By Hardi Hindocha, Growth Marketing Lead
A UX researcher at a healthcare SaaS company was asked to understand why patients were not completing the onboarding flow for a new medication tracking app. She ran twelve in-depth interviews, coded the transcripts, produced a theme report, and presented to product leadership. The themes were clear: patients felt confused by the terminology, frustrated by the number of steps, and uncertain whether the app was secure.
Leadership nodded and asked the one question she had not planned for. "But why do different types of patients experience this so differently? The elderly users seem confused, the younger ones seem frustrated, the chronically ill ones seem distrustful: are these the same problem or three different problems?"
She had designed her study as a set of undifferentiated interviews. She had not chosen a research design that would let her distinguish between population segments or understand whether the variation was meaningful. She had collected good data. She had designed a study that could not answer the question her stakeholders actually needed answered.
That is what poor qualitative research design looks like in practice. Not bad fieldwork. A mismatch between the research question, the design, and what the resulting data can legitimately support.
What Qualitative Research Design Actually Means, And What Most Guides Miss?
Qualitative research design is the set of decisions about how you will structure a study to answer a research question that cannot be resolved by counting or measuring. It is the blueprint that connects your question to your data collection approach, your participant selection, and your analysis method.
Most definitions stop there. Here is what most guides leave out.
A qualitative research design is not a data collection method. Interviews, focus groups, and observation are data collection methods. Phenomenology, grounded theory, and ethnography are research designs. The design comes first and determines which methods are appropriate. A researcher who picks interviews before choosing a design has started in the wrong place, it is like deciding to use a hammer before knowing what you are building.
Design also determines what your findings can legitimately claim. A phenomenological study produces findings about the shared essence of a lived experience. A grounded theory study produces a theoretical model that explains a social process. A case study produces a rich, bounded account of one specific context. These are not the same kind of knowledge, and they are not interchangeable. Using phenomenological data collection with a grounded theory analysis framework produces findings that satisfy neither standard.
The design choice also determines your sample size, your recruitment logic, and your saturation criteria. Ethnographic research with five participants across two sites is appropriate. Grounded theory with five participants is usually not because the sample size is too small to build theory from.
Types Of Qualitative Research Design: The Five Core Approaches
Phenomenology: understanding the lived experience
Phenomenological research design asks what a particular experience is like for the people who have lived it. The goal is to identify the essential, shared structure of an experience across participants, what Husserl called the invariant features.
It is the right design when your research question is "what is it like to go through X?" The emphasis is on subjective experience, meaning, and the perspective of the person inside the phenomenon.
What it is not suited for: building theory about social processes, comparing contexts, or producing a detailed account of one specific case. If your question is more "why does this happen differently in different places?" or "what theory can explain this pattern?", phenomenology is the wrong framework.
What the Phenomenology analysis looks like: Researchers identify significant statements from transcripts, cluster them into meaning units, and derive themes that capture the shared structure of the experience. The output is a description of the essence of the experience, not a list of findings about what different people said.
A real example Phenomenology design and analysis: A researcher studying how intensive care unit nurses experience moral distress would use phenomenological design. She wants to understand what moral distress feels like, not build a theory of why it varies across hospitals, not describe one unit in detail. She recruits nurses with experience of moral distress, interviews them in depth, and builds a structural description of the shared experience.
Grounded theory: building theory from the ground up
Grounded theory research design produces a theoretical model that explains a social process or phenomenon. It was developed by Glaser and Strauss to generate theory from data rather than testing pre-existing theory against data.
The distinctive feature is theoretical sampling. In grounded theory, you do not decide your final sample at the start. You collect data, begin analysis, identify gaps in your emerging theory, and recruit additional participants specifically to fill those gaps. You continue until theoretical saturation, the point where new data adds nothing to your theoretical model.
It is the right design when your question is "what is happening here, and what theory can explain it?" It is well suited for understanding processes, particularly social or organisational processes where the dynamics between actors matter.
What it is not suited for: studying the subjective quality of a personal experience, producing a detailed account of a specific bounded case, or understanding a narrative arc in individual lives.
What the Grounded theory analysis looks like: Open coding first (breaking data into small units and labelling them), then axial coding (identifying relationships between categories), then selective coding (integrating categories around a core category that explains the process). The output is a theoretical model with explanatory power, not just a set of themes.
