Thematic analysis - how many themes to have in Qualitative Data Analysis?

TL;DR

A fixed universal number of themes doesn’t exist; theme count should support a coherent, evidence-based narrative with enough depth.

Briefing Cornell Notes

Briefing

How many themes is “enough” in qualitative analysis? There’s no universal number—most researchers land in a practical range, and the deciding factor is how to tell a coherent, detailed story from the data. In day-to-day work, Dr Kriukow reports typically generating anywhere from two to eight (sometimes ten) themes per study, with an “average” often closer to three to six themes for a standard-length write-up. That narrower band aligns with Braun and Clarke’s commonly cited guidance for thematic analysis in a roughly 10,000-word journal article, where going much beyond a handful of themes can make it hard to provide enough depth and specificity.

Still, the guidance shouldn’t be treated as a rigid rule tied to publication length. A larger study can legitimately produce more themes; what matters is selecting and structuring themes so the write-up remains detailed and meaningful. The core criterion is the narrative purpose of qualitative reporting: findings are essentially a story built from the data. Different researchers working with similar (or even the same) dataset may choose different “building blocks” (themes) because they emphasize different aspects of the story—while staying anchored in the evidence.

The transcript frames two common extremes. At one end, a study might produce only one theme—possible in very small, highly focused projects—but that often fails to exhaust the dataset or fully convey the complexity of participants’ accounts. At the other end, many students and early-career researchers generate a large number of themes (even 20 or 30) after coding deeply, then get stuck trying to decide what to keep, what to merge, and what to cut so the final analysis becomes readable and persuasive.

A concrete example illustrates how theme counts can change without changing the underlying data. Consider interviews with social workers (including those serving minority groups) aimed at producing recommendations for practice. One researcher might organize the analysis around a single overarching theme: recommendations for improving practice, with subthemes such as more training, increased staffing, and cultural/linguistic awareness training. Another researcher might restructure the same material into three themes: (1) challenges social workers face (e.g., lack of staff, heavy workload, limited cultural and language understanding), (2) good practices or facilitators (e.g., adequate staffing and realistic workload planning), and (3) suggestions/recommendations. Neither framework is inherently “better” or “worse”—they differ because they emphasize different parts of the story.

The transcript’s bottom line is that theme selection should serve implications and relevance: does the analysis answer the research question, and will it produce findings that are useful for practice or policy? Even if “defining good practice” isn’t explicitly a research question, a theme can still be added if it helps explain how participants’ beliefs connect to their reported challenges and recommendations. In short, the “right” number of themes is the one that supports a coherent, evidence-based narrative with enough depth to make the findings actionable.

Cornell Notes

Qualitative thematic analysis doesn’t have a fixed “correct” number of themes. Practical experience suggests many studies end up with roughly two to eight (sometimes ten) themes, while a common average for a typical write-up is about three to six themes—consistent with Braun and Clarke’s guidance for detailed reporting. The key isn’t publication length; it’s whether the themes let researchers tell a coherent story from the data with sufficient depth. The same dataset can support different thematic frameworks (e.g., one theme focused on recommendations versus multiple themes separating challenges, facilitators, and suggestions). Ultimately, theme count should be driven by research questions, narrative structure, and the usefulness of the implications for practice or policy.

Why does the transcript reject a single universal “number of themes” for qualitative studies?

It argues that the right theme count depends on how researchers need to structure a coherent narrative from the data. Even when the dataset is the same, researchers may emphasize different aspects—recommendations, challenges, facilitators—based on how they want the findings to serve the study’s purpose and implications. That narrative goal determines how many distinct building blocks (themes) are necessary to communicate the evidence with adequate depth.

What practical range does Dr Kriukow report generating in real qualitative work?

From ongoing work as a data analyst, Dr Kriukow says she typically works on at least two studies at a time, yielding a sample of studies to estimate an average. She reports generating anywhere between two and eight, sometimes ten, themes per study. When narrowing to a likely average for a study write-up, she suggests roughly three to six themes.

How does Braun and Clarke’s guidance fit into this discussion—and what caution is added?

The transcript notes that Braun and Clarke describe an average of about two to six themes for a 10,000-word report or journal article, reasoning that more themes can be difficult to discuss in sufficient depth. But it warns against treating publication type or length as the main criterion. A larger study can still justify more themes; the selection should be about what best supports the story and depth, not about hitting a numeric target.

What is the “story” criterion for deciding theme count?

Qualitative reporting is framed as telling a story of the data. Themes function as building blocks that convey the message. Because researchers differ in preferences for how to tell that story—while still staying grounded in the data—theme count can shift depending on whether the narrative centers on recommendations, struggles, facilitators, or definitions of good practice.

How can the same dataset lead to different theme numbers in the social worker example?

The transcript describes two thematic frameworks built from the same interview material. One approach uses one overarching theme—recommendations for improving practice—with subthemes like more training, more staff, and cultural/linguistic awareness training. Another approach uses three themes: challenges (e.g., lack of staff, heavy workload, limited cultural/language understanding), good practices/facilitators (e.g., adequate staffing and realistic workload planning), and suggestions/recommendations. Both frameworks aim at practice improvement, but they organize the evidence differently.

What role do implications and usefulness play in deciding whether to add or merge themes?

Theme decisions should be tied to whether the findings answer the research question and produce actionable implications for practice or policy. The transcript even suggests adding a theme not explicitly asked for—such as participants’ beliefs about what counts as good practice—if it helps connect challenges to recommendations. That relevance can justify additional themes when they strengthen the interpretive and practical payoff.

Review Questions

If a study produces 20 themes, what specific narrative or depth problem might that create, and how could theme merging or restructuring address it?
In the social worker scenario, what changes in theme structure when the analysis shifts from “recommendations-only” to “challenges plus facilitators plus recommendations”?
Why does publication length alone fail as a criterion for theme count, according to the transcript?

Key Points

1
A fixed universal number of themes doesn’t exist; theme count should support a coherent, evidence-based narrative with enough depth.
2
Practical experience suggests roughly two to eight (sometimes ten) themes per study, with a common average around three to six themes for typical write-ups.
3
Braun and Clarke’s two-to-six theme guidance for ~10,000-word articles is about depth and detail, not a strict rule to follow regardless of study design.
4
Publication length/type should not be the main criterion; the dataset and the intended story structure matter more.
5
Theme selection should be driven by research purpose and implications—whether findings will be useful for practice or policy.
6
The same dataset can yield different theme counts depending on whether the analysis centers recommendations, challenges, facilitators, or definitions of good practice.
7
When theme counts explode (e.g., 20–30), the task becomes reducing and reorganizing themes so the final account remains readable and persuasive.

Highlights

The “right” number of themes is the one that lets researchers tell a detailed, coherent story from the data—not a number derived from word count alone.

Dr Kriukow reports typically generating two to eight (sometimes ten) themes per study, with three to six as a common average for standard write-ups.

A single dataset can support either one recommendations-focused theme or a three-theme structure separating challenges, facilitators, and suggestions.

Braun and Clarke’s guidance is framed as a depth constraint: too many themes can make it hard to discuss each with sufficient specificity.

Theme decisions should be judged by usefulness—how well the analysis answers the research question and informs practice or policy.

Topics

Theme Count
Thematic Analysis
Braun and Clarke
Qualitative Coding
Research Implications

Mentioned

Braun Clarke
Dr Kriukow