The residential division of Prism's high-speed Internet service uses one advertising agency, while its commercial division uses
another. Two analysts, Andy and Brad, are asked to test the effectiveness of the two agencies. Andy proposes an A/B test that compares the click-through rates per ad of the two agencies. Brad proposes a difference-in-difference test in which the budgets for both agencies are increased by 50%, and the percentage changes in the click-through rates are compared.True or False: Andy's proposal is more likely than Brad's to suffer from bias due to issues with "representativeness."
Andy's proposal is more likely than Brad's to suffer from bias due to issues with "representativeness."
The above statement is True.
Explanation:
AB testing compares the click-through rate of the two agencies only. This test runs on random basis. It is a straightforward exercise that affects the business in the short-run and its effect are temporary. While difference-in-difference testing studies the non-randomized data. The rates are higher but it covers the whole data.
Andy's proposal is more likely than Brad's to suffer from bias to issues with 'representatives'. Andy opts for A/B testing while Brad opts for difference-in-difference testing. Therefore, Andy's proposal is more likely to bias due to limited verification.
