What is considered early? Forum

[leroyjenkins6969]

Hey all,

I know that, in general, it is best to submit your apps as early as possible.

What is considered "early" (in terms of significantly boosting your chances of admission)? Before November 15? Before the new year?

What do you all think?

[ZVBXRPL]

Lord have mercy.

[leroyjenkins6969]

May any and all lords have mercy.

Let's consider the mortal world for now: what kind of mercy will admissions officers have on those of us who apply "early" enough?

So, what do you all think is considered early?

[outlawscr10]

It would be interesting to see data on when applications are normally sent.

[leroyjenkins6969]

That would be interesting.

I think Yale's website says that applicants who submit before November 15 are statistically more likely to be accepted.

[outlawscr10]

To be fair though, it's in their best interest to have applications come in evenly, rather than in a deluge at the end. It's in everyone's best interest. I've always kinda thought it was a pragmatic stance for schools to take on the matter.

[leroyjenkins6969]

Good point.

Also, from what I can remember, most applications are sent in after January 1st (source: https://www.facebook.com/events/1401294676752926/ Not sure if there is a video out there ).

[Nova]

Before thx giving is early

Before Christmas is on time

IMO

[drevo]

That would be interesting.

I think Yale's website says that applicants who submit before November 15 are statistically more likely to be accepted.

Yale is sort of known for not using rolling admissions in the way every other law school does. Now your statement could more correspond to the top tier students having their shit ready to go at the beginning of the cycle as opposed to those who submit later. Even on the TLS profile of Yale it states, " However, it should be noted that applying earlier in the cycle does not provide any boost to an applicant’s chances of being admitted."

[elterrible78]

That would be interesting.

I think Yale's website says that applicants who submit before November 15 are statistically more likely to be accepted.

Yale is sort of known for not using rolling admissions in the way every other law school does. Now your statement could more correspond to the top tier students having their shit ready to go at the beginning of the cycle as opposed to those who submit later. Even on the TLS profile of Yale it states, " However, it should be noted that applying earlier in the cycle does not provide any boost to an applicant’s chances of being admitted."

Based on my statistical analysis, it definitely doesn't.

[drawstring]

That would be interesting.

I think Yale's website says that applicants who submit before November 15 are statistically more likely to be accepted.

Yale is sort of known for not using rolling admissions in the way every other law school does. Now your statement could more correspond to the top tier students having their shit ready to go at the beginning of the cycle as opposed to those who submit later. Even on the TLS profile of Yale it states, " However, it should be noted that applying earlier in the cycle does not provide any boost to an applicant’s chances of being admitted."

Based on my statistical analysis, it definitely doesn't.

What about the rest of HYSCCN?

[drevo]

What about the rest of HYSCCN?

I think that conventional wisdom for the rest of all law schools is that the earlier the better. Now in regards to the poll and Before Nov. 15, Thanksgiving, Dec, etc. I haven't done any in depth reading into that but I can almost guarantee someone on TLS has and can hopefully chime in or link you to it.

[Nova]

El terrible. Plz gimme a job in a decade. Plz plz.

(/5am drunk nova)

[elterrible78]

That would be interesting.

I think Yale's website says that applicants who submit before November 15 are statistically more likely to be accepted.

Yale is sort of known for not using rolling admissions in the way every other law school does. Now your statement could more correspond to the top tier students having their shit ready to go at the beginning of the cycle as opposed to those who submit later. Even on the TLS profile of Yale it states, " However, it should be noted that applying earlier in the cycle does not provide any boost to an applicant’s chances of being admitted."

Based on my statistical analysis, it definitely doesn't.

What about the rest of HYSCCN?

Using LSN data, and controlling for LSAT, GPA, timing of decision, ED (where applicable), gender, non-trad status, and URM status, including only the 2010-11 through 2012-2013 cycles, and using only people whose final outcome was either "accept" or "reject" (the number is the % increase in your chances of an acceptance for each earlier-month you submitted - so Oct v. Nov, for example):

Yale: Not statistically significant
Harvard: 23.6%
Stanford: Not statistically significant
Chicago: 126.5% (not a typo)
Columbia: Not statistically significant
NYU: Not statistically significant
UVA: 30.7%
Penn: 45.9%
Berkeley: 46.4%
Michigan: 50.3%
Duke: Not statistically significant
Northwestern: 29.1%
Cornell: 27.4%
GULC: 22.4%

I just used the 2010-2013 data because 2010 is when the number of applicants started dropping sharply. If you include all the cycles available from LSN, a lot more of these schools look like they give a boost for applying early.

[elterrible78]

El terrible. Plz gimme a job in a decade. Plz plz.

(/5am drunk nova)

If I have one myself, I'll see what I can do.

