To test the effectiveness of two Web advertising agencies, you increase your ad purchase with agency A by 50% without changing your purchase through agency B. The referrals to your Web site from agency A increased by only 34% but the referrals from agency B fell by 21%. Why might the difference-in-difference estimate of the referrals per dollar through agency A be biased?

Answer rating: 100% (QA)

Difference in differences is a statistical technique used in econometrics and quantitative soci... View the full answer