According to a study by the EHU, testing oneself during learning, even if it leads to getting it wrong, is a more effective strategy than simply re-reading

A study by the group of Eugenia Marín has confirmed that doing tests when acquiring knowledge is more effective than simply re-reading the subject matter, because errors made in assessments, if followed by corrective feedback, promote memorisation. The conclusion was that the benefits occur no matter whether the tests are administered before or after revising. The work highlights the need to change the perception of error in learning.

Although failure is an intrinsic part of learning, education environments have traditionally tended to avoid errors for fear that they will interfere with memory. However, there is growing evidence that making mistakes, far from being harmful, can actually enhance learning. This was confirmed by research led by Eugenia Marín: “The study shows that learning from mistakes is a solid, effective strategy. We analysed how incorporating testing when acquiring new knowledge, allowing or even encouraging mistakes during learning, yields better results than methods based solely on re-reading the subject matter,” said the researcher in the Learning and Cognition group at the University of the Basque Country (EHU).

This conclusion was drawn from several experiments conducted specifically to find out the effect of incorporating testing into learning sessions. In this respect, the EHU research aimed to determine the temporal circumstances in which errors improve memory. The subject matter that the participants had to memorise consisted of 108 pairs of words that were weakly related semantically (e.g., seal–ball). “We wanted to see whether it is better to conduct the tests after exposure to the subject content (post-test) or before (pre-test), thus activating prior related knowledge and expectations, even though the content was unknown, which could lead to errors being made,” explained Marín.

The results showed that the order of the tests is not a determining factor: “The people who did the pre-test or post-test obtained similar results to each other and, in both cases, better results than the participants in the group that did not do any tests,” she added. So, the study reveals that those who used the pre-test or post-test as a learning strategy obtained more correct answers in the final exam than those who limited themselves to reading the lexical pairs during the revision phase.

So the findings of the EHU study confirm that learning from mistakes is a robust strategy that can be widely applied, and various effective ways of doing so are indicated. “In the classroom, you can give a test after presenting content to reinforce what has been learned, or have a quiz at the beginning of the class on a subject that has not yet been covered. What is more, in another study we found that even when the pre-test is given at the end of a class and feedback is provided in another session up to two days later, it remains an effective strategy,” explained Marín.

The researcher also stressed that these methods can be applied not just by teachers, and encourages students to use the tests on an individual level: “For example, before starting on new subject matter someone who is revising can look at what topic will be covered next and ask themselves what they know about it. One can also test oneself after having worked on the subject matter, as part of the preparation.

Tendency to underestimate error-based learning strategies

The EHU research also gathered the learners’ perceptions of different learning strategies. Before concluding the experiments, the participants were asked whether they believed it was more beneficial to experience mistakes or correct answers, and the majority opted for learning without mistakes. In other words, using methods such as re-reading the information.

“The potential of incorporating testing and making mistakes into the learning phase tends to be underestimated. This is because when we re-revise or re-read something, we have a sense of fluency, and feel that we have all the information accessible in our minds and that we have learned it. However, testing ourselves and discovering that we have suffered the impact of forgetting is unpleasant because it shows that there are things we have not managed to learn. That is why, although the learning strategies that require the most effort are those that yield the best results in terms of long-term memory, in metacognitive terms, we believe that they are not as effective,” explained Dr Marín.

The author of the research emphasised that the best long-term learning involves effort: “Although we may feel more at ease when we reread or underline, what really leaves a significant long-term impact are the methods that force us to make a greater effort; for example, when we try to recall and confront our own mistakes. Errors are a driving force behind knowledge, and it is important to normalise them in the learning process.”

Further information

Eugenia Marín has a PhD in Psychology, and is a lecturer and researcher in the Department of Basic Psychological Processes and their Development. She teaches on the Degree course in Psychology and the Master's in Cognitive Neuroscience of Language. Right now, her main line of research focusses on examining the effectiveness of different learning strategies. This research is part of a project funded by the Spanish Ministry of Science, Innovation and Universities entitled “Testing, errors and learning” in which Eugenia Marín is the lead researcher. The paper was published together with Yeray Mera and Nataliya Dianova. Mera currently lectures on the Degree Course in Psychology and this paper is part of her PhD thesis read in 2025 and supervised by Marín; and Dianova is a pre-PhD researcher in the Faculty of Philosophy under the supervision of Marín.

Bibliographic reference

Yeray Mera, Nataliya Dianova & Eugenia Marin-Garcia

Testing before learning; Exploring the robustness of the pretesting effect

Memory and Cognition

DOI:10.3758/s13421-025-01813-x