Considering Types of Data

14 Managing Qualitative Research Data

Dr. Joel T. Minion

Learning Outcomes

By the end of this chapter you should be able to:

  1. Identify what distinguishes qualitative data from other forms of research data.
  2. Understand the iterative processes by which qualitative researchers generate and manage data.
  3. Describe ways Research Data Management could better encompass qualitative data and the needs of qualitative researchers.
  4. Advocate for greater inclusivity of all types of research data in Research Data Management principles, policies, strategies, and practices.


Sound data management is essential to research excellence. Most higher-learning institutions support initiatives in this area, but few such efforts focus on qualitative data or its researchers. Attend the average training session and you’ll be forgiven for thinking that Research Data Management (RDM) applies primarily to data involving numbers or geospatial images. While exceptions exist (e.g., the First Nation Principles of OCAP® — ownership, control, access, possession), acknowledgement of qualitative research data often feels like an afterthought. This is probably because qualitative data are highly descriptive, typically text or voice based, and collected solely from humans, which makes such data particularly identifiable. Qualitative data also require researchers to account for social context and relationships, and are commonly generated by studies involving sensitive topics and marginalized communities. Such challenges mean qualitative research data seldom fit neatly into prevailing RDM frameworks.

In this chapter, we’ll consider why this area of RDM remains underexamined and how deficiencies might be addressed. The content reflects what I have learned over 25 years spent variously as a librarian, qualitative health researcher, data manager, and educator in Canada and Europe. Like many of my qualitative colleagues, I struggle to fit myself into existing RDM principles, policies, strategies, and practices. There are few experts on this topic and only limited resources, so this chapter is not a how-to guide.

There are many forms of qualitative data and myriad ways in which they can be generated, analyzed, organized, archived, shared, and in some cases, reused. We’re going to discuss where the management of qualitative data fits within the research process. If you’re a researcher, this means exploring how to think about and organize your data more effectively. If you’re a librarian, archivist, or other type of data specialist, the discussion should augment your information management skills with a stronger understanding of how qualitative data come to be.

The chapter is divided into three sections: (1) the nature of qualitative research data, (2) how such data reflect the qualitative research process, and (3) RDM-related challenges when collecting qualitative data. Finally, we’ll discuss how to improve the management of qualitative research data.

The Nature of Qualitative Data

Qualitative data are created and analyzed in ways dissimilar to quantitative and digital humanities data. This doesn’t mean that different types of research data are mutually exclusive or cannot work together. Many researchers use multiple methods in their research, such as combining interviews with psychometric testing to answer a question, like “How is clinical depression experienced by individuals caring for a partner living with early stage dementia?” Such approaches illustrate the interconnectedness of different types of research data.

What Makes Research Data Qualitative?

Qualitative data share no single philosophy or set of methodological principles. They are data generated by research examining social aspects of the human condition using descriptive methods rather than measurement. Researchers can engage with and observe individuals in a multitude of ways to understand how people interact and make sense of their world in different settings: at home, work, in the community, while receiving healthcare, and so on. Qualitative research has its roots in the social sciences, particularly anthropology, sociology, and psychology, though researchers from other disciplines also work from a qualitative perspective. For example, researchers in nursing commonly use qualitative data to examine patients’ lived experiences.

Qualitative data can be collected during a single point of contact or through interaction over an extended period. What is captured is always filtered through the researcher and their experience and interpretation of interactions with participants. In this way, the researcher becomes an integral part of the data. Qualitative data are important because they provide information that cannot otherwise be measured or counted, such as how Afghani refugees make sense of government services when arriving in Canada or what it’s like to compete as a Paralympian or why some people are drawn to alt-right movements.

What Do Qualitative Data Look Like?

The most common ways to generate qualitative data are interviews, focus groups, and observations (e.g., you could interview refugees one on one about their experiences, hold focus groups with athletes, or watch what happens during an alt-right gathering). These are often used because they are relatively straightforward to learn and practice at a basic level. Other methods include oral histories, participant diaries, photography/videography, document analysis, artifacts (e.g., food, clothing), and open-ended survey questions.

