A: I don't trust non-parametric econometrics.
B: You mean you don't understand non-parametric econometrics.
A: No...I don't trust econometricians.
B: What do you mean? You don't trust people in general?
A: I don't trust anyone who analyzes data that they did not have any hand in creating. It just seems foolish. Why would anyone attempt to answer a question or understand a problem through someone else's eyes? There seems to be a step missing to me.
Subscribe to:
Post Comments (Atom)
4 comments:
Are they equally distrustful of full-parametric econometricians?
One thing that I really admire about the hard sciences is that you can't push the envelope too hard because someone can always go back and try to re-create your experiment. Any study that starts with striclty observational data is a study that starts with two strikes against it.
I think they were mostly distrustful of the data economists use, so they didn't think much of parametric econometrics either.
You know I work up here with statisticians. This year there are a few who care about econometrics (economics) too. In some ways it is scary.
The guy next to me is big in to experimental design.
I didn't really complete that thought, but right now, I can't seem to do it anyway.
The problem with all of this nonsense is that most the data is meaningless to begin with and tells you nothing. Call bullshit all you want but at the end of the day any moron could sit down and look at your data set and draw as much inference as you.
I've said it before but I'll say it again - all data is bad and there is no such thing as a "good dataset", at least as far as the theory is concerned. Look at the Fed and Congress - they're just now saying that the housing bubble pop and that we're in serious deep shit. The smack fiend I used to buy dope from could have told you the same thing four years ago. Think he knows any of this shit? No.
Post a Comment