Col­umn: Lies and Statis­tics

The Standard Journal - - COMMENTARY - KEVIN MYRICK Ed­i­tor

The other day, I posted a link to a New York Times story from June on the Fish Wrap’s Face­book page. The story pre­sented ev­ery county in the United States along with some var­i­ous sta­tis­ti­cal ev­i­dence to go along with the Times’ the­ory that, many ar­eas of the na­tion are do­ing bet­ter than oth­ers in this lat­est post-re­ces­sion re­cov­ery.

Their num­bers, when all com­puted to­gether, placed Polk County as 2,595 out of 3,135 coun­ties in the United States. All the num­bers can be found by go­ing to http://tinyurl. com/kcvf5hm, which will take you to the Times’ story.

Another in­ter­est­ing note in all of this is the No. 1 rank­ing for the top county in the United States: Los Alamos County, N.M. They reached the top spot due largely to the high num­ber of col­lege grad­u­ates and work­ers with masters and doc­toral de­grees.

The story takes a lot of dif­fer­ent statis­tics to cal­cu­late the ranks, from the num­ber of grad­u­ates with high de­grees to how many peo­ple in the com­mu­nity are on dis­abil­ity. Here’s my big­gest prob­lem with this story and all based on statis­tics: when were the num­bers gen­er­ated?

Not once did the writ­ers of this story men­tion when th­ese num­bers were com­piled, or even how they were gen­er­ated. How is one to truly un­der­stand where we are to­day if we’re us­ing num­bers that might be older than Christ­mas, or fresh from some lat­est study re­leased in the New Eng­land Jour­nal of Medicine?

In my opin­ion, with­out that con­text it’s hard to make com­par­isons. It also makes us all too eas­ily ma­nip­u­lated by those statis­tics and how they are pre­sented.

I think one of the com­ments left on the story hit the nail right on the head too: what is their def­i­ni­tion of easy or hard?

It’s par­tic­u­larly hard to think about this – or any story in­volv­ing statis­tics – as be­ing patently hon­est con­sid­er­ing peo­ple be­lieve just about any statis­tic they want to in the end.

I read about a survey on Bloomberg News just last week that hit the nail right on the ham­mer with my ar­gu­ment. Re­sults of a survey con­ducted by by U.K. poll­ster Ip­sos Mori found that Americans rank just above Ital­ians in their un­der­stand­ing of statis­tics.

Th­ese re­sults claim Americans be­lieve 15 per­cent of the pop­u­la­tion iden­ti­fies them­selves as Mus­lim. In re­al­ity, only 1 per­cent of the pop­u­la­tion iden­ti­fies them­selves as fol­low­ers of Is­lam.

Then there’s my fa­vorite of the lot, that goes to show we all watch too much tele­vi­sion if this re­sponse is to be be­lieved.

“What per­cent­age of Amer­i­can girls aged be­tween 15 and 19 years give birth each year?” Survey says: 24 per­cent. In re­al­ity, its only 3.1 per­cent.

Then again, when were the “true” an­swer survey re­sults com­piled? Once again, that seems to have been left out of the story.

Also an in­ter­est­ing side­note about this, only 11,527 peo­ple were in­ter­viewed in the poll over- all. Which means they based the thoughts of an en­tire set of coun­tries off what would be a molec­u­lar sized amount of peo­ple in the gi­ant pool of the hu­man pop­u­la­tion.

Polling data for pol­i­tics works the same way, and so the small­est num­ber of peo­ple com­pared to the over­all vot­ing pop­u­la­tion ends up telling our na­tion’s lead­ers what to think and how to vote. It tells cable news com­pa­nies what to think vot­ers will do in elec­tions.

In an age when data mat­ters more than ever – from all those math­e­mat­i­cal cal­cu­la­tions used by the com­puter just to type this col­umn to the bil­lions of phone calls be­ing scooped up by the NSA daily – not telling the en­tire truth with statis­tics is sin­ful.

If we can’t even get the num­bers of what we are right, then what’s the point of re­port­ing them, much less re­ly­ing on them as a gauge to de­fine who and what we are.

Statis­tics are im­por­tant, but only if used in an hon­est and straight­for­ward way. We rely on data too heav­ily in decision mak­ing, mainly in re­sponse to what we see as a short term nu­mer­i­cal prob­lems.

We also need to shame those who seek to ma­nip­u­late statis­tics for their own gain. Those char­la­tans of cal­cu­la­tion de­serve ridicule for their reck­less abuse of the num­bers.

What we for­get is that all the data in the world is great, but re­ly­ing on it too heav­ily forces us to be­come too fo­cused on the trees to see the whole for­est. There’s the big­ger pic­ture, the in­di­vid­ual and the force of in­ter­con­nect­ed­ness driv­ing our world to­day just as much as the num­bers. If we re­mem­ber that, we’ll use that data to make bet­ter de­ci­sions over­all.

Newspapers in English

Newspapers from USA

© PressReader. All rights reserved.