Interpretative Phenomenological Analysis (IPA): 7 Steps
Qualitati Research Team · 2026-09-29 · 11 min read
Short answer: Interpretative phenomenological analysis (IPA) is a qualitative method for studying how particular people make sense of a significant lived experience. It combines phenomenology, hermeneutics, and a case-by-case (idiographic) focus. You analyse each transcript in depth, build personal experiential themes for each participant, and only then look across cases for group experiential themes.
Last updated: September 29, 2026
Interpretative phenomenological analysis is one of the most widely used qualitative methods in psychology, health, education, and management research, and one of the most often done badly. Many theses call their analysis "IPA" but follow thematic analysis steps with a phenomenological label on top. This guide explains what IPA is, when it fits, how many participants it needs, and the seven analytic steps in the current (2022) terminology, with a short worked example and a quality checklist you can use before submission.
Key takeaways
- IPA asks how a specific person makes sense of a specific experience. It is idiographic: each case is analysed fully before any cross-case comparison.
- The 2022 second edition of the core textbook renamed the building blocks: experiential statements (formerly emergent themes), personal experiential themes (PETs), and group experiential themes (GETs).
- Samples are small and homogeneous. Three participants is a common default for a master's project; doctoral studies typically use more, but depth matters more than number.
- High-quality IPA tells an unfolding narrative, stays close to participants' words, and reports divergence as well as convergence.
- AI can speed up transcription and organisation, but the interpretative layer of IPA (the researcher making sense of the participant making sense) cannot be delegated.
What is interpretative phenomenological analysis?
Interpretative phenomenological analysis is a qualitative approach developed by Jonathan A. Smith in health psychology from the mid-1990s. It is now set out in Interpretative Phenomenological Analysis: Theory, Method and Research by Smith, Flowers, and Larkin (SAGE, second edition 2022) and in the shorter Essentials of Interpretative Phenomenological Analysis by Smith and Nizza (APA, 2022).
IPA rests on three theoretical commitments:
- Phenomenology. The focus is on experience as it is lived and understood by the person, not on causes or frequencies.
- Hermeneutics. Analysis is interpretation. IPA describes a double hermeneutic: the researcher is trying to make sense of the participant trying to make sense of what is happening to them.
- Idiography. The particular case comes first. You understand one person's experience in detail before claiming anything about a group.
When IPA fits (and when it does not)
IPA fits research questions about the meaning of a significant, often life-changing, experience for a defined group, for example "How do first-generation PhD students make sense of leaving academia?" or "What is the experience of returning to work after a cancer diagnosis?" It is a poor fit when you want to explain a social process (use grounded theory), map patterns across a large dataset (use thematic analysis), study how stories are structured (use narrative analysis), or evaluate a program. Our guide to qualitative research questions by design shows how to phrase a phenomenological question.
IPA vs thematic analysis vs grounded theory vs narrative analysis
The most common reviewer complaint about IPA studies is that the analysis is indistinguishable from thematic analysis. This table shows where the methods actually differ.
| Method | Core question | Unit of analysis | Typical sample | Main output |
|---|
| IPA | How does this person make sense of this experience? | The individual case, then the group | Small, homogeneous | Personal and group experiential themes, with divergence |
| Reflexive thematic analysis | What patterns of meaning run across the dataset? | The dataset | Flexible, often larger | Themes across participants |
| Grounded theory | How does a social process work? | Incidents and categories | Theoretical sampling until saturation | An explanatory theory or model |
| Narrative analysis | How do people construct and tell their stories? | The story as a whole | Small to moderate | Story structures, plots, positionings |
For the other methods, see thematic analysis vs grounded theory and our narrative analysis guide.
How many participants does an IPA study need?
IPA uses small, purposive, fairly homogeneous samples, because each case must be analysed in depth. Smith, Flowers, and Larkin suggest three participants as a reasonable default for student projects, and a single case can be defensible when it is rich enough. Professional doctorates and PhDs usually include more interviews, but the textbook authors are clear that more cases are not automatically better: a larger sample can thin the idiographic depth that makes IPA worth doing.
