Is self-grading effective, and worth it? All signs point to “yes.” Some research findings appear at the end of this post.

Along with the minimum 50, self-grading is another high-leverage practice often found in an ungrading approach that keeps the focus on learning. In practice, though, self-grading is often misunderstood. If anyone hears about students giving themselves a grade and imagines a kid with their head on the desk all quarter who suddenly pops up and says “I get an A,” that’s dead wrong. With a solid self-grading practice that maximizes teacher prep time and empowers students to evaluate their learning, this student would lack evidence to make such a claim. And that’s one focus of this post (i.e., making a claim). Let’s first start with what teachers have been doing—historically—to make a claim about students’ grades so we can explain self-grading…

Gradebooks Making The Claim
Teachers assign work, collect it, evaluate it, score it, perhaps provide feedback, and record it in the gradebook. That is, they’ve been the ones in charge of collecting learning evidence, and once they have that evidence they make a claim about a student’s grade. More accurately, they have the computer make a claim using questionable practices, such as averaging all that evidence together, regardless of type, quality, timing during the learning process, circumstance, etc.

Right away, let’s recognize how this is all a great deal of work for the teacher. The logistics alone can sap a considerable amount of one’s planning time, which is one common reason teachers don’t want to change their grading practices. The irony! Next, let’s acknowledge that a teacher’s professional judgment is undone by most gradebooks. Considering all the teacher knows about their students, the gradebook just gets a series of numbers. It doesn’t know if one number represents a bad day, another represents a good day, another represents an absolute stellar day, etc. The gradebook takes them all, and it skews the representation of learning by showing an average at a level the student probably never understood/performed at. For example, giving the computer scores of 100, 95, 95, 65, and 55 spits out an 82. Let’s say the teacher got those original scores from six possible levels of understanding (i.e., 100, 95, 85, 75, 65, 55). Each of those numbers mean something according to specific criteria at each level, but notice how the student never did anything at the 85 level in the original scores. The closest are actually two scores of 95. That averaged 82 from the computer is misleading.

Teachers should find it somewhat insulting that all this work of grading gets them no closer to a more valid way of representing learning. And they’d be justified in feeling that way. Now, imagine the impact on the biggest stakeholders in grading: students themselves. So, let’s eliminate all that extra work and get a better representation of learning by having students be the ones to collect and evaluate their work that also builds reasoning skills (i.e., claim, evidence, reason).

What To Grade/Self-Grade?!
If you’re on board, the first step is to determine what to have students self-grade. Broadly stated, there are three categories of everything teachers have ever graded: process, product, and progress. I’ve heard teachers say that their students have a hard time reporting what they know and can do, so they wouldn’t want to trust a student’s self-assessment and self-grade when it comes to product. While research doesn’t necessarily back that up in every case, if you also feel this way, then don’t grade product. You’ve got two other options waiting for ya. Just determine what a student should be doing in order to learn, or compare growth over the quarter. EZPZ. Next, let’s look at the workflow…

Self-Grading Workflow
Quite simply, self-grading transfers most of the same traditional grading steps to the student end, with the teacher holding onto feedback on theirs. Students put something from the week (or whatever) into a portfolio (digital Google Classroom, Drive, or physical folder), then review and self-grade at the end of the quarter. The magic is that the teacher is no longer bogged down by assigning specific products, then collecting and evaluating them. The student does all that. Even better, neither the teacher nor student has to deal with scoring anymore! Numbers aren’t necessary when students are focus on learning throughout the quarter.

Over my decade in the classroom, I found that the best system was to have a Google Classroom assignment for record-keeping, even if physical folders were also used. That way, absent students could see that the rest of their class added learning evidence on Wednesday, the whatever of September, and could see that it was marked as “missing.” In my experience, many students never checked the official school gradebook, so our collection/delivery system of Google Classroom—also AMAZING for absent teachers to upload directions/resources/links to so the sub just plays crowd control—took the role of organizing that information.

