What (Exactly) Is Thematic Analysis?
Plain-Language Explanation & Definition (With Examples)
By: Jenna Crosley (PhD). Expert Reviewed By: Dr Eunice Rautenbach | April 2021
Thematic analysis is one of the most popular qualitative analysis techniques we see students opting for at Grad Coach – and for good reason. Despite its relative simplicity, thematic analysis can be a very powerful analysis technique when used correctly. In this post, we’ll unpack thematic analysis using plain language (and loads of examples) so that you can conquer your analysis with confidence.
First, the lingo…
Before we begin, let’s first lay down some terminology. When undertaking thematic analysis, you’ll make use of codes. A code is a label assigned to a piece of text, and the aim of using a code is to identify and summarise important concepts within a set of data, such as an interview transcript.
For example, if you had the sentence, “My rabbit ate my shoes”, you could use the codes “rabbit” or “shoes” to highlight these two concepts. The process of assigning codes is called coding. If this is a new concept to you, be sure to check out our detailed post about qualitative coding.
Codes are vital as they lay a foundation for themes. But what exactly is a theme? Simply put, a theme is a pattern that can be identified within a data set. In other words, it’s a topic or concept that pops up repeatedly throughout your data. Grouping your codes into themes serves as a way of summarising sections of your data in a useful way that helps you answer your research question(s) and achieve your research aim(s).
Alright – with that out of the way, let’s jump into the wonderful world of thematic analysis…
What is thematic analysis?
Thematic analysis is the study of patterns to uncover meaning. In other words, it’s about analysing the patterns and themes within your data set to identify the underlying meaning. Importantly, this process is driven by your research aims and questions, so it’s not necessary to identify every possible theme in the data, but rather to focus on the key aspects that relate to your research questions.
Although the research questions are a driving force in thematic analysis (and pretty much all analysis methods), it’s important to remember that these questions are not necessarily fixed. As thematic analysis tends to be a bit of an exploratory process, research questions can evolve as you progress with your coding and theme identification.
When should you use thematic analysis?
There are many potential qualitative analysis methods that you can use to analyse a dataset. For example, content analysis, discourse analysis, and narrative analysis are popular choices. So why use thematic analysis?
Thematic analysis is highly beneficial when working with large bodies of data, as it allows you to divide and categorise large amounts of data in a way that makes it easier to digest. Thematic analysis is particularly useful when looking for subjective information, such as a participant’s experiences, views, and opinions. For this reason, thematic analysis is often conducted on data derived from interviews, conversations, open-ended survey responses, and social media posts.
Your research questions can also give you an idea of whether you should use thematic analysis or not. For example, if your research questions were to be along the lines of:
- How do dog walkers perceive rules and regulations on dog-friendly beaches?
- What are students’ experiences with the shift to online learning?
- What opinions do health professionals hold about the Hippocratic code?
- How is gender constructed in a high school classroom setting?
These examples are all research questions centering on the subjective experiences of participants and aim to assess experiences, views, and opinions. Therefore, thematic analysis presents a possible approach.
In short, thematic analysis is a good choice when you are wanting to categorise large bodies of data (although the data doesn’t necessarily have to be large), particularly when you are interested in subjective experiences.
What are the main approaches?
Broadly speaking, there are two overarching approaches to thematic analysis: inductive and deductive. The approach you take will depend on what is most suitable in light of your research aims and questions. Let’s have a look at the options.
The inductive approach
The inductive approach involves deriving meaning and creating themes from data without any preconceptions. In other words, you’d dive into your analysis without any idea of what codes and themes will emerge, and thus allow these to emerge from the data.
For example, if you’re investigating typical lunchtime conversational topics in a university faculty, you’d enter the research without any preconceived codes, themes or expected outcomes. Of course, you may have thoughts about what might be discussed (e.g., academic matters because it’s an academic setting), but the objective is to not let these preconceptions inform your analysis.
The inductive approach is best suited to research aims and questions that are exploratory in nature, and cases where there is little existing research on the topic of interest.
The deductive approach
In contrast to the inductive approach, a deductive approach involves jumping into your analysis with a pre-determined set of codes. Usually, this approach is informed by prior knowledge and/or existing theory or empirical research (which you’d cover in your literature review).
For example, a researcher examining the impact of a specific psychological intervention on mental health outcomes may draw on an existing theoretical framework that includes concepts such as coping strategies, social support, and self-efficacy, using these as a basis for a set of pre-determined codes.
The deductive approach is best suited to research aims and questions that are confirmatory in nature, and cases where there is a lot of existing research on the topic of interest.
Regardless of whether you take the inductive or deductive approach, you’ll also need to decide what level of content your analysis will focus on – specifically, the semantic level or the latent level.
A semantic-level focus ignores the underlying meaning of data, and identifies themes based only on what is explicitly or overtly stated or written – in other words, things are taken at face value.
In contrast, a latent-level focus concentrates on the underlying meanings and looks at the reasons for semantic content. Furthermore, in contrast to the semantic approach, a latent approach involves an element of interpretation, where data is not just taken at face value, but meanings are also theorised.
