# Why average Multiple Scores

##### Why average Multiple Scores
« on: April 15, 2005, 06:18:45 PM »
Can anyone explain to me the rationale behind averaging multiple LSAT scores. To me it might make sense if the scores are within a single standard deviation. As the scores difference increases the likelihood that the average is a good indicator of where you stand decreases (the variance increases).

Also given a large enough difference the stastically sound conclusion would be to accept one score and disregard the other entirely. For example suppose you move up 9 points. This is three standard deviations meaning that the likelihood that the increase is due to chance is less than 1%. So as a matter of policy including the lower score as part of consideration is likely to be a poor choice 99% of the time.

So why do schools do it? I have a few theories:

1) In reporting numbers to US-News, LSAC they have to average scores. Not averaging scores means they decrease their numbers.

-This leaves open the question of why LSAC/US-news want averages.

2) From a low score it is reasonable to infer that this is a person who will crack under pressure/sickness/extenuating circumstance.

According to this explaination law schools value Badasses who never have a bad day above people who are smarter/more able but might have an off day.

-It should be noted that GPA would either back this up. If you panick in high pressure tests then you probably panic in final exams also and there should be some eveidence in the GPA. All the same, it seems foolish to me.

3) Additional information. Law Schools have so many people already, if they can just throw out some applications and justify it they might as well.

This theory says that law schools are forced to make often arbitrary discriminations. The argument would be that averaging is arbitrary, but at least it is based on something tangible (a number) as opposed to the other arbitrary decisions admissions officers make.

Response: this would seem to be even more arbitrary.  Subtracting points from a persons LSAT score when there is a 99% chance that it will be incorrect to do so would be something like randomly taking 99 applications and subtracting some number of points from them just to lighten the burden of applications.  Ultimately the goal of admissions is to make a non-arbitrary decision -- this means trying to discern relavent considerations from a record even with the possibility of error.  If it is really optimal to average big LSAT differences on the grounds of arbitrariness alone then this is a tacit acknowledgement that the admissions process is worse than chance at judging intangibles.  I don't believe this to actually be the case.

Any thoughts?

I would especially like to hear from anyone more official than the average college student.

#### dave303

• 531
• La plus belle pour aller danser
##### Re: Why average Multiple Scores
« Reply #1 on: April 15, 2005, 06:26:01 PM »
Because if they let only the highest socre count people would keep taking it over and over till they got their best score.

##### Re: Why average Multiple Scores
« Reply #2 on: April 15, 2005, 06:50:05 PM »
Reply to: people will take the LSAT again and again and again.

1) So What? -- if really everyone does this then the only effect would be to weed out flukish scores.
1.5) Most colleges take higher SATs and there isn't really a huge epidemic of taking the SATs as many times as you want.
2) LSAC limits people to taking the test three times in a (I think) 2 year period.
3) No, because I am arguing that scores should be averaged if the difference is small (this means people risk going down) and it doesn't really have any significant impact unless you think your first score was a fluke
4) OK, even if your right that people take it again and again, that is a forward looking piece of analysis about having the policy very public and well known, but in judging any individual applicant who has taken it twice it wouldn't really be wrong to consider the score increase.  In fact it would be acknowledging that really the picture of the applicant is not the right one, but the applicant's score increase is disregarded to obtain whatever is good about people not taking it twice.
5) This doesn't justify averaging -- for example you could pay a penalty for the the second time you take the LSAT. The higher LSAT score could auotomatically be reduced by 2 points -- this would discourage people from taking it twice unless they thought they could go up more.  Consider the policy in the extreme case where you average a 140 with a 170.  Do you honestly think tht

#### dave303

• 531
• La plus belle pour aller danser
##### Re: Why average Multiple Scores
« Reply #3 on: April 15, 2005, 06:56:24 PM »
Many many people take the SAT multiple times. One of the purposes of averaging the LSAT is to artificialy weed out the applicant pool. The best universities get probably something like a hundred applications for every single spot. Anything they can do to seperate one applicant from another they will.

Med schools do this far worse than law schools. Averaging LSAT scores will not be going away anytime in the near future.

##### Re: Why average Multiple Scores
« Reply #4 on: April 15, 2005, 07:02:35 PM »
There are many distinctions that Law Schools might draw to weed down the applicant pool.  I don't think that it is valid to say "anything to weed down the ziae of the pool."

Some things are better than others for weeding down the pool.  This, especially in the case of a very large score difference, is a particularly arbitrary thing to use.  It would be less arbritrary to weed down the pool with any of a number of other factors, so why use this one?

#### Julie Fern

• 25797
• hillary clinton say "boo!"
##### Re: Why average Multiple Scores
« Reply #5 on: April 15, 2005, 07:11:50 PM »

#### dave303

• 531
• La plus belle pour aller danser
##### Re: Why average Multiple Scores
« Reply #6 on: April 15, 2005, 07:14:06 PM »

#### Julie Fern

• 25797
• hillary clinton say "boo!"
##### Re: Why average Multiple Scores
« Reply #7 on: April 15, 2005, 07:19:27 PM »
good one!

#### Fernando

##### Re: Why average Multiple Scores
« Reply #8 on: April 15, 2005, 07:31:03 PM »
There are many distinctions that Law Schools might draw to weed down the applicant pool.  I don't think that it is valid to say "anything to weed down the ziae of the pool."

Some things are better than others for weeding down the pool.  This, especially in the case of a very large score difference, is a particularly arbitrary thing to use.  It would be less arbritrary to weed down the pool with any of a number of other factors, so why use this one?

Strictly speaking, that makes sense.

#### Troy McClure

##### Re: Why average Multiple Scores
« Reply #9 on: April 17, 2005, 09:08:44 AM »
They average scores because the point of the test is to see how well people do testing in a timed, stressful environment much like 1L exams.  If you bomb for whatever reason in 1L, you don't get do-overs.  It's all about how you perform on that one day.  The LSAT schedule is posted months in advance.

We all have all the time and opportunity to attend classes, read books, and take every LSAT already given.  If the person wanders in unprepared, why should they get to retake it?  That's their loss.  I didn't really prepare that much, and I didn't score as well as I could had I put in the months of prep many people here do.

It's very obvious to all how important the test is.  By averaging, they're taking into account your performance on your best and worst days much like any normal school exam would.