman University. The dataset (i.e., interview transcripts) is publicly available through Figshare, making it a potential source for secondary qualitative data analysis, subject to relevant ethical and data-use considerations.
Step 1: clarifying the research goal and plan
In reanalysis-based secondary qualitative data analysis, the first step is to clarify the research goals and questions and to design a research plan accordingly. As Cheong et al. (Reference Cheong, Lyons, Houghton and Majumdar2023) noted, formulating clear research goals and questions is a crucial starting point in secondary qualitative data analysis, where researchers work with existing data rather than collecting new data. The formulation of research goals and questions in secondary qualitative data analysis can take various forms. For instance, some L2 researchers may examine whether different researchers generate different findings and interpretations when analyzing the same dataset, either under the same research goals and questions or from different analytical perspectives. Alternatively, secondary researchers may reanalyze existing datasets collected by other researchers to address new research questions or apply different theoretical frameworks. Our illustrative example is situated in the latter context, as we assume that secondary researchers reuse the dataset originally generated by Sunardi et al. (Reference Sunardi, Rahayu, Rusmawaty, Hermagustiana and Sunggingwati2025). Their dataset is publicly shared through a journal platform and Figshare; therefore, this type of data sharing can be categorized as formal data sharing (cf. Heaton, Reference Heaton2008). In their original study, Sunardi et al. investigated the emotional experiences of graduate preservice teachers during teaching practicums in Indonesia through the lens of Hargreaves’ (Reference Hargreaves1998, Reference Hargreaves2001) emotional geography framework. In this hypothetical secondary analysis, researchers reanalyze the interview transcripts through the lens of emotion regulation to explore how graduate preservice teachers regulate their emotions and what emotion regulation strategies they use in response to their emotional experiences.
In this step, researchers must also consider potential ethical and legal challenges associated with secondary analysis (e.g., Campbell et al., Reference Campbell, Javorka, Engleton, Fishwick, Gregory and Goodman-Williams2023; Chauvette et al., Reference Chauvette, Schick-Makaroff and Molzahn2019; Feldman & Shaw, Reference Feldman and Shaw2019; Long-Sutehall et al., Reference Long-Sutehall, Sque and Addington-Hall2011; Mozersky et al., Reference Mozersky, Parsons, Walsh, Baldwin, McIntosh and DuBois2020; Tamminen et al., Reference Tamminen, Bundon, Smith, McDonough, Poucher and Atkinson2021). Because Sunardi et al. (Reference Sunardi, Rahayu, Rusmawaty, Hermagustiana and Sunggingwati2025) openly shared their data through an appropriate platform (i.e., Figshare), it is likely that relevant procedures for data sharing were considered. Nevertheless, such procedures are not always applicable to all studies. In particular, Long-Sutehall et al. (Reference Long-Sutehall, Sque and Addington-Hall2011) argued that informed consent obtained in the primary study cannot be automatically assumed to cover secondary data analysis. Thus, L2 researchers should verify whether the original informed consent permits secondary use or requires additional approval, and establish a data governance plan that protects confidentiality, prevents identification, and accounts for the sensitivity of identity-related data.
Beyond clarifying research goals and questions and addressing ethical and legal considerations, researchers should also carefully plan how secondary analysis will be conducted. Such planning involves determining the analytical perspectives to be adopted, the types of data required, and considerations of researchers’ positionality and contributorship. For instance, because this hypothetical reanalysis-based secondary qualitative data analysis aims to explore how graduate preservice teachers regulate their emotions and what emotion regulation strategies they use in response to their emotional experiences, Gross’s (Reference Gross and Gross2014) framework could be applied. Researchers should also consider the roles of those involved in the research process, including who collected, analyzed, and interpreted the original data and findings. Importantly, data selection for reanalysis-based secondary qualitative data analysis should be primarily guided by the intended research purposes and analytical approaches rather than by data availability alone. For example, if researchers aim to explore how graduate preservice teachers regulate their emotions in response to their emotional experiences, interview data may be particularly suitable because they provide access to participants’ reflections on and interpretations of their emotional experiences and regulation strategies. In our illustrative example, the interview transcripts originally collected by Sunardi et al. (Reference Sunardi, Rahayu, Rusmawaty, Hermagustiana and Sunggingwati2025) are used as the dataset because they contain participants’ accounts of their emotional experiences during teaching practicums. Whereas the original study examined these experiences through the lens of Hargreaves’ (Reference Hargreaves1998, Reference Hargreaves2001) emotional geography framework, the hypothetical reanalysis-based secondary qualitative data analysis reexamines the same data through the lens of emotion regulation, focusing on how graduate preservice teachers regulate their emotions and what emotion regulation strategies they use in response to these experiences.
