Navigating the challenges of assessment frequency

The problems with assessing too much or too little, and how to find a balance

Photo by Shubham Dhage on Unsplash

One of the many reasons assessment is hard, is determining when and how often to do it. Assess students too little, and the assessments become high-stakes and anxiety-inducing. Assess too much, and the course becomes a joyless march from one quiz to another. Alternative grading does not provide a solution to this problem. What is the right balance?

That question took a practical form recently as I was preparing a talk on alternative grading to faculty at Cardiff University in Cardiff, Wales¹. My background research involved learning about grading practices in higher education in the United Kingdom. Briefly and to the best of my understanding, the typical UK approach is a descendant of the Oxford tutorial model, where students attend lectures and tutorials, but there are few graded assignments, sometimes none, until a single final examination at the end of the term. A survey of syllabi from the Cardiff University math department confirmed this: Most of them, regardless of the level of the course, allotted between 70% and 100% of the course grade to a final exam.

When I asked current and former UK-based faculty how they felt about this, opinions varied, but they all communicated a common concern: We want to avoid over-assessing students. A concern with alternative grading approaches like we promote at this blog is that allowing reattempts without penalty necessitates having lots of assessment, which creates the very anxiety that it is trying to alleviate. I think my UK colleagues have a good point here, and it’s an issue I’ve been dealing with for some time, so here are some thoughts about it.

Assessment versus grading

I want briefly to make a distinction between assessment and grading. Many times these are used as synonyms, but they aren’t. Assessing a student means simply determining what a student knows and what misconceptions they may have². Grading a student on the other hand means evaluating their work and assigning some kind of mark to it. These are not the same. We can assess students without grading them³.

In fact, decoupling assessment from grading is the fundamental principle behind “ungrading” or collaborative grading; and while many UK courses may have only one graded item, they are often supplemented with numerous non-graded formative assessments that take place during tutorial sessions. (In this way, the old-school UK models were very much a progenitor of modern collaborative grading practices.)

When I refer to “under-” or “over-assessment”, I am thinking about graded work like quizzes or homework or projects, not the basic human act of coming alongside a student to see what they know. Although, an instructor can certainly fail to do the latter often enough, resulting in an impoverished relationship between instructor and student and the many follow-on issues that arise.

The dangers of under-assessment

Over-assessment may have its problems, which we’ll consider later, but what about the opposite, which is when you assess students very infrequently (as seen, in my view, by many of the university courses I’ve looked at in the UK)? Under-assessment in the grading sense seems to carry two main issues.

First, the less frequent assessments become, the more “valuable” they are in terms of the economy of the course grade (because they are “worth” more) and the greater the consequences for making mistakes. This is just simple supply-and-demand, and everyday experience from school. And, the greater the consequences of failure, the more anxiety is induced in students, which ironically makes them more likely to fail; and the less likely a student is to take even the most basic risks that are attendant with learning.

Second, less frequent assessments means fewer iterations of the feedback loop that drives all learning. We should think of reattempts of assessments as assessments themselves, because that’s what they are; so, here, I’m referring to situations where not only the original assessment is infrequent but the reattempts are limited. Final exams are like this, being one-time-only assessments with little or no reattempt opportunities. Midterm exams with only 1-2 reattempts possible are similar. Either way, the feedback loop never has a chance to get going.

A hybrid approach like the one frequently used in the UK, where there are a small number of high “value” graded items supplemented by frequent non-graded formative assessments, is well-intentioned and seems like it could be done well, and probably is done well in many places. But it also does not seem to really avoid either of these issues in my view. The grade in the course is still largely determined by one event (or a handful of them) and I have to believe students will feel the anxiety and pressure attendant with that low frequency. And if that event is a final exam, there are no do-overs, which isn’t helpful for learning.

The benefits of frequent assessment

If we make assessments frequent, there are certain upsides that we can expect.

One of those is retrieval practice or the “testing effect”. This refers to the phenomenon where retrieving information from memory (for example, during an assessment) improves long-term memory and recall more effectively than passively studying material. A 2011 study by Karpicke and Blunt, for example, showed that students who read a passage for five minutes, then put the text away and spent 10 minutes recalling everything they could on a blank page, outperformed other students on a memory task who simply read the passage, or read it and then re-read it later. I don’t think we want our students to take assessments every 5-10 minutes! But engaging in frequent graded assessments gives more opportunities for retrieval practice.

Another is simply having more opportunities for engaging in a feedback loop. When students are assessed frequently, as long as they are getting helpful feedback, it translates into more opportunities to iterate and build a strong framework of understanding — one that truly represents understanding, not just a one-time performance result. More frequent iteration through feedback loops also builds the specific mental muscle of knowing how to convert feedback into improvement — a meta-skill that is highly valuable and yet little discussed or practiced, seemingly.

A third benefit is the lowering of the stakes on any individual assessment and thereby making the whole class environment more failure-tolerant and less brittle. If a student who otherwise is progressing well in understanding class material happens to have a bad day when there is an assessment, more frequent assessments will provide a cushion, much like an error-correcting code can self-recover if there is a glitch in the message being sent.

