... and how to deal with and/or avoid them
It’s nearing the middle of the semester, and that means many of us are dealing with midterm grades. If you’re using an assessment system where the focus is on feedback and revision, this can cause some trouble. How do you determine a midterm grade when a student’s learning is truly in progress?
In this post, I’ll talk a bit about what I do, and how you can navigate different reporting requirements.
What are midterm grades for, anyway?
One of the biggest problems is that midterm grades try to serve so many purposes. They are overloaded with “meaning”. Here are just a few of the common things that midterm grades are expected to do:
Act as a progress check for students: What grade do they have right now?
Serve as formative feedback for students. Are they on track? If a midterm grade isn’t where a student would like it to be, what can they do to improve?
Help institutions identify students who are floundering in their classes and need support.
Track the progress of first-year students or students on academic probation.
Report grades for students who have special academic requirements, especially athletes.
So midterms are both summative reporting and formative feedback intended for students, offices, and institutions. Simple, right?
I vary my approach depending on who requests the grade. This usually determines what information is actually needed, and how much leeway I have in reporting it.
Midterm reporting for students
I regularly make progress reports for students, but I rarely report a letter grade in them. That’s because I find that what students really want1 is formative feedback: Am I headed in the direction I want to go, and if not, what do I need to change?
A single letter grade isn’t actually useful for this, since all it can say is either “things look good” or “something isn’t going well — good luck figuring out what!”
So the information I try to give to students is clear, actionable feedback to help steer the rest of their semester: What are they doing successfully, and what do they need to work on?
If you’re using a form of iterative grading like standards-based grading or specifications grading, your records already have what you need to make good progress feedback. Some things I usually give feedback on:
Has the student been meeting objectives regularly? If not, I give advice on which objectives to focus on, how or when they will have an opportunity to do that, and what resources are available for review. This might involve an invitation to come to an office hour and talk about a specific topic.
Has the student been taking advantage of reassessments (assuming they have things that need to be reassessed)? If not, then that’s a major piece of steering advice. Students might not be used to reassessments, much less seeing them as an expected and regular part of their workload. This is a chance for a reminder about how reassessments work and why they are important.
Is the student doing other, non-graded things that enable success, such as: Completing prep work, attending class, meeting with teams, etc.? It’s probably worth reminding students about the expectations, benefits, and reasoning for doing these things.
Some of this can happen informally during class or office hours. But I also like to give concrete feedback to every student, especially so that I don’t miss those who are shy or have a misunderstanding about their own progress. I either email feedback to students or post something in the gradebook in our LMS.
I automate some of this feedback. Here’s a typical example of written feedback from a class that I recently taught:
Here is your mid-semester grade check-in. The comments below give my feedback on your progress in each grade category.
Daily Prep (10/12 completed): OK -- be careful not to miss any more, or these will drop your grade by a “-”. These are due 2 hours before each class on Blackboard. I recommend starting daily preps the night before they are due.
Quizzes (M’s on 11/11 objectives so far): Excellent work! Keep it up on the last few quizzes.
Homework (70% average): Your homework average is low mainly because you haven’t revised any homework problems yet. You can revise each homework assignment once, and the new grade replaces the old one. This is a great way to show me what you’ve learned during the semester! Please start doing this soon. You can find instructions for how to revise on Blackboard.
I created general comments for common types of progress. For example, students who completed 10 or 11 of the quiz objectives all got the same comment, and a simple spreadsheet command2 drew that from a table in my gradebook and inserted it into this comment. After a quick skim through each part of my gradebook, I wrote personalized comments for students in unusual quiz situations, but that was relatively few. This student had automatic comments for Daily Prep and Quizzes, and a customized homework comment.
In the end, students often still want to know: “What grade am I headed towards?” It’s an understandable question, one that has been enabled and encouraged by years of being able to check out their “current grade” in a traditional weighted-average system. Reporting a letter grade, whether to students or the institution, requires some more work that I’ll discuss next.
Midterm reporting for institutions
Sometimes, it’s impossible to avoid a midterm letter grade. For example, my institution requires midterm grades for all first-year students, and the form I enter them in only accepts letter grades — no feedback at all!
So, when you have to determine a letter grade, how can you do it? Here are some common options:
Make a projection based on current progress
Suppose you have 20 course objectives, but students have only attempted 10 of them by midterm time. That’s half of the objectives, and so a student who has met 7 objectives now might be expected to complete about 14 by the end of the semester. You can use your final grade requirements and report the grade the student would earn for 14 objectives.
Just like in a weighted-average system, this “current grade” is not truly a valid grade, because it represents only part of the whole semester’s work. It is a projection: a guess at what a student might earn if they somehow continue to work in exactly the same way but on different topics.
In a weighted average system, this guess might be moderately predictive, if only for the reason that averaging tends to “lock in” a student’s grade, giving past attempts at least as much weight as future improvement.
But in a system that values reassessments with full forgiveness, things can change dramatically before final grades are due. A student might reasonably meet all 10 current objectives by the end of the semester, earning full credit. The 70% projection becomes effectively meaningless!
