Mind · Strategy
Experiment
Structured causal discovery via variable isolation
Definition
Mechanism
A structured Protocol for causal discovery. Unlike Verification, which confirms a claim already made, Experiment generates a Solution by comparing a control group against a treatment group differing in the variable under test and, so far as declared, nothing else.
Invariants
- Control and treatment differ in the variable under test; every other difference is eliminated or written down.
- The design commits in advance to learning from either outcome.
Risks
Failure modes
Undeclared difference
something varied that the design did not name, so no cause is established.
Mechanism assumed
the comparison shows the variable mattered, not why.
Relationships
Builds on
References
Infrastructure
Communication standard
Confirming alignment with truth or spec
Yields
Infrastructure
The fractal unit of completed work
Used by
References
Calculating the economic value of reducing uncertainty
Composes with
Bayesian belief updating via falsification rungs
Reference