AI Auto-Analysis
Overview
After transcription, FieldSocio’s AI pipeline automatically analyzes your interview data using Gemini 2.0 Flash. The analysis is tailored to your project’s chosen methodology.
What the AI Generates
Summary
A concise overview of the interview covering:
- Key topics discussed
- Main arguments or viewpoints expressed
- Notable quotes or insights
Auto-Coding
Qualitative codes are generated based on your project’s methodology:
| Methodology | Coding Approach |
|---|---|
| Grounded Theory | Open codes → axial codes → selective codes |
| Thematic Analysis | Initial codes → themes → sub-themes |
| Phenomenology | Meaning units → textural descriptions → structural descriptions |
| Framework Analysis | Codes mapped to pre-defined framework categories |
Entity Extraction
The AI identifies and extracts:
- People — Names, roles, relationships mentioned
- Places — Locations, landmarks, institutions
- Organizations — Government bodies, NGOs, companies
- Concepts — Key themes and abstract ideas
Sentiment Analysis
- Overall interview sentiment
- Per-segment sentiment tracking
- Useful for understanding emotional tone across topics
Using AI Results
AI-generated analysis is a starting point, not a final analysis:
- Review all auto-generated codes before accepting
- Modify codes to match your analytical framework
- Reject codes that don’t fit your analysis
- Add your own codes alongside AI suggestions
- Use summaries for quick reference, not as substitutes for close reading