Skip to content

Overtesting in K-12 is ruining one of the best things we can do for college students, and we need to talk about it.

Posted in K-12

Have you ever had the experience of sharing something that you knew was exciting and important, something that inspired you and led you to change how you do your work? This is how I feel when I talk about retrieval practice and its key outcome, the testing effect. In short, this effect consists of a robust improvement in memory for information that you’ve been quizzed on (or answered questions about, or quizzed yourself on – the details are remarkably unimportant).

The discovery that retrieval boosts memory is something that dream teams of cognitive psychologists have written whole books about. It is something that I’ve written about too from the angle of technology, because that is one of the key things ed tech does for us – it makes it possible, easy even, to put into practice the long-standing advice that we should quiz students frequently and with a low emphasis on grades.

When I talk to faculty about the testing effect, many share my enthusiasm. And more of them already know about it, far more than when I first started speaking and writing about this topic 10 or so years ago. This is inspiring, and quite a testament to how far we have come in incorporating learning science into teaching in higher education.

Let’s be clear: Quizzing helps. Short answer quizzes are better than multiple-choice ones, but both kinds help. It helps regardless of where exactly the quiz questions happen during study, and can even help when there’s no feedback given. Incorporating frequent low-stakes quizzes may even reduce achievement gaps among college students – in other words, helping the students who need it the most.

And in contrast to what critics may assume, the learning acquired through quizzing doesn’t just hold for a few types of artificial laboratory tests, nor does it evaporate in a matter of hours or days. Also contrary to the idea that tests don’t tap “real learning,” quizzing actually helps students transfer what they know into new contexts, which fits my definition of what substantive, lasting learning really is. Nor do tests automatically feed anxiety, as demonstrated by one study of middle and high schoolers who reported less test anxiety after doing regular retrieval practice in class.

Get this: Retrieval practice even helps us remember things we haven’t studied yet, a phenomenon called the forward testing effect. And through the process of test-potentiated learning, quizzes boost the effectiveness of studying that we have yet to do.

All of these benefits stand in stark contrast to the underwhelming impacts of rereading, which tends to be the default method of studying in college. In general, students are remarkably unaware of what retrieval practice can do for them, so it’s up to us to build it into our courses, and to tell them why we’re doing this, so that they can start doing more of it on their own.

When I set up faculty professional development experiences, I often incorporate quizzing in one form or another. In that setting, faculty participants really like it – they say it brings an element of fun or even suspense, that it heightens the sense of surprise, wonder, and importance of what we are doing.

So why do so many still recoil when I suggest that we do the same for our students?

There are philosophical reasons – a belief that spending time on tests trivializes what’s being learned, or debases the classroom atmosphere. In some classes, a case can be made that too much, or maybe any, quizzing sets up a false impression that there are clear-cut answers to questions within the discipline. I’d have to agree that if a major priority for your teaching is to highlight ambiguity and debate, tests are not the way to do that.

But another set of reasons has to do with a big, angry elephant in the room: K-12 overtesting. As most everyone knows by now, we in the United States have gone all in for standardized tests as a way to evaluate teachers, schools, school systems, and kids themselves. Used as a brute-force instrument to rank and punish educators, these time-consuming exams took the focus off learning and off children’s social and emotional development, crowding out instruction and crushing innovation.


The spectacularly unpopular and ineffective set of education policy decisions attached to high-stakes testing all make up the school environment that most of our students experienced when they were growing up. Many of us in higher education specifically try to undo the damage wrought by this system. We want to show students that learning can be more than that: more engaging, more unpredictable, less of a chore and more a joy.

So even when we know that quizzing could be great for our students, we shy away, frightened by the specter of constant, stressful examinations that reduce learning to memorization and reduce teaching to an exercise in quality control.

If you teach in the United States, this is the reality. Merely denying it won’t make it go away. So is there anything we can do to redeem tests as a method for accelerating learning?

Some of it comes down to basic PR and framing. We can point to the fundamental differences between testing-for-learning and testing-for-ranking as we’ve come to know it today. The reason for K-12 standardized tests isn’t, and never was, to promote learning. There’s little or no feedback on specific items that you got wrong. The tests themselves are often exhausting, stressful, and high-stakes – all the things that properly deployed tests-for-learning should not be.

We can also continue to spread the word about the research basis for quizzing, which is remarkably coherent as well as sizeable by now. When we do this, we should be mindful about how we translate data to interpretation to application, because a lot can get lost along the way. In particular, we should stress that these ideas aren’t predictions, or opinions, about what might happen or ought to happen as a function of quizzing. They’re actual empirical findings, and while they are not guaranteed to replicate in absolutely any environment on every single occasion, that should elevate them to a different level as a basis for advice to teachers.

We can make tests-for-learning look and feel different from tests-for-ranking. To do this, we can focus less on the ritual of the test (put all papers away! no talking! have two sharpened #2 pencils or else!) and more on eliciting the underlying cognitive processes that make tests work.

These processes are many and myriad. For one, tests make very clear what we already know, and what we need to go back and review. When I give my students practice quizzes (using the Kahoot! app, which is wonderful for the purpose), this is what they want to get out of it: a reading on where they stand and where the trouble spots are before they go into a higher-stakes exam. Without something like a quiz, it’s easy to let familiarity seduce you into believing you know more than you do, something that especially trips up students with weaker metacognitive skills. Testing also elicits the effortful, attentive processing that is the basis for forming new connections and new memories. It’s not the only way, but it’s one tried and true way.

Even though we want students to be putting in focused effort, that doesn’t mean it has to be a chore. Retrieval practice is just that, practice, and there’s no reason why it shouldn’t feel more like a relaxed rehearsal session than an examination. Gamification is one way to do this. In small doses, features like competition and points for speed (both features of Kahoot!, by the way) can set a tone of low stress and high engagement.

And most importantly, we can teach students about retrieval practice as part of the metacognitive skills that all lifelong learners need to have.  In this way, tests-for-learning can become something that’s shared with students rather than something that’s done to them.

My hunch is that students themselves tend to get their heads around barriers to quizzing quicker than faculty do. They may not love the idea of sitting down once a week or more to do the work, and not every quiz-based assignment may spark their love of learning, but they can appreciate the payoff it gives for time invested, and learn to incorporate it into their own study routines.

That brings me back to what it’s like to talk to other faculty about this issue. I love to do it, but the flip side of my excitement is that it sets me up for a pretty rude shock when I do run into resistance. When someone tells me that they would never reduce learning to such a rote exercise, that doing so extends the woeful legacy of terrible educational policy, when they tell me with absolute certainty that multiple choice tests have no bearing whatsoever on what a person knows – I feel it, and it’s uncomfortable.

But, that comfortable cocoon of familiar research is probably good to challenge from time to time, and something I know I can’t stay inside if I’m going to open up more and better conversations about teaching and learning.

We in higher education cannot ignore what retrieval practice can do to accelerate learning, and we owe it to our students to build it into our classes. This is hard when we, and our students, are lugging around so much baggage from the slow-unfolding disaster that is K-12 standardized testing in the United States. As in so many things, I think deep and direct conversation with our own colleagues is the first step forward.