Homogeneity matters because IPA looks for convergence and divergence within a group that shares the experience in question. If your participants differ on too many dimensions at once, it becomes hard to say what varies and why. See our overview of qualitative sampling strategies and why data saturation is a poor stopping rule for idiographic work.
How to do IPA: 7 steps (2022 terminology)
Data usually come from semi-structured interviews that invite detailed first-person accounts. Transcribe verbatim, then work through the steps below one case at a time.
- Read and re-read the first transcript. Listen to the audio while reading. Note your own first impressions separately, so they do not quietly shape the analysis.
- Make exploratory notes. Annotate the transcript line by line with three kinds of comment: descriptive (what the participant says), linguistic (how they say it: metaphors, pauses, repetition, pronouns), and conceptual (what it might mean, questions you have).
- Construct experiential statements. Turn your notes into short statements that capture what the experience means for the participant. Each should stay close to the text but carry an interpretation.
- Search for connections across experiential statements. Cluster statements that belong together. Some will be dropped; that is expected.
- Name the personal experiential themes (PETs). Give each cluster a title, organise them in a table with transcript line references and key quotes, and write a short account of that participant.
- Repeat for each remaining case. Treat each new transcript on its own terms. Try to bracket the themes from earlier cases rather than looking for them.
- Develop group experiential themes (GETs). Compare PETs across cases. Identify what is shared, how it is shared differently, and what is unique. The final table shows which participants contribute to each GET.
These steps follow the second-edition structure. Older papers use "emergent themes", "superordinate themes", and "master themes" for roughly the same ideas; if your supervisor learned IPA from the 2009 edition, agree on terminology early.
A short worked example
The excerpt below is illustrative, written for this guide, not real participant data. Imagine a study of early-career academics who left university jobs for industry research.
"I kept saying I was just trying it out. For a year I didn't tell anyone at my old department. It felt like I'd failed a test nobody else had to take."
- Descriptive note: Delayed telling former colleagues; framed the move as temporary.
- Linguistic note: "Just trying it out" minimises; "a test nobody else had to take" signals isolation and unfairness.
- Conceptual note: Leaving is lived as a verdict on the self, not a career choice.
- Experiential statement: "Keeping the departure provisional to protect an academic identity."
Across the transcript, this statement might cluster with others into a PET such as "Leaving as a private failure". At the group stage, you might find that three participants share this sense of failure while a fourth describes leaving as relief. That divergence is not noise; in IPA it is part of the finding.
The IPA Quality Checklist
This Qualitati editorial checklist draws on Smith's criteria for evaluating IPA studies (Health Psychology Review, 2011) and the four markers of high-quality IPA described by Nizza, Farr, and Smith (Qualitative Research in Psychology, 2021). Score each item yes or no before you submit.
| Check | What good looks like |
|---|
| Phenomenological question | The research question asks about the meaning of an experience, not causes or frequencies. |
| Homogeneous sample | Participants share the experience and the context that matters; the inclusion rationale is stated. |
| Case-by-case analysis | Each transcript was analysed to PETs before cross-case work began. |
| Unfolding narrative | The findings read as an argued account, not a list of themes with quotes attached. |
| Vigorous experiential account | Themes describe what the experience is like, not just topics participants mentioned. |
| Close analytic reading | Quotes are interpreted, including language, metaphor, and hesitation, not merely presented. |
| Convergence and divergence | Each GET shows how participants share it differently, and where someone does not. |
| Prevalence reported | A table shows which participants contribute to each theme. |
| Reflexivity | The researcher's position and its likely influence are documented. See our guide to reflexivity in qualitative research. |
Can AI help with interpretative phenomenological analysis?
Partly. Recent studies show both the promise and the limits. Martínez-Pernía and colleagues used GPT-4 to support a phenomenological analysis of interviews with 28 adults and reported that the AI-assisted analysis took about 20 hours against about 70 for the human-coded analysis of the same data, but also that it missed some bodily sensations the human analysis caught and did not fully match more complex experiential structures (Frontiers in Psychology, May 2025). Their study used descriptive and micro-phenomenological methods rather than IPA, but the lessons transfer. Separately, Ashwin, Chhabra, and Rao found that large language models can introduce systematic bias when annotating qualitative interview data (Sociological Methods & Research, 2025).