A real example of Grounded theory design and analysis: A researcher studying how remote workers negotiate professional boundaries between work and home life would use grounded theory design. She wants to understand the social process at work and how norms are established, challenged, renegotiated, not just what the experience feels like. She begins with a set of interviews, identifies emerging patterns, recruits additional participants who represent gaps in the emerging model (people who have resolved the tension successfully, people who have not), and builds a theory of the boundary-negotiation process.
Ethnography: understanding a culture or community from the inside
Ethnographic research design seeks to understand the shared beliefs, values, practices, and meanings of a cultural or social group. The defining feature is prolonged engagement where the researcher spends extended time in the field, typically combining observation, informal conversation, and formal interviews.
It is the right design when your question is "how does this group understand and organise their world?" It requires time investment that shorter qualitative designs do not: weeks or months, not days.
What it is not suited for: studying a specific individual's experience, building a general theory, or producing bounded case analysis. Ethnography produces what Clifford Geertz called "thick description" - which is richly contextualised accounts of cultural practice.
What the ethnography analysis looks like: Field notes are coded alongside interview data. Analysis focuses on cultural themes and looks at recurring patterns of meaning across observations and conversations. The output is a cultural portrait of the group.
A real example of ethnography design and analysis: A researcher studying how agile development teams at technology companies construct shared understanding of quality would use ethnographic design. She embeds with two or three teams over several months, attending standups, retrospectives, and informal conversations. Her findings describe the cultural logic through which these teams define and maintain quality standards.
Case study: deep understanding of a bounded situation
Case study research design produces a rich, detailed account of a specific, bounded case like a person, an organisation, a programme, an event, or a community. The emphasis is on depth within the boundaries of that case, rather than generalisability across cases.
Robert Yin's definition is useful: a case study investigates a contemporary phenomenon in depth, within its real-world context, especially when the boundaries between phenomenon and context are not clearly evident.
Single-case designs are appropriate when the case is critical, extreme, revelatory, or longitudinal. Multiple-case designs allow for comparison across cases and strengthen the basis for analytical generalisation.
It is the right design when your question is "what happened in this specific situation, and what can we learn from it?" or "how does this phenomenon play out across these three specific contexts?"
What it is not suited for: building general theory from scratch (grounded theory does that), capturing the subjective essence of an experience (phenomenology does that), or producing a cultural portrait of a community (ethnography does that).
What the case study analysis looks like: Cross-case pattern matching, explanation building, and the search for rival explanations. The output is a case narrative with analytical conclusions, findings that are specific to the cases studied but analytically transferable to similar situations.
A real example of case study design and analysis: A researcher studying how a single NHS trust implemented a new patient record system, including the decisions made, the staff responses, and the outcomes, would use a single case study design. She is not trying to build a general theory of NHS system implementation. She is trying to understand what happened in this specific context and what it reveals about the challenges of health IT implementation.
Narrative inquiry: understanding experience through story
Narrative inquiry treats stories as the primary unit of analysis. It holds that people make meaning of their experiences by constructing narratives about them, and that understanding how someone tells their story reveals as much as the events the story describes.
It is the right design when your question is "how does this person make sense of their experience over time?" or "what does the arc of this person's life or career reveal about a broader social or cultural phenomenon?"
What it is not suited for: understanding a cultural group, building theory, or studying a bounded situation. Narrative inquiry is the most individually focused of the core qualitative designs.
What the narrative inquiry analysis looks like: Researchers read transcripts for plot, character, turning points, and narrative structure. Analysis attends to how the story is told, not just what it says, the sequence the narrator chooses, the explanations they construct, what is foregrounded and what is left out.
A real example of narrative inquiry design and analysis: A researcher studying how founders of failed startups make sense of their experience would use narrative inquiry. She is interested in the stories founders tell, how they construct meaning from failure, what narrative frames they use and not in building a theory of why startups fail.
Participatory action research: designing with, not just for
Participatory action research (PAR) is a design approach where the researcher and the participants are co-researchers. The community or group being studied participates actively in defining the research questions, collecting data, interpreting findings, and determining what action to take. It is most used in community-based, educational, and social justice research contexts.
It is the right design when the research goal is not just to understand a situation but to change it, and when the change must be led by the people most affected. It is less commonly used in commercial research contexts but increasingly relevant in patient-centred healthcare research and co-design projects.
Examples Of Qualitative Research Design In Applied Research Contexts
Examples of qualitative research design in UX research
UX researchers use qualitative research design most often in one of three modes:
Discovery research, where the question is "what is the user's experience of this domain and what problems exist?", typically uses ethnographic or contextual inquiry approaches.
Evaluative research, where the question is "how do users experience this specific product or feature?", often uses phenomenological framing without naming it as such.
