| Previous studies have examined the relative contribution of domain-general and domain-specific mechanisms supporting the development of mathematics cognition, yielding mixed results. Two etiological origins have been proposed for children struggling with mathematics, the first resulting from a neurobiological deficit in the parietal region of the brain, leading to a pure mathematical disability, specific to mathematical processing; the second, resulting from a deficit in domain-general cognitive mechanisms related to working memory, visual-spatial processing, and attentional control. The precise etiology has yet to be determined. It is clear, however, that children who experience mathematical difficulties (MD) use less developmentally mature strategies and commit more errors than their typically achieving (TA) peers and are less proficient at recognizing when an error has occurred. This study investigated the neural correlates of error detection capabilities in arithmetic problems using a combined behavioral and functional Magnetic Resonance Imaging (fMRI) design. Error detection abilities were examined in a group of 21 adolescents, 7 of whom were identified as MD. Participants engaged in a novel Error Detection task, which consisted of 40 addition, subtraction, and multiplication problems presented along with a proposed solution. All problems were presented in the center of the visual field (50% vertical). Participants identified if the solution was correct, indicating Yes or No, via button press (e.g., 5 + 9 = 14; 3 x 8 = 32). As this task is novel to neuroimaging studies of mathematics, it is positioned to provide preliminary data regarding behavioral performance and patterns of neural activation and deactivation related to error detection abilities in mathematics. The results indicate that the TA group significantly ( Z = 2.72, p = 0.006) outperformed the MD group on task accuracy. While no differences in activation patterns were found between MD and TA participants, the TA group had significantly more deactivation in the default mode network (DMN) as compared to the MD group, most notably in the amygdala. Failure to suppress DMN activation during cognitively demanding tasks (e.g., mathematical calculations) may leave a child susceptible to interference from extraneous, non-task related activation; thereby, increasing the potential for lower accuracy scores and less efficient neural processing. Although participants in the current study are still undergoing developmental changes, these results suggest that the differences between TA and MD adolescents' ability to detect mathematical errors may be related to differences in the DMN and that at least in the case of error detection, domain-general cognitive processing deficits may contribute to an inability to detect mathematical errors. These results have implications for both educational programming and intervention practices. By understanding the neural networks supporting arithmetic processing, educators, intervention specialists, and administrators will be better equipped to make curricular and programmatic decisions that address the deficits encountered by children who struggle with mathematics learning. |
