Logo: Deutsches Institut für Internationale Pädagogische Forschung

Press Releases

Losing in a quiz but winning in learning

People who initially predict the possibly right answer in a knowledge test are subsequently better able to memorize it, particularly when the predictions turn out to be false. A new study by the German Institute for International Educational Research (Deutsches Institut für Internationale Pädagogische Forschung – DIPF) and the University of California, Berkeley has revealed this finding, recently published in the journal “Learning and Instruction“. The researchers demonstrated that the element of surprise when predictions are falsified is implicated in the increased learning success.

“What do you think is the correct answer?“: Many students are familiar with this question from their classrooms and lecture halls. After all, teachers and lecturers have made good experiences when adding a playful quiz component to cramming. Psychological researchers now report systematic evidence for the beneficial role of such predictions in activating prior knowledge and improving learning. “Our study presents the first systematic approach to investigating whether and how predictions facilitate learning“, said Professor Garvin Brod from DIPF, who was in charge of this recent study.

Comparisons of countries, football results and dilated pupils

36 students took part in this study, and the team of researchers conducted two computer-based tasks with them. In the first task, the test focused on geography. In the second task, students were asked about football results, an interest in football being conditional to their participation.

In each task, tests were run in two variations. Regarding geography, test-takers were first of all requested to predict which of two European countries had the larger population size. Students were then shown the correct answer. In a second variation of the test, the students again saw pairs of European countries. This time, the correct solution was immediately given and the students were then asked to state which solution they had expected. In order to find out which method was more conducive to learning, the students‘ knowledge was once assessed right at the beginning of the test, and again at the end. Here, the students were asked to sort all the countries that appeared in the test variations by size of population. They turned out to be more successful regarding the countries for which they had previously stated predictions.

The football tests went along the same line, in that students were asked to predict results from the German first league. Again, formulating predictions led to a higher learning achievement. This time however, the researchers also aimed to find out which predictions were particularly conducive to learning, the correct ones or the false ones. The outcome was clear, as Professor Brod points out: “Particularly the false predictions contribute to the method’s success.“

The researchers hypothesized that false predictions elicit surprise, which itself might be the reason underlying an enhanced learning effect. This hypothesis was tested using an eye-tracking camera that recorded changes in participants’ pupil sizes throughout the test series. And indeed: only when participants first stated a prediction and then found out they had been wrong, their pupils dilated – and this was taken to be an indicator of surprise.

Conclusion and outlook

While the study as such does not prove that surprise is indeed the relevant factor for improved learning, such a conclusion seems plausible because a correlation was also demonstrated between the extent of dilation and the increase in learning. Brod: “For example, one might assume the invocation of surprise to enhance attention, and inspire learners to reconsider their assumptions.“ Such a correlation needs to be further investigated. The group of researchers furthermore plans to study the effect of false predictions on learning that was thus demonstrated in a laboratory situation in real-life lesson settings. Even at this point in time, Brod draws a first conclusion: “There is a benefit to using quiz-type elements in learning scenarios“.

Garvin Brod has written an article on this study, together with DIPF colleague Professor Marcus Hasselhorn and Professor Silvia Bunge from the University of California, Berkeley. The article is already accessible as an Online-First edition:

Brod, G., Hasselhorn, M. & Bunge S. A. (2018). When generating a prediction boosts learning: The element of surprise. Learning and Instruction, 55, 22–31. doi: 10.1016/j.learninstruc.2018.01.013


Study: Prof. Dr. Garvin Brod, DIPF, +49 (0)69 24708-139, R2FydmluLkJyb2RAZGlwZi5kZQ==

Press: Philip Stirm, DIPF, +49 (0)69 24708-123, c3Rpcm1AZGlwZi5kZQ==, www.dipf.de/en/

About DIPF:

The German Institute for International Educational Research (DIPF) is situated in Frankfurt am Main and in Berlin. DIPF generates empirical educational research, digital infrastructures and targeted knowledge transfer, thus contributing to coping with challenges in education. The Leibniz Institute prepares and documents knowledge for education, to support scientists, policy-makers and practitioners in education– to the benefit of society.

last modified Apr 25, 2018