I enjoy learning new things myself, so this past weekend I not only spoke at ResearchED NYC but also attended several sessions. I would like to briefly reflect on Ben Lovett’s talk about test anxiety. We had a very small audience for him, which surprised me. I consider the subject to be very topical, and the man clearly knew what he was talking about. Here is a brief report of his presentation.

For anyone who talks to teachers will run into this. Anyone who listens to students, even faster. And the numbers don’t lie either: a large group of young people experience exams as something that is accompanied at least “sometimes” to “often” by nervousness, worrying, and physical tension. That ties in nicely with the broader narrative about an “ anxious generation ”. But as Lovett put it, it became particularly clear that we are taking one crucial step too quickly. We see fear, and we conclude: therefore, poorer performance.

Only… that turns out not to be entirely correct. One of the central questions in his lecture was remarkably simple: would the same student perform better if he were less anxious? The answer he gave, based on extensive research, was nuanced, to say the least. In fact, it often turns out that this is simply not the case.

Yes, there is a negative correlation between test anxiety and performance. But it is small. And more importantly, as soon as you account for differences in knowledge and skills, that connection largely disappears. Students often perform approximately the same on low-intensity tests as on high-intensity tests.

To be honest, this surprised me, because it is not an easy conclusion. It means that part of our explanation is actually a misinterpretation. We see that weaker students report anxiety more often, and we reverse the direction. Whereas it could just as easily be that the lower skill level helps explain the anxiety, and not the other way around. Lovett made it even clearer by saying: do not simply assume that a student who reports test anxiety performs worse as a result.

But that is not the end of the story. On the contrary. Where his lecture became truly interesting was in the question: if test anxiety does not directly undermine performance, where does the effect lie? The answer, according to Lovett, is in behaviour.

Students with more test anxiety study differently. They procrastinate more often. They are quicker to resort to superficial strategies, such as rereading or reviewing summaries. And they avoid precisely those activities that have the most impact, such as practising with questions or solving problems. In research on math anxiety, for example, you see that students spend less time doing practice exercises and more time looking at worked examples.

In other words, the impact of test anxiety lies not so much in the moment of the test, but in the process leading up to it. And that has consequences for what we do in schools.

A first reflex is often to want to eliminate the fear itself. Breathing exercises, relaxation techniques, perhaps even extra time or separate spaces. Lovett was remarkably cautious about this. Not because those things never work, but because they run the risk of reinforcing something else: avoidance. I immediately thought of Philippe Geubels’ program about anxiety and how to deal with it. Avoidance was never the best option there.

That idea of ​​an anxiety-avoidance cycle came up several times. The more you avoid situations that evoke tension, the less chance you have of experiencing that you can handle it. And so the anxiety persists, or even becomes stronger. That makes the narrative a lot less comfortable. Because it means that well-intentioned adjustments sometimes perpetuate the problem. Something that also came up frequently in his presentation. Often, what we do at school to combat test anxiety actually perpetuates that test anxiety.

So what does work? Again, surprisingly little that is spectacular. Good education. It seems to be a recurring theme lately. Think of tests that are clearly structured and have clear expectations. Examples of questions so students know what to expect. And above all: studying in a way that resembles the test itself. Practising recall, answering questions without a textbook, and regular short quizzes that spread out studying. I personally introduced constructive alignment in the session, something Lovett was less familiar with.

And all this is not because it would “solve the fear,” but because it better prepares students. And that preparation proves to be more robust, even under stress.

There was another, slightly more painful part to the lecture. It is not just students who experience test anxiety. Parents do too. There even exists a scale that measures the extent to which parents are tense on exam days, worry about making mistakes, or experience physical stress around their child’s tests. That tension rarely remains without effect.

According to Lovett, what schools and parents can do is not act as if tests don’t matter. Instead, they should normalise that tension is part of the process, while simultaneously making it clear that this tension does not mean you will perform worse.

In concrete terms: Test anxiety is real. It is widespread. And it deserves attention. But it is not the simple mechanism we sometimes make it out to be. It does not automatically undermine performance. It works indirectly, through behaviour, through preparation, and through expectations.