Possible Statistcal Analysis of 2012-2013 Cycle...Thoughts? Forum

[Yardbird]

So looking at a chart posted at the LSATBlog (Link), I can't but help think that this year's cycle might be more similar to the 2005-2006 application cycle (or even the 2001-2002 cycle) than last year's cycle due to the significant drop in test takers. In fact, there were 8,000 less test takers in 2011-2012 than in 2005-2006. There is likely to be a small drop in testers for 2012-2013 cycle as well if the trend continues from 2010. What does this all mean?

If we conservatively predict that we lose about 15k testers this year (114,925 total, an 11.5% decrease from 2011-2012), we get the following number of testers scoring:
Above a 170 (97.5%): ~2875
Above a 174 (99.3%): ~800
Above a 178 (99.9%): ~115

Compare this to 2009-2010, the all time high for LSAT test takers (171,514):
Above a 170 (97.5%): ~4290
Above a 174 (99.3%): ~1200
Above a 178 (99.9%): ~170

2010-2011, the next high for LSAT test takers (155,050):
Above a 170 (97.5%): ~3875
Above a 174 (99.3%): ~1085
Above a 178 (99.9%): ~155

and 2011-2012, last year's numbers (129,925 testers):
Above a 170 (97.5%): ~3250
Above a 174 (99.3%): ~910
Above a 178 (99.9%): ~130

Looking at the class sizes and LSAT medians for the current Top 10 for 2012, we see the following:

(#)S: Class Size | LSAT Median
(1) Y: 214 | 173
(2) H: 559 | 173
(3) S: 170 | 172
(4) Co: 397 | 171
(5) Ch: 191 | 171
(6) N: 450 | 171
(7) P: 255 | 170
(7) V: 368 | 170
(7) B: 292 | 168
(10)M: 371 | 169
(Total Class Size for T10 is 3267)

Compare this to numbers of the Fall 2004 entering class (I couldn't find 2005 numbers, but they aren't likely to have changed much in 1 year seeing as they havent changed much in 8 years):

(#)S: Class Size | LSAT Median
(1) Y: 189 | 171
(2) H: 554 | 170
(3) S: 166 | 168
(4) Co: 374 | 171
(5) N: 440 | 169
(6) Ch: 192 | 170
(7) P: 257 | 170
(7) M: 381 | 167
(9) V: 360 | 169
(13)B: 270 | 165
(Total class size for T10 is 3183)

From these we can calculate the % change in Class Size and LSAT Medians from 2004-2012 (all %ages reported as increase/decrease from 2004):

(#)S: Class Size | LSAT Median
(1) Y: +13.23% | +1.170%
(2) H: +0.903% | +1.765%
(3) S: +2.410% | +2.381%
(4) Co: +6.150% | +0.000%
(5) Ch: -0.521% | +0.588%
(6) N: +2.273% | +1.183%
(7) P: -0.778% | +0.000%
(7) V: +2.222% | +0.592%
(7) B: +8.148% | +1.818%
(10)M: -2.625% | +1.198%

Class sizes for the most part have increased while none of the LSAT Medians decreased from their 2004 values (most increasing with Columbia and Penn remaining the same). With only ~2875 test takers scoring above a 170 and ~3250-3300 seats available in the T10, there aren't enough top scorers to go around. This will likely impact the law schools admissions for this year (or soon in the future)

If we assume that schools want to keep current class sizes there are a few possibilities I see:
-------(1) Schools will take a hit on their LSAT Median in order to maintain their GPA Median
-------(2) Schools will take a hit on their GPA Median if they take more High LSAT Low GPA Splitters to maintain their LSAT Medians, possibly reaching past previous "GPA floors"

If we assume that schools will alter class sizes with the decrease in applicants, I again see a few possibilities:
-------(1) Schools will accept fewer lower LSAT score applicants, possibly admitting High LSAT splitters taking a hit on their GPA Median
-------(2) Schools will accept fewer higher LSAT score applicants, possibly admitting High GPA Low LSAT splitters taking a hit on their LSAT Median
-------(3) Schools will accept fewer applicants overall

I think the most likely scenario for keeping current class sizes will be #2 and the most likely for changing class sizes will be either #1 or #3. I don't think schools will want to take a hit on their LSAT Medians over their GPA Medians.

