Step 1 of 5
Overview & The Decision Framework
Qualitative research is the study of meaning, experiences, and social processes.
Research 101 Modules
Quantifying Qualitative Data
Learn how to systematically classify, code, and analyze non-numerical data (words, experiences, interviews) to draw rigorous scientific conclusions.
Module complete. It is checked off in your course progress.
Qualitative Method Selector
- Goal
- Sample
- Method
Goal
What is the primary goal of your qualitative analysis?
The most important factor in choosing a qualitative analysis method is what you are trying to achieve. Methods differ fundamentally in their epistemological assumptions, their procedures, and the kinds of claims they allow you to make (Creswell, 2018; Braun & Clarke, 2006).
Sample
What is your sample size and what type of data do you have?
Both thematic analysis and content analysis identify patterns in qualitative data, but they differ in their approach to meaning. Thematic analysis is interpretive and meaning-focused; content analysis is more systematic and can include frequency counts. Sample size also matters: content analysis handles large corpora well, while thematic analysis works best with richer but smaller datasets.
Pattern identification
Reflexive Thematic Analysis
Braun & Clarke (2006, 2019) — the most widely used qualitative analysis method
A method for identifying, analysing, and reporting patterns (themes) within qualitative data. Reflexive thematic analysis (Braun & Clarke, 2006, 2019) is theoretically flexible — it is not tied to a specific epistemological framework — making it the most accessible and widely applicable qualitative analysis approach. It prioritises the researcher's active role in constructing themes from the data.
Typical sample size
Flexible — typically 6-30 participants
Data types
Interviews, focus groups, open-ended surveys, documents
Approach
Inductive or deductive
- 1Familiarise yourself with the dataRead and re-read all transcripts or data. Take initial notes. Do not code yet. The goal is immersion — you should know your data deeply before analysis begins.
- 2Generate initial codesGo through the data systematically and label segments with short descriptive codes capturing what is meaningful or interesting. Code inclusively — capture more rather than less at this stage.
- 3Search for themesSort codes into potential themes by grouping codes that share a common central concept. Create a thematic map showing how codes and themes relate to each other.
- 4Review themesCheck candidate themes against the coded data (do the codes cohere?) and the full dataset (does the theme represent the data as a whole?). Merge, split, or discard themes as needed.
- 5Define and name themesWrite a clear definition for each theme capturing its essence and scope. Name themes with evocative, analytical labels — not just descriptive summaries of the data content.
- 6Write the reportWeave together analytic narrative and data extracts (quotes). Each theme should tell an analytical story, not just describe what participants said. Reflect on your own role in constructing the analysis.
Strengths
- Theoretically flexible — applicable across epistemologies and data types
- Accessible to novice qualitative researchers (Braun & Clarke, 2006)
- Can be used inductively (data-driven) or deductively (theory-driven)
- Produces rich, nuanced findings that capture meaning across a dataset
Limitations
- Themes are researcher-constructed — requires transparency about interpretive choices
- Does not generate theory (unlike grounded theory) or focus on individual experience (unlike IPA)
- The flexibility can be a weakness — requires strong researcher reflexivity to avoid superficial analysis
- Difficult to fully standardise across research teams
Example in practice
A researcher interviews 20 undergraduate students about their experiences of transitioning into a research lab. Using reflexive thematic analysis, they identify themes such as 'navigating uncertainty,' 'the role of mentorship in building confidence,' and 'belonging and identity in research culture.'
Sources: Braun, V. & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in Psychology, 3(2), 77–101. | Braun, V. & Clarke, V. (2019). Reflecting on reflexive thematic analysis. Qualitative Research in Sport, Exercise and Health, 11(4), 589–597.
Systematic coding
Content Analysis
Quantitative or qualitative — both count and interpret textual content
A systematic, replicable technique for compressing large amounts of text into fewer content categories based on explicit coding rules. Content analysis can be quantitative (counting the frequency of predefined categories) or qualitative (interpreting the meaning of categories). It is more structured and transparent than thematic analysis, and is particularly suited to large text corpora and comparative studies.
