Grading Issue

| | Comments (11)

I'm grading an exam that has some unusual directions. 10% of the exam is multiple choice, and I grade that unusually also, but the issue I'm facing has to do with the other 90%. I give my students freedom to choose among a number of questions to fill out that 90%. There are 10-point, 20-point, 30-point, and 40-point questions. They can skip one whole section, but they need to do at least one question from each of the other three sections. The easiest way to do it is to do one 20-point, one 30-point, and one 40-point, but there are lots of permutations to get to 90 points while doing something from each of three sections.

I've had students answer 100 points of questions. If one of their questions is a 10-pointer, then I just don't grade that one. If not, then I'll count a 20-pointer at half value (i.e. if they would have gotten an 18 out of 20, then it's 9 out of 10). I've also had students just not answer enough questions, in which case they end up with a lower score than they might have had. But for the first time (this is the third time I've used this model -- really the fourth and fifth times, if you count each section as a separate time) I've got a problem with a student answering the wrong point total where I can't think of a good solution.

This student answered three questions besides the multiple choice. She did two 40-point questions and a 10-point question. That does lead to the right point total, but it wasn't from three sections. The questions were taken from two of the four sections instead of the required three. So how do I handle this? I could simply grade the questions she did and ignore it, but that sends a message that the exam rules aren't really important. I do think some point loss is required. It's not clear how many points to penalize her, though. It isn't as if there's a question amount over the total that she's done, since she's done the right number of points. But I could claim that two 40-pointers is impossible within the rules I stipulated and then count the 40-pointer as a 30-pointer, something akin to what I do if they do too many points worth total without any 10-pointers to ignore. That 40-pointer should have been two 20-pointers or a 30-pointer and a 10-pointer. Those are the only ways for her to have followed the directions that get as close as possible to the questions she actually answered. The more generous option of those two,, then, would count one of the 40-pointers (the second one, to be precise, since that's the one that took it over the top) as a 30-pointer and insist that she should have done a 30-pointer instead and then a missing 10-pointer.

The result would be penalizing her 10 points, a whole letter grade, for not answering the right combination of questions. Does that seem too harsh? Does it seem easy enough to explain to her when justifying why I penalized her this amount? I'm not entirely satisfied with this way to handle it, but I'm also at a loss for how I might better handle this case.


It seems to me that the issue is that she didn't follow directions. A letter grade may be harsh, but what really matters is that you're consistent. You should penalize her the same amount you penalize someone for doing too many points--they also didn't follow directions. The problem is that it sounds like you don't penalize students for doing too many points--you simply ignore, or scale down, the extra points, but that doesn't necessarily reduce their total grades. Since that's the case, I'd let her off with a warning, and then set a clear and consistent policy: if anyone does not follow directions (does too many points, does two from the same section, etc.), then you will reduce their grade down to the right number of points, and furthermore subtract an appropriate penalty from their total (10, or maybe 5, if that's too harsh). It's different if they fail to do enough sections--that may be failure to follow directions, but it could also just be a case of running out of times. Not having enough points to add up to 100 should be penalty enough in that case.

I don't penalize anyone for answering too many questions. They just did extra work that won't count. Extra work doesn't deserve point loss. In cases where they do too many points but removing one question won't work, I scale down one question to make the questions they answered (which is all I have to grade after all) fit into the point value that I can assign. It would be unfair for me to deduct points from those who simply do too many questions that I deduct from those who don't do enough, but that's what you seem to be saying (before then saying that would be different, so I'm not sure what you mean).

Jeremy - Donald Crankshaw has a point when he indicates that doing too many points breaks the rules too. I've had students to this--they've answered all of the questions in a section where they were only supposed to answer some. While it seems fine, the problem is when they get some of them wrong. I suspect their hope is that I will throw out the wrong ones and keep the right ones (of which there are usually enough to satify the requirements). Your scaling system is pretty good, but still works to their advantage. My instinct is to keep the wrong answers and then add in right answers until they meet the requirement. This is bad for the student, but that's their penalty for breaking the rules.

As for doing 2 from one section, normally I'd just mark the second one wrong. However, that'd be overly punitive in a case where the second question is a 40 point essay (when I encounter it, it's usually a 2 point question). So in this case, I'd say slap them with a standard penalty (say 5 or 10 points) and then grade as normal.

With things like breaking the rules (non-cheating), I feel like the goal is to give enough of a penalty that they'll follow the rules in the future, as opposed to the goal of reducing their grade (which may be a side effect). So a letter-grade seems to be the max necessary to achieve that.

I have a clear rule when students do too many multiple choice questions. I just grade the first ten and leave the rest ungraded. So everyone has their first ten counted.

The difference with this is that the questions are of different point values. If they did ten points too many, and there are two ten-point questions, I remove the second one. If there's just one, my decision is also automatic.

