Top Law Schools

hoos89 wrote:"You can't prove me wrong" was in response to someone pointing out that I can't prove that those things are correlated

LOL that never happened. You're a real piece of work.

Bob_loblaw34 wrote:
First, you committed the appeal to ignorance fallacy at least twice that I have seen (though perhaps regarding the same issue). And questioning whether they are correlated is a subset of questioning plausibility and legitimacy. In other words, if the correlation you blindly assert and assume to exist does, then you explanation seems much more plausible and legitimate. If the correlation actually does not exist, then your argument is pretty worthless (i.e., implausible and illegitimate) isn't it? Again I just take issue with someone retorting with a clear logical fallacy. Actually trying to help you out.

I think you're misinterpreting what I meant. I'm not saying that my assertion is correct because it hasn't been disproven. My point isn't to prove that the correlations I've put forward are correct, but to call into question the assertion that because grades are correlated with length and ratings by English majors, law school exams are illegitimate. This isn't exactly a ground breaking method to question correlations: because they aren't experimental, potential confounding factors should be considered when drawing inferences from them.

hoos89 wrote:

Bob_loblaw34 wrote:
First, you committed the appeal to ignorance fallacy at least twice that I have seen (though perhaps regarding the same issue). And questioning whether they are correlated is a subset of questioning plausibility and legitimacy. In other words, if the correlation you blindly assert and assume to exist does, then you explanation seems much more plausible and legitimate. If the correlation actually does not exist, then your argument is pretty worthless (i.e., implausible and illegitimate) isn't it? Again I just take issue with someone retorting with a clear logical fallacy. Actually trying to help you out.

I think you're misinterpreting what I meant. I'm not saying that my assertion is correct because it hasn't been disproven. My point isn't to prove that the correlations I've put forward are correct, but to call into question the assertion that because grades are correlated with length and ratings by English majors, law school exams are illegitimate. This isn't exactly a ground breaking method to question correlations: because they aren't experimental, potential confounding factors should be considered when drawing inferences from them.

Right, and the way your are attempting to call the assertion into question is by offering plausible, legitimate explanations for the correlation. It is these explanations which you assert, and then say haven't been proven wrong and therefore are deserving of some sort of credibility. If you want to offer an explanation to a phenomenon, anecdotal evidence is almost as unimpressive as saying "well you can't prove my explanation wrong!" In other words, you are trying to find a legitimate justification for the correlation between typing speed and exam performance (or whatever the actual correlation Cougar cited was) such that the mere presence of this correlation does not destroy the validity of Law School exams. However, anecdotal evidence is basically meaningless in isolation, and responding by saying "you cant prove my explanation is wrong!" is just as bad. So your explanation remains unjustified. Or at least thats where my head is at.

Law school exams are flawed in the sense that they fail to award points for the penetrative comments I make during class. I really don't understand why. They contribute to everyone's learning experience and demonstrate my thorough understanding of the material.

Bob_loblaw34 wrote:

hoos89 wrote:

Bob_loblaw34 wrote:
First, you committed the appeal to ignorance fallacy at least twice that I have seen (though perhaps regarding the same issue). And questioning whether they are correlated is a subset of questioning plausibility and legitimacy. In other words, if the correlation you blindly assert and assume to exist does, then you explanation seems much more plausible and legitimate. If the correlation actually does not exist, then your argument is pretty worthless (i.e., implausible and illegitimate) isn't it? Again I just take issue with someone retorting with a clear logical fallacy. Actually trying to help you out.

I think you're misinterpreting what I meant. I'm not saying that my assertion is correct because it hasn't been disproven. My point isn't to prove that the correlations I've put forward are correct, but to call into question the assertion that because grades are correlated with length and ratings by English majors, law school exams are illegitimate. This isn't exactly a ground breaking method to question correlations: because they aren't experimental, potential confounding factors should be considered when drawing inferences from them.

Right, and the way your are attempting to call the assertion into question is by offering plausible, legitimate explanations for the correlation. It is these explanations which you assert, and then say haven't been proven wrong and therefore are deserving of some sort of credibility. If you want to offer an explanation to a phenomenon, anecdotal evidence is almost as unimpressive as saying "well you can't prove my explanation wrong!" In other words, you are trying to find a legitimate justification for the correlation between typing speed and exam performance (or whatever the actual correlation Cougar cited was) such that the mere presence of this correlation does not destroy the validity of Law School exams. However, anecdotal evidence is basically meaningless in isolation, and responding by saying "you cant prove my explanation is wrong!" is just as bad. So your explanation remains unjustified. Or at least thats where my head is at.

