Research register
Cortana publishes study status, planned sample, evidence rules, quality controls, and limitations so a research claim can be judged on more than presentation.
STUDY REGISTER
There are no completed Cortana field studies in the public register yet.
DRAFT PROTOCOL 0.1
A proposed field study of the operating, technical, economic, and organizational conditions that stop AI initiatives from moving from experimentation into reliable use.
EVIDENCE STANDARD
A percentage without a denominator, a model without assumptions, or a client result without permission is not treated as credible proof.
Planned work, active fieldwork, analysis, and published findings are never presented as the same thing.
Counts accompany percentages, and missing or excluded records remain visible.
Observed data, participant reports, management assumptions, and external benchmarks stay separate.
Recruitment bias, sample constraints, uncertainty, and conflicting evidence belong beside the conclusion.
Research consent is separate from commercial follow-up, and identifiable material requires written release.
PUBLICATION NOTE
Planned sample sizes and model outputs are design inputs, not findings. Cortana will not present them as completed research or client performance.