It is important to note that secondary qualitative data analysis can take different forms depending on who conducts the reanalysis (cf. Heaton, Reference Heaton2008). In some cases, original researchers may reuse their own datasets for secondary purposes. In other cases, as in our hypothetical scenario, researchers who were not involved in the original data collection must identify and obtain access to suitable datasets for analysis. Considering these options, three main routes may be available for accessing qualitative data in L2 research. The first route is to contact the original researchers and request access to their qualitative datasets. The second route is to use datasets available through repositories such as IRIS and Figshare, which often provide documentation and metadata that facilitate data sharing and transparent reanalysis. The third route is to systematically identify relevant datasets through database searches (e.g., Cheong et al., Reference Cheong, Lyons, Houghton and Majumdar2023; Chong & Plonsky, Reference Chong and Plonsky2021). Although this approach is particularly relevant for synthesis-based secondary analyses, it may also help researchers locate potential datasets for reanalysis-based secondary qualitative data analysis.
Importantly, when using qualitative data collected by others without participants’ prior consent for secondary use, it is essential to contact the original researchers and, where appropriate, the participants through the original researchers to explain the research goals, analytical methods, and intended use of the findings in the reanalysis-based secondary qualitative data analysis.Footnote 4 Such communication can help mitigate the ethical and legal risks inherent in secondary qualitative data analysis. Additionally, even when original researchers reuse data that they themselves collected, additional explanation and consent from participants may be warranted if the data are used to address research purposes or questions that extend beyond those described during the original consent process, particularly when the resulting findings could potentially have adverse consequences for the original participants (e.g., findings that may stigmatize learners’ language-learning difficulties and negatively portray teachers’ instructional practices). If consent cannot be obtained from the original participants, particularly when the reuse of the data may result in adverse consequences for them, reanalysis-based secondary qualitative data analysis may not be ethically permissible. Therefore, before conducting reanalysis-based secondary qualitative data analysis, researchers should carefully evaluate whether their proposed research questions and purposes extend beyond the scope of the original consent and whether the intended analysis could potentially result in adverse consequences for the original participants.
Taken together, in preparing for reanalysis-based secondary qualitative data analysis, researchers should adopt a systematic initial step that includes formulating appropriate research goals and questions (e.g., examining new hypotheses or reanalyzing existing data using different theoretical frameworks), identifying the type and source of data, research planning (e.g., data collection, use, analysis, and management with ethical considerations), determining whether the original researchers will be involved, and clarifying contributorship.
Step 2: preregistering research (optional but recommended)
Once a research plan or proposal has been created (e.g., research goals or questions, target data, and types of data analyses), researchers may consider preregistering their study on dedicated platforms. Such preregistration may include information on research goals or questions, data collection methods, the involvement of original researchers, data analyses and theoretical frameworks adopted, researchers’ positionality or contributorship (i.e., the roles of each researcher), as well as ethical, legal, and epistemological considerations regarding the data and the sharing of results. For reanalysis-based secondary qualitative data analysis, preregistration may be particularly challenging because researchers may sometimes need to examine the availability, scope, and characteristics of existing datasets before finalizing their research questions and analytical approaches. Therefore, preregistration may be considered a voluntary rather than mandatory step. Nevertheless, preregistration can provide a useful means of documenting research plans and enhancing transparency in reanalysis-based secondary qualitative data analysis.
It is worth noting that in recent years, leading journals in applied linguistics, such as Language Learning, have introduced registered reports, which allow researchers to preregister their study proposals prior to data collection to enhance research transparency.Footnote 5 However, because this model presupposes that data collection has not yet occurred, it does not directly correspond to the practice of reanalysis-based secondary qualitative data analysis, in which the dataset already exists. In this respect, online repositories such as OSF and Figshare may offer a more suitable alternative: unlike journal-based registered report systems, which involve formal peer review of the proposal prior to data collection, these repositories allow researchers to document and share their analytical plans for existing datasets without requiring prior review, thereby enhancing transparency in a way that better accommodates the nature of reanalysis-based secondary qualitative data analysis.
