Interpretive phenomenological analysis (IPA) is a qualitative methodology for exploring how individuals make sense of significant personal experiences. Developed by Jonathan Smith in 1996, IPA uses small purposive samples (typically 4-10 participants), semi-structured interviews, and a close, iterative analysis of each individual case before looking for patterns across cases. The "double hermeneutic" at its heart — the researcher interpreting a participant who is interpreting their own experience — distinguishes it from descriptive methods that aim to simply report what people said.
IPA is widely used in psychology, nursing, health sciences, education, and social work. Its focus on meaning, identity, and the texture of lived experience makes it well suited to questions that quantitative methods cannot answer: what is it like to receive a chronic illness diagnosis? How do early-career teachers make sense of professional identity? What does the experience of caring for a parent with dementia actually involve, day to day?
This guide covers IPA's philosophical foundations, its 6-step analytic process, how to design an IPA study, and when IPA is the right choice for your research.
What are the philosophical foundations of IPA?
IPA draws on three intersecting traditions:
Phenomenology — originating with Edmund Husserl and developed by Martin Heidegger and Maurice Merleau-Ponty — aims to describe the structure of experience as it is lived, rather than as theoretical constructs would predict it. Husserl's notion of "bracketing" (setting aside assumptions to attend to the phenomenon itself) runs through IPA, though IPA rejects the idea that full bracketing is possible.
Hermeneutics is the theory of interpretation. Heidegger's hermeneutic phenomenology held that interpretation is inescapable — we always bring our prior understanding to any experience. IPA applies this through the double hermeneutic: the participant interprets their experience; the researcher then interprets the participant's interpretation. This layered process is not a problem to overcome but the source of IPA's analytical depth.
Symbolic interactionism contributes IPA's interest in how people understand and represent themselves and their experiences to others. It positions meaning as socially situated and evolving.
The result is a methodology that is neither purely descriptive (just reporting what people said) nor purely theoretical (applying a framework to data). IPA analyses are always interpretive — the researcher's perspective actively shapes the findings.
When should you use IPA?
IPA is well suited to a specific type of research question. The clearest test: does your question ask about the quality and meaning of an individual's experience of something significant?
Good IPA research questions:
- "How do people living with type 2 diabetes make sense of changes to their identity?"
- "What is the experience of returning to work after a mental health crisis?"
- "How do first-generation university students understand their place in academic culture?"
These questions share three features: they focus on a specific experience, they ask about meaning and sense-making (not just what happened), and they are best explored through depth rather than breadth.
IPA is less well suited to:
- Research questions that require large-N generalisation
- Descriptive questions that could be answered by thematic analysis without the interpretive depth
- Studies where the goal is to test a hypothesis or evaluate an intervention
- Topics where participants' accounts are unlikely to vary meaningfully (where one or two cases would suffice)
If your research question is about patterns across a population rather than depth within individual cases, thematic analysis or grounded theory is likely more appropriate. See the comparison table below for a fuller picture.
How does IPA compare to thematic analysis and grounded theory?
| IPA | Reflexive thematic analysis | Grounded theory | |
|---|---|---|---|
| Primary goal | Understand lived experience in depth for each individual | Identify patterns of meaning across a dataset | Generate theory about a process or phenomenon |
| Sample size | Small and purposive (4-10 typically) | More flexible; typically 15-30+ | Driven by theoretical saturation; 15-60+ |
| Unit of analysis | The individual case, before cross-case analysis | The dataset as a whole | Evolving theoretical categories |
| Data collection | Semi-structured interviews | Interviews, documents, focus groups | Anything; guided by theoretical sampling |
| Iterative data collection? | No — data collected before analysis | No — data collected before analysis | Yes — theoretical sampling requires ongoing collection |
| Epistemological stance | Interpretive, constructionist | Broadly constructionist (reflexive TA) | Post-positivist to constructionist depending on variant |
| Key output | Themes capturing shared experiential meaning, with individual variation noted | Themes as patterns of meaning across participants | A substantive theory of how/why something happens |
What is the 6-step IPA analytic process?
IPA analysis is conducted case by case — you analyse your first participant fully before moving to the second. This preserves the idiographic commitment of the methodology: the individual is the unit of analysis, not the dataset.
Step 1: Read and re-read. Begin with your first transcript. Read it multiple times, immersing yourself in the participant's account. Note anything interesting: observations, questions, resonances, tentative interpretations. Nothing is too small.
Step 2: Initial noting. Annotate the transcript with three types of comment: descriptive notes (what the participant is doing or talking about), linguistic notes (how they say it — metaphors, hedges, emphases), and conceptual notes (your emerging interpretations, connections to theory). This is the most time-intensive step and should not be rushed.
Step 3: Develop experiential themes. Move from the annotated transcript to a set of experiential themes — phrases that capture the essence of each key finding in the data. These are more abstract than your initial notes but still grounded in the participant's account. You might generate 15-30 themes for a single case at this stage.
Step 4: Search for connections. Examine your experiential themes and look for how they cluster and connect. Do some themes have a hierarchical relationship? Do some seem to be conditions for others? Begin to map the relationships.
Step 5: Move to the next case. Repeat steps 1-4 for each participant independently. Resist the temptation to pattern-match to your first case — each participant's account deserves its own analysis.