These methods can be used in combination, resulting in interconnected datasets. A researcher interested in how climate scientists collaborate may conduct observations at a conference, where they also interview attendees and gather presentation handouts. Qualitative researchers often keep reflexive journals to reflect on their place within a project, to capture emergent ideas, and to identify new lines of inquiry. Researchers may also turn to social media for data, such as examining online discussions where people interact independent of the researcher. While a qualitative lens is increasingly applied in this way, the focus of this chapter is on data collected by a researcher.

Exercise: Working with Less Common Forms of Qualitative Data

Imagine you are the data librarian at a university. A graduate student asks for advice on how to manage data collected for a study of people undergoing treatment for cervical cancer. The methods will involve interviews combined with photovoice, an approach about which you know little. Skim the paper below and identify questions to ask about the photographs being collected and how these might be managed.

Wang, C., & Burris, M. A. (1997). Photovoice: Concept, methodology, and use for participatory needs assessment. Health Education & Behavior24(3), 369-387.

While qualitative data come in a variety of forms, text is by far the most common. Interviews and focus groups are typically audio-recorded using hand-held devices, with recordings then transcribed for analysis.

The Complexity of Transcribing

Transcription is time consuming and challenging to do without proper equipment (e.g., good headphones, specialized software). Many qualitative researchers outsource this work, though doing so can raise concerns about cost and a possible need to transfer data out of Canada.

The process also requires researchers to decide how much detail to transcribe. Is every “umm,” “err,” or false start to be captured (often referred to as “full verbatim”)? Or is the goal simply to produce a readable version of what was said (“verbatim”)? Such decisions are critical because different forms of qualitative analysis require specific levels of transcription.

Finally, all transcripts must be verified for accuracy prior to analysis. This involves listening to each recording while reading the transcript to catch mistakes and omissions.

Video recording is less common, in part because some participants find it more intrusive, so it may require a higher level of buy-in before people agree to take part. Video can also be more demanding to analyze. Observational data is usually captured using handwritten, typed, and/or audio field notes depending on circumstance and version (e.g., handwritten in the field with dictated audio notes created and transcribed later).

Less common forms of qualitative data vary widely in how they are handled. Hard copies, like meeting minutes or conference handouts, likely require scanning prior to storage and analysis. Participant diaries may need to be typed before they can be analyzed. Digital photographs can be stored in different formats depending on a researcher’s needs and preferences, while artifacts may be photographed or worked with in their original form using notes.

How Qualitative Data Become New Knowledge

Effective management of qualitative data requires understanding the analytic process. A biologist measuring fish populations in northern lakes will probably use a software program (e.g., SPSS, Stata, R) to analyze their data statistically, but how does a sociologist extract meaning from an interview transcript? Most often, doing so involves working inductively upwards from the data to identify higher order concepts and meanings. The goal is to look beyond what was said, heard, observed, photographed, and so on, to recognize ideas crosscutting an entire dataset. With text-based data, analysis may involve coding and use of qualitative data analysis software (e.g., NVivo, Quirkos). While software can handle large volumes of data, the programs themselves do not analyze data. That’s the researcher’s job. Furthermore, not all qualitative researchers code or use software. Some prefer using paper copies of transcripts, highlighters, pens, and index cards.

In some respects, data can be the most straightforward part of the qualitative research process. After all, transcripts from different focus group studies generally look the same: pages of text capturing what was said and by whom. The content, however, will reflect who was conducting each study and why. An anthropologist and a psychologist are likely to approach the same topic differently and ask contrasting questions. Data only take on meaning when they are analyzed. This process is complex because there is no single ontology, epistemology, theory, or mode of analysis crosscutting all forms of qualitative research. Researchers work from their own perspectives, so the same data could be interpreted in different ways depending on who is conducting the analysis and to what end.

Understanding Qualitative Research

To effectively manage qualitative data, you need to understand how qualitative research takes place. We’ll consider qualitative research practices from a data perspective. The aim is to link qualitative data to three key elements in the research process: how research teams are structured, the implications of such structures for data generation, and the place of participants in qualitative research.

