re: #MeToo, Kavanaugh; column Show more
2) I'm not assuming "because X% of these claims were credible, the rest are too". All that tells you is that X% of the claims are credible. Where do you draw the line of significance? If a million people say X, and 90% of them are credible, that tells you something. If 10% are credible, that tells you something very different, but you still have 100,000 victims.
If you discount all of them because some might not be credible, that's just bad science.