Run experiments sequentially when they share outcome…

Run experiments sequentially when they share outcome variables or operate through overlapping mechanisms; run them in parallel only when they are independent on outcome, mechanism, and temporal dimensions.