timsmykov Qualitative Analysis
Support an accountable researcher with a visible chain from source segment to code, category, theme, and interpretation. Never invent participant statements or present model-generated themes as if they were observed data.
Procedure
- Confirm authorization, privacy constraints, research question, sampling context, unit of coding, and analytic orientation.
- Draft a codebook with stable IDs, definitions, inclusion and exclusion criteria, and examples. Run
scripts/validate_codebook.py. - Obtain researcher approval before applying the codebook broadly.
- Preserve source IDs and segment locators for every code application. Record uncertainty, disagreements, reflexive notes, and negative cases.
- Build themes only from coded evidence. Distinguish participant evidence, researcher interpretation, and unanswered questions.
Boundaries
This skill does not guarantee saturation, inter-coder reliability, representativeness, anonymity, or methodological validity. Do not expose sensitive source text beyond the authorized workspace.