Measuring the Unmeasurable
Organizations generate enormous amounts of qualitative knowledge: meeting notes, emails, interviews, audit findings, and conversations, which rarely influence analytical models. This research investigates whether large language models can transform qualitative evidence into measurable variables while preserving nuance and context.
Research Question
Can qualitative organizational evidence be converted into structured variables that hold up inside quantitative decision models?
Why This Matters
Most decisions in real organizations are made on a blend of numbers and narrative. Models that ignore the narrative half optimize against a partial picture of the business.
Current Direction
Reading across qualitative research methodology and recent work on LLM-based information extraction, looking for the smallest viable schema that preserves nuance.
Early Notes
Early hypothesis: the unit that matters is not the sentence or the document, but the claim, with provenance and confidence attached.