Module 15 · 3 lessons

Qualitative analysis

Turning a hundred pages of talk into findings somebody else can check. Collecting interviews is the easy part.

Free — no account needed to read

Turning a hundred pages of talk into findings somebody else can check.

Why it matters

Collecting interviews is the easy part. Most qualitative work fails in analysis, where "themes emerged" stands in for a process nobody can inspect.

What it teaches

• Coding: open, axial, selective, and codebooks
• Thematic, content, narrative and discourse analysis
• Memoing and the analytic trail
• Saturation, and how to claim it honestly
• Triangulation and working with more than one coder
• Reflexivity and positionality

When you reach for it

• You have transcripts and do not know what to do with them
• You need to defend how you got from data to findings
• A quote is vivid and you are about to build a theme on it

The common misreading

That analysis means finding good quotes. Quotes are evidence for a finding; the finding has to come from a process applied to everything, including the material that does not fit.

What this module will not tell you

This module will not tell you which analytic tradition your field expects. Thematic analysis in psychology and grounded theory in sociology have different conventions and different reporting standards — read three recent papers in your own venue.

Think first

Nineteen participants describe a process as frustrating. One describes it as fine. What do you do with the one?

Imagine twenty interviews in front of you. Write down your first three moves. Then ask how somebody else would check that you did them fairly.

The lessons

• Coding: from text to categories — A code needs a definition and a boundary. Without both it expands until it means nothing. (9 min)
• Thematic, content, narrative, discourse — A theme makes a claim about meaning. A topic just names what came up. (8 min)
• Saturation, triangulation, reflexivity — Rigour here is a visible trail: memos, negative cases, and an honest account of how categories were built. (9 min)

Practise it: Code one transcript properly

One piece of material, worked through carefully, teaches more than skimming ten.

1. Open-code one transcript closely, staying near what was said.
2. Write a codebook entry for your three most-used codes: name, definition, example, and when not to apply it.
3. Group codes that belong together and name the category.
4. Write a memo: why these categories, what surprised you, what you are unsure about.
5. Find one segment that does not fit, and write what it does to your account.
6. Turn one category into a theme — a sentence that makes a claim somebody could dispute.

What a considered answer tends to contain (not a mark scheme): A considered answer has at least one code that had to be split, a memo recording a decision, and a theme stated as a claim rather than a topic.

Use it on real work: Start the analytic trail on real material

Do this on the material from a project you are actually running.

1. Create your codebook with definitions and boundaries before coding widely.
2. Code at least three pieces of material and memo as you go.
3. Deliberately look for negative cases and record what you did with them.
4. Write your themes as claims, with the evidence behind each.
5. Write your positionality: who you are to this material, and what it gave and cost you.

Reflect

What did the material that did not fit tell you that the pattern could not?