Chances at Cornell? Forum

[hos9903]

.

[Mojosodope]

Cornell is obviously a reach for me, 3.37/167. My question is whether there's enough of a shot at riding the WL to acceptance for me to justify paying the application fee. On LSN there were only a few applicants in my range, most of them got WL but that's the extent of the data (I assume a lot of them just didn't follow up with the site).


Thanks for any input

My advice would be to retake.

[hephaestus]

Cornell is obviously a reach for me, 3.37/167. My question is whether there's enough of a shot at riding the WL to acceptance for me to justify paying the application fee. On LSN there were only a few applicants in my range, most of them got WL but that's the extent of the data (I assume a lot of them just didn't follow up with the site).


Thanks for any input

My advice would be to retake.

I agree. I think you should retake, because Cornell is very expensive at sticker.

[Arcticlynx]

Deleted, because law school admission probably do not fit a normal distribution. See http://admissionsbythenumbers.blogspot.com/.

[hos9903]

Thanks for the responses guys, I appreciate it. Retake is definitely still on the table...I may take a year off depending on how this cycle works out and retake for next. There are still some schools on my list I'd definitely attend if the $$ is right.


I guess that leads me to a bigger question. I'm not really sure how WLs work with Cornell. Is movement/acceptance atypical? Or if you ride it out and write some LOCI is there a good shot you may end up accepted, but stuck with sticker? (sorry if this now makes my thread off-topic in the chances forum)

Thanks again

[hephaestus]

Thanks for the responses guys, I appreciate it. Retake is definitely still on the table...I may take a year off depending on how this cycle works out and retake for next. There are still some schools on my list I'd definitely attend if the $$ is right.


I guess that leads me to a bigger question. I'm not really sure how WLs work with Cornell. Is movement/acceptance atypical? Or if you ride it out and write some LOCI is there a good shot you may end up accepted, but stuck with sticker? (sorry if this now makes my thread off-topic in the chances forum)

Thanks again

Cornell puts a significant group of students on "Reserve," which seems like the equivalent of a priority waitlist at some other schools. A normal wait list is basically a rejection.

[shifty_eyed]

Have you graduated yet? If not, waiting a year and loading up on easy classes to boost your GPA seems like a good idea.

[hos9903]

Have you graduated yet? If not, waiting a year and loading up on easy classes to boost your GPA seems like a good idea.

I haven't graduated yet, which is why I am open to waiting a cycle. But I also have two majors, neither of which leave my schedule open to loading up on easy classes. Realistically I could boost my gpa to somewhere in the 3.43-3.48 range with the time I have left, which isn't much better than a 3.37. Both are ~25% or lower at any schools I'm looking at (T30, T20, maybe GULC and Cornell).

[Arcticlynx]

Deleted, because law school admission probably do not fit a normal distribution. See http://admissionsbythenumbers.blogspot.com/.

[hos9903]

+1 point on the LSAT is worth about 0.022 to 0.033 points of GPA if your use linear regression to fit GPA/LSAT data for top-50 schools depending on where you are along the GPA and LSAT spectrum. A GPA boost from 3.6 to 3.7 is worth a whole lot less for someone with a 180 than someone with a 160. But as a general guide I would figure that raising you GPA from 3.37 to 3.43 is the same as raising your LSAT from 167 to 169, and 3.37 to 3.47 would be equivalent to going from a 167 to 170. So if you can do it, then getting your GPA up would be worth it for sure.

I would actually advocate saving the time you would otherwise invest in law school applications and studying for the LSAT to focus on your classes as much as possible. I mean, go ahead and run the numbers for the 170/3.37 and a 167/3.47 but I would think the latter would make for a better application. Plus if you decide you still want to retake next October/December that's always and option.

SUMMARY OUTPUT

Regression Statistics
Multiple R 0.857759689
R Square 0.735751685
Adjusted R Square 0.728609839
Standard Error 0.073910621
Observations 39

ANOVA
df SS MS F Significance F
Regression 1 0.562774577 0.562774577 103.0198142 3.05386E-12
Residual 37 0.202122859 0.00546278
Total 38 0.764897436

Coefficients Standard Error t Stat P-value Lower 95% Upper 95% Lower 95.0% Upper 95.0%
Intercept -0.830008823 0.443539594 -1.871329716 0.069222631 -1.728705406 0.06868776 -1.728705406 0.06868776
X Variable 1 0.027291299 0.002688833 10.1498677 3.05386E-12 0.021843206 0.032739392 0.021843206 0.032739392

For the 39 schools that I have analyzed (ie 39 of the t-50), +0.027 (± 0.002) GPA is worth +1 LSAT.

