3 ms·
"The population is not uniform, yet our sample is? SOMETHING IS WRONG." http://www.empiricalzeal.com/2012/12/21/what-does-randomness-look-like/ http://www.emp
by invalidOrTaken 14y ago
"The population is not uniform, yet our sample is? SOMETHING IS WRONG."
http://www.empiricalzeal.com/2012/12/21/what-does-randomness-look-like/ http://www.empiricalzeal.com/2012/12/21/what-does-randomness...
Finding a cluster doesn't necessitate a bias.
- naftaliharris 14y agoWell, possibly. Let's compute some simple p-values: Suppose that some fraction p of potential speakers are female. Then the probability that all 22 of the speakers are male would be (1-p)^22. For p = 0.1, we have a p-value of 0.098, which could reasonably arise by chance. But for p = 0.20, we have a p-value of 0.0074, which is not so likely to arise by chance. So if you assume that at least 20-25% of potential, qualified speakers were female--a reasonable assumption, I think--then this "cluster" does indeed suggest a bias.