UCLA Admissions Data

(Applications Advice, Letters of Recommendation . . . )
rictheruler
Posts: 27
Joined: Sun Nov 22, 2015 5:33 am

UCLA Admissions Data

Postby rictheruler » Mon Oct 30, 2017 4:30 pm

Randomly stumbled across this link:

https://www.lsac.org/officialguide/2013/lsac_4837.asp

If you scroll to the bottom you can see the number of applicants admitted with certain GPA/ LSAT combos. What's interesting is how far off these numbers are from LSN data and from conventional wisdom. For example, if you look at the 3.5-3.75/ 165-169 range less than a quarter of applicants were accepted, which is weird because that's right in line with this year's class (GPA is slightly below 25th- median, LSAT is slightly below median- above 75th). Of course, we don't know exactly how these numbers were distributed, but if you do a search on mylsn for the same range, it says 65% were admitted.

Why the discrepancy? Does LSN paint too rosy a picture because of self-selection? Can we expect other schools have a similar discrepancy?

User avatar
UVA2B
Posts: 2936
Joined: Sun May 22, 2016 10:48 pm

Re: UCLA Admissions Data

Postby UVA2B » Mon Oct 30, 2017 4:35 pm

To put it simply, those grids are painting an extremely broad brush to raise specific concerns about the chances in a given range. Those ranges of GPA and LSAT are HUGE for admissions purposes. Within each one of those range matches, there's probably a climbing chance of admission within it. So it's really likely nearly every 3.5/165 is getting WLed or denied (possibly with some URM admits), while nearly every 3.75/169 is getting admitted. You have no idea within that data set where the specific data points fall.

There isn't nearly enough granularity in those data sets, especially because we're talking about huge amounts of statistical bunching that would be impossible to draw any decent conclusion without knowing A LOT more about the data set.

lobsicle
Posts: 38
Joined: Fri Oct 06, 2017 1:33 pm

Re: UCLA Admissions Data

Postby lobsicle » Mon Oct 30, 2017 4:38 pm

UVA2B wrote:To put it simply, those grids are painting an extremely broad brush to raise specific concerns about the chances in a given range. Those ranges of GPA and LSAT are HUGE for admissions purposes. Within each one of those range matches, there's probably a climbing chance of admission within it. So it's really likely nearly every 3.5/165 is getting WLed or denied (possibly with some URM admits), while nearly every 3.75/169 is getting admitted. You have no idea within that data set where the specific data points fall.

There isn't nearly enough granularity in those data sets, especially because we're talking about huge amounts of statistical bunching that would be impossible to draw any decent conclusion without knowing A LOT more about the data set.


It is true the cells are not very granular, but, still, only about two-thirds of those with 3.75+ and 170+ are admitted. That is much lower than what you'd expect and certainly not nearly everyone. Maybe YP in some cases?

User avatar
UVA2B
Posts: 2936
Joined: Sun May 22, 2016 10:48 pm

Re: UCLA Admissions Data

Postby UVA2B » Mon Oct 30, 2017 4:43 pm

lobsicle wrote:
UVA2B wrote:To put it simply, those grids are painting an extremely broad brush to raise specific concerns about the chances in a given range. Those ranges of GPA and LSAT are HUGE for admissions purposes. Within each one of those range matches, there's probably a climbing chance of admission within it. So it's really likely nearly every 3.5/165 is getting WLed or denied (possibly with some URM admits), while nearly every 3.75/169 is getting admitted. You have no idea within that data set where the specific data points fall.

There isn't nearly enough granularity in those data sets, especially because we're talking about huge amounts of statistical bunching that would be impossible to draw any decent conclusion without knowing A LOT more about the data set.


It is true the cells are not very granular, but, still, only about two-thirds of those with 3.75+ and 170+ are admitted. That is much lower than what you'd expect and certainly not nearly everyone. Maybe YP in some cases?


It's certainly reasonable that it could be YP, or it could be a litany of other reasons.

I think there is probably a disconnect between what LSN provides in self-reported data and where reality of chances of admissions lie, but that's mostly the self-inflicted nature of self-reported data. LSN is only a weather gauge for chances, not an accurate predictor of an individual's chances of admission. There are regularly people who get in to a given school with numbers you wouldn't expect, while some will not get in for one reason or another (or multiple reasons that compound to cause a WL or rejection).

Data in admissions should be read in context always. We can reasonably believe a 3.75/171 should get into UCLA, there will be people with that number profile who don't get in every single year. Applicants aren't a monolith that can be directly compared solely on two dimensions (even though those two dimensions are by far the biggest dimensions).

rictheruler
Posts: 27
Joined: Sun Nov 22, 2015 5:33 am

Re: UCLA Admissions Data

Postby rictheruler » Mon Oct 30, 2017 4:50 pm

UVA2B wrote:To put it simply, those grids are painting an extremely broad brush to raise specific concerns about the chances in a given range. Those ranges of GPA and LSAT are HUGE for admissions purposes. Within each one of those range matches, there's probably a climbing chance of admission within it. So it's really likely nearly every 3.5/165 is getting WLed or denied (possibly with some URM admits), while nearly every 3.75/169 is getting admitted. You have no idea within that data set where the specific data points fall.

