Qualitative Data Analysis

Author: Matt Carrano | Last edit: July 07, 2025

What is qualitative analysis and synthesis?

Many user research techniques, including user interviews, usability tests, and other direct user engagements generate qualitative data - data that is not numerical in nature or easily analyzed using statistical methods. The data that you collect in the form of notes or recordings coming out of a simple study with even a few participants can be overwhelming. This method introduces structured techniques for reducing that data to a smaller set of insights that can be actionable by your team. The approach used here is designed to simplify the process while maintaining a degree of objectivity and rigor to connect direct observations and/or verbatim comments with expressions of desired user outcomes. 

Why perform qualitative analysis?

Analysis and synthesis is an essential final step in any research activity. It’s a means for making sense out of the large amount of raw data that is generated from the process.

When should you perform qualitative analysis?

Analysis typically takes place at the end of the research process, after all data has been collected. However, there are things you can do along the way to get started with the process and make it easier to complete at the end.

How to perform qualitative analysis and synthesis

This section outlines a step-by-step process for performing qualitative data analysis using a thematic analysis approach. While there are many ways to approach the analysis problem, thematic analysis tends to by the simplest and most widely used approach for tackling this problem.

Process overview

The challenge for qualitative data analysis and synthesis is to map what you’ve learned to study goals and generate new insights based on your evidence. In his book, Head First Data Analysis, Michael Milton characterizes the process like this:

The approach followed here is referred to as a “thematic analysis” and is the most common way to disassemble and evaluate qualitative data coming from user research studies. Other approaches are described in this article from User Interviews.com. Here is an illustration of what that process might look like:

From: How to Analyze Qualitative Data from UX Research: Thematic Analysis, Maria Rosala, Nielsen Norman Group, 2022.

The sections below lay out these steps in more detail and provide an example of how you can put this methodology into action.

Gather and review your data

Your analysis process starts with your raw data. This will take the form of notes or recordings captured during or immediately after the research activity. If you are working with video or audio recordings, transcribing or distilling those recordings into text-based notes is a necessary first step. This article from User Interviews provides a great overview of different note-taking techniques and templates you can use depending on the type of study you are running. Transcription can be performed by feeding your video capture into EnjoyHQ (available to licensed Red Hat users) or by using the transcription tools in Google Meet or similar platforms.

If multiple researchers participated in the study, it will be useful for everyone on the team to review all transcripts before beginning the analysis. That will ensure everyone participating in the analysis process is starting with a common understanding of the data that was collected.

Tagging your data

After you have gathered all of your raw data, the next step is to distill that information into a set of tags (sometimes called “codes”) that will facilitate categorization of findings into themes. A tag or code is simply a word or phrase that acts as a label for a segment of text. Tags are like keywords that help you search for an article or hashtags that can identify a tweet. Tags can be used to normalize data by extracting the important information from a more verbose snippet of raw data. They will help us aggregate and categorize data points later on in the process. Multiple tags may be assigned to the same piece of data. 

For example, suppose a verbatim comment that you recorded during a set of interviews about an online service was, “the signup process takes too long.” There are two important aspects of this statement that may be related but are separate: the signup process and things take too long. We can define these as tags and apply them to this comment as well as to any other comments where similar sentiments are expressed. In this way we can aggregate all of the comments related to the signup process separately from comments about things taking a long time to complete. 

What to tag

The following are some tips for knowing when to tag or code a comment or note (from https://uxbooth.com/articles/a-guide-to-user-research-analysis/):

• It aligns directly to a project/research goal.
• The participant specifically said or implied that something is very important.
• Repetition – a thing is said or heard multiple times.
• Patterns – when certain observations are related or important to other tags and themes already established in the project goals or research.

Experienced researchers often will code their raw data immediately following each session. This both gives them a head start on the analysis phase to take place later and lets them consider the meaning of what they just heard while the conversation is still fresh in their mind.

Here’s an example of how you might take a verbatim response to a question about climate change and convert it into a set of codes:

Adapted from: How to Do Thematic Analysis | Step-by-Step Guide & Examples, Jack Caufield, Scribbr.com, 2023.

Notice how responses recorded from a single participant can be distilled into a simple idea, like “distrust of experts” that could be applied to a variety of statements across multiple interviews.

Identifying themes

Thematic analysis is the process of grouping your coded data into themes. Some themes may be taken directly from your research goals, i.e., what you were trying to learn from this study. Other themes may emerge from looking at the data. Themes might include things like:

• Demographics that characterize a set of users.
• Activities that people perform.
• Preferences or opinions.
• Behavioral variables.
• Pain points, frustrations, or problems that people experience.

What you are trying to do through this process is to find relationships between different data points to paint a bigger picture. Consider what the data associated with each theme is telling you about user behavior. Is there a consistent pattern in participant responses? Are responses segmented, i.e., is it possible that there are two or more groups of users that have differing or conflicting opinions around a particular theme?

Here is an example from the same climate change study referenced above that shows how tags can be turned into themes:

Adapted from: How to Do Thematic Analysis | Step-by-Step Guide & Examples, Jack Caufield, Scribbr.com, 2023.

Notice how themes are more general and more encompassing than codes and that multiple codes might all suggest a common theme.

