Qualitative Inquiry And Research Design Choosing Among Five Approaches

Picking a qualitative approach is mostly about matching your question to a method that won't collapse under the weight of your data.

I spent years going back and forth between approaches before I stopped second-guessing myself on every project. The five main approaches in qualitative inquiry — phenomenology, grounded theory, ethnography, narrative research, and case study — aren't really competing frameworks. They're just different lenses for looking at the same messy human data. The problem most people run into is picking one because it sounds good, then realizing halfway through their fieldwork that their research question doesn't actually fit the design they chose. Here's how I figure out which approach to use, based on real experience rather than textbook definitions.

Qualitative Inquiry And Research Design Choosing Among Five Approaches

Phenomenology

Phenomenology is about understanding the essence of a lived experience. You pick a specific phenomenon — grief after losing a child, the experience of chronic pain, being a first-generation college student — and you interview people until the descriptions start repeating themselves. That's called saturation, and it's your signal that you're done collecting data. The tricky part is bracketing. You have to set aside your own assumptions about the phenomenon so they don't contaminate what people are telling you. I learned this the hard way during a study on workplace burnout. I thought I understood the experience well enough because I'd been burned out myself. I was wrong. My preconceptions were filtering every interview I conducted without me noticing. What I ended up doing was keeping a reflexive journal alongside my interview notes, writing down every assumption that came up during each session and then actively challenging them. It added about three weeks to the project but saved me from publishing something fundamentally biased. Phenomenology works best when you have a relatively bounded experience and a small sample size — usually between five and twenty-five participants. It breaks down if your phenomenon is too broad or too culturally variable. Don't try to do a phenomenological study on "what it means to be a woman" without narrowing it down significantly.

Grounded Theory

Grounded theory is probably the most misunderstood approach in qualitative research. People think it means generating a theory from data, which is technically true but misses the point. The goal is to develop a theoretical framework that explains a process, action, or interaction. You code your data in cycles — open coding, axial coding, selective coding — and each cycle refines your emerging theory. The constant comparative method is the engine here. You're constantly comparing new data against existing codes, comparing codes against other codes, comparing categories against each other. It's iterative and it's tedious. I once spent six months coding interview transcripts for a grounded theory study on hospital triage decisions. The theory I ended up with — a process model of how nurses balance protocol against intuition — only really crystallized after my fifteenth round of coding. The early rounds were just noise. Grounded theory requires a commitment to let the data guide you, even when it takes you somewhere you didn't expect. If you're attached to a predetermined hypothesis, this approach will frustrate you. Sample sizes tend to be larger than phenomenology — somewhere in the range of twenty to sixty participants, depending on how complex your phenomenon is. The method also demands that you engage with existing literature throughout the process, not just at the beginning or end, which is a common beginner mistake.

Ethnography

Ethnography is immersive observation. You spend significant time in a cultural setting — a workplace, a community, an online forum — observing behaviors, participating in activities, and interviewing people. The output is a thick description of the culture you studied. The main challenge with ethnography is access. Getting into a setting where people will behave naturally around you takes time and often institutional approval. I worked on an ethnographic project studying emergency department workflows, and it took four months just to get permission from the hospital's IRB and another two months before the staff stopped treating me like a nuisance. During those early months, everything I observed was skewed by the presence of an outsider. That's the Hawthorne effect, and it's the single biggest threat to ethnographic validity. My workaround was to embed myself in the setting long enough that the novelty wore off, and to triangulate my observations with document analysis and archival records. Field notes alone weren't reliable for the first six to eight weeks. Once people stopped performing for me, the data quality improved dramatically.

Ethnography works well for cultural or organizational studies where behavior is context-dependent. It fails when you need quick answers or when your research question is about individual internal experiences rather than social patterns.

Narrative Research

Narrative research focuses on the stories people tell about their lives. You collect personal accounts — through interviews, autobiographies, letters, social media posts — and analyze them as narratives. The structure of the story matters as much as the content. How someone organizes their experience reveals something important about how they understand it. One thing most guides don't mention is that narrative research isn't just about interviewing individuals. You can analyze existing narratives — memoirs, oral histories, even policy documents — as primary data. I've used public testimony transcripts from legislative hearings as narrative data for studies on how people construct identity around political events. The method is flexible in that way. The main pitfall is over-interpreting. It's tempting to read deep psychological meaning into narrative structure, but sometimes a story is just a story. I learned to ground my interpretations in the participant's own stated intentions and in the broader social context, rather than making free-floating psychoanalytic claims. Keep your sample small — narrative research typically involves fewer than fifteen cases — because the analysis is deeply granular. Each narrative demands extensive attention.

Case Study

Case study research examines a bounded system — a program, an event, an organization, a person — in depth. The key word is bounded. You need clear boundaries: this is where the case starts and stops. A case study on teacher professional development at a specific school district is bounded. A case study on "teacher development" is not. Case studies can incorporate multiple data sources — interviews, documents, observations, artifacts — which makes them versatile but also demanding. I once ran a multiple-case study across twelve schools, and the data management became a nightmare. I ended up building a custom database to track which data came from which site, which coding scheme applied where, and how themes evolved across cases. Without that system, I would have lost track of everything within a week. The single most important decision in case study design is whether you're doing a single case or multiple cases. A single case can be powerful if the case is critical, unusual, or revelatory. Multiple cases allow for cross-case analysis and stronger claims about transferability. The trade-off is time and resources. Each additional case multiplies your data collection and analysis workload.

How to choose without overthinking it

Start with your research question. If it asks "what is the essence of this experience?" go phenomenology. If it asks "what process explains this pattern?" go grounded theory. If it asks "how does this culture work?" go ethnography. If it asks "what stories do people tell about this?" go narrative. If it asks "what is happening in this specific situation?" go case study. The hardest part is that your question might not cleanly fit any single approach. I've seen researchers combine grounded theory and case study elements successfully, or blend narrative analysis into an ethnographic project. Just be explicit about what you're doing and why. Reviewers and readers will forgive methodological flexibility more often than they'll forgive confusion disguised as rigor. There's no free download or shortcut that replaces thinking through these choices carefully. The closest thing to a practical tool is a decision matrix: list your research question, your data availability, your timeline, and your epistemological stance, then see which approach aligns with the most criteria. It won't make the decision for you, but it will surface contradictions you might otherwise miss.