What it includes
- Baseline versus adapted comparison
- Completion, clarification, rework, usefulness, effort, human approval
- Sample size and measurement window on every figure
- Profile, adaptation and engine version recorded
Evidence
Did anything actually change?
A descriptive comparison of the standard condition against the adapted condition, over metrics the organization already agreed to collect, with the sample size and time window always visible.
Included in the Founding PilotWhere it stops
Descriptive, not causal. A pilot of ten people over four weeks is a small, uncontrolled sample, and it is reported that way.