Writing up qualitative research

There are specific challenges to writing up research based on qualitative data: there is a lot of data to present, complex stories to tell and conclusions that took a long time to develop. But some simple guidelines can help structuring any kind of research output

Writing up qualitative research

There are specific challenges to writing up research based on qualitative data: there is a lot of data to present, complex stories to tell and conclusions that took a long time to develop. But some simple guidelines can help structuring any kind of research output based on qualitative methods, and help communicate the rich results that come from qualitative data.


There’s a very general statement whenever you are writing or speaking – consider your audience. Who is likely to read your research output? Usually, you would have several from a project, such as a summary for the participants or special interest groups, a research paper for the scientific community (and the journal reviewers) and a thesis which is written to demonstrate knowledge and research skills for an examining panel. Understanding who you are writing for should dictate much of the specifics below, especially how much you need to write to explain certain aspects. Will your readers understand qualitative methodologies? A challenge is often writing for reviewers (or supervisors) who don’t have a qualitative background and don’t understand the methods and small sampling approaches in qualitative research.

This guide is designed to be general enough to apply to a research article or thesis chapter, but note that each will have it’s own specific guidelines for structure and content that should be applied. Some research journals have very specific structures and formatting guidelines to follow, where others allow for more flexibility. Regardless, the key points below should give a general overview of ‘best practice’ in writing up qualitative research.

Theory and Methodology

If you are coming from quantitative disciplines, this is the first section that seems unusual when writing up a research project. Although not explicit, all research has a epistemological and theoretical background and set of assumptions that guide the collection methods and interpretation. However, qualitative research has a strong tradition of making those explicit and justified. The researcher(s) should state their philosophical beliefs on knowledge that have shaped the project.

Describing the epistemology in detail would entail writing about how the researchers see the notion of ‘truth’, positivity, and how much can be understood in the world. Usually this need go no further than adopt one an existing critique or interpretation, such as feminist research methods, or post-positivism. But it should also explain why the qualitative approach was appropriate for this research question, and thus what the appropriate methodology was.

While often connected most clearly to method, the methodology should also touch on the approach selected for analysis and interpretation, noting that adopting something common like grounded theory actually implies a whole series of theoretical and epistemological underpinnings. In a journal article, this will be short: typically only a few sentences, unless the paper is adopting or proposing a new or unusual take on some theoretical stance. But in a qualitative-based thesis, this could well be a chapter in it’s own right.

Finally, researchers should consider some sort of engagement with ‘reflexivity’ and explaining the role and influence of the researcher on the research project. I’ve got a whole article on this, but being open and explicit about the fact that research is not ‘value-free’ and that the background and positionality of the researcher influences the project is usually a key principle in qualitative research.

Methods and analysis

Qualitative methodologies have great variety in the type of methods that can be used, and subtleties as to how they are applied. Therefore it is probably more important to describe the method in detail than in quantitative based research projects, which often have very standard formula for applying tools and measurements. It’s rarely enough to say ‘We interviewed’ or ‘We ran focus groups’ as these common methods can be applied in many different ways. So specify not just whether you used semi-structured or unstructured interviews, but consider describing the settings, interview guides, and any underlining theoretical or methodological stances.

As a reviewer, I also like to see details of both the sampling and recruitment categories – and as I’ve written before, I think these are distinct and important things. When using the small sample sizes that are hallmarks of most qualitative research, it is especially important to know who these people are, where they are from and how they were selected. Although we might get details about individual respondents lives and backgrounds later, its still important to give an overview of the people who make up the data – just remember it’s not always necessary to reduce these to statistical averages, such as the percentage of people in each age range.

It’s also very important to give a good level of detail about the analytical approach. I’ve written a LOT about this, but suffice it to say that just saying ‘We used grounded theory’ or ‘We applied thematic analysis’ is rarely specific enough – again due to the many different ways these terms are interpreted and enacted. Usually the qualitative analysis process is exploratory, non-linear and fraught with missteps and overlapping approaches, so is worthy of more detail. As qualitative research relies so much on the interpretation of the researcher, authors deserve some insight into the analytic process. If using coding, a list of codes, or illustration of some of the main ones can be insightful, even in an appendix.

