Harvard Law c/o 2020 Applicants (2016-2017) Forum

[R. Jeeves]

guyz as can be ascertained from the last couple posts I think there's a silent majority of those not possessing JS1s. we just don't post everyday to announce our continued status because that would be errr kinda pathetic

hang in there buds

+1

Announcing that I'm KJD above both 75ths and haven't heard anything, despite going complete in September, gets a bit redundant after a while.

i feel your pain. i have to do a similar thing at weddings

[unrelated]

guyz as can be ascertained from the last couple posts I think there's a silent majority of those not possessing JS1s. we just don't post everyday to announce our continued status because that would be errr kinda pathetic

hang in there buds

+1

Announcing that I'm KJD above both 75ths and haven't heard anything, despite going complete in September, gets a bit redundant after a while.

i feel your pain. i have to do a similar thing at weddings

Hang in there, I hear dating is mostly a black box

[potterotter]

---

[AJ1010]

Anyone with a lower LSAT score (below the 25th) receive a JS1 yet?

I have had a JS1 and i am KJD and below both 25ths.

[addie1412]

Anyone with a lower LSAT score (below the 25th) receive a JS1 yet?

I have had a JS1 and i am KJD and below both 25ths.

Are you a URM?

[AJ1010]

Anyone with a lower LSAT score (below the 25th) receive a JS1 yet?

I have had a JS1 and i am KJD and below both 25ths.

Are you a URM?

Yes. Sorry, I should have included that.

[Mackgal]

Anyone with a lower LSAT score (below the 25th) receive a JS1 yet?

I have had a JS1 and i am KJD and below both 25ths.

Are you a URM?

Yes. Sorry, I should have included that.

When did you get your JS1?

[AJ1010]

Anyone with a lower LSAT score (below the 25th) receive a JS1 yet?

I have had a JS1 and i am KJD and below both 25ths.

Are you a URM?

Yes. Sorry, I should have included that.

When did you get your JS1?

received JS1 request 12/2; interviewed last week

[galeatus]

Ok, got insanely bored at work and decided to do some (neurotic) analysis on the cycle data, based on the application profiles on LSN. The data may have some errors here and there (I'm really bad at counting), but I believe the general trend is correct, barring all the usual errors (self-reported data/unrepresentative dataset etc.)

This year's applicant pool so far (excluding URM) looks like this:
175+ (above 75%): 24 (20% of total)
173-174 (median to 75%): 20 (17% of total)
170-172 (25% to median): 45 (38% of total)
170- (below 25%): 30 (25% of total)
Total: 119

Last year's applicant pool up to 16/12/2015 was:
175+ (above 75%): 36 (30% of total)
173-174 (median to 75%): 22 (18% of total)
170-172 (25% to median): 32 (27% of total)
170- (below 25%): 30 (25% of total)
Total: 120

The data conforms to the massive drop in high LSAT scores that the LSAC data and Spivey suggested (33% drop in 75%+ scorers!).
Not gonna read into this much, as it is very much possible that many people are holding for their Dec scores and an easy Dec test may offset all this, but if the Dec test doesn't produce an abnormally high amount of applicants with top scores, it sounds like good news for everyone.

[harmcharm]

Ok, got insanely bored at work and decided to do some (neurotic) analysis on the cycle data, based on the application profiles on LSN. The data may have some errors here and there (I'm really bad at counting), but I believe the general trend is correct, barring all the usual errors (self-reported data/unrepresentative dataset etc.)

This year's applicant pool so far (excluding URM) looks like this:
175+ (above 75%): 24 (20% of total)
173-174 (median to 75%): 20 (17% of total)
170-172 (25% to median): 45 (38% of total)
170- (below 25%): 30 (25% of total)
Total: 119

Last year's applicant pool up to 16/12/2015 was:
175+ (above 75%): 36 (30% of total)
173-174 (median to 75%): 22 (18% of total)
170-172 (25% to median): 32 (27% of total)
170- (below 25%): 30 (25% of total)
Total: 120

The data conforms to the massive drop in high LSAT scores that the LSAC data and Spivey suggested (33% drop in 75%+ scorers!).
Not gonna read into this much, as it is very much possible that many people are holding for their Dec scores and an easy Dec test may offset all this, but if the Dec test doesn't produce an abnormally high amount of applicants with top scores, it sounds like good news for everyone.

Wow, you're amazing. And while I didn't take the December test, I know a lot of people who did after underperforming in September. They said December was basically terrible (especially in comparison to September, which I heard wasn't too great itself), so hopefully good news for the rest of us!

[poptart123]

Wow, you're amazing. And while I didn't take the December test, I know a lot of people who did after underperforming in September. They said December was basically terrible (especially in comparison to September, which I heard wasn't too great itself), so hopefully good news for the rest of us!

You heard December was worse than September?? I heard September was already pretty bad

[Future Ex-Engineer]

Haven't quite finished my app (will submit next week and pray that my Dec score is 170+), but I took both September and Dec.