None of that is really anecdotal evidence though. It's true that I have previously presented anecdotal evidence that I did well without writing a ton, but me saying that it seems plausible that length would be correlated with other factors that seem legit to me (e.g. spotting more issues, performing more thorough analysis) is not anecdotal. It's not evidence at all so that's just kind of a weird label to put on it. And just because I can't prove my alternate explanation doesn't make it reasonable to take the similarly unproven original explanation (i.e. "if you write a lot in a way that appeals to an English major you'll do well") at face value. My whole purpose was to call JCougar's assertion into question, not to prove my own. I can't prove that his explanation is wrong, and he can't prove that mine is. Both are at least plausible. Therefore we can't take from the mere existence of the correlations he presented that law school exams are invalid.

Some people are just better than others at law school exams. No one is saying those people will be better lawyers.

hoos89 wrote:None of that is really anecdotal evidence though. It's true that I have previously presented anecdotal evidence that I did well without writing a ton, but me saying that it seems plausible that length would be correlated with other factors that seem legit to me (e.g. spotting more issues, performing more thorough analysis) is not anecdotal. It's not evidence at all so that's just kind of a weird label to put on it. And just because I can't prove my alternate explanation doesn't make it reasonable to take the similarly unproven original explanation (i.e. "if you write a lot in a way that appeals to an English major you'll do well") at face value. My whole purpose was to call JCougar's assertion into question, not to prove my own. I can't prove that his explanation is wrong, and he can't prove that mine is. Both are at least plausible. Therefore we can't take from the mere existence of the correlations he presented that law school exams are invalid.

...except his point is backed by a study that controlled for your assumptions by administering tests in different formats. If any professors share your "plausible explanation" (read: rampant speculation), it could possibly bias how they grade tests, reinforcing the very effect you're attempting to explain away. I doubt any are that dim though, they're poor assumptions. It's my opinion that students who don't know the law are more likely to throw a wall of text at it hoping that something in there earns them points.

whatwhatinthebutt wrote:

hoos89 wrote:None of that is really anecdotal evidence though. It's true that I have previously presented anecdotal evidence that I did well without writing a ton, but me saying that it seems plausible that length would be correlated with other factors that seem legit to me (e.g. spotting more issues, performing more thorough analysis) is not anecdotal. It's not evidence at all so that's just kind of a weird label to put on it. And just because I can't prove my alternate explanation doesn't make it reasonable to take the similarly unproven original explanation (i.e. "if you write a lot in a way that appeals to an English major you'll do well") at face value. My whole purpose was to call JCougar's assertion into question, not to prove my own. I can't prove that his explanation is wrong, and he can't prove that mine is. Both are at least plausible. Therefore we can't take from the mere existence of the correlations he presented that law school exams are invalid.

...except his point is backed by a study that controlled for your assumptions by administering tests in different formats. If any professors share your "plausible explanation" (read: rampant speculation), it could possibly bias how they grade tests, reinforcing the very effect you're attempting to explain away. I doubt any are that dim though, they're poor assumptions. It's my opinion that students who don't know the law are more likely to throw a wall of text at it hoping that something in there earns them points.

I can't find the article for free online, and I don't see anything about this in the preview of the article. That said, consider me skeptical that administering tests in different formats controls for that. Also you're taking my statements to the extreme. I'm not saying that wall of text = good exam. I'm saying that longer exams are likely to hit more issues and do more analysis, on average, and that this relationship should be more pronounced at the lower end of the spectrum (i.e. particularly short exams are just not going to be able to cover everything). That there exist counter examples (word vomit) doesn't mean that correlation isn't likely to exist at all.

hoos89 wrote: I can't find the article for free online, and I don't see anything about this in the preview of the article. That said, consider me skeptical that administering tests in different formats controls for that. Also you're taking my statements to the extreme. I'm not saying that wall of text = good exam. I'm saying that longer exams are likely to hit more issues and do more analysis, on average, and that this relationship should be more pronounced at the lower end of the spectrum (i.e. particularly short exams are just not going to be able to cover everything). That there exist counter examples (word vomit) doesn't mean that correlation isn't likely to exist at all.

Nobody's taking your statements to the extreme. You have a habit of altering your statements, misrepresenting what others have said, and now are claiming you haven't even read what you originally responded to. You're right to be skeptical as to whether the authors properly control for what they say they do. There are assumptions built into every model and they have to be justified. I'm not entirely convinced either, but your assertions that it's more likely your way until proven wrong are ridiculous. You've got a poorly thought out hypothesis that you deem to be the most likely representation of reality, I'm just pointing out that it's speculation and a possible source of bias if any professors subscribe to your view.

whatwhatinthebutt wrote:

hoos89 wrote: I can't find the article for free online, and I don't see anything about this in the preview of the article. That said, consider me skeptical that administering tests in different formats controls for that. Also you're taking my statements to the extreme. I'm not saying that wall of text = good exam. I'm saying that longer exams are likely to hit more issues and do more analysis, on average, and that this relationship should be more pronounced at the lower end of the spectrum (i.e. particularly short exams are just not going to be able to cover everything). That there exist counter examples (word vomit) doesn't mean that correlation isn't likely to exist at all.

