10,000 laptops a week lost at US airports? I don't believe it
A story circulating in the news over the weekend could well throw business travelers into panic; according to the Ponemon Institute, as many as 10,000 laptop computers are lost in American airports every week. I've learned, however, to question statistics that seem unreasonable, and this one struck me as 14K b*****t. Looking further into the referenced study, I believe I might well be right.
The first question I had was, just how was this data measured and collected? The answer, from the study- "Laptop loss frequencies were collected from a confidential field survey as either a direct weekly estimate or as a range variable as reported by airport officials. Exact loss frequencies were typically not calculated or available for review." The 'range variables' of the responses were not included.
At this point, I had a vision of a harried airport official with a survey and a hundred other papers in his/her hands writing down a WAG just to get it out of the way. But do the statistics support this suspicion? Take a look.
Atlanta: 89 million passengers; 430 lost laptops; 4.8 lost laptops/million passengers
Chicago: 76 million, 825 lost: 10.85 ll/m
Los Angeles: 62 million, 1,200: 19.4 ll/m
Dallas: 60 million, 250: 4.16 ll/m
Denver: 50 million, 175 3.5 ll/m
JFK New York: 48 million, 900; 18.5 ll/m
Las Vegas 48 million, 240: 5.0 ll/m
Houston: 43 million, 175: 4.0 ll/m
Phoenix: 42 million, 200 4.8 ll/m
Orlando: 36 million, 50: 1.4 ll/m
I find it very hard to believe that passengers in the Los Angeles airport lose five times more laptops than those in Dallas, Denver and Houston, and therefore I wouldn't credit this report until a more objective metric substantiates it.
My second question was, to what end was this report prepared? Not surprisingly, the study was commissioned by Dell, and is prominently featured in marketing for the company's Prosupport Services, a package that purports to help business travelers maintain data security, including that found on lost laptop computers.
You'll no doubt see this study quoted widely in the press, since it makes a nicely shocking headline. However, statistics are a marketer's best friend, as moldable as paper-mache and as confusing to most people as three-card monte. In this case, they also have the aroma of week-old fish.