Themes should lead to key insights. Ask “what patterns emerge around each of these themes and how do they help to answer the original research questions? How do these themes suggest a larger story?” The insights are what your stakeholders will be most interested in hearing as a result of your research. They should include a statement of what you’ve learned and a reference to the data and themes that support that statement.

Again, returning to our example about climate change, we can conclude that misinformation, conflicting information, and a general mistrust of expertise leads to a situation where some segments of the public are reluctant to support or practice behaviors that might stem the threat of climate change.

Thematic analysis simplified

As you might imagine, the process described above can be tedious if you are working with large amounts of data. There are software applications that can help simplify and automate this process. Two popular applications include Dovetail and EnjoyHQ. Another popular approach that works well with a team is to employ affinity mapping. Using an affinity mapping approach, the team performing the analysis will:

1. Read through each interview transcript to understand key ideas expressed relative to each question.
    
2. Distill these ideas into tags that capture key takeaways. In this case, tags are insights that might correspond directly to questions asked during the interview, but they are not required to. Look for significant statements that may have been made by a participant even if they were not directly related to the question at hand.
3. Write a short statement that summarizes each insight on a sticky note and add it to the affinity mapping board. This could be a physical board, if the activity is being done in person, or a virtual board using a tool like Miro. Use a separate color to represent each participant so that individual insights can be mapped back to participants. 
4. Group stickies that seem to go together. This should be an iterative process. It’s also okay to duplicate sticky notes that feel like they could belong to multiple groups.
5. Once you are satisfied with your groupings, identify a theme that can be applied to each group. Ask, what do the insights have in common? The theme should be summarized in a short statement that you use to label the group of stickies.
6. For each theme, write one or two sentences that elaborates on and summarizes what we can conclude from the data as it relates to the theme at hand. You may choose to go back to the raw data and find quotes that support each theme and add them to your summary. This will help complete the connection between your analysis and the words of actual target users.

Example of thematic analysis with affinity mapping in practice

Let’s apply thematic analysis to draw conclusions from a set of qualitative interviews of college students to understand what would motivate them to use a software application to adopt more environmentally friendly lifestyles. Goals of the study were to understand:

1. What are students' primary concerns about the environment?
2. What things do students currently do and what would motivate them to do more?
3. What are current pain points or demotivators?
4. What are current technology preferences for tracking personal behavioral trends?

Questions and raw data collected from the student interviews is summarized in the spreadsheet here: User Interview Analysis - Example Data

An affinity mapping approach was used to identify common themes emerging from the data and to summarize key findings relative to those themes. Miro was used as a tool for performing the affinity mapping activity. 

Following the initial coding activity, the Miro board shows insights from the raw interview notes copied onto sticky notes for each participant.

The sticky notes were then grouped and labeled like so:

Finally the emergent themes were summarized into a table, as follows, along with quotes from the original transcripts to connect insights back to participant verbatims.

The completed analysis gives us a more consolidated overview of the major trends that emerged from the research and weaves a story based on finding common ground between participants. This will provide a solid foundation for further downstream deliverables like personas and journey maps that place the user’s experience in context and illustrate future possibilities. 

Summary

In summary, there is not one correct or fool-proof way to do qualitative data analysis. But however you choose to approach this problem, you should always keep the following in mind:

• Your analysis should start by distilling the large amount of data you collected into a more concise set of takeaways that can be more easily grouped together and compared to lead to more global insights.
   
• Your primary objective should be to identify trends. You cannot possibly react to everything one participant says, but rather you are looking for common ideas and themes that you’ve heard from multiple people.
   
• It’s important to maintain your objectivity. Be careful to not infer things that participants did not actually say and make sure that your conclusions are supported by data. Linking your insights back to verbatim comments is a good way to preserve that connection.
   
• Keep your initial research questions in mind while performing your analysis, but also leave yourself open to emergent trends. Answering only the questions at hand might lead us to miss important insights. Let your findings emerge from the data.

The resources and references listed below provide some additional insights into how one can address the analysis problem.

Resources and references

A short course on performing qualitative data analysis and synthesis entitled, “Making Sense, Stories, and Successes from Data,” was offered to UXDers in early 2024. This addresses the problem of making sense of qualitative data and is recommended for PWDRs attempting this activity for the first time. Slides and a recording link (approximately 3 hours) are available.

Other references:

[1] A Guide to UX Research Analysis 

[2] Analysis in UX Research 

[3] How to Do Thematic Analysis | Step-by-Step Guide & Examples, Jack Caufield, Scribbr.com, 2023.

[4] How to Analyze Qualitative Data from UX Research: Thematic Analysis, Maria Rosala, Nielsen Norman Group, 2022.

[5] Qualitative Coding Tutorial: How To Code Qualitative Data For Analysis (4 Steps + Examples) 

[6] Qualitative Data Analysis 101 Tutorial: 6 Analysis Methods + Examples 

[7] Qualitative Coding: How to Turn Complex Data into Conclusive Insights, Lizzy Burnam, User Interviews, Inc, 2022.

Get in Touch!

Spotted a typo? Any feedback you want to share? Do you want to collaborate? Get in touch with the UXD Methods working group using one of the links below!

Drop us a message on the #uxd-hub Slack channel

Submit Feedback