Nobody's taking your statements to the extreme. You have a habit of altering your statements, misrepresenting what others have said, and now are claiming you haven't even read what you originally responded to. You're right to be skeptical as to whether the authors properly control for what they say they do. There are assumptions built into every model and they have to be justified. I'm not entirely convinced either, but your assertions that it's more likely your way until proven wrong are ridiculous. You've got a poorly thought out hypothesis that you deem to be the most likely representation of reality, I'm just pointing out that it's speculation and a possible source of bias if any professors subscribe to your view.

I've read what I had access to (the preview in the google book), but I'm not going to pay for access to that article to win an internet argument. Speaking of misrepresenting what others have said: I haven't said that my explanation is the most likely explanation. Nor have I said that it's more likely my way until I'm proven wrong. I have said that we should be careful about inferring too much from correlations with so many potential confounding factors present. A pretty basic premise of statistics is that correlation does not equal causation, so yes the presence of correlation is often SOME evidence of a point you may want to make, a major aspect of critically analyzing any correlation is to consider potential confounding factors. Assigning any meaning to a correlation usually involves some level of speculation. It is just as speculative to say that length is correlated because word vomit gets rewarded as it is to say that it's correlated because people who spot more issues are going to tend to have longer essays on average.

hoos89 wrote:I'm saying that longer exams are likely to hit more issues and do more analysis, on average

hoos89 wrote: Speaking of misrepresenting what others have said: I haven't said that my explanation is the most likely explanation. Nor have I said that it's more likely my way until I'm proven wrong. I have said that we should be careful about inferring too much from correlations with so many potential confounding factors present. A pretty basic premise of statistics is that correlation does not equal causation, so yes the presence of correlation is often SOME evidence of a point you may want to make, a major aspect of critically analyzing any correlation is to consider potential confounding factors. Assigning any meaning to a correlation usually involves some level of speculation.

You've said many times how likely it is, as well as "prove me wrong" more than once. A pretty basic concept in statistics is multicollinearity, yet you're so unfamiliar with the concept you keep saying "potential confounding factors" when referring to correlation between independent variables. Tell me more, you've clearly got much to teach about statistical analysis and the studies you haven't read.

Likely != most likely. I also never flatly said "prove me wrong," but that you couldn't prove me wrong. I meant that as a statement of fact, not as a taunt (and it was accompanied by the fact that I also can't prove you wrong). Also collinearity and confounding factors are closely related but not entirely overlapping. What I'm talking about is actually more accurately described as a confounding factor. Collinearity is the extent to which multiple explanatory variables are related to each other in a multiple regression (and to my knowledge, the factors I'm talking about weren't actually in the regression). I'm not referring to a correlation between independent variables, but between dependent variables and variables that were not (to my knowledge) in the regression.

Here comes the "I don't use the proper terminology because I don't want to confuse laypeople" excuse.

whatwhatinthebutt wrote:Here comes the "I don't use the proper terminology because I don't want to confuse laypeople" excuse.

Confounding is the proper terminology here not collinearity.

I got median grades at WUSTL, and I STRONGLY disagree that law school exams are entirely random or just a function of typing speed. The difference between a 4.3 and a 4.2 on an exam may come down to typing speed on very specific exams(torts, civ pro maybe?), but the difference in the QUALITY of answers between the 4.0s and a 3.5s on an almost any law school exam is going to be substantial. In 3 years, with grades ranging from the A+s to failing(shoutout to Examsoft ), I could have predicted my exam grade after taking it within .3-.4 on every exam except three. All three of those were extremely easy, and obviously it was difficult to establish any true differentiation.

That's not to say law school exams don't necessary test silly things. Many of the open book exams essentially test how well you make an outline and practice using it, rather than any sort of understanding of the material. One particular professor made an exam where 20% of it was effectively testing whether you attended specific classes. Still, the exam grade is based off the quality of your actual exam answers, and they do vary substantially.

hoos89 wrote:What I'm talking about is actually more accurately described as a confounding factor.

Get the fuck out of here if you think "a good grasp of the nuances of the law", "sound analysis", and "length" aren't relevant to your grade. None of them are extraneous.

whatwhatinthebutt wrote:

hoos89 wrote:What I'm talking about is actually more accurately described as a confounding factor.

Get the fuck out of here if you think "a good grasp of the nuances of the law", "sound analysis", and "length" aren't relevant to your grade. None of them are extraneous.

None of them besides length were actual independent variables, so they are by definition extraneous.

E: That they are relevant to your grade is what makes them confounding. Extraneous doesn't mean irrelevant.

hoos89 wrote:

whatwhatinthebutt wrote:

hoos89 wrote:What I'm talking about is actually more accurately described as a confounding factor.