Unlike its quantitative counterpart, which frequently turns to well-established processes (e.g., randomized clinical trials in medical research, validated survey instruments in psychology), qualitative research is more evolutionary and flexible. For instance, interview questions can evolve significantly into new, follow-up lines of inquiry across discussions with multiple participants. It’s even possible to add or remove data types once a study is underway (which would be the case, for example, if photographs prove not to be as useful as anticipated). Such decisions are never taken lightly, but that such changes are possible is characteristic to qualitative research.

Qualitative Research Teams

There can be considerable variation in the composition of research teams that use qualitative approaches to collect data. The range includes the following:

  • A researcher or graduate student (e.g., someone with a small grant to conduct 20 interviews about how single parents manage childcare concerns)
  • A group of researchers within a university or department (e.g., a senior academic and two postdoctoral fellows using focus groups and city planning documents to study proposed changes in urban traffic patterns)
  • A larger team from different disciplines at multiple universities (e.g., six mid-career researchers from energy engineering, business, and organizational psychology using observations and interviews to explore communication networks in teams installing offshore wind farms)
  • An international, multidisciplinary group of researchers collaborating across countries and field sites (e.g., three dozen researchers, graduate students, clinicians, and patient partners studying the impact of long COVID-19 on health outcomes in Canada, the United States, and the United Kingdom using interviews, analysis of medical records, and a longitudinal survey)

Over their careers, researchers may be involved in several such arrangements, though most will likely develop a preference or specialized skill set for one or two approaches in particular.

Any time a study team involves more than one researcher, there will almost certainly be relational hierarchies and contrasting levels of expertise, which translate into differing roles and responsibilities. As in quantitative research, a principal investigator (PI) is the researcher who leads a project and who may be supported by one or more co- or sub-investigators. The PI holds authority for a study and is accountable to institutions, funders, and ethics boards for how a project unfolds. In large teams, the PI may have little direct involvement in day-to-day research activities, including data generation and management. Early career researchers (e.g., graduate students, postdoctoral fellows) are frequently responsible for collecting, processing, organizing, and securing data.

The Relationship Between Research Teams and Data

The structure of a study team has implications for how qualitative research is conducted and what data are generated. Two elements in the relationship between teams and data are worth highlighting.

The first concerns the iterative nature of qualitative research and how this affects the data. Qualitative data are often analyzed from the moment they are collected, meaning data now influence data to come. For example, a researcher will use what they learn in one interview to determine what to ask in the next. Changes could be minor (e.g., a question reframed to make it clearer) or substantial (e.g., an entire line of questioning is added or dropped). Larger studies often rely on peer debriefing throughout data collection to generate new insights, address practical challenges, and enhance researchers’ skills. In this way, qualitative data are collected in a reflexive, progressive manner.

The second element in the relationship involves the division of roles and responsibilities within teams and how these can imprint on data. Because qualitative data are closely tied with the circumstances of their collection, who collects the data will impact what is collected. Unless details about the collection process are captured, qualitative data can lose their capacity to support rigorous analysis. For instance, focus group transcripts require particulars about the participants (e.g., age, profession, level of education) as well as field notes about the tone of the discussion and how participants interacted (e.g., rolled eyes are not caught on audio recordings).

Allowing each researcher on a team to capture and process their own contextual data introduces potential variation and possible omission of critical details. One way to avoid this is to assign a few people to serve as data stewards (ideally more than one in case that person leaves the project) who help ensure data management (e.g., file naming, folder structures) is consistent.

Exercise: Capturing Context

Congratulations! You’ve just been hired as a postdoctoral fellow for a study involving the observation of verbal and behavioural exchanges in Canadian courtrooms. The aim is to explore differences in interactions involving judges, lawyers, plaintiffs, and defendants from visible minority populations. The project team comprises a PI, one co-investigator at another university, two other recent PhD graduates located elsewhere, a study coordinator, and one master’s student.

The study will require up to 600 hours of observations by four team members in five cities. Because none of you will be able to speak to or record anyone being observed, data collection is limited to handwritten fieldnotes that each researcher will type and share afterwards.

