• Enter your email address to follow this blog and receive notifications of new posts by email.

    Join 41 other followers

  • February 2013
    S M T W T F S
    « Jan   Mar »
  • Recent Bookmarks:

  • Archives

  • Categories

  • Advertisements

Government Rule by Statistical Studies

David Brooks, NYT columnist, has been writing a lot lately about statistics and all the studies being churned out by social and psychological researchers. And, based on an earlier column of his I read this year, he is going to spend a lot more column space on these topics in the future. He is just one member of a growing movement of data-based-decision makers, or “quants”, who intend to use statistics in every corner of their lives, and subsequently in every corner of ours.

There are plenty of reasons to oppose this sort of philosophy from the libertarian or conservative perspective. The lack of freedom and destruction of tradition in the name of statistics is still upheaval and oppression, even if it’s supposedly good for us. It hurts seeing this from a guy like Brooks, who is from the right and with whom I agree more than disagree, but I admit the idea of a “statistically perfect” life is alluring to some, including me. While the “philosophy of data” might help us in our understanding of why we behave the way we do, we need to tread carefully, it’s not all that effective. And the more you use “studies” to try to make changes in the real world, the less effective it will be.

One of the studies Brooks’ cites in his column was research showing changing the shape of a glass can get people to drink less while thinking they drink more. It’s basically an optical illusion, but the study showed people who used very tall and narrow glasses consumed less than a control group using standard glasses, but they (the tall-glass people) ‘felt’ they had consumed more liquid than they actually consumed. And the liberal fascist immediately thinks “hey, we should make a law saying all carbonated beverages should be consumed in these tall and narrow glasses, certainly this will end our obesity problem.”

It won’t, because these little effects disappear at the macro level, over the longterm.

I used statistical evidence for a long time as I tried to gain control over my weight. I knew the research and did my best to use small plates (a study showed people eat less when using small plates), I never allowed myself to be distracted while eating (another study), I took smaller bites (another study), drank a big glass of water before eating (another study), always ate breakfast (another study), and did some other things that were supposed to help me control my eating. And none of it worked, because my body knew how much it was eating, even if my brain thought I was eating enough because my plates were small and I always waited 20 minutes after eating before having a dessert.

The research I was using was fine, it wasn’t fraudelent, it’s just the effects found in those studies didn’t have staying power. Most of the time these studies are done over very short periods of time (it should be obvious the tall/thin glass effect and the small plate effect weren’t going to undo a billion years of evolution telling us to eat as much as we can now because tomorrow we might not have any food); it’s common sense, but common sense is the first thing lost when using data to make decisions about your life.

I’m not saying data is bad or that the studies are meaningless. But I am saying that studies try to control all the factors to find an effect, but it’s impossible to control all the factors in the complex system (real life) where you’re going to try to enact the changes you found in the study. Life is complicated, lab research is simple. It’s as straightforward as that. Use data when you can, but understand its limitations. (What’s truly ironic is you have to be very statistically savvy to understand exactly why you should skeptical of every new ‘study’ that’s thrown around the popular press, but that’s a topic for another post.)

We’re going to see more of this in the future. Why? because it works. People feel better about anything when the “science” label is put on it, and the left has now claimed the “science” ground even though they have no more claim to it than anyone else (it doesn’t help that some conservatives have abandoned science to the left, to the detriment of conservatism.) The best cure is to learn as much about the scientific method and as much about statistics and probability as possible. Science doesn’t belong to any political movement, and no more than a few decades ago statistics was considered a “right-wing” way of looking at the world because it was devoid of liberal values like compassion. If liberals can have data and keep their values of compassion, there’s no reason why conservatives can’t embrace the same while still adhering to our values of tradition and freedom.*

[One interesting area where statistical research has changed my life is shopping. I once read research showing that if a person touches something, they are much more likely to purchase the item. I try to be frugal, so every time I touch an object while shopping, I reflect on this research and re-evaluate whether I really need the object. I even made a rule, if I touch an object and I don’t absolutely need it, I put it back. Sometimes I come back on another trip and buy it, but not normally. From the opposite side, whenever I’m selling something (I normally work retail), I always find a way to get the customer to touch the object.]

*Yes, I know there’s tension between freedom and tradition, but it’s not as severe a divide as some people would like to think. Again, a topic for another post.


From where does the word “Statistic” originate? Wiktionary says:

“From German Statistik, from New Latin statisticum (“of the state”) and Italian statista (“statesman, politician”). Statistik introduced by Gottfried Achenwall (1749), originally designated the analysis of data about the state.”

So we shouldn’t be surprised, and in fact it’s been a common force in left-wing thinking, that statistics will be used as a rational for growing The State. Statistics, both real and otherwise, will always be used to help The State in its growth into the leviathan.

%d bloggers like this: