PDA

View Full Version : untitled's Aggregate Results and Analysis - 2012



untitled
02-22-2012, 05:48 PM
Hello all,

Now that there are some results coming out, I'm updating the aggregate results page (the index is here (http://tmanalysis.econwiki.com/)). That means for schools where results exist, you can look at what has been posted to TM. Also you can watch current results change the elo difficulty of admission, here (http://tmanalysis.econwiki.com/analysis/difficulty.html) (remember, this is not difficulty of the program, just of the admission). Over the past few years most of my results have become stable. I am confused if GRE scoring changed, it sounds like it didn't, although the maybe the test has become more difficult? Or scoring an 800 is more difficult? So that will change the graphics a bit, but I expect the elo scores to remain stable (although I love it when.

Also, I have updated the much maligned flyout ranker (http://tmanalysis.econwiki.com/flyouts/flyout.html), which scrapes and aggregates the bluwiki flyout (http://bluwiki.com/go/Econjobmarket) pages.

Please let me know if you find any mistakes. I usually try to fix them. And do comment, I always enjoy hearing ideas about different things I could do to make things better. Best of luck to everyone.

delimitless
02-23-2012, 04:25 PM
Thanks very much for continuing with this project, untitled.

Also, I'd like to encourage everyone to post their profiles in the roll call section to better improve the data. While I wouldn't quite call it a moral hazard, posting your profile is a good way to give back if you've benefited from the site.

apo
02-23-2012, 05:45 PM
thanks untitled. you are doing great job!

mneconomist
02-23-2012, 08:00 PM
such a great thing. highly appreciate your work!

buzios
02-23-2012, 09:10 PM
Hello all,

Now that there are some results coming out, I'm updating the aggregate results page (the index is here (http://tmanalysis.econwiki.com/)). That means for schools where results exist, you can look at what has been posted to TM. Also you can watch current results change the elo difficulty of admission, here (http://tmanalysis.econwiki.com/analysis/difficulty.html) (remember, this is not difficulty of the program, just of the admission). Over the past few years most of my results have become stable. I am confused if GRE scoring changed, it sounds like it didn't, although the maybe the test has become more difficult? Or scoring an 800 is more difficult? So that will change the graphics a bit, but I expect the elo scores to remain stable (although I love it when.

Also, I have updated the much maligned flyout ranker (http://tmanalysis.econwiki.com/flyouts/flyout.html), which scrapes and aggregates the bluwiki flyout (http://bluwiki.com/go/Econjobmarket) pages.

Please let me know if you find any mistakes. I usually try to fix them. And do comment, I always enjoy hearing ideas about different things I could do to make things better. Best of luck to everyone.

Here's to many more years of applicant anxiety!

econsd
02-23-2012, 09:33 PM
This information has been very informative when I was applying this year. I hope our results help your analysis.

gemmy
02-23-2012, 09:41 PM
May I ask what 'flyout' is? I'm sorry;;;

untitled
02-24-2012, 04:06 AM
May I ask what 'flyout' is? I'm sorry;;;

It is the third-ish step in the job application process. First you apply to about or over 100 schools. Then you interview at a regional or national or field conference (usually the big AEA conference which is often held in Chicago over winter break - each school with an opening may interview 20-50 candidates, sometime sitting on beds in hotel rooms). Then they fly 3-8 applicants to the school to give a job talk (this is the flyout). Flyouts are posted by candidates sometimes, it is an extremely stressful process, there is some bad information and lots of bias in those postings, so be careful with that ranking.

resource
02-24-2012, 12:25 PM
It is the third-ish step in the job application process. First you apply to about or over 100 schools. Then you interview at a regional or national or field conference (usually the big AEA conference which is often held in Chicago over winter break - they interview 20-50 candidates, sometime sitting on beds in hotel rooms). Then they fly 3-8 applicants to the school to give a job talk (this is the flyout). Flyouts are posted by candidates sometimes, it is an extremely stressful process, there is some bad information and lots of bias in those postings, so be careful with that ranking.