The team has discussed which exchanges they are most interested in capturing. It becomes apparent, however, that the data collectors also need to capture context details more consistently. Your task in this exercise is to identify: (1) which information researchers should be recording beyond the exchanges themselves, (2) how such information might be collected systematically, and (3) how to link the contextual data to the courtroom data.

An audit trail is another practice helpful in qualitative research. This documentation tracks activity and decision making throughout the life of a project, detailing what took place, when, and why. Some of this will be captured in the data itself, but on larger projects a separate document accessible by multiple team members can be critical. The information recorded connects what is taking place at the team level to data generation in real time. For example, an audit trail helps a team avoid trying to recall when and why they decided midway through a project to introduce a new field site or data collection method. Unfortunately, there are few standards on how to create and manage such documents in qualitative research. As we will see in the final section, bringing together researchers and data/information specialists can address such challenges.

The Social Dimension of Qualitative Research

No overview of the qualitative research process is complete without acknowledging study participants. Qualitative research is relational, meaning researchers often interact directly with individuals taking part in a study. This relationship may be fleeting, as in the case of one-off interviews conducted by phone, though even these connections can require efforts to develop rapport with individuals during recruitment. Relationship building is critical in studies where contact is substantial and prolonged, and when researchers are interacting with individuals from marginalized or stigmatized communities, who otherwise may be hesitant to take part in research for fear of disclosure or out of fear that their data will be used against personal or community interests. While ethical oversight governs some aspects of these relationships, they can be complex for researchers in very practical ways. How close is too close? Can the researcher believe what they are being told? Is a key informant representative of a community — or are they an outlier? Addressing such concerns requires researchers to think critically about their role in the research process and the impact this can have on the data.

Data Generation

Qualitative data generation has its own challenges. Researchers must adhere to requirements (e.g., ethical, institutional, professional) governing what they can and cannot do, and studying humans in naturalistic settings can be messy. We’ll talk about governance of qualitative data generation and consider three specific issues that impact how data are gathered: participant recruitment, field site location, and the evolving relationship between study participants and their data.

Governance of Data Generation

In qualitative research, data are never generated without permission or some form of exemption. Besides the informed consent of participants, the most important authorization comes through ethics approval, whereby researchers submit study particulars to an independent ethics body, detailing among other things what data will be gathered, how it will be organized and stored, and how people will make informed decisions on whether to contribute data. For studies involving humans, researchers in Canada (including graduate and undergraduate students) usually need to complete the TCPS 2: CORE-2022 course before applying for ethics approval. This training details researchers’ obligations when collecting and handling data as well as the rights of study participants.

Once a study receives ethics approval, researchers must do what they said they would do. Any modifications (e.g., changes in recruitment methods, expansion of the data collected) require submission and approval of an ethics amendment prior to being implemented. Ethics approval also sits alongside other data-related requirements, such as those maintained by universities (e.g., how long data must be stored). Data generated outside of such precepts is unusable.

Sometimes data collection doesn’t require ethics approval, such as when a qualitative study is conducted as a service evaluation or quality improvement initiative. This approach may be the case in health research that doesn’t involve the public and is of low risk to participants (e.g., clinicians) — for example, in a study of physiotherapists about their experiences treating clients who use wheelchairs. Screening tools may be used to determine whether a full ethics review is required (e.g., ARECCI). Data generation for service evaluations is generally less rigid (e.g., informed consent may not be required) but seldom less rigorous. The data typically look like those from any other qualitative study and are analyzed and reported in much the same way.

Common Challenges

Data generation does not always unfold smoothly. Two challenges, recruitment and field site location, are well known, while a third is still evolving: participants’ relationship with their research data.


Without participants, there can be no qualitative data. How individuals are recruited into a study and tracked becomes reflected in their data. Potential participants must first be identified, a process that can be demanding and slow. Study samples may need to be balanced along factors such as age, sex, or education. Capturing how recruitment unfolds can help make sense of the data. Which details are recorded will vary by study, but they will likely include the following:

  • who has been contacted, how many times, and how they have responded
  • individuals’ professional/personal particulars (e.g., clinical role, preferred pronouns)
  • date(s), time(s), and details of data collection (e.g., researcher name, field location)
  • state of the data (e.g., being transcribed, being anonymized, ready for analysis)
  • restrictions on how data can be used (e.g., if a participant does want to be quoted directly)
  • whether recontact is permitted

All recruitment records must be kept confidential and separate from the data to prevent re-identification. While conventionally not seen as data, recruitment details can be a critical form of metadata. For example, such information can highlight when a key informant entered a study or whether an interview was with a teacher or teacher’s assistant. This level of detail is not always captured in the data.

