. "Cases and controls are matched with regard to characteristics such as age, gender, and socioeconomic status, so as to eliminate these characteristics as the cause of observed differences in past exposure." . . .