Grades in this class arise from a mix of many different ways to judge your work, NOT solely from your performance on exams. Be sure you understand the components! Here is the breakdown.
| Component activity | Total points (scaled if needed) |
Percent |
| Homework | 300 | 23% |
| Midterm exams (100 pts each) | 200 | 15% |
| Final exam | 200 | 15% |
| Lab Reports | 165 | 13% |
| Quizzes | 100 | 8% |
| Readings/ Pre-Lab Questions/Surveys | 100 | 8% |
| Lab/Recitation Participation | 100 | 8% |
| Lecture Participation (TopHat) | 135 | 10% |
| Total | 1300 | 100% |
These divisions are not guaranteed. We may adjust due to unforeseen circumstances that cancel classes or HW - snow, tornadoes, etc.
Component grades
Each of the components will be assigned grading patterns. On exams and quizzes, scores above 75% are considered to be A, above 60% a B, and above 45% a C. On other components there are other grade assignments. Since we expect you to attend all classes, recitations, and labs, an A on participation might require 95%, a B 90%, and a C 80%. Homework grade assignments depend on how the class responds to the particular problems assigned, but in some previous years, an A has required about 85% and a B about 75%. Lab report grades are assigned at the end of the term, since the specific assignment may depend on the way individual TA's grade. Lab grading may be adjusted for differences among the patterns of individual TAs.
Overall grades
The overall grade
boundaries are created by adding the grade boundaries for each
component. Since it's only the total points that matters, if you
are a few points below a grade boundary in one component, you can
make it up by being above the grade boundary in another component
by that number of points.
This means that good scores in some categories can compensate somewhat for weaker scores in others. But there are two things to note:
- Some categories have higher grade boundaries than others. To get an A in an exam requires 75%, but to get an A in lecture participation may require 95%.
- Though one category can compensate for another, being WAY below a boundary in any category can be very hard to overcome. Missing a lot of classes or not doing a lot of homework assignments can seriously hurt your grade.
It's hard to know exactly where the final grade cutoffs
will lie at the end of the semester, since some of the boundaries
depend on how the class performs during the term. But in the
past, the boundary for an A typically falls somewhere between
80-85% of the total points, a B typically falls somewhere between
70--75%, a C in the 60-65% range, and passing, about 50%. NOTE:
THIS MEANS THAT AT THE END OF THIS CLASS A BOUNDARY WILL BE
SET THAT IS LIKELY TO FALL IN THESE RANGES, NOT THAT IF
YOU FALL ANYWHERE IN THOSE RANGES YOU WILL GET THAT GRADE!
What about my grade as shown on ELMS?
We try to have our total points on ELMS (and the resultant percentage that ELMS calculates) give some indication of how you are doing, but it CANNOT be taken as a guarantee of a particular grade. If you are concerned about your grade or cumulative points, please do not hesitate to talk to your lecturer in person.
Grading fairness
In a class of this size and with this much hand-graded work, your instructor can’t handle all the grading himself. He has graduate teaching assistants (TAs) to help. They will not only run the recitations and labs, they will grade each week’s paper homework, one WebAssign problem that needs individual reading and evaluation, exam problems, and your lab reports. From our experience, we know that some TAs are harsh graders and some are easier graders. How can we arrange things so that all students receive equally fair evaluations and some students’ grades do not suffer from getting a particularly severe TA?
We handle this in two ways. For homework and exam grading, each problem graded is graded by a single TA. For example, this means that each TA grades two paper homeworks each term for the entire class. This way, every student gets the same mix of easy and hard graders and on the same problems.
Laboratories are a bit more difficult. We once tried having a single TA grade all the labs in the class for each lab but it did not work out well. Many students did not like not having good communication with the lab grader and felt that the direct feedback from their TA was very important. So instead, we have chosen to create fairness by rescaling each TAs lab grades at the end of the semester. Here’s how it works.
TAs typically give different average grades and different spreads. Different TA’s average grades can differ by as much as 15% (25 points out of 165 total). What we do to make them fair is to calculate the mean and standard deviation of grades for each TA. Then each student’s score is converted to a "z score" — the fraction of a standard deviation your score is away from the mean. If your lab score is s and your TA’s mean and standard deviation are m and σ respectively, your z score is z = (s-m)/σ. Thus a z-score of 0.5 means you are half a standard deviation above your section’s mean and a z-score of -0.5 means you are half a standard deviation below.
A new common score is then created by transforming back to scores from z-scores but choosing a common mean and standard deviation, m0 and σ0, for the entire class. Your renormalized score is then s’ = m0 + z σ0. This makes every section have the same mean and spread of scores. We try to choose m0 and σ0 so that no one loses any points, but that isn’t alway possible. If your TA is an easy grader, your final lab grade may be renormalized down slightly! We try to make this rare and as small as possible. This isn’t ideal, but it is fair, giving all students a lab grade that they have earned while reducing the effect of TAs that grade differently as much as possible.