Testing is an effective way to improve learning
Over the last few months, we’ve been talking about competences, knowledge and school, but never about assessment.
On this topic, have you ever heard of the Roediger and Karpicke experiment?
You can read the full study or this article summarising the salient points.
In this experiment, three groups are defined, alternating study (S for Study) and test (T for Test).
- The first group studied for four periods = SSSS
- In the second group, we study for three periods and then carry out a test = SSST
- In the last group, we have one study period followed by three tests = STTT
The study shows that the more we evaluate, the better the retention rate.
Moreover, here’s what I think is also important to remember from this experience.
- Multiple re-readings, which are widely used by students to learn, lead to the illusion of competence. In other words, after reading something, you think you know it, but this belief is false, at least in the long term.
- A test is generally perceived as a means of measuring what has been learned, not as a means of learning.
- However, as the authors state, “Tests are an effective means of improving learning, not just assessing it”. When you assess (or assess yourself), you know that you don’t know. You understand that you need to revise the lesson, that you need to practise, and so on. As for the teacher, he or she discovers that this or that notion needs to be revisited, and so on.
And the best part of all? The study shows that testing yourself, even without feedback, encourages long-term retention of information. This is what is known as the “testing effect”, which engages the student in a process of reactivating and consolidating knowledge.
Interesting, isn’t it? What’s more, AI now makes it possible to produce on-the-fly assessments in less time than it takes to say it. In short, let’s assess, assess, assess.
