Framework Analysis: Ritchie & Spencer’s Method and the Framework Matrix

Reviewed by Prof. Dr. Alexander Holtermann · Updated 16 October 2026 · 6 min read

Framework analysis organises qualitative data in a matrix of cases (rows) and codes or themes (columns), to compare within and across cases systematically.

Key takeaways

Framework analysis was developed by Jane Ritchie and Liz Spencer at a UK social research institute in the 1980s for applied policy research and published in 1994.
Its hallmark is the framework matrix: one row per participant, one column per code or theme, and summarised data in each cell.
Ritchie and Spencer describe five stages: familiarisation, identifying a thematic framework, indexing, charting, and mapping and interpretation.
Gale et al. (2013) refined the procedure into seven stages, which are widely used in health research.
The method suits questions with predefined topics, team projects and comparisons between groups; it combines deductive and inductive elements.

What is framework analysis?

Framework analysis (also called the framework method) is a systematic approach to qualitative data analysis based on a matrix that summarises data by case and by code or theme. It was developed by Jane Ritchie and Liz Spencer at Social and Community Planning Research (today the National Centre for Social Research) in London, where researchers needed transparent, timely analyses for government and policy clients. Their 1994 chapter in *Analyzing Qualitative Data*, edited by Bryman and Burgess, is the classic reference.

Framework analysis is not tied to a particular epistemology. It is particularly popular in health services research, nursing, public health and social policy in the UK, and Braun and Clarke (2022) count it among the codebook approaches to thematic analysis. Its main advantage is that you can see both the individual case (by reading across a row) and the pattern across cases (by reading down a column).

Running example: in your study of 12 first-generation students, your university’s widening participation office wants to know which support measures help students feel they belong. Framework analysis lets you compare students with and without paid work, living at home or on campus, across predefined topics.

The five stages according to Ritchie and Spencer (1994)

The original description has five interconnected stages. They are applied systematically but, like all qualitative analysis, involve going back and forth.

Stage	What you do
1 Familiarisation	Immerse yourself in the data: read transcripts, listen to recordings, note recurring ideas
2 Identifying a thematic framework	Develop a set of key issues, concepts and themes from the research aims and the data
3 Indexing	Apply the framework systematically to all data, marking which passages relate to which theme
4 Charting	Summarise the indexed data for each case and theme and place the summaries in charts (matrices)
5 Mapping and interpretation	Use the charts to define concepts, map the range of phenomena, create typologies and find associations and explanations

The seven stages according to Gale et al. (2013)

Gale and colleagues published a step-by-step guide in *BMC Medical Research Methodology* that is now one of the most cited descriptions of the method, especially in health research. It makes the steps more explicit for multidisciplinary teams and students.

Stage	Description	Belonging study
1 Transcription	Verbatim transcripts as the basis	12 anonymised transcripts
2 Familiarisation with the interview	Read, listen, note impressions	Case summary and margin notes for each participant
3 Coding	Code the first transcripts openly or with predefined codes	3 transcripts coded line by line
4 Developing a working analytical framework	Group codes into categories, agree a set of codes with definitions	5 categories, 24 codes
5 Applying the analytical framework	Index all transcripts with the framework	All 12 transcripts coded
6 Charting data into the framework matrix	Summarise data per participant and code in matrix cells, with references to quotes	Matrix 12 × 24
7 Interpreting the data	Identify characteristics of and differences between data, map connections, develop explanations	Typology of belonging pathways

The framework matrix: how to chart data

The matrix is the heart of the method. Each row is a case (usually a participant), each column a code or theme, and each cell contains a short summary of what that participant said about that theme, often with a reference to illustrative quotes (for example transcript line numbers). Charting is an analytic step: you reduce data while keeping the participant’s language and meaning, which means you constantly decide what is essential.

Good cells are concise but specific. Avoid pasting whole quotes into cells (the matrix becomes unreadable) and avoid summaries so abstract that the original meaning is lost.

