Now that we're on the topic of econometrics . . . somebody recommended to me a book by Deirdre McCloskey. I can't remember who gave me this recommendation, but the name did ring a bell, and then I remembered I wrote some other things about her work a couple years ago. See here.
And, because not everyone likes to click through, here it all is again:
After I posted this discussion of articles by McCloskey, Ziliak, Hoover, and Siegler, I received several interesting comments, which I'll address below. The main point I want to make is that the underlying problem--inference for small effects--is hard, and this is what drives much of the struggles with statistically significance. See here for more discussion of this point.
Scott Cunningham writes,
Today I was rereading Deirdre McCloskey and Ziliak's JEL paper on statistical significance, and then reading for the first time their detailed response to a critic who challenged their original paper. I was wondering what opinion you had about this debate. Is statistical significance and Fisher tests of significance as maligned and problematic as McCloskey and Ziliak claim? In your professional opinion, what is the proper use of seeking to scientifically prove that a result is valid and important?
Dsquared's comment on this entry mentioned the economist Deirdre McCloskey, whom I googled and found this paraphrase of a quote from Don Boudreaux, "that no one was ever convinced by raw data of the truth of a proposition that he or she did not already hold to be true."
I wonder what the original form of the quotation was. Or maybe what I really wonder is, what is the point of demarcation for which Boudreaux's statement is true? As written it is certainly false. I know this because I myself get convinced by raw data of propositions on which I held no prior opinion. Here are just a few recent examples:
- Rich people and poor people differ more in their political preferences in poor states than in rich states
- The NYC police department stopped more minorities than whites, even after controlling for neighborhoods and previous crime rates
- Americans have, on average, about 700 acquaintances each
- and many, many others.
And, as a bonus, here's an example where an analysis of raw data left me unconvinced of a hypothesis (that a local election was rigged).
I certainly did not "already hold" these propositions to be true--or false. In fact, in many cases (for example, the rich and poor states), the interesting proposition didn't even come to me until the data whacked me on the head with it.
Different sorts of propositions?
How, then, could Boudreaux (and McCloskey?) say such a thing? There must be different sorts of propositions. The propositions that I study tend to be technical--even the studies of voting and police stops are emprical questions, not value judgments. But for the big social questions, maybe it's hard for people to be persuaded by data because they already have such strong opinions. One thing I like about emprical science (including social science) is that I can be "convinced by raw data of the truth of a proposition that he or she did not already hold to be true"--and it happens all the time.
Recent Comments