skip to primary navigationskip to content

Maths Learning Difficulties Can Be Related to Two Different Types of Working Memory Problems

For many years it has remained a mystery why certain people struggle greatly with mathematics even though they are motivated to learn it and have access to appropriate learning resources. Cognitive scientists have explored whether Mathematics Learning Disability (MLD) could be due to a brain region or function which has developed differently in people who struggle with maths. Most commonly, MLD has been linked to problems with working memory, i.e. the brain’s ability to hold and manipulate information over a short period of time.

Working memory was initially thought to be domain-general, meaning that it is the same regardless of whether the content is related to mathematics, reading, or some other subject. Some studies, though, have shown that working memory may be domain specific—different subtypes of working memory operate for different types of tasks. Dénes Szűcs, a neuroscientist at the University of Cambridge, combined the data from 32 commonly cited studies which reported on the working memory function of children with MLD. He then conducted a meta-analysis (an analysis of data obtained from multiple research studies) to examine the role working memory plays in mathematics performance.

Szűcs concluded that there may be more than just one cause for MLD by showing that the MLD participants could be divided into two groups in which each had a different type of working memory deficit. One of the two groups with MLD showed poor reading skills and scored poorly on verbal working memory tasks (tasks involving holding in memory and manipulating verbal information). The other group had purer deficits in mathematics and scored poorly on visuospatial working memory tasks (tasks involving holding and manipulating spatial and visual information). This seems to indicate that there could be at least two different causes of MLD, both of which are subtypes of working memory.

Szűcs concluded that mathematics requires an extensive network of brain activities and that a problem with any one of these could lead to MLD. Since verbal processing seems to be required for the brain to conduct mathematics, underdeveloped verbal processing could lead to MLD. In the same way, the brain also seems to use visuospatial processing for mathematics, so a deficit in this area could also lead to MLD.

Whilst the symptoms of MLD can look similar, the problem may arise from different sources. Since mathematics is such a rigorous (requiring a lot of mental resources) and exacting (requiring one perfect answer in many cases) discipline, it could reveal brain deficits that would not show up with other disciplines. This could be why participants who score poorly on visuospatial working memory tasks can do well in other subjects besides mathematics. The fact that MLD can arise because of multiple distinct deficits may be of great importance both to mathematics teachers and researchers investigating causes of MLD in children.

Szücs, D. (2016). Approximate number system training does not show any specific transfer benefits: A critical assessment of design, facts, bias and inference in the number sense training literature (in submission)