[Nova]

Tyty

[cotiger]

Using LSN data, and controlling for LSAT, GPA, timing of decision, ED (where applicable), gender, non-trad status, and URM status, including only the 2010-11 through 2012-2013 cycles, and using only people whose final outcome was either "accept" or "reject" (the number is the % increase in your chances of an acceptance for each earlier-month you submitted - so Oct v. Nov, for example):

Yale: Not statistically significant
Harvard: 23.6%
Stanford: Not statistically significant
Chicago: 126.5% (not a typo)
Columbia: Not statistically significant
NYU: Not statistically significant
UVA: 30.7%
Penn: 45.9%
Berkeley: 46.4%
Michigan: 50.3%
Duke: Not statistically significant
Northwestern: 29.1%
Cornell: 27.4%
GULC: 22.4%

I just used the 2010-2013 data because 2010 is when the number of applicants started dropping sharply. If you include all the cycles available from LSN, a lot more of these schools look like they give a boost for applying early.

Would you mind posting the full results table? What are the results and p-values of those deemed not statistically significant?

There was some confusion on another thread where some people thought NSS implied a near-zero effect, which is not necessarily true.

The full table for ED would also be super useful because people are trying to make decisions based on the size of those effects.

[elterrible78]

Using LSN data, and controlling for LSAT, GPA, timing of decision, ED (where applicable), gender, non-trad status, and URM status, including only the 2010-11 through 2012-2013 cycles, and using only people whose final outcome was either "accept" or "reject" (the number is the % increase in your chances of an acceptance for each earlier-month you submitted - so Oct v. Nov, for example):

Yale: Not statistically significant
Harvard: 23.6%
Stanford: Not statistically significant
Chicago: 126.5% (not a typo)
Columbia: Not statistically significant
NYU: Not statistically significant
UVA: 30.7%
Penn: 45.9%
Berkeley: 46.4%
Michigan: 50.3%
Duke: Not statistically significant
Northwestern: 29.1%
Cornell: 27.4%
GULC: 22.4%

I just used the 2010-2013 data because 2010 is when the number of applicants started dropping sharply. If you include all the cycles available from LSN, a lot more of these schools look like they give a boost for applying early.

Would you mind posting the full results table? What are the results and p-values of those deemed not statistically significant?

There was some confusion on another thread where some people thought NSS implied a near-zero effect, which is not necessarily true.

The full table for ED would also be super useful because people are trying to make decisions based on the size of those effects.

The P-values on most of them are not even close to being in a long line to buy a ticket to the ballpark, but tell me exactly what info you want, and I'll whip something up. I'm probably not going to post a full regression results table for each school, though.

[cotiger]

Would you mind posting the full results table? What are the results and p-values of those deemed not statistically significant?

There was some confusion on another thread where some people thought NSS implied a near-zero effect, which is not necessarily true.

The full table for ED would also be super useful because people are trying to make decisions based on the size of those effects.

The P-values on most of them are not even close to being in a long line to buy a ticket to the ballpark, but tell me exactly what info you want, and I'll whip something up. I'm probably not going to post a full regression results table for each school, though.

I'm primarily just interested in the coefficients and p-values for the ED results. Lots of people ask about where to ED or whether it makes a difference, and it would be awesome to be able to tell them something other than merely "the effect at Penn is not provable."

Thanks for all the work you've done!

[elterrible78]

Would you mind posting the full results table? What are the results and p-values of those deemed not statistically significant?

There was some confusion on another thread where some people thought NSS implied a near-zero effect, which is not necessarily true.

The full table for ED would also be super useful because people are trying to make decisions based on the size of those effects.

The P-values on most of them are not even close to being in a long line to buy a ticket to the ballpark, but tell me exactly what info you want, and I'll whip something up. I'm probably not going to post a full regression results table for each school, though.

I'm primarily just interested in the coefficients and p-values for the ED results. Lots of people ask about where to ED or whether it makes a difference, and it would be awesome to be able to tell them something other than merely "the effect at Penn is not provable."

Thanks for all the work you've done!

No problem! Here's what I have. Just as a disclaimer, this is based on LSN numbers (and doesn't include the verrrrrry tail end of last cycle), and the results are for ALL applicants (splitters, reverse, URM, EVERYONE) for the 2010-11, 2011-12, and 2012-13 cycles. If anyone has any questions about what it means, I'll be happy to give my best (non-expert law student) explanation, and welcome input from anyone else.

[cotiger]

Would you mind posting the full results table? What are the results and p-values of those deemed not statistically significant?

There was some confusion on another thread where some people thought NSS implied a near-zero effect, which is not necessarily true.

The full table for ED would also be super useful because people are trying to make decisions based on the size of those effects.

The P-values on most of them are not even close to being in a long line to buy a ticket to the ballpark, but tell me exactly what info you want, and I'll whip something up. I'm probably not going to post a full regression results table for each school, though.

