Training Log Archive: iansmith

I listened to a delightful NPR "On Point" interview with economist Emily Oster about her book about analyzing data concerning what actions should and should not be taken by pregnant women. Also in attendance was Dr. Jeanne Conry, the president of the American Congress of Obstetricians and Gynecologists. In general, I found Oster's reasoning much more compelling and substantive than Conry's. For instance, concerning whether to consume deli meat and the risk of listeria, Conry cited a 20x increase in pregnant women in the chance of infection over non-pregnant people as an argument for avoiding deli meat. Oster's reply essentially remarked that the probability is a product of likelihood and the prior, and for sufficiently small priors, an increase in likelihood has negligible effect on expectation value. Pwn'd. My n = 1 conclusion is that economists are better at decision theory than obstetricians.

C • No one... 2