What is the difference between thematic analysis and grounded theory?

Thematic analysis is a method for identifying patterns in data without requiring a specific theoretical framework, while grounded theory aims to build new theory from data through a more prescribed iterative process. Thematic analysis is more flexible and accessible, making it suitable for applied UX research. Grounded theory involves additional steps like theoretical sampling and saturation that go beyond pattern identification.

How long does thematic analysis typically take?

The time varies significantly based on data volume and depth of analysis. For a typical UX study with 10 to 15 interview transcripts, expect 2 to 4 days for thorough analysis including familiarization, coding, theme development, and report writing. Larger datasets or more rigorous academic analysis can take weeks. Using analysis software can speed up the coding process considerably.

Should I use an inductive or deductive approach?

Use an inductive approach when you want themes to emerge from the data without preconceptions, which is ideal for exploratory research. Use a deductive approach when you have specific hypotheses or frameworks you want to test against the data. Many UX researchers use a hybrid approach, starting inductively while remaining aware of relevant frameworks and design heuristics.

How many themes should a good thematic analysis produce?

There is no fixed number, but most analyses produce 4 to 8 major themes. Fewer than 3 themes suggests the analysis is too broad and misses important distinctions. More than 10 themes may indicate insufficient abstraction. Each theme should be distinct, internally coherent, and supported by multiple data points from different participants.

What tools work best for thematic analysis?

For small datasets, physical sticky notes on a wall or a simple spreadsheet work well. For larger projects, dedicated qualitative analysis tools like Dovetail, NVivo, Atlas.ti, or Dedoose offer features like code management, search, and visualization. Collaborative tools like Miro or FigJam can support team-based coding sessions, especially for remote teams.

Can I do thematic analysis with a small team or solo?

Yes, solo thematic analysis is common in UX practice, though it requires extra diligence to avoid personal bias. Keep a reflexivity journal documenting your assumptions and decisions. If possible, have a colleague review a portion of your coding to check reliability. Team-based analysis is preferred for academic rigor but practical constraints in industry often necessitate solo work.

Methods Articles Compare Find a Method About

ParticipatoryTesting & ValidationMixed-Methods ResearchIntermediate

Thematic Analysis

Identify and interpret recurring patterns in qualitative data to transform raw observations into actionable design insights.

Thematic analysis identifies recurring patterns and themes in qualitative data like interviews and surveys, transforming raw observations into structured insights.

Duration60 minutes or more.

MaterialsSorting materials, post-its, writing tools, wall.

People1 designer or more.

InvolvementIndirect User Involvement

Thematic Analysis is a systematic qualitative research method for identifying, analyzing, and interpreting recurring patterns and themes within data such as interview transcripts, survey responses, diary entries, or field notes. UX researchers, academic researchers, and product teams use it to move from raw, unstructured observations to clearly defined insights that can inform design decisions and stakeholder communication. The process involves immersing yourself in the data, generating initial codes that label meaningful segments, grouping those codes into broader themes, and refining the thematic structure until it tells a coherent story about what the data reveals. Thematic Analysis is flexible enough to work with almost any type of qualitative data and can be applied using either an inductive approach, where themes emerge from the data itself, or a deductive approach, where themes are guided by existing frameworks or research questions. This versatility makes it one of the most widely used analysis methods in UX research, social sciences, and healthcare research. When done well, thematic analysis transforms overwhelming volumes of qualitative information into a structured, evidence-based narrative that teams can act on confidently.

WHEN TO USE

After conducting interviews, focus groups, or diary studies when you need to make sense of large volumes of qualitative data.
When synthesizing findings across multiple research methods or data sources into a unified set of insights.
During the analysis phase of any qualitative study to identify patterns that inform design decisions and priorities.
When you need to present research findings to stakeholders in a structured, evidence-based narrative format.
After open-ended survey questions generate too much textual data for simple reading and require systematic analysis.

WHEN NOT TO USE

×When you have only quantitative data that is better analyzed with statistical methods than qualitative coding approaches.
×For very small datasets of fewer than 3 to 4 participants where there is insufficient data to identify meaningful patterns.
×When you need real-time insights during a live research session and cannot invest the time required for thorough coding.
×If the research question requires preserving individual narratives intact rather than fragmenting data into coded segments.

HOW TO RUN

Step-by-Step Process

1. Data Familiarization

Begin by thoroughly reading and immersing yourself in the data to become familiar with the content. This can include reading transcripts, watching videos, or going through notes multiple times. This step helps in understanding the participants' perspectives, their experiences, and the context surrounding their statements.

2. Generating Initial Codes

Systematically work through the data and generate descriptive codes, labels or tags to define key features or ideas found in the dataset. The codes should be brief, accurate, and energetically descriptive. This process is iterative and should remain flexible as you move through the data, as new insights may require revising or creating new codes.

3. Searching for Themes

Start analyzing the codes and look for potential patterns or relationships that can combine them into larger themes. Themes can be more abstract in nature than codes, capturing concepts or ideas underlying the data. At this stage, you can use visual tools like mind-maps or diagrams to help explore connections and potential thematic structures.

4. Reviewing Themes

Review your identified themes and make sure they accurately represent the data. This includes checking if there are any discrepancies or inconsistencies within the themes, and whether any refinements, merging, or splitting of themes are necessary. Additionally, some themes may be discarded if they do not contribute to the overall understanding of the research topic.

