Chris Hibbert writes:
>Look at the numbers for Drinking and for weight! Heavy drinkers and
>overweight people have the lowest mortality! Underweight people and
>teatotallers have higher mortality. That's directly contradictory to
>all the reports we've been seeing which would lead us to believe that
>underweight might be better than "normal" (i.e. average?), which
>should be much better than overweight. Moderate drinking is supposed
>to be better than teatotalling, which is much better than heavy
>drinking. Can these numbers be right? Or am I misreading the whole
>table?
You're not misreading it. But I didn't include the confidence intervals from the paper. The 95% confidence interval for the heavy drinking coeff of .85 is .46 to 1.59, and for the overweight coef of .94 is .72 to 1.23. You shouldn't take small differences very seriously here.
Robin Hanson
hanson@econ.berkeley.edu http://hanson.berkeley.edu/
RWJF Health Policy Scholar, Sch. of Public Health 510-643-1884
140 Warren Hall, UC Berkeley, CA 94720-7360 FAX: 510-643-2627