Statistics. Are they all blarney? (I used a little Irish lingo in honor of St. Patrick’s Day)
The story is told of a statistician who was asked of his findings, “What does this mean?” He replied, “It depends on the lens through which they are viewed.”
Statistics and the research that compiles them are interesting. They’re useful, and they shouldn’t be disregarded. But should they be believed? Well let’s say, “Conditionally, they should.”
It was recently stated in a column by Jason Moreno ("Two neighborhoods, two grocery stores, two shootings and two responses," March 6) that Elkhart was in the bottom 4 percent of safe cities. In other words, 96 percent of the nation’s cities were safer than our fair locality. That’s shocking, so I thought I’d check it out and see if I could learn more.
The 4 percent rating comes from a web site called Neighborhood Scout (www.neighborhoodscout.com). It turns out that there’s good news and there’s bad news. Elkhart does rate very badly on that website. But maybe there’s a little wiggle room in just how bad.
I did a quick Internet search as to what was the safest city in the U.S. And I found that it all depends on who you ask. Househunt.com says it’s Sugar Land, Texas. Forbes magazine says it’s Plano, Texas. Parenting Magazine says it’s Burlington, Vt. Movoto ranks Glendale, Calif., No. 1.
I then ran those cities through Neighborhood Scout’s listings. Sugar Land got a 45, Plano got a 33, Burlington got a 20, and Glendale got a 53.
All of the Neighborhood Scout ratings put the allegedly “safest cities” in the middle or, in the case of Plano, in the bottom third of the nation’s safe cities. One would think if others thought they were the safest, that they’d at least be in the high 90-percentile bracket at N.S.
Statistics are a helpful tool. But they should be viewed with care.
Some time ago, I wrote critically about statistics that were in an opinion poll the president was using to promote his push for gun control regulations following the Newtown, Conn., tragedy.
Some folks complained that, in that column, I was saying President Obama was lying that there was such a poll, but that wasn’t the case. I had said that I didn’t believe the poll had value, and that Obama’s use of it was flawed. I also quoted other polls that had found different statistics.
Few things illustrate the interpretive nature of statistics better than the nation’s unemployment stats. They are famous for the their “it depends” inaccuracy. The Bureau of Labor Statistics has six classifications of unemployment, U-1 through U-6. U-3 numbers are most reported in the press.
The U-3 numbers reflect the unemployed who have actively looked for work within the past four weeks. It does not count discouraged workers who aren’t seeking work, part-timers who cannot find full time, or “marginally attached workers.”
U-6 numbers include the categories that U-3 does not, and so it shows a much higher rate of unemployment. And there are also adjusted, unadjusted, and seasonally adjusted versions.
Plus there’s a Gallup Unemployment rating system that moves alongside, but does not match, the U-3.
All of these compilations give politicians and activists plenty of information to use to take credit, place blame, or “prove” a claim, whatever it may be.
Statistics. It all depends on the lens through which they are viewed. My favorite distorted statistic is the one humorist Garrison Keillor uses in describing his mythical hometown of Lake Wobegon, where he says, “The children are all above average.”
Former Elkhart furniture store owner Richard Leib has served on planning committees in several industries. An avid auto fan, he raced in the 1972 coast-to-coast Cannonball Run. He has written on a wide range of subjects.