# Should tests given after the drop date be made less difficult in order to help the remaining students raise their grades? [closed]

It is that time of the semester again (after the final drop date) and my colleagues and I are once again having the discussion on whether or not the two remaining tests should be made so they are less difficult in order to help the remaining students raise their grades.

It is department policy not to curve grades and the current average of the remaining students is 61%; furthermore, the grade distribution is skewed left and not normal. I argue that as long as we are not compromising the goals of the class, and that the students can demonstrate that they understand the Calculus, then we should make the individual test problems a little less difficult and let them raise their grades (my colleagues disagree). Is this a valid solution or would it be better to weight the tests less, and the homework and quizzes more?

• I think it's very unclear what you are asking. Of course it is "valid" for you and your colleagues to choose whatever difficulty of problems you like. And you can also choose whatever distribution of grades you like. You need to clarify what exactly you want to accomplish and what are the two or three alternatives you are considering. – Chris Cunningham Apr 22 '14 at 12:29
• For example, why would you want them to "raise their grades?" Do you see independent benefit in the students' grades being high? Why? What is the counterargument? – Chris Cunningham Apr 22 '14 at 12:29
• I've voted to close the question because the answers, while good advice in general, are mostly just general education advice. I believe this was caused by the original question not being clear enough to generate definitive answers. The key fact of the situation was later revealed in comments to an answer, but even if that fact was added to the question it would still not make very much sense. Just to be clear, this is obviously only my opinion. – Chris Cunningham Apr 24 '14 at 20:37

The situation you describe is pretty dire. At most US insitutions, an uncurved 61 is a D or F. If your mean is a D or F, and the median is below that (I think that's what you mean by leftward skew?) then most of the class is likely to fail.

I think the first thing you should ask yourself is whether the tests were fairly evaluating the goals you set for your students. If they were too hard for other reasons (not enough time given, numerical problems too difficult, poor directions), then you should fix these problems in future tests, which will have the effect of making your tests less difficult. You will probably also need to adjust your grading scheme to deal with the damaged tests, probably by weighting them less.

If you decide the tests were fair, then your students are not meeting the goals you set for them. I would encourage you to lower your goals. There might be a few cases were it is legitimate for an instructor to fail half a class. (A flight instructor shouldn't certify anyone who can't fly.) But these cases are really rare. Normally, I think that you should adjust the goals of the class to a level that most students can meet, so they are getting some benefit from the class. Lowering your goals for the class will have the effect of making exams more easy, but it will have the more important effect of making it relevant for your students.

I also think you should speak to your chair about whether this policy against curving is absolute. Solving the problems with your test and curricular goals, as described above, are the big tasks, but a curve could also be a helpful bandaid, and your administration doesn't want half your class to fail.

ADDED 4/24/2014 I just wanted to note that I wrote this answer before the OP clarified that this was a course with a common curriculum over several years and a consistent testing history, whose students had recently become unable (for whatever reason) to meet the previous standards. That's an incredibly messy situation, and I don't claim to have advice for it without actually knowing what larger forces are causing this. The only advice I would give strongly in that setting is that the OP should notify his higher ups of what is going on.

• Aside from anything else, if you look at the distribution of grades and then give students an easier test in order to bring the grades up to what you want it to be, then you're probably achieving by the backdoor the thing that the rule against curving grades was designed to prevent -- a "weak" group of students coming out with the same grades as a "strong" group. Aren't you? Of course the cost of not doing this is that a "hard" exam fails more students than an "easy" one the previous year, which is almost certainly worse. – Steve Jessop Apr 22 '14 at 15:17
• I don't know. I have taken courses where instructors gave incredibly hard exams (median below 40%) and then adjusted everything with a huge curve at the end; maybe that is what the policy is intended to prevent. (Whether this is something to be prevented or not is an interesting conversation, but probably too long to belong in this comment thread.) – DES-SupportsMonicaAndTransfolk Apr 22 '14 at 15:20
• ... in an ideal world you adjust the mark scheme on the questions when you realise that this year's questions are next to impossible given the course content, but naturally there's some logistics required to achieve that, and they aren't necessarily in place. – Steve Jessop Apr 22 '14 at 15:22
• In this lies the problem, the test given this semester are basically the same as those given previous semesters. However over the last two years we are seeing that student scores are dropping. – Todd Thomas Apr 22 '14 at 16:48
• @ToddThomas If student scores are dropping, then you are dealing with students with poorer backgrounds, poorer intellects, poorer studying habits, or all of the above. All of these things are bigger issues than you can address on your own, but you could do a lot for your students by giving them lots of practice problems, holding recitation sessions, and doing anything else you can think of to help them prepare for the exam. It seems to me that you can't fairly say that your current students performed equally to your past students, but you should definitely help them if you can! – 2rs2ts Apr 22 '14 at 19:56

Fairness suggests that the grade a student gets shouldn't depend on the semester they happen to take a course in. So the first question I'd ask is what sorts of grades have students typically gotten in this course. If they haven't typically been averaging 61% at this point, then you need to know what's changed: are these actually a weaker crop of students, or were the tests harder than before, or did something else cause this difference.

The most likely explanation (though not the only) is that the tests were a bit harder this year. (This is the most likely explanation because if you're teaching a large number of students, the law of large numbers says they should be pretty similar to last year's---and 100 or so is usually large enough. On the other hand it's very easy to drop the average on an exam by 10 points by making one or two questions just a little harder in ways that cause students to waste time on them.) In that case you should absolutely reconsider the remaining tests, and probably make them easier to compensate. (Actually, you should curve the earlier tests, but apparently that's not allowed.)

I would like to complement Speyer's answer with this: encourage their feedback about the course goals, tests and related activities. This is a critical time to take in account what they have to say as well. As inarticulated their opinions may be, they may carry (part of) a solution to what's going on.

Lowering your goals and/or reevaluating how your tests are measuring them will have greater effect if you manage to see from their perspective as well.

I consider it my duty to certify that whoever passes, has a decent grasp of the subject. Within that constraint, you can reevaluate what you are doing. Are exams too hard? Do students waste much time doing computations? (In my first-year chemistry class, before calculators and such, it was a hard rule that numerical results were worth 5%. Everything was set up so you could leave the numbers and operations written down and do them at the end if you had time.) You could consider giving extra homework (read and critizice a section of the lecture notes, or a recent paper in the area, ...).

I'm against changing the rules (i.e., adding an extra exam covering earlier material or such), but in extreme cases it might be warranted (I had a repeat exam once, when the results of the original were beyond horrible).

I'd seek guidance from more senior colleagues locally.