College ROI

Payscale has an attempt to calculate ROI of a college education. The most significant thing you can take from it is:
If you are going to stop with a Bachelors degree, make sure it is in Engineering. 9 of the top 10 schools are engineering. And number 10? Stanford, a nice elite school sitting in the heart of Silicon Valley.

Other than that, you probably have more questions than answers.
In the top 20, a few non-engineering schools appear. They are located in Boston, Silicon Valley, Los Angeles, and New Hampshire. So, it pays to go to a school in a big city where you can get big wages, or to at least go to a rural ivy where all the jobs are in the big cities.

A degree that pays well makes sense. After all, the cost of college is very similar, no matter what you study. I think there were some small additional "lab fees" or increased tuition costs for engineering and science classes when I went to college. However, even then (when a $1000 scholarship easily covered a year's tuition), the fees were barely a rounding error in the total cost of tuition.

A university located (or connected to) a large job market is also a big plus. Big City U may charge a little more for tuition than Podunk U. However, Big City U grads have plenty of high paying big city jobs. Podunk U may have a few recruiters come from the big city, but most students will stay in podunkia.

Some lawyer friends described their law firms as focussing recruiting in the Ivy League, but getting one or two of the top students from the local university. Similarly, some of the big consulting companies will focus their recruiting on the prestigious universities. If you are going to be paying similar private college tuition, might as well make it count. Alas, not everyone can get in there. However, they can choose a lower tier college that has a good location or employment connections.

The methodology of the study had a gazillion holes. Tuition and using only terminal bachelors are just a couple. The total tuition cost was estimated to be higher for people who took longer to graduate. However, many of these students took a year off, or went to school part time. Thus, total tuition costs would be similar to regular graduates. I graduated in five years after starting. However, I took two years off for a mission, then finished up by doing a study abroad my last summer. (I even debated going back for one more semester, just to use my full four year scholarship.)

Including only people with "terminal" bachelors degrees is also a significant issue. At some schools, most people may end with a 4 year degree. However, the top students may go on for more education. This would cause stats to look worse. Alternatively, a school may send most people to graduate school. Only people that had a great career would not pursue more education. (And how would people like the Google founders be treated? They went to grad school, but didn't finish. And Mr. Facebook? He never graduated from Harvard, but if he didn't go there, he wouldn't have started the company.)

These, and many other problems made the calculation of ROI and ranking of individual schools very error prone, and not terribly reliable. However, the big picture observations look more reliable.