This is incredible. Did you just use STATA to answer my chances thread? Awesome. Thank you, really really appreciate it. I have actually already applied to 11 schools (and really made this whole thread to see if Cornell was worth it since I'm in the 'what the hell?' phase of school selection), so if it is starting to look like my GPA will get that boost as the year goes on, I may defer and see where I end up next cycle after a June '14 retake.

[Arcticlynx]

Deleted, because law school admission probably do not fit a normal distribution. See http://admissionsbythenumbers.blogspot.com/.

[jn7]

+1 point on the LSAT is worth about 0.022 to 0.033 points of GPA if your use linear regression to fit GPA/LSAT data for top-50 schools depending on where you are along the GPA and LSAT spectrum. A GPA boost from 3.6 to 3.7 is worth a whole lot less for someone with a 180 than someone with a 160. But as a general guide I would figure that raising you GPA from 3.37 to 3.43 is the same as raising your LSAT from 167 to 169, and 3.37 to 3.47 would be equivalent to going from a 167 to 170. So if you can do it, then getting your GPA up would be worth it for sure.

I would actually advocate saving the time you would otherwise invest in law school applications and studying for the LSAT to focus on your classes as much as possible. I mean, go ahead and run the numbers for the 170/3.37 and a 167/3.47 but I would think the latter would make for a better application. Plus if you decide you still want to retake next October/December that's always and option.

SUMMARY OUTPUT

Regression Statistics
Multiple R 0.857759689
R Square 0.735751685
Adjusted R Square 0.728609839
Standard Error 0.073910621
Observations 39

ANOVA
df SS MS F Significance F
Regression 1 0.562774577 0.562774577 103.0198142 3.05386E-12
Residual 37 0.202122859 0.00546278
Total 38 0.764897436

Coefficients Standard Error t Stat P-value Lower 95% Upper 95% Lower 95.0% Upper 95.0%
Intercept -0.830008823 0.443539594 -1.871329716 0.069222631 -1.728705406 0.06868776 -1.728705406 0.06868776
X Variable 1 0.027291299 0.002688833 10.1498677 3.05386E-12 0.021843206 0.032739392 0.021843206 0.032739392

For the 39 schools that I have analyzed (ie 39 of the t-50), +0.027 (± 0.002) GPA is worth +1 LSAT.

What happens statistically if he applies and sends a few emails indicating his specific interest in the school etc.? This is bananas.

[Lavitz]

Retake / try to raise your gpa.

But if you already applied to 11 schools that are presumably lower ranked than Cornell, I don't see why you wouldn't want to apply and see what happens. You'll likely be put on Reserve. There tends to be a significant amount of movement over the summer. People get off WLs elsewhere and they try to find people on the reserve list who help at least one median and are still willing to pay sticker to attend.

If you've retaken by then, haven't improved, get off reserve, and decide you're willing to pay sticker to attend this year, then it will have been worth it to apply to Cornell. If you improved on a retake by then, you can just sit out and reapply next cycle.

[cotiger]

This is incredible. Did you just use STATA to answer my chances thread? Awesome. Thank you, really really appreciate it. I have actually already applied to 11 schools (and really made this whole thread to see if Cornell was worth it since I'm in the 'what the hell?' phase of school selection), so if it is starting to look like my GPA will get that boost as the year goes on, I may defer and see where I end up next cycle after a June '14 retake.

I'm not sure exactly what Arcticlynx is doing (regressing GPA on LSAT?), but whatever it is, it isn't even close to appropriate. If I'm reading his results correctly, all he's doing is saying that each additional LSAT point predicts an increase in .03 uGPA. Nothing to do with how schools view the two metrics.

If you want a statistical treatment of the relative importance of uGPA and LSAT, go here: http://admissionsbythenumbers.blogspot.com/

If you want to know your chances, go to mylsn.

[Arcticlynx]

This is incredible. Did you just use STATA to answer my chances thread? Awesome. Thank you, really really appreciate it. I have actually already applied to 11 schools (and really made this whole thread to see if Cornell was worth it since I'm in the 'what the hell?' phase of school selection), so if it is starting to look like my GPA will get that boost as the year goes on, I may defer and see where I end up next cycle after a June '14 retake.

