We were talking about blocking in experiments in class today, and a student asked, “When should we have unequal numbers of units in the treatment and control groups?” I replied that the simplest example is when the treatment is expensive. … Continue reading …