How to Analyse Qualitative Data: Thematic Analysis Step By Step

Written By

Andrew Prescott

✔️ 97% Satisfaction | ⏰ 97% On Time | ⚡ 8+ Hour Delivery

How to Analyse Qualitative Data: Thematic Analysis Step By Step

Thematic analysis is the most commonly used approach to qualitative analysis in UK dissertations. It's accessible to researchers new to qualitative work. It's flexible enough to apply to many different questions. It's genuinely rigorous when done properly. Many students find it more intuitive than other qualitative approaches because it mirrors how humans naturally find patterns in information.

But thematic analysis requires systematic work. Doing it properly is basic different from simply reading your data and deciding what it means.

Before Analysis Begins: Transcription

If you've conducted interviews, you need transcripts. Full word for word transcripts. This transcription is tedious work. It's also key. You can't analyse what you haven't transcribed.

Transcribe accurately. Include pauses. Include repetitions. Include filler words. Your transcript is your data. If you clean it up too much, you're losing information.

For focus groups, transcribe who said what. For interviews, you need less complexity but still full accuracy.

Once transcribed, read through your transcripts while listening to your audio recording. Check for errors. Correct them. Your transcript is your foundation for all subsequent analysis.

Initial Coding: The Most Important Step

Read your first transcript carefully. As you read, identify segments of text that seem meaningful. These might be particularly vivid quotations. They might be ideas that appear multiple times. They might be unexpected statements. They might be answers to your research questions.

Mark these segments. Label them with a code. A code is a short word or phrase that captures the meaning of the segment. These are inductive codes, generated from the data rather than applied from theoretical frameworks.

Your first transcript might generate 40 or 50 codes. That's normal. You're generating as many codes as the data suggests.

Code your second transcript similarly. You'll probably use some of your previous codes but you'll also generate new codes. That's fine. Your coding framework is still developing.

Continue through all your transcripts. By the time you've coded all interviews, you'll have a substantial list of codes. You might have 80 or 100. The exact number varies.

Academic writing at degree level demands a level of critical engagement with sources that goes beyond simply reporting what other researchers have found in their studies. You need to evaluate the quality and relevance of each source you use, considering factors such as the methodological rigour of the study, the date of publication, and the credibility of the journal or publisher involved. When you compare and contrast the findings of different researchers, you demonstrate to your marker that you have a genuine understanding of the debates and controversies within your field of study. Building a habit of critical reading from the early stages of your research will save you considerable time during the writing phase, as you will already have formed considered views on the key texts in your area.

Organising and Reviewing Codes

Once you've coded all your data, look at your complete list of codes. You'll notice patterns. Some codes are similar. Some overlap. Some are really specific examples of more general codes.

Start grouping similar codes together. You might have coded segments with labels like "communication barriers," "difficulty sharing information," and "messaging challenges." These are all variations on a communication theme. Group them.

This is where you start moving from 100 codes towards 10 to 15 more coherent groupings. These groupings are preliminary themes.

Identifying Themes

A theme is more than just a grouping of similar codes. A theme represents a pattern in your data that relates to your research question. A code describes what participants said. A theme interprets what that pattern means.

If you have coded segments about lack of communication channels, difficulty reaching colleagues, and feelings of isolation, the theme might be "communication barriers in remote working environments." The theme brings the codes together around a central idea.

Develop a brief description of each theme. What does this theme mean? What pattern does it represent? How do the codes within it fit together?

You might end up with themes like: structural barriers to adoption, organisational culture challenges, individual resistance, managerial support, training gaps, implementation successes.

Refining Themes: The key Iterative Process

Once you've identified themes, refine them. Go back to your coded data. Do all the segments coded within a theme really belong together? Or do some fit better in a different theme?

Are your themes distinct from each other? Or do some overlap? Can you sharpen the boundaries?

Are there codes that don't fit well into any theme? Do you need an additional theme? Or do those segments not actually belong in your analysis?

