Constant Comparative Method — Constant Comparative Analysis

CCM is the analytic core of grounded theory but is not identical to the full grounded theory methodology. Grounded theory also includes theoretical sampling, memo-writing, and a specific theory-building logic. CCM has been extracted and used as a stand-alone analytic technique in thematic analysis and content analysis, though critics note that without the accompanying sampling and saturation logic, the comparison becomes less theoretically powerful.

Theoretical saturation is reached when new data — collected from deliberately varied sources chosen to challenge the emerging theory — no longer produce new categories, no longer alter the properties or dimensions of existing categories, and no longer reveal new relationships among categories. In practice, researchers typically document three to five consecutive interviews that add nothing new before declaring saturation. Saturation depends on the richness of the data and the complexity of the phenomenon, not on a fixed participant count.

Yes. CCM was developed with a broad conception of data. It has been applied to field observation notes, focus group transcripts, historical documents, social media text, policy papers, and case records. The logic of comparing incident to incident within a category is data-type neutral, provided the data are sufficiently rich to support conceptual comparison.

Both methods involve coding and grouping qualitative data, but their goals differ. Thematic analysis aims to identify, organise, and describe patterns of meaning across a dataset. CCM aims to develop or refine a theoretical explanation by systematically comparing every incident within categories, tracking properties and dimensions, and driving sampling decisions based on theoretical gaps. CCM is inherently iterative and theory-building; thematic analysis is primarily descriptive and can be applied in a single analytical pass.

Software such as NVivo, ATLAS.ti, or MAXQDA is helpful for managing large numbers of codes, linking codes to memos, and retrieving incidents within categories for comparison, but it is not required. The intellectual work of comparison, category refinement, and theoretical integration must be done by the analyst. For smaller datasets (fewer than 20 interviews), systematic manual coding with a well-organised memo system is entirely workable.