The construct ‘mathematical anxiety’ (MA) has been under discussion for 60 years [Dowker et al., 2016]. It is known, that MA is positively correlated with trait anxiety (TА) [Wang et al., 2014], but the differences of... [ view full abstract ]
The construct ‘mathematical anxiety’ (MA) has been under discussion for 60 years [Dowker et al., 2016]. It is known, that MA is positively correlated with trait anxiety (TА) [Wang et al., 2014], but the differences of their effects on brain activity are not clear. This study examined cortical topography of brain activity changes relevant to MA and TA during error-recognition tasks. We assumed that MA and TA are different in their neuronal mechanisms.
The study included 52 healthy volunteers (age 24.6 ± 2.91, 1 left-handed). Levels of MA (MAL) were measured with sMARS test. To assess TA level (TAL), we used the STAI test. EEGs were recorded using 128 channels (127 EEG + VEOG) digitized at 1000 Hz. The EEG electrodes were placed according to 10–10 system and referred to Cz. Electrodes impedance was maintained below 20 kΩ. Participants judged whether or not one of the presented algebra, arithmetic and lexical tasks contained an error (90 tasks, 30 of each type). Event-related spectral perturbations (ERSP; Delorme and Makeig, 2004) were used to assess the spectral-power changes in brain activity. ERSP values were averaged across trials, time-frequency points and electrodes within respective regions separately for each participant and were used in repeated measures ANOVA.
The effects of MA and TA were found in theta band (4–8 Hz) in time 150–350 ms after task onset. The interaction between factors of MAL on sagittality, F(2,98)=3.42; p=0.045, was significant for correct tasks. The effect of MAL was significant in anterior (low MAL 0.80±0.21, high MAL 1.11±0.20, p<0.05) and posterior cortical areas (low MAL 1.07±0.25, high MAL 1.41±0.23, p<0.05), but non-significant in central cortex (low MAL 0.91±0.18, high MAL 0.91±0.17, p=0.69). The interaction between factors of TAL on laterality, F(2,98)=3.70; p=0.029 was significant for tasks with error. Significant differences (p<0.05) between people with low and high TAL were found in right hemisphere (low TAL: 0.86±0.17, high TAL: 1.20±0.19) and in medial cortex (low TAL: 1.07±0.19, high TAL 1.24±0.20), but were not found in left hemisphere (low TAL: 0.77±0.17, high TAL: 0.83±0.18).
We found that the effect of MA was strongly expressed in the anterior and posterior cortical regions, consistent with the cortical topography of the emotional-related reactions [Etkin et al., 2015]. At the same time, the TA level influenced the relationship between the left and right hemispheres. In concusion, the findings of this study give evidence that even if psychometric measures of TA and MA are correlated they have dissimilarities based on their neuronal mechanisms.
