in a , a researcher finds a certain number of people in several categories—e.g., owners of cats versus owners of dogs. respondents are not picked on probability sampling criteria.