5. Defining and Naming Themes

Further refine and develop a clear definition and name for each theme. This should describe the core essence of what the theme represents and the aspects of the data it covers. A well-defined theme should be able to explain the overall research question and tell the story of the data.

6. Report Writing

Write a detailed report that represents the findings of the thematic analysis. This includes a thorough description of each theme, evidence from the data, and quotes or examples to support your claims. Ensure your report is coherent, logical, and clearly communicates the findings to the reader. Tie your themes back to the research question and provide any relevant insights for your research objectives.

EXPECTED OUTCOME

What to Expect

After completing a thematic analysis, your team will have a structured set of clearly defined themes supported by evidence from the data. Each theme will be named, defined, and illustrated with representative participant quotes that make the findings tangible and compelling. The thematic map will show how themes relate to each other and to the original research questions, providing a coherent narrative that moves beyond individual data points to reveal broader patterns. Stakeholders will receive a research report that translates qualitative complexity into actionable insights, identifying what users need, where they struggle, and what opportunities exist. The codebook created during analysis serves as a reusable reference for future research, enabling consistent coding across studies and building institutional knowledge over time.

PRO TIPS

Expert Advice

Start with initial codes that stay close to the data, then gradually abstract to higher-level themes over multiple passes.

Keep a codebook that defines each code clearly to ensure consistency, especially when multiple researchers are involved.

Document your decision-making process for merging or splitting themes to maintain analytical rigor and transparency.

Look for both semantic themes on the surface and latent themes that reveal underlying assumptions and deeper meanings.

Use participant quotes to ground each theme in evidence, making findings more compelling and credible for stakeholders.

Involve a second researcher in coding a subset of data to check for inter-rater reliability and reduce individual bias.

Set rules for discussion in advance when sorting as a team to prevent debates from derailing the analysis process.

Use digital tools like Dovetail or spreadsheets for remote analysis while keeping physical post-its for in-person sessions.

COMMON MISTAKES

Pitfalls to Avoid

Themes matching interview questions

Using interview questions as themes rather than letting themes emerge from the data produces superficial analysis. Themes should capture patterns across questions, not mirror the structure of your discussion guide.

Coding at too high a level

Starting with broad categories rather than specific codes misses important nuances in the data. Begin with fine-grained codes close to the data, then progressively abstract to higher-level themes through iterative passes.

Single-coder bias

Having only one person code all the data introduces individual interpretation bias. Use a second coder on at least a subset of data to check inter-rater reliability and discuss discrepancies.

Insufficient data immersion

Rushing through the familiarization phase leads to surface-level themes that miss deeper insights. Read through all data at least twice before starting to code, noting initial impressions and patterns.

Dropping outlier data points

Excluding data that does not fit neatly into themes weakens the analysis. Outliers often reveal important edge cases or emerging trends that deserve attention rather than dismissal.

DELIVERABLES

What You'll Produce

Research Plan

Document outlining objectives, scope, methodology, and timeline for analysis.

Interview Guide

Prepared questions and prompts for engaging participants in conversation.

Transcripts

Written records of interviews capturing dialogue and non-verbal cues.

Codebook

Systematic organization of codes, themes, and sub-themes from the data.

Coding Matrix

Visual representation showing patterns and relationships between codes.

Thematic Map

Visual diagram of main themes and their relationships within the data.

Analysis Report

Comprehensive document detailing themes, patterns, and implications.

Findings Presentation

Stakeholder presentation summarizing key insights and recommendations.

FAQ

Frequently Asked Questions

METHOD DETAILS

Goal: Testing & Validation
Sub-category: Affinity Diagramming
Tags: thematic analysisqualitative researchcodingthemespattern recognitiondata analysisaffinity diagrammingresearch synthesisinterview analysiscategorization
Related Topics: Qualitative Research MethodsAffinity DiagrammingGrounded TheoryContent AnalysisUser ResearchResearch Synthesis

HISTORY

Thematic analysis has roots in early qualitative research traditions but was formalized as a distinct method by Virginia Braun and Victoria Clarke in their influential 2006 paper 'Using Thematic Analysis in Psychology.' Before this publication, thematic analysis was widely practiced but poorly defined, often treated as an unnamed step within other methodologies like grounded theory or content analysis. Braun and Clarke's contribution was to articulate a clear six-phase process and establish thematic analysis as a method in its own right, independent of any particular theoretical framework. Their work has since been cited over 100,000 times, making it one of the most influential methodology papers in the social sciences. In UX research, thematic analysis gained prominence in the 2010s as the field matured and practitioners needed rigorous yet practical methods for analyzing qualitative data from user interviews and field studies. Today it is considered a foundational skill for UX researchers worldwide.

SUITABLE FOR

Synthesizing interview transcripts, field notes, and open-ended survey responses into coherent narratives
Discovering meaningful patterns and connections across large sets of qualitative research data
Building evidence-based arguments for design decisions grounded in user research findings
Processing data from multiple research methods into a unified set of insights and themes
Creating research reports that communicate complex findings to stakeholders clearly
Identifying unexpected themes that challenge existing assumptions and reveal new opportunities
Analyzing open-ended survey responses at scale when quantitative analysis alone is insufficient
Organizing brainstorming session outputs into structured categories for action planning

RESOURCES

RELATED METHODS