Jackson Free Press

French author André Gide wrote: "Trust those who are seeking the truth; doubt those who find it." I wrote the quote on a white sticky in magenta ink yesterday and attached it to my laptop… "doubt those who find it." After sitting in a hearing room for three days this past week listening to hours of testimony, it's apt. The law, I think, is all about finding the "truth," subjective as it usually is.

My father, a brilliant man and grad-school professor of subjects-I-never-wanted-anything-to-do-with, raised me to think critically and question assumptions. It's a way of thinking that made me the bane of many a teacher's existence. I asked too many questions; "I don't understand," was their cue to change the subject, fast. I remember a lengthy debate in elementary school about the concept of zero, for example, that only ended when my teacher's faced looked a lot like my mother's when I pushed "why" too far: veins popping out on her neck and forehead, eyes narrowing and lips forming an angry line. It was a concept I had to figure out on my own.

Because I naturally question assumptions, I have an aversion to statistics—they're so easy to manipulate. To think like a statistician means connecting all the dots and then declaring a "truth" about where those dots led. But, as we all do from time to time (including me), people tend to start with a conclusion, then attempt to find evidence to support it. Unless you happen to be a theoretical physicist, that's just the wrong way look for the truth.

The problem is that backwards research methodology also can produce some spectacular half-truths: the more spectacular they are, the more we want to believe them. It's the old "don't-look-at-the-man-behind-the-curtain" trick, sleight-of-hand used by politicians, marketers, magicians and apparently, lawyers.

I admit it; I look at the tabloids in the grocery store. If I have 15 minutes to spare, I'll go through an issue of People magazine to get my rumor fix, and hearing something juicy about someone I know is heady stuff, especially when I don't particularly like the person. I try hard not to gossip, but I get sucked in, too. It's not very difficult to redirect my attention with ambiguities, pretty colors and platitudes. I have my biases: agree with them, and you've got me hooked—at least for the moment.

Believing can be satisfyingly easy. It's doubting and questioning that's hard. Taking down the curtain and confronting what's behind it can be scary and painful, but in an age of ten-second sound bites and headline news, it's incumbent on all of us to raise our awareness of what's really going on. Question assumptions, including your own. Consider the source of your information: is there an ulterior motive? Do you have all the facts or just the ones you've been given?

Doubt those who claim to have found the truth.

---------------------------------------------

If you think you're pretty good at concentrating and can't really be distracted, watch this b-ball practice video and tell me how many times the players in the white t-shirts pass the basketball.

Here's another little video about the town of Allopath, an allegory about how faulty conclusions can become "fact."

Here's an interesting and easy-to-read article on understanding statistics and how they can be manipulated.

Previous Commentsshow

Statistics are like bikinis. What they reveal is suggestive, but what they conceal is vital. ~Aaron Levenstein

Brent Cox

Feb 18 2007

The history of our race, and each individual's experience, are sown thick with evidence that a truth is not hard to kill and that a lie told well is immortal. Mark Twain (1835 - 1910), Advice to Youth He also did the "There are lies, damn lies, and statistics." quote as well. :) We got that one in the first class of my Stat class in college.

Ironghost

Feb 18 2007

Just curious if anyone watched the b-ball video... Without going back to look, did you notice anything interesting or out of place, like an animal?

Ronni_Mott

Feb 18 2007

It wouldn't open for me...

Rico

Feb 18 2007

I noticed a gorilla. However I watched the video about four times before I saw it because I was trying only to watch the ball being passed. I don't know that this proves a dishonestly selective focus, but rather a focus that intentionally excluded as much as possible in order only to see what I was trying to see. Okay, so maybe that's a good definition of a dishonest focus in some situations. Damn you and your gorilla!

Brent Cox

Feb 18 2007

Ronnie & Iron, Good point about statistics. I'd tweak Twain's saying a bit by saying "There are lies, damned lies, and misused statistics". Furthermore, my statistics professor in college even told the class "Statistics can never prove anything. They can only suggest". Even so, statistics are a powerful tool. One of the JFP's darlings, Richard Florida, also supplied an example, namely knowing exactly what the stats are designed to measure Several years ago, when he came with his "Bohemian Index", New Orleans ranked only 40-something out of the 49 largest metro areas - below even Salt Lake City!! Naturally, New Orleans guffawed and gave Florida an earful, basically saying Florida doesn't know what he's talking about. But as Florida later explained, the index measured ONLY people who were employed as artists, writers, musicians, etc. - meaning they listed their bohemian activity as their primary source of income. So its not that Florida was wrong, it's just that his method of measuring bohemianism can't detect moonlighting bohemians (which obviously NOLA has lots of). This is statistic's version of "Read the Fine Print!" Examples of a bad use of statistic: Assuming that every portion accurately reflects the whole especially when the stats for the whole obscure what the parts might be like (e.g., the state of Mississippi vs. Fondren/Belhaven, the Memphis metro area vs. Cooper-Young, etc.). Assuming that the part represents the whole is also wrong (London vs the rest of the UK; Rio and Sao Paulo vs the rest of Brazil, etc.) Example of a good use of statistics: It's way too complex to explain here, but having the following is a good running start: * appropriate sample size (usually at minimum 30 of X, afterward the more of X the better) *proper use of the Pearson R, The T-Test, the Phi Coefficient - namely which test is the most appropriate to use; especially knowning which to use in a "non-nomral distribution" (this is usually the case) *proper knowledge of causality theory is also a big help here (for reasons stated above

Philip

2007-02-18T22:46:05-06:00

ID

110429

Comment

Philip, thanks for your comments. Good stuff. It's unfortunate that most of us don't have a good grasp of how statistics really work, and so are likely to be led down whatever garden path happens to be popular or expedient for the presenter. The whole point of the b-ball video is to demonstrate how easily our attention can be diverted by something completely irrelevant. After watching it for the first time, I was amazed to find out that there was a gorilla walking through the scene and I never saw it -- I was concentrating on counting white t-shirt passes, just like I was told. It's telling that most people miss the big black gorilla walking through. Sort of like "don't pay attention to Iraq, pay attention to gay marriage." It's frightening when you consider the implications.

Author

Ronni_Mott

Date

2007-02-19T18:44:04-06:00