A defensible division of labour for IPA:
- Reasonable to automate: transcription, anonymisation, organising quotes by transcript line, drafting a PET table from statements you wrote.
- Use with care: AI-suggested clusters of experiential statements, treated as a prompt to check against your own reading.
- Keep human: exploratory notes, experiential statements, and every interpretative claim. The double hermeneutic requires a researcher who is the one making sense.
Where Qualitati fits
Qualitati is a European qualitative research platform for universities and research companies, with GDPR and data privacy as central priorities. It does not perform IPA for you, and it should not. It supports the parts of an IPA project around the interpretation:
- Data collection. Run semi-structured interviews yourself with Active Listener mode, which gives real-time prompts and section tracking, or run AI-moderated interviews in text or voice for pilot or exploratory phases.
- Transcripts in 10 languages. Useful for cross-national phenomenological studies, with transcripts you can review line by line.
- Case-level organisation. QDA Workspace supports inductive coding and codebook building, which you can use to tag your own experiential statements per transcript and compare them across cases.
The free tier includes 30 monthly credits with no credit card required. See pricing for current rates, or compare approaches in our NVivo alternatives guide.
Limitations and methodological concerns
- Limited generalisation. IPA offers theoretical transferability, not statistical generalisation. Say so plainly in your limitations section.
- Dependence on articulate accounts. IPA works best when participants can reflect on and describe their experience. Adapt interviews for participants who find this hard, or consider another method.
- Researcher influence is the point and the risk. Interpretation is central, so a transparent audit trail and reflexive notes are essential.
- Terminology drift. Mixing 2009 and 2022 terms in one thesis confuses examiners. Pick one and cite it.
Who this is for, and when not to use this approach
This guide is for PhD and master's students, supervisors, and researchers in psychology, health, education, and management who are planning or writing up an IPA study. Do not choose IPA when your goal is to explain a process, test a hypothesis, compare many heterogeneous groups, or summarise opinions across a large sample. Reflexive thematic analysis or grounded theory will usually serve those aims better.
FAQ
What is the difference between IPA and phenomenology?
Phenomenology is a philosophical tradition and a family of research methods. IPA is one specific method within it that adds an explicit interpretative (hermeneutic) stance and a case-by-case idiographic focus. Descriptive phenomenology, by contrast, aims to describe the essential structure of an experience.
How many participants do you need for IPA?
Small, homogeneous samples are the norm. Three participants is a common default for master's projects, and doctoral studies usually include more. Depth of analysis per case matters more than the total number.
What are personal and group experiential themes?
Personal experiential themes (PETs) summarise one participant's experience. Group experiential themes (GETs) are developed by comparing PETs across participants and show both shared and divergent patterns. The terms come from the 2022 second edition of Smith, Flowers, and Larkin.
Is IPA the same as thematic analysis?
No. Both produce themes, but IPA analyses each case in depth before comparing, focuses on the meaning of a lived experience, and reports divergence between participants. Thematic analysis works across the whole dataset from the start.
Can I use ChatGPT or other AI tools for IPA?
For transcription and organisation, yes, with ethics approval and data protection in place. For experiential statements and interpretation, AI output should be treated as a prompt at most, because the method requires the researcher's own sense-making.
Bottom line
Interpretative phenomenological analysis is the right choice when you want to understand how a small, specific group makes sense of a significant experience. Analyse case by case, use the 2022 terminology consistently, report divergence, and let the interpretation stay yours. When your interview guide is ready, start free with 30 credits to collect and transcribe interviews in Qualitati and organise your experiential statements across cases.
Human-review note: the worked example and the IPA Quality Checklist are editorial tools from the Qualitati Research Team, not a validated instrument. Confirm sample size and reporting conventions with your supervisor and the IPA texts cited above.