Qualitative analysis overview

Qualitative analysis helps you structure interviews, workshops, and sessions. You mark relevant statements, assign codes, and examine patterns across several sessions. StoryVault connects each coded passage to its transcript, speaker, and recording position.

If a codebook already exists, open a session belonging to that project and switch to its Coding tab. Select a passage and assign a code from the project’s codebook. The tab belongs to the session view; the project provides the codebook.

Note

The Coding module is enabled per tenant. Ask an administrator if Codebook, Coding Reports, or the Coding tab is missing.

End-to-end workflow

1. Define the research question and which sessions belong to the material.
2. Build a project codebook with clear definitions and examples.
3. Code relevant passages manually or review AI suggestions one by one.
4. Record decisions and borderline cases in memos.
5. Review samples as a team and refine the codebook deliberately.
6. Select a document set in Coding Reports and inspect matrix, frequency, quotes, and co-occurrence.
7. Create a snapshot for reproducible results and export the required tables and quotes.

Build the codebookDefine, organize, and maintain project codes.

Code transcriptsMark passages, combine codes, write memos, and review suggestions.

Analyze reportsCompare corpora, verify quotes, interpret patterns, and export results.

Deductive, inductive, or combined

A deductive codebook starts with categories from a guide or quality model. An inductive approach develops codes from the material. Many projects combine both: initial codes provide direction and observations add new categories.

The Discovery template starts with Pain Point, Desired Benefit, Job-to-be-Done, Workaround, and ZDF Signal. It is a practical starting point, not a prescribed method.

One passage can carry several codes. Memos document interpretation. Open AI suggestions are excluded from reports until a person accepts them.

For defensible results, define inclusion and exclusion criteria, keep segments as short as possible while preserving context, review samples after codebook changes, and use document sets and snapshots. Frequency indicates where to look; it does not establish importance, and co-occurrence does not establish cause.

Accepted coding and memos remain after re-transcription. StoryVault remaps segments and flags uncertain matches. Earlier text remains available in Transcript versions.