3 In comparing test results against the norm, is the stated significant difference level in scores appropriate? From a statistical testing point of view, is type I or type I error more important here? Advertising Testing Services, Inc. (ATS) is a small marketing research supplier special- izing in the copy testing of television commercials ATSI is located in Princeton, New Jersey, and numbers among its clients some of the nation's largest advertisers. ATSI tests both animated and finished versions of commercials Procedure ATSI's typical testing procedure is as follows: 1 Approximately 150 respondents in total are recruited from five to seven different shopping centers around the country. Typical locations would be St. Louis, Missouri; Hartford, Connecticut: Tampa, Florida: Denver, Colorado; and Charlotte, North Carolina. Potential respondents are screened to make sure they are in the target audience of the commercial being tested and are users of the product category being tested. This is done by asking potential respondents whether or not they have bought or used prod- ucts in many product categories, including the one of interest. 2 Respondents are taken to a viewing area where they are shown the test com- mercial along with six other commercials for noncompetitive products. 3 After viewing all the commercials once, the respondent is given a personal interview.