Generative research that produces a theory of user motivation or decision-making benefits from grounded theory principles, including theoretical sampling and category saturation.
A UX researcher at a fintech company studying how first-time investors make portfolio decisions would make different design choices depending on the research question.
If the question is "what is it like to make your first investment decision?" - the design is phenomenological.
If the question is "what theory can explain how novice investors develop confidence over time?" - the design is grounded theory.
If the question is "what happened in the journey of this particular user who started, stopped, and restarted their investing behaviour over six months?" - the design is a case study.
The research question drives the design. Not the other way around.
Examples of qualitative research design in market research
Commercial market research uses qualitative designs less formally than academic research, but the underlying logic is the same. Brand perception studies that explore how consumers hold a brand in their minds draw on phenomenological principles. Innovation research that tracks how a new product category develops consumer meaning over time benefits from grounded theory principles. Segmentation research that produces rich portraits of consumer types draws on ethnographic methods and case study logic.
A CPG brand studying how consumers decide which cooking oils to buy from a supermarket shelf would apply phenomenological design to understand the full experiential dimension of the moment of choice, including the felt sense of quality, the cultural associations triggered by different oils, and the tacit knowledge being applied. A list of purchase criteria would not capture that.
Examples of qualitative research design in healthcare and social research
Healthcare researchers studying patient experience draw on all five core designs depending on the question. Studies of what it is like to live with chronic pain use phenomenological design. Studies of how patients and clinicians negotiate treatment decisions together use grounded theory. Studies of care delivery within a specific ward or unit use case study or ethnographic design. Studies of how a patient's illness shapes the story they tell about their life use narrative inquiry.
The design type directly determines what kind of evidence you can produce and how strong a claim you can make.
Qualitative Research Design Methods: What Data Collection Looks Like Under Each Design?
Each qualitative research design is associated with particular data collection methods, but the relationship is not exclusive. Here is how the alignment typically works:
Phenomenology relies primarily on in-depth, semi-structured interviews, often conducted in two phases to allow for follow-up. The goal is to elicit rich, reflective accounts of the experience. Journaling and written reflection are sometimes used alongside interviews. Observation is not usually primary, the emphasis is on the inner, subjective dimension, which observation alone cannot access.
Grounded theory uses semi-structured interviews as the primary data collection method, often supplemented by observation and document analysis. The critical methodological feature is concurrent data collection and analysis, you begin analysing your first interviews before you have completed your last ones, because the analysis determines who you interview next.
Ethnography is defined by observation, specifically participant observation where the researcher is present in the community or setting over an extended period. Informal conversations, formal interviews, and document analysis supplement the observation data. Field notes are the primary data artefact.
Case study is the most methodologically flexible design. It draws on interviews, observation, documents, artefacts, and archival data as evidence. The flexibility is a feature - the richness of a case study comes from triangulating across multiple evidence sources. Yin's principle is that case studies should use at least three evidence sources.
Narrative inquiry uses biographical and life history interviews as the primary method, often with minimal structure to allow the participant's own narrative logic to emerge. Photo elicitation, asking participants to bring photographs that represent significant moments, is a useful adjunct method in narrative designs.
How To Choose A Qualitative Research Design? A Decision Framework
The question most guides do not answer is how to actually decide. Here is a practical decision framework built around the research question.
Start with what you want to produce. Not what you want to learn in the abstract, but what the output of the study will look like. Is it -
A description of an experience,
A theoretical model,
A rich account of one situation,
A life story, or
An action plan co-created with a community.
Because each of the output maps to a design.
To decide precisely, ask four diagnostic questions:
Am I studying the inner, subjective quality of an experience that people have had? If yes, phenomenology is the primary candidate.
Am I trying to build or develop a theoretical model that explains a social process? If yes, grounded theory is the primary candidate.
Am I trying to understand a cultural group or community from the inside, and do I have time for extended fieldwork? If yes, ethnography is the primary candidate.
Am I trying to produce a deep, rich account of one specific bounded situation, organisation, event, or small group of comparable situations? If yes, case study design is the primary candidate.
If the answer is none of the above, and the question is about how a person makes sense of their experience over time, narrative inquiry is the candidate.
Check the match between your question, your access, and your timeline. Ethnography requires extended fieldwork. Grounded theory requires theoretical sampling, which means you cannot finalise your sample at the start. Case study requires access to multiple evidence sources within the case. Phenomenology requires participants who can reflect on and articulate their experience. If there is a mismatch between the ideal design and your constraints, acknowledge it explicitly and choose the design that is closest to appropriate given your actual constraints.
