LSN and Harvard

(Applications Advice, Letters of Recommendation . . . )
XtraordinaryMachine
Posts: 34
Joined: Tue Dec 22, 2009 3:30 am

LSN and Harvard

Postby XtraordinaryMachine » Mon Dec 06, 2010 5:21 am

Two Considerations:

1) The number of users for Harvard on LSN for the 09-10 cycle was ~735.
735/7610 (total #of Harvard applicants last cycle) = <10%; therefore LSN reflects less than 10% o fthe applicant pool, making it highly unlikely that it is a representative sample. Not sure what way its skewed if it really is. (Thoughts? Personal Experience?)

2) With all the emphasis on applying early, it seems to me it benefits mostly reverse splitters, (at least that is what I've gathered from LSN 09-10 cycle, looking at only those who have gone complete by Oct 1 since that is when I'm planning on applying).

**

User avatar
Eugenie Danglars
Posts: 2353
Joined: Wed Jul 07, 2010 12:04 pm

Re: LSN and Harvard

Postby Eugenie Danglars » Mon Dec 06, 2010 5:25 am

LSN is not a representative sample. But it's better than no sample? You just have to keep in mind it's not the full picture.

I would imagine that in general (maybe not for H, but for others), the graphs are skewed positive. People are more likely to report a success than a failure, right?

whymeohgodno
Posts: 2508
Joined: Mon Jul 19, 2010 8:15 pm

Re: LSN and Harvard

Postby whymeohgodno » Mon Dec 06, 2010 5:25 am

LSN isn't random sampling so it's not completely accurate but there is nothing wrong with it's sampling size. The only thing you could criticize about it is that it's a completely self selecting sample which could result in systematic sampling error.

XtraordinaryMachine
Posts: 34
Joined: Tue Dec 22, 2009 3:30 am

Re: LSN and Harvard

Postby XtraordinaryMachine » Mon Dec 06, 2010 5:30 am

Completely agree that people are more likely to report a success than failure, but I have a feeling the people on LSN are all overachievers and obsessive (I'm obviously not here to judge;)

Although I may be telling myself that for false hope.

whymeohgodno
Posts: 2508
Joined: Mon Jul 19, 2010 8:15 pm

Re: LSN and Harvard

Postby whymeohgodno » Mon Dec 06, 2010 5:34 am

XtraordinaryMachine wrote:Completely agree that people are more likely to report a success than failure, but I have a feeling the people on LSN are all overachievers and obsessive (I'm obviously not here to judge;)

Although I may be telling myself that for false hope.


False hope? If they were all overachivers+obsessive that would make them more likely to be more qualified applicants soft wise which would make LSN more positively skewed which would mean that if anything it would lessen your hope, not give you false hope.

Unless you hope LSN is overly optimistic, but I can't imagine why you would.

Also as to more likely to report success than failure, I agree. I usually view some ranges of numbers with skepticism if there are a ton of Pendings.

stilles
Posts: 93
Joined: Sun Nov 28, 2010 12:15 am

Re: LSN and Harvard

Postby stilles » Mon Dec 06, 2010 5:54 am

whymeohgodno wrote:
XtraordinaryMachine wrote:Completely agree that people are more likely to report a success than failure, but I have a feeling the people on LSN are all overachievers and obsessive (I'm obviously not here to judge;)

Although I may be telling myself that for false hope.


False hope? If they were all overachivers+obsessive that would make them more likely to be more qualified applicants soft wise which would make LSN more positively skewed which would mean that if anything it would lessen your hope, not give you false hope.

Unless you hope LSN is overly optimistic, but I can't imagine why you would.

Also as to more likely to report success than failure, I agree. I usually view some ranges of numbers with skepticism if there are a ton of Pendings.


I think OP meant to say LSN iis full of people who are 'type a' and thus already have high numbers and/or are extremely tenacious in their monitoring of LSN and such. However this assumption may be something he (or she) is telling him/herself for a delusional chance of acceptance. That's what I get out of it, b/c NO ONE wants LSN to be over optimistic.

