Explanation of grading system for lecture courses
Grades in lecture courses are adjusted in three ways: (1) increased weight of best 2 exams and decreased weight of worst exam (included before setting curve), (2) extra credit earned from exams (included after setting curve), and (3) reward for improvement from the first exam to the final exam (included after setting curve). Below the logic of these weighting criteria is described.
All students have experienced a bad exam day where a particular exam does not accurately reflect knowledge for the material and the grade is unusually low compared to the other exams. To reduce the impact of the ‘bad day’ grade, we employed a correction which boosts the contribution of the highest exam grades to the final grade and lowers the contribution of the lowest exam grade. For a class with 3 exams, the two best exams would count for a greater percentage of the overall grade and the worst exam score would count for less when calculating the final grade. This correction is intended to mitigate the negative effects of a single bad exam. It does not provide extra credit; the boost in score is a function of actual performance (i.e., the scores on the best performed exams).
Extra credit can hurt students who don’t do it if the curve is set with the extra credit included. Imagine ten students each averaged 88% in the class but five of them earned 3% of extra credit. If the extra credit were included before the curve was set, then the five students who earned extra credit would have 91% and make it less likely that the curve would help the five who did not earn extra credit. To prevent this, the curve in my classes is set before including extra credit (if any is available), so that the extra credit can only boost grades for those who do it and not hurt the grades of those who do not.
A part of any course is ‘learning how to learn.’ A student who shows improvement from the first exam to the last exam is demonstrating some mastery of the process of learning the particular material in the class. Straight weighting of exam grades does not reward this type of improvement. I use an algorithm to reward improvement over time (but not punish for declines or steady state). Like the extra credit, this correction is not included until after setting the curve so it can only help those who improved and does not hurt those who did not. Also, the algorithm takes into account the fact that it is somewhat more difficult to improve an already high score than a lower score. For example, improving a grade of 25% to 30% is presumably much easier than improving a grade of 90% to 95%. The algorithm provides more credit to the latter performance than to the former.
For a more detailed explanation of the algorithm, feel free
to contact me.