A one-line decision test: Write out your research question in one sentence. The first word usually tells you the design. "What is it like to...?" is phenomenological. "How does [social process] develop or evolve?" points to grounded theory. "What happened when...?" points to case study. "How does [cultural group] understand and practise...?" points to ethnography. "How does [person] make sense of their [experience] over time?" points to narrative inquiry.
Common Mistakes In Qualitative Research Design That Undermine Findings
Here are some common mistakes that you need to avoid while planning your qualitative research as they may not bring you the best results and insights.
Choosing a method before choosing a design
This is the most common mistake. A researcher decides to run interviews, then works backwards to justify a design. The problem is that the choice of interviews shapes the analysis: interview data is well suited to phenomenological and grounded theory analysis but less well suited to producing the thick contextual description that ethnographic design requires. If you decide on interviews first, you may discover mid-analysis that your data cannot support the claims you want to make.
The fix: Write the research question, apply the diagnostic framework above, choose the design, then choose the methods the design requires.
Treating qualitative research design as a generic category
Researchers who write "this study uses a qualitative research design" without specifying the design type are doing two things wrong. They are signalling to reviewers that they have not thought through the epistemological and methodological implications of their choices. And they are giving themselves no analytical framework to work from when the data is collected.
The fix: Name the design type explicitly in the research proposal and methodology section. Justify the choice by connecting the design to the research question. Cite the methodological literature that defines the design you are using.
Underestimating what each design requires
Grounded theory requires iterative sampling until theoretical saturation and not a fixed sample decided in advance.
Ethnography requires prolonged fieldwork and not a few hours of observation.
Case study requires multiple evidence sources, and not just interviews with one or two informants within the case.
When researchers use the label of a design without applying its defining features, they lose the analytical advantages the design provides and open themselves to methodological criticism.
The fix: Read the foundational literature for the design you are using. For phenomenology, Moustakas (1994) is the standard reference. For grounded theory, Strauss and Corbin (1998) and the subsequent Charmaz (2006) constructivist grounded theory text. For case study, Yin (2018). For ethnography, Spradley (1979) and more recently Pink (2015). For narrative inquiry, Clandinin and Connelly (2000).
Designing the analysis separately from the design
The analysis approach must be consistent with the research design. Thematic analysis applied to phenomenological data is legitimate but requires a specific form and interpretative phenomenological analysis (IPA) or Moustakas's transcendental phenomenological analysis. Standard thematic analysis applied to grounded theory data misses the iterative theoretical development that grounded theory requires.
Knowing which analysis approach fits your design before you begin data collection prevents the post-hoc analysis mismatch that produces methodologically incoherent findings.
The fix: Before data collection, specify not just the design but the specific analysis method you will use. Name it, cite it, and confirm it is epistemologically compatible with the design you have chosen.
How AI Fits Into Qualitative Research Design, And Where Researcher Judgment Still Leads?
The design choice itself remains entirely within the researcher's judgment. No AI tool can determine whether your research question requires phenomenological or grounded theory design. That decision requires interpretive judgment about the nature of the knowledge you are trying to produce.
What AI analysis tools have changed is the analysis stage that follows data collection, and that change is significant.
Manual thematic analysis of twenty interviews used to take days. A researcher would read and re-read transcripts, generate initial codes, develop themes, review themes against the data, name and define themes, and produce a final report. The intellectual work was substantial, but a large proportion of the time was consumed by mechanical organisation of material.
Tools like DoReveal work differently. A researcher uploads interview recordings or transcripts, and DoReveal processes the full content of each interview using a conversation understanding engine, producing themes, participant quotes anchored to each theme, JTBD breakdowns, emotional laddering, and journey maps.
In a documented benchmark, fourteen interviews were analysed in thirty-four seconds. In a comparative test against three other tools, using a real COVID-19 healthcare study, DoReveal ranked first across coverage, analytical depth, voice of participant, usefulness, and novel insights.
The parts that still require the researcher's judgment are exactly the parts that the design choice determines. Deciding whether the emerging themes represent a genuinely shared structure of experience (phenomenological conclusion) or the beginning of an explanatory theory (grounded theory conclusion) requires the researcher's interpretive framework. Deciding which findings are worth elevating to the top line of a client report requires knowledge of what the client needs to act on. Deciding whether an AI-identified emotional pattern is a genuine signal or a transcript artefact requires the qualitative researcher's ear.