Another pet hate or mine is papers that just refer to any qualitative analysis software as ‘We analysed the results in Quirkos/Nvivo’. Again, this tells the reader no more useful information than saying ‘We wrote the results using a blue pen’ – it’s not the tool, but how is is applied that affects and defines the analytic approach!

Quoting Data

In the results, findings and discussion sections, a lot of outputs based on qualitative research include extracts from the qualitative data: often from the transcripts from interviews, focus groups, diaries etc. Including quotes like this is a key feature in qualitative research outputs, they give the reader a chance to hear the participants in their own words, giving some validation of the researcher’s own interpretations. They also are effective at illustrating key points in the data and moving on the flow, and providing a framing device for discussing findings. Qualitative software like Quirkos can also help greatly at this stage, helping you keep track of quotes on particular topics, or organised by key quotes so you can quickly find them. However, there is the challenge of the ‘wall of text’.

There is a risk of the paper or chapter becoming over-run with quotations, which coming from different participants with different perspectives can become overwhelming for the reader. In general quotes should be used sparingly, for illustrative snapshots, or colourful instances where the quote says something better than the researcher can. In general, quotes from transcripts won’t be as short and generalisable as a sentence from the writer.

There’s also a temptation to use a quote to prove every conclusion or statement that the author wants to make, in the same way you would have a reference to back up each statement made in a literature review or introduction. But this is rarely necessary, as it will also break the flow of the writing and lead to issues with space and length. The writer should use just enough quotes to establish trust from the reader of their interpretations, saving quotes for when the author is making an important or surprising finding.

There is also the possibility of using two different types of quotes: short in-text quips, and longer multiple-sentence extractions. Shorter quotes are best for using just a few words, or a sentence at most that fits within the discussion. For example, one person may comment, ‘Time is always precious when you have children’. Here it’s not always necessary to attribute the quote to a particular individual (unless what they say is particular to their identity or case).

Longer quotes break the flow of the text, and should usually be formatted as an indented paragraph. These should almost always have a separate line after the quote giving the participant’s number, pseudonym and any relevant information about them, for example age, location or gender if it is relevant to understanding the context of the quote. Usually these should be kept as short as necessary, using … to truncate any sections of the quote that are not relevant. They also need to make sense alone, so any references to ‘it’ or ‘them’ or ‘that’ which are not obvious in the extract should be replaced with a definition in square brackets, for example [my feelings], [my husband] or [my car].

This is even true of other data types, especially pictures. Although most research products like journal articles and theses unfortunately make it very difficult to include audio and video extracts, pictures are possible. Yet these also break the flow of text, and take up a lot of space. They should also be used sparingly, and be well integrated and illustrative of the argument being made.

There can also be additional confidentiality considerations including any multimedia extracts. Text quotations automatically remove the audible voice and face of participants, but this might have to be done manually for participants. Permission to share photos, especially those showing faces and other sensitive or personal situations should also be explicitly sought, ideally for each type of output that might be produced.

Finally, quotes can’t do everything, many important aspects of subtext, and the unspoken (whether non-verbal, or not willing to share) are important context to give the reader. It’s up to the writer to connect the dots, be the expert that was part of the whole process and summarise a huge amount of rich qualitative data into a coherent story.

Discussion and other considerations

Latterly, any output needs to summarise the data and process, and refer back to the research questions and the wider literature. There will often be new questions posed by the data, as well as any conclusions from the data. Finally, consider being honest about what went wrong in the project – either in the data collection stage, or the analysis when theories and assumptions were challenged and changed.

Remember, just as good qualitative research has a story at it’s centre, the same should be true of communicating the findings. Partly it’s a story of the research project (and the role of the researchers in it), but also the story of the respondents – their own journey, experiences and lives. That is why creating a single story that encapsulates a common and differing tales across many participants can be a challenge.

Quirkos is a unique tool for qualitative analysis that helps you explore, manage and discover stories in your qualitative data. We designed it to help us write up our own qualitative data easier and better, and you can try yourself for free right here! There’s no registration needed for the offline version, and the cloud subscription service lets researchers, work, collaborate and share project data with ease.