Use when developing and densifying categories with properties, dimensions, conditions, and consequences.
A category is a higher-order concept that groups related incidents under an abstract label. Dense categories specify properties (characteristics) and dimensions (ranges along which properties vary), plus conditions under which patterns hold and consequences that follow.
Move up a level when multiple codes repeatedly co-occur or share a latent pattern. Rename categories using participant-relevant language when possible (in vivo lift), then refine to conceptual clarity.
Example: if “time pressure” is a category, properties might include (institutional vs interpersonal) and (acute vs chronic); dimensions map variation across cases.
Ask:
Use conditional matrices (see visual-modeling) when relationships multiply.
Compare incidents within the same category to discover new properties. Compare across categories to locate boundaries (what this category is not).
A thin category is a label without variation spelled out. A thick category has:
Saturation is about categories: you stop sampling for a category when fresh data no longer reveals new properties/dimensions relevant to your emerging theory. Some peripheral categories may remain thin if they are not theoretically central—justify that choice.
Category: “Patching workarounds.”