Helen, you and Peterson are making a mathematical error here. See page 74 of @causalinf's textbook: http://scunning.com/cunningham_mixtape.pdf …
-
-
Replying to @besttrousers @HPluckrose and
What you are doing is called "controlling for a collider". Doing it means your regression has no causal implications. You are *assuming* no discrimination, not proving it.
1 reply 0 retweets 0 likes -
Replying to @besttrousers @gztstatistics and
I'm not assuming anything. Just saying that the fact that men and women earn different amounts of money doesn't tell us anything about the cause if the only variable considered is gender and we don't account for different jobs chosen and different hours worked.
1 reply 0 retweets 0 likes -
Replying to @HPluckrose @gztstatistics and
You are assuming it by the nature of the regression you are suggesting. That you are unaware of it doesn't make it untrue.
1 reply 0 retweets 0 likes -
Replying to @besttrousers @HPluckrose and
No. The people you are arguing against already know the simple notion that the choices women make could be affected by discriminatory behaviour/systems. Which is why they spend a large amount of time showing evidence that this effect is weak, and other causes are more likely.
2 replies 0 retweets 0 likes -
Replying to @amiguello1 @HPluckrose and
I'm sure they understand that. What they don't understand is that the form of the regression they are advocating for doesn't allow for such choices.
1 reply 0 retweets 0 likes -
Replying to @besttrousers @HPluckrose and
lol no they obviously do, it's a very simple point. They are already in the next step of the discussion. This regression is just to show that the typical point of "being paid less for the same job" is incorrect.
1 reply 0 retweets 1 like -
Replying to @amiguello1 @HPluckrose and
Nope. Such a regression is invalid. Read mostly Harmless Econometrics:pic.twitter.com/RnalEouATe
2 replies 0 retweets 2 likes -
Replying to @besttrousers @amiguello1 and
Yes, yes, you have explained that stats are worthless here. Therefore, your expertise is irrelevant to the problem at hand. We accept that you can't help so you don't need to remain in the conversation.
1 reply 0 retweets 0 likes -
Replying to @HPluckrose @amiguello1 and
Stats are not worthless- the specific regression you were proposing can not do what you wanted it to do.
1 reply 0 retweets 0 likes
OK, then, but, in reality, it does matter whether the pay gap is caused by women working less and working different jobs or by discrimination so stats are not of any use for discussing this. You can stop talking about them any time now.
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.