Sept cold I wound up with a 164. I then actually studied for Dec. Think I landed between 168-172 (huge range I know).

My analysis:
Sept games slightly harder than Dec games (granted this may have to do with the fact I didn't study for Sept)
Sept RC slightly harder than Dec RC (I didn't find Eileen Grey particularly difficult, missed 2 Q's on her section, -8 overall on RC in Sept)
Sept LR 1 (or 2, doesn't matter) equivalent difficulty to Dec LR 1
Sept LR 2 far easier than Dec LR 2 (went -1 on Sept LR 2)

As such, I expect the curve to be -11 or -12. Expecting I missed somewhere between 8 and 14 problems.

[unrelated]

Ok, got insanely bored at work and decided to do some (neurotic) analysis on the cycle data, based on the application profiles on LSN. The data may have some errors here and there (I'm really bad at counting), but I believe the general trend is correct, barring all the usual errors (self-reported data/unrepresentative dataset etc.)

This year's applicant pool so far (excluding URM) looks like this:
175+ (above 75%): 24 (20% of total)
173-174 (median to 75%): 20 (17% of total)
170-172 (25% to median): 45 (38% of total)
170- (below 25%): 30 (25% of total)
Total: 119

Last year's applicant pool up to 16/12/2015 was:
175+ (above 75%): 36 (30% of total)
173-174 (median to 75%): 22 (18% of total)
170-172 (25% to median): 32 (27% of total)
170- (below 25%): 30 (25% of total)
Total: 120

The data conforms to the massive drop in high LSAT scores that the LSAC data and Spivey suggested (33% drop in 75%+ scorers!).
Not gonna read into this much, as it is very much possible that many people are holding for their Dec scores and an easy Dec test may offset all this, but if the Dec test doesn't produce an abnormally high amount of applicants with top scores, it sounds like good news for everyone.

Hey, I too am neurotic. Not sure if this has been addressed somewhere else on the forum, but using the Spivey data to recreate last year's applicant numbers YTD, I found the percent change of applicants with 170+ to be -16% this cycle YTD. Percent change of applicants with 172+ is -24% YTD.

You can see the scrawlings of my neurosis here: https://docs.google.com/spreadsheets/d/ ... edit#gid=0

Matriculation numbers are based off of available class profiles for T14. Some may be inaccurate and some conclusions not entirely fleshed out because while I am neurotic, I am still lazy. Let me know if you find any flaws in calculations.

[R. Jeeves]

Hey, I too am neurotic. Not sure if this has been addressed somewhere else on the forum, but using the Spivey data to recreate last year's applicant numbers YTD, I found the percent change of applicants with 170+ to be -16% this cycle YTD. Percent change of applicants with 172+ is -24% YTD.

You can see the scrawlings of my neurosis here: https://docs.google.com/spreadsheets/d/ ... edit#gid=0

Matriculation numbers are based off of available class profiles for T14. Some may be inaccurate and some conclusions not entirely fleshed out because while I am neurotic, I am still lazy. Let me know if you find any flaws in calculations.

Since you're playing with the data, you wanna test out my hypothesis from the Spivey thread?

http://www.top-law-schools.com/forums/v ... start=5325

[unrelated]

Hey, I too am neurotic. Not sure if this has been addressed somewhere else on the forum, but using the Spivey data to recreate last year's applicant numbers YTD, I found the percent change of applicants with 170+ to be -16% this cycle YTD. Percent change of applicants with 172+ is -24% YTD.

You can see the scrawlings of my neurosis here: https://docs.google.com/spreadsheets/d/ ... edit#gid=0

Matriculation numbers are based off of available class profiles for T14. Some may be inaccurate and some conclusions not entirely fleshed out because while I am neurotic, I am still lazy. Let me know if you find any flaws in calculations.

Since you're playing with the data, you wanna test out my hypothesis from the Spivey thread?

http://www.top-law-schools.com/forums/v ... start=5325

Hadn't seen that thread before!

Not sure how you spied that, but you sir, are definitely correct. The score bands Spivey cited were formed using formula Percentage Change = (Previous # - Current #)/(Current #) and this is incorrect. It should, as you suggest and I used in my excel, look like Percent Change = (Current - Previous)/(Previous).

I used the same formula as you to backwards calculate the Previous numbers (after algebra it looked like Percent Change = (Current/Previous)-1). To verify, I did the 165-169 score band and I got -24% for a score band %change instead of the reported -18.8%. When I changed to your hypothesis formula to backwards calculate the Previous numbers and the score band %change, I got the same as Spivey. So you were spot on!

The good news here is that this means the YTD %change and YTD raw data are not NECESSARILY (heh) flawed, but rather, Spivey (or LSAC) just fucked up his (their) algebra in trying to figure out score band changes. This also means that the score band numbers are down even more than Spivey reported!

