Grounded Theory: Definition, Types, and When to Use It

What Is Grounded Theory?

Imagine you are studying how first-year PhD students cope with the pressure of their new life. You have no hypothesis to test and no ready-made theory that quite fits what you are seeing. Instead of forcing your data into someone else’s framework, you collect interviews, look closely at what people actually say, and let a theory grow out of the data itself. That is grounded theory.

Grounded theory is a qualitative research method in which theory is built from data rather than tested against it. Instead of starting with a hypothesis and checking whether it is true, the researcher starts with real-world data — interviews, observations, documents — and develops concepts, categories, and finally a theory that explains what is happening in that setting. The theory is “grounded” because it is rooted in the evidence, not borrowed from a textbook.

A Short History: Where Grounded Theory Came From

Grounded theory was created in the 1960s by two American sociologists, Barney Glaser and Anselm Strauss. Their book, The Discovery of Grounded Theory (1967), was a reaction against a research culture that treated theory-building as the job of a few “great theorists” and expected everyone else to simply test existing ideas. Glaser and Strauss argued the opposite: ordinary researchers, working closely with real data, could generate useful theory themselves — and their method gave qualitative researchers a systematic, defensible way to do it.

Over the decades, the two founders disagreed about how the method should develop, and a third approach later emerged. That is why today we talk about types of grounded theory rather than a single method — more on those below.

The Core Idea: Theory Emerges from Data

Most research you meet as a student follows a deductive path: start with a theory, derive a hypothesis, collect data, and see whether the data support it. Grounded theory flips this around. It is inductive: start with data, look for patterns, and build a theory that explains them.

Three ideas sit at the heart of the method:

• Constant comparison. Every new piece of data is compared with the data and codes you already have. Similarities and differences sharpen your categories as you go.
• Theoretical sampling. You do not fix your whole sample in advance. You collect some data, analyse it, and then decide whom to interview or what to observe next — based on what your emerging theory still needs.
• Theoretical saturation. You stop collecting data when new interviews stop adding anything new to your categories.

The Three Types of Grounded Theory

When students first meet grounded theory, the different versions can be confusing. The three main approaches — classic (Glaserian), Straussian, and constructivist — share the same core logic but differ in philosophy and procedure.

1. Classic (Glaserian) Grounded Theory

Glaser’s version stays closest to the 1967 original. The researcher avoids reading too much of the existing literature before starting, so that preconceived concepts do not “force” the data. Coding begins openly — you label what you see — and a core category gradually emerges that ties everything together. Once that core category is clear, you code selectively around it. Glaser’s message was essentially: trust emergence, and do not impose a structure too early.

2. Straussian Grounded Theory

Strauss, working with Juliet Corbin, developed a more structured version that many students find easier to follow. It adds an intermediate stage called axial coding, in which you systematically relate categories to their subcategories using a “coding paradigm”: conditions, actions/interactions, and consequences. Because its steps are explicit and teachable, the Straussian approach is the version most often found in textbooks and methods courses.

3. Constructivist Grounded Theory

Kathy Charmaz’s constructivist version, set out in Constructing Grounded Theory (2006), takes a different philosophical position. Where Glaser saw theory as something discovered in the data, Charmaz argues that theory is constructed: the researcher and the participants build it together, and the researcher’s own background and questions shape the result. This version is popular in psychology, education, and health research, where lived experience and meaning are central.

Which type should you choose? The one your supervisor knows and your department accepts, and the one whose philosophy matches your research question. State your choice clearly in your methodology chapter and follow it consistently — mixing procedures from different versions without explanation is one of the most common examiner criticisms.

How Grounded Theory Works: The Steps

Whatever version you follow, the working process follows a similar rhythm — you move back and forth between collecting and analysing — in these stages:

1. Start with a research interest, not a hypothesis. Frame a broad question, such as “How do new PhD students experience their first year?”
2. Collect your first data. Usually in-depth interviews, but observation, focus groups, and documents all count. Start small — five or six interviews.
3. Code the data line by line (open coding). Break the text into small pieces and label what is happening in each one. Use short, active labels that stay close to the participants’ words: “seeking reassurance,” “hiding confusion,” “comparing self to others.”
4. Compare constantly and group codes into categories. Which codes belong together? “Seeking reassurance” and “asking seniors for feedback” might form a category called “borrowing confidence.” Write memos — short analytic notes — as you think this through. Memos are where your theory starts to take shape.
5. Sample theoretically. Your early analysis reveals gaps — perhaps lab-based students cope differently from humanities students. Interview them next: each new round of data collection is guided by the emerging theory.
6. Relate the categories (axial coding in the Straussian version). Work out how your categories connect: what conditions give rise to them, what strategies people use, and what consequences follow.
7. Identify a core category (selective coding). One central idea should explain most of what is going on, with everything else relating to it. In our PhD example, the core category might be “proving you belong” — the process by which new students manage the fear of being exposed as not good enough.
8. Stop at saturation. When new interviews keep confirming what you already know and no new categories appear, you have enough data. Write up the theory: the core category, its related categories, the conditions and consequences, and the evidence behind each claim.

