Grounded Theory: A Practical Introduction

Learning Goals

Lesson

Check Understanding

What is the difference between open coding and axial coding in grounded theory?

Open coding is the initial pass through the data where you attach labels to segments of text without worrying about how those labels relate to each other. Axial coding comes after you have a set of open codes and involves connecting them: asking what causes this concept, what are the conditions under which it appears, and what consequences it has. Open coding produces a vocabulary; axial coding produces structure.

A research report states: "We used grounded theory. We first collected all 30 interviews and transcribed them, then coded the transcripts over two weeks." What is wrong with this and what should the researchers have done instead?

Grounded theory requires that data collection and analysis happen in parallel. Constant comparison means coding each interview and revising your categories before recruiting or interviewing the next participant. By collecting all 30 interviews before any analysis, the researchers eliminated the mechanism that makes GT iterative: the ability for emerging theory to guide who you talk to next and what you ask. What they describe is thematic analysis on a fixed dataset, not grounded theory. They should have begun coding after the first two or three interviews, revised their sampling and interview guide as new concepts emerged, and stopped collecting data when new interviews added no new open codes.

What does "theoretical saturation" mean, and how would you report evidence for it in a paper?

Theoretical saturation is the point at which new data stops producing new concepts. It means the emerging theory is sufficiently developed that additional interviews are unlikely to change it. To report evidence for it, you might record how many new open codes appeared in each interview and show that the count dropped to zero by interview N. A table with interview number and new-codes-added is more convincing than the phrase "we reached saturation" on its own.

How do you tell whether a paper that claims to use grounded theory actually used grounded theory?

Look for three things. First, check whether data collection and analysis overlapped: GT requires coding to begin before all data is collected. If the methods section describes sequential phases ("we collected, then we coded"), that is a red flag. Second, look for evidence that early findings changed the sampling strategy or interview guide, which is how emerging theory is supposed to drive data collection. Third, look for a description of theoretical saturation with some evidence — not just the claim. Papers that pass all three checks are rare.

Exercises

Core category in Zieris & Prechelt

Read the abstract of Zieris & Prechelt [Zieris2021]. Identify the core category they report — the concept that integrates the rest of their model — and write one sentence explaining how that core category connects to at least two other concepts they identify. Then write one sentence about whether the core category could have been specified in advance as a survey question.

Proposing axial codes

You interview five developers about how they decide which bugs to fix first. Your open codes include "customer pressure," "ease of fix," "affects many users," "personal interest," and "manager says so." Propose two axial codes that group these into broader categories. Write one sentence for each axial code explaining which open codes fall under it and why the grouping makes sense conceptually. Then write one sentence identifying which open code is hardest to place and why.

Diagnosing a GT violation

A paper claims to use grounded theory but collected all data before any coding began. Explain in two sentences why this violates a core GT principle. Then explain in one sentence what consequence that violation has for the resulting theory: specifically, what kind of claim is the paper not entitled to make?

Designing a saturation check

Theoretical saturation is often invoked without evidence. Design a simple record-keeping procedure that a researcher could follow during data collection that would let a reader judge whether saturation was actually reached. Your procedure should specify what you record after each interview, how you would display the evidence in a published paper, and what pattern in the data would convince you that collection should stop.

Comparing methods

Compare thematic analysis (Lesson 20) and grounded theory (this lesson) on three dimensions: starting point, output, and appropriate use case. Write one sentence per dimension per method, for six sentences total. Then write a prompt you could give to an LLM to help you choose between the two methods for a specific research question:

[your prompt here]