Working with interviews for a thesis involves preparing a transcript, analysing what participants say and reporting findings linked to the research questions. A transcript is material for analysis. A results chapter takes shape when you explain what the accounts show and what supports your interpretation.
Do you have several recordings, dozens of pages of text and no idea where to start? You do not need to buy specialist software straight away. First, decide how you will work with the material. The example below shows how to move from an extract to a code and a cautious conclusion, without turning the interviews into a collection of random quotations.
Interviews are a way to collect data. How will you analyse them?
In your thesis, identify both how you conducted the interviews and the method you used to analyse them. Thematic analysis, narrative analysis, qualitative content analysis and grounded theory methodology are not interchangeable terms for the same activity.
In this guide, we show how to organise work with codes and themes using a simple example. This is not a complete protocol for every qualitative method. If you are analysing how people construct narratives of their own lives, dividing their accounts into themes may not be enough. When you adopt a particular method, apply its assumptions consistently and describe your own analytical decisions.
In the study by Mandes and Karlińska (2024), qualitative coding involved reading documents, interpreting extracts and checking their meaning across the whole dataset. The authors also show how this work differs from automated topic modelling. This is a useful distinction: a list of frequently occurring words is not, in itself, an interpretation of interviews.
1. Prepare a transcript you can work with
Give participants stable identifiers, such as R01 and R02. Separate their responses from the interviewer's questions. Add paragraph numbers or timestamps so that you can later locate an extract in the recording. Keep the source version separate from your working file.
The level of detail in your transcript depends on the aim of the study. If you are interested in the content of people's experiences, you may use a different convention from one suited to analysing interaction, pauses and overlapping speech. Set the rules before processing all the material. Do not change the meaning of an account to make it sound more academic.
- Check an automated transcript against the recording, especially negations, numbers and specialist terms.
- Mark unclear passages according to your chosen convention rather than filling in what you think was said.
- Replace identifying information with neutral labels, such as [institution name]. Removing a first name alone may not be enough.
- Retain important information about the context of a question so that you do not interpret the response in isolation from the conversation.
When sharing material, take account of the agreed rules on confidentiality and use of the recording. Do not upload a full interview to another tool just because it offers free transcription. First, check its terms of use and what material you can entrust to it.
2. Read everything before choosing quotations
After the first reading, write a short memo about each interview: main topics, uncertainties, tensions and questions that require you to return to the material. This is a working memo, not a finished finding. Distinguish the participant's words from your own suppositions.
Three questions can help: what does the participant say about the issue being studied, in what situation do they place their experience, and what in their account does not fit your initial expectations? This helps you avoid limiting the analysis to sentences that confirm your first interpretation.
3. Assign codes to meaningful extracts
A code is a working label for an extract that is relevant to the analysis. It may refer to an action, experience, assessment or situational condition. The extract you code should be long enough to preserve its meaning. There is no need to automatically assign a separate code to every sentence.
From an account to a code: an original example
A fictional statement from R03: “At first, I did not know who to contact. It took a friend to show me where to find the mentor’s office hours.”
Possible working codes: unclear contact route; peer support; finding information about office hours.
Analytical memo: the participant describes obtaining information through informal help. The extract does not prove that formal information was unavailable. The context and other accounts need to be checked.
Keep a simple code list: name, working meaning, example and limits of application. Record changes. If two codes begin to mean the same thing, you can merge them, but revisit the earlier extracts when you do. The level of detail in this documentation and the way a team agrees on codes should fit the chosen method.
4. Examine the connections between codes
An analytical theme should help answer a research question. The heading of a question in the interview guide, such as “Assessment of placements”, may simply name a topic of conversation. A more precise analytical finding might be “Access to help depends on informal contacts”. However, you need to check this theme against the whole dataset rather than derive it from one convenient quotation.
Compare extracts within each interview and across interviews. Look for contrasting cases too. If R03 describes help from a friend while R07 received clear instructions from his mentor, the difference may prompt a question about organisational conditions. You do not need to remove it to make the chapter more uniform.
Do not equate the number of coded extracts with the number of people. One person may return to the same issue many times. If you report counts, explain what you counted. The frequency of a topic alone does not determine its importance for the interpretation.
5. Report a finding that connects interpretation with the material
A practical structure for a passage in the chapter is: an analytical finding, an explanation of its scope, a relevant quotation and commentary. You can change the order, but the reader should understand why you included the quotation.
An example paragraph that limits the conclusion
In R03’s account, peer support acts as an intermediary in contacting the mentor. She explains that “it took a friend to show me where to find the mentor’s office hours”. This example describes a way of overcoming a difficulty in finding information. It does not establish whether the cause was that the instructions were unavailable, unclear or published somewhere the participant did not know about.
In the completed chapter, address the relationships between themes, differing experiences and the limits of the material. Do not write “all participants” when you mean only some of them. Do not alter a quotation to remove a participant’s hesitation or strengthen your interpretation.
After presenting the findings, you can relate them to the literature. The difference between these tasks is explained in the guide to writing a discussion of results.
Do you need software to analyse interviews?
For a small amount of material, you can organise your work in a word processor with a structured register of extracts. With a larger dataset, analysis software helps you search, assign codes and return to the context. Choosing a tool does not replace choosing a method.
Niedbalski (2014) discusses the capabilities and limitations of NVivo and Atlas.ti in grounded theory research. His paper concerns a particular methodology; you should not transfer every procedure it describes to every interview analysis. You need to justify your own decisions regardless of the name of the software you use.
Check your chapter before submitting it
- Have you named the analysis method and actually applied it?
- Can you trace every quotation to its source?
- Does the chapter contain your analysis rather than just summaries of successive interviews?
- Does it address material that does not fit the initial interpretation?
- Are participant identifiers consistent and identifying information protected?
- Do the conclusions answer the research questions without going beyond the material?
Questions about analysing interviews
Does the service cover interviews and open-ended questions?
Yes, the scope can include organising participant accounts, coding, identifying categories or themes, and describing them with examples from the material. We adapt our approach to the chosen qualitative method and the thesis research questions.
Can I submit only one section of the thesis?
Yes. You can submit an individual chapter, questionnaire, methodology section, results, references, footnotes or another section. Support can also cover presentation content or editing a finished text.
When sending a section, include your thesis topic and outline, plus any feedback from your supervisor. This helps us understand how it fits into the full document.
Sources and the nature of the examples
The publications below were used to distinguish interpretation, tools and analytical approaches. The statements, memos and paragraph presented here are original teaching examples.
- Mandes, S., Karlińska, A. (2024). W stronę nowej metodologii analizy treści. Podobieństwa i różnice pomiędzy modelowaniem tematycznym i jakościową analizą treści. Przegląd Socjologii Jakościowej, 20(4). DOI: 10.18778/1733-8069.20.4.06.
- Niedbalski, J. (2014). Zastosowanie oprogramowania Atlas.ti i NVivo w realizacji badań opartych na metodologii teorii ugruntowanej. Przegląd Socjologii Jakościowej, 10(2), 60-80. DOI: 10.18778/1733-8069.10.2.04.