“But how do I know when to use what approach?”, I hear you ask.
Well, this all depends on the type of data you’re analysing and what you’re trying to achieve with your analysis. For example, if you’re aiming to analyse explicit opinions expressed in interviews and you know what you’re looking for ahead of time (based on a collection of prior studies), you may choose to take a deductive approach with a semantic-level focus.
On the other hand, if you’re looking to explore the underlying meaning expressed by participants in a focus group, and you don’t have any preconceptions about what to expect, you’ll likely opt for an inductive approach with a latent-level focus.
Simply put, the nature and focus of your research, especially your research aims, objectives and questions will inform the approach you take to thematic analysis.
What are the types of thematic analysis?
Now that you’ve got an understanding of the overarching approaches to thematic analysis, it’s time to have a look at the different types of thematic analysis you can conduct. Broadly speaking, there are three “types” of thematic analysis:
- Reflexive thematic analysis
- Codebook thematic analysis
- Coding reliability thematic analysis
Let’s have a look at each of these:
Reflexive thematic analysis takes an inductive approach, letting the codes and themes emerge from that data. This type of thematic analysis is very flexible, as it allows researchers to change, remove, and add codes as they work through the data. As the name suggests, reflexive thematic analysis emphasizes the active engagement of the researcher in critically reflecting on their assumptions, biases, and interpretations, and how these may shape the analysis.
Reflexive thematic analysis typically involves iterative and reflexive cycles of coding, interpreting, and reflecting on data, with the aim of producing nuanced and contextually sensitive insights into the research topic, while at the same time recognising and addressing the subjective nature of the research process.
Codebook thematic analysis, on the other hand, lays on the opposite end of the spectrum. Taking a deductive approach, this type of thematic analysis makes use of structured codebooks containing clearly defined, predetermined codes. These codes are typically drawn from a combination of existing theoretical theories, empirical studies and prior knowledge of the situation.
Codebook thematic analysis aims to produce reliable and consistent findings. Therefore, it’s often used in studies where a clear and predefined coding framework is desired to ensure rigour and consistency in data analysis.
Coding reliability thematic analysis necessitates the work of multiple coders, and the design is specifically intended for research teams. With this type of analysis, codebooks are typically fixed and are rarely altered.
The benefit of this form of analysis is that it brings an element of intercoder reliability where coders need to agree upon the codes used, which means that the outcome is more rigorous as the element of subjectivity is reduced. In other words, multiple coders discuss which codes should be used and which shouldn’t, and this consensus reduces the bias of having one individual coder decide upon themes.
Quick Recap: Thematic analysis approaches and types
To recap, the two main approaches to thematic analysis are inductive, and deductive. Then we have the three types of thematic analysis: reflexive, codebook and coding reliability. Which type of thematic analysis you opt for will need to be informed by factors such as:
- The approach you are taking. For example, if you opt for an inductive approach, you’ll likely utilise reflexive thematic analysis.
- Whether you’re working alone or in a group. It’s likely that, if you’re doing research as part of your postgraduate studies, you’ll be working alone. This means that you’ll need to choose between reflexive and codebook thematic analysis.
Now that we’ve covered the “what” in terms of thematic analysis approaches and types, it’s time to look at the “how” of thematic analysis.
Need a helping hand?
See how Grad Coach can help you...
How to “do” thematic analysis
At this point, you’re ready to get going with your analysis, so let’s dive right into the thematic analysis process. Keep in mind that what we’ll cover here is a generic process, and the relevant steps will vary depending on the approach and type of thematic analysis you opt for.
The first step in your thematic analysis involves getting a feel for your data and seeing what general themes pop up. If you’re working with audio data, this is where you’ll do the transcription, converting audio to text.
At this stage, you’ll want to come up with preliminary thoughts about what you’ll code, what codes you’ll use for them, and what codes will accurately describe your content. It’s a good idea to revisit your research topic, and your aims and objectives at this stage. For example, if you’re looking at what people feel about different types of dogs, you can code according to when different breeds are mentioned (e.g., border collie, Labrador, corgi) and when certain feelings/emotions are brought up.
As a general tip, it’s a good idea to keep a reflexivity journal. This is where you’ll write down how you coded your data, why you coded your data in that particular way, and what the outcomes of this data coding are. Using a reflexive journal from the start will benefit you greatly in the final stages of your analysis because you can reflect on the coding process and assess whether you have coded in a manner that is reliable and whether your codes and themes support your findings.
As you can imagine, a reflexivity journal helps to increase reliability as it allows you to analyse your data systematically and consistently. If you choose to make use of a reflexivity journal, this is the stage where you’ll want to take notes about your initial codes and list them in your journal so that you’ll have an idea of what exactly is being reflected in your data. At a later stage in the analysis, this data can be more thoroughly coded, or the identified codes can be divided into more specific ones.