So what does this all mean? Potentially a very good cycle for High LSAT Low GPA splitters scoring above 170 or a bad cycle for everyone if the schools accept less applicants overall in order to cut down class sizes while maintaining numbers.

Thoughts?

[Edit] Fixed some values (changed percentiles for above 170 to 97.5 from 98 and adjusted numbers accordingly)

[nyremy]

Good work. Just a few comments:

1. I doubt schools will decrease class sizes. This would lower their revenue and would thus likely negatively affect their profits. I see no reason why a school would choose to do that just because there are a lack of "qualified" (numerically) applicants, especially when it is a problem that should affect all peer schools equally.

2. I think it is slightly inappropriate to use 2004's data as a comparison point because I believe, and I may be wrong, that there are more 170+scorers than before. I believe back then a 170 was the 98th percentile while now it is the 97th. Thus while the total number of applicants may have been comparable, the amount of top scorers would still be higher today.

3. Where did you get the median LSAT scores? I'm not doubting them, I've just been looking for that information!

[Law Sauce]

In 2004, the % points matched up with the # scores differently. In the later 1990s, i believe, 167 was the top 98th percentile. Now thats like a 171.

[Yardbird]

1. I doubt schools will decrease class sizes. This would lower their revenue and would thus likely negatively affect their profits. I see no reason why a school would choose to do that just because there are a lack of "qualified" (numerically) applicants, especially when it is a problem that should affect all peer schools equally.

My thoughts exactly.

2. I think it is slightly inappropriate to use 2004's data as a comparison point because I believe, and I may be wrong, that there are more 170+scorers than before. I believe back then a 170 was the 98th percentile while now it is the 97th. Thus while the total number of applicants may have been comparable, the amount of top scorers would still be higher today.

I believe its been 97.5 for a while now (including 2004). I rounded up to overestimate prospects for top scorers. In all likelihood the number of top scorers will be lower than in 2004 since there will be less test takers overall compared to 2004.

3. Where did you get the median LSAT scores? I'm not doubting them, I've just been looking for that information!

I found an interesting document from a firm's hiring pamphlet from 2006 that detailed all information about the 2004 class. LSAT medians, attrition rates, gpa medians, rank, etc.

[KevinP]

Don't forget that in 2004 the ABA required schools to report the average of matriculants' LSAT scores, whereas now they require schools to report the highest.

I'm anxiously awaiting the number of test takers for June in order to see if this decline continues into next cycle.

[Tiago Splitter]

Until 2006 schools were required to report the average LSAT score, not the highest. As a result retaking was far less popular. This also means that a high LSAT was worth more in those days than it is now.

[Yardbird]

Don't forget that in 2004 the ABA required schools to report the average of matriculants' LSAT scores, whereas now they require schools to report the highest.

I'm anxiously awaiting the number of test takers for June in order to see if this decline continues into next cycle.

I'm interested in seeing if the trend continues as well and by how much. If its truly significant than it may bode well for everyone this cycle.

Until 2006 schools were required to report the average LSAT score, not the highest. As a result retaking was far less popular. This also means that a high LSAT was worth more in those days than it is now.

True, but reporting isnt the same as evaluating and a good portion of schools still evaluate based on averages/holistically and not based on the highest LSAT. All the more reason now too since there are less high scorers overall so the high LSAT is worth a lot now than it was the past few years.

[Tiago Splitter]

Until 2006 schools were required to report the average LSAT score, not the highest. As a result retaking was far less popular. This also means that a high LSAT was worth more in those days than it is now.

