Why we should test our students more

Assessment is learning, and students value it. Katy Burgess explains the psychology of testing

Katy Burgess is a Senior Lecturer in psychology at Cardiff University.

Testing has a bad reputation.

The word itself creates images like the one above. Of students crammed in exam halls, stressed and anxious.

There is a drive in education to reduce these kinds of exams, and more generally reduce how much we assess our students. Aside from the stress it causes for students, there are numerous other arguments for a reduction in assessment.

For instance, testing is not the best way of assessing learning. Students often cram and regurgitate what they have been taught, rather than building higher order skills we seek to develop. Cramming does not lead to long-term memories, meaning shortly after the exam, the students have forgotten the course content. And for academics and professional services, high assessment burden can mean it is difficult to keep on top of workload.

Testing is learning

However, testing, otherwise known as retrieval practice, is a fundamental part of learning and long-term memory formation. You will likely have experienced a failure in retrieval with a word you don’t often use. You can perhaps remember the first letter, but you are not able to retrieve the word itself. The word is on the tip of your tongue (this is actually called the tip of the tongue phenomenon in the literature).

The reason we experience tip of the tongue is because we haven’t retrieved that information enough. Like a muscle, the more often we use the muscle, the stronger it gets. The more we retrieve information, the stronger the memory. It’s why you can remember every lyric from a song you sang hundreds of times as a teenager but haven’t heard for years. Every time you sang along to that song you were retrieving the lyrics and the melody. Thus, your memory for that information is strong.

Teaching in higher education is a balance between teaching skills and subject-specific knowledge. For example, you expect a medical student to be able to recall what symptoms are associated with certain illnesses. Of course, students need to develop higher order skills such as critical analysis, but these come after having foundational knowledge and being able to apply and manipulate that knowledge in new situations. If you want to build knowledge and skills across years, students have to remember what they were taught in the first place. This won’t be the case if they sat one exam or assessment, and were never given opportunities to retrieve the information.

Student use of testing

Psychological and educational studies have evidenced that retrieval is one of the most powerful learning tools we have. And yet, it is underutilised by both educators and learners. For example, staff often don’t offer opportunities for retrieval during teaching sessions. Students tend to revise content by watching back over lecture recordings, making notes, highlighting, mind mapping and so on. They often don’t test themselves until right before an exam, as they use testing to check what they know.

This is unsurprising, given research exploring people’s judgements of learning. Learners are bad at identifying what strategies will lead to better memory, and often use techniques that increase familiarity and comfort with the material (like going over the content again) but not memory of it. Learners tend to avoid making errors, when errors are in fact very beneficial for learning – as long as they are aware they made an error.

The role of educators in testing

This is where educators come in. If we are testing student knowledge throughout our teaching, we can identify where students are making errors and give feedback. For example, if I am teaching my students about a concept, I can run a quick multiple choice or open-ended question using polling software and see how many people are getting it right, and if they aren’t, what types of errors are they making. I can provide immediate feedback, help students who made errors to see what is wrong about their answer and why. If lots of students are showing they are struggling to understand , I can tailor my subsequent teaching to go over this concept again. This is testing, but not as we typically think of it.

There are numerous ways testing can be embedded as part of the curriculum. How we embed it is less important than it being done in the first place. The benefit of retrieving information holds whether there are motivational consequences or not. That is, students will learn the same amount retrieving for a high stakes test than a no-stakes test. When thinking about the best way to provide opportunities to retrieve, it is worth considering the usual factors, such as cohort size, topic, student behaviour and so on. For larger cohorts, in person testing can help build community and encourage students to discuss answers they were unsure of with one another. Regular low stakes summative testing can help students to feel empowered and like they are chipping away at their degree throughout the semester. Regular testing also reduces test anxiety, but does create more anxiety than other learning strategies.

We need to change the narrative around testing. Testing is not a way to assess what people have learnt. It is a key part of the learning process.

If we want our students to be able to build their skills, we need them to be comfortable with regularly retrieving information, and applying it in novel contexts. Perhaps surprisingly student feedback on additional testing is positive.

Students have said they have learnt more from lectures with embedded retrieval than any other. This is because retrieval is learning.

4 responses to “Why we should test our students more

  1. Thanks, Katy
    I have experienced similar in my own teaching. I’ve found Audience Response Technology useful in assessing real-time retrieval. It has also prompted immediate changes when retrieval was lacking in places. The students loved it, especially when done in groups, but generating personalised feedback.

  2. A great article, thanks Katy. This speaks to something fundamental about the art and science of what we do in the classroom. I still think we overassess in HE, in terms of high stakes summative assessment. I wonder to what extent that is driven by a misunderstanding of the point you make here – that testing is vitally important to learning, but not all testing has to be (and nor should it be) in service of ‘getting a grade’.

  3. Agree – as was the case 20+ years ago when the concept of formative assessment as an engine for learning came into being – just a new description for an existing concept! The trick, though, is to convince students of the value – many do not because it isn’t “real” assessment. I have seen increasing differential attainment in those students who engage in formative activities of this kind and those who don’t. Despite all manner of approaches, including showing data of previous cohorts engagement and performance in subsequently learning, even for modules several years later, I have made little progress. So, maybe the key question is how can we encourage students to see what neuroscientists have known for decades?

Leave a Reply