Is the Bar Graded on a Curve? Forum

[lawschoolwoohoo]

Just wondering. Barbri told us it was, and that as long as you're above a certain percentile, you're probably in business. Meanwhile, my law school's dean sent us a pretty threatening email a couple days ago about how the bar is not graded on a curve and how people that could've passed in 2013 now cannot pass because of a reverse curve that makes people's scores go down. That e-mail really got in my head, so I just wanted to know you all's take on it?

[rmhco]

(I don't have an answer I'm confident giving in your question)

No offense but that seems like a really shitty thing for your law school to do a few days before the bar! Maybe at the outset, but a psych out when we're counting the hours?

[0lol]

wtf is a reverse curve

[Lemming]

Your dean is correct. The bar exam is equated based on the MBE scores, not curved. Equating assumes that the testing pool has roughly the same ability year after year. Therefore if 2016 bar takers as a whole do worse on the MBE than previous years then the bar examiners will give less points to everyones MPT and MEE scores.

This is an extreme simplification but assume if last year the median raw score on the MBE was 130 and then this year the MBE median was a 125, then administrators will apply a lower scaled score to the essays and MPT's. So if someone writes exactly the same essay in both years, the essay will get a lower score in the year that the testing pool as a whole scored lower. This is a problem because as we've known for a few years LSAT scores are down and less qualified people are going to law school.

So yes, expect a decline in bar passage rates for the next couple years unless NCBEX fixes something.

[lawschoolwoohoo]

Your dean is correct. The bar exam is equated based on the MBE scores, not curved. Equating assumes that the testing pool has roughly the same ability year after year. Therefore if 2016 bar takers as a whole do worse on the MBE than previous years then the bar examiners will give less points to everyones MPT and MEE scores.

This is an extreme simplification but assume if last year the median raw score on the MBE was 130 and then this year the MBE median was a 125, then administrators will apply a lower scaled score to the essays and MPT's. So if someone writes exactly the same essay in both years, the essay will get a lower score in the year that the testing pool as a whole scored lower. This is a problem because as we've known for a few years LSAT scores are down and less qualified people are going to law school.

So yes, expect a decline in bar passage rates for the next couple years unless NCBEX fixes something.

Thank you! So how concerned do I need to be? Is this basically a foregone conclusion of failing?

And to the above poster, yes, it is shitty of the dean. My school had an over 10% drop in bar passage last year so we've been getting some "encouraging" emails all spring. I wish she would've sent it a little earlier before bar week.

[not guilty]

good move by dean, just poor timing.

that was an early June email.

[Lemming]

Looking at the difference in scores between 2014 and 2015 as well as bar passage rate decline at certain schools, there was an decrease in overall bar passage rates but it seems that decrease was confined to the T2 and lower schools. The CA bar exam saw about a 2% decrease in bar passage rates from 2014 to 2015. Stanford, Berkley, USC, and UCLA saw minuscule drops if not increases in the bar pass rate over the same period where as the lower ranked schools saw big declines in pass rate.

So top schools have top students and the decrease in bar passage rates are unlikely to have an effect, but schools that have a lot of students on the bubble (passing with only a few points to spare) is at risk of large declines in pass rates.

[lawschoolwoohoo]

Looking at the difference in scores between 2014 and 2015 as well as bar passage rate decline at certain schools, there was an decrease in overall bar passage rates but it seems that decrease was confined to the T2 and lower schools. The CA bar exam saw about a 2% decrease in bar passage rates from 2014 to 2015. Stanford, Berkley, USC, and UCLA saw minuscule drops if not increases in the bar pass rate over the same period where as the lower ranked schools saw big declines in pass rate.

So top schools have top students and the decrease in bar passage rates are unlikely to have an effect, but schools that have a lot of students on the bubble (passing with only a few points to spare) is at risk of large declines in pass rates.

Thank you. That's very helpful. My school is (not sure anymore as of 2016) but in 2015 (and before) a T-20 but had a big drop. I graduated in the top 20% and have done all of Barbri/a lot of adaptibar so I hope I'm okay, but that is great information to have. Yeah, I definitely think it was good to let us know, but really not right now. At this point, less than a week out, ignorance is bliss.

[SLS_AMG]

Your dean is correct. The bar exam is equated based on the MBE scores, not curved. Equating assumes that the testing pool has roughly the same ability year after year. Therefore if 2016 bar takers as a whole do worse on the MBE than previous years then the bar examiners will give less points to everyones MPT and MEE scores.

