Educators, researchers, and the general public are not sure whether math anxiety is the result of poor performance or poor performance is the result of math anxiety. It is therefore important to explore the causal relationship between the two for planning instruction and remediation.

The anxiety-performance link has two possible causal directions. They have been extended into the specific field of mathematics anxiety. The **first** direction is explained by the ** Deficit Theory**. Mathematics performance deficits, for example on mathematics tests, generate mild to extreme mathematics anxiety, which may lead to higher anxiety in similar situations. For example, students who have not mastered non-negotiable skills with efficient strategies at their grade level (number concept—Kindergarten, additive reasoning by the end of second-grade, multiplicative reasoning by the end of fourth-grade, etc.)[1] attempt fewer problems on tests, thereby lowering their score. For example, if a student does not have the mastery of (a) multiplication tables, (b) divisibility rules, (c) short-division, and (d) prime factorization before they do operations on fractions, they will have difficulty in mastering them; they will, therefore, be afraid of proportional reasoning (fractions, decimals, percents, ratio, proportion, scale factor, etc.) and then algebraic operations. To turn lower performance into high-level of math anxiety requires time. But, in the case of a vulnerable child (e.g., learning disability, lower cognitive and executive functions, etc.), anxiety may take less time to manifest and may escalate quickly. Thus, children with mathematical learning disabilities are often found to have disproportionately higher levels of mathematics anxiety than typically developing children, supporting the Deficit Theory.

The **second** causal direction is that anxiety, particularly math anxiety, reduces mathematics performance by affecting *any or all of these processes*:

- the
*pre-processing*(initiating or responding to mathematics tasks—attitudinally and cognitively, negative predisposition for mathematics in general and particular mathematics), *processing*(making sense of the problem—linguistically, conceptually, and/or procedurally, connecting multiple presentations of the problem—data into table, graph, or diagram, etc.),*retrieval of information*(relevant prior knowledge—formulae, definition, equations, concept, or skills),*comprehending*(understanding the problem, making connections between the incoming information and prior information and knowledge, translating the words and expressions into mathematical expressions and equations, etc.), and*perseverance*(engaging and staying with the problem and showing interest in the outcome of the problem),

thereby reducing the level of performance. This is referred to as the ** Debilitating Anxiety Model**.

**The Deficit Theory Model of Anxiety
**In at least some cases, having especially poor mathematics performance in early childhood could elicit mathematics anxiety. This poor performance could be the result of environmental factors such as poor math teaching, lack of resources and experiences.

Studies of developmental dyscalculia and mathematical learning disabilities indicate that specific cases of mathematics anxiety are related to poor performance, but that poor performance could be attributed to these deficits and then the resultant math anxiety. However, only 6-8% of the population suffers from developmental dyscalculia and such findings cannot be generalized to the typically developing child. It should also be noted that cognitive resources are not the only possible deficits that could cause poor mathematics performance and math anxiety. For example, self-regulation (one of the components of executive function) deficits have been associated both with mathematics anxiety and decreased mathematics performance. The condition of *acquired dyscalculia* (e.g., children without learning disabilities who show gaps in their mathematics learning will fall in this category) is a clear example of poor mathematics performance.

It has been found that significant correlations exist between a student’s mathematics performance, both at elementary and adolescent age, in one year and their mathematics anxiety in the following year. These correlations are stronger than those found between a student’s mathematics anxiety in one year and their academic performance in the following year, indicating that mathematics performance may cause mathematics anxiety, thus providing support for the Deficit Theory.

**The Debilitating Anxiety Model
**Mathematics anxiety can impact performance at the stages of pre-processing, processing and retrieval of mathematics knowledge. Recent research suggests that anticipation of mathematics tasks causes activation of the neural ‘pain network’ in high math anxiety individuals, which may help to explain why high math anxiety individuals are inclined to avoid mathematics. In young children, task-avoidant behaviors have been found to reduce mathematics performance. Similarly, many adolescents with mathematics anxiety avoid math-related situations, suggesting that mathematics anxiety influences performance by reducing learning opportunities.

Adults with high mathematics anxiety answer mathematics questions less accurately but more quickly than those with lower levels showing that mathematics anxiety is associated with decreased cognitive reflection during mathematics word problems. Because of poor numerical skills, adults do not have resources to check their answers for correctness. Such data suggest that adults with mathematics anxiety may avoid processing mathematical problems altogether, which could lead both to reduced mathematics learning and to lower mathematics performance due to rushing, lack of engagement, and lack of comprehension. Adults with mathematics anxiety are less likely to enroll in college or university courses involving mathematics.

