Using CANTAB to compare the effects of augmented reality N-back training and traditional two-dimensional N-back training on working memory

We recently caught up with Dr Bo Zhang from East China Normal University to discuss her research: A Comparison of the Effects of Augmented Reality N-Back Training and Traditional Two-Dimensional N-Back Training on Working Memory

Can you tell us a little about yourself?

I received my PhD from SMARTLab, University College Dublin in 2017. I have always been interested in using technology to make society more inclusive, for example, by using virtual worlds to help international students at university, or software-based education and training for people with intellectual disabilities or age-related cognitive decline. I currently work at East China Normal University, where I independently lead two projects: National Youth Project of National Education Sciences Planning and Shanghai Education Youth Project of Philosophy and Social Sciences Planning. Meantime, I am joining a Major Project of the National Social Science Foundation: at Department of Education Information Technology, Faculty of Education. Our team group is mainly focused on using Artificial Intelligence to promote the future development of education in schools, colleges and universities.

Tell us a little about the background and methods of your research.

While there has been a lot of research on the effectiveness of cognitive training software, there is still debate about whether it works. One issue is that there are many kinds of cognitive training software, and they all have different characteristics. In recent work, Nigel Robb (co-author of the study) and I have been focusing on the specific characteristics of games and other software that might contribute to specific cognitive effects, for example, whether they have 2D or 3D graphics. In the current study, we wanted to investigate how adding markerless augmented reality to a cognitive training program would affect the outcomes. To test this, we developed two versions of an N-back task, which differed only in the graphical presentation. One was a traditional, 2D N-back, presented from a top-down perspective on the screen (the control version). The other version presented the N-back test using augmented reality, so that the test was viewed in 3D, as if projected into the surroundings (the experimental version). We had two different groups, one using each version of the software. Both groups were tested using CANTAB Spatial Working Memory (SWM) before and after the N-back training. We compared their improvement on the SWM after training.

What were your key findings and implications of your study?

Our main finding is that there may be a trade-off in outcomes when specific characteristics of a training program are altered. On the easiest levels of the SWM, participants who trained on the traditional 2D N-back showed slightly more improvement that those who trained on the AR version. On the other hand, the AR version was rated as more engaging by participants, which could affect how long they would persevere with a training program. Our results may also support the Common Demands Theory of training transfer, which suggest that improvements on cognitive tests after training depend on similarities between the training and the tests. The presentation of the 2D N-back was more like the SWM than the AR version, which might explain why participants in the 2D training group showed more improvement on the easier levels.

Why did you choose CANTAB for your study?

We chose the CANTAB SWM because it has been used widely in previous research, there are multiple levels of difficulty, and it is very straightforward to use, both in terms of administering the test and collecting and analyzing the data.

Do you have any future research planned?

Completed the tests with healthy people groups, I am currently continuing to explore the application of markerless AR cognitive training to help children with autism to improve their cognitive abilities in China, while Nigel is continuing research on how altering specific characteristics of training programs using control and experimental version of the same software can shed more light on how these characteristics contribute to training effects. We hope to be able to combine these two research strands and collaborate further in the future.

Finally, 2020/21 has been a difficult time for everyone, with many tough challenges in our lives, study, and work. But I think that we can always find a way to use technology to persevere and keep trying to make the world a better, more inclusive place.

Bo Zhang - East China Normal University