There isn't nearly enough granularity in those data sets, especially because we're talking about huge amounts of statistical bunching that would be impossible to draw any decent conclusion without knowing A LOT more about the data set.


This is a good point, but then I would assume searching with the exact same range on mylsn would yield similar results, no? Unless for some reason there are just fewer 3.5/ 165s and more 3.75/ 169s who post on LSN than who actually applied, the data should be more similar.

And I take your point about admissions chances being about more than just numbers, but I still feel like this discrepancy is larger than one would expect. If you're 170+ and 3.75+, and you're not being YP'd, the conventional wisdom is you'd have to really screw something up in your application not to be admitted.

cavalier1138
Posts: 4481
Joined: Fri Mar 25, 2016 8:01 pm

Re: UCLA Admissions Data

Postby cavalier1138 » Mon Oct 30, 2017 6:03 pm

The issue is that MyLSN can actually have lower accuracy as you increase the ranges, due to the issues with self-reporting. It's great for giving a semi-accurate picture of a very narrow window of applicants. Also, the data you have is from 2013. UCLA's median dropped a point and their 25th percentile dropped 2 points from 2012-13. So it's probably not the best year to look at for admissions data.

I get that there's this strong desire to show everyone that the "system" isn't right and these schools must be averaging your scores or calling references or performing rituals of darkest magick to delve your very soul, but your numbers are pretty much all that matter. Stop worrying about the stuff on the fringes that won't actually affect your app.

User avatar
MT Cicero
Posts: 645
Joined: Sun Oct 14, 2012 7:40 pm

Re: UCLA Admissions Data

Postby MT Cicero » Mon Oct 30, 2017 6:05 pm

https://www.lsac.org/officialguide/2016/lsac_4837.asp

Here's the one for the 2013-14 cycle, starting in Fall 2014 (c/o 2017). Just thought it'd be funny to see if I was the lowest GPA accepted that cycle. Looks like it was either me or the other person in my (low) GPA band.

rictheruler
Posts: 27
Joined: Sun Nov 22, 2015 5:33 am

Re: UCLA Admissions Data

Postby rictheruler » Mon Oct 30, 2017 6:12 pm

cavalier1138 wrote:I get that there's this strong desire to show everyone that the "system" isn't right and these schools must be averaging your scores or calling references or performing rituals of darkest magick to delve your very soul, but your numbers are pretty much all that matter. Stop worrying about the stuff on the fringes that won't actually affect your app.


That's not my desire, and I actually hope you're right. I'm more worried about my softs being very subpar and that everyone assuring me I'll do well is taking too simplistic an approach to admissions. The numbers game is really the only reason I have a good shot at a top school.

MT Cicero wrote:https://www.lsac.org/officialguide/2016/lsac_4837.asp

Here's the one for the 2013-14 cycle, starting in Fall 2014 (c/o 2017). Just thought it'd be funny to see if I was the lowest GPA accepted that cycle. Looks like it was either me or the other person in my (low) GPA band.


Interesting, at first glance it looks like these numbers actually align better with what you would expect than the link I posted. Would be interested to see more recent data.

cavalier1138
Posts: 4481
Joined: Fri Mar 25, 2016 8:01 pm

Re: UCLA Admissions Data

Postby cavalier1138 » Mon Oct 30, 2017 6:14 pm

rictheruler wrote:
cavalier1138 wrote:I get that there's this strong desire to show everyone that the "system" isn't right and these schools must be averaging your scores or calling references or performing rituals of darkest magick to delve your very soul, but your numbers are pretty much all that matter. Stop worrying about the stuff on the fringes that won't actually affect your app.


That's not my desire, and I actually hope you're right. I'm more worried about my softs being very subpar and that everyone assuring me I'll do well is taking too simplistic an approach to admissions. The numbers game is really the only reason I have a good shot at a top school.


Ok, so say you're right. What then?

You can't make your softs sparkle. You can't make adcomms care about an addendum. So stop focusing on this stuff and just get your apps in.

User avatar
MT Cicero
Posts: 645
Joined: Sun Oct 14, 2012 7:40 pm

Re: UCLA Admissions Data

Postby MT Cicero » Mon Oct 30, 2017 6:16 pm

rictheruler wrote:Interesting, at first glance it looks like these numbers actually align better with what you would expect than the link I posted. Would be interested to see more recent data.


Just change the date in the web address bar. Goes all the way to "2018," which is for Fall 2016 matriculants (c/o 2019).

rictheruler
Posts: 27
Joined: Sun Nov 22, 2015 5:33 am

Re: UCLA Admissions Data

Postby rictheruler » Mon Oct 30, 2017 7:17 pm

MT Cicero wrote:
rictheruler wrote:Interesting, at first glance it looks like these numbers actually align better with what you would expect than the link I posted. Would be interested to see more recent data.


Just change the date in the web address bar. Goes all the way to "2018," which is for Fall 2016 matriculants (c/o 2019).


Genius.




Return to “Law School Admissions Forum”

Who is online

Users browsing this forum: cannonballer, cavalier1138, Dcc617, hdunlop, pleaseberkeley and 9 guests