Reminds me of this essay by Scott/Yvain where he mentions a reddit thread of over 10,000 comments specifically looking for people who opposed gay marriage, but with practically nobody who opposed gay marriage participating.
A great reminder that while increasing sample size decreases random bias, it does nothing to reduce systemic bias.
True. Hmm...
Reminds me of this essay by Scott/Yvain where he mentions a reddit thread of over 10,000 comments specifically looking for people who opposed gay marriage, but with practically nobody who opposed gay marriage participating.
A great reminder that while increasing sample size decreases random bias, it does nothing to reduce systemic bias.
True. Hmm...