Theater tests of advertising effectiveness sometimes involve having several groups of people observe a pilot film for a television series which hasa number ofcommercials interspersed. Several groups see a commercial for the brand under test whereas other groups do not see this commercial. After the showing, the viewers complete questionnaire which, in part, measure their attitude toward the test brand. Differences in attitudes between those who saw the commercial and those who did not are a measure of the commercial's effectiveness. What type of experiment is this? What type of errors are likely to affect it? What changes, if any, would you suggest?