Thing

DISCLAIMER


Welcome! The goal of this blog is to share my analysis of the free, publicly available user-reported law school applicant data from Law School Numbers. Using the data from Law School Numbers is problematic for a variety of reasons (such as users misreporting their actual information, users creating fake accounts, selection bias, etc.) and if I had access to it, I'd much rather work with the data that schools themselves have on applicants. We have what we have, though. Also, while I do have some facility with the type of statistical analysis I employ in my blog posts, I am far from being a professional statistician. I am doing this solely for the purpose of providing my analysis to interested readers, getting feedback, and generating discussion. What I am not doing is prescribing courses of action for law school applicants, or pretending to actually know what goes on behind closed doors in law school admission committees' meetings. I am, however, interested in looking at the story the numbers seem to portray, and sharing that with people with similar interests. I think I'll be able to provide a lot of interesting, and perhaps even helpful, analysis here, but at the end of the day, it is up to the individual law school applicant to put together applications and application strategies tailored to his or her own hopes and goals.

Columbia Law School Profile

Model 1: Waitlists Included Model

In the table below, you will find a list of variables that theoretically play a role in determining whether a law school accepts you or not.  The dataset used in this model includes users on Law School Numbers who reported being either accepted, rejected, or waitlisted at the school in question (those who were first waitlisted and then accepted or rejected are treated as acceptances or rejections, respectively).  For this model, I used an Ordered Logistic Regression, which allows you to look at how each of the independent variables affects the chances that you will be:
  • Accepted rather than (combined waitlisted/rejected)
  • (Combined accepted/waitlisted) rather than rejected
By way of example: in this model for Columbia, for two otherwise identical candidates, an additional point on the LSAT increases an applicant's chances of being accepted rather than either waitlisted or rejected by 44.8% (in other words, a 173 is 44.8% more likely than a 172 to get accepted rather than waitlisted or rejected, all else being equal).  A 173 is also 44.8% more likely to get either accepted or waitlisted than rejected than is a 172, all else equal.  If you have any questions, just e-mail me, or check out this link to the awesome UCLA stats site. 

In any case, here are the results from this first model, in which I test the impact of LSAT score, GPA, each earlier month the application is sent (and by earlier month, I don't mean month earlier...I mean September vs. October, or October vs. November), URM status, non-traditional status, and female applicant status.  Since Columbia does have a binding-ED option, I have included it.  Everything is based on LSN data from the 2003/2004 cycle to the present.

                              

URM candidates get a decent boost at Columbia (although nothing approaching what we see at other schools), being a little more than 8 times more likely to be accepted than an otherwise identical non-URM candidate.  One thing to always keep in mind when interpreting this is that there are almost certainly different "floors" for LSAT and GPA for URM applicants than non-URM applicants.  This matters because below the numbers "floors" for non-URM applicants, a URM is pretty much infinitely more likely to get in.  The URM boost in terms of its LSAT equivalent is pretty standard stuff, at 5.7 points, but the half-point GPA equivalent is pretty big.  The relative boosts Columbia gives for one extra LSAT point vs. .10 extra GPA points is actually pretty even.  Columbia gives no advantage to non-traditional students, but female applicants are 68.4% more likely to get in than otherwise identical male counterparts, which is a relatively big boost. Something else to note is that there is no statistically significant effect of ED application.   Finally, there seems to be a definite advantage to applying early to Columbia, with your chances of admission increased by almost 15% for each month earlier you apply.

Model 2: Waitlists Excluded Model

The next model excludes waitlists, including only applicants that reported either being accepted or rejected (whether that was directly, or after first being waitlisted).  The results in the table should be interpreted in the same way, but the interpretation is a little easier.  The number given for each variable is simply the increase in likelihood of being accepted rather than rejected.

                              

When we drop waitlists and only look at straight accepts or rejects, the boosts get significantly bigger, especially the URM boost, with URM candidates more than 40 times more likely to get in than an otherwise identical non-URM candidate.  Again, please bear in mind the caveat from the first model about disparate "numbers floors" for URM and non-URM candidates.  Female applicants are more than twice as likely to get in, all else equal, and the LSAT, GPA, and boosts for earlier applicants also increase significantly.  It's interesting to note that even as the LSAT and GPA boosts increase, though, the URM boost increases even more, as evidenced by the significantly larger URM boost in terms of LSAT and GPA equivalents. Again, no statistically significant effect of early decision applications.

