Things have been a bit quiet here because I have been busy working on an expansion of phds.org: a new set of graduate school rankings. Take a look!

Here is what's new:

• Program profiles - We now have detailed profiles of ~6,000 graduate programs at 418 US universities; altogether there are more than 25,000 pages of data!

• Outcome data! - For years, people (myself included) have been calling for programs to make data on post-PhD outcomes available to prospective students. Relatively few programs collect this information, and even fewer make it available. However, it turns out that the National Science Foundation has been collecting outcome data for decades, but no one has ever looked at it at the program level. So for the first time ever, we have detailed, program-level information on placement rates at graduation, employment sector for recent graduates, time to degree, source of funding, graduate debt, and more.

• More data - In addition to the outcome data, we also have comprehensive information on costs, student demographics, and more.

• New UI - The user interface for the rankings has undergone a major overhaul. It's now even easier to generate rankings designed just for you.

So why do this?

There have been a number of papers that show that the US News rankings influence university behavior. See, for example, Monks and Ehrenberg and the introduction to this great paper by Avery et al. (they also cheat -- see this Wall Street Journal article ). If universities respond to rankings by changing their behavior, why not introduce elements into the rankings that encourage positive changes? Student outcomes should not be very difficult to improve - investments in better career services, processes that get students to think about their careers earlier in the PhD process, and curricula that better match what employers are looking for should all pay off. And the results benefit everyone: students get good jobs, departments get happy alumni, and the country gets scholars who are productively employed.

Another reason is to see if other providers of rankings can be influenced. If, say, US News decided to copy some of these ideas for their own rankings, then again, everyone comes out ahead (except for me, perhaps).

US News's rankings come out tomorrow - this should be interesting!