next »

[grad_geek]

Thanks to both Magistra and Scene!

I think I'll start with the advice from that UMN web site, and see if that helps me out with the grading next semester.

(I suspect that grading will always be the part of teaching that I hate...)

[iomhaigh]

I'm not sure that this method would apply for your discipline, but I have found that grading one page at a time causes me to grade much more consistently.  Once I am past the first page, I have no idea whose paper I have in front of me (except for the rare student whose handwriting is easily identifiable) and I can concentrate solely on the content.

I do this on exams.  I grade all answers to essay 1, then all to essay 2, then all to essay 3, etc. 

...

I do the same, only exactly the opposite. I grade the last question first. This has the added benefit of "hiding" the name until I get to the first page. (NB I tell my students to only put their name on the first page. I am thinking of changing to a cover page for name, etc next time so I don't see their name until I am done completely). I picked up this technique from an u/g professor, so it is not at all original.

And like others, I use rubrics heaviliy for the longer problems (like writing code or short essays). The yes/no understand/no clue questions are easy.  :-)

I use blue books (turn the page and the name is gone forever!) and forbid them from writing their name inside it.    The latter doesn't work all the time, but the big old sharpied out blobs that they see when they get it back reinforces that I wasn't kidding.

Of course, these are essay exams.  The other bonus to this method is that I get far far fewer whining complaints about grades because I tell them how I grade and that if they want to discuss their exam, then they need to bring it to my office hours because I honestly have no idea who wrote what on which question, much less how what one person did as a whole on the exam because I don't grade them straight through.  The lack of through-line makes it fairer.  It also means the lazy whiners just don't show up to complain because they can't do it right then and there without my saying "I don't know what you did wrong.  I don't know what you wrote."

[trabb]

I probably do this to some extent.  Here's why I don't worry about it:

1.  For each set of papers, I begin at a different point in the alphabet and grade alphabetically; thus everyone at one point or another gets their papers read early, late, and middle.

2.  The effect is negligible.  I'm not going to give what should be "A" work an "F" because I'm grumpy at the end of a big stack of essays.  At most, it's going to drop an "A" paper to an "A-" or "B+." That shouldn't happen, but over the course of the semester, the odds that this will have any effect on the final grade are pretty slim.

[finallyfullprof]

Since I don't deal with "paper" papers anymore, I have my own electronic system. I do first read-throughs of everything online just to get a gist of how the students did. Then I arrange them into groups of however many papers I need to grade each day to return the papers in a timely manner.  I try to set them up so that I have at least one A or B paper in each group and at least one god-awful paper in each group so that those don't cluster into any one day and skewer my perspective.

I use both comments and a rubric for out-of-class essays and comments only for in-class essays.  All points go into WebCT as I grade so students can see the points right away. I generally follow with the comments within 24 hours.

[anon4now]

I do something similar to what one prev poster suggests.  I sit down and read through the whole stack, fast, impressionistically, sorting into +, ok, - stacks (the good, the bad, and the ugly?).  Then I rest, do something else for a brief time, exercise, clean house, whatever, for a half hour but not much more. Then I come back with rubric and pen, and adjust my impression/stacking as I go, because often my second read, a slower read-through, is different from the first. It honestly is worth the extra time---but only if you read that first time very fast and without marking anything. 

The other thing is, you can do this on line with the comment functions in wordprocessing software, but the temptation is to get too longwinded----they will not read many comments, and the physical space of the margin keeps you limited in a good way to concise, on-point comments.  I also agree completely that the good students need more attn than they usually get, and the bad ones less, so I grade the top papers first, and spend more time on them; by the time I get to the bottom/ugly stack, I am circling things on the rubric and scrawling "WHAT EVIDENCE SUPPORTS THIS CLAIM? IS THERE A THESIS STATEMENT IN THIS ESSAY? F" or "S-V AGREEMENT THROUGHOUT; D" and other such "global" comments.

Don't worry---over the years you do actually get better and faster at this.

[yemaya]

And like others, I use rubrics heaviliy for the longer problems (like writing code or short essays). The yes/no understand/no clue questions are easy.  :-)

I use rubrics too.  It saves a lot of time and keeps me consistant as I become tired and/or frustrated toward the end of a run of papers.  If anything, I could see myself getting irritated and starting to grade harder toward the end if it were a bad batch of papers.  I think we're all human and all prone to this problem, but there are ways to make sure that the grading doesn't suffer.

[chicklet]

I mark the papers on the first pass, then go back and assign a grade based on the totality of information from what I've read, for the reasons OP pointed out (did I not explain the assignment well, was the first paper I read unusually good or bad?).

[galactic_hedgehog]

I'm not sure that this method would apply for your discipline, but I have found that grading one page at a time causes me to grade much more consistently.  Once I am past the first page, I have no idea whose paper I have in front of me (except for the rare student whose handwriting is easily identifiable) and I can concentrate solely on the content.

I do something similar. I try to grade one or two problems at a time so I can be as consistent as possible. If I change my mind about something, I then go through all the previous ones to verify/change the way I graded them.

Alan

Ditto for me.  I also find that this allows me to get through everything faster, in that I get into a groove for those questions and don't have to keep looking back my notes or answer key.

I also try not look at whose exam or lab I'm grading (of course, like everyone, I have those students with distinct handwriting) but sometimes the answer is so abysmal (or, more rarely, so excellent) that I just gotta see who it was.

[yemaya]

I'm not sure that this method would apply for your discipline, but I have found that grading one page at a time causes me to grade much more consistently.  Once I am past the first page, I have no idea whose paper I have in front of me (except for the rare student whose handwriting is easily identifiable) and I can concentrate solely on the content.

I do something similar. I try to grade one or two problems at a time so I can be as consistent as possible. If I change my mind about something, I then go through all the previous ones to verify/change the way I graded them.

Alan

Ditto for me.  I also find that this allows me to get through everything faster, in that I get into a groove for those questions and don't have to keep looking back my notes or answer key.

I also try not look at whose exam or lab I'm grading (of course, like everyone, I have those students with distinct handwriting) but sometimes the answer is so abysmal (or, more rarely, so excellent) that I just gotta see who it was.

This may not help with lab reports or hand-written assignments, but I require my students to submit their papers with a cover sheet and have all their info just on their cover sheet.  I flip over the cover sheets and don't look at names until I log their grades.