True, but reporting isnt the same as evaluating and a good portion of schools still evaluate based on averages/holistically and not based on the highest LSAT. All the more reason now too since there are less high scorers overall so the high LSAT is worth a lot now as well as in the past.

I'm going to completely disagree with this. There is almost no evidence that schools care about much beyond the highest LSAT and GPA. I do agree that there are fewer high scores compared to, say, 2009-2010.

Also, look at the cycles people were having in 2004-2006 compared to this year. Even though things got easier this year, it was still tougher than before the ABA changed its policy. Columbia was accepting people with sub 3.3 GPAs. Harvard was accepting people with sub-3.6s. Things are getting easier, but they haven't gotten back to those earlier levels yet, even with the same number of applicants.

[Yardbird]

Until 2006 schools were required to report the average LSAT score, not the highest. As a result retaking was far less popular. This also means that a high LSAT was worth more in those days than it is now.

True, but reporting isnt the same as evaluating and a good portion of schools still evaluate based on averages/holistically and not based on the highest LSAT. All the more reason now too since there are less high scorers overall so the high LSAT is worth a lot now as well as in the past.

I'm going to completely disagree with this. There is almost no evidence that schools care about much beyond the highest LSAT and GPA. I do agree that there are fewer high scores compared to, say, 2009-2010.

Also, look at the cycles people were having in 2004-2006 compared to this year. Even though things got easier this year, it was still tougher than before the ABA changed its policy. Columbia was accepting people with sub 3.3 GPAs. Harvard was accepting people with sub-3.6s. Things are getting easier, but they haven't gotten back to those earlier levels yet, even with the same number of applicants.

I have talked to many admissions reps and a bunch at T10 schools have mentioned they average scores unless an applicant can point out a reason why. Example, Berkeley will average both unless they differ by more than 8 points, GULC will average unless they differ by more than 3 and there's an addendum. For the most part, they will look at the higher one with an addendum but most will average unless you say otherwise. I do get what you're saying though.

Regarding the things getting easier, no one thought the numbers would drop as much as they did last year. Last year's cycle would not have felt the effects of last year's drop. This year will. The fact that the numbers dropped to an all-time low for the past decade will be an important consideration for schools this year. They couldn't have foreseen the numbers dropping when the cycle started in September with 3 tests still unadministered (Oct, Dec, Feb). They'll keep a close eye on the June numbers this year and keep last year's numbers in mind as well. That is just my opinion though.

[Edit] Harvard did wait-list a bunch of sub 3.6's and admitted a non-URM 3.4 last year. Columbia waitlisted 3 sub 3.3 non-URMs compared to one in 2010. I would say this year they will probably wait list a bunch more. Also keep in mind that LSN and TLS are very intertwined in that top performers use both sites. There are splitters out there with a high lsat and low gpa who may have gotten in whose numbers are not on LSN. In fact, there's a good likelihood they don't know about it or this site.

[Samara]

I have talked to many admissions reps and a bunch at T10 schools have mentioned they average scores unless an applicant can point out a reason why. Example, Berkeley will average both unless they differ by more than 8 points, GULC will average unless they differ by more than 3 and there's an addendum. For the most part, they will look at the higher one with an addendum but most will average unless you say otherwise. I do get what you're saying though.

OP is tl;dr, but there is absolutely no reason to believe that schools average scores and plenty of reasons to believe that they don't. And I'm sorry, but a statement from an admissions representative should not at all be construed as representative of the overarching policy. They say a lot of things to pretend like it's not all about the numbers, but for 95% of applicants, it is.

[Tiago Splitter]

I have talked to many admissions reps and a bunch at T10 schools have mentioned they average scores unless an applicant can point out a reason why. Example, Berkeley will average both unless they differ by more than 8 points, GULC will average unless they differ by more than 3 and there's an addendum. For the most part, they will look at the higher one with an addendum but most will average unless you say otherwise. I do get what you're saying though.

No addendum for me, 12 point increase and it didn't seem to matter for my cycle. Countless others on TLS can share a similar story.

