
(((tedfrank))) @tedfrank
@lefineder @monitoringbias Yes, 16 cases in the databank, & they’d disproportionately consist of people in the subpopulation. Might be something silly like 3/40 vs 13/450,000 in which case the small sample size is a good reason to withhold. More info needed, including whether cousin marriage confounds. — PolitiTweet.org