I'm not sure exactly what Arcticlynx is doing (regressing GPA on LSAT?), but whatever it is, it isn't even close to appropriate. If I'm reading his results correctly, all he's doing is saying that each additional LSAT point predicts an increase in .03 uGPA. Nothing to do with how schools view the two metrics.

If you want a statistical treatment of the relative importance of uGPA and LSAT, go here: http://admissionsbythenumbers.blogspot.com/

If you want to know your chances, go to mylsn.

Yes it is a correlation index, based on a matrix of two covariant factors LSAT and GPA relative to a nonlinear model (see multifactor nonlinear regression). I guess you could call it a sliding scale, but in 3 dimensions.

However, as http://admissionsbythenumbers.blogspot.com/ points out, law schools don’t act like most physical variables (it’s not like predicting wind-speed or molecular activity) and probably doesn’t fit a normal distribution. So I am going to delete these posts, because without an understanding of the model and it’s limitations they are very likely to be misinterpreted and become counterproductive.

However, I would also caution against myLSN (which has huge flaws) relies on a small sample size that is very likely to be biased and not representative of population. The law school medians are probably a better predictor than myLSN.

[elterrible78]

This is incredible. Did you just use STATA to answer my chances thread? Awesome. Thank you, really really appreciate it. I have actually already applied to 11 schools (and really made this whole thread to see if Cornell was worth it since I'm in the 'what the hell?' phase of school selection), so if it is starting to look like my GPA will get that boost as the year goes on, I may defer and see where I end up next cycle after a June '14 retake.

I'm not sure exactly what Arcticlynx is doing (regressing GPA on LSAT?), but whatever it is, it isn't even close to appropriate. If I'm reading his results correctly, all he's doing is saying that each additional LSAT point predicts an increase in .03 uGPA. Nothing to do with how schools view the two metrics.

If you want a statistical treatment of the relative importance of uGPA and LSAT, go here: http://admissionsbythenumbers.blogspot.com/

If you want to know your chances, go to mylsn.

Yes it is a correlation index, based on a matrix of two covariant factors LSAT and GPA relative to a nonlinear model (see multifactor nonlinear regression). I guess you could call it a sliding scale, but in 3 dimensions.

However, as http://admissionsbythenumbers.blogspot.com/ points out, law schools don’t act like most physical variables (it’s not like predicting wind-speed or molecular activity) and probably doesn’t fit a normal distribution. So I am going to delete these posts, because without an understanding of the model and it’s limitations they are very likely to be misinterpreted and become counterproductive.

However, I would also caution against myLSN (which has huge flaws) relies on a small sample size that is very likely to be biased and not representative of population. The law school medians are probably a better predictor than myLSN.

I'm the Admissions By the Numbers guy, and I'd like to point out that I use the exact same data that myLSN does (actually, I get the data FROM the guy who does myLSN), so the same sample size problems apply to my own. I don't generally think that the sample size is small (actually, I'll tell you right now it's not), and although I was at some point skeptical that the sample skewed towards "higher end" applicants, I'm not really completely convinced of that, either (the medians for accepted applicants on LSN are slightly higher than schools' medians, but that's probably because they are for ACCEPTED applicants, and not applicants who necessarily matriculated).

myLSN is probably the most solid "What are my Chances?" tool out there for people. The real place that it runs into trouble is not the site itself (which acknowledges its limitations), but the inability of users to think critically about what they are looking at.

A great example of this is trying to use the site to gauge whether or not to apply ED. A lot of people will select a GPA range and an LSAT range, and then compare the results of ED and RD applicants. This is very suspect for a couple reasons. First, if the accepted ED applicants all had numbers at the higher ends of those ranges, then a higher acceptance rate for them would not be all that surprising, and would definitely not indicate that EDing conferred a benefit. Additionally, there are all kinds of things that matter (despite the IT'S ALL GPA AND LSAT, BRAH! echo-chamber around here) that myLSN does its best to account for, but given the limitations of its design, can only go so far.

In short, myLSN is great for what it is: giving you a quick snapshot of the LSN outcomes for the parameters you set. For my money, it's a whole lot more useful than LSP or any of the other "predictors" out there. myLSN doesn't claim to predict anything (neither do I, by the way); it simply shows you what has happened.