Regarding the things getting easier, no one thought the numbers would drop as much as they did last year. Last year's cycle would not have felt the effects of last year's drop. This year will. The fact that the numbers dropped to an all-time low for the past decade will be an important consideration for schools this year. They couldn't have foreseen the numbers dropping when the cycle started in September with 3 tests still unadministered (Oct, Dec, Feb). They'll keep a close eye on the June numbers this year and keep last year's numbers in mind as well. That is just my opinion though.

I do think schools were caught a bit off guard, but they aren't completely stupid. There has been quite a bit of wait list activity and now that we're in July you can see how low the schools had to go to fill their classes. It was certainly lower than the last couple of years, but we haven't dropped to the 2004 levels yet. To be honest, a big reason why I paid the money to apply to Harvard was because of the apparently mistaken belief that things would return to the levels seen in 2003-2004 when people with my numbers had no problem getting accepted there.

But things are getting easier. Hopefully there will be another 20% drop in apps and you guys can enjoy great cycles.

[ahnhub]

LSAC discloses how many applicants have a certain score (highest):

People thought this would be a really good cycle for splitters as well. The cycle's definitely been easier, but I don't know if it's been any better for splitters than anybody else. I think the effect has been most dramatic in the T25-ish range; those schools everyone thought were a secure bet before detailed employment info came out. T14 definitely relaxed their standards, but not quite to the level of schools like GW.

There has actually been a collapse of applicants the past two years--like a 25% fall in people applying. But the proportion of high scores has been increasing relatively steadily along with more retakers, so the corresponding fall in high scores is less dramatic. I marvel at the world law school applicants pre-2005 lived in. Someone with a 3.9 GPA who was nailing every PT at 170+ might have a bad day and end up with a 163. They'd end up paying sticker at BU instead of breezing into Harvard, like they could probably do today.

[Yardbird]

Regarding the things getting easier, no one thought the numbers would drop as much as they did last year. Last year's cycle would not have felt the effects of last year's drop. This year will. The fact that the numbers dropped to an all-time low for the past decade will be an important consideration for schools this year. They couldn't have foreseen the numbers dropping when the cycle started in September with 3 tests still unadministered (Oct, Dec, Feb). They'll keep a close eye on the June numbers this year and keep last year's numbers in mind as well. That is just my opinion though.

I do think schools were caught a bit off guard, but they aren't completely stupid. There has been quite a bit of wait list activity and now that we're in July you can see how low the schools had to go to fill their classes. It was certainly lower than the last couple of years, but we haven't dropped to the 2004 levels yet. To be honest, a big reason why I paid the money to apply to Harvard was because of the apparently mistaken belief that things would return to the levels seen in 2003-2004 when people with my numbers had no problem getting accepted there.

But things are getting easier. Hopefully there will be another 20% drop in apps and you guys can enjoy great cycles.

As a sub 3.3 high LSAT splitter aiming for Columbia I am hoping for this

[KevinP]

1. I doubt schools will decrease class sizes. This would lower their revenue and would thus likely negatively affect their profits. I see no reason why a school would choose to do that just because there are a lack of "qualified" (numerically) applicants, especially when it is a problem that should affect all peer schools equally.

I think this depends on the school. Although the decline will affect most peer schools uniformly, some schools will use this as an opportunity to maintain or increase its medians by decreasing class sizes and hence increase its US News ranking. Last cycle also saw a noticeable decrease in applicants, and class sizes dropped significantly at a number of schools (from a historical high).

[Yardbird]

1. I doubt schools will decrease class sizes. This would lower their revenue and would thus likely negatively affect their profits. I see no reason why a school would choose to do that just because there are a lack of "qualified" (numerically) applicants, especially when it is a problem that should affect all peer schools equally.

I think this depends on the school. Although the decline will affect most peer schools uniformly, some schools will use this as an opportunity to maintain or increase its medians by decreasing class sizes and hence increase its US News ranking. Last cycle also saw a noticeable decrease in applicants, and class sizes dropped significantly at a number of schools (from a historical high).

