Have any of you ever taken the "Pepsi Challenge?" Read the following example of a two-sample design, including the conclusion(s) drawn by the experimenter.

Psychologists working for food and beverage companies have always played a critical role in product development. In one experiment conducted by a consumer psychologist, subject preference for two types of cola was measured. The company had noticed that in one marketplace its brand of cola performed significantly worse than its leading competitor's cola. These data were particularly puzzling as on a nationwide basis its cola performed significantly better than its competitor's cola.

The researchers were concerned that some local condition may have contributed to the rejection of their cola, so they set out to test this hypothesis. The experimental design was a repeated measure design in which each subject tasted two colas. One cola was marked Q (the competitor's brand) and the other was marked M (their brand). A random sample of citizens between the ages of 14 and 62 were asked to participate in the experiment. All subjects tasted brand Q and then brand M, and then gave their preference. Much to the surprise of the experimenters, the subjects reported an overwhelming preference for brand M. The authors concluded that the sample preferred their company's brand and that advertising must have contributed to the consumption of the competitor's brand in that area. Therefore, they suggested a multimillion-dollar advertising campaign to rectify the situation.

• Indicate what you see as other possible explanations for the results of their follow-up study.
• How might you design a study to determine if this "other explanation" is really the reason for the their preference?
• Can you think of any other possible examples used in marketing where an obvious confound was present?