Numerical cognition and math anxiety
➗ 🧠 😰 So now it seems like a real thing, hehehe.
Mathematics anxiety is defined as a feeling of tension, apprehension and even fear, ranging from mild discomfort to extreme avoidance, which interferes with the ordinary manipulation of numbers and the solving of math problems.
Math anxiety is highly prevalent and has detrimental consequences for learning and mastering mathematics.1 It is defined as feelings of tension that some individuals suffer in situations where they have to deal with numbers and can impair not only their academic achievement in math but also their performance in daily activities (e.g., calculating money for purchases or evaluating the economic conditions when applying for a loan). It is noteworthy that highly math–anxious (HMA) individuals usually avoid the science, technology, engineering, and mathematics (STEM) disciplines, so this type of anxiety negatively affects career choices and, consequently, professional success and economic incomes.
Key findings:
- Our findings also support the predictions of ACT (Attentional control theory) suggesting that math anxiety impairs executive cognitive control, specifically response inhibition in numerical tasks, a function that requires goal-driven attention allocation.
- Our results confirmed the hypothesis that math anxiety is related to a domain-specific behavioral inhibition deficit.
- In the current study, differences were found between tasks in both ERPs and hit rate for Nogo trials in the HMA but not in the LMA group. Regarding ERPs, our results confirmed those of previous studies, recording the frontal Nogo-N2 and Nogo-P3 Go & Nogo tasks for the LMA group in both tasks. This result suggests that this group inhibits a planned response in Nogo trials both in numerical and non-numerical tasks.
Behavioral performance and electroencephalogram activity were recorded while 28 highly math–anxious (HMA) and 28 low math–anxious (LMA) individuals performed both a numerical and a non-numerical Go/Nogo task. In the numerical task, single-digit numbers were presented, and participants were asked to press a button if the number was even. In the non-numerical task, letters were presented, and the button had to be pressed if the letter was a vowel.
Nogo trials were answered less accurately and elicited larger Nogo-N2 and Nogo-P3 than Go trials in both tasks and both groups. Importantly, behavioral and brain response differences between tasks were only found in the HMA group.
First, they were more error-prone in numerical Nogo than in non-numerical Nogo trials; and second, their Nogo-N2 and N2d (NogoGo difference) were smaller in the numerical task than in the non-numerical task. No differences were found in the LMA group. These results suggest that HMA individuals’ response inhibition is impaired specifically when dealing with numbers, which could contribute to their low achievement in math tasks.
Several explanations have been put forward to account for why math anxiety is negatively related to math achievement,3 with one of the most investigated claiming that HMA people might have a deficit in executive functions.
Executive functions (also called executive control or cognitive control) are top-down mental processes essential for all types of cognitive performance, since they allow individuals to solve problems, shift strategies flexibly, ignore distractors, inhibit irrelevant impulses, and monitor their actions. According to Miyake et al.,4 there are three core executive functions: inhibition, updating, and shifting. Inhibition is the ability to ignore dominant, automatic, or prepotent responses or information that are irrelevant to task processing.
It includes behavioral inhibition an overbearing response or an inappropriate action in a given context) and cognitive inhibition5,6 (i.e., interference control). Updating is the ability to take information into account or manipulate it and work with it, updating working memory representations. People have to encode new information and must decide what content should be removed from working memory.4 Last, shifting is the ability to shift attention between multiple tasks or change perspective during problem-solving.4 This ability requires the use of updating and inhibition functions,5 since to change the perspective, we must deactivate or inhibit the previous perspective and activate the new one in working memory.
Attentional control theory7,8 (Attentional control theory), developed from processing efficiency theory,9 accounts for the adverse effects that general anxiety may have on executive functions. According to ACT, anxiety decreases the efficiency of attentional control, with inhibition and shifting being the most impaired executive functions in threatening conditions. Moreover, ACT claims that anxiety affects processing efficiency (i.e., the use of cognitive resources to perform the task; usually measured by response time or neural activity) more than performance effectiveness (i.e., the ability to perform the task at a standard level; usually measured by response accuracy). So, under some circumstances, anxious people might recruit additional cognitive processes (i.e., they make more effort) to avoid performance impairments in tasks that require attentional control.