Field Site Location

A second challenge in data generation is field site location. Qualitative researchers routinely go where participants are, which can present a variety of obstacles. Picture yourself as a researcher in an unusual location (e.g., a remote Arctic community, an urban tent encampment, or a hospital emergency department at 2:00 a.m.) and ask yourself:

  • How will I capture data? (e.g., audio-recorder, pen and paper, photographs)
  • What if my chosen approach fails? (e.g., batteries run out, a pen will not write in cold weather)
  • How can I digitize, secure, and/or back up my data? (e.g., scan fieldnotes, remove files from audio recorders, copy data to the cloud)
  • Can I transfer data for processing? (e.g., sending recordings to a transcriber requires a reliable Internet link)
  • How do I share data with other team members or my academic supervisor?
  • Will data need to be translated? How can I ensure confidentiality and integrity throughout this process?

While quantitative researchers may encounter similar problems, their qualitative counterparts only collect identifiable and potentially sensitive human data, which can make field site challenges particularly difficult to address.

Exercise: Collecting Data Far from Home

Dr. James Cummings is a British sociologist who conducted an ethnographic study of gay men’s experiences in Hainan, China. In a newspaper article, he reflected on the challenges of working with research participants who needed to keep aspects of their lives invisible. Read the article and consider Dr. Cummings’s experience generating and managing research data. What obstacles would he have faced? How might these be similar or different for a researcher studying similar communities in a Canadian setting? How might RDM practice be improved to better support this type of field site?

Cummings, J. (2018). The double lives of gay men in China’s Hainan province. The Conversation.

Participants and Their Data

Data are no longer seen as something over which participants have little control. Some research participants (e.g., those taking part in a professional capacity, such as clinicians or public officials) ask to review and amend their data before giving consent for its use. Because such requests are not common, tracking changes made to the data (e.g., edits to an interview transcript) can be tricky, and there are few best practice guidelines.

This connection between participants and their data is shifting significantly. There has been a growing ethical argument that individuals who take part in research have a right to be informed of any findings arising from their data. Questions have also been raised about making qualitative data available for secondary analysis. How much say should research participants have in how their data are used now or in the future? And what are the ethical, legal and social implications (ELSI) for researchers in accommodating such choices?

One approach is the use of dynamic consent, which allows research participants to remain engaged (if they wish) with their data over longer periods of time and to revisit their original decisions about consent. Where interview transcripts are archived in a repository or data library (e.g., Borealis in Canada, the Qualitative Data Repository at Syracuse University, or the UK Data Service) access is frequently restricted given the identifiability of the material. Dynamic consent allows future researchers to recontact former study participants for permission to reuse their data in new ways. Patients and families involved in some fields of research (e.g., rare diseases) are often interested in maximizing use of their data if such efforts improve the likelihood of a medical breakthrough. While dynamic consent is used primarily for quantitative data, the underlying concept reflects broader shifts in the relationship between research participants and all forms of data.

Qualitative Research Data Meets RDM

This final section returns the discussion to where we started: acknowledging the need for RDM principles, policies, strategies, and practices that speak specifically to qualitative data.

The Processing of Interview Data

Qualitative data do not arrive ready for analysis. They almost always require considerable processing, and each step can create additional versions of the same underlying data, so RDM practice can be almost as iterative as the research it supports.

In the following worked example, the table tracks modifications to a single interview between the time a discussion is recorded to when the data are ready to be analyzed (in this case, coded using NVivo software). Each row represents the creation of a new file.