User avatar
Na_Swatch
Posts: 472
Joined: Sun Sep 20, 2009 4:40 pm

Re: LSN and Harvard

Postby Na_Swatch » Mon Dec 06, 2010 5:55 am

XtraordinaryMachine wrote:Two Considerations:

1) The number of users for Harvard on LSN for the 09-10 cycle was ~735.
735/7610 (total #of Harvard applicants last cycle) = <10%; therefore LSN reflects less than 10% o fthe applicant pool, making it highly unlikely that it is a representative sample. Not sure what way its skewed if it really is. (Thoughts? Personal Experience?)

2) With all the emphasis on applying early, it seems to me it benefits mostly reverse splitters, (at least that is what I've gathered from LSN 09-10 cycle, looking at only those who have gone complete by Oct 1 since that is when I'm planning on applying).

**


haha never expect TLS to get anything involving statistics correct:

1) As everybody has said before there is nothing wrong with the sample size... only sampling error would cause differences and these aren't likely to be too large, consider:

-People who wind up with the LSAT / GPA to be near HLS medians are, on the whole, more organized & motivated in general. Thus likely a decent amount WILL find LSN... sampling error will be based more on their own willingness to post information which would appear to be relatively random.

-Averages/ Numbers from LSN seem to match up at least fairly close to the actual numbers for HLS's stats.

-You do need to consider just basic/ logical approaches to reading LSN numbers, such as a high chance for longterm Pending accounts = Rejected/Waitlisted.

2) Here it seems to be of MORE help to reverse splitters (well i dont know if you would call a 4.0, 170 a reverse splitter but i guess for hls it is) but still there is substantial benefit it seems to everyone who applies early.

stilles
Posts: 93
Joined: Sun Nov 28, 2010 12:15 am

Re: LSN and Harvard

Postby stilles » Mon Dec 06, 2010 6:11 am

-People who wind up with the LSAT / GPA to be near HLS medians are, on the whole, more organized & motivated in general. Thus likely a decent amount WILL find LSN... sampling error will be based more on their own willingness to post information which would appear to be relatively random.-Great Point

-Averages/ Numbers from LSN seem to match up at least fairly close to the actual numbers for HLS's stats. For the 09-10 cycle OP mentioned, it seems a little harsh actually, however nothing too out of the ordinary

-You do need to consider just basic/ logical approaches to reading LSN numbers, such as a high chance for longterm Pending accounts = Rejected/Waitlisted. Can you explain what effect this can have (for us non-math/stats majors:).[/quote]

User avatar
Deuce
Posts: 1048
Joined: Fri Jun 25, 2010 11:12 am

Re: LSN and Harvard

Postby Deuce » Mon Dec 06, 2010 6:36 am

Na_Swatch wrote:
XtraordinaryMachine wrote:Two Considerations:

1) The number of users for Harvard on LSN for the 09-10 cycle was ~735.
735/7610 (total #of Harvard applicants last cycle) = <10%; therefore LSN reflects less than 10% o fthe applicant pool, making it highly unlikely that it is a representative sample. Not sure what way its skewed if it really is. (Thoughts? Personal Experience?)

2) With all the emphasis on applying early, it seems to me it benefits mostly reverse splitters, (at least that is what I've gathered from LSN 09-10 cycle, looking at only those who have gone complete by Oct 1 since that is when I'm planning on applying).

**


haha never expect TLS to get anything involving statistics correct:

1) As everybody has said before there is nothing wrong with the sample size... only sampling error would cause differences and these aren't likely to be too large, consider:

-People who wind up with the LSAT / GPA to be near HLS medians are, on the whole, more organized & motivated in general. Thus likely a decent amount WILL find LSN... sampling error will be based more on their own willingness to post information which would appear to be relatively random.

-Averages/ Numbers from LSN seem to match up at least fairly close to the actual numbers for HLS's stats.

-You do need to consider just basic/ logical approaches to reading LSN numbers, such as a high chance for longterm Pending accounts = Rejected/Waitlisted.