Typical sample size
Any — well-suited to large text corpora
Data types
Documents, media, social media, interview transcripts, policy texts
Approach
Deductive (predefined categories) or inductive
- 1Define your research question and unit of analysisDecide what you are counting or categorising — words, sentences, paragraphs, documents, or themes. Define your unit of analysis precisely before starting.
- 2Select your sample of textIdentify the corpus of text you will analyse (e.g. all interview transcripts, a sample of news articles, policy documents from a defined period). Justify your sampling strategy.
- 3Develop your coding frameworkFor deductive content analysis: derive categories from theory or prior literature. For inductive content analysis: derive categories from the data itself. Define each category with clear operational definitions and decision rules.
- 4Code the dataApply your coding framework systematically to all texts. Use at least two coders where possible to assess inter-rater reliability (calculate Cohen's kappa or percentage agreement).
- 5Calculate inter-rater reliabilityIf using multiple coders, calculate a reliability statistic (Cohen's kappa ≥0.70 is generally considered acceptable). Resolve disagreements through discussion and refine category definitions.
- 6Analyse and reportSummarise the frequency, distribution, or patterns of categories. For quantitative content analysis, apply appropriate statistical tests. For qualitative content analysis, interpret the meaning of categories and their relationships.
Strengths
- Highly systematic and transparent — coding decisions can be audited
- Can handle very large text corpora efficiently
- Inter-rater reliability provides a measure of rigour
- Compatible with mixed-methods designs — quantitative and qualitative together
Limitations
- Risk of decontextualising data by focusing on surface features
- Inter-rater reliability can be difficult to achieve for interpretive categories
- Frequency of occurrence does not necessarily indicate importance
- Predefined categories may miss unexpected or nuanced content
Example in practice
A researcher analyses 50 published papers' methods sections to categorise the types of statistical tests used in undergraduate STEM research. Using a deductive coding framework, they code each paper and calculate the frequency of each test type across disciplines.
Sources: Hsieh, H.F. & Shannon, S.E. (2005). Three approaches to qualitative content analysis. Qualitative Health Research, 15(9), 1277–1288. | Krippendorff, K. (2018). Content Analysis: An Introduction to its Methodology (4th ed.). SAGE.
Lived experience
Interpretative Phenomenological Analysis (IPA)
Smith, Flowers & Larkin (2009) — for in-depth individual experience
An approach to qualitative inquiry committed to examining how people make sense of their major life experiences. IPA has a dual interpretative focus: it asks how participants make sense of their world, and how the researcher makes sense of the participant making sense of their world (the 'double hermeneutic'). It works case-by-case, analysing each participant's account in full before moving across the group.
Typical sample size
Small purposive samples — typically 3-8 participants
Data types
In-depth interviews, personal diaries, think-aloud protocols
Approach
Inductive — data-driven, no prior theory
- 1Select a purposive, homogeneous sampleRecruit participants who all share the experience you are studying. Homogeneity (similar background, role, or experience) allows you to focus on the phenomenon rather than group differences.
- 2Conduct in-depth interviewsUse semi-structured, open-ended interview questions that invite participants to describe their experience in their own words. Interviews are typically 45-90 minutes and should be audio-recorded and transcribed verbatim.
- 3Analyse the first transcriptRead and re-read the first transcript. Make exploratory notes in the margin. Then produce more refined experiential statements — concise interpretive labels that capture meaning in the participant's account.
- 4Identify personal experiential themesCluster related experiential statements into personal experiential themes for this participant. Write a narrative summary of the participant's experience.
- 5Move to subsequent transcriptsRepeat steps 3-4 for each participant independently, bracketing your knowledge of previous participants' accounts to honour each person's unique experience.
- 6Identify group experiential themesLook across all participants' personal themes for convergences and divergences. Identify group-level themes that are present across cases, noting variations. Document which participants exemplify each theme and quote directly.
