The way I've graded for a few years goes like this:
1. Gather what you need to grade: papers, books for checking citations, etc.
2. Get yourself a "cool tool" or two. For me, it means this:
- Filling up pens with an interesting color of ink (green, purple) for the paper versions.
- Download e-versions to grade electronically on the iPad (iAnnotate has improved exponentially lately!).
- Or, if it's early in the semester where I'm still giving lots of explanations about things, open up the file of auto-text or cut-and-paste entries so that I can use those for routine things and spend more time really writing comments about the content.
3. Write down the students' last names in some kind of order. I mix it up so that I don't read the same students' papers first or last every time. This serves two purposes: (1) you can't avoid a student's paper and (2) you get to cross the names off the list. If you are at all the "cross it off the list is very satisfying" kind of person, this really helps.
4. Get a timer and figure out how long you're going to allot for each paper. You may need to adjust the time after the first few, but if you've been teaching for a lot of years, you should have a pretty good idea of how long they should take you. If you're tempted to take longer, ask yourself this: "Is the student going to benefit from this additional comment or correction?" Sometimes it's "yes," but often the answer is "no," and you have to move on.
5. Build in some breaks or changes in activity. Flavia recommends taking a break every 6 papers, and that sounds good. I also change it up by grading X number of electronic versions and then X number of paper versions. A change may not be as good as a rest, but it helps.
I have colleagues who prefer the "10 a day, every day" system, and if that works for them, that's great. Since I am an ace procrastinator, what this meant was that I would spend a couple of hours dreading grading, a couple of hours grading, and then a few hours trying to settle down to writing or reading because my mind was still back with the papers. Where grading is concerned, I'm a monotasker and definitely not a multitasker.
Another advantage is that for me, there's a norming process that goes on so that I can grade more consistently from paper to paper, since the overall features of the whole set and its issues are in my head somewhere.
Grading still takes longer than I think it ought to, given this system, but the end result is what Flavia talks about: once it's done, it is done, and you don't have to think about it any more until the next set. That's incredibly satisfying.