W I T N E S S E T H:

Adding up law school exams

I advise a lot of law school faculty members about how to compute scores for exams. It sounds like a simple thing to add up the grades for an exam, but one of the problems that comes up over and over is how to combine different portions of an exam that are on different point scales. For example, a multiple choice portion of the exam might have 40 possible points and an essay portion might have 100 possible points. One common approach is to simply compute the percent of possible points for each part and then combine the two percentages. The problem is that if the two parts are supposed to have set weights (say, 50%-50% or 70%-30%) and they have different variances (as is likely), then that method actually makes one portion count more than it should.

One approach that will work well for those with a basic knowledge of Excel is to do what bar exams typically do, which is to scale one portion of the exam to the other. The idea of scaling is to put the two portions of the exam on the same scale so that they have approximately the same mean and standard deviation. The approach, which the California bar explains in this resource, takes each score to be converted (say the essay portion), subtracts off the mean of the scores for that portion, divides by its standard deviation, multiplies that quantity by the standard deviation for the other portion, and then adds the mean for the other portion.

To make this a bit more concrete, say we want to convert multiple choice scores (which had 40 possible points) into the same scale as the essay portion (which had 100 possible points). Suppose a particular student scored a 32 on the multiple choice portion. If the mean and standard deviation for the multiple choice was 30 and 10, respectively, and the mean and standard deviation for the essay portion was 70 and 20, respectively, then you would calculate [(32-30)/10] x 20 + 70 = 74. Then the student's scaled score of 74 on the MC can be combined with the essay score with whatever weight is desired. The scaled multiple choice scores will have the same mean and standard deviation as the essays, making them directly comparable.

Let me know what you think in the comments!

The GRE movement in U.S. law schools

The latest reporting over at TaxProf indicates that the number of law schools accepting the GRE has grown to 34. The movement toward the GRE has been somewhat slow, in part because the ABA has dragged its feet in allowing wholesale use of the test under Standard 503. As it stands, the ABA position is that each law school needs to conduct its own study of the validity and reliability of the GRE before using the test for admission. 

That position doesn't make any sense.  The GRE's reliability is well established and is not specific to where it's used (law school versus other graduate school). The validity of the GRE for predicting law school performance is clear and has been demonstrated in every law school study publicly available to date. I have personally analyzed data related to the validity of the GRE for law school admission. There is overwhelming evidence that it is as good a predictor as the LSAT for law school performance.

This means that eventually substantially all law schools will eventually accept the GRE, either because the ABA will eventually act and revise standard 503 or because those schools will conduct individualized studies or go ahead without them. Indeed, any law schools that haven't started this process are already behind the curve. That will be a liability in future law student recruiting, as other schools will have an advantage as test takers begin realizing they have alternatives for admission to law school.

Here's why this matters for students. Some students I've spoken with are apprehensive about the idea of the GRE for law school admissions. They believe that accepting both tests constitutes a "watering down" of the admission requirements that will hurt the value of their degrees. The fact is that the opposite is true. Schools that lag behind in migrating toward the GRE will face a competitive disadvantage in law student recruiting. This means those schools will recruit the "leftover" applicants that similarly selective schools didn't admit, or they will need to pay more in scholarships to get the top students. Once a law school slides relative to its competitors in admissions, it can cause a self-propelling process that leads to even fewer strong applications the next cycle, and so forth. That is what will diminish the value of your degree.

In short, law students should support the idea of their schools moving toward the GRE. It is not a "watered down" test; indeed, the opposite claim could be made as the GRE requires some quantitative literacy. Schools leading the charge on the GRE will have a competitive advantage in admissions in the coming years. Only a handful of students are being admitted on a "GRE-only" basis right now, so there is still some time for law schools to move toward the new test. So if your law school is making this change, that is a good thing for the value of your degree.

A different take on the transfer market in law schools

The law school transfer "market" is a controversial topic. The rise in transfers arguably contributed to undermining the soundness of many schools during the financial crisis and its aftermath. Brian Leiter sees the transfer market as an artifact of the US News system, which may well be true. Adding to this, Jerry Organ is out with analysis of new new ABA data on law school transfers showing that the rate of transfer has declined somewhat. Most academics see that decline as a good thing.

There is another side to the transfer market that often goes unseen, however. Indeed, there is an argument that a healthy and active transfer market could contribute to breaking down artificial barriers to law schools (primarily LSAT). The key fact that some law professors don't know is that the traditional admissions criteria--primarily LSAT and undergraduate GPA--are quite noisy predictors of law school success. If you look at a scatterplot of first-year law school GPA (what the LSAT is designed to predict) versus LSAT score for any given law school, you will find the dots all over the map. There is a clear relationship between LSAT and law school GPA, but it is far, far less deterministic than many law students and even law professors believe. (Part of this is due to the statistical restriction of range of using data within a single law school, which is why such internal analyses are often flawed, but that's for another day).

