So there's a book by an economist named Thomas Picketty which is very big right now, on capital and RoI trends over time. Now, I have not yet read this book so I'm going off of snippets and reviews (positive and negative), but since I've been reading Richard Heuer's /The Psychology of Intelligence Analysis/ (originally written for the CIA) one thing that is big on my mind now is that lots of data doesn't have to lead to good analysis. Heuer reports the interesting result that doctors who emphasize accumulating data over hypothesis testing tend to be worse at diagnosis; and experiments on experts in various domains (from horserace handicapping to social science) reveals that giving an expert more data does not improve the accuracy of his results, but does increase his CONFIDENCE in his own accuracy by a lot. That is, if you make a best-guess based on initial data, you're very aware that it's only a guess and your estimate of your own accuracy is pretty good; if someone gives you a whole bunch of information backing up the initial data, your accuracy doesn't improve but your impression of your own accuracy does--the extra data just makes you overconfident. It turns out that a multiple-hypotheses approach, concentrating on indicators that can differentiate between hypotheses (basically Bayesianism), is one of the best available approaches for avoiding overconfident estimates.
So, again emphasizing that I have only read reviews: Picketty's book, like Herrnstein and Murray's /The Bell Curve/ from two decades ago, appears to be heavily data-driven. Therefore, when and if I read Picketty's book I will be keeping a sharp and skeptical eye out for the following question: what hypotheses are you using your data to evaluate, and which pieces of evidence support some or all of those hypotheses?
--
Deck thyself now with majesty and excellency; and array thyself with glory and beauty.
Cast abroad the rage of thy wrath: and behold every one that is proud, and abase him.
Look on every one that is proud, and bring him low; and tread down the wicked in their place.
Hide them in the dust together; and bind their faces in secret.
Then will I also confess unto thee that thine own right hand can save thee.
I could not love thee, dear, so much,
Loved I not Honor more.
So, again emphasizing that I have only read reviews: Picketty's book, like Herrnstein and Murray's /The Bell Curve/ from two decades ago, appears to be heavily data-driven. Therefore, when and if I read Picketty's book I will be keeping a sharp and skeptical eye out for the following question: what hypotheses are you using your data to evaluate, and which pieces of evidence support some or all of those hypotheses?
--
Deck thyself now with majesty and excellency; and array thyself with glory and beauty.
Cast abroad the rage of thy wrath: and behold every one that is proud, and abase him.
Look on every one that is proud, and bring him low; and tread down the wicked in their place.
Hide them in the dust together; and bind their faces in secret.
Then will I also confess unto thee that thine own right hand can save thee.
I could not love thee, dear, so much,
Loved I not Honor more.
No comments:
Post a Comment