[velis_vel]

> Regression analysis was used to estimate wage differences, after controlling for the following choices and characteristics: graduates’ occupation, economic sector, hours worked, employment status (having multiple jobs as opposed to one full-time job), months unemployed since graduation, grade point average, undergraduate major, kind of institution attended, age, geographical region, and marital status.

The problem with controlling for all those things is that you leave out other factors; for example, if men were preferentially hired over women, that wouldn't show up in this data.

[ameister14]

That's true, but we're talking about equal pay for equal work.

Such a case would definitely be discrimination, it just wouldn't be germane.