Using nonrandom data, or in other words biased data, is the most basic kind of error a study like this can commit. It is not up to us to show that the salaries of the one category of people that the Feds used here are not representative of the whole class of college graduates, it is up to them to show that it is. Short of that, there is no choice except to be highly suspect of that data as being a fair representation.
@fallenchemist, I don’t see the website making the claim that the data is fair or representative. Seems to me like you and others may be projecting your feelings on to what is simply data.
I’m assuming they already know that they have data-challenged educators in charge of the rollout. 
Excuse me for saying this, but you must be joking. The GOVERNMENT is putting this out as a scorecard, as an authoritative way to compare colleges and other institutions. What in the world is the point of such an exercise if the data is no good, i.e. not fair or representative? Especially something that carries the imprimatur of the Federal Government? It would be beyond naive to think that the vast majority of the public isn’t going to assume that because it is a study and publication of the Department of Education, it can be assumed that the data is valid. They know that.
@fallenchemist, BTW, salary isn’t the only criteria that I see on that website. It seems to me that you can sort by a few more if you want to.
As for government imprimatur, etc. . . . Maybe it’s because I tend to be irreverent, bur I just don’t see what the big deal about the data being from the government is.
As for the data being good; that seems like a judgement call. For those folks who want to know how much money those students who get Federal loans make on median, the data is fine.
Maybe it’s because I’m laissez-faire when it comes to data and ultimately trust people to be fine handling data (if you’re libertarian, don’t you almost have to have that belief? If you believe that people are easily misled, then does that not call for a higher authority to tell people what is good and what is bad?)
PurpleTitan, people want to know how much money grads make, not just those who got aid. I presume the gov started the study to quantify the value of college education. They chose earnings to represent an outcome. Even if you accept that earnings represent success, the study doesn’t give you much since their earnings are from a biased sample.
A few? I see graduation rate, which has been available for a long time. And is out of date already, btw. They are not even using their own latest data.
As for the rest, I already stated how I think, based on many years of watching how people react to government studies and position papers, this will be thought of by the general public. Laissez-faire really doesn’t cover propaganda, lying, misleading, sloppy, incorrect, and all sorts of other categories when it comes to this kind of thing. FYI, here is the first definition of laissez-faire from dictionary.com and I think you will agree that makes your use of the term particularly ironic.