Research Article: Worrying Thoughts Limit Working Memory Capacity in Math Anxiety

Date Published: October 27, 2016

Publisher: Public Library of Science

Author(s): Zhan Shi, Peiru Liu, Michael A Motes.


Sixty-one high-math-anxious persons and sixty-one low-math-anxious persons completed a modified working memory capacity task, designed to measure working memory capacity under a dysfunctional math-related context and working memory capacity under a valence-neutral context. Participants were required to perform simple tasks with emotionally benign material (i.e., lists of letters) over short intervals while simultaneously reading and making judgments about sentences describing dysfunctional math-related thoughts or sentences describing emotionally-neutral facts about the world. Working memory capacity for letters under the dysfunctional math-related context, relative to working memory capacity performance under the valence-neutral context, was poorer overall in the high-math-anxious group compared with the low-math-anxious group. The findings show a particular difficulty employing working memory in math-related contexts in high-math-anxious participants. Theories that can provide reasonable interpretations for these findings and interventions that can reduce anxiety-induced worrying intrusive thoughts or improve working memory capacity for math anxiety are discussed.

Partial Text

Mathematics anxiety (MA) is defined as a feeling of tension and anxiety that interferes with the manipulation of numbers and the solving of mathematical problems in a wide variety of ordinary life and academic situations [1]. Apart from enjoying mathematics less, having lower perceptions of their mathematical abilities, and ignoring the value of mathematics in everyday life [2–4], people with MA are characterized by increased worries about math failure [5], an avoidance of math and/or numerical tasks [6], and even a negative emotional response to the prospect of doing math [4]. Despite normal performance in most thinking and reasoning tasks, individuals with MA perform poorly when numerical information is involved [3, 4, 7, 8]. Specifically, empirical evidence has been found for performance differences as a function of MA. These differences typically are not observed when engaged in simple addition or multiplication but are observed when engaged in more difficult arithmetic problems involving carry operation [9, 10]. Executing the carry operation is thought to be controlled by working memory (WM) [11, 12], which is defined as a limited resource cognitive system used to actively maintain information in the face of ongoing processing and/or distraction [13–17]. A large number of theoretical and empirical studies have thus developed a generally accepted interpretation of the effect of MA on mathematics, which claims that MA detrimentally impacts mathematical performance by disrupting WM resources otherwise devoted to skill execution [18–22]. For example, in a seminal study investigating the relationship between MA and WM, high-math-anxious (HMA) and low-math-anxious (LMA) participants were asked to perform two-column addition problems requiring a carry operation, which placed a load on WM [18]. Participants performed these problems while completing a secondary letter memory task involving the maintenance of either 2-letter strings or 6-letter strings in memory. Math error rates under the 2-letter load for HMA participants were only slightly larger than LMA participants. However, HMA participants produced significantly more math errors under the 6-letter load than for LMA participants. Based on processing efficiency theory (PET) [20], Ashcraft and Kirk [18] concluded that the high letter load was detrimental to HMA participants’ WM performance because their worries about math disrupted WM resources that might otherwise have been available to perform the difficult math problems. However, as admitted by Ashcraft and Kirk [18], they “do not test the specifics of Eysenck and Calvo’s prediction [20] here, which states that it is intrusive thoughts and worry (in this case, about math) that detract from available working memory capacity (WMC).” Instead, they “assess the more general prediction that MA disrupts WM processing when the cognitive task involves arithmetic or math-related processes” (p. 226) [18].

The aim of this study was to examine WMC in a math-related context as well as WMC in a valence-neutral context in MA by developing a modified RS task. Consistent with the prediction, the results demonstrated that, on one hand, HMA participants performed worse on the WMC task under the context of math-related sentences (i.e., remembering the letter lists in the context of math-related sentences) than LMA participants; on the other hand, there was no group (HMA versus LMA) difference in performance on the WMC task under the context of valence-neutral sentences (remembering the letter lists in the context of valence-neutral sentences). These findings provide evidence that MA, like test anxiety [26], is characterized by impaired WMC under the context of sentences describing dysfunctional math-related thoughts. Moreover, the present study found more detailed results—WM load effects, that is, impaired WMC in MA emerged only in high WM load trials (i.e., size 4 and 5) but not in low WM load trials (size 3). This result indicated that the effects of MA on WMC were not all-or-none effects but might be accumulated effects. Additionally, no effects of MA on reaction time and accuracy were discovered in the present study.