At Cambridge, we believe it is essential that evidence should drive developments in educational policy and practice. We also recognise that it’s important to validate hypotheses – and to interpret results carefully and appropriately before using them to inform teaching and learning. In this blog, we explore how theory and evidence in relation to cognitive science has the potential to benefit learning and teacher practices, which is one of our 12 principles for the future of education. But we also note how ignoring evidence in favour of theory alone can create challenges. Cognitive science is here broadly defined as the study of the human mind including the mental processes and its architecture.
Examples of how cognitive science evidence can be used in education
Many studies have investigated the potential for evidence from cognitive science to inform teaching and learning. Some of the most interesting relate to the usefulness of retrieval practice in learning school science, and the use of cognitive science as a means of improving mathematics teaching and learning. Here we take a look at a few examples to see what benefits can arise, but also what can happen when theory overshadows the evidence.
Using retrieval practice to enhance science learning
Evidence from cognitive science has challenged the commonly held belief that elaborative study is better for learning than retrieval practice. (1) ‘Elaborative study methods are those that require learners to organize new material and distinguish unique features of terms and concepts in the service of future retrieval.’ (page 1) (2) An example of such a method involves learners drawing a concept map of how they understand information and make connections between concepts, whilst using reference material such as a textbook. Retrieval practice, on the other hand, involves retrieving knowledge from memory, that is, reactivating knowledge in response to a cue – such as a question. Retrieval is not simply reproducing knowledge; it is a more complex process that involves reconstructing knowledge which has been brought to mind. The issue here is that teachers often devote a lot of learning time to elaborative study methods, as this is commonly thought to result in meaningful long-term learning. But there is strong evidence that this is not an optimum approach.
retrieval practice is a promising approach to improving learning
A series of experiments were run to test the benefits of elaborative study and retrieval practice. (1) In these experiments, learners studied using various combinations of three methods: study (reading a text); retrieval practice; and elaborative concept mapping. A week later their conceptual knowledge in science was assessed using a short-answer test of verbatim questions and inference questions, to investigate the effectiveness of the methods. The research showed that retrieval practice alone was more effective than both study and elaborative concept mapping, and that retrieval practice can be an effective tool for promoting learning. A recent review (3) of research in education settings has concluded that retrieval practice is a promising approach to improving learning. The review also concluded that supplementary research is needed for the development of specific approaches in teaching and learning – an important process of turning experimental findings into precise guidance for teaching. Practice examination questions have been suggested as one retrieval practice strategy that could be used to facilitate learning. (4)
Using cognitive science evidence to improve teaching and learning in mathematics
Rittle-Johnson and colleagues (5) have focused on how cognitive science and classroom-based research can be used to maximise mathematical learning. They have drawn specifically on two sets of cognitive science findings:
(a) when considering just one thing, people tend to focus on surface features, (6) and
(b) when considering a pair of things people tend to note key structural features of each thing as well as similarities and differences between the two things – a process that aids learning. (7) (8)
comparing correct versus incorrect examples is more conducive to learning
These findings from cognitive science have been supported by classroom-based research. For example, in one experiment (9) involving two groups of students, Group 1 compared correct versus incorrect solutions to mathematics problems, while Group 2 compared different types of correct solutions only. The students in the two groups were asked to talk through what they were doing, and later the researchers qualitatively coded the articulations. Group 1 discussed correct concepts more frequently than Group 2, suggesting that comparing correct versus incorrect examples is more conducive to learning.
Rittle-Johnson and colleagues (5) drew on this study (9) as well as other research on applying cognitive science to mathematics learning, to create materials for teaching, teacher CPD, and curriculum development. These materials were then used in practice, evaluated, and subsequently refined as part of a trial. Provisional findings indicate improvements in teaching and learning in response to the trial. Overall, this work illustrates that cognitive science together with other evidence, in this case classroom-based research, can be used to improve teaching and learning.
A separate group of researchers (10) investigated the effectiveness of applying a number of cognitive science principles for improving mathematics teaching and learning:
- comparing things can improve learning
- guiding students in mathematical conventions such as noticing and practising graphical conventions and interpreting visualisations (e.g. graphs), can improve comprehension of visualisations
- revisiting material learned earlier and actively retrieving material from long-term memory in a low stakes environment can improve the durability of learning
- scaffolding (adjusting support to reflect the learners’ level of performance, such as using hints) can promote deeper learning.
Ninety middle schools in a state in the US were randomly assigned to different experimental conditions. In one of these conditions the curriculum was modified based on the above cognitive science principles, and teachers were given professional development in implementing the curriculum. The academic performance of students in this modified curriculum condition was then compared with that of students in a control condition (where no changes were made to usual education practices). In general, students who received instruction informed by cognitive science were found to perform slightly better than their peers who received ‘normal’ instruction. Even though further work is needed before any definitive conclusions can be drawn, basing teacher practice on cognitive science principles may be a promising strategy for improving learning.
