How reliable are the placement statistics on phds.org?

[gooddogeverywhere]

Hi everyone, I am starting my PhD this fall. I have got a number of offers but haven't visited any yet. I would like to look into schools early, though.

To find out about the placement statistics, I have to go to the group websites and check where the alumni ended up. Most programs do not post such information. Recently a friend of mine tells me about a website called phds.org, which provides nice placement statistics like (see the "outcomes"):

http://www.phds.org/university/wisc/program/ranking/chemistry/3771

I really wonder how reliable this website is. Where are these data come from?

Edited January 20, 2016 by gooddogeverywhere

[Eigen]

Placement statistics can be telling, or can be worthless information. 

From the outside looking in, you have no idea if the people that weren't placed were successful or not. As my department chair likes to say "We will advocate as well as we can for you- but you have to give us material to advocate with".

In other words, you can't infer your likelihood of placement from stats alone.

[St Andrews Lynx]

I'd deal with the placement issue on a lab-by-lab, rather than a school-by-school basis. It's hard to tell from broad statistics whether (I) there are some labs that are better at placing their graduates than others (ii) if the people who wanted to get a tenure-track academic job (for instance) were able to do so. Some PIs have good connections with industry or seen as good breeding/training grounds for future faculty. 

You can look on group websites or LinkedIn to try and find out if people found good jobs quickly. And try to ask grad students at the recruitment weekend (they're more likely to be honest and actually know about placements than the admin or PIs). 

[Argon]

The information on phds.org is probably not something to believe at face-value, in my opinion. Firstly, the data is quite old. It's from 2010. That's 6 years ago. A lot can change in the span of 6 years at a university. However, the data itself is legitimate (even if it's out-dated now) and is collected directly from the institution by a non-profit organization. 

Equally important, while some statistics on phds.org are very insightful, some are less informative. You said you're interested in the placement statistics. However, phds.org doesn't tell you how many people got what they wanted. For instance, in the link you sent, 65% of the people in the program went in R&D. However, did only 65% of the people want to go into R&D? If so, that's a 100% success rate. The success rate is not present on phds.org though. So I don't think those kinds of outcome statistics are very relevant. 

This is especially true because it varies much more on a lab-by-lab basis than a university basis. As others have suggested, you should check out the "People" or "Alumni" page of group websites. Sometimes you will see where they ended up. If a page doesn't have this (or even if it does), it is worth speaking with the professors you're interested in working with and inquiring about where their students are placed. That is likely your best bet.