Abstract:

The approximate number system (ANS) underlies the ability to approximately represent numerical magnitude or to estimate the number of a quantity without counting. ANS acuity is a reliable predictive factor of mathematical achievements and is most likely not influenced by educational or cultural factors of general cognitive processing. It is well established that individuals with mathematics anxiety (MA) perform normally in general cognitive tasks but significantly below-average in mathematical tasks such as numerical magnitude processing, computation and mathematical problem solving. Previous behavioral study reported significant negative correlations between MA level and ANS acuity but it remained unclear whether MA individuals are deficient in their ANS functioning.

The current study explored the approximate number processing in individuals with mathematics anxiety at a neural level and gathered electrophysiological evidences for ANS deficiency in MA individuals. We selected 31 individuals with high math anxiety and 29 individuals with low math anxiety who were matched for mean general intelligence, rapid visual perception ability, visual searching ability and mean general anxiety level. The participants completed a dot array task both actively and passively as their electroencephalogram (EEG) data were recorded. In the active dot array task, the participants were required to make judgements on the relative quantities of blue and yellow dots while the ratio of the dots were manipulated to be high or low. In the passive viewing task, the participants were required to complete an irrelevant task to ensure that proper attention was paid to the stimuli while the ratio was also manipulated.

We found that the behavioral performances of the two groups were comparable in both active and passive viewing tasks. However, event-related potential (ERP) analysis revealed that the high anxiety group had larger P2p amplitudes at the occipital electrodes than the low anxiety group in both active and passive viewing tasks. Furthermore, time-frequency analysis was performed and significant ratio effects were found in the low anxiety group in δ band (1~5 Hz) event-related synchronization (ERS) and β band (29~34 Hz) event-related desynchronization (ERD) while no ratios effects were found in the low anxiety group.

The current study provided electrophysiological evidences for ANS deficiencies in individuals with MA. The findings of this study have important theoretical implications for the causes of the impaired mathematical abilities in MA individuals as well as for effective interventions for MA.

Key words: mathematics anxiety, approximate number system, EEG, time-frequency analysis