In the field of math anxiety, previous studies have shown that HMA individuals, compared with lowly math–anxious (LMA) people, show inefficient attentional control.11–13 HMA individuals showed worse interference control than their less math-anxious peers in a numerical Stroop task14 and in an emotional Stroop task when mathrelated words were presented15 (although other studies16,17 found no association between math anxiety and interference control in emotional Stroop tasks), so it has been suggested that they are more vulnerable to distraction in these tasks.
More recently, a less efficient shifting function in HMA individuals who had difficulties when switching between arithmetical operations (i.e., additions and subtractions) was reported.11 Therefore, to date, math anxiety has been related to impairments in the shifting and interference control functions of attentional control. However, it remains to be determined whether HMA individuals may also have impairments in behavioral inhibition, another executive function that is necessary to perform math tasks, which, according to ACT, might also be impaired by anxiety. The main objective of this study was to fill this gap by investigating whether HMA individuals may also suffer a deficit in the behavioral inhibition system. Response inhibition deficits have been found, for example, in children with attention-deficit/hyperactivity disorder (ADHD),18 which, though not considered a learning disability, undoubtedly makes learning difficult. An inefficient withholding of incorrect responses in HMA individuals could be because they adopt rigid and inflexible strategies when performing mathematics tasks, which could lead them to produce incorrect answers or be inefficient in math assessment situations.
To study behavioral inhibition (i.e., the ability to suppress inappropriate actions), one of the most used paradigms is the Go & Nogo tasks task.
In this task, participants are presented with a series of stimuli of frequent Go trials, in which they have to respond, and infrequent Nogo trials, in which they have to withhold their response. Go trials are more frequent than Nogo trials, usually at a ratio of 3:1, to create a prepotent tendency to respond that allows the measurement of behavioral response inhibition. s
The incidence of false alarms in this task (i.e., commission errors in Nogo trials) is the standard measure for behavioral inhibition: a higher rate of false alarms indicates that there are difficulties in motor inhibition.
Behavioral inhibition can also be studied by recording event-related brain potentials (ERPs) in Go/Nogo tasks. There are two EEG/ERP components elicited by Nogo trials at the frontocentral electrodes, known as Nogo-N2 and Nogo-P3, which are suggested to reflect brain activity associated with inhibitory control.19,20
Nogo-N2 is a frontocentral negative wave with a latency of around 200–300 ms post-stimulus, which is followed by Nogo-P3, a positive wave with a latency of around 300–500 ms after stimulus onset.
Nogo-N2 and the subsequent Nogo-P3 have been linked to attention and cognitive control: NogoN2 is thought to reflect the first stage of the inhibition of a planned response before the actual motor process and is related to conflict monitoring,20 whereas Nogo-P3 is thought to reflect the actual inhibition of the motor response in a later stage of the inhibition process21 or, alternatively, the monitoring of the outcome of inhibition22 (i.e., an evaluation process). Nogo-N2 amplitudes are larger in individuals with lower false alarm rates20 and smaller in children with ADHD, in contrast with the nonclinical population,18,23,24 so it has been suggested that Nogo-N2 is an index of successful response inhibition.20,25 Source-localization studies have found that Nogo-N2 and Nogo-P3 are generated in prefrontal areas, including the orbitofrontal cortex and the anterior cingulate cortex.26
Regarding math anxiety, it is important to understand not only whether it is associated with an impairment in the behavioral inhibition function but also, if so, to disentangle whether it is a domain-general or domain-specific deficit when dealing with numbers. Previous studies have found contradictory results concerning HMA deficits in cognitive inhibition: some found that they had worse interference control (i.e., worse cognitive inhibition) in tasks containing math-related stimuli,14,15 while others found they had impaired interference control in Stroop and flanker tasks that did not contain numerical or mathematical stimuli.32,33
The aim of this study was to examine whether HMA individuals suffer from a deficit in the behavioral inhibition system and to determine whether this is a specific numerical inhibition deficit or a general inhibition deficit, compared to their LMA counterparts. To address these questions, inhibition skills were assessed in HMA and LMA individuals while they performed a numerical and a non-numerical Go/Nogo task. Behavioral and ERP measures were analyzed, focusing on differences between tasks in both groups. Based on the previously discussed research, our predictions are as follows. If math-anxious individuals have a general inhibition deficit, we expected they would show an increase in error rate in Nogo trials (i.e., more false alarms) both in the numerical and the non-numerical tasks as compared to their less math-anxious peers; similarly, the HMA group should show a smaller Nogo-N2 than their less math-anxious counterparts in both tasks. However, if task differences were found in the HMA group (i.e., more false alarms in Nogo trials and less negative Nogo-N2 in the numerical than in the non-numerical task), this would indicate that math-anxious individuals have a specific numerical inhibition deficit.
