A few ‘graphs from my weekend National Post column:
In the last six months, 85,619 cell phones were left behind in Chicago taxis, according to a recent survey. The situation is little better in London, with passengers having lost more than 60,000 cellphones, up significantly from a similar period in 2001.
One of the perils of our ardor for ever-tinier electronic devices is that they are easy to lose. A cellphone salesperson recently told me that the second most likely reason someone comes in to buy a new cellphone is that they lost their old one.
Where do all the phones go? Sometimes we know, like in an airport, a taxi, or a hotel. But all-too-often, we have no idea where the darn things went — perhaps to the same place as unmatched socks.
[deletia]
But a columnist in the Wall Street Journal is now questioning the accuracy of the “lost phone” survey. Writer Carl Bialik complains about three things in the study (issued by mobile security vendorPointsec):
- Given that the study was funded by a company which sells tools forprotecting data on lost mobile devices, it is questionable from the get-go.
- The taxi drivers were non-random, and there were only 100 drivers in each of nine cities.
- The number of devices lost in each cab was small.
Let’s deal with each of these points in turn. First I am conscious that Pointsec, a mobile security company, issued the study, but courtesy of Mr. Bialik I too now have a copy of the company’s data spreadsheet, so I can assess the quality of the claims independent of Pointsec’s press releases. There is no need to reject it out-of-hand.
With respect to Mr. Bialik’s second point, it is true that the study is small and not chosen randomly. In a perfect world, the sample size might be larger, but then again maybe not: It is a statistical misconception that sample size is directly tied to size of the population of interest. The Pointsec folks sampled 100 drivers in each city, and, given that drivers, in aggregate, reported losing multiple phones in the period, the size of the phenomenon was eminently observable.
[deletia]
Skepticism about statistical studies, especially vendor-sponsored ones, is laudable. But knee-jerk condemnation of such studies is as wrong-headed as fawning over every release that comes arcing in the window. Lost mobile devices is a big security problem,and the fact that it was a self-serving vendor saying so doesn’t change the truth of the matter.
Related posts: