Framework Analysis

Framework analysis is a systematic approach to qualitative data analysis that uses a structured matrix (a "framework") to organize and interpret data according to key themes and cases. Developed by Jane Ritchie and Liz Spencer in the 1980s for applied social policy research, it is particularly well suited to research with specific questions, predefined issues, or policy relevant objectives.

Qualitative research can produce large volumes of unstructured data (interview transcripts, field notes, focus group recordings) that are difficult to analyze in a rigorous and transparent way. Framework analysis provides a clear, systematic method for managing this complexity. Unlike some qualitative approaches that are designed for theory generation (such as grounded theory), framework analysis is designed for applied research where the questions are known in advance and the goal is to describe and interpret patterns related to specific issues. Its matrix format makes it easier for research teams to work together on the analysis and for readers to trace how conclusions were derived from the data. Developed for social policy research, it is now widely used in other fields as well, including health services research and public health.

In Ritchie and Spencer's account, framework analysis proceeds through five stages. The first stage is familiarization: the researcher reads and re reads the transcripts and field notes to become thoroughly immersed in the data. The second stage is identifying a thematic framework: the researcher develops an initial set of codes and themes, drawing on the research questions, the topic guide, and issues that emerge from the data itself. This framework is provisional and can be modified as analysis proceeds.

The third stage is indexing: the researcher systematically applies the thematic framework to the data, labeling sections of each transcript with the relevant code(s). The fourth stage is charting: the data are rearranged from their original context and organized into a matrix (chart or framework) with themes as columns and cases (individual participants or units) as rows. Each cell contains a summary of what that participant said about that theme, along with references to the original data.

The fifth stage is mapping and interpretation: the researcher examines the completed matrix to identify patterns, associations, and explanations. This stage involves looking across cases within a theme (to identify the range of views on a topic) and across themes within a case (to understand individual experiences holistically).

The matrix format is the hallmark of framework analysis and what distinguishes it from other approaches. It provides a structured overview of the entire dataset that allows both within case and cross case analysis, and it creates an audit trail that enhances transparency and rigor.

Framework analysis is most appropriate for applied qualitative research with relatively clear questions or objectives, when the data come from semi structured interviews or focus groups with a defined topic guide, when the researcher needs to compare and contrast across cases, when multiple researchers are working on the analysis, and when the findings need to be communicated to policy audiences who value systematic and transparent methods. It is less appropriate for highly inductive research aimed at generating novel theory from data, where grounded theory or interpretative phenomenological analysis might be better choices.

A common error is applying the thematic framework too rigidly, forcing data into predefined categories and missing emergent themes that the framework does not capture. The framework should be treated as a living tool that evolves as analysis proceeds. Another mistake is reducing the charting stage to a mechanical exercise, producing summaries that are too brief or too literal and losing the richness of the original data. Researchers sometimes skip the familiarization stage and proceed directly to coding, which risks superficial indexing. Finally, some researchers describe their analysis as "framework analysis" when they have merely organized data into a table without following the systematic stages of the method, which undermines the rigor that the approach is designed to provide.

A health services research team uses framework analysis to explore barriers to accessing mental health care among 30 refugees in three cities. After conducting semi structured interviews and familiarizing themselves with the transcripts, they develop a thematic framework with five themes: awareness of services, language barriers, cultural stigma, trust in providers, and systemic barriers (waiting times, referral processes). The charting stage produces a 30 x 5 matrix. Mapping the matrix reveals that language barriers and cultural stigma interact differently across the three cities: in City A, interpreting services are available but stigma is high; in City B, stigma is lower but no interpreting services exist. These city level patterns inform targeted recommendations for each location.