Table 1. The iterations of a qualitative interview.
Data File File Name Observations Complications


(data as originally captured)






Highly identifiable data that are seldom shared beyond study team



Data may be broken into 2+ files if interview is interrupted or long; two similar recordings may exist if back-up recorder is used



Transcript — original

(version received from transcriber)






Likely to contain multiple transcription errors



May require re-formatting for consistency if different transcribers are used



Transcript — verified

(version after being checked against original recording)






Track changes useful but can result in sub-versions (i.e., tracked, accepted)



Variation more likely if same person does not verify all interviews



Transcript — edited

(version after being changed at participant’s request)






Likely to require notes about edits; usually done using verified version of data



Could force decision about whether to include data if requested edits are significant



Transcript — anonymized

[See more about anonymization in chapter 13, “Sensitive Data.”]






Must decide whether to anonymize interviews individually or collectively



Anonymization keys are highly disclosive and must be kept separate from data



Transcript — NVivo

(version imported into software and edited further)






Copy edits in NVivo are not captured in earlier versions



Version resides within NVivo ecosystem unless downloaded


This table demonstrates how one transcript can exist in multiple versions. Most are transitional, although this worked example is fairly basic. Any number of factors could complicate how the interview data in question are handled. These include the following:

  • participants being interviewed more than once
  • interviews requiring translation during or after transcription
  • transcripts needing to be linked to other data files (e.g., field notes or photographs associated with the same participant)

Exercise: Interview a Qualitative Researcher

This exercise invites you to interview a qualitative researcher about how they manage their data. Start by identifying someone who routinely uses qualitative methods and has a reasonable working knowledge of processing qualitative data (so perhaps not a graduate student). Ask to see their file folder structure (for ethical reasons, you will probably not see specific data). Have the researcher walk you through the types of files they are keeping. Consider how the folders and files have been organized and named. Ask questions about what the researcher has kept, where, and why. Reflect on what you have learned and, if appropriate, propose ways to improve the researcher’s current approach to data management.

Data processing is not always as complex as in the worked example. Qualitative research has been conducted successfully for decades using simpler approaches that still manage to get the job done. Nevertheless, researchers can always improve, particularly as new RDM requirements emerge. Open scholarship demands that, wherever possible, qualitative researchers begin to manage their data in ways compatible not just with research excellence but with an eye to possible sharing and reuse. This transition has implications for two practices still not common among qualitative researchers: metadata and data archiving.

Attaching metadata to qualitative research data can be problematic because qualitative data require contextual detail, but context is disclosive. How do researchers describe data adequately while maintaining confidentiality? For example, metadata indicating that data come from a study of clinicians’ perspectives on providing compression therapy in a community clinic are likely too simple. Recording that the participants were nurses with the same specialist training, that the clinic was at the forefront of developing an innovative approach to compression garments, and that the patients all lived with type 2 diabetes increases the usefulness of the data, though such information heightens the risk of disclosure and re-identification. This is less of an issue for metadata used by individual researchers or within study teams when conducting primary data collection and analysis. But what about metadata to facilitate secondary analysis by external researchers? Metadata standards specific to qualitative data are difficult to find. This isn’t a significant issue in 2023, because qualitative data are seldom placed in repositories, much less made openly available without restrictions.

Many qualitative researchers remain hesitant to archive their data and open it to reuse, and funders don’t demand that researchers do so. Sharing qualitative data also raises issues for recruitment because most researchers tell participants that their data won’t be accessible to anyone outside the study team. Such practices are likely to change as open data principles become embedded in more qualitative-centric disciplines and as funders’ expectations shift. We see this already in the Indigenous data sovereignty movement, which raises fundamental questions concerning metadata and ownership. (For more information see chapter 3, “Indigenous Data Sovereignty.”) Many of the same concerns are being raised by and about other identifiable groups within society. For example, who must be consulted when making RDM-related decisions about data collected from religious or minority ethnic communities? Who gets to decide how those data should be described, archived, and potentially reused? Read chapter 12, “Planning for Open Science Workflows,” for more about open data.

Finally, the most significant challenge illustrated by the worked example is determining which version of the data is definitive. Original recordings are the most accurate and descriptive, but they are highly disclosive. Verified, anonymized transcripts seem the likely choice, but how can researchers ensure identifying details have been removed? Are intermediate versions kept and for how long? If a host institution requires data be stored for five years following completion of a study, does this apply to all versions, or can some be deleted? Such questions can be asked about every data type generated in a qualitative study, making the management of qualitative data remarkably complex.

