A gov’t scorecard system that advises poor students that Harvard is the most affordable school for them, is about as useful as advising zoos that they should all get a panda or unicorn exhibit. The probability of it happening is remote to none.
The devil’s in the details. I’m afraid new parents & applicants won’t have the experience to parse the data correctly. You really have to click on the blue “i” circles to get a fuller picture. For example:
*the scorecard only lists federal student loan debt under “Typical Total Debt;” it excludes parent PLUS loans and private loans. Thus, it should not be taken as a picture of the typical total debt. Parents who took out a home equity loan wouldn’t show up, etc.
*Graduation rates are for first time, full time students. I assume transfers don’t show up as graduates. A college which accepts transfers may have a better graduation rate than published. Some (many?) public universities accept transfers from community colleges. If there are a core of community college transfers who choose nursing or engineering, for example, such data reporting would seem to depress the graduation rate as well as the figures for graduate income.
Do they count community college transfers once, for the original community college, or for the final college?
I do think, though, that the tool could be useful once families have a rough starting group of colleges. It is eye opening to see the different costs by income levels for different colleges. It is especially poignant to see how much for-profit colleges cost, in comparison to non-profit colleges, both private and public.
When I search schools by name, nothing comes up!
What value judgement? I just see data. How people interpret that data is them applying their own value judgement.
@panpacific, ??? But there are many classes. Take in many classes and the sample size isn’t small.
OK, it works for me in Firefox nut not in Explorer! ???
BTW, when comparing salary data, regional cost differences should be taken in to account as well as potential major differences. I don’t have a problem with only using a dataset that has students who were Federal loan borrowers, though. Arguably, how successful those who come from wealthy backgrounds do isn’t very relevant to the rest of us, and that percentage can vary a lot by school.
Those numbers should be taken for the same age, though (say, earnings at 30 or 40), and homemakers and those who don’t work will bring the median down.
Deciding what data is worth including is the initial value judgment, and that filter is going to often be invisible to the users of the site.
@dfbdfb, true of any data, and certainly something that anyone looking at data should keep in mind.
Easy to say, but if the intent is to target uneducated adults and teenagers, the feds have a duty to make clear exactly what this data is – AND isn’t.
They can’t just sit behind their educated gubmint tower and assume everyone has passed AP Stats, in English.
The feds should have a higher threshold than say, the marketing/advertising firms on Madison Avenue.
Do the data include professional schools? It seems to show medical schools. I would think that would affect the earnings greatly.
Misrepresentation is not information. Stats should be chosen carefully to represent the infomation we seek. Isn’t that where we need brains?
I wonder how much the “gubmint” spent on this website? I still prefer College Navigator…
No, @PurpleTitan. The sample size is still small when multiple classes are counted in because the population pool where the sample is taken from increases as well. If 7% of 5000 is considered a small sample size, 7% of 100,000 is small too. Besides, the sample is taken by applying an arbitrary filter, which is students receiving federal aid. While it can be used as what it is, it should not be used to evaluate the college as a whole, which seems to be what the tool is meant to be.
@panpacific, ever taken a stats class? Note that polling organizations don’t poll anywhere close to 7% of the voting population yet can do a decent job of predicting results within a certain margin of error. Now, there may be structural issues with using only the population taking out Federal loans (if some schools provide their poorest students with 100% grants and other schools don’t), but I think it’s a better comparison of like with like than if full-pays are mixed in as well.
BTW, where are you getting that 7% figure from, again?
^Unless they did a study that shows earnings of students receiving federal aid represent the overall earnings within margin of error, they have no business calling it earnings of graduates. That’s plain misrepresentation. To me, it is the worst example of bureaucratic mind, it didn’t matter it is meaningfull/useful or not, they had to produce something.
@PurpleTitan: Consider that 7% hypothetical. If you think you are an expert in statistics, run with it. One last piece of reminder - this is not a poll.
@panpacific, using a hypothetical number to make a point makes for a rather unconvincing argument. For example, I could make up percentages. 90% and 0.01%. Yet I think we can agree that a median based on 90% is more meaningful than a median based on 0.01%.
@Iglooo, arguably, no median is meaningful (except for bragging rights and determining which schools have lots of unsuccessful grads) as someone likely isn’t going to be the median at all schools.
Post on CC by the U.S. Under Secretary of Education. I guess “We’re famous, sort of :-)”
Not true, actually. Just for starters, 7% of 5k gives a 6% margin of error at a 95% confidence level, 7% margin at 99% confidence. On the other hand, 7% of 100k gives only a 2% margin at 95% confidence, and even at 99% confidence the margin of error is still only 2%—much, much better and more reliable results.
That said, the issue of representativeness of the sample is still a huge issue—those numbers assume that the sample is representative, which may well not be true in this case. But as far as how reliable a sample is by size, the actual number of samples is important along with the proportion of the entire population.
If you want to give input into the website, click on http://talk.collegeconfidential.com/college-search-selection/1810817-we-ve-made-a-lot-of-changes-to-the-college-scorecard-and-we-want-to-know-what-you-think.html#latest.
The US Undersecretary of Education is asking for comments!
@Iglooo I meant my “information is information” statement in a general sense. The weakness of this particular information had already been described and discussed. I agree with everyone, including myself, that this data is poor information and is misleading even for what it purports to represent, not to mention that by making salary essentially the only benchmark for what makes for a successful outcome, it is making a value decision that I think a great many will vehemently disagree with.