To address this, some instructors invent various ways to guess at future progress. For example, if you use a multi-level rubric, you might include “partial credit” for students who have made some progress (e.g. reaching “progressing” rather than “not yet”). This is dangerously close to using points and partial credit, and can encourage students to think about your class’s grades in those ways when that’s not at all what they mean.
Alternately, you might “fuzz” the final grade criteria a bit for the sake of determining current grades, giving students benefit of the doubt. But in the end, none of these approaches are very satisfying to me, and I tend to avoid them.
Give the letter grades meaning
One of the biggest advantages of standards- or specifications-based grading systems is that they directly tie student grades to real, concrete achievements. Letter grades have a definite meaning. Projections or “partial credit” calculations throw away this meaning and replace it with guesswork.
So here’s another option: Give midterm grades meaning too. Just as you can write “bundles” or narrative descriptions for final grades, do the same for midterm grades. There are lots of ways to do this, but here’s the simplest. Assign only three midterm grades: A, C, and F. Before you start assigning, write out specific definitions for each, for example:
A: This student is making satisfactory progress and is taking advantage of reassessments as needed. I have no significant concerns about their progress.
C: I have a significant concern about this student’s progress for at least one reason, such as: Meeting very few objectives; Not taking advantage of reassessments when needed; Consistently not completing important ungraded work (such as pre-class prep).
F: This student has stopped coming to class; has missed a significant number of assignments; or is otherwise in a situation that suggests they will not pass the class.
You could easily define B’s and D’s, but I like these three. Just as with all-or-nothing assessments, it is easier to make consistent decisions about a limited number of options. Be sure to share these definitions with students, because they will likely interpret them as a projection, even if they aren’t meant to be read that way.
No matter what, I avoid assigning +/- or “half” grades (such as AB). The fewer categories you have, the more clearly you should be able to determine where a student’s progress fits.
Go with your gut
I’m not usually a fan of “going with your gut”, especially on grades. If I “know good work when I see it”, then I should also be able to write down some clear descriptions of what that good work is, so that students can know it when they see it as well. This clarity of expectations is a central part of standards-based and specifications grading systems.
But for midterms, it might be worth letting your gut have a go at things. Take a look at each student’s overall work and make a judgment, including current progress and how well they are using resources.
If you go this way, I recommend two things:
Tell students that this is what you’re doing. Let them know that they can (and should!) talk with you about any questions they have about your judgment. It’s also worth reminding them that these grades truly are based on a current status, and that they are neither a promise nor a threat about the final grades.
Limit the number of grade options you assign. Much like the A/C/F scale I described above, it’s easier to make consistent judgments when you have fewer options. Avoid the 11-level yea/boo meter of +/- grades, and don’t pretend that you (or anybody!) can consistently distinguish between a B and B- at midterm time.
Finally — hear me out for a second — you could refuse to enter letter grades. An entirely honest and consistent response is “There is no grade for this student. They are still learning.”
Of course, if you go this direction, be aware of potential consequences. If a student will suffer consequences due to external rules, or if the university will put a hold on their account because unentered grades count as F’s, I will do something. That might be finding a way to invent a midterm grade (and it will probably be an A in all but the strangest circumstances). Or it might be having a conversation with whoever is requesting the grade, with a bit of feedback for the student and some pointed questions about the purpose of using midterm grades in this way.
There might also be consequences for you as well, and only you can decide if those will be acceptable to you. I know what the consequences of not entering grades are for me and I’m ready to handle that — especially because, as a tenured professor, an annoyed email from the registrar is likely to be the worst of it. If you’re in a situation like mine, why not use that bit of extra power you have and see what happens?
I was going to write something thoughtful here, but instead this is what came out: Midterm grades are a pain. They distract from what I love about teaching and force me to do things that I don’t believe in. But, as long as I am giving students useful, actionable feedback, I’ll keep dealing with midterm grades as well.
Although not necessarily what students ask for. “What grade am I at?” is usually a coded way of asking if a student is meeting expectations in a way that will lead to a good-enough final grade. I sometimes have one-on-one discussions with students who insist that they want a single letter grade rather than feedback. Those are excellent opportunities to remind students that I want them to succeed, that their grades are directly tied to concrete things they can do, and perhaps discuss the mechanics how grades and reassessments work.
Here are some more details for those interested. This assumes familiarity with basic spreadsheet commands, specifically VLOOKUP. For numeric data like the number of objectives completed, I create a small table off to one side in my gradebook. Its first column contains the lower end of each range that I want to leave a comment about, and its second column is the corresponding comment:
I have a summary column in the gradebook that contains each student’s total number of objectives completed. Next to that, I add a new column that uses VLOOKUP to insert the comment from the table that corresponds to that student’s total. Now, each student has an automated comment next to their total. I do a quick read through and identify any students who I want to give a special comment, and replace the text in this column as needed. If I have multiple categories that I want to leave comments on, I repeat this process for each summary column.
Finally, I create a cell that stitches together all of these comments together with boilerplate text to create a personalized comment:
As you can tell, my gradebook tends to have many tabs (one per major category of assignment), but that’s just my personal style — this would work well in any grade spreadsheet.