Score Bands Forum

[ConsideringLawSchool]

Why is 174-179 the score band for 177? Thanks.

[musicfor18]

It's a predictive model of the range of scores you would likely have if you took the LSAT many times, since any single measurement can have errors. I'm not sure how much schools pay attention to it, at all.

[JasonR]

I'm not sure how much schools pay attention to it, at all.

None.

[Flanker1067]

It doesn't take much to figure out that if someone took the LSAT multiple times in a row that they probably won't score the exact same, but would be in the same range.

[skip james]

i think the asymmetrical shape is due to the fact that, not just in terms of difficulty, but in terms of physical possibility, it's just easier to score lower than higher at 177. i would bet that a person who scored a 123 would have a similar band in the other direction, i.e. 121- 126.

[Richie Tenenbaum]

Why is 174-179 the score band for 177? Thanks.

Hrmmm, interesting. No clue why they would not put 174-180. Maybe the amount of 180s are much lower then all the rest of 170s, so they decided to tweak the score band a little? That really does not make too much since though since isn't it only around .1 points between 179 and 180? Like 99.8x%ile to 99.99%ile? These %iles could be off though.

[sdv]

It's been mentioned before, but basically once you hit the 99th percentile (around 173 I think???) a certain amount of luck becomes a factor in differentiating scores (much more so than in the lower ranges). at that point a person who scores 1 or two points higher or lower isn't necessarily better at the test, but rather managed to avoid more of the "oh crap i read that question wrong" moments that pretty much everyone has at that level. As to why the score band doesn't go to 180, it's probably so statistically insignificant that it can't be included in the score band.

I'm a personal believer in the score band; my 174 had a band of 171-177, which I think is extremely accurate, since on easier practice tests (which are less forgiving of mistakes) I tended to score as low as 171, while at the same time after looking at my test I made 2 inexcusably stupid mistakes that should have given me a 176. So, I personally find those bands to be eerily accurate.

[ConsideringLawSchool]

It's been mentioned before, but basically once you hit the 99th percentile (around 173 I think???) a certain amount of luck becomes a factor in differentiating scores (much more so than in the lower ranges). at that point a person who scores 1 or two points higher or lower isn't necessarily better at the test, but rather managed to avoid more of the "oh crap i read that question wrong" moments that pretty much everyone has at that level. As to why the score band doesn't go to 180, it's probably so statistically insignificant that it can't be included in the score band.

I'm a personal believer in the score band; my 174 had a band of 171-177, which I think is extremely accurate, since on easier practice tests (which are less forgiving of mistakes) I tended to score as low as 171, while at the same time after looking at my test I made 2 inexcusably stupid mistakes that should have given me a 176. So, I personally find those bands to be eerily accurate.

All makes sense, but your score band is -3/+3. My question is why the 177 band is -3/+2

[musicfor18]

Bands for scores that are at the upper or lower extreme of the score range are calculated differently. I think SDV gave a pretty good explanation of why, at the upper end, the band would be weighted more toward the left than right.

[ConsideringLawSchool]

Bands for scores that are at the upper or lower extreme of the score range are calculated differently. I think SDV gave a pretty good explanation of why, at the upper end, the band would be weighted more toward the left than right.

Is 180 ever in the band?

[skip james]

Bands for scores that are at the upper or lower extreme of the score range are calculated differently. I think SDV gave a pretty good explanation of why, at the upper end, the band would be weighted more toward the left than right.

Is 180 ever in the band?

i'm guessing for 178+ it is. but just a guess.

[HiLine]

All makes sense, but your score band is -3/+3. My question is why the 177 band is -3/+2

That means your chance of scoring -3 from 177 is the same as that for +2. And that is because the percentage of test-takers who score 180 is disproportionately low.

[PDaddy]

It's a predictive model of the range of scores you would likely have if you took the LSAT many times, since any single measurement can have errors. I'm not sure how much schools pay attention to it, at all.

Sad part is, that's what they are supposed to be paying attention to. Score bands (as opposed to individual numerical scores) are everything in "comparative admissions", which most adcoms claim is their model. Statistically, two people with similar GPA's and subjective features cannot be distinguished by a difference of a few points on the LSAT. But despite repeatedly being warned about this, adcoms refuse to stop taking shortcuts, and continue to use 1-3 point differences in LSAT scores as if they're actually meaningful.

[legallybound]

It's a predictive model of the range of scores you would likely have if you took the LSAT many times, since any single measurement can have errors. I'm not sure how much schools pay attention to it, at all.

Sad part is, that's what they are supposed to be paying attention to. Statistically, two people with similar GPA's and subjective features cannot be distinguished by a difference of a few points on the LSAT. But despite repeatedly being warned about this, adcoms refuse to stop taking shortcuts, and continue to use 1-3 point differences in LSAT scores as if they're actually meaningful.