AI analysis is a powerful tool inside a well-designed study. It does not substitute for the design decision that determines what the data can mean.
For more on how AI-assisted analysis works in practice, see our guide to qualitative data analysis tools and AI thematic analysis.
Running a qualitative study and need your interview data analysed by design type, themes, JTBD, emotional laddering, without spending three days in a spreadsheet? Try DoReveal on three interviews at no cost. No credit card. No demo call. See what it finds in your data.
Start free or
!Frequently Asked Questions About Qualitative Research Design
What is qualitative research design?
Qualitative research design is the blueprint that connects a research question to the strategy for studying it. It determines how the study will be structured, who will be recruited, what data will be collected, and how that data will be analysed. The five core qualitative research designs are phenomenology (studying lived experience), grounded theory (building theory from data), ethnography (understanding a cultural group), case study (producing a rich account of a bounded situation), and narrative inquiry (understanding experience through story). The design must be chosen before the data collection method, because the design determines which methods are appropriate.
What are the main types of qualitative research designs?
The five most widely used qualitative research design types are phenomenology, grounded theory, ethnography, case study, and narrative inquiry. Participatory action research (PAR) is a sixth design used primarily in community-based and social justice research contexts. Each design type is associated with a different kind of research question, a different data collection approach, and a different analysis framework. They are not interchangeable, using the methods of one design with the analytical framework of another produces methodologically incoherent findings.
What is an example of qualitative research design?
A hospital studying how nurses experience moral distress would use phenomenological qualitative research design, the question is about the shared structure of a lived experience. A researcher studying how a specific NHS trust implemented a new electronic health record system would use case study design, the question is about what happened in a bounded, specific context.
A researcher studying how professional communities of practice develop shared norms would use grounded theory design, the question is about a social process that requires a theoretical explanation. Each example uses a different design because each research question asks for a fundamentally different kind of knowledge.
What is the difference between qualitative research design and qualitative research methods?
Research design is the overall strategy for studying a question. Research methods are the specific data collection techniques used within that strategy. Interviews, focus groups, observation, and document analysis are all research methods. Phenomenology, grounded theory, ethnography, and case study are all research designs. The design comes first and determines which methods are appropriate. A common mistake is choosing methods before choosing a design, which can produce data that does not fit the analysis framework the researcher needs to answer their question.
How do I choose between qualitative research design types?
The choice follows from the research question. If you want to understand the subjective quality of a lived experience, use phenomenology. If you want to build a theory that explains a social process, use grounded theory. If you want to understand how a cultural group or community sees and organises their world, use ethnography. If you want to produce a rich, detailed account of a specific bounded situation, use case study. If you want to understand how a person makes sense of their experience over time through narrative, use narrative inquiry. Write your research question in one sentence and ask what kind of knowledge the question requires, that usually identifies the right design.
What is basic qualitative research design?
The phrase "basic qualitative research design" usually refers to a study that uses qualitative data collection methods, such as interviews or focus groups, without committing to a specific methodological tradition. While this approach is common in applied and commercial research contexts where the formal rigour of academic qualitative methodology is not required, it has a significant disadvantage: without a named design framework, the researcher has no established analysis procedure to follow and no methodological literature to draw on when justifying their analytical decisions. Even in applied research, naming and following one of the five core designs produces more defensible findings.
How does qualitative research design affect analysis?
The design determines the analysis framework. Phenomenological design requires analysis methods that identify the shared structural features of an experience, interpretative phenomenological analysis or Moustakas's transcendental phenomenological analysis.
Grounded theory design requires iterative coding through open, axial, and selective coding stages, building toward a theoretical model.
Case study design requires cross-source triangulation, pattern matching, and explanation building. Ethnographic design requires cultural theme analysis that integrates field notes, interview data, and artefacts.
Using the wrong analysis framework for a given design produces findings that do not meet the standards of any recognised qualitative methodology.
Can AI tools help with qualitative research design?
AI tools do not help with the design choice itself, that requires the researcher's interpretive judgment about the kind of knowledge the research question requires.
Where AI tools contribute is in the analysis stage that follows data collection. Once a qualitative study has been designed and interviews or observations completed, AI analysis platforms can process transcripts to surface themes, participant quotes, emotional patterns, and structured frameworks like JTBD and emotional laddering significantly faster than manual analysis. DoReveal, for example, applies multiple analytical frameworks natively and in a documented benchmark analysed fourteen interviews in thirty-four seconds. The researcher's role remains central in interpreting what the AI surfaces and deciding how findings translate into conclusions that are consistent with the chosen design.