The most prestigious law school admissions discussion board in the world.
|
 |
 
|
"Final" employment study draft uploaded
Poast new message in this thread
Date: March 23rd, 2005 11:15 AM
http://www.autoadmit.com/studies/ciolli
Comments welcome. |
 |
Date: March 23rd, 2005 11:16 AM I was pre-sticky. Thats prestigious. |
|
Date: March 23rd, 2005 11:47 AM
"while the average Cravath equity partner earns $2,110,000 per year, the average Hughes equity partner earns only $450,000 per year."
While this is the final copy, when you edit for publication, you should add emphasis on only. :) |
|
Date: March 23rd, 2005 2:09 PM |
|
Date: March 23rd, 2005 2:13 PM
As an unknown you can't take any chances -- especially if you're an unknown, a law student, wrote an empirical piece, *and* the piece includes rankings that might make many low ranked schools look pretty bad. In other words, I had 4 strikes against me before I even submitted to anyone.
Way I see it, I'll probably either get dinged everywhere, or get accepted someplace really prestigious. |
|
Date: March 23rd, 2005 2:31 PM
And what do you mean by accepted? It's a research article. |
|
Date: March 23rd, 2005 2:32 PM |
|
Date: March 23rd, 2005 2:36 PM |
|
Date: March 23rd, 2005 2:37 PM
It might. We'll see. |
|
Date: March 23rd, 2005 3:12 PM |
|
Date: March 23rd, 2005 3:19 PM |
|
Date: March 24th, 2005 1:06 AM |
|
Date: March 24th, 2005 1:08 AM |
|
Date: March 24th, 2005 1:11 AM |
|
Date: March 24th, 2005 1:13 AM |
|
Date: March 24th, 2005 1:15 AM
My article got picked up pretty quickly. It was more 'legal' than yours though. Still, I bet you'll have better luck than expected. Don't be surprised if you get stuck with a TTT journal; usually your first few publications will be in journals no one's heard of, but once you have a few, it'll be easier to get into more prestigious journals. |
|
Date: March 24th, 2005 1:18 AM
How long did it take before you got your first acceptance? I assume you submitted as a student? |
|
Date: March 24th, 2005 1:21 AM
I think it took about a month. Most journals start accepting submissions in early February and making decisions in March-April. I had 10 dings before I got an acceptance. I submitted to 25 journals.
edit- I withdrew from the other 14 journals, so I only got the one acceptance. |
|
Date: March 24th, 2005 1:24 AM
I hope submitting in late March hasn't fucked me over.
Alas, I don't know of any free library I can print from here. |
|
Date: March 24th, 2005 1:28 AM
I don't think it will fuck you over at all. I know my friends on journals at my school are still looking at new articles. I definitely submitted early (early Feb). I doubt a law review will publish yours because it's very non-legal, but an employment journal will probably like it.
One problem I did run into is some journals wouldn't even consider it since I submitted it concurrently to different journals. Did you watch out for that? |
|
Date: March 24th, 2005 1:30 AM |
|
Date: March 24th, 2005 2:16 AM |
|
Date: March 23rd, 2005 12:02 PM |
|
Date: March 23rd, 2005 12:37 PM
This isn't a ranking study, it's a determinants of employment placement study. The "rankings" were just created as a way to measure employment placement so I could do the regression, which was the goal all along.
"School Y is ranked higher than Schools X and Z" isn't very helpful -- what's helpful is knowing WHY School Y is ranked higher than Schools X and Z. Is it reputation? Is it curriculum? Class rank disclosure policies? Class size? Grading system? That's what I'm trying to answer here. |
|
Date: March 23rd, 2005 1:46 PM |
|
Date: March 23rd, 2005 1:51 PM |
|
Date: March 23rd, 2005 2:06 PM |
|
Date: March 23rd, 2005 2:09 PM |
|
Date: March 23rd, 2005 3:21 PM |
|
Date: March 23rd, 2005 2:56 PM |
|
Date: March 24th, 2005 2:01 AM |
|
Date: March 24th, 2005 10:41 AM |
|
Date: March 24th, 2005 10:42 AM |
|
Date: March 24th, 2005 12:16 PM |
|
Date: March 24th, 2005 12:18 PM |
|
Date: March 24th, 2005 12:24 PM |
|
Date: March 24th, 2005 12:26 PM
http://www.ilr.cornell.edu/cheri/
As I said in the beginning, the purpose is for students and other stakeholders to identify characteristics that are positively associated with stronger national employment placement at elite firms. |
|
Date: March 24th, 2005 12:27 PM |
|
Date: March 24th, 2005 12:29 PM
If you've looked at the appendices, you'll see that the traditional HYS / CCN / MVP / etc. categories just don't seem to hold. Reputation isn't everything. |
|
Date: March 24th, 2005 12:31 PM
Anyway, like I said, not questioning or attacking you. Just curious. Good luck to you. |
|
Date: March 24th, 2005 12:33 PM |
|
Date: March 24th, 2005 12:41 PM |
|
Date: March 24th, 2005 4:22 PM |
|
Date: March 24th, 2005 9:11 PM |
|
Date: March 24th, 2005 9:16 PM |
|
Date: March 24th, 2005 6:30 PM |
|
Date: March 24th, 2005 9:43 PM
This comment was in a footnote: "The U group received Vault and PPP scores of 250 – this number is completely arbitrary and has no effect on the rankings, for any number greater than the lowest possible Vault or PPP rank attainable by a firm would accomplish the same thing." I don't the "no effect" comment can be defended. Consider how badly John Marshall's TQS score would suffer by a lower U since over 90% in both regions in which it appears receive this ranking. I think your study does a poor job accounting for schools that don't have a large placement into top firms. |
|
Date: March 24th, 2005 9:50 PM |
|
Date: March 24th, 2005 10:56 PM |
|
Date: March 24th, 2005 10:53 PM
Let me give a simple example. School A has 70% employed in Region X and School B has 50% employed in Region X, and both have a class size of 100. To make it even more simple, let's assume that all these guys are working at the exact same firm, and that firm's Vault*PPP composite comes out to 100.
