No, they actually bought ads on three types of Web sites (adult, gift shopping and news) and tracked how many users of each type of site were in private mode when they landed on the ads. The researchers shelled out $120 to buy a total of 155,216 ad impressions, split evenly among the three Web site categories. That gave their survey a far better statistical sample than many high-priced analyst studies use.
This week's study on Web-browser privacy has one especially interesting element: how the university researchers reached their results on the number of people who use browsers in private mode while shopping online. Did the researchers from Stanford University and Carnegie Mellon University just extrapolate from available Web survey numbers, the way you might expect cash-strapped academics to do?