It should be noted that there is also a 'flyout' in the Phd admissions process, too. This is an all/some expenses paid trip to visit a phd program to which you've been admitted to get a feel for the department. These usually occur sometime in march. Details are usually provided when/after you are admitted.

apo
02-24-2012, 12:52 PM
It should be noted that there is also a 'flyout' in the Phd admissions process, too. This is an all/some expenses paid trip to visit a phd program to which you've been admitted to get a feel for the department. These usually occur sometime in march. Details are usually provided when/after you are admitted.

Is it really paid? Do all universities have fly-outs? So If i am from another country, universities can pay all or some of the expenses?

untitled
02-24-2012, 02:23 PM
Someone noticed that if a profile doesn't label GRE scores as verbal and quantitative (so my scores could have been written 600/800/5.5 and 590/800/6), the code guesses the first number is the Q score, which in my case is obviously incorrect (no econ PhD applicant is going to be that good at verbal and bad a quantitative and get in to a school in the top 75 like UW). One work around is to always take the larger number, which assumes econ students are better at Q than V. I have heard that an 800 V is nearly impossible, so I set it so if one score is 800, then that must be the Q. Should I have it look for the larger number? Should I have it do something else?

exs
02-24-2012, 03:39 PM
Someone noticed that if a profile doesn't label GRE scores as verbal and quantitative (so my scores could have been written 600/800/5.5 and 590/800/6), the code guesses the first number is the Q score, which in my case is obviously incorrect (no econ PhD applicant is going to be that good at verbal and bad a quantitative and get in to a school in the top 75 like UW). One work around is to always take the larger number, which assumes econ students are better at Q than V. I have heard that an 800 V is nearly impossible, so I set it so if one score is 800, then that must be the Q. Should I have it look for the larger number? Should I have it do something else?

I'm not sure about using the larger number: on the new scale, my verbal score was higher than my quant score. (How're you handling the new scores, anyway?)

Of course, I'm at a loss for a more constructive suggestion. I guess you'll have to balance the possibility for the kind of error you're worried about (reversed Q and V scores) versus making a new kind of error (perhaps untenable assumptions about the relative Q and V scores of Econ applicants).

to2012
02-24-2012, 03:45 PM
My Q and V scores only differed by one increment; I don't think it is that outlandish that there would be occasional applicants with V > Q.

Peot
02-24-2012, 05:03 PM
untitled, my profile appears as my message encouraging people to post in Roll Call... I can edit that post if you'd like.

Marshall
02-25-2012, 09:25 AM
Is it really paid? Do all universities have fly-outs? So If i am from another country, universities can pay all or some of the expenses?

Every place is different. Generally, every school has a open house day they invite all their prospective students to come and visit them on. One school has invited me, but given no funding. Another has offered $500 to reimburse me. I have to fly clear across the country so this won't cut it, and obviously even less so for some internationals, but if you're lucky you may get multiple flyout reimbursement offers. Then pool the money and hit all the schools you can in one trip, and it can workout. My understanding is not every school gives money for fly-outs and not everyone will get offers of flyout money.

Untitled I wouldn't worry too much about the whole Q and V issue. It may be messed up a bit some places, but honestly, that's not the most important/useful part of the profiles. I really appreciate all the hard you work you put into this every year.

untitled
02-25-2012, 07:59 PM
untitled, my profile appears as my message encouraging people to post in Roll Call... I can edit that post if you'd like.

Don't worry about it. I'm going out of town next week, but I have a work around for that and should be able to implement it for your post when I get back.

untitled
02-25-2012, 08:06 PM
I'm not sure about using the larger number: on the new scale, my verbal score was higher than my quant score. (How're you handling the new scores, anyway?)

Of course, I'm at a loss for a more constructive suggestion. I guess you'll have to balance the possibility for the kind of error you're worried about (reversed Q and V scores) versus making a new kind of error (perhaps untenable assumptions about the relative Q and V scores of Econ applicants).