The importance and benefits of preregistration, such as enhancing the transparency of research processes and obtaining timely feedback on research designs (if such feedback procedures or systems are available), have also been highlighted by Branney et al. (Reference Branney, Brooks, Kilby, Newman, Norris, Pownall, Talbot, Treharne and Whitaker2023) in the field of psychology. In line with this perspective, we include preregistration as a step in the framework and encourage researchers to undertake it for two main reasons. First, because reanalysis-based secondary qualitative data analysis is a relatively novel approach in L2 research, preregistration may provide an opportunity to obtain early feedback on research goals, ethical considerations, and data governance before the study begins. In this hypothetical scenario, researchers could preregister a research plan for a reanalysis-based secondary qualitative data analysis examining how graduate preservice teachers regulate their emotions and what emotion regulation strategies they use in response to their emotional experiences, provided that suitable preregistration or registered report options are available. Such a plan may include, for instance, what data will be used and how they will be accessed (i.e., interview transcripts from Sunardi et al., Reference Sunardi, Rahayu, Rusmawaty, Hermagustiana and Sunggingwati2025); how ethical and legal issues will be addressed, including how consent from the original researchers and participants will be obtained where necessary; what research goals and questions will be examined (e.g., how graduate preservice teachers regulate their emotions and what emotion regulation strategies they use in response to their emotional experiences); how the data will be analyzed (e.g., through semantic coding, theoretical coding, and thematic analysis); and how the findings will be shared publicly. Through preregistration, researchers may receive feedback on their research plans (if such opportunities are available through preregistration systems). In this way, researchers can refine their study design and carefully consider how to implement this relatively novel methodological approach.
A second reason is that preregistration can enhance transparency by clarifying researchers’ epistemological and methodological orientations. This transparency includes documenting key aspects of the research process, such as how data are collected, how data are analyzed, and which theoretical or analytical frameworks are used. Such information can provide a clearer account of how reanalysis-based secondary qualitative data analysis is conducted. Preregistration may also contribute to clearer contributorship statements (Branney et al., Reference Branney, Brooks, Kilby, Newman, Norris, Pownall, Talbot, Treharne and Whitaker2023), which help clarify how each researcher will contribute to different aspects of the research process. In other words, clarifying researchers’ contributions and positionalities can enhance transparency regarding who will be involved in the generation and interpretation of research findings and how those findings will be produced throughout the research process.
Step 3: conducting research
At this stage, researchers collect, analyze, and interpret data in accordance with predefined roles (i.e., contributorship; cf. Branney et al., Reference Branney, Brooks, Kilby, Newman, Norris, Pownall, Talbot, Treharne and Whitaker2023), thereby ensuring transparency in the construction of findings. Transparency also requires clarity about how data are selected and collected, which, as noted earlier, should be guided by research goals and questions. Although researchers differ in their views on what data are suitable for secondary qualitative data analysis (e.g., Cheong et al., Reference Cheong, Lyons, Houghton and Majumdar2023; Heaton, Reference Heaton2008), we adopt the position of Cheong et al. (Reference Cheong, Lyons, Houghton and Majumdar2023) that both research and non-research data can be used for secondary purposes (even though our hypothetical scenario targeted research data collected from research participants).
Once the target data have been identified and obtained, researchers can proceed with the analysis while considering relevant analytical considerations such as analytical approaches, lenses, and frameworks. When researchers who were not involved in the original data collection conduct the analysis, consulting the original researchers to interpret the findings in light of the epistemological stance and sociocultural context of the original study is especially important. In our hypothetical scenario, researchers seek to explore how graduate preservice teachers regulate their emotions and what emotion regulation strategies they use in response to their emotional experiences. To address this aim, the data from Sunardi et al. (Reference Sunardi, Rahayu, Rusmawaty, Hermagustiana and Sunggingwati2025) may be coded and categorized according to themes related to participants’ emotional experiences and emotion regulation strategies. The coded data may then be further analyzed using Gross’s (Reference Gross and Gross2014) process model of emotion regulation to examine how these strategies are reflected in participants’ accounts of their emotional experiences.
Finally, the findings should be evaluated in terms of their trustworthiness, using criteria such as credibility, dependability, and transferability (Cheong et al., Reference Cheong, Lyons, Houghton and Majumdar2023).Footnote 6 To ensure trustworthiness, two aspects are particularly important: transparency in the analytical process and critical reflection on the researcher’s positionality. Transparency enables readers to understand how interpretations were generated, while critical reflection helps account for the potential influence of researchers’ prior knowledge and assumptions, particularly when the analyst was not involved in the original data collection. Accordingly, researchers should explicitly articulate their analytical procedures and positionality. In the hypothetical scenario, the findings and interpretations may also vary depending on who analyzes the data (e.g., researchers involved in preservice teacher education or those without such involvement) and which theoretical framework and analytical approach are adopted (e.g., Gross’s process model of emotion regulation as the theoretical framework, combined with thematic analysis, narrative analysis, or grounded theory as the analytical approach). Given this variability, researchers need to justify their choice of theoretical framework and analytical approach, while also making their own positionality explicit. In particular, researchers’ familiarity with the research context, as well as their experience and expertise in preservice teacher education and emotion regulation, may influence how they interpret the data. Therefore, explicitly clarifying researchers’ positionalities, alongside the rationale for their analytical choices, is important for enhancing the transparency and trustworthiness of the analysis. By making their relationships to the research context and roles in the original and secondary studies explicit, researchers can help readers understand how their perspectives and prior knowledge may have shaped the coding, interpretation, and development of findings. Such transparency is particularly important in reanalysis-based secondary qualitative data analysis, where researchers may interpret data collected by others without having been directly involved in the original data collection process.
