Toothpaste manufacturers have found consistently that if they ask for detailed information on the frequency with which people brush their teeth, and then make minimal assumptions as to the quantity of toothpaste used on each occasion, as well as spillage and failure to squeeze the tube empty, the result is a serious overstatement of toothpaste consumption. How would you explain this phenomenon? Would it be possible to design a study to overcome these problems and obtain more accurate estimates of consumption? Describe how such a study could be conducted.