A Co-Production Model of RDM and Qualitative Data

The worked example raises the question of whether effective management of qualitative data is a realistic expectation of RDM principles, policies, strategies, and practices. The advent of the First Nations Principles of OCAP® — a significant and important framework still being translated into practice and detailed more in chapter 3, “Indigenous Data Sovereignty — suggests that it is. So how might such a goal be achieved? As a rule, qualitative researchers don’t have the information management expertise needed to develop RDM best practice. Conversely, librarians, archivists, and data managers often can’t speak to the complexities of qualitative data and their associated research processes.

In 2020, while working on a study examining co-production in healthcare, I attended yet another RDM training session that didn’t speak to my type of research or my data management concerns. But a light bulb went on when I realized that researchers and data/information specialists have complementary skill sets. If they worked together, the result could be a better system for managing qualitative data.

For co-production to be effective, it would need to be highly collaborative and draw upon the best of both worlds. Our discussion ends with a possible roadmap for how such cooperation might be enacted:

  • Qualitative researchers would be responsible for the following:
    • ensuring all RDM partners understand qualitative data and research processes
    • guaranteeing that data management practices in study teams are consistent and maximize the analytic value of data
    • securing funds to underwrite RDM-associated project costs (e.g., hiring a digital archivist or suitably skilled research associate)
    • advocating for research cultures to support data sharing wherever possible
    • using their professional status and networks to communicate to funders and institutions the challenges and costs inherent in managing qualitative data
  • Data librarians, archivists, and other data specialists would be responsible for the following:
    • applying library and information/data science principles and best practice to the management of qualitative data
    • helping researchers create final datasets (with associated metadata) that meet or exceed the requirements for research excellence
    • using their professional links to stay abreast of and disseminate developments in qualitative RDM practice
  • Together, both groups would be responsible for the following:
    • establishing and advancing effective standards for managing qualitative data
    • developing and delivering RDM training
    • advocating that future RDM principles, policies, strategies, and practices embrace all forms of research data


This is both an exciting and frustrating time to be involved in the management of qualitative research data. Opportunities abound to drive forward new principles, policies, strategies, and practices. At the same time, most qualitative researchers struggle to locate themselves in existing RDM frameworks. Institutions, funders, and RDM practitioners are each grappling with how to address the needs of research communities. While qualitative data are not wholly exceptional (they are, after all, frequently used in conjunction with other types of research data), they remain distinct in many respects. Such complexities highlight the limitations of broad-brush approaches to RDM as well as the need to expand data management to better incorporate all disciplines, fields of research, and methods of inquiry.


Reflective Questions


Reflective Questions

  1. Identify at least three notable characteristics of qualitative data.
  2. In addition to interviews, focus groups, and observations, name two other forms of qualitative data.
  3. What is the purpose of an audit trail?
  4. Name two data-related challenges a qualitative researcher might encounter in a remote field location.
  5. Interview data typically exist in multiple versions between collection and analysis. Identify two such versions.
  6. In a sentence, describe the goal of a co-production model for qualitative RDM.

View Solutions for answers.


Key Takeaways

  • Data generated through qualitative research are complex because they are human based, iterative, context dependent, and highly challenging to de-identify.
  • Such data are difficult to situate within existing RDM principles, policies, strategies, and practices.
  • Effective management of qualitative research data must understand and reflect the research processes at play, including changing expectations around data archiving and reuse, and shifting responsibilities to study participants.
  • Together, researchers and data/information specialists are well positioned to co-produce new approaches to RDM that better meet the needs of qualitative researchers and their data.


Dr. Minion sincerely thanks Dr. Naomi Adelson and Dr. Tamara McCarron for their invaluable feedback on earlier versions of this chapter.

Additional Readings and Resources

Adelson, N., & Mickelson, S. (2022). The Miiyupimatisiiun research data archives project: Putting OCAP® principles into practice. Digital Library Perspectives, 38(4), 508-520.

