How Evidano was used in the study
The thesis, on the contestation and localisation of gender equality and LGBT norms in Georgia, followed Braun and Clarke’s six-phase thematic analysis. Here is where Evidano fit in, exactly as the thesis reports.
Initial coding (Step 2): The author employed Evidano to process transcripts and suggest preliminary codes across the dataset, giving a broad first overview of the material.
Human review of every code: AI-suggested codes were manually reviewed, then broadened, narrowed, merged, or discarded against the research question. Evidano was the starting point, not the final word.
Reducing researcher bias: In the limitations, the author reports using Evidano to receive a less biased interpretation of the data. Ethics approval came from the University of Glasgow.
How Evidano fit the workflow
From transcripts to published themes
- 112 interview transcripts
From academia, international organisations and civil society, conducted in Tbilisi and online in early 2025.
- 2Evidano suggests preliminary codes
Across the full dataset.
- 3Researcher refines, merges and discards
Codes by hand.
- 46 themes
Defined, named and analysed.
- 5Published thesis
With AI use disclosed in the methods and limitations.
“To manage the volume of qualitative data efficiently and generate a broad overview of potential codes, the artificial intelligence tool AILYZE […] was employed.”
“Transcripts were processed by the tool to suggest preliminary codes.”
“To mitigate this issue, an artificial intelligence tool […] was used to receive a less biased interpretation.”
Where the time goes
Hand-generate initial codes across every transcript. Slow start on large datasets. Researcher judgment throughout.
A broad first pass of codes from your transcripts. A fast overview to build and refine from. Researcher judgment throughout.