She could use a control to compare her data to and make sure it is cleaner. Ex. Use the old detergent on ten shirts and then use the new detergent on another ten shirts. She would then compare the shirts she did each detergent on and come up with her conclusion.
