I have written a lot on this blog about grading, about my longing for “ungrading”, my qualms about heading that direction, and the things I’ve decided to do instead. This is a more specific post about the latter, and my gradual shift in emphasis I’m calling “Modified Ungrading”.
We know the arguments against A-F grades. They encourage a focus on marks rather than learning, and they can be discouraging. We know there are also arguments in favor. They rank performance against a standard, they inform the student of their level of success, and they mark improvement. (Many years ago I heard a speaker say he did not want to drive across a bridge built by someone who got a C in engineering, or be operating on by someone who got Cs in medical school. I get that.)
This is an area where faculty workload and student desires can profitably intersect. Most of us teach large groups of students, and at community college we do it without teaching assistants. The big argument in favor of ungrading is that students benefit more from individualized feedback than from grades. But in this environment, tutorial-style feedback and individualized learning isn’t possible. Auto-grading, with current technology, is possible, and helps fill the gap between grading and ungrading.
In most learning management systems, when a student submits something, a certain number of points can be automatically assigned. The arguments for and against doing this are the same as those for grading in general. But in addition, auto-grading provides immediate marking and a sense of completion, two factors which are often ignored in discussions of grades vs feedback. 
So far, I’ve implemented auto-grading for two types of assignments, both of which are submitted only to me: lecture notes, and document paragraphs. What this enables me to do is have 2 points automatically assigned, then go back and look at the work itself. Sure, I could assign the points when I look, but auto-grading instead changes things, for me and for the students.
For the students, they get their two points immediately, so it’s confirming they did it. For me, I then have ample time to go back and peruse their work. And with the points already given, I can pay more attention to my Assignment Comments (feedback). The tendency is to talk to them first, then think about the grade, instead of the reverse.
I’ve had to change the grade sometimes, but I’ve had the chance to explain individually before doing that, and to give an opportunity for corrections.
This shift to Modified Ungrading may seem ridiculously subtle, but it’s quite real. Mentally I can shift from “processing” the work of 40 students per section, to responding to it personally.
In fall, I intend to apply this to discussion forums, which we use for posting primary sources. Ultimately, I would like everything auto-scored except writing assignments, on which I can spend more individual time.
This way, I can focus my work more on communication than grading. And when students see their points immediately, and rarely changed, I’m hoping they will do the same.
