« Blog Symposium on Radin's Boilerplate | Main | Learning from exams »

Thursday, May 16, 2013

Spreading out grading

I am happy to say I have finished grading for the semester and it was as thrilling an experience as ever. I experimented for the first time with a mixed short-answer/multiple choice format for the final in Civ Pro and liked it a lot as a testing mechanism; it gave me a good sense of what students did and didn't know (I will have more to say about that in a later post). I also did not find grading it overly burdensome.

The real struggle for me was grading the take-home essay portion--that is the part that feels overwhelming. And it struck me this cycle that the source of the struggle is several-fold: 1) the sheer number of essays to read all at one time, 2) that they all say basically the same thing (things actually, since students wrote on one of 3 questions), and 3) the fairly short time window (about a week) to get them all read, which even if sufficient time, feels crunched. So while it is perhaps too soon (my grades have not yet posted and I have not yet met the deluge of questioning 1Ls), I am thinking about alternative approaches for next spring.

One thought is if and how to spread written assignments, and thus grading, over the course of the semester. So: At the end of each portion of the course (for example, Pleading or Subject Matter Jurisdiction), I would assign a group of students to write an essay on that topic, due 7-10 days later. This would mean I am grading more regularly during the semester--I would have a group of papers to read every other week or so (more frequently if I sub-divide a longer portion of the class, such as pleading, into sub-parts). But I would be reading fewer of them at once and on less of a deadline Maybe I am completely wrong, but it feels like that would be more manageable and less of a slog than reading 60 papers all saying the same thing all at once (or even reading 30 papers on one topic at the same time I'm reading 30 papers on another) with a week to turn them in while also reading and grading their in-class exams. And I also believe (again, perhaps wrongly) that I will do a better job of reading and grading with more time and fewer papers.

I see a couple of obvious drawbacks to this. One is that students might balk at the "unfairness" of having their workload at a different time than their classmates, with every student believing that her time--early in the semester, late in the semester, close to legal-writing time--is disadvantageous. But I think randomizing it might help alleviate the objection. Another is the trick of making sure I can fairly compare grades across different assignments on different topics, but I've been doing a version of that for a few years, so it is just a matter of careful problem selection. It also may be more difficult to assign two essays each semester, as I've been doing the past few years.

Does anyone do something like this? And how do you find it works? Are there other drawbacks I'm missing?

Posted by Howard Wasserman on May 16, 2013 at 09:31 AM in Howard Wasserman, Teaching Law | Permalink


TrackBack URL for this entry:

Listed below are links to weblogs that reference Spreading out grading:


Howard, one other drawback (that might be more pertinent to Civ Pro than other subjects) is that atomized problems deprive you of your ability to test (and encourage the students to learn) the way in which things fit together from different parts of the course. To give one example, how the joinder rules interact with supplemental jurisdiction.

Posted by: I. Glenn Cohen | May 16, 2013 10:00:29 AM

Well, I at least can test *some* of the students on that if I use an supplemental jurisdiction essay. I also can do a short-answer question to test that (I usually manage to include at least one on the exam), so I should still be able to get around to it.

Posted by: Howard Wasserman | May 16, 2013 10:09:05 AM

I am not sure it is realistic to believe that you can grade 60 essays every month, on a quick-enough turnaround to prevent complaints, with enough substantive comments to satisfy 1Ls who want significant feedback. I am not saying it is impossible, but only that it is a massive time commitment and may - counterintuitively (in my mind) - result in less satisfaction with your course.

Posted by: Anon | May 16, 2013 10:59:15 AM

Anon: That's not what I'm proposing--quite the opposite, actually. I am talking about doing, say, 10, every two weeks or so.

Posted by: Howard Wasserman | May 16, 2013 11:18:18 AM

Have you thought about designing a rubric that you could share with the class and have them grade each others? Sharing your rubric would allow students a better sense of what you expect their answers to look like. Having students grade each other's anonymous work would allow them to see what other students are doing well and what they are doing poorly. Finally, it speeds up the grading process.

Posted by: Matthew Bruckner | May 16, 2013 11:48:55 AM

Howard, I think this is an interesting idea, and like you, I've been thinking about how to better teach and test students. I resisted a mid-point assessment earlier in my career in part because I feel there is something about the courses I teach that isn't completely processed until the end of the course. I would expect students to write differently on contract law after the third week than after the tenth week (whether I start with consideration, damages, or something else). Thus, to grade an assignment at different points in the semester, on the same or different material, seems like it would inherently favor students who have been with you for the whole semester.

An added difficulty of different material is that, in some areas, some topics are much more difficult. As bad as a memo assignment on 2-207 and the battle of the forms might be, letting half the class write on promises grounded in the past instead seems destined to leave half the class doing less work for the same reward, or getting a greater reward for the same work.

Posted by: Jake Linford | May 16, 2013 2:23:08 PM

Matthew: peer review is valuable; peer grading is not. Unless you're grading based on the very low-level mechanics of the writing (which I hope law school courses rise above) no rubric can possibly be specific enough to eliminate spurious variation in the grades. Courses I've taken where essays were graded by a team of TAs had a hard enough time synchronizing their grades. It'd be impossible in a course where every student is a grader.

Besides, you'd practically have to allow students to appeal their grades to you, which means you'd probably end up spending almost as much time reading and grading as if you just did it yourself in the first place.

Posted by: Andrew MacKie-Mason | May 21, 2013 10:56:00 AM

Post a comment