Qualitative analysis
Codebooks, coding and the line between suggestion and evidence.
Who this is for
Analysts and research officers coding transcripts.
No screenshot on this pageThe transcript surface was captured and NOT published. With no interview selected it shows "No interview selected", which is an empty state rather than the coding screen this guide describes, and a picture of an empty screen under a caption about codebooks would be a caption making a claim the image does not support.
Codebooks and coding
Codes live in a codebook attached to a study. A coding must point at a real span of a real transcript, and the quotation must match the transcript at that span — a quotation that has drifted from its source is rejected.
AI suggestions are not findings
Not citable until a human confirms itProvenance records which codings were machine-suggested, which were human-authored and which were confirmed, and a qualitative summary states that split rather than presenting one undifferentiated total.
What this does not do
Stated here rather than discovered later
- A quotation stays attached to the response it came from. A translation is never presented as the original.
- Key informant interviews, in-depth interviews and focus groups are formats within qualitative research. No separate module, route or table exists for any of them.
- The platform does not translate. It governs translations produced elsewhere.