Oh joy, it’s time to grade more papers!
At the moment, in fact, I have two batches of papers (approximately 400 words each, approximately 100 papers per batch) to grade, since I hadn’t finished marking the earlier ones before the next ones came due. And of course, owing to the piles of smoking rubble that constitute our budget at the state universities right now, there are no funds at present for graders.
I’ve blogged before about my strategies for grading fairly and consistently without taking a million years to finish the job. I’m still more or less using these strategies. But today, I’m trying to work out a more specific question:
What is the optimal number of papers for me to grade in a sitting?
There’s surely some number of papers beyond which not taking a break in grading is likely to be bad for me (either making me stiffen up from sitting so long, or making me grumpy or fatigued enough that I actually nod off in the middle of writing a comment; yes, this has happened) or for the students whose papers I’m grading (because a grumpy grader might be less generous with partial credit than the even-keeled grader she was N papers earlier in the stack, and because comments written by a fatigued grader are not always coherent enough to be useful).
However, there also seems to be some minimum number of papers that one must grade to get “in the swing” of the grading activity and to achieve the quickest turn-around per paper. (I find this is true even if I’ve already, in earlier sittings, graded a significant portion of the total stack. To get back into efficient grading mode takes a few papers at least.) Another consideration is that if the number of papers you grade per sitting is too small, you never get done with the task.
I have a strong sense that the minimum number of papers (of the 400 word length) that ought to be graded per sitting is no lower than 5. I suspect the maximum number of papers I ought to be grading before a break is around 20. But I’m interested to hear from others in the grading trenches about how you figure out the optimal number of papers to get graded in a sitting.
Yes, I do have my suspicions that considering this as an optimization problem may be a way to avoid actually sitting down and grading the damn papers. But there may be folks out there interested in discussing this particular issue of educational practice — and some of them may also want to put off digging in to their stack of grading for the weekend.
I think the answer to this depends upon the length and nature of the papers. The papers in my advanced classes are a bit longer (15-25 pages). My students only write one paper in the entire class, so the consequence of the paper for their grade is quite large. For that reason, I try extra hard to develop strategies for grading them evenly.
Early on in my career, I read them all twice – each one quickly to get the gist and then a second time slowly to assign a grade. Even though this is likely the most fair way to evaluate the whole set before assigning scores, I eventually came to a situation where I simply did not have the time to do it this way.
As you noted, reviewing multiple papers is better (in my case, I don’t sit down unless I can do at least 3). I find that if I sort through them first, I can get more done. Based upon my interest in the subject (my students write on a diversity of topics), my impression of the likely quality of the paper (based upon my experience with the student) and the length, I organize them into triads (trying to combine good, average and hard reads into single triads). I find that I can make it farther if I mix up the papers in this way. I think the risk of going through the good papers and getting stuck with a run of less well-done ones is too high not to try and even this out a bit.
In addition, I try to keep myself going by rewarding myself (short jaunt to Starbucks, a game of fetch with the dogs, a cookie) in between each triad (my own personal homage to Skinner). That seems to help me to stay on task, though if I read a paper that is really poorly done or otherwise disappoints/antagonizes me to an extreme degree, I know well enough to put them all down and walk away.
In general, when I do it this way, I can read 3-4 triads per sitting and get them done in a few days (I would typically have about 40 to do).
I’m tackling a grading assignment of similar size this weekend (100 exam pages, each with two “short answer” questions that fill the page). I’ve been working on them for 1-1.5 hours at a stretch, which is enough time to do about 20-25 pages. Before I started grading I also read through the first dozen or so exams to get a sense of what the average response was (and flipped through the whole stack looking for great and terrible papers, to find the range). It’s been going well so far, although I do find myself looking back at previous grades, to make sure I’m doing everything fairly (giving the same grade to two answers that made the same sorts of errors).
There is no fixed number since there are a bunch of variables in play that are very hard to quantify. These boil down to how much energy and concentration can I muster for the current run of grading. The better you feel the more you can take on. That said David has a good comment with the take a break doing something that has absolutely nothing to do with the papers (but you know that already, right? :).
Once or twice, around finals time, I graded 35-45 papers. This was back in my TA days. I was totally burned out at the end and just wanted to be done. These were 3-5 page papers, and I was cranking through them at 8-10 minutes a paper or better. By the end, I’m afraid I really wasn’t paying as much attention to the papers as I should have.
I’d definitely say 20 is a good max, depending on the length. Better yet, your max is when you’ve stopped actually paying attention to the paper and/or when you notice that you’re taking significantly less or more time to grade the paper.
I found I graded best when I set a timer and allotted myself 10 minutes per paper (for those 3-5 pages; 400 words could probably be done in about 5 minutes). I’d definitely block out at least an hour straight to do papers, and I’d break up hours with a little walk around break (bathroom, make tea, etc). I’d typically also do a short walk around in the middle of the hour (water bottle filled or something) as a mini-brain break.
If I found that I started taking lots more or lots less time, I’d just take a good break for some longer thing and pledge to return back to the papers. It worked for me, when I could find the time to do it, and I was always able to justify my grades and reasoning well when it came time to handing back to the students.
That’s a tough question; I have struggled with it as well. With longer papers (800-1000 words), I have found that the optimal number is 10 to 15 in a sitting, with at least a 45-minute break between sittings. With short ones like you’re dealing with, 25 is a good number.
But I do preparation work. I will pick 10 or so papers out of the bunch and spend awhile perusing them to create a kind of ideal paper in my mind that is derived from the good points I find, good points I don’t find, and doesn’t include the average, poorly thought-out statements and arguments that the sample papers contain. That way I can move more quickly through the rest, looking out for the keywords and phrases that denote a proper understanding of the subject they’re writing about. This is possible because even if they had some leeway on the topic of the paper, I always give them, orally or in the assignment itself clearly-defined expectations; even before I start marking, I know roughly how the points will break down. Incidentally, this not only makes marking easier, it allows me to dismiss most of the complaints coming from the inevitable whiny students who want their grade raised…it is hard for them to argue when they can go back to the assignment and see exactly where they didn’t follow my instructions.
I’ve also noticed that BAD papers are much more difficult to grade than good ones, because in the absence of what should be there, you have to explain to them why what they wrote is wrong or irrelevant. I can get a good 400-word paper marked in a minute or two when I’ve done the prep-work right and marked a few papers already. A bad paper could take 5-10 minutes. Typically a 25-paper sitting would take 2 hours at most. However, I always stop at the 2 hour mark regardless of how many I’ve done because frankly, after 2 hours, my concentration and focus go out the window.