A manufacturer wanted to get opinions from 4,000 hardware store managers on a new type of lawn mower. An associate provided a list of such stores, divided into 400 large and 3,600 small stores. He drew a random sample of 200 stores and was disappointed to find only 19 large stores in the sample, since they represented more than 30 percent of the potential volume. A friend suggested that he draw a second sample. What do you recommend? What other pieces of information would you like to have?