Emil O W Kirkegaard

@KirkegaardEmil do you have s factors computed by county somewhere handy?

@RCAFDM No, but I gave you the data before, so you can calculate them yourself.

@KirkegaardEmil didn't you say the data needed to be adjusted for age or something?

@RCAFDM Yes, but I don't have age data. Remember, I asked you to dig it up. ;) You probably need non-linear adjusting.

@KirkegaardEmil but yeah.... you think adjust against median age or something else?

@RCAFDM More variables than one and use LOESS.

@KirkegaardEmil you mean normalize against multiple age distribution points or against multiple S-factor variables (edu, $, etc)?

@RCAFDM @KirkegaardEmil didn't Andrew Gelman say you could just weight equally by each year of birth and it would accomplish the same thing?