An example of the misuse of cognitive science
The above examples indicate the potential power of cognitive science. However, when not interpreted or used appropriately, cognitive science research may not generate positive outcomes for education. One example of the problematic use of cognitive science in education comes from the widespread notion of ‘learning styles’. This notion suggests that there are different learning modes that we can identify and categorise (e.g., visual, auditory) and that individuals will learn better if taught in their preferred mode. This is a popularly held belief in education practice, and there are various hypotheses and theoretical models of learning styles published by researchers. (11) While most researchers have consistently debunked the concept of learning styles and numerous empirical studies have found no evidence that teaching to learning styles improves learning, (12) belief in and use of learning styles remains widespread amongst educational practitioners and programmes. (13) This is concerning because at best the use of learning styles might be irrelevant and a waste of teacher time, at worst there is evidence they may have a detrimental effect and pigeonhole students. (14) This highlights that cognitive science theories can sometimes be misapplied or used in practice before there is sufficient evidence to support their use.
So, what did we conclude?
Overall, cognitive science is an area of study that can usefully inform education policy and practice. However, for evidence from cognitive science to have the desired effects of maximising learning and improving teaching, it needs to be interpreted and used appropriately. In addition, its effectiveness in educational settings needs to be first tested, before it can form the basis of educational policy and practice. As the figure above shows, cognitive scientists and educators should collaborate to translate cognitive science findings into potential teaching practices. These potential practices should then be trialled in education settings to ascertain their effectiveness. If satisfactory evidence of their effectiveness is generated from the trials, they could then be applied more widely in education practice.
About the authors:
The authors are all members of the Cambridge Assessment Research Division, from left to right:
Jackie Greatorex, Principal Research Officer
Melissa Mouthaan, Research Officer
Tori Coleman, Research Officer
Filio Constantinou, Senior Research Officer
References:
1↩ Karpicke, J. D., and Blunt, J.R. (2011) Retrieval practice produces more learning than elaborative studying with concept mapping, Science, 331, 772-775. doi: 10.1126/science.1199327
2↩ O’Day, G. M., and Karpicke, J. D. (2020, 18 June). Comparing and combining retrieval practice and concept mapping. Journal of Educational Psychology, Advance online publication. doi: 10.1037/edu0000486
3↩ Moreira, B. F. T., Pinto, T. S. S., Starling, D. S. V., & Jaeger, A. (2019, 8 February). Retrieval practice in classroom settings: a review of applied research. Frontiers in Education, 4, 5. doi: 10.3389/feduc.2019.00005
4↩ Oates, T. (2021, 13 May) Assessment – perhaps it’s just about good questions.
5↩ Rittle-Johnson, B. Star, J.R. and Durkin, K. (2020) How can cognitive-science research help improve education? The case of comparing multiple strategies to improve mathematics learning and teaching, Current Directions in Psychological Science, 29(6), 599-609. doi:10.1177/0963721420969365
6↩ Gick, M. L., Holyoak, K. J. (1983). Schema induction and analogical transfer. Cognitive Psychology, 15, 1–38. doi:10.1016/0010-0285(83)90002-6 in Rittle-Johnson et al (2020)
7↩ Gentner, D., Loewenstein, J., Thompson, L. (2003). Learning and transfer: A general role for analogical encoding. Journal of Educational Psychology, 95, 393–405. doi:10.1037/0022-0663.95.2.393 in Rittle-Johnson et al (2020)
8↩ Schwartz, D. L., Bransford, J. D. (1998). A time for telling. Cognition and Instruction, 16, 475–522. doi:10.1207/s1532690xci1604_4 in Rittle-Johnson et al (2020)
9↩ Durkin, K., Rittle-Johnson, B. (2012). The effectiveness of using incorrect examples to support learning about decimal magnitude. Learning and Instruction, 22, 206–214. doi:10.1016/j.learninstruc.2011.11.001 in Rittle-Johnson et al (2020)
10↩ Yang, R., Porter, A. C., Massey, C. M., Merlino, J. F., & Desimone, L. M. (2019) Curriculum-based teacher professional development in middle school science: A comparison of training focused on cognitive science principles versus content knowledge, Journal of Research in Science Teaching, 57(4), 536-566. doi: 10.1002/tea.21605
11↩ Pashler, H., McDaniel, M., Rohrer, D., & Bjork, R. (2008). Learning styles: Concepts and evidence. Psychological science in the public interest, 9(3), 105-119. doi: 10.1111/j.1539-6053.2009.01038.x
12↩ Willingham, D. T., Hughes, E. M., & Dobolyi, D. G. (2015). The scientific status of learning styles theories. Teaching of Psychology, 42(3), 266-271. doi: 10.1177/0098628315589505
13↩ Newton, P. M., & Mahallad M. (2017, 27 March) Evidence-based higher education–is the learning styles ‘myth’ important? Frontiers in psychology, 8, 444. doi: 10.3389/fpsyg.2017.00444
14↩ Kirschner, P. A., & van Merriënboer, J. J. (2013). Do learners really know best? Urban legends in education. Educational psychologist, 48(3), 169-183. doi:10.1080/00461520.2013.804395