Design rigorous experiments — hypotheses, controls, randomization, power analysis, and preregistration. Use when planning any study where you need trustworthy causal conclusions.
