orgtheory.net

stupid statistician tricks

Here’s a conversation I’ve had a few times with statisticians:

Statistician: ” … and these simulations show how my results work.”

Me: “What does your research tell us about a sample of, say, a few hundred cases?”

Statistician: “That’s not important. My result works as n–> 00.”

Me: “Sure, that’s a fine mathematical result, but I have to estimate the model with, like, totally finite data. I need inference, not limits. Maybe the estimate doesn’t work out so well for small n.”

Statistician: “Sure, but if you have a few million cases, it’ll work in the limit.”

Me: “Whoa. Have you ever collected, like, real world network data? A million cases is hard to get.”

Statistician: “The Internet is a network with millions of nodes.”

Me: “Sure, but the Internet is one specific network. Most real world networks have hundreds or thousands of nodes. Like a school, or firms that trade with each other. Network data is expensive to collect. Some famous social science papers analyze networks of dozens of people.”

Statistician: “Um… the Internet! Scaling! Big networks! The Internet is a network! Facebook! FACE. BOOK!”

Me (rolls eyes): “What-EVER!”

This illustrates a fundamental issue in statistics (and other sciences). One you formalize a model and work mathematically, you are tempted to focus on what is mathematically interesting instead of the underlying problem motivating the science. An economist works on another equilibrium theorem rather than, say, taxes. The physicist works on the mathematics of super string theory, even when the experimental evidence isn’t there.

We have the same issue in statistics. “Statistics” can mean “the mathematics of distributions and other functions arising in statistical models.” Or it can mean the traditional problems of statistics like inference, measurement, model estimation, sampling, data collection/management, forecasting, and description. The problem for a guy like me (a social scientist with real data) is that the label “statistician” often denotes someone who is actually a mathematician who happens to be interested in distributions. That’s why they are happy with limit theorems, because limits smooth out hard problems and produce elegant results.

What I really want is a nuts and bolts person to help me solve problems. I may tease economists for their bizarre obsession is identification at the expense of all else, but at least identification is a real issue that needs to be taken seriously.

Adverts: From Black Power/Grad Skool Rulz

Written by fabiorojas

April 27, 2012 at 12:01 am

Posted in fabio, mere empirics

11 Responses

Subscribe to comments with RSS.

  1. […] Fabio Rojas reports on “a conversation I [Rojas] have had a few times with statisticians”: Rojas: “What […]

    Like

  2. Strawmen make the best interlocutors.

    Like

    pretendous

    April 27, 2012 at 12:51 am

  3. […] stupid statistician tricks « orgtheory.net Me: “Sure, but the Internet is one specific network. Most real world networks have hundreds or thousands of nodes. Like a school, or firms that trade with each other. Network data is expensive to collect. Some famous social … stupid statistician tricks « orgtheory.net […]

    Like

  4. Fabio, I think you missed the point entirely. Your assumption is that you and your work really should matter to this person. Appears that she/he could not care less, and I don’t see any reason why she or he should care. If you care that much, why don’t you figure it out on your own? If you don’t have those sorts of skills, then maybe find someome who does and who shares your interest. Just my $.02 worth. Oh, and, “whatever”ing someone is going to really get you far on stuff like this.

    Liked by 1 person

    Scott R. Eliason

    April 27, 2012 at 6:14 am

  5. I can’t tell if Scott was being sarcastic or sincere.

    Liked by 1 person

    brayden king

    April 27, 2012 at 2:56 pm

  6. From what Scott wrote over on Gelman’s blog, I’d say he’s sincere. Nerd fight!!!

    Like

    Rich

    April 27, 2012 at 3:57 pm

  7. Who would be so daring so as to believe statistics ought to be applicable to real-world problems?

    Like

    Guillermo

    April 27, 2012 at 5:47 pm

  8. In the econometrics literature, authors usually perform Monte Carlo studies to determine the finite sample performance of new estimators, especially when several consistent estimators are proposed in the same study. Is this not the case in applied stats?

    Like

    Nate

    April 27, 2012 at 7:01 pm

  9. […] Friday, I write a semi-humorous post about the interaction between statisticians and non-statisticians. The issue that brought it up was […]

    Like

  10. Did someone say nerd fight? Where’s the nerd fight? Did I miss it? Scott was all like, “Nobody cares about you Fabio!” Then he was like, “Ugh, wish I hadn’t pressed ‘send’ on that.” And Fabio was all like, “Ugh, you hurt my feelings Scott!” And Scott was like, “let me say that again on some other blog.” And then again Scott was like, “I really wish I hadn’t pressed ‘send’ on that either.” Then Mr. Linnemeyer, the metal shop teacher, came back in the room and…that was that. What a pathetic nerd fight!

    Like

    annie

    May 2, 2012 at 11:05 am

  11. @annie: Yea, I don’t think Linnemeyer even had to physically get between them. Soon as he came back into the shop room, they were both right back to the lathe. Fabio was like, “I’m making a wicked one hitter!” Scott was like, “Dude, I’m making a blackjack and when I’m done…let’s just say you’ll understand the law of large numbers!”

    Like

    Rich

    May 3, 2012 at 12:31 am


Comments are closed.

%d bloggers like this: