Policies / Grading
1 Expectations and Policies
We expect that you will attend each course lecture. If you plan to or inadvertently miss a lecture, please let us know. Medical absences will of course be excused, but we want to ensure that we can work together to get you back on track in the course.
2 Extentions
Completing projects on time is an essential requirement for progressing along in the course. Because this course (like most programming courses) is highly incremental, we will make our best effort to ensure that adequate time is given to complete projects and labs. In light of that, we will be hesitant to move deadlines for the whole class.
Instead, each student will receive four lag days throughout the whole semester in which a project or lab may be turned in late, not using more than two days on any one assignment. For exmaple, you may turn in the first two labs two days late (each). But after that, no subsequent work will be accepted late.
3 Grading
Category | Allocation |
Labs / Projects | 50% |
Exams | 40% |
Participation | 10% |
After assigning a percentage grade based on the above table, we will assign grades from 1.0-4.0. We consider a 50% being the lowest percentage that earns a 1.0, and 100% being the highest that gets a 4.0. The spreadsheet formula is:
\dfrac{(percentage-0.5) \cdot 3.3}{0.5} + 0.85
We then round to the closest grade we can report to the registrar, i.e. 3.46 becomes a 3.3, but 3.54 becomes a 3.7. Note that Haverford grading policy will not allow us to round to the nearest tenth, much as we might want.
The combined effect of the above makes for the following rough cutoffs:
Lowest percent to get grade of 1.0 is about... | 50% |
2.0 | 65% |
2.3 | 70% |
2.7 | 75% |
3.0 | 80% |
3.3 | 85% |
3.7 | 90% |
4.0 | 95% |
Parting thoughts from Dave on grading
This typically gives a fairly satisfactory distribution of grades at the top end of the scale, but the lower end of the scale often goes too far into the low grades; in such cases I often either "curve up" the lower end of the overall scale or the lower end of one of the components that is having a dominating effect on the overall grade (i.e., if the variance in exam grades is much greater than that of the lab grades, I may curve the exams before averaging rather than the overall percentages); I apply this curve only to work that was actually submitted, so it doesn’t help those who just didn’t hand in anything; I almost never "curve up" the scores above 85%; I don’t think I’ve ever "curved down" the top scores. I just take this as a sign that we’ve had a good class or perhaps I can add another very challenging lab assignment or something. The average grade of students who didn’t "disappear" from my class (i.e., end up with several 0’s for labs) and then go on to fail the course, is typically in the 3.1-3.3 range, though of course this can vary a lot for small classes.