Step 2! You’re going strong. In this step, you’ll want to look out for patterns or themes in your codes. Moving from codes to themes is not necessarily a smooth or linear process. As you become more and more familiar with the data, you may find that you need to assign different codes or themes according to new elements you find. For example, if you were analysing a text talking about wildlife, you may come across the codes, “pigeon”, “canary” and “budgerigar” which can fall under the theme of birds.
As you work through the data, you may start to identify subthemes, which are subdivisions of themes that focus specifically on an aspect within the theme that is significant or relevant to your research question. For example, if your theme is a university, your subthemes could be faculties or departments at that university.
In this stage of the analysis, your reflexivity journal entries need to reflect how codes were interpreted and combined to form themes.
By now you’ll have a good idea of your codes, themes, and potentially subthemes. Now it’s time to review all the themes you’ve identified. In this step, you’ll want to check that everything you’ve categorised as a theme actually fits the data, whether the themes do indeed exist in the data, whether there are any themes missing, and whether you can move on to the next step knowing that you’ve coded all your themes accurately and comprehensively. If you find that your themes have become too broad and there is far too much information under one theme, it may be useful to split this into more themes so that you’re able to be more specific with your analysis.
In your reflexivity journal, you’ll want to write about how you understood the themes and how they are supported by evidence, as well as how the themes fit in with your codes. At this point, you’ll also want to revisit your research questions and make sure that the data and themes you’ve identified are directly relevant to these questions.
By this point, your analysis will really start to take shape. In the previous step, you reviewed and refined your themes, and now it’s time to label and finalise them. It’s important to note here that, just because you’ve moved onto the next step, it doesn’t mean that you can’t go back and revise or rework your themes. In contrast to the previous step, finalising your themes means spelling out what exactly the themes consist of, and describe them in detail. If you struggle with this, you may want to return to your data to make sure that your data and coding do represent the themes, and if you need to divide your themes into more themes (i.e., return to step 3).
When you name your themes, make sure that you select labels that accurately encapsulate the properties of the theme. For example, a theme name such as “enthusiasm in professionals” leaves the question of “who are the professionals?”, so you’d want to be more specific and label the theme as something along the lines of “enthusiasm in healthcare professionals”.
It is very important at this stage that you make sure that your themes align with your research aims and questions. When you’re finalising your themes, you’re also nearing the end of your analysis and need to keep in mind that your final report (discussed in the next step) will need to fit in with the aims and objectives of your research.
In your reflexivity journal, you’ll want to write down a few sentences describing your themes and how you decided on these. Here, you’ll also want to mention how the theme will contribute to the outcomes of your research, and also what it means in relation to your research questions and focus of your research.
By the end of this stage, you’ll be done with your themes – meaning it’s time to write up your findings and produce a report.
You’re nearly done! Now that you’ve analysed your data, it’s time to report on your findings. A typical thematic analysis report consists of:
- An introduction
- A methodology section
- Your results and findings
- A conclusion
When writing your report, make sure that you provide enough information for a reader to be able to evaluate the rigour of your analysis. In other words, the reader needs to know the exact process you followed when analysing your data and why. The questions of “what”, “how”, “why”, “who”, and “when” may be useful in this section.
So, what did you investigate? How did you investigate it? Why did you choose this particular method? Who does your research focus on, and who are your participants? When did you conduct your research, when did you collect your data, and when was the data produced? Your reflexivity journal will come in handy here as within it you’ve already labelled, described, and supported your themes.
If you’re undertaking a thematic analysis as part of a dissertation or thesis, this discussion will be split across your methodology, results and discussion chapters. For more information about those chapters, check out our detailed post about dissertation structure.
It’s absolutely vital that, when writing up your results, you back up every single one of your findings with quotations. The reader needs to be able to see that what you’re reporting actually exists within the results. Also make sure that, when reporting your findings, you tie them back to your research questions. You don’t want your reader to be looking through your findings and asking, “So what?”, so make sure that every finding you represent is relevant to your research topic and questions.
Quick Recap: How to “do” thematic analysis
Getting familiar with your data: Here you’ll read through your data and get a general overview of what you’re working with. At this stage, you may identify a few general codes and themes that you’ll make use of in the next step.
Search for patterns or themes in your codes: Here you’ll dive into your data and pick out the themes and codes relevant to your research question(s).
Review themes: In this step, you’ll revisit your codes and themes to make sure that they are all truly representative of the data, and that you can use them in your final report.
Finalise themes: Here’s where you “solidify” your analysis and make it report-ready by describing and defining your themes.
Produce your report: This is the final step of your thematic analysis process, where you put everything you’ve found together and report on your findings.
Tips & Suggestions
In the video below, we share 6 time-saving tips and tricks to help you approach your thematic analysis as effectively and efficiently as possible.
In this article, we’ve covered the basics of thematic analysis – what it is, when to use it, the different approaches and types of thematic analysis, and how to perform a thematic analysis.
If you have any questions about thematic analysis, drop a comment below and we’ll do our best to assist. If you’d like 1-on-1 support with your thematic analysis, be sure to check out our research coaching services here.
Psst… there’s more (for free)
This post is part of our dissertation mini-course, which covers everything you need to get started with your dissertation, thesis or research project.