Short answer: Narrative analysis is a qualitative method that treats whole stories, not coded fragments, as the unit of analysis. It asks what people tell, how they structure the telling, and for whom. Catherine Kohler Riessman describes four types: thematic, structural, dialogic/performance, and visual. Use it when sequence, turning points, and identity matter more than cross-case patterns.
Last updated: September 26, 2026
Narrative analysis is the method researchers reach for when participants do not just report facts but tell stories: how they left academia, how they lived with a diagnosis, how they migrated. This guide defines narrative analysis in qualitative research, explains the four types, compares it with thematic analysis, and walks through a six-step process with a short worked example you can adapt for a thesis or project.
Key takeaways
- Narrative analysis keeps each story intact and analyzes its content, structure, and context together.
- Riessman's four types (thematic, structural, dialogic/performance, visual) answer different questions. Pick one as your primary lens before coding.
- Narrative analysis differs from thematic analysis in its unit of analysis: the case and its sequence, not patterns across cases.
- Labov's six-part model (abstract, orientation, complicating action, evaluation, resolution, coda) is the most widely taught tool for structural analysis.
- AI can help transcribe, segment, and summarize stories, but interpretation of meaning, positioning, and audience stays with the researcher.
What is narrative analysis?
Narrative analysis is a family of qualitative approaches that take the story as the unit of analysis. Instead of breaking an interview into many short codes, the analyst reads each account as a whole and asks how events are selected, ordered, and evaluated, and what the telling does for the narrator. Stories are treated as meaningful constructions, not transparent records of what happened.
Two terms are often confused. Narrative inquiry, associated with D. Jean Clandinin and F. Michael Connelly, usually names a whole research methodology, from design to writing. Narrative analysis names the analytic methods applied to storied data. Catherine Kohler Riessman's Narrative Methods for the Human Sciences (SAGE, 2008) is the standard reference for the second.
When to use narrative analysis
- Your research question asks how people make sense of change, disruption, or transition over time.
- Identity is central: how participants present themselves, and to whom.
- Your data are naturally storied: life-story or biographical interviews, oral histories, diaries, long open-ended accounts.
- You have a small sample and want depth per case rather than breadth across many.
If you need to write the question first, see our qualitative research questions examples, which include narrative inquiry questions.
The four types of narrative analysis
Riessman organizes the field into four analytic approaches. The table summarizes what each focuses on and the question it answers.
| Type | Focus | Core question | Typical data |
|---|---|---|---|
| Thematic | What is told | What themes recur within and across whole stories? | Interviews, written accounts, archives |
| Structural | How it is told | How is the story organized, and where does the narrator evaluate events? | Verbatim transcripts with pauses and repairs |
| Dialogic/performance | To whom and why | How is the story co-produced with an audience, and what identity does it perform? | Interviews including interviewer turns, conversations, group talk |
| Visual | Images as narrative | How do photographs, drawings, or video tell or support a story? | Photo-elicitation, drawings, video diaries |
Thematic narrative analysis
Thematic narrative analysis attends to content, but unlike standard coding it keeps each story whole and reads themes in their place within the plot. You might find that "being misrecognized" appears in many career-change stories, yet it sits at the start of some accounts and at the turning point of others. That placement is part of the finding.
Structural narrative analysis
Structural analysis examines form. The most common tool is William Labov and Joshua Waletzky's model of oral narrative, reprinted in the Journal of Narrative and Life History (1997). It identifies six elements:
- Abstract: what the story is about.
- Orientation: who, when, where.
- Complicating action: what happened next.
- Evaluation: why it matters, in the narrator's view.
- Resolution: how it turned out.
- Coda: the return to the present.
Evaluation clauses are often the most revealing, because that is where narrators tell the listener how to read the events.
Dialogic/performance analysis
Dialogic analysis treats storytelling as a social act. It includes the interviewer's questions in the transcript and asks how the audience, setting, and wider discourses shape what can be said. A participant who knows the interviewer is also an academic may tell a departure story differently than to a friend in industry.
Visual narrative analysis
Visual analysis extends the approach to images that participants make, choose, or are shown. It considers the image itself, how it was produced, and how people talk about it, and is common in photo-elicitation and arts-based studies.
Narrative analysis vs thematic analysis
Both are interpretive, and they can be combined, but they answer different questions. Thematic analysis, as described by Braun and Clarke (Qualitative Research in Psychology, 2006), identifies patterns of meaning across a dataset. Narrative analysis preserves the sequence and context of each account.
| Criterion | Narrative analysis | Thematic analysis |
|---|---|---|
| Unit of analysis | The whole story or case | Coded segments across cases |
| Treatment of sequence | Central: order and turning points are data | Usually set aside |
| Typical sample | Small; deep per case | Small to large; breadth across cases |
| Main output | Case portraits, story types, structural patterns | A set of themes with supporting extracts |
| Best for | Identity, change, meaning-making over time | Shared experiences, views, and practices |
For how thematic analysis compares with grounded theory, see thematic analysis vs grounded theory; for a full thematic walkthrough, see our thematic analysis guide.
Two neighbouring methods are easy to confuse with narrative analysis. Discourse analysis asks how language constructs social reality rather than how one person's story is built; interpretative phenomenological analysis (IPA) examines the lived experience of a specific event in a small, homogeneous sample.
How to do narrative analysis: six steps
- Collect storied data. Use open prompts that invite accounts ("Tell me how you came to leave your post") rather than tightly structured questions. A loose semi-structured interview with few, broad questions works well.
- Transcribe with detail. Keep pauses, repetitions, and interviewer turns, especially for structural or dialogic work.