On a tangential note, I tried going back to find last cycle's data from Spivey's blog, and did, here: http://blog.spiveyconsulting.com/every- ... ts-change/

The numbers don't match up as they should according to the formula, but of course that is because both data sets are YTD, and the dates are different (off by about 9 days). But it might give you a rough way to verify that the raw numbers and %change we were given aren't flawed.

Update: Just did the other score bands using Hypothesis formula and got same numbers as Spivey, so that's definitely what happened.

[R. Jeeves]

The good news here is that this means the YTD %change and YTD raw data are not NECESSARILY (heh) flawed, but rather, Spivey (or LSAC) just fucked up his (their) algebra in trying to figure out score band changes. This also means that the score band numbers are down even more than Spivey reported!

ok, good. so we've confirmed that the data is definitely wrong somewhere.

I believe the numbers came from LSAC - Spivey only posted them.

I think it is very unlikely that they ONLY screwed up the calculation for the bands. For that to be the case they would have had to do some very roundabout calculations given the fact that they have the raw data. I think the % YTD changes they reported for every score is also wrong.

[R. Jeeves]

Update: Just did the other score bands using Hypothesis formula and got same numbers as Spivey, so that's definitely what happened.

LSAC: We are experts in psychometrics and use extremely sophisticated methods in data analytics to ensure the consistency of our tests! Statistical analysis is totally in our wheelhouse!

*cannot calculate simple % changes without crowdsourcing it to a forum full of lib arts students*

[+] Spoiler

ok tbf it was prob just an arithmetic error, but LSAC is the worst and I like ripping on them

[sigree]

Ok - what does JS1 stand for? Also, were all of those accepted ED's? Finally, did anyone here submit in December and has had no status changes?

[Aquinas]

Ok - what does JS1 stand for? Also, were all of those accepted ED's? Finally, did anyone here submit in December and has had no status changes?

JS1 means an interview with admissions; no one has been accepted yet; HLS doesn't have an ED track; no one's status checker changes at all once their app is marked complete (at least until they receive a decision).

[xnsch]

Ok - what does JS1 stand for? Also, were all of those accepted ED's? Finally, did anyone here submit in December and has had no status changes?

Nobody has been admitted yet (or at least not anyone on this forum) and I'm pretty sure Harvard doesn't have an ED program. JS1 is the jargon for the interview invitation whereas JS2 is the acceptance call/notification

[sigree]

Thanks so much - I'm sort of doing this alone so have no idea about the process.

[unrelated]

I think it is very unlikely that they ONLY screwed up the calculation for the bands. For that to be the case they would have had to do some very roundabout calculations given the fact that they have the raw data. I think the % YTD changes they reported for every score is also wrong.

Well, if we think YTD% was calculated wrong, but we have the formula they used to find YTD% using their raw Current and Previous data, then it shouldn't matter if the YTD% they gave us is wrong, right?

We can still backwards calculate the Previous data and then use that with accurate Current data to figure out real % changes. Don't really want to calculate that though.

[R. Jeeves]

I think it is very unlikely that they ONLY screwed up the calculation for the bands. For that to be the case they would have had to do some very roundabout calculations given the fact that they have the raw data. I think the % YTD changes they reported for every score is also wrong.

Well, if we think YTD% was calculated wrong, but we have the formula they used to find YTD% using their raw Current and Previous data, then it shouldn't matter if the YTD% they gave us is wrong, right?

We can still backwards calculate the Previous data and then use that with accurate Current data to figure out real % changes. Don't really want to calculate that though.

yeah we can. that's actually what i ended up doing haha. I can post my spreadsheet with that information later if ppl want.

[unrelated]

yeah we can. that's actually what i ended up doing haha. I can post my spreadsheet with that information later if ppl want.

I just did it--wasn't as difficult as I thought it would be. I don't think they did the YTD% wrong now that I'm looking at the changes. Unless I did it wrong too, this would actually suggest that all the sign changes were wrong in what was originally reported by Spivey--meaning that for every score, applicants actually went up (except for 170 and 179). I don' think that's likely, but I also just flat out don't want it to be true lol.

[R. Jeeves]

yeah we can. that's actually what i ended up doing haha. I can post my spreadsheet with that information later if ppl want.

I just did it--wasn't as difficult as I thought it would be. I don't think they did the YTD% wrong now that I'm looking at the changes. Unless I did it wrong too, this would actually suggest that all the sign changes were wrong in what was originally reported by Spivey--meaning that for every score, applicants actually went up (except for 170 and 179). I don' think that's likely, but I also just flat out don't want it to be true lol.

i dont think they did

Percentage Change = (Previous # - Current #)/(Current #)

I think they did

Percentage Change = (Current # - Previous #)/(Current #)

eta: and realized i mistyped that in the spivey thread - just edited it.

so they only switched Current # and Previous # in the denominator, their numerator is fine, so the signs should come out right i think. but ill check this again a bit later, since I cant access your google doc while Im at work.