Table 1.
Selected papers focusing on the neural correlates of math anxiety.
| Study | Sample | Math anxiety | Groups | Tasks | Controlled variables | Results |
|---|---|---|---|---|---|---|
| fMRI studies | ||||||
| Lyons and Beilock (2012a) | 28 adults | sMARS | 20% (out of N = 108) | Single-digit arithmetic verification task (simple/ complex, subtraction/ multiplication); control word task; cues before task | Trait anxiety, working memory | High math-anxious individuals show with decreasing math deficits increasing differences in IFJ and IPL before the math task and in right NAc and left hippocampus during math task (increasing emotion regulation leads to compensation of math deficit) |
| Lyons and Beilock (2012b) | 28 adults | sMARS | ca. 50 % | Single-digit arithmetic verification task (simple/ complex, subtraction/ multiplication); control word task; cues before task | Trait anxiety, working memory | High math-anxious individuals show higher activation in bilateral dorso-posterior insula, MCC and right CSd (pain-related activity before math task) |
| Young et al. (2012) | 46 7- to 9-year-old children | SEMA | 50% | Single-digit arithmetic verification task (simple/ complex; addition/ subtraction); control number identification and passive fixation task | IQ, working memory, reading ability, math ability, trait anxiety | High math-anxious individuals show hyperactivity and abnormal effective connectivity of right amygdala extending into anterior hippocampus (processing negative emotions), less activation in IPS, right DLPFC, basal ganglia (less efficient task processing) and greater deactivation in VMPFC (emotion regulation) |
| Pletzer et al. (2015) | 36 adults | MARS30-brief | 25% (out of N = 127) | Two-digit number magnitude comparison task, number bisection task; control mental rotation task, control verbal reasoning task | Math ability | Low math-anxious individuals show moderately stronger deactivation within the task-related default mode network than high math-anxious individuals (less efficient processing), high math-anxious individuals did not activate left DLPFC, left inferior frontal gyrus and insula (task-irrelevant instead of task-relevant inhibitory control) |
| tDCS study | ||||||
| Sarkar et al. (2014) | 45 adults | Brief version of MARS | 25% (out of N = 165) | Single-digit arithmetic verification task with affective primes (addition, subtraction, multiplication, division); control flanker task | Age, gender | Stimulating the DLPFC leads to better math performance and decreased cortisol concentrations in high math-anxious individuals (less stress), but impaired math performance and prevented cortisol decrease in low math-anxious individuals |
| ERP studies | ||||||
| Suárez-Pellicioni et al. (2013a) | 26 adults | sMARS | 25% (out of N = 342) | Single-digit arithmetic verification task (split effect; addition) | Math ability, trait anxiety, spatial visualization, reasoning ability, verbal comprehension ability, gender distribution | High math-anxious individuals show an enhanced and delayed P600/3b for large-split solutions (difficulty in inhibiting processing of irrelevant information and less processing efficiency) |
| Suárez-Pellicioni et al. (2013b) | 34 adults | sMARS | 25% (out of N = 452) | Single-digit numerical Stroop task; control classical Stroop task | Trait and state anxiety, years of formal education, age, handedness, ethnicity, gender distribution | High math-anxious individuals show an enhanced ERN (abnormal error monitoring), but no difference for CRN or Pe (normal generic response monitoring processes) |
| Suárez-Pellicioni et al. (2014) | 34 adults | sMARS | 25% (out of N = 490) | Single-digit numerical Stroop task | Trait and state anxiety, simple math ability, years of formal education, age, handedness, ethnicity, gender distribution | High math-anxious individuals show a tendency for an enhanced conflict sustained potential (stimulus conflict processing), no enhanced N450 and greater conflict sustained potential amplitude in conflict adaptation (attentional control deficit and distractibility) |
| Núñez-Peña and Suárez-Pellicioni (2014) | 53 adults | sMARS | 25% (out of N = 629) | Single-digit number magnitude comparison task (size effect, distance effect) | Trait anxiety, age, years of formal education, gender distribution, handedness, ethnicity | High math-anxious individuals show an enhanced ERP distance and size effect (less precise number magnitude representation) |
The studies are sorted by method and include information on the analyzed sample size, the math anxiety questionnaire used, the cut-off value for a possible prescreening group for low vs. high math-anxious individuals, the numeric and the control task, controlled or matched variables between groups and the main results of the study.
ACC, anterior cingulate gyrus; CRN, correct-response negativity; CSd, dorsal segment of central sulcus; DLPFC, dorsolateral prefrontal cortex; ERN, error-related negativity; ERP, event-related potential; fMRI, functional magnetic resonance imaging; IFJ, inferior frontal junction; IPL, inferior parietal lobe; IPS, intraparietal sulcus; IQ, intelligence quotient; MCC, midcingulate cortex; MARS, math anxiety rating scale; NAc, nucleus accumbens; Pe, error-related positivity; SEMA, scale for early mathematics anxiety; tDCS, transcranial direct current stimulation; VMPFC, ventromedial prefrontal cortex.