Order Effects You observe that your friends felt exhilarated after riding a roller coaster without loops (which turn riders upside-down), then felt nauseated after riding a roller coaster with loops. You conclude that loops lead to nausea. The problem is that there could be an order effect. Perhaps your friends would have felt nauseated after the second roller coaster ride whether or not it had loops. Counterbalancing would avoid this confound. Half of your friends would be randomly assigned to ride the one without loops first, then the one with loops; half of them would be randomly assigned to ride the one with loops first, then the one without loops. (Your second author gets queasy just thinking about roller coasters and would not participate in this experiment no matter what experimental controls were in place! But your first author would volunteer!)