Participant	Paid work	Help-seeking	Peer support	Family
P01 (lives at home)	15 h/week retail; misses evening events (l. 112)	Avoids office hours, “didn’t want to look stupid” (l. 40)	Relies on one school friend at same university	Parents proud but “don’t get it” (l. 201)
P04 (campus)	None	Saw help-seeking as failure until term 2 (l. 55–70)	Flatmates key; study group from week 6	Weekly calls, pressure to succeed
P07 (commuter)	Two jobs, 25 h/week (l. 18)	Asked by email only	Study group after lecturer suggested it (l. 140)	Siblings’ role model

When is framework analysis a good choice?

Framework analysis works best when your research has clear, predefined questions (for example from a commissioned evaluation), when you want to compare groups or cases systematically, when several researchers analyse the data together and when you need a highly transparent audit trail. It is less suitable for highly heterogeneous data, for exploring lived experience in depth (IPA is better) or for developing theory of a process (grounded theory is better).

Suitable: evaluation of a support programme, patient or student experience studies, policy research
Suitable: mixed deductive-inductive designs where the interview guide defines main topics
Less suitable: very small samples where each case should be analysed in depth on its own terms
Less suitable: highly interpretive or discourse-oriented questions

Tools, write-up and supervisor expectations

You can chart the matrix in a spreadsheet, in a table in your thesis document or in QDA software with matrix functions. Our free coding tool at /codieren lets you build the analytical framework as a category system with definitions, anchor examples and coding rules, code your transcripts and export a code matrix and the coded segments; you then write the summaries for each cell.

In your methods chapter, cite Ritchie and Spencer (1994) and/or Gale et al. (2013), state which stages you followed, how the framework was developed (deductive sources and inductive additions) and how many researchers were involved. Put the framework and an excerpt of the matrix in the appendix. In the findings, present themes with quotes and highlight comparisons between cases or groups. Some departments, particularly in nursing and public health, expect a specific version of the method; follow your university’s guidelines (you can upload them to Professor Digital).

Frequently asked questions

What are the stages of framework analysis?

Ritchie and Spencer (1994): familiarisation, identifying a thematic framework, indexing, charting, mapping and interpretation. Gale et al. (2013) describe seven: transcription, familiarisation, coding, developing a working analytical framework, applying it, charting into the matrix, and interpreting.

What is the difference between framework analysis and thematic analysis?

Framework analysis uses a structured framework and a case-by-theme matrix, which makes comparisons explicit. Reflexive thematic analysis develops themes more organically without a matrix. Braun and Clarke classify framework analysis as a codebook approach to TA.

Is framework analysis inductive or deductive?

Both. The framework often starts from the research questions and interview guide and is extended with themes from the data.

Can I do framework analysis in a spreadsheet?

Yes. A spreadsheet with participants in rows and codes in columns is a common and practical way to chart a framework matrix for a dissertation.

Who developed framework analysis?

Jane Ritchie and Liz Spencer at Social and Community Planning Research (now the National Centre for Social Research) in the UK, published in 1994.

Can I do framework analysis in NVivo or Excel?

Yes. NVivo has framework matrices that link cases to themes with summaries; MAXQDA offers similar summary tables. A spreadsheet with one row per participant and one column per theme works just as well for smaller projects.

Sources and further reading

Ritchie, J., & Spencer, L. (1994). Qualitative data analysis for applied policy research. In A. Bryman & R. G. Burgess (Eds.), Analyzing Qualitative Data. Routledge.
Gale, N. K., Heath, G., Cameron, E., Rashid, S., & Redwood, S. (2013). Using the framework method for the analysis of qualitative data in multi-disciplinary health research. BMC Medical Research Methodology, 13, 117.
Braun, V., & Clarke, V. (2022). Thematic Analysis: A Practical Guide. Sage.

Your university’s examination regulations and style guide always take precedence.