The grades earned in first-year law school, however, are in another category. Although they too contain noise, they are far better predictors of bar exam passage (and upper division grades) than is LSAT. Indeed, in models I've run, even one semester of law school grades dominates LSAT to the degree that LSAT is no longer predictive of bar passage at all. Although it's true that law professors don't typically conduct rigorous assessments of the validity of law school grades (in the test theory sense), those grades are viewed as the single most important piece of information in large law firm hiring.

So what's the point? If larger numbers of students with high law school GPAs consider transfers, and if schools judiciously accept more transfer students based on their law school grades, a sorting process will occur that is better than the existing LSAT/UGPA criteria. There are many students who cannot achieve high scores on the LSAT but who are capable of strong performance in law school. A system that encourages a robust transfer market gives those students the opportunity to rise through the law school pedigree hierarchy. Lower-ranked schools will need to dip deeper into the admissions predictors to secure a first-year class, and will lose many of their top students, diminishing their bar passage numbers. But that can be counteracted with the "other end" of transfers, academic attrition. 

In short, there is an argument that increased rates of transfers along with more aggressive admissions and academic attrition at the lower end would increase opportunity for those who are truly capable of high law school performance, not merely those who excel on a single LSAT exam.

 

Law professor blogging and practice-ready scholarship, a six year assessment

The announced demise of the Concurring Opinions Blog, once a mainstay of law professor blogs, has got me (and many others) thinking about the value (or not) of law professor blogging. Some believe the era of the law professor blog is over, cannibalized at the one end by mircoblogging on Twitter and on the other end by the rise of shorter online pieces in law reviews. 

I started this blog in 2012 with the idea of producing "practice ready scholarship." The idea was that a law professor blog could be fully committed to the creation of new knowledge and the communication of that knowledge to those who could use it in a form they could use it. Here is an excerpt from what I wrote then in my very first post.

I have started this blog based on the conviction that there is a need for law professors, especially in corporate law, corporate finance, and M&A, to devote more resources toward producing practice-ready scholarship.

By “practice-ready scholarship,” I mean scholarly research that has the potential to influence the practical aspects of transactional law practice. Both parts of the term are important. I plan to focus on research that is “practice ready” in the sense that it can have a tangible impact on legal practice at the concrete, rather than merely an abstract or theoretical level. On the other hand I plan to focus on work that is “scholarship” in the sense that it aims to generate new knowledge rather than merely summarizing doctrine or recapitulating existing information. Ultimately, the hope is that this blog will be a small contribution toward bridging the gulf between law schools and lawyers in transactional practices.

I see quantitative academic scholarship as uniquely positioned to provide valuable information to the bench and bar that is otherwise unavailable. I believe that insights from empirical analysis of deals are desperately needed by practitioners. I also believe that law professors have a comparative advantage in performing this empirical analysis because practitioners lack the incentives, time, and inclination to collect, analyze, and report on data for a public audience. Equally important is the existence of a forum to translate these theoretical and empirical insights into terms that can be directly applied by practitioners. The typical transactional lawyer often needs quick answers and cannot persuade clients to pay them to read 60-page law review articles to indulge the lawyer’s intellectual curiosity.

I started this blog with the hope of providing a forum to bridge the gap between scholarly inquiry and practice—especially transactional practice in corporate law, corporate finance, and M&A, where I practiced myself. This blog is designed to communicate useful results from my own original research in progress, as well as commenting on insights from other academics’ research that might otherwise be overlooked by busy practitioners. The goal, overall, is to bring data and theory to practice, without sacrificing either scholarly rigor or practical usefulness.

One of the problems with group blogs is that they often eventually devolve into a single active author, reducing the volume of posts below the point where they keep readers coming back. Yes, readers will follow links from aggregator blogs to interesting material (at least when aggregator blogs don't Control-C Control-V the entire content--the business model for some blogs). But I don't want to recycle (steal?) other people's content; I want to create new content. And readers won't keep visiting a blog if they don't get new content at least 50% of the days they visit the site. WITNESSETH has always been a single author blog, so there have been times when the volume of posts has dropped too low to sustain readership.

I have reflected on my own blogging experience and determined that there is a continued need for what I provide here at WITNESSETH. As a result, I am committing myself in an early new year's resolution to at least one post every day to keep people coming back. If I were rewriting my original "launch" post today, I would have included my commentary on law schools, the law profession, and legal technology along with transactional law and M&A as my points of focus. That's what this blog will provide starting now and through 2019. I would like to try to more effectively realize my original goal of creating and communicating practice-ready scholarship of my own and publicizing that of others. I hope this will translate into increased reader engagement in 2019 and beyond.

Unearthing the Household International poison pill

In the course of a project I'm working on related to "rights agreements" (poison pills), I discovered that the ThomsonOne service has some limited availability of paper filings from the pre-EDGAR era. I searched for the earliest poison pills and was unable to find them for Lenox and El Paso (believed to be the first poison pills). However, I found the one from Household International that was the subject of the landmark Supreme Court case validating the poison pill (Moran v. Household International).

I uploaded here a  copy of the 8-A filing with the pill document attached in case readers are interested in delving into this little piece of M&A history. I inserted below the first page of the pill from the document. Many of the terms are almost verbatim identical to those used today, with the exception that the Household Pill did not contain the now-standard "flip in" feature.