Re:U.S. sues Oracle, alleges salary and hiring discrimination
by euyyn on Wednesday January 18, 2017 @08:31PM
(#13432182)
If it's just "white males are more reluctant to work for us", then surely that would also show in them hiring more females and more non-whites than expected, as they are more willing to work for them? It's not as if you're either a white male, or you're Indian.
[ Parent
| on HN ]
Re:U.S. sues Oracle, alleges salary and hiring discrimination
by explainthisth on Wednesday January 18, 2017 @08:46PM
(#13432272)
The evidence "group A was hired more than than group B, and the effect was statistically significant" doesn't imply we should find a significant effect when comparing any other set of groups A, B, ~A, C, C^B, etc.
It's generally hard to prove things with statistics without a randomized experiment.