Step 6: Look for patterns across cases. Only after completing individual case analyses do you look across all cases for shared experiential themes. What appears in multiple cases? Note both the patterns and the individual variation — where one participant's experience departs from the others is often analytically important.
How small is a good IPA sample?
IPA is deliberately idiographic — its sample sizes are small because depth, not breadth, is the goal. Published IPA studies typically use 4-10 participants, with 6 being common for a doctoral study. Smith and colleagues suggest that 3 is a reasonable minimum for a journal article if the analysis is genuinely deep.
The key principle is homogeneity: participants should share the defining feature of the experience under study but not be identical in other ways. A study of new nurse graduates' professional identity experiences would recruit people who are new nurse graduates — not a convenience sample of anyone who has started a job.
See our guide on qualitative research sample size for a broader discussion of how sample size decisions relate to research questions.
What does a good IPA write-up look like?
An IPA results section presents 3-5 superordinate (main) themes, each with 2-3 subordinate themes, illustrated by verbatim quotes from multiple participants. The writing moves between participants' words and the researcher's interpretation — the interpretive layer is what distinguishes IPA from case study reporting.
Each theme should be named analytically, not just descriptively. "Negotiating the boundaries of illness" is an IPA theme. "Participants mentioned boundaries" is not.
Quote presentation includes a pseudonym or case identifier, allowing the reader to track how different individuals relate to the theme. Where a finding holds for most but not all participants, this variation should be acknowledged rather than smoothed over.
How does AI-assisted analysis fit with IPA?
IPA's case-by-case process is time-intensive, and for small datasets (6-8 participants) this is as it should be — the depth is the point. Where AI tools add value is in managing the materials: keeping your annotated transcripts, notes, and emerging themes organised in a way that supports rather than shortcuts the analysis.
Skimle's document view and analyst notes features are designed for close annotation of individual texts. You can work through each case systematically, adding notes and codes directly to the transcript, and then use the categories view to compare how themes emerge across cases once individual analysis is complete.
The interpretive work in IPA is non-delegatable. But the mechanics of organising, annotating, and comparing 8 densely coded transcripts become much more manageable with structured tooling.
For academic researchers across psychology and health sciences, the academic researchers use-case page outlines how Skimle fits into rigorous qualitative workflows.
Frequently asked questions
How is IPA different from phenomenology?
IPA is a specific qualitative methodology that draws on phenomenological philosophy but is not phenomenology itself. Descriptive phenomenology (following Giorgi's method) aims to bracket the researcher's assumptions and produce a pure description of the essence of an experience. IPA is interpretive — it incorporates the researcher's interpretation and does not claim to access pure experience. This is the "interpretive" in IPA.
Can IPA be used for focus group data?
IPA was designed for individual interviews, and its logic — detailed case analysis before cross-case comparison — fits individual accounts. Focus groups can be used if the research question specifically concerns how people make sense of experience in a social context. However, the group dynamics complicate the individual analysis that is central to IPA. Most IPA practitioners recommend individual interviews.
How do I know when I have enough data in IPA?
Unlike grounded theory, IPA does not use saturation as a stopping criterion. You recruit a purposive sample defined by your research question, complete the analysis, and present findings from that sample. If your analysis reveals very high inter-participant variation that makes cross-case analysis difficult, your sample may be too heterogeneous rather than too small.
What citation should I use for IPA?
The foundational citation is: Smith, J. A. (1996). Beyond the divide between cognition and discourse: Using interpretative phenomenological analysis in health psychology. Psychology and Health, 11, 261-271. For a methodological reference, use the 2009 book: Smith, J. A., Flowers, P., & Larkin, M. (2009). Interpretative Phenomenological Analysis: Theory, Method and Research. SAGE.
Is IPA appropriate for online qualitative data (social media, forums)?
Yes, with care. IPA has been applied to naturally occurring texts such as online forum posts, diaries, and autobiographical accounts where the author is reflecting on a significant personal experience. The key requirement is that the material provides sufficient depth and personal reflection to support the double hermeneutic. Brief social media posts are unlikely to work; longer reflective posts or threads may.
Ready to manage close, case-by-case qualitative annotation without losing your notes across transcripts? Try Skimle for free and see how structured annotation and analysis tooling supports deep interpretive work.
Related reading:
- The complete guide to thematic analysis
- Grounded theory methodology: a practical guide
- Qualitative research sample size: how many participants do you need?
About the authors
Henri Schildt is a Professor of Strategy at Aalto University School of Business and co-founder of Skimle. He has published over a dozen peer-reviewed articles using qualitative methods, including work in Academy of Management Journal, Organisation Science, and Strategic Management Journal. His research focuses on organisational strategy, innovation, and qualitative methodology. Google Scholar profile
Olli Salo is a former Partner at McKinsey & Company where he spent 18 years helping clients understand the markets and themselves, develop winning strategies and improve their operating models. He has done over 1000 client interviews and published over 10 articles on McKinsey.com and beyond. LinkedIn profile
Sources
- Beyond the divide between cognition and discourse — Smith (1996), Psychology and Health
- Interpretative Phenomenological Analysis: Theory, Method and Research — Smith, Flowers & Larkin (2009), SAGE
- Interpretative phenomenological analysis as a useful methodology for research on the lived experience of pain — PMC