Fifty-six healthy volunteers participated in the study, 28 of whom had low scores on the Shortened Mathematics Anxiety Rating Scale34 (SMARS), and 28 had high scores.
Fig. 1. Example of Go (left) and Nogo (right) trials in the numerical (top) and non-numerical (bottom) tasks.
Fig. 2. Raw grand average ERPs in Go and Nogo trials at Fz in the LMA (A) and the HMA (B) groups for the numerical (left) and non-numerical (right) tasks. Abbreviations: ERP, event-related brain potential; HMA, highly math-anxious; LMA, low math-anxious.
Fig. 3. (A) Difference waves (Nogo minus Go trials) at Fz in the numerical and non-numerical tasks for the LMA and HMA groups. (B) Topographic maps of N2d (250–350 ms) for the LMA and HMA groups in both tasks. Abbreviations: HMA, highly math-anxious; LMA, low math-anxious.
Our results confirmed the hypothesis that math anxiety is related to a domain-specific behavioral inhibition deficit.
In the current study, differences were found between tasks in both ERPs and hit rate for Nogo trials in the HMA but not in the LMA group. Regarding ERPs, our results confirmed those of previous studies,20 recording the frontal Nogo-N2 and Nogo-P3 for the LMA group in both tasks. This result suggests that this group inhibits a planned response in Nogo trials both in numerical and non-numerical tasks. Importantly, the present study revealed smaller Nogo-N2 (i.e., less negative amplitude) in the numerical than in the non-numerical task in the HMA group, but no task differences in the LMA group. In the same vein, the dN2 (Nogo–Go amplitude difference) was smaller in the HMA group than in the LMA group only in the numerical task.
The amplitude of Nogo-N2 has been associated with successful response inhibition,20 so, although both groups inhibited their planned response in Nogo trials in the non-numerical task, the HMA group might have exhibited worse inhibition of premature responses than their LMA peers in tasks involving numbers. It is noteworthy that, as outlined in our introduction, a reduced Nogo-N2 has also been found in ADHD as compared to nonclinical children,18,23,24 and that response inhibition deficits are prominent in children with this type of disorder.46
Interestingly, task differences emerged in the hit rate in Nogo trials (as a measure of response inhibition) only in the HMA group. HMA and LMA individuals were more error-prone in Nogo trials, where they needed to suppress the prepotent Go response, than in Go trials in both tasks, reproducing previous studies in the general population.
The LMA group showed no differences between tasks in terms of Nogo trial accuracy. These results again suggest inefficient domain-specific behavioral inhibition in the HMA group when they have to inhibit a prepotent response in a numerical task.
Our findings also support the predictions of ACT,8 suggesting that math anxiety impairs executive cognitive control, specifically response inhibition in numerical tasks, a function that requires goal-driven attention allocation.
source: https://nyaspubs.onlinelibrary.wiley.com/doi/10.1111/nyas.15216
http://www.ub.edu/brainlab/research-post/numerical-cognition-and-math-anxiety/