Law School Predictors

Ohnoeshalpme

How reliable are law school predictors in general? Which ones are the most reliable? Which ones did you use?

Kateryna

Look at your median scores and it will give you a clear picture. I would suggest lsac website as most reliable. To get general idea use lawshoolnumbers

Seeking Perfection

Law school numbers is good especially since it also provides scholarship info. http://mylsn.info/dispresults.php is a nice way to automatically sort through some of the data from law school numbers.
The main problem with law school numbers is that it is all self reported which introduces problems of response bias.

There is also an LSAC predictor which is based on less specific data. I think it is based on the medians and maybe the 25ths and 75th percentiles. Schools can opt out so this won't work for all schools. It's better because you don't have to worry about bias and because it's a regression,but worse because the data is less specific and you can't see things like the fact that WUSTL offers nearly everyone with a high LSAT a big scholarship.
https://officialguide.lsac.org/release/ugpalsat/ugpalsat.aspx

Hourmd looks like it used to be good but uses outdated data now. It also used the law school numbers data, but tried to run a regression on it. This regression would have suffered from bias, but might have gave more info than just the numbers.

All of these have the core problem that if the next cycle is different than the previous ones, the predictors can't tell you. For instance there was an uptivk in high scoring LSAT takers this cycle so lots of people have been underperforming their numbers. Most also can't account for work experience.

Ohnoeshalpme

@"Seeking Perfection" said:
Law school numbers is good especially since it also provides scholarship info. http://mylsn.info/dispresults.php is a nice way to automatically sort through some of the data from law school numbers.
The main problem with law school numbers is that it is all self reported which introduces problems of response bias.

There is also an LSAC predictor which is based on less specific data. I think it is based on the medians and maybe the 25ths and 75th percentiles. Schools can opt out so this won't work for all schools. It's better because you don't have to worry about bias and because it's a regression,but worse because the data is less specific and you can't see things like the fact that WUSTL offers nearly everyone with a high LSAT a big scholarship.
https://officialguide.lsac.org/release/ugpalsat/ugpalsat.aspx

Hourmd looks like it used to be good but uses outdated data now. It also used the law school numbers data, but tried to run a regression on it. This regression would have suffered from bias, but might have gave more info than just the numbers.

All of these have the core problem that if the next cycle is dufferent than the previous ones, the predictors can't tell you. For instance there was an uptivk in high scoring LSAT takers this cycle so lots of people have been underperforming their numbers. Most also can't account for work experience.

this is really good, thanks for the updated info

nathanieljschwartz

@lelinwek @"Dillon A. Wright" pretty funny post but pretty sure its a bot