I’d also recommend putting a collected/not collected note in that official gradebook for everyone who is looking at it, which includes other supports in the building, and perhaps at home. For example, in PowerTeacher, you can create an assignment as “collect only,” and then add combinations of symbols, such as collected, missing (i.e., was in class but didn’t turn in), absent & missing (i.e., not in class, but possible to turn in later), and absent & exempt (i.e., not in class, but not really possible to turn in later, such as a specific activity that can’t really be recreated). This is also done for organization, and to show those DAPS (Department Leaders, Admin, Parents, Students) what’s being collected in class. N.B. There’s no need to assign ANY scores in Google Classroom if you’re using some other official gradebook. Why do things twice?! That feature is for schools that are completely Google and use Classroom as their gradebook.

So, assignments are posted to Google Classroom for organization and a place for students to submit their work (or share what they just put in their physical folder), and the official gradebook can give additional information about the learning evidence that’s being collected. EZPZ.

Students Making A Claim
This part deserves the most attention, but is well worth it. Being able to rationalize one’s decisions and actions is an important skill. Students will probably struggle to connect the grade they chose to specific things they’ve done in class. The best advice I can give is to look at an example rationale as a whole class, and maybe provide some logical sentence starters (e.g., “My grade should be ___ because ___ shows that I ___”). Other than that, you’ll be mostly addressing individual students who need more support.

Claims are submitted to Google Classroom so feedback is easily given as a reply. A quick copy/paste of the student’s claim (i.e., “You said…”), then a reply that their learning evidence doesn’t support their statement, and what to do about it (e.g., “add something to your portfolio that shows…”). Even a single sentence from a student shows whether they’re understanding the connection between what they submitted an what they think it represents. This connection provides all the basis we need for justifying the grade a student thought best represents their learning. See this post on distinguishing between evidence of learning and documentation of just being present, and this one on guiding students to use their learning evidence to back up a grade claim.

FAQs
Q: Don’t Students Overgrade themselves?
A: Not really. A few will do so at a certain point, and some of the earliest research shows that was the case. Regardless, that’s why you give feedback on what students are submitting, and then help them make a claim, or see how their claim lacks support. In fact, I found that students undergraded themselves more often, also appearing in the research, probably from low confidence due to common pitfalls of a points-based grading system.

Q: Isn’t It Way More Work?
A: Not in my experience. If you hate communicating with students, even asynchronously, then you won’t enjoy giving the feedback that students benefit from, period. Students do need guidance on self-grading since they have far more autonomy than they’re used to. The “work” you’re thinking of in the Q also depends on how you imagine what grading should be. If you THRIVE in taking a Sunday morning to head to a coffee shop with stacks of papers, and you enjoy the mundane red pen thing, then under this new system of giving a sentence or two of feedback on a Google Classroom assignment it all might feel like more writing to you than in the past. If you call that “more work” than all the assigning, collecting, evaluating, and scoring, then yes, this system will feel like more work. If so, find a workflow that highlights its reduced work so you can feel how much easier everything goes. Maybe you start by only commenting on students who submit questionable learning evidence. Once you’ve got that down (and realize it doesn’t take as much effort as you thought), start giving feedback on the good learning evidence that students submit, too, focusing on the content (e.g., “I really like that you…I notice that…etc,.”).

Some Findings:
While Davis & Rand (1980) found one of the larger discrepancies I’ve seen between students grading themselves and what the teacher would have graded them (i.e., 28 of 87 college students had different scores), the students overall reported enjoying class more, and scored higher on two of three major assessments. Sadler & Good (2006) found a very high correlation (0.91 to 0.94) between students grading themselves and what the teacher would have graded students. The meta-analysis of Sanchez et al. (2017) found that self-grading had long-term effects (.34) across 33 studies, and that frequency didn’t matter (i.e., don’t self-grade any more often than necessary, which is like just once per quarter or progress report). Unlike the early Davis & Rand (1980) student, fewer than 5 out of 50 students suggested different grades from the teacher’s own expectations in Guberman (2021).