In a perfect world, we'd all have access to all the actual applicant data from the schools, as well as a meaningful way to measure resume strength, strength of LORs, strength of PS, etc. etc. etc. Unfortunately we don't (and probably never will), so we make due with what we have. It's when users of these sites have unrealistic perceptions of what they are looking at that problems occur.

Just my two cents.

[LSATSCORES2012]

This is incredible. Did you just use STATA to answer my chances thread? Awesome. Thank you, really really appreciate it. I have actually already applied to 11 schools (and really made this whole thread to see if Cornell was worth it since I'm in the 'what the hell?' phase of school selection), so if it is starting to look like my GPA will get that boost as the year goes on, I may defer and see where I end up next cycle after a June '14 retake.

I'm not sure exactly what Arcticlynx is doing (regressing GPA on LSAT?), but whatever it is, it isn't even close to appropriate. If I'm reading his results correctly, all he's doing is saying that each additional LSAT point predicts an increase in .03 uGPA. Nothing to do with how schools view the two metrics.

If you want a statistical treatment of the relative importance of uGPA and LSAT, go here: http://admissionsbythenumbers.blogspot.com/

If you want to know your chances, go to mylsn.

Yes it is a correlation index, based on a matrix of two covariant factors LSAT and GPA relative to a nonlinear model (see multifactor nonlinear regression). I guess you could call it a sliding scale, but in 3 dimensions.

However, as http://admissionsbythenumbers.blogspot.com/ points out, law schools don’t act like most physical variables (it’s not like predicting wind-speed or molecular activity) and probably doesn’t fit a normal distribution. So I am going to delete these posts, because without an understanding of the model and it’s limitations they are very likely to be misinterpreted and become counterproductive.

However, I would also caution against myLSN (which has huge flaws) relies on a small sample size that is very likely to be biased and not representative of population. The law school medians are probably a better predictor than myLSN.

I'm the Admissions By the Numbers guy, and I'd like to point out that I use the exact same data that myLSN does (actually, I get the data FROM the guy who does myLSN), so the same sample size problems apply to my own. I don't generally think that the sample size is small (actually, I'll tell you right now it's not), and although I was at some point skeptical that the sample skewed towards "higher end" applicants, I'm not really completely convinced of that, either (the medians for accepted applicants on LSN are slightly higher than schools' medians, but that's probably because they are for ACCEPTED applicants, and not applicants who necessarily matriculated).

myLSN is probably the most solid "What are my Chances?" tool out there for people. The real place that it runs into trouble is not the site itself (which acknowledges its limitations), but the inability of users to think critically about what they are looking at.

A great example of this is trying to use the site to gauge whether or not to apply ED. A lot of people will select a GPA range and an LSAT range, and then compare the results of ED and RD applicants. This is very suspect for a couple reasons. First, if the accepted ED applicants all had numbers at the higher ends of those ranges, then a higher acceptance rate for them would not be all that surprising, and would definitely not indicate that EDing conferred a benefit. Additionally, there are all kinds of things that matter (despite the IT'S ALL GPA AND LSAT, BRAH! echo-chamber around here) that myLSN does its best to account for, but given the limitations of its design, can only go so far.

In short, myLSN is great for what it is: giving you a quick snapshot of the LSN outcomes for the parameters you set. For my money, it's a whole lot more useful than LSP or any of the other "predictors" out there. myLSN doesn't claim to predict anything (neither do I, by the way); it simply shows you what has happened.

In a perfect world, we'd all have access to all the actual applicant data from the schools, as well as a meaningful way to measure resume strength, strength of LORs, strength of PS, etc. etc. etc. Unfortunately we don't (and probably never will), so we make due with what we have. It's when users of these sites have unrealistic perceptions of what they are looking at that problems occur.

Just my two cents.

ElTerrible78, you're clearly in cahoots with the myLSN gay, otherwise you would never support such a terrible website, man. Besides, aren't you a 1L? You should be studying.

[elterrible78]

ElTerrible78, you're clearly in cahoots with the myLSN guy, otherwise you would never support such a terrible website, man. Besides, aren't you a 1L? You should be studying.[/quote]

Would not sully my reputation by associating with such a dirtbag. Especially not at pub trivia, where he gets even the biology questions wrong. Get back to reading, chump.