The dangers of over-assessment

But it is possible to have too much of a good thing. I learned a lesson about over-assessment recently in a course I taught which, in a rare situation, had a fairly light schedule of topics to cover. I decided to use the final 25 minutes of every Friday for quizzes over our Learning Targets. These quizzes consisted of a collection of small problems (really more like “exercises”) over one Learning Target at a time, and each quiz would cover several targets. I thought having these weekly would be really beneficial for students, since there would be plenty of chances for students to show what they knew and reattempt things that needed work, and being so frequent no single one would be such a big deal.

But it turned out they were a big deal, in a way I didn’t anticipate. What happened was that the class took on a culture of constant quizzing. Quizzes were so common that every class meeting and every activity became about preparing for the next quiz; and every question turned into “On a quiz or a test, will we be asked to….?” The class stopped being about math at all, and became a class about how to pass math quizzes.

And although having these weekly made them less scarce, and thereby lowered the “value” of each one for students, I expected students to be less anxious — but they weren’t. Students were just as anxious before the weekly quizzes as they used to be when the assessments were once a month, or less. In fact they may have become more anxious about these because they were frequent despite their being small, and mostly consisting of reattempts of previous learning targets. I’m not sure why this is. I think there is just something about timed testing that makes timed assessments not subject to the laws of division, where you would expect a greater “denominator” (the frequency) to lower the overall amount of test anxiety.

At the same time, while the anxiety was the same or even higher, the high frequency lowered the perceived value of each quiz — like inflation devalues money or grades. Many students admitted to not putting in the same level of effort to prepare for a weekly quiz as they would a monthly test, because they knew if they messed up on a quiz they’d just retake it later. We sometimes take this attitude as a good thing, but in this instance I don’t think it was good.

Having weekly quizzes also made it hard for students to have time to engage meaningfully with feedback loops. It takes time to grade those quizzes, and by the time they are returned, there are just a few days until the next one, which isn’t a lot of time for some students to process the information in my feedback or put it to use in practice.

And finally, somebody has to grade those quizzes, and the sheer volume of grading — weekly, times 60 students, and the length of the quizzes grows during the semester as students have more Learning Targets to reattempt — was just too much.

All of what I just described are real dangers of assessing too frequently, even if the assessments are small in proportion to their frequency.

Striking a balance

Back to the question at the beginning: What’s the right balance? There’s no scientific or magical formula for this, but I’ve found the following guidelines to be helpful in finding that balance:

• It’s OK not to assess as often as possible. For example, just because you can give assessments weekly, or even more often, it doesn’t mean you should — or that students would benefit from it. A little scarcity can be good, and more opportunities to attempt and reattempt assessments does not always translate into more mindful engagement with feedback. It could just as easily lead to overwhelm (for students and for you) and that culture of constant quizzing I mentioned.

• Design a minimal cadence for assessment, then add extra opportunities as needed. The semester following the one I described earlier, I dialed back the quiz schedule to every other week, then used the 12-week plan to use almost all of the last two weeks of the semester for additional reattempts as needed — either by appointment, or in a targeted way (for example if there was a “core” Learning Target that needed its own quiz). I was worried at first that cutting the frequency in half would severely curtail student performance, but in fact there was almost no impact at all, and the “quiz culture” went away. Later, I dialed it back even further to monthly assessments with extra reattempts in the last two weeks, and I have found this to be far less work for everyone and just as effective for students. (You can always schedule “extra” attempts if you start with a minimal schedule; but you cannot start with a full schedule and then remove attempts.)

• Be open to alternative forms of assessment. If a student wants to be assessed on a Learning Target in some way other than a written quiz (which is the default) then I’ll listen to what they propose. I can’t always accommodate them, but if they’re willing to do (for example) an oral quiz in my office then since it’s their idea, they are invested in it and should feel less anxious about it and more motivated to do it. (I should note that it’s worthwhile to proceed with caution on this route, since being fair requires you to consider any student’s request for an alternative with equal seriousness, and it can lead to a different form of overwhelm for you if you have normal quizzes and alternative assessments for every single student.)

Finding the right balance in assessment is more art than science, and it’s clear that neither over-assessment nor under-assessment is the ideal path. The key is to design a thoughtful and flexible approach that allows for both frequent feedback and the opportunity for students to reflect and improve without feeling overwhelmed. It’s essential to consider not just the frequency of assessments but also their quality, the opportunities for reattempts, and the overall learning experience. By carefully adjusting the cadence and being open to alternative forms of assessment, we can foster a more effective and supportive environment for student learning. Ultimately, the goal is to create a space where assessments serve their true purpose: to help students grow and deepen their understanding, without adding unnecessary stress to their educational journey.

1. Unfortunately for me, this was an online talk.

2. In educational use, it comes from the Latin assidere which means “to sit beside”.

3. I’m not sure if you can grade student work without assessing it, but I think it’s possible if we grade mindlessly.