Experiments
- Empirical designs relying on deliberate, randomized intervention assignments, the gold standard for causal inference.
- Experiments remove confounding by design and allow direct estimation of treatment effects. Common in ecology when feasible.
Overlap: Closely related to Quasi-Experiments (when randomization is partial or approximate) and Adjusted Methods (when additional covariate control is needed)
Key feature: Random assignment of treatment
Usage: For establishing causality under minimal assumptions