The Relationship between Working Memory and Arithmetic in Primary School Children: A Meta-Analysis

Brain Active Areas Associated with a Mental Arithmetic Task: An eLORETA Study

The neural underpinnings of mental calculation, the fundamentals of arithmetic representations and processes, and the development of arithmetic abilities have been explored by researchers over the years. In the present work, we report a study that analyzes the brain-activated areas of a group of 35 healthy subjects (9 males, 26 females, mean age ± SD = 18.23 ± 2.20 years) who performed a serial subtraction arithmetic task. In contrast to most of the studies in the literature based on fMRI, we performed the brain active source reconstruction starting from EEG signals by means of the eLORETA method. In particular, the subjects were classified as bad counters or good counters, according to the results of the task, and the brain activity of the two groups was compared. The results were statistically significant only in the beta band, revealing that the left limbic lobe was found to be more active in people showing better performance. The limbic lobe is involved in visuospatial processing, memory, arithmetic fact retrieval, and emotions. However, the role of the limbic lobe in mental arithmetic has been barely explored, so these interesting findings could represent a starting point for future in-depth analyses. Since there is evidence in the literature that the motor system is affected by the execution of arithmetic tasks, a more extensive knowledge of the brain activation associated with arithmetic tasks could be exploited not only for the assessment of mathematical skills but also in the evaluation of motor impairments and, consequently, in rehabilitation for motor disorders.

The componential nature of arithmetical cognition: some important questions

Research on typically developing children and adults and people with developmental and acquired dyscalculia converges in indicating that arithmetical ability is not unitary but is made up of many different components. Categories of components include non-symbolic quantity representation and processing; symbolic quantity representation and processing; counting procedures and principles; arithmetic operations; arithmetical knowledge and understanding; multiple forms and applications of conceptual knowledge of arithmetic; and domain-general abilities such as attention, executive functions and working memory. There is much evidence that different components can and often do show considerable functional independence, not only in developmental and acquired dyscalculia, but in typically achieving children and adults. At the same time, it is possible to find complex interactions and bidirectional relationships between the different components, including between domain-specific and apparently domain-general abilities. There is a great deal that still needs to be discovered. In particular, we need to learn more about the origins in infancy of subitizing and approximate magnitude comparison, the extent to which these interact, the extent to which they may be further divisible, and the extent and ways in which they themselves may develop with age and the extent to which they may influence later-developing components. There also needs to be a lot more research on exactly how domain-general and domain-specific abilities contribute to mathematical development, and how they interact with one another.

The Structure of Working Memory and Its Relationship with Intelligence in Japanese Children

There is a host of research on the structure of working memory (WM) and its relationship with intelligence in adults, but only a few studies have involved children. In this paper, several different WM models were tested on 170 Japanese school children (from 7 years and 5 months to 11 years and 6 months). Results showed that a model distinguishing between modalities (i.e., verbal and spatial WM) fitted the data well and was therefore selected. Notably, a bi-factor model distinguishing between modalities, but also including a common WM factor, presented with a very good fit, but was less parsimonious. Subsequently, we tested the predictive power of the verbal and spatial WM factors on fluid and crystallized intelligence. Results indicated that the shared contribution of WM explained the largest portion of variance of fluid intelligence, with verbal and spatial WM independently explaining a residual portion of the variance. Concerning crystallized intelligence, however, verbal WM explained the largest portion of the variance, with the joint contribution of verbal and spatial WM explaining the residual part. The distinction between verbal and spatial WM could be important in clinical settings (e.g., children with atypical development might struggle selectively on some WM components) and in school settings (e.g., verbal and spatial WM might be differently implicated in mathematical achievement).