NOTE: These are unadjusted figures, so in this case, a smaller TQS means greater placement.
Scenario 1: Have Unemployed = 250
School A's Region X TQS would be:
(70*100 + 30*250)/100 = (7000+7500)/100 = 145
School B's Region X TQS would be:
(50*100 + 50*250)/100 = (5000+12500)/100 = 175
Okay, so School A places better than School B. Now, let's change 250 to 2500:
School A's Region X TQS would be:
(70*100 + 30*2500)/100 = (7000+75000)/100 = 820
School B's Region X TQS would be:
(50*100 + 50*2500)/100 = (5000+125000)/100 = 1300
SCHOOL A WILL STILL BE RANKED HIGHER THAN SCHOOL B. Nothing changes. It doesn't matter whether you use 250, 2500, 250000000000, 227, or 892... Just as long as the number is greater than 200, you'll get the same set of ordinal rankings. John Marshall doesn't benefit or suffer depending on what value U has -- since the U value is the same for *every* school. |
|
Date: March 24th, 2005 11:45 PM |
|
Date: March 24th, 2005 11:58 PM
School A's Region X TQS would be:
(70*100 + 30*250)/100 = (7000+7500)/100 = 145
School B's Region X TQS would be:
(50*25 + 50*250)/100 = (1250+12500)/100 = 12750/100 = 127.5
Okay, so School A places better than School B. Now, let's change 250 to 2500:
School A's Region X TQS would be:
(70*100 + 30*2500)/100 = (7000+75000)/100 = 820
School B's Region X TQS would be:
(50*25 + 50*2500)/100 = (1250+125000)/100 = 126250/100 = 1262.5
You are a tool.
|
|
Date: March 25th, 2005 12:27 AM
Any suggestions on how to fix this?
With 250:
F, L, B, D, A, K, C, J, I, N, M, E, H, G, P, O
60.82F 81.29L 93.42B 95.27D 110.12A 130.03K 137.07C 138.97J 139.27I 148.51N 149.37M 152.19E 154.40H 172.98G 179.17P 206.00O
With 2500:
F, L, B, D, A, I, J, K, C, H, N, M, E, G, P, O
182.58F 311.38L 540.75B 548.99D 758.48A 904.87I 956.03J 1022.50K 1032.77C 1156.03H 1162.14N 1184.81M 1279.04E 1542.73G 1602.95P 1883.96O |
|
Date: March 25th, 2005 12:41 AM |
|
Date: March 25th, 2005 12:43 AM
EDIT: Upon much closer reflection, it actually looks like this isn't even that big of a deal -- in practice, all this really does is bring about very slight changes at the margins that only have an effect on ordinal rank, but negligible impact on absolutes. I'll look into seeing if I can come up with a more efficient formula that minimizes this, but honestly it really doesn't seem like it's necessary since the only schools this would have a noticable effect on are lower ranked schools, and those weren't even used in the regression anyway. |
|
Date: March 25th, 2005 1:19 PM
The fact that you fail to recognize the effects your variable have on the TQS scores indicates that you have no idea what the TQS measures. You just cobbled together an opaque formula because you wanted to use regression analysis. TQS score effectively means nothing. Maybe try including natural logs or something, that would look impressive.
You say that ". . . the only schools this would have a noticable effect on are lower ranked schools, and those weren't even used in the regression anyway." Do you say which schools you used in the regression in the study? Didn't you criticize Leiter for arbitrarily limiting his study to the schools he considered national? Isn't what you're doing equally aribitrary and not even disclosed?
|
|
Date: March 26th, 2005 12:20 AM |
|
Date: August 3rd, 2007 4:42 AM |