There isn't any rule against doing too many in a given section. Some students did four ten-point questions. Some did two twenty-point. Some did two thirty-point. It's just that you can't do two forty-point and still follow all the other rules. In this case, the only rule broken was to do at least one question from each of three sections. She would have had to do a twenty-point (along with a second 20 or two 10s) or a thirty-point (along with a 10) to meet the requirements, and she did a longer question for forty points instead.

Another option (and I don't know if you have time or the leeway to do this) is to refuse to grade the exam altogether for the student until the student follows the rules to the letter: the right number of points from the right number of questions from the right sections. It will be a "retake" of sorts for the student, with appropriate amount of time allotted for doing the missing question. Then count the extra answers as "wrong" answers (because the "right" answers would have been blank, i.e. not answered) and deduct points appropriately, and there is your final grade. (What would have been harsh is to treat both missing answers and extra answers as "wrong", thereby serving up a double whammy.)

I had a chemistry teacher back in high school who gave organic chemistry exams in the following way. We were supposed to come up with chemical structures for different compounds, and if we ever broke the rule that carbon is supposed to have exactly four (4) bonds, he would mark the infracting answer with the letters "ISH" (short for "I Stop Here"), and tally all the points to that mark and no further. Students have had entire sections wiped clean because they committed the unpardonable early on in the exam, and produced some truly epic fails.

Since they took this exam on Halloween, I'm not at this point having anyone do any of it over. If I were to do that, I might have this student do a 30-point and a 10-point and not count the second 40-point. I wouldn't require a whole do-over. But it's too far gone to go back and do any of it again at this point.

The chemistry teacher's response seems to me to be thoroughly immoral. It's one thing to count what they did in not following directions as a zero but to credit the rest of the exam as legitimate. It's quite another to select the point where they make a mistake arbitrarily as the cutoff beyond which they won't grade, when the student actually did all that later work and may have had an A exam except for one very early mistake, none of which counts, while another student has a mistake at the very end and gets most of the exam counted for having done all the work.

If the chemistry teacher fairly apprised his students prior to the exam of the consequences of the unpardonable sin as well as the reasons why that sin was unpardonable, then there is nothing immoral about the penalty imposed.

The fact that you are so unsure of what to do leads me to wonder whether your student might have been less than fully apprised of the consequences of her choice. If so, you should err on the side of generosity. Since the situation has never arisen before, I do not see any issue of consistency.

That's right. If the penalty was 100% clear beforehand, I don't have a moral objection. If it's unfair in terms of some being penalized more for the same mistake, but the unfairness was clear at the outset, then I wouldn't object.

You're right that I didn't publicize any penalty for this exact situation, because it hadn't occurred to me that anyone would do this exact thing. I expected some people simply not to do enough or to do too many, but I didn't expect them to answer two 40-point questions and not enough of another section while still getting the point total correct.

As I see it, it depends on why you have the particular rule that they must answer questions from at least three sections. What distinguishes the sections? If it's nothing more than relative difficulty, then a possible reason for your rule is to prevent the students from avoiding the more challenging questions. In this case, the student fulfilled the intent, if not the letter, of the rule and shouldn't be penalized at all. On the other hand, perhaps the sections are primarily distinguished by subject matter, and you want to assess your students on the broadest range possible. If so, this student has failed one of these sections by not doing it and should be penalized at least the minimum point value of one of the missed questions.

It's not subject matter, but it's not difficulty either. It's length of answer. The higher point values indicate that they should be spending a longer time preparing for it beforehand and answering it when they actually take the exam. I do think there are different skills involved in answering questions of different length, though. The details are much more important for the shorter questions, whereas broader connections between different points will play a bigger role in longer questions. I intended to have some balance between those things while also finding a nice scheme to give them some choice within limits while allowing questions to write themselves regardless of length, since I can sort them into categories according to length later.

When I've done this sort of test option array in the past, and a student here and there would do what you recount here, I start at the top and work down, and when I get to the max for a particular allotment, I stop, and then I don't even look at the rest.

Students who have ignored instructions as you describe have, in my experience, tended to feel entitled to get what they want. They struck me as trying to "work the system." It's a lot like playing Mom and Dad off each other to get what you want.
But that's just my guideline, each case is different.
It helps if you have too much work to do, too many papers to grade.
Bruce Meyer, Arlington, MA

Leave a comment


    The Parablemen are: , , and .



Books I'm Reading

Fiction I've Finished Recently

Non-Fiction I've Finished Recently

Books I've Been Referring To

I've Been Listening To

Games I've Been Playing

Other Stuff


    thinking blogger
    thinking blogger

    Dr. Seuss Pro

    Search or read the Bible

    Example: John 1 or love one another (ESV)

  • Link Policy
Powered by Movable Type 5.04