A Beginner’s Guide to Thematic Analysis of Interviews

Thematic analysis looks for meaningful patterns in qualitative material. It is more than highlighting frequently repeated words: a theme should help answer the research question and be grounded in the data. A student project can use a transparent, modest approach without claiming that a few interviews represent every person in a population. Keep a record of how codes changed, retain contrary examples and explain what interpretation you bring to the material.

Familiarise yourself with the material

Read each transcript or permitted set of notes in full before creating a polished list of themes. Note the setting, participants’ own terms and passages that surprise you. Correct transcription errors against the recording if authorised and available. Protect identifying details in your working files. A memorable quotation can attract attention but should not determine the whole analysis before you examine the other accounts.

Code for meaning

A code is a short label describing what a passage contributes. “Unclear deadline instructions” is more analytical than “deadlines” when the question concerns students’ planning. Some passages may carry more than one code. Keep definitions and examples for codes that recur; merge near-duplicates and split codes that hide distinct experiences. Do not force every response into your initial categories if new ideas emerge.

Build and test themes

Look across codes for a larger pattern that answers the question. A theme such as “feedback is valued when it suggests an action” may combine comments about examples, timing and clarity. Test it against all relevant material, including accounts that do not fit. A theme should have boundaries and an explanatory point; a list of interview topics is not automatically a set of themes.

Use quotations responsibly

Choose short excerpts that illuminate your interpretation, then explain why each one matters. Remove or mask identifying details according to the approved plan. Do not let one participant speak for everyone. Show variation where it changes the conclusion. State how many interviews or documents were analysed, but do not turn a small purposive set into a population percentage.

Write the method and reflexive limits

Describe recruitment, data handling, coding steps and how you reviewed themes. Reflect on your role: your questions and expectations may shape what participants shared and what you noticed. Explain what the analysis can illuminate and what it cannot. If you used software, name it as an organisational aid, not as the author of the interpretation. Clear decisions are more valuable than a claim of perfect objectivity.

Worked example

Five interviews may mention that feedback arrived late, yet the underlying pattern could be different for each student: one could not act before the deadline, another had no example to follow, and a third had already moved to another task. A theme called “timing affects usability” should explain these mechanisms and include any interview where late feedback was still useful. Counting the word “late” alone misses the point.

Test a theme against a contrary account

After drafting a theme, search for the case that fits it least well. Suppose you propose “late feedback is ignored”, but one participant used late feedback on a later assignment. The exception may show that transfer to a future task matters more than lateness alone. Refine the theme to explain when timing reduces usefulness and when the advice remains usable. Record the change in a short analysis memo. This is stronger than hiding the exception, and more informative than counting how many people said “late”. In the findings section, pair the refined theme with a quotation that illustrates its central mechanism and a brief mention of the contrasting account. The reader can then see how interpretation emerged from the material.

Checklist before you submit

Read whole accounts before finalising codes.
Define codes and retain contradictory passages.
Give each theme a clear explanatory claim.
Use quotations with context and approved privacy protection.

Questions students ask

How many themes should a small project report?

Enough to answer the question within the available evidence and word limit; avoid both a single vague theme and a long list of topics.

Do I need a coding tool?

No. A table or annotated document can work if decisions are traceable and material is handled securely.