Are there themes so small they're not really themes at all? Can they be merged into larger themes?

This refinement is iterative. You're reading the data. You're moving codes between themes. You're creating new themes. You're eliminating weak themes.

This is the most important part of analysis. Rigorous analysis requires this careful, repeated refinement.

Developing Clear Theme Definitions

One of the advantages of starting your writing early is that it gives you the chance to discover gaps in your knowledge while you still have time to fill them through additional reading or further data collection.

Once your themes are relatively stable, develop clear definitions for each. What exactly does this theme encompass? What's included? What's explicitly not included?

A good theme definition is concise but thorough. It should be clear enough that someone else could understand what belongs in this theme.

For example: "Organisational Culture Barriers: perceived organisational resistance to change, perceived lack of leadership support, beliefs that circular economy practices don't fit with existing business models, fear that change will disrupt established processes."

This definition is clear. You could take another dataset and apply this definition to identify segments related to organisational culture barriers.

Your abstract should be written last and should provide a clear and accurate summary of your entire dissertation, including your research question, methods, key findings, and the main conclusion you reached.

Using Data Extracts to Illustrate Themes

For each theme, select data extracts that illustrate it. These are actual quotations from participants that exemplify the theme. Usually you'll have several extracts per theme, showing different facets or different ways participants expressed the theme.

These extracts aren't just pretty examples. They're evidence that the theme actually exists in your data. They show your reader where you got your theme from.

Good extracts are relatively short (two to three sentences) unless a longer quotation is particularly vivid. They should illustrate the point without requiring extensive context.

Checking Your Analysis

Once you've completed thematic analysis, step back. Do the themes make sense? Do they relate to your research question? Do they represent genuine patterns in your data?

Taking the time to understand your data thoroughly before you begin writing about it ensures that your analysis is grounded in what the evidence actually shows rather than what you hoped or expected it would reveal.

If you have access to a second coder, have them independently code a subset of your data. Do their codes align with yours? Where they disagree, why? This agreement check strengthens your credibility.

You might also return to participants if possible, sharing your themes and asking whether they feel you've captured their views accurately. This member checking strengthens analysis credibility.

When you begin writing your dissertation, the most important thing you can do is develop a clear research question that is both specific enough to be answerable and broad enough to generate meaningful findings. A vague or overly ambitious research question will create problems throughout every chapter of your dissertation, making it difficult to maintain a coherent argument and frustrating both you and your markers. The process of refining your research question often involves reviewing the existing literature carefully to understand what has already been studied and where the genuine gaps in knowledge lie. Once you have a focused and well-grounded research question, the rest of your dissertation structure tends to fall into place more naturally, since each chapter can be organised around answering that central question.

Writing About Your Themes

Once you've completed analysis, present your findings using your themes. Typically you'll have one section per theme. You'll briefly describe the theme. You'll include several illustrative quotations. You'll note how many participants mentioned this theme or how frequently it appeared.

This presentation shows your reader both the theme itself and the evidence supporting it.

Common Mistakes

Don't finish coding and then skip the refinement step. Initial codes are rough. They need to be refined into genuine themes.

When you consider the relationship between your data analysis and your overall argument, the connections should feel natural to anyone reading your dissertation from beginning to end, which means every section needs to earn its place within the broader structure you have chosen to present.

Don't create themes that are too large. If your theme encompasses completely different patterns, split it into separate themes.

Don't create themes that are too small. If a theme contains only a few segments from one or two participants, it's probably not a genuine theme.

Don't cherry pick quotations. Include extracts that illustrate the theme even if they complicate your narrative.

Don't claim you've done thematic analysis when you've simply read your data and decided what seems important. Systematic coding and refinement are key.

If you're uncertain how to code your data systematically or how to refine your preliminary codes into rigorous themes, professional services like dissertationhomework.com can guide you through thematic analysis step by step.

===

Need Expert Help With Your Dissertation?

Our UK based experts are ready to assist you with your academic writing needs.

Order Now