2) Here it seems to be of MORE help to reverse splitters (well i dont know if you would call a 4.0, 170 a reverse splitter but i guess for hls it is) but still there is substantial benefit it seems to everyone who applies early.

:lol: :lol: :lol: :lol: :lol: :lol: :lol: :lol: :lol: :lol: :lol: :lol: :lol: :lol:

User avatar
Nulli Secundus
Posts: 2625
Joined: Mon Jun 21, 2010 7:19 am

Re: LSN and Harvard

Postby Nulli Secundus » Mon Dec 06, 2010 6:51 am

Yeah definitely people prefer reporting success. For example, my HLS application status will remain forever "Pending" in the event of a ding (99%). I will become a legend, grandfathers will show my profile to their grandchildren and say "This guy applied to HLS with me and it is said that he is still refreshing his status checker to this day." and will add "What a loser lol".

bdubs
Posts: 3729
Joined: Sat Feb 06, 2010 2:23 pm

Re: LSN and Harvard

Postby bdubs » Mon Dec 06, 2010 1:32 pm

152 acceptances on LSN for last cycle out of 833 total acceptances (18%). That is about double the rate of acceptance relative to the total applicant pool.

LSN may be a bit more biased for Harvard than for others. Lots of people who have no shot at Harvard and who don't know anything about admissions send in applications hoping that they will get in because they are special snowflakes.

whymeohgodno
Posts: 2508
Joined: Mon Jul 19, 2010 8:15 pm

Re: LSN and Harvard

Postby whymeohgodno » Mon Dec 06, 2010 2:40 pm

bdubs wrote:152 acceptances on LSN for last cycle out of 833 total acceptances (18%). That is about double the rate of acceptance relative to the total applicant pool.

LSN may be a bit more biased for Harvard than for others. Lots of people who have no shot at Harvard and who don't know anything about admissions send in applications hoping that they will get in because they are special snowflakes.


Take all the pendings and add them in as rejections lolol.

User avatar
Na_Swatch
Posts: 472
Joined: Sun Sep 20, 2009 4:40 pm

Re: LSN and Harvard

Postby Na_Swatch » Mon Dec 06, 2010 6:30 pm

stilles wrote:
Na_swatch wrote:-People who wind up with the LSAT / GPA to be near HLS medians are, on the whole, more organized & motivated in general. Thus likely a decent amount WILL find LSN... sampling error will be based more on their own willingness to post information which would appear to be relatively random.-Great Point

-Averages/ Numbers from LSN seem to match up at least fairly close to the actual numbers for HLS's stats. For the 09-10 cycle OP mentioned, it seems a little harsh actually, however nothing too out of the ordinary

-You do need to consider just basic/ logical approaches to reading LSN numbers, such as a high chance for longterm Pending accounts = Rejected/Waitlisted. Can you explain what effect this can have (for us non-math/stats majors:).


Hmm some general pointers about my last point:

-LSN will have a higher rate of acceptance on a whole due to the fact that LSN's entire sample represents a more organized group so you will have a lot less the applying to HLS with say 3.5, 165 types than in reality.

-However this previous fact about applicant pool doesn't really mess with the accuracy of the data as median and quartiles are calculated from the accepted pool, not applicant pool.

-Then, as mentioned before, you need to count pending's as rejections and also take extreme outliers with a grain of salt (the usual URM flames on LSN, the 4.0, 180 fake accounts, etc.)

-Finally LSN's numbers are a little higher as a whole as waitlisted applicant data is a bit under-represented... those who get in off the waitlist very late are probably not as prone to update/ post information as during the main cycle due to rush/ long time passing/ forgetfulness/ etc.

The Takeaway: LSN is the best data you can get for sure [I used it during my application cycle for 09-10 and found it quite accurate] but on the margins or for things like waitlist chances it has some deficiencies




Return to “Law School Admissions Forum”

Who is online

Users browsing this forum: Bing [Bot], chargers21, iluvlawschool, Purple Post It Note and 8 guests