There will be weekly problem sets, which will contain both quantitative problems and essay-type questions. Policies for lateness and collaboration on problem sets are described in this short essay on the theory and practice of problem sets. There will be two in-class tests (on February 5 and March 26), and a final exam. All of these exams will be open book, but electronic aids, including calculators, are not allowed. The discussion sections are required and will form a crucial part of the course: part of each section will be devoted to understanding the current problem set. There will be an optional 6-12 page paper.

There are a total of 600 points in the course, which will be determined as follows: 30% problem sets (9 problem sets of 20 points each for a total of 180 points); 30% in-class exams (120 points for the stronger test, 60 points for the weaker - we will automatically figure out which test is better for each student, and multiply accordingly); 30% final exam (180 points); 10% other activities (mostly sections and in-class questions). The optional paper will be worth 15% of the grade (90 points), and will reduce the weight of the weakest major portion of your grade from 30% to 15% (from 180 to 90 points). However, no problems sets or tests can be dropped altogether.

Students sometimes ask whether this course is graded "on a curve". This depends on what you mean by grading on a curve. I do not use the high-school grading scheme of 90-100%=A, 80-90%=B etc. Such a scheme essentially means that all good students should be able to answer all questions correctly, and that prevents asking interesting questions. So compared to traditional high-school grades, one might say the course has a curve. But I do not have fixed fractions in mind for how many students will get As, Bs etc. So students are not competing against each other; it does not hurt your grade to help someone else. Everyone who deserves an A will get one, regardless of how other students perform. In that sense, the course is not graded on a bell curve, or any other kind of curve.

The exact relationship between the numerical points given on the tests and problem sets and the letter grades varies from year to year - this is necessary because it can be hard to know in advance whether the class will find a particular problem to be hard, or not. However, in the past few years, the grading scheme has come out close to the following:

90-100%: A

80-90%: B+/A-

70-80%: B

60-70%: C+/B-

50-60%: C-/C

less than 50%: D or F

This scheme does not apply to any given exercise - some are harder than others. Also, I do not guarantee that this year will work out this way, so please don't overinterpret the first numerical scores you get. As the semester goes on, I'll give occasional updates on how to interpret your scores more precisely so that you know where you stand.