Budin-Ljøsne, I., Teare, H. J. A., Kaye, J., Beck, S., Bentzen, H. B., Caenazzo, L., Collett, C., D’Abramo, F., Felzmann, H., Finlay, T., Javaid, M.K., Jones, E., Katić, V., Simpson, A., & Mascalzoni, D. (2017). Dynamic consent: A potential solution to some of the challenges of modern biomedical research. BMC Medical Ethics18(1), 1-10.

Chauvette, A., Schick-Makaroff, K., & Molzahn, A. E. (2019). Open data in qualitative research. International Journal of Qualitative Methods, 18, 1-6.

Corti, L. (2019). Archiving qualitative data. In P. Atkinson, S. Delamont, A. Cernat, J.W. Sakshaug, & R.A. Williams (Eds.), SAGE research methods foundations. SAGE Publications.

Cummings, J. (2018). The double lives of gay men in China’s Hainan province. The Conversation.

Diaz, P. (2021). Introduction: Archiving qualitative data in practice: Ethical feedback. Bulletin of Sociological Methodology150(1), 7-27.

DuBois, J. M., Strait, M., & Walsh, H. (2018). Is it time to share qualitative research data? Qualitative Psychology5(3), 380-393.

Mannheimer, S., Pienta, A., Kirilova, D., Elman C., & Wutich, A. (2019). Qualitative data sharing: Data repositories and academic libraries as key partners in addressing challenges. American Behavioral Scientist, 63(5): 643-664.

Moon, K., & Blackman, D. (2014). A guide to understanding social science research for natural scientists. Conservation Biology28(5), 1167-1177.

Pels, P., Boog, I., Florusbosch, J. H., Kripe, Z., Minter, T., Potsma, M., Sleeboom-Faulkner, M., Simpson, B., Dilger, H., Schönhuth, M., Poser, A., Castillo, R. C. A., Lederman, r., & Richards-Rissetoo, H. (2018). Data management in anthropology: The next phase in ethics governance? Social Anthropology26(3), 391-396.

Saunders, B., Kitzinger, J., & Kitzinger, C. (2015). Anonymising interview data: Challenges and compromise in practice. Qualitative Research15(5), 616-632.

Steinhardt, I., Fischer, C., Heimstädt, M., Hirsbrunner, S. D., Ikiz-Akinci, D., Kressin, L., Kretzer, S., Möllekamp, A., Porzelt, M., Rahal, R., Schimmler, S., Wilke, R., & Wünsche, H. (2021). Opening up and sharing data from qualitative research: A primer.

Suter, W. N. (2012). Qualitative data, analysis, and design. Introduction to Educational Research2, 342-386.

Van den Eynden, V., & Chatsiou, K. (2011). Data management for qualitative data using NVivo 9.

Wang, C., & Burris, M. A. (1997). Photovoice: Concept, methodology, and use for participatory needs assessment. Health Education & Behavior24(3), 369-387. 10.1177/109019819702400309


About the author

Dr. Joel T. Minion, PhD MLIS MA BA (Hons) is a qualitative health researcher, librarian, data manager, and educator with experience in Research Data Management (RDM) in both Canada and Europe. He is currently a Research Scientist with the Faculty of Nursing’s Translating Research in Elder Care (TREC) research program at the University of Alberta, where he is responsible for legacy planning and asset protection of TREC’s longitudinal data. Joel was previously Qualitative Research Lead for the University of Calgary’s Health Technology Assessment Unit in the O’Brien Institute for Public Health, and before that a Senior Research Associate with Newcastle University’s Policy, Ethics and Life Sciences (PEALS) Research Centre in the UK. He holds a PhD in health informatics from the University of Sheffield and a MLIS degree from Western University. Since 2010, Joel has been actively involved in managing qualitative research data and ongoing efforts to integrate it into broader RDM frameworks.



Icon for the Creative Commons Attribution-NonCommercial 4.0 International License

Research Data Management in the Canadian Context Copyright © 2023 by Dr. Joel T. Minion is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License, except where otherwise noted.

Digital Object Identifier (DOI)

Share This Book