How to do thematic analysis

Uncovering themes in data requires a systematic approach. Thematic analysis organizes data so you can easily recognize the context.

Thematic analysis is a method for analyzing qualitative data that involves reading through a data set and looking for patterns to derive themes. The researcher's subjective experience plays a central role in finding meaning within the data.

There are several approaches to thematic analysis. The one you use will depend on what’s most suitable for your research design.

Inductive thematic analysis approach

Inductive thematic analysis entails deriving meaning and identifying themes from data with no preconceptions. You analyze the data without any expected outcomes.

Deductive thematic analysis approach

In the deductive approach, you analyze data with a set of expected themes. Prior knowledge, research, or existing theory informs this approach.

Semantic thematic analysis approach

With the semantic approach, you ignore the underlying meaning of data. You take identifying themes at face value based on what is written or explicitly stated.

Latent thematic analysis approach

Unlike the semantic approach, the latent approach focuses on underlying meanings in data and looks at the reasons for semantic content. It involves an element of interpretation where you theorize meanings and don’t just take data at face value.

Thematic analysis is beneficial when you’re working with large bodies of data. It allows you to divide and categorize huge quantities of data in a way that makes it far easier to digest.  

The following scenarios warrant the use of thematic analysis:

• You’re new to qualitative analysis

• You need to identify patterns in data

• You want to involve participants in the process

Thematic analysis is particularly useful when you’re looking for subjective information such as experiences and opinions in surveys, interviews, conversations, or social media posts. 

Thematic analysis is a highly flexible approach to qualitative data analysis that you can modify to meet the needs of many studies. It enables you to generate new insights and concepts from data. 

Beginner researchers who are just learning how to analyze data will find thematic analysis very accessible. It’s easy for most people to grasp and can be relatively quick to learn.

The flexibility of thematic analysis can also be a disadvantage. It can feel intimidating to decide what’s important to emphasize, as there are many ways to interpret meaning from a data set.

Since it focuses on looking for patterns across data, you may overlook phenomena that occur in only one instance. Also, thematic analysis doesn't use theoretical frameworks. This could limit the interpretive power of an analysis. 

The basic thematic analysis process requires recognizing codes and themes within a data set. A code is a label assigned to a piece of data that you use to identify and summarize important concepts within a data set. A theme is a pattern that you identify within the data. Relevant steps may vary based on the approach and type of thematic analysis, but these are the general steps you’d take:

1. Familiarize yourself with the data(pre-coding work)

Before you can successfully work with data, you need to understand it. Get a feel for the data to see what general themes pop up. Transcribe audio files and observe any meanings and patterns across the data set. Read through the transcript, and jot down notes about potential codes to create. 

2. Create the initial codes (open code work)

Create a set of initial codes to represent the patterns and meanings in the data. Make a codebook to keep track of the codes. Read through the data again to identify interesting excerpts and apply the appropriate codes. You should use the same code to represent excerpts with the same meaning. 

3. Collate codes with supporting data (clustering of initial code)

Now it's time to group all excerpts associated with a particular code. If you’re doing this manually, cut out codes and put them together. Thematic analysis software will automatically collate them.

4. Group codes into themes (clustering of selective codes)

Once you’ve finalized the codes, you can sort them into potential themes. Themes reflect trends and patterns in data. You can combine some codes to create sub-themes.

5. Review, revise, and finalize the themes (final revision)

Now you’ve decided upon the initial themes, you can review and adjust them as needed. Each theme should be distinct, with enough data to support it. You can merge similar themes and remove those lacking sufficient supportive data. Begin formulating themes into a narrative. 

6. Write the report

The final step of telling the story of a set of data is writing the report. You should fully consider the themes to communicate the validity of your analysis.

A typical thematic analysis report contains the following:

• An introduction

• A methodology section

• Results and findings

• A conclusion

Your narrative must be coherent, and it should include vivid quotes that can back up points. It should also include an interpretive analysis and argument for your claims. In addition, consider reporting your findings in a flowchart or tree diagram, which can be independent of or part of your report.  

In conclusion, a thematic analysis is a method of analyzing qualitative data. By following the six steps, you will identify common themes from a large set of texts. This method can help you find rich and useful insights about people’s experiences, behaviors, and nuanced opinions.

Qualitative data analysis is the process of organizing, analyzing, and interpreting non-numerical and subjective data. The goal is to capture themes and patterns, answer questions, and identify the best actions to take based on that data. 

Researchers can use qualitative data to understand people’s thoughts, feelings, and attitudes. For example, qualitative researchers can help business owners draw reliable conclusions about customers’ opinions and discover areas that need improvement. 

In addition to thematic analysis, you can analyze qualitative data using the following:

Content analysis

Content analysis examines and counts the presence of certain words, subjects, and contexts in documents and communication artifacts, such as: 

• Text in various formats

• Pictures

• Videos

• Audio

This method transforms qualitative input into quantitative data. You can do it manually or with electronic tools that recognize patterns to make connections between concepts.  

Narrative analysis

Narrative analysis interprets research participants' stories from testimonials, case studies, interviews, and other text or visual data. It provides valuable insights into the complexity of people's feelings, beliefs, and behaviors.

Discourse analysis

In discourse analysis, you analyze the underlying meaning of qualitative data in a particular context, including: 

• Social

• Cultural

• Political

• Business

• Historical 

This approach allows us to study how people use language in text, audio, and video to unravel social issues, power dynamics, or inequalities. 

For example, you can look at how people communicate with their coworkers versus their bosses. Discourse analysis goes beyond the literal meaning of words to examine social reality.

Grounded theory analysis

In grounded theory analysis, you develop theories by examining real-world data. The process involves creating hypotheses and theories by systematically collecting and evaluating this data. While this approach is helpful for studying lesser-known phenomena, it might be overwhelming for a novice researcher. 

While qualitative data can answer questions that quantitative data can't, it still comes with challenges.

1. If done manually, qualitative data analysis is very time-consuming.

2. It can be hard to choose a method. 

3. Avoiding bias is difficult.

4. Human error affects accuracy and consistency.

To overcome these challenges, you should fine-tune your methods by using the appropriate tools in collaboration with teammates.

Learn more about thematic analysis software

Thematic analysis is a method of analyzing qualitative data. It is applied to texts, such as interviews or transcripts. The researcher closely examines the data to identify common patterns and themes.

You can do thematic analysis manually, but it is very time-consuming without the help of software.

The two main types of thematic analysis include codebook thematic analysis and reflexive thematic analysis.

Codebook thematic analysis uses predetermined codes and structured codebooks to analyze from a deductive perspective. You draw codes from a review of the data or an initial analysis to produce the codebooks.

Reflexive thematic analysis is more flexible and does not use a codebook. Researchers can change, remove, and add codes as they work through the data. 

The goal of thematic analysis is more than simply summarizing data; it's about identifying important themes. Good thematic analysis interprets, makes sense of data, and explains it. It produces trustworthy and insightful findings that are easy to understand and apply. 

Grouping codes into themes summarize sections of data in a useful way to answer research questions and achieve objectives. A theme identifies an area of data and tells the reader something about it. A good theme can sit alone without requiring descriptive text beneath it.

For example, if you were analyzing data on wildlife, codes might be owls, hawks, and falcons. These codes might fall beneath the theme of birds of prey. If your data were about the latest trends for teenage girls, codes such as mini skirts, leggings, and distressed jeans would fall under fashion.  

Thematic analysis is straightforward and intuitive enough that most people have no trouble applying it.