Which schools did the ckass size drop at?

[ahnhub]

Also, an addendum to my earlier post: I just remembered there was something very strange about the proportion of high-to-low LSAT scores. The number of high-scoring applicants was actually falling FASTER than the number of low-scoring ones: i.e., the exact opposite thing you would expect to happen when people are finding out low-ranked schools have horrible job prospects.

I can't really think of any great explanations for this. Maybe 1) people with good academic credentials are just not applying because they have better alternatives, 2)there was a "drag effect" on the previous cycle, which still had a glut of high scorers from before the collapse in applicants which has mostly cleared out now.

[Kring345]

I have talked to many admissions reps and a bunch at T10 schools have mentioned they average scores unless an applicant can point out a reason why.

I had a 162 and 168. I wouldnt have gotten into a single t14 if they averaged, but I got into CLS, Berk, Mich, George, Cornell, and WL everywhere else. They dont average. At most, they 'keep in mind' initial scores.

[KevinP]

Which schools did the ckass size drop at?

You can browse this thread to for a more comprehensive list: http://www.top-law-schools.com/forums/v ... 2&t=163250

From the top of my head Duke's entering class size dropped from 238 to 211. NYU dropped from 476 to 450. GW dropped from 523 to 490.

[Yardbird]

I have talked to many admissions reps and a bunch at T10 schools have mentioned they average scores unless an applicant can point out a reason why.

I had a 162 and 168. I wouldnt have gotten into a single t14 if they averaged, but I got into CLS, Berk, Mich, George, Cornell, and WL everywhere else. They dont average. At most, they 'keep in mind' initial scores.

URM? GPA? The stats here are mostly affecting nonURM splitters.

[KevinP]

.

[top30man]

I have talked to many admissions reps and a bunch at T10 schools have mentioned they average scores unless an applicant can point out a reason why.

I had a 162 and 168. I wouldnt have gotten into a single t14 if they averaged, but I got into CLS, Berk, Mich, George, Cornell, and WL everywhere else. They dont average. At most, they 'keep in mind' initial scores.

URM? GPA? The stats here are mostly affecting nonURM splitters.

The schools that told you this likely weren't telling the truth. Or a version of the truth they wish you to believe. Schools may look at it but they dot average it. I literally wouldn't have broken then t20, let alone the t14 if they averaged. I performed this cycle as if my only lsat was my third.

[Kring345]

I have talked to many admissions reps and a bunch at T10 schools have mentioned they average scores unless an applicant can point out a reason why.

I had a 162 and 168. I wouldnt have gotten into a single t14 if they averaged, but I got into CLS, Berk, Mich, George, Cornell, and WL everywhere else. They dont average. At most, they 'keep in mind' initial scores.

URM? GPA? The stats here are mostly affecting nonURM splitters.

Not a URM but veteran. We don't receive that much of a bump though so it doesn't explain much probably.

[sunynp]

1. I doubt schools will decrease class sizes. This would lower their revenue and would thus likely negatively affect their profits. I see no reason why a school would choose to do that just because there are a lack of "qualified" (numerically) applicants, especially when it is a problem that should affect all peer schools equally.

I think this depends on the school. Although the decline will affect most peer schools uniformly, some schools will use this as an opportunity to maintain or increase its medians by decreasing class sizes and hence increase its US News ranking. Last cycle also saw a noticeable decrease in applicants, and class sizes dropped significantly at a number of schools (from a historical high).

Haven't some schools already announced they are reducing class size? They are more concerned about keeping the numbers for the rankings then losing a few seats. If they lose ranking, they will lose a lot more status.

[2014]

7 point lsat jump with no reason other than "I tried harder" so no addendum. Got in where an average would have precluded me.

This data makes me wish I deferred a year even though I'm thrilled about where I'm going. Didn't realize I could be such a commodity.

[PaulKriske]

This data makes me wish I deferred a year even though I'm thrilled about where I'm going.