+1

[musicfor18]

It's a predictive model of the range of scores you would likely have if you took the LSAT many times, since any single measurement can have errors. I'm not sure how much schools pay attention to it, at all.

Sad part is, that's what they are supposed to be paying attention to. Score bands (as opposed to individual numerical scores) are everything in "comparative admissions", which most adcoms claim is their model. Statistically, two people with similar GPA's and subjective features cannot be distinguished by a difference of a few points on the LSAT. But despite repeatedly being warned about this, adcoms refuse to stop taking shortcuts, and continue to use 1-3 point differences in LSAT scores as if they're actually meaningful.

+1

1 LSAT point can also mean several thousands of dollars more in scholarship money, too.

[PDaddy]

...right? If schools would do their homework more, subjective factors would play a larger part in admissions. Ironically, the model used on URM's is the ideal. All students should get that same treatment. But then the NULaw admissions model (i.e., the "B-School Model") would have to become THE model in law school admissions because most straight-out-of-UG students do not have the intangibles to allow schools to efficiently reduce the weight applied to the LSAT and focus more on soft factors. This is the problem with the misuse of the LSAT. It really disadvantages everyone.

[HiLine]

But despite repeatedly being warned about this, adcoms refuse to stop taking shortcuts, and continue to use 1-3 point differences in LSAT scores as if they're actually meaningful.

When adcoms have to make a decision on which candidates within that band should be accepted, they employ the most commonsense criterion: choose those at the top of the band, and blame the denials of the rest on mishap.

[tomwatts]

Bands for scores that are at the upper or lower extreme of the score range are calculated differently. I think SDV gave a pretty good explanation of why, at the upper end, the band would be weighted more toward the left than right.

Is 180 ever in the band?

I'm not sure, but my "band for average score" (from a 176 and a 180, so an average of 178) is 176-179.

[PDaddy]

But despite repeatedly being warned about this, adcoms refuse to stop taking shortcuts, and continue to use 1-3 point differences in LSAT scores as if they're actually meaningful.

When adcoms have to make a decision on which candidates within that band should be accepted, they employ the most commonsense criterion: choose those at the top of the band, and blame the denials of the rest on mishap.

Based on resources and manpower, that makes sense. That's exactly the problem, they should be able to compare students within bands if they look similar, then find subtle differences that will put certain students ahead of others in terms of "fit" within the class, diversity needs, etc. At its best, the LSAT would be used in that way.

[vanwinkle]

That's exactly the problem, they should be able to compare students within bands if they look similar, then find subtle differences that will put certain students ahead of others in terms of "fit" within the class, diversity needs, etc.

How is this not what they do now?

[booboo]

That's exactly the problem, they should be able to compare students within bands if they look similar, then find subtle differences that will put certain students ahead of others in terms of "fit" within the class, diversity needs, etc.

How is this not what they do now?

I was under the impression that each T14 school had one LSAT score as their 25th/Median/75th for matriculates.

[vanwinkle]

I was under the impression that each T14 school had one LSAT score as their 25th/Median/75th for matriculates.

I can't even figure out what this means.

[r6_philly]

I was under the impression that each T14 school had one LSAT score as their 25th/Median/75th for matriculates.

I can't even figure out what this means.

Only a single numerical score is reported in US News rankings. Or, no one cares about the bands.

Personally I felt like it's a gimmick by LSAT to say "well we don't want to say that your score is as an accurate predictor of your actual score potential, but the schools do whatever they want with it even though we don't agree."

[vanwinkle]

Only a single numerical score is reported in US News rankings.

But schools take more than just that one score. A school that has a 170 median is going to accept many students with LSAT scores above 170, as well as a number with LSAT scores below 170. Most of them will be within a few points of that median. In effect this gives a school a band that it does seriously consider, and within that band it weighs the qualifications of different applicants to choose some and decline others. This goes back to the question I posed earlier, how is such a practice different from what PDaddy is advocating for as a "solution":

That's exactly the problem, they should be able to compare students within bands if they look similar, then find subtle differences that will put certain students ahead of others in terms of "fit" within the class, diversity needs, etc.

The correct answer is there is no difference. That's already what law schools are doing. PDaddy is just on another one of his illogical and misguided rants about law school admissions again.

[r6_philly]

3 x 168 are all within 2 points of 170. 3 x 170 are all within 2 points of 170.

if school take 3 168, median will be 168
if school take 3 172, median will be 172
if school take 2 168, 1 172, median will be 168
if school take 1 168, 2 172, median will be 172

so 6 students all within 2 points of 170 and are all capable of scoring 170 per their own score bands. If a school needs to choose 3, which three do you think the school will favor (probably a lot regardless of softs)?

Maybe score band matters more in the middle 2/3 of the sample where the percentile doesn't vary as much. But for the outliers like T14 applicants, 4 points means 5%tile and makes a huge difference.

For the record, I like the fact that LSAC is admitting that the single score is not as accurate as it looks. However, there rest of the system render it pointless.