I'm primarily just interested in the coefficients and p-values for the ED results. Lots of people ask about where to ED or whether it makes a difference, and it would be awesome to be able to tell them something other than merely "the effect at Penn is not provable."

Thanks for all the work you've done!

No problem! Here's what I have. Just as a disclaimer, this is based on LSN numbers (and doesn't include the verrrrrry tail end of last cycle), and the results are for ALL applicants (splitters, reverse, URM, EVERYONE) for the 2010-11, 2011-12, and 2012-13 cycles. If anyone has any questions about what it means, I'll be happy to give my best (non-expert law student) explanation, and welcome input from anyone else.

Awesome. Thanks so much!

[MoMettaMonk]

Would you mind posting the full results table? What are the results and p-values of those deemed not statistically significant?

There was some confusion on another thread where some people thought NSS implied a near-zero effect, which is not necessarily true.

The full table for ED would also be super useful because people are trying to make decisions based on the size of those effects.

The P-values on most of them are not even close to being in a long line to buy a ticket to the ballpark, but tell me exactly what info you want, and I'll whip something up. I'm probably not going to post a full regression results table for each school, though.

I'm primarily just interested in the coefficients and p-values for the ED results. Lots of people ask about where to ED or whether it makes a difference, and it would be awesome to be able to tell them something other than merely "the effect at Penn is not provable."

Thanks for all the work you've done!

No problem! Here's what I have. Just as a disclaimer, this is based on LSN numbers (and doesn't include the verrrrrry tail end of last cycle), and the results are for ALL applicants (splitters, reverse, URM, EVERYONE) for the 2010-11, 2011-12, and 2012-13 cycles. If anyone has any questions about what it means, I'll be happy to give my best (non-expert law student) explanation, and welcome input from anyone else.

As someone who has no idea what these numbers mean, a brief explanation would be greatly appreciated.

[cotiger]

Actually, one last request. Could we get the chart for ED splitters, too? Lots of potential ED people here are big-time splitters.

[cotiger]

As someone who has no idea what these numbers mean, a brief explanation would be greatly appreciated.

Basically, the coefficient is our estimation of what the effect on admissions likelihood is for either applying one month earlier (on the left) or ED (on the right). The p-values to the right of the coefficients are a representation of how sure we are that effect is different from zero. The lower the p-value, the more certain we are that the observed effect really exists.

We call a result with a very large p-value "Not statistically significant" because it is very unlikely that the effect is real. For instance, superficially, the coefficients indicate that applying to Duke earlier in the cycle or EDing to NYU actually make it less likely to be accepted. However, the large p-values tell us what we intuitively know: that that's ridiculous and really much more likely that there's just no effect.

[elterrible78]

As someone who has no idea what these numbers mean, a brief explanation would be greatly appreciated.[/quote]

Heh, no problem!

There are two different factors here: earlier application submission (Early) and binding-early decision application (ED) - you probably got that, though. I do want to point out that the "early" application means an earlier month, not a month early. So we're broadly comparing October v. November applicants, for example, no matter when in the month they applied. That needs to be clarified.

The coefficient for each variable is that substantive effect. In my analysis, the dependent variable is either 1 (applicant was admitted) or 0 (applicant was rejected). The coefficient shows you the "effect" of an independent variable (here, either "early" or "ed").

The coefficient for early = the % increase in your likelihood of being admitted for each 1-point increase in the value of the dependent variable (for "early", a one-point increase is equal to an earlier month the application was sent...September is a-point increase over October).

The coefficient for ED is the same, only here the independent variable is also binary, so it's either a 0 or a 1. In other words, the ED coefficient shows you the % increase in your likelihood of being admitted if you apply ED, period.

The P-values for each is basically the chance that the result is just...well, by chance, and not because of any actual causal effect. The lower the P-value, the higher the probability that there is actually something going on there rather than just random chance. In statistics, there are different "levels of confidence" that researchers use, but the basic cutoff that is used is .10 (beyond that, nobody really is comfortable saying that this is a statistically significant result). Some studies use .05, and those who want to be really strict use .01.

So, in this example, let's take Chicago: Applicants are 126.5% more likely to be admitted for each earlier month they apply, and there is a 0.000% likelihood that this is just random chance rather than an actual effect. Applicants are also 536% more likely to be admitted applying ED vs. regular decision, and there is a 2% chance that this result is just random chance, rather than an actual effect.

Again, I want to throw in that this is all controlling for LSAT, GPA, URM status, nontraditional status, and gender. In other words, don't take your 158 & 2.95 combo and try to convince yourself that if you apply ED, you're 5 times more likely to be admitted than a 175 / 3.8 regular-decision applicant.

Hope that helped!