On Tue, 6 Feb 2007, Michael Scoles went:
Demographic information regarding subject pools, including information on gender and ethnicity, is a common component of grant applications.
Yeah, but the ethnic categories don't make a lot of sense. NIH follows the bizarre standards of US Census Bureau, coding "race" and "ethnicity" orthogonally (i.e. independently of one another). Race is pretty much what you'd expect. But ethnicity is coded only as "Hispanic" or "not Hispanic"! In 2005, I emailed the Census Bureau about that:
Subject: why only two ethnicity categories? What is the logic in recording ethnicity as simply "Hispanic" or "not Hispanic"? Why do your ethnic categories not include, for example, English, Irish, Scottish, Welsh, French, Danish, Swedish, German, Italian, Polish, Ukrainian, Greek, Turkish (and all I've done here is list some of the obvious European ones; there are obviously hundreds more). I infer that you record all such ethnicities as "not Hispanic," though they're just as clearly "not Korean" or "not Australian." So...why have you chosen such a strange dichotomy in what would otherwise be a very rich dataset?
The relevant portion of their reply was:
The Office of Management and Budget sets the standards of race and ethnicity. Questions on the census form are usually always needed for some form of funding enacted by Congress.
OK, I guess, but then the category shouldn't be called "ethnicity." It should be called "Hispanic," taking on a value of either "y" or "n." --David Epstein [EMAIL PROTECTED] --- To make changes to your subscription go to: http://acsun.frostburg.edu/cgi-bin/lyris.pl?enter=tips&text_mode=0&lang=english
