I'm actually surprised that your students didn't raise the question before the exam and force you to think this issue through up front.

Jared, have you taken an exam recently (or given one) in which this was the system? I thought about it briefly while creating the exam, but dismissed it. In thinking back upon my education, I remember exam administrators (PSAT, SAT, etc.) saying "wrong answers won't be penalized," but I honestly don't remember taking an exam where wrong answers were penalized. I was just suprised that the testing center asked as a matter of course, somehow signalling that the structure was common. How common is it?

I can't think of a discipline where it is worse to be wrong and not know it than law. Well, medicine, maybe.

Knowing when you are unsure and being able to admit it to yourself is a critical skill.

Apply the "wrong penalty."

On the SAT, 1/4 of a point is subtracted for wrong answers. This is the basis of the Princeton Review's "Joe Bloggs" technique (named pre-blogs)--as they say:

"Instead of trying to find the right answer, try to find the wrong answers. By eliminating wrong answers, you greatly improve your chances of getting the question right because even if you can't narrow your choices to a single answer at the end, you will have only two or three to choose from instead of all five. Physically cross out the wrong answer choices in your test booklet, and then guess among whichever answer choices remain.

Only a quarter point is subtracted for every wrong answer, while a full point is added for every right answer. So, if you can eliminate at least one answer choice, guess among the two, three, or four remaining choices."

http://tinyurl.com/qjn24

(I taught Princeton Review before law school--it's not that I remember this from my long, long ago high school days.)

If you don't take off points for wrong answers on essay exames, I don't see why you would do it for a MC exam.

This semester, I had my first law school exam that took off points for wrong answers. I personally think it's a bit silly to do that.

There is another option. I remember one multiple choice exam where the professor gave half credit for the second best answer.

There's a better way to do give a law school exam in a survey course: ask short answer questions rather than multiple choice. Limit students to 100 words or so per answer. That way, you can hit all the topics you want without going to an all-or-nothing system like multiple choice.

At Texas, a well-known BA (and Securities Regulation) professor awards most of his points on True/False questions, deducting for wrong answers. This is much worse than multiple-choice. It was the most arbitrary exam that I've ever taken. Ex. "A University of Chicago professor wrote . . . ." Correct answer: False, it was a Harvard professor.

On the other hand, I've had some very well constructed multiple-choice exams that IMHO did a good job of testing students' comprehension of the material. FWIW, I've seen multiple choice work really well on Evidence and Federal Income Tax, and really poorly on Civ. Pro. I don't know if the difference is the subject or the author, but I suspect the latter.

I disagree with "Former Lawyer" that the wrong answer penalty should be applied. First, if it was not mentioned prior to the exam, it cannot be applied. To do so would be unfair to the students who did not take this into consideration before the exam. Law school exams are just as much about method as arriving at the correct answer. Second, many students will dismiss the answers that are clearly wrong and end up with two answers, one of which is clearly right (in the Professor's viewpoint) and one in which the student could justify by simply applying the answer (perhaps in a context/manner not considered by the Professor).

To answer Jared's question about why students did not raise this question prior to the exam; the answer is simple. Students are generally taking four to five law school exams in less than six or seven days. They rely on the Professor to lay out the intricacies of that particular Professor's exam. Students do not have the time or energy to devote themselves to considering questions that have not crossed their minds since taking the SAT. Most law school professors do not employ the "wrong answer penalty." Most law professors at Marquette University Law School do not employ the "wrong answer penalty." It is not a question that students consider, unless raised by another student, professor, or perhaps, blog.

I also like the suggestion by "ac." Second best answers get a partial point deduction, and would satisfy both goals: to penalize "guessers" and to give partial credit to students who may have arrived at their answer in an unconventional manner (i.e. employing a method that the Professor had not considered).

I also like the suggestion by "PK." But, as a law student, I got three of my best grades on short answer exams, so I may be a bit biased.

"Students are generally taking four to five law school exams in less than six or seven days."

Where? Who? Hyperbole?