Step 4: presenting and reflecting on the results
After completing the analysis, interpretation, and evaluation, researchers move on to the final step: selecting findings to report, clarifying the contribution of each researcher, specifying the involvement of the original researchers (if any), and determining data-sharing procedures (where applicable). Ideally, this process should involve consultation with the researchers who collected the original data, given the ethical and legal concerns discussed above. In the hypothetical scenario, the findings are assumed to show how graduate preservice teachers regulate their emotions and what emotion regulation strategies they use in response to their emotional experiences. To ensure that these findings are reported responsibly, researchers should first carefully consider whether reporting certain findings could result in adverse consequences for the original participants, for instance, by revealing identifiable information, disclosing sensitive personal experiences, stigmatizing or negatively portraying individual participants or groups, or violating the conditions under which the original data were collected and shared. Where such risks are identified, researchers may consult the original researchers and, where appropriate, the participants before disseminating the results. Based on this assessment, researchers should determine not only whether such findings can be reported, but also how they should be presented. For example, they may need to remove or modify potentially identifying details, avoid reporting findings that could cause unnecessary harm to participants, or provide sufficient contextual information to prevent misleading or overly negative interpretations of the data.
Reanalysis-based secondary qualitative data analysis concludes by addressing the core research questions formulated by the secondary analysts. However, if new issues emerge and further investigation is required, researchers may return to the initial stage of the framework and begin a new cycle of reanalysis-based secondary qualitative data analysis.
Our framework is intended to serve as a preliminary model for reanalysis-based secondary qualitative data analysis in L2 research. While further refinement and adaptation are necessary, this article offers a practical guide for L2 researchers to engage in this type of inquiry. Using a hypothetical scenario involving graduate preservice student teachers during their teaching practicums at the Teacher Training and Education Faculty of Mulawarman University (Sunardi et al., Reference Sunardi, Rahayu, Rusmawaty, Hermagustiana and Sunggingwati2025), we illustrate how data collected in an original study can be reused for secondary qualitative data analysis and how such analysis can address new research aims and perspectives. Although the concept and practice of reanalysis-based secondary qualitative data analysis did not originate in L2 research, our simplified framework draws on key insights from other fields to illustrate how this approach can be applied to L2 research.
Conclusion
This article introduces the potential of conducting reanalysis-based secondary qualitative data analysis in L2 research. As noted by Chong and Plonsky (Reference Chong and Plonsky2021), although secondary qualitative research is well established in other fields, it remains rarely applied in L2 studies. While their work primarily focused on synthesis-based qualitative secondary research, this article presents an alternative approach: reanalysis-based secondary qualitative data analysis. Despite several ethical, legal, and epistemological challenges associated with secondary qualitative data analysis, it offers substantial benefits, such as enabling researchers to reanalyze existing datasets through new theoretical lenses and reducing the demands associated with primary data collection. Drawing on insights from previous research and acknowledging these potential risks, this article proposes a framework for reanalysis-based secondary qualitative data analysis, illustrated through a hypothetical scenario based on the research context of graduate preservice student teachers during their teaching practicums at the Teacher Training and Education Faculty of Mulawarman University (Sunardi et al., Reference Sunardi, Rahayu, Rusmawaty, Hermagustiana and Sunggingwati2025). Although this framework is preliminary and needs further refinement, it aims to establish a more systematic, transparent, and ethical foundation for reanalysis-based secondary qualitative data analysis in L2 research. We hope that this work will encourage researchers to further explore the potential of reanalysis-based secondary qualitative data analysis as a valuable and underutilized methodological approach in future L2 studies.
Data availability statement
This article does not have any original data available for sharing or reuse.
Acknowledgments
We sincerely appreciate the journal editor, Luke Plonsky, and the guest editors of this special issue, Meng Liu, Ali H. Al-Hoorie, and Phil Hiver, for their kind guidance and generous support. We would also like to thank the anonymous reviewers for providing us with very helpful comments and suggestions, all of which were important in improving the quality of our article.
Funding
This work was supported by JSPS KAKENHI Grant Number JP25K04295.
Competing interests
The authors have no competing interests to declare.
Declaration of generative AI and AI-assisted technologies in the writing process
During the preparation of this work, the authors used ChatGPT (GPT-5.6 Luna and earlier versions) to improve the grammatical accuracy, clarity, and coherence of English sentences. After using the tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication.
Open access