Get the fuck out of here if you think "a good grasp of the nuances of the law", "sound analysis", and "length" aren't relevant to your grade. None of them are extraneous.

No, but none of them besides length were actual independent variables, so they are by definition extraneous.

Extraneous doesn't mean not included in the regression, if they're not included but relevant they result in omitted variable bias. You have no clue what you're talking about. Besides, they're controlled in the first posted Crane study. I'm done with you.

How would you all feel about having all MC exams in law school?

Hikikomorist wrote:How would you all feel about having all MC exams in law school?

I didn't mind the ones that I had. Some subjects are better suited to it. I am good at MC though, so I may be biased.

I think the mixed MC/Essay tests are a good comprimise though.

I do fucking hate policy questions though.

sublime wrote:

Hikikomorist wrote:How would you all feel about having all MC exams in law school?

I didn't mind the ones that I had. Some subjects are better suited to it. I am good at MC though, so I may be biased.

I think the mixed MC/Essay tests are a good comprimise though.

I do fucking hate policy questions though.

Which ones do you think are better suited? The only class where I really couldn't see the whole thing being MC is Con Law.

Hikikomorist wrote:

sublime wrote:

Hikikomorist wrote:How would you all feel about having all MC exams in law school?

I didn't mind the ones that I had. Some subjects are better suited to it. I am good at MC though, so I may be biased.

I think the mixed MC/Essay tests are a good comprimise though.

I do fucking hate policy questions though.

Which ones do you think are better suited? The only class where I really couldn't see the whole thing being MC is Con Law.

The more statutory a subject is, the better. So UCC 2,3,4,9, Evidence, Bankruptcy, Tax.

For case based stuff, it is really hard to write good MC questions that are both not simple facts, and not too ambiguous. A question like "which of the following is not an element of battery" is worthless.


And you can do it, as you can tell by the bar exam, however the bar examiners obviously spend more time coming up with the tests than your average professor ever will. I think some essay writing is also good, for as long as it is required for the bar.

sublime wrote:

Hikikomorist wrote:

sublime wrote:

Hikikomorist wrote:How would you all feel about having all MC exams in law school?

I didn't mind the ones that I had. Some subjects are better suited to it. I am good at MC though, so I may be biased.

I think the mixed MC/Essay tests are a good comprimise though.

I do fucking hate policy questions though.

Which ones do you think are better suited? The only class where I really couldn't see the whole thing being MC is Con Law.

The more statutory a subject is, the better. So UCC 2,3,4,9, Evidence, Bankruptcy, Tax.

For case based stuff, it is really hard to write good MC questions that are both not simple facts, and not too ambiguous. A question like "which of the following is not an element of battery" is worthless.


And you can do it, as you can tell by the bar exam, however the bar examiners obviously spend more time coming up with the tests than your average professor ever will. I think some essay writing is also good, for as long as it is required for the bar.

So you would agree then that a Torts MC test is utterly worthless?

sublime wrote:

Hikikomorist wrote:How would you all feel about having all MC exams in law school?

I didn't mind the ones that I had. Some subjects are better suited to it. I am good at MC though, so I may be biased.

I think the mixed MC/Essay tests are a good comprimise though.

I do fucking hate policy questions though.

The bar exam itself manages to test all 1L subjects perfectly fine and well via multiple choice. There's benefits and drawbacks to MC. The main benefit is that at least it's objective. But that's one main drawback, too--not every question in law has an objectively right answer.

Testing your writing is important, because writing is a critical skill that basically every lawyer uses. Of course, they don't use it they way law exams make you use it (typing speed contest).

I have a lot of background in the testing industry, and tests where the main constraint is time (and not difficulty) are notoriously bad at assessing talent. Those timed math tests they sometimes give you in third grade or whatever where you have to see how many multiplication problems you can do in three minutes have zero validity, and actually end up discouraging smart students. Most educators no longer use them anymore, because studies on their effectiveness have told educators not to.

The best way to administer a challenging test is not to shorten the time constraints, but to actually make it harder. As in the complexity of the logic/reasoning skills needed to solve it. Law school exams, even in essay format, could be a lot better, and less of a crapshoot, if they doubled the time needed to take them, but implemented a word limit that nobody would have a problem reaching as long as they studied and kept up in class. Something like 6 hours for 3,000 words, but make the fact patterns longer and the legal issues involved twice as complex.

It's easier just to copy n' paste the same exams from years past, though, and change a few names around here and there. Plus legal pedagogy and the actual search for truth have been strangers from the beginning. I don't expect them to get together anytime soon. Legal education is mostly a money-making racket. Teaching you anything of value or assessing your abilities honestly have never really been priorities, and they probably never will be.

acr wrote: So you would agree then that a Torts MC test is utterly worthless?

That torts MC section saved my ass in that class. I fucked up that essay so hard