Strengths
- Produces rich, in-depth insights into individual lived experience
- Honours the uniqueness of each participant's account
- Rigorous and well-established methodology with clear procedural guidelines
- Particularly powerful for under-researched or sensitive experiences
Limitations
- Small samples limit generalisability — findings cannot be statistically representative
- Time-intensive data collection and analysis — each transcript requires several hours of detailed work
- Requires a phenomenological epistemological commitment — not appropriate for all research questions
- Demands extensive researcher reflexivity about how interpretation shapes findings
Example in practice
A researcher conducts in-depth interviews with six first-generation university students about their experience of entering a research lab for the first time. IPA reveals themes of 'feeling like an outsider,' 'learning an unwritten language,' and 'the turning point of being trusted with independent tasks.'
Sources: Smith, J.A., Flowers, P. & Larkin, M. (2009). Interpretative Phenomenological Analysis: Theory, Method and Research. SAGE. | Pietkiewicz, I. & Smith, J.A. (2014). A practical guide to using IPA. Psychological Journal, 20(1), 7–14.
Theory generation
Grounded Theory
Glaser & Strauss (1967); Charmaz (2006) — constructivist version
A systematic methodology for generating theory grounded in data. Grounded theory is used when no adequate existing theory explains a phenomenon, or when the researcher wants to develop a new conceptual model. Data collection and analysis proceed simultaneously and iteratively — new data are collected based on what the emerging analysis reveals (theoretical sampling). Analysis continues until theoretical saturation is reached (no new concepts are emerging).
Typical sample size
Determined by theoretical saturation — typically 15-30 participants
Data types
Interviews, observations, documents — collected iteratively
Approach
Inductive — theory emerges from data
- 1Begin with open-ended data collectionConduct initial interviews or observations without rigid hypotheses. Ask broad, open questions about participants' experiences and processes. Transcribe and begin analysis promptly.
- 2Open codingCode the first data in a line-by-line, granular fashion. Assign short, descriptive codes to every meaningful unit. Use gerunds (action words) to capture processes: 'managing uncertainty,' 'seeking validation.'
- 3Theoretical samplingUse your emerging analysis to decide who or what to sample next. Collect new data to develop, test, and refine emerging concepts. This distinguishes grounded theory from conventional sampling strategies.
- 4Focused codingIdentify the most frequent and analytically significant codes from open coding. Use these focused codes to categorise larger segments of data across the full dataset.
- 5Memo writingWrite analytic memos throughout the analysis — these are notes to yourself about concepts, relationships between categories, and emerging theoretical ideas. Memos are the bridge between coding and theory.
- 6Theoretical saturation and theory buildingContinue sampling and coding until no new concepts, properties, or relationships are emerging (theoretical saturation). Write up the theoretical model, showing how categories relate to a central or core category.
Strengths
- Produces original, data-grounded theory rather than merely describing findings
- Iterative data collection and analysis is methodologically rigorous
- Theoretical sampling ensures the theory is grounded in the full range of the phenomenon
- Well-suited to under-researched areas where existing frameworks are inadequate
Limitations
- Extremely time-intensive — not appropriate for short-term or resource-limited projects
- Requires a deep understanding of the methodology — easily misapplied
- True theoretical saturation can be difficult to determine and justify
- The researcher must bracket prior theoretical assumptions (especially in classical grounded theory)
Example in practice
A researcher studies the process by which undergraduate students develop a research identity. Through iterative interviews with 22 students and mentors, they develop a grounded theory of 'identity anchoring' — a model of how students move from peripheral participation to confident researcher self-concept.
Sources: Charmaz, K. (2006). Constructing Grounded Theory: A Practical Guide through Qualitative Analysis. SAGE. | Glaser, B.G. & Strauss, A.L. (1967). The Discovery of Grounded Theory. Aldine. | Creswell, J.W. (2018). Research Design (5th ed.). SAGE.