This is an extreme simplification but assume if last year the median raw score on the MBE was 130 and then this year the MBE median was a 125, then administrators will apply a lower scaled score to the essays and MPT's. So if someone writes exactly the same essay in both years, the essay will get a lower score in the year that the testing pool as a whole scored lower. This is a problem because as we've known for a few years LSAT scores are down and less qualified people are going to law school.

So yes, expect a decline in bar passage rates for the next couple years unless NCBEX fixes something.

The equating is definitely correct re: the MBE, but how certain are you that it's also used with the MEE? I don't see how it could be. The MBE reuses questions so it can compare performance over time, but MEE essays aren't used in every MBE jurisdiction and are new with each exam. It would seem impossible to "equate" the essays based on these discrepancies.

[mvp99]

I read there is also another problem: equating is only reliable when you compare very similar exams and the new exam added civ pro. The new scaled score cannot be compared reliable to old scaled scores.

[Lemming]

Your dean is correct. The bar exam is equated based on the MBE scores, not curved. Equating assumes that the testing pool has roughly the same ability year after year. Therefore if 2016 bar takers as a whole do worse on the MBE than previous years then the bar examiners will give less points to everyones MPT and MEE scores.

This is an extreme simplification but assume if last year the median raw score on the MBE was 130 and then this year the MBE median was a 125, then administrators will apply a lower scaled score to the essays and MPT's. So if someone writes exactly the same essay in both years, the essay will get a lower score in the year that the testing pool as a whole scored lower. This is a problem because as we've known for a few years LSAT scores are down and less qualified people are going to law school.

So yes, expect a decline in bar passage rates for the next couple years unless NCBEX fixes something.

The equating is definitely correct re: the MBE, but how certain are you that it's also used with the MEE? I don't see how it could be. The MBE reuses questions so it can compare performance over time, but MEE essays aren't used in every MBE jurisdiction and are new with each exam. It would seem impossible to "equate" the essays based on these discrepancies.

http://www.ncbex.org/assets/media_files ... column.pdf

Here's an article that explains how they scale the MEE off the MBE. Obviously states with their own essays don't have to follow the same scaling of essays.

[whitecollar23]

There's a lot of stats there, but it seems that the purpose of the scaling is to ensure the distribution of the MEE scores are the same as that of the MBE. The goal would be to ensure that the MEE difficulty is equated to the MBE.

What that means is that if the MEE is for some reason really easy one year, it will be scaled down to line up with the MBE mean. Further, if the MEE is too difficult, it will be scaled upward to line up with the MBE mean. The scaling is done to the scores across the board, if I'm not mistaken. It's not necessarily a reverse-curve. Rather, it's an alignment curve, to align MEE with MBE.

I read through it quickly, but I believe with this curve, it will be more difficult to pass in February where the average MBE score is lower.

Thoughts?

[whitecollar23]

tl;dr

It's not a reverse curve. Rather, it's an equalizing curve. If the written portion is too easy, it will be curve the mean score down and equalize the standard deviation. If the written portion is too difficult, it will curve the mean score up and also equalize the deviation.

So if you come across some really tough essays, realize that the portion of the exam might be curved up.

[whitecollar23]

Correction:
The scaling does not affect pass rates. I still don't completely understand the article, but from what I've read, it's not some reverse-curve and not something we should worry about.

If someone with a stronger background in statistics can explain better, please do so. Thanks!

[rmhco]

Correction:
The scaling does not affect pass rates. I still don't completely understand the article, but from what I've read, it's not some reverse-curve and not something we should worry about.

If someone with a stronger background in statistics can explain better, please do so. Thanks!

That is reassuring - I was going into a complete panic. Hopefully someone can confirm

[whitecollar23]

Correction:
The scaling does not affect pass rates. I still don't completely understand the article, but from what I've read, it's not some reverse-curve and not something we should worry about.

If someone with a stronger background in statistics can explain better, please do so. Thanks!

That is reassuring - I was going into a complete panic. Hopefully someone can confirm

Though sorta seems like doing terribly on the essays will be scaled down and that doing really well will be scaled up. Which is odd. But that might just be the example and in reality the SD might be higher with the essays, so the reverse of that would be the case to scale toward the MBE.

STATS EXPERT PLEASE COME HERE AND CLARIFY ALL OF THIS.

[BigZuck]

There has got to be better things you can do with your time right now than worrying about this. Just do your best and let the chips fall where they may.

Keep grinding.

Or take a break and get some rest. But don't worry about this.

[Calvin Murphy]

There has got to be better things you can do with your time right now than worrying about this. Just do your best and let the chips fall where they may.

Keep grinding.

Or take a break and get some rest. But don't worry about this.

So you're saying we should hit up the casino and the club?