- Identify the stories. Mark where each narrative begins and ends inside the interview. Not every answer is a story.
- Apply your primary lens. Map Labov elements, trace themes in sequence, or annotate audience and positioning, depending on your type.
- Write a case summary. For each participant, summarize the plot, the turning point, and how the narrator positions themselves.
- Compare across cases, then return to the whole. Look for story types or shared turning points, and check every claim against the complete account, not an isolated quote.
A worked example
The excerpt below is a constructed, illustrative example, not real participant data. Imagine a mid-career lecturer explaining why they left a university for an industry research role.
"So basically it's the story of how I stopped waiting. I'd been on fixed-term contracts for six years, teaching full loads. Then in the spring the permanent post went to an external candidate, and my head of department told me in the corridor. That was the moment, honestly, I thought: I have been auditioning for a job that doesn't exist. I applied to three companies that summer. Now I run studies I actually design, and I sleep better."
| Labov element | Text | Analytic note |
|---|---|---|
| Abstract | "the story of how I stopped waiting" | Frames the account as agency, not loss |
| Orientation | Six years of fixed-term contracts, full teaching loads | Establishes precarity as background |
| Complicating action | Post goes to an external candidate; news given "in the corridor" | The informal setting signals disregard |
| Evaluation | "auditioning for a job that doesn't exist" | The key reinterpretation of the past |
| Resolution | Applied to three companies; now designs own studies | Recovered professional identity |
| Coda | "I sleep better" | Returns to the present with a verdict |
A thematic reading would code "precarity" and "lack of recognition." A narrative reading adds that the narrator places the turning point in a single scene and recasts six years of work as an audition. Across several such cases, you might compare "stopped waiting" stories with "pushed out" stories and ask what distinguishes them.
The Narrative Analysis Fit Check
This Qualitati editorial checklist helps you decide whether narrative analysis suits your study. Answer yes or no.
- Does your research question involve change, transition, or identity over time?
- Will participants have room to tell extended accounts in your data collection?
- Is your sample small enough to analyze each case as a whole?
- Have you chosen one primary type (thematic, structural, dialogic/performance, or visual)?
- Will your transcripts preserve sequence and interviewer turns?
- Can you report findings as cases or story types, not only as a list of themes?
Four or more "yes" answers suggest a good fit. Fewer suggest thematic analysis may serve your question better.
Where Qualitati fits
Qualitati is a European qualitative research platform for universities and research companies, with GDPR and data privacy as central priorities. It supports the practical steps around narrative analysis, while the interpretation stays with you:
- Collecting stories. Run AI-moderated interviews in text or voice with open, story-eliciting prompts, or use Active Listener mode to get real-time prompts and section tracking while you interview in person.
- Working with transcripts. QDA Workspace supports inductive and deductive coding and codebook generation, which you can use to mark story boundaries and Labov elements.
- Cross-case comparison. ThemeLens maps codes to research questions across up to 100 transcripts with participant-anchored quotes, useful for the comparison step after case-level reading.
- Multilingual studies. Interviews and analysis run in 10 languages.
New to the platform? The tutorial walks through a first project. The free tier includes 30 monthly credits with no credit card required; see pricing.
Limitations and methodological concerns
- Interpretation is contestable. Two analysts can read one story differently. Keep an audit trail and a reflexive note; see our guide to reflexivity in qualitative research.
- Transferability is limited. Small, deep samples support theoretical insight, not population estimates.
- The interview shapes the story. Tightly structured guides, including AI-moderated ones, can cut stories short. Pilot your prompts for narrative space.
- AI is an assistant, not an analyst. Jenner and colleagues compared human, Claude, and GPT-o1 interpretations across steps of narrative analysis and concluded the models show promise as assistants (Methods in Psychology, 2025). A 2026 study of open-source models on 34 patient interviews found only partial alignment with human coders (International Journal of Qualitative Methods, 2026). Check every AI summary against the full story.
Who this is for, and when not to use it
This guide is for PhD students, academics, and research teams studying life events, careers, illness, migration, or organizational change through in-depth interviews. Do not use narrative analysis when you need to compare many short responses, such as open-ended survey answers, or when your question is about shared views rather than individual trajectories.
FAQ
What are the four types of narrative analysis?
Following Riessman (2008): thematic (what is told), structural (how it is told), dialogic/performance (to whom and for what purpose), and visual (how images tell or support stories).
What is the difference between narrative analysis and thematic analysis?
Narrative analysis keeps each story whole and treats its sequence as data. Thematic analysis identifies patterns of meaning across segments from many participants.
How many participants do you need for narrative analysis?
There is no fixed number. Narrative studies are often small because each case is analyzed in depth; justify your sample by the aim and depth of analysis rather than a quota.
What is an example of narrative analysis?
Mapping a career-change interview onto Labov's six elements to show where the narrator locates the turning point and how they evaluate it, as in the worked example above.
Can AI do narrative analysis?
AI can help with transcription, locating stories, and first-pass summaries. Published studies so far support using it as an assistant under researcher oversight, not as a replacement for interpretation.
Bottom line
Narrative analysis is the right method when the shape of a story, its turning points, and its audience matter as much as its content. Choose one of Riessman's four types, collect accounts with room to unfold, and analyze each case whole before comparing. When you are ready, start free with 30 credits to run story-eliciting interviews and organize your narrative analysis in Qualitati.
Human-review note: the worked example is constructed for illustration, and the Fit Check is an editorial tool from the Qualitati Research Team, not a validated instrument. Confirm methodological choices with your supervisor or the methods literature in your discipline.