Find testable structure in public data without inventing new facts or laundering exploratory analysis into confirmation.
PUBLIC PROGRAM
Retrospective Data Lab
Public dataset mining · statistical discovery
“Old data. New instrument. Different question.”
Re-analysis of public datasets for signals that become measurable only after better questions, controls, or instruments exist.
ActivePredeclared re-analysis plans, holdout sets, sensitivity maps, provenance tracking, and versioned public methods.
01Retrospective analysis
02Public datasets
03Sensitivity mapping
04Provenance
Public material is limited to the surface that can be stated precisely without exposing protected implementation, sensitive relationships, unfinished claims, or invention strategy.
ASSOCIATED RESEARCH
NO PUBLIC PAPER BINDING.
Program relationships are canonical register bindings. They do not imply publication, outside validation, or unrestricted access.
CONTROLLED PROGRAM
This program currently exposes its mission and posture without a public research-file binding.