Do you know of any studies that show that female _applicants_ are less likely to be _hired_ for higher paying jobs? I. e. not just that men are more likely to _work_ in higher paying jobs.
Scroll Down to Read Other Opinions
You miss so many explanatory variables that calling your post "misleading" would be an understatement.
This link will actually work:jamanetwork.com/.../2532788
It boggles my mind that someone would downvote -without comment!- a post that simply reports the findings of a published study. Maybe I'm naive, but I would assume that if someone thought that they had valid reasons to believe that the study wap crap they would state those reason. It seems that some people just don't like actual evidence.
As the old saying goes, the plural of "anecdote" is not "evidence".
If that's true, you shouldn't have any problem enlightening us as to how the authors of this study screwed up: jamanetwork. com/.../2532788
Umm... that should be jamanetwork.com/.../2532788
@JenSCDC That study makes absolutely no sense, it does not take into account the experience of the physicians e. g. a 25 year old physician vs a 50 year old physician, also does not take into account there are more male physicians than female physicians, so obviously the more male physicians with more experience earning more money will make the difference seem like they get paid less. Obviously if this were true they could sue their companies as their companies are breaking the Equal Pay Act of 1963
"it does not take into account the experience of the physicians e. g. a 25 year old physician vs a 50 year old physician"It actually _shouldn't_ include experience because it includes clinical billing and publication record. I don't know if billing rates would be affected by experience, but publication record certainly would. I. e. the effect of experience is indeed included in the study. Because of the dependence of publication record on experience, including a separate variable for experience would be a bad thing because it would introduce multicollinearity."also does not take into account there are more male physicians than female physicians" Huh?
@JenSCDC You didn't know that more men go into the physician profession than women?
No, it's that I didn't see the relevance. As long as there's a decent number of women in the study it's not a problem.
@JenSCDC It is a problem... more men means more men getting promotions, then paid more. Even men are more likely to get promotions anyway because women are less likely to ask for one this just makes the likeliness even greater. Also there are also less women physicians because most women after they have a child will settle down and become stay-at-home wives.
"It is a problem... more men means more men getting promotions, then paid more."That makes absolutely no sense whatsoever.The study used regression to analyse the data. When you do a regression analysis you learn what effect each explanatory variable has on the outcome variable given the effect of the other explanatory variables. The number of data points is only important in that it increases the precision of your estimates of the sizes of the variables' effects. The male/female ratio would only be relevant if it were extremely high. While I'm sure that the ratio is on the high side, there's no way that it could be anywhere close to high enough to be a problem.
@JenSCDC On average, women in the physician field are younger, less likely to be full professors and have fewer scientific publications than men. Also, on average they work less hours than men. However, there are studies that say you are less likely to die if you are treated by a female physician but at the same time they also treat less patients on average. It's no surprise that if you do less work, have less experience, less hours you'd get paid less.
Crap. I must be doing a lousy job of explaining things. Let's see if this makes sense:The ideal study -of anything, in any field- would only consist of subjects who are absolutely identical in every way but one. Obviously, that's impossible, so let's move on to the next best scenario- there are only two ways in which the subjects differ, and for each of two ways there are only two possible values. I. e. we have variable #1 and variable #2, and #1 can be only either A or B and variable #2 can only be C orD.We obviously know what the average value of outcome variable -let's call it Y- for all subjects where variable #1 is A, the average when #1 is B, #2 is C, and #2 is D. But if Y is larger when variable #1 is A than when it's B, do we actually know that this is because variable #1 is A? No, because we're ignoring variable #2, and so we don't know its effect on Y. (continued)
However, it's possible to use math to calculate variable #1's effect on Y _after_variable_#2's_effect_has_been_accounted_for_.The is what the authors of the study have done- they've been able to isolate the effect of sex on income. It's that effect which is being studied, not the actual incomes of the doctors.Am I making sense?