Model
Variables, controls, and causal model
Run 3 controlled trials. Record the changed variable, controls, measured response, and uncertainty before choosing a mechanism.
- Keep fixed
- starting system state and observation interval
- Measure
- Observable change
- Uncertainty
- Use repeated trials and report the observed range before comparing models.
- Model limit
- The Molecule Forge model supports claims only within the tested variables and boundary conditions.
Evidence
Dataset and uncertainty
Commit a falsifiable model and run the repeated trials to generate the dataset.