Yuck to the new scale. For a while, everyone was posting old scale scores, even this year so I was hoping I could just ignore it and it would go away...

Eventually I'll have to convert all old scale scores to the new scale, I think. During the transition I am missing lots of scores because people put so much information in (such as score ranges and explanations of the new system) and my script can't read it. I think that will persist, but next year when more people post new score range scores I'll have to take some time and make sure that the scraping is still pretty accurate.

As for figuring out which number is Q versus V, if your profile tells me by including the letters Q (or M for math, some people used to do that, I think) and V, then it will be fine. The problem is when people just post numbers I have to guess which is which. So my guess is the first is Q, unless the second is 800, the guess Q is 800. I agree that assuming the Q is the higher score might not be perfect, and I'm just going to go with the 800 rule for now. I think most people include letters, and there is no problem for these profiles.

untitled
03-12-2012, 04:45 PM
I'm updating the aggregate results again (I did it a couple times last week, so there isn't a big change). As people are making their decisions, I wanted to point out a couple of tools. Along with the Publish or Perish (http://www.harzing.com/pop.htm) tool and whatever use you might find from increased use of google scholar personalized pages (http://scholar.google.com/citations?hl=en&view_op=search_authors&mauthors=economics) among publishing authors, RePEc has produced a nice tool for exploring connections between professors called CollEc, discussed here (http://blog.repec.org/2012/03/11/a-new-repec-service-collec/). I'm still very unhappy with RePEc due to the number of authors that do not take part, and actually suggest trying the American Mathematical Society collaboration distance page (http://www.ams.org/mathscinet/collaborationDistance.html) for a similar effort (certainly the AMS version is less effective than CollEc for most links between economists).

untitled
03-12-2012, 10:01 PM
I just received a note that some of my NC State results were incorrectly given to UNC. This error stretches to all my results, so North Carolina has generally been seen as less selective than it really is, moving up about 15 spots to tied for the 77th most selective program. Thanks to everyone who finds bugs, and to that user in particular.

I also want to remind everyone that people use the aggregate results quite a bit, and it is very generous and helpful for people to post their profiles, and not just their results. As we approach April 15, it is common for participation on these forums to drop, so try to post your profiles soon, in case you get busy elsewhere and forget to do it. That said, posting profiles and results once you have everything is even better if possible.

Thanks again everyone for your help, and please let me know if you have any issues.

untitled
03-15-2012, 06:38 PM
Hey all.


Congratulations to everyone as always. I hope the best to you all. I just wanted to let everyone know that I made a tweak to the flyout ranking based on a problem with people with the same name listed in different years (not necessarily the same person, though). There aren't a lot of differences, but one that I think is interesting is that Cornell has snuck into the top 20. While this may say something about Cornell, it says more about the fickleness of this ranking (which extends to all rankings).


As for the aggregate results, let me know if you see any problems. Good luck to everyone with the rest of your results and on your decision.

untitled
03-22-2012, 05:30 PM
Thank you everybody for all of your results and keep them coming. I just ran the script and updated the pages, please let me know if you see anything wrong.

These days my favorite thing to look at is the selectivity/difficulty ranking. Since my last update, we've had a number of new results, and two trends have shown up. One, Tilburg was more selective this year than in years past. The other is now we have enough results from Florida State that they are ranked by selectivity, and they come out decently selective, coming at 101, right behind the University of Washington and Oregon State.

econsd
03-22-2012, 05:45 PM
Hey untitled, UNC is still picking up NCSU acceptances for 2012.

untitled
03-22-2012, 10:03 PM
Are there any UNC results so far this year?

econsd
03-22-2012, 10:07 PM
I haven't seen any on TM and the applicants got an email saying we should get our decisions on the 26th or 27th.

untitled
03-22-2012, 10:12 PM
Hmm, I guess the graphic doesn't update to NULL, it can only be replaced once I get new results I think. I'll delete it, so there isn't any confusion.