A Worked Example: Studying How New PhD Students Cope

Suppose you interview eight first-year PhD students about their experience. You notice students repeatedly describing the same moments: the first supervision meeting, the first seminar where they did not understand the discussion, the first time a classmate seemed to know far more than they did.

Your open codes might include: “feeling behind,” “avoiding questions,” “working late to catch up,” “seeking praise from supervisor,” “comparing self to peers,” and “relief when others admit confusion.” Comparing these, you group them into two categories: hiding struggle and borrowing confidence.

As you interview more students, you notice a pattern: students who find one trusted peer early on move faster from hiding struggle to asking for help openly. You sample theoretically — interviewing students who work alone versus those in labs — and the pattern holds. The core category that emerges is “earning the right to belong”: new PhD students manage self-doubt through a cycle of hiding struggle, borrowing confidence from others, and gradually testing whether they can be open about not knowing. That is a grounded theory: a small, testable explanation built from real data, not imported from a textbook.

Grounded Theory vs Other Qualitative Designs

Students often confuse grounded theory with other qualitative approaches. This table draws the lines:

The key difference is the goal. Phenomenology describes experience; ethnography describes culture; grounded theory explains a process by building a theory. If your question is “What is it like to be a first-year PhD student?” you probably want phenomenology. If it is “How do first-year PhD students move from self-doubt to confidence?” you probably want grounded theory.

When to Use Grounded Theory — and When Not To

Use grounded theory when:

• Little is known about your topic, and no existing theory fits it well.
• Your question is about a process — how something happens or develops over time.
• You want to build a theory rather than test one.
• You can collect data in rounds, returning to the field as your analysis develops.

Choose something else when:

• Your question is “does X cause Y?” — that needs an experimental or correlational design.
• You need to describe what an experience is like, not explain the process behind it (phenomenology fits better).
• You have a single fixed sample and no chance to collect more data in response to your analysis.
• Your timeline is very short — grounded theory’s cycle of collecting, coding, and re-collecting takes time.

Common Mistakes to Avoid

• Forcing the data into a favourite theory. If you decided the answer before you started, you are not doing grounded theory.
• Calling any qualitative study “grounded theory.” Coding interview data and reporting themes is thematic analysis, not grounded theory. Grounded theory must end in a theory — a core category with related categories and stated conditions.
• Skipping theoretical sampling. Interviewing twenty people chosen in advance and coding everything at the end misses the method’s defining feature: letting analysis guide the next round of data collection.
• Stopping before saturation. A theory built on six interviews with no sign of saturation is a sketch, not a grounded theory. Be honest about this limitation if your sample is small.

Frequently Asked Questions

How is grounded theory different from thematic analysis?

Both involve coding qualitative data, but their goals differ. Thematic analysis (for example, Braun and Clarke’s method) aims to identify and report patterns — themes — across the data. Grounded theory goes further: it builds an explanatory theory, with a core category and relationships between categories, through theoretical sampling and constant comparison. Thematic analysis is a method of analysis; grounded theory is a full methodology.

How many participants do I need for a grounded theory study?

There is no magic number — saturation decides. Most grounded theory theses use roughly 15–30 interviews, but the honest answer is: keep going until new data stop adding to your categories. For a student project with a fixed deadline, plan for at least 12–15 interviews and report saturation honestly.

Can I use grounded theory with a literature review?

Yes — it depends on the version you follow. Classic Glaserian grounded theory advises delaying the literature review until the theory emerges. Straussian and constructivist versions treat the literature as a sensitising resource you can use throughout. Either way, the literature should inform your thinking without dictating your categories.

Is grounded theory qualitative or quantitative?

It is a qualitative methodology. It works with words, observations, and documents rather than numbers. (There are rare mixed-methods adaptations, but the method as Glaser and Strauss designed it is qualitative through and through.)

Key Takeaways

• Grounded theory builds theory from data instead of testing a theory against data.
• Its three main types — classic (Glaserian), Straussian, and constructivist (Charmaz) — share the same logic but differ in philosophy and procedure.
• Constant comparison, theoretical sampling, and theoretical saturation are the engine of the method.
• Use it when little is known about a process and you want to explain how it works — not when you need to test a hypothesis or describe an experience.
• A real grounded theory study ends in a theory: a core category with related categories, conditions, and consequences, all supported by evidence.

If your research question asks how a process unfolds and no existing theory quite explains it, grounded theory gives you a disciplined way to build the answer from the ground up — one interview, one code, and one memo at a time.