Non-Splitters, Splitters, and Reverse-Splitters: Acceptance/WLs/Rejections, and Means

Last, I'm including a table that breaks down how non-splitters, splitters, and reverse-splitters are represented in the data.  This one you really have to be careful with, because the data on LSN does skew towards higher-caliber applicants, and so acceptances are more highly represented than they are in the applicant pool.  Really, the value of this kind of thing will become more clear when we can compare schools, because that same "higher-caliber" applicant caveat will apply across the board, so we can probably draw somewhat valid conclusions by comparing schools.  For now, I'll include it for interested parties, but please do not look at this and say to yourself, "Self, as a splitter I have an X% chance of getting into Columbia!"



Columbia waitlisted a pretty significant percentage of its applicants, especially splitters and reverse-splitters (check out that waitlist percentage for reverse-splitters!)  This is probably to make sure numbers balance out to maintain medians as people withdraw.  The trends for LSAT and GPA are pretty predictable for non-splitters, with numbers declining as you move from admitted to waitlisted, and waitlisted to rejected.  For splitters and reverse-splitters though, it's pretty clear that once you've reached the 75th percentile for the score in question (LSAT for splitters, GPA for reverse), that score itself doesn't matter nearly as much as your below-25th number, as you can see by the fact that splitters' LSATs actually increase from the "accepted" to the "WL/rejected" categories, but their GPAs drop pretty precipitously as you move across.  For reverse-splitters, it's the other way around...the GPAs are pretty standard across the categories, but the LSATs drop (although the drop between accept and WL is relatively small).

ED for Splitters and Reverse-Splitters, non-URM and URM tables

Because people are often curious as to whether they should ED as a splitter or reverse-splitter, and because the number of both of those groups that actually do apply early decision is so small that you can't get meaningful results through regression analysis, I am including the following tables that gives the number of observations for each group for RD and ED, the numbers and percentages of each group broken down by outcome, as well as the mean LSAT and GPA for each group broken down by outcome (accept, waitlist, reject). The first table on tabulates non-URM candidates, and the second URM-candidates. I am doing them separately because they are primarily helpful because they allow us to compare the mean LSAT and GPA of each candidate type between ED and RD applications, and lumping them all in together would throw off those means (because we expect the means for URM candidates to be lower). Here they are:


Non-URM Applicants 



It's so hard to say anything about this with so few instances of splitter and reverse-splitter ED.  I mean, we can definitely say that no splitters who applied ED to Columbia were rejected, and 4 out of 5 were accepted, both those who were accepted also had almost a point higher on the LSAT, and marginally lower GPAs than splitters who applied regular decision.  In terms of percentage of success, reverse-splitters who applied early decision also did better than their regular decision counterparts, but again, they also had significantly better numbers. If I were a betting man, I'd say that those splitters and non-splitters who applied ED would have gotten in even had they applied RD.  And, when you run the numbers, there is no statistically significant effect of ED application for non-splitters (although if you eyeball the numbers, it seems like although ED applicants had less success, percentage-wise, they generally did get in with lower numbers than RD candidates).  If and when I can get any kind of an idea on how ED affects scholarship outcomes, I will make sure to post it here, because I think that could play a major role in deciding whether or not to apply ED.

URM Applicants 



If it was hard to say anything about ED for the non-URM table because of lack of data, it's next to impossible to say anything about the URM table.  There isn't a single URM splitter who applied ED in the data, and there was only one URM reverse-splitter (he/she got waitlisted with essentially the same GPA and a much higher LSAT score than the average URM regular decision reverse-splitters,  so it's pretty likely that applying ED gave no advantage here).  When we look at the non-splitter ED applicants, the four who were accepted squeaked in with LSATs that were considerably below the mean for RD applicants, but the GPAs were the same.  Since there were only four of them anyway, I'm not sure any meaningful conclusions can be drawn.  And, again, ED applications had no statistically significant effects in the regressions I ran in models 1 or 2.

So, in sum, at the point that your LSAT and GPA are set in stone, the best you can do is apply early, at least in terms of the factors we look at here.  Of course, there is a lot that these models don't account for, so I think it would be quite unwise to prioritize applying early over crafting the strongest possible application, because factors such as personal statements and letters of recommendation obviously matter.  But you've got all summer, right?  Have that stellar application ready to go on the first day applications open, if you can.

No comments:

Post a Comment