The worry induced by mathematics anxiety impairs mathematics performance during mathematics processing by taxing processing resources and minimizing their impact. Worry reduces working memory’s processing and storage capacity, thus reducing performance. For instance, research shows a negative correlation between college students’ math anxiety levels and their working memory span. Further, there is an interaction between adults’ mathematics anxiety and their performance on high and low working-memory load mathematics problems, with high working-memory load questions being more affected by mathematics anxiety. Thus, mathematics anxiety appears to affect performance by compromising the working-memory functions of those with high math anxiety.

Mathematics anxiety also affects strategy selection, leading individuals to choose simpler and less effective problem-solving strategies and thus impairing their performance on questions with a high working-memory load. This is supported by evidence suggesting that those with high working-memory, who usually use working-memory intensive strategies, are more impaired under pressure than those who tend to use simpler strategies.

Mathematics anxiety may manifest as (a) lack of willingness to engage in the activity because of previous negative impressions of mathematics, (b) poor reception and information processing, therefore disposing individuals to avoid mathematics related situations, (c) poor comprehension of mathematics information in mathematics learning tasks, thereby abandoning the tasks prematurely and giving up too easily, and (d) later, at the stages of processing and recall, mathematics anxiety may influence performance by cognitive interference. Math anxiety, thus, may negatively tax executive function resources, such as working memory, which are vital for the processing and retrieval of mathematical facts and methods. All of these affected behaviors impact the *Standards of Mathematics Practice (SMP) [2]* identified and recommended by the framers of

*Common Core State Standards in Mathematics (CCSS-M)*.

**[3]**On the other hand, positive emotions enhance learning by increasing the willingness to initiate tasks, develop persistence, use effective strategies and recruit cognitive resources. The idea that emotions have an effect on general achievement and particularly on math achievement is strongly supported by studies across all ages that manipulate anxiety to reveal either a decrement or improvement in performance. This effect of mathematics anxiety on performance is likely through executive function skills. This is particularly so in the case of working memory. The working memory functioning is impaired by the intrusive negative thoughts, negative talk, and poor self-esteem generated by math anxiety.

The mechanisms of influence of math anxiety, particularly cognitive interference, may be more immediate than from one academic year to the next. Since the effect of anxiety on recall would cause a fairly immediate performance decrement in those with high mathematics anxiety, this supports the debilitating anxiety model—the impact of math anxiety on performance.

To conclude, the evidence for the relationship between math anxiety and mathematics performance is mixed. Neither theory can fully explain the relationship observed between mathematics anxiety and mathematics performance. While some studies provide data, which fit the Deficit Theory, others provide more support for the Debilitating Anxiety Model. The mixture of evidence suggests a bidirectional relationship between mathematics anxiety and mathematics performance, in which poor performance can trigger mathematics anxiety in some individuals and mathematics anxiety can further reduce performance in a vicious cycle.

The belief about a causal relationship should prompt articulating educational policy, program planning in mathematics education, developing initial and then remedial mathematics instruction, assessment, particularly for those who suffer from math anxiety.

For example, if policy-makers share the belief that math anxiety is just another name for ‘bad at math,’ to reduce students’ math anxiety, effort and money will be targeted at courses to improve their mathematics performance. It will involve searching alternative teaching methods to mitigate this situation. In some cases, this may be (a) the development of computer-adaptive programs that may offer a way to ensure that students do not experience excessive failure in their math learning, (b) adjusting the difficulty level of mathematics tasks to an individual student’s ability, or (c) adapting remediation to student’s mathematics level and his/her mathematics learning personality[4].

If the relationship is in fact in the other direction, such efforts are likely to be ineffective and it would be better to focus on alleviating mathematics anxiety in order to improve mathematics performance. Then, it is important to understand the nature of classroom teaching that may produce math anxiety and focus on remediation of math anxiety. This will focus, particularly on methods, which may be undertaken in the mathematics classroom and during interventions. For example, writing about emotions prior to a math test has been seen to increase performance in those with high math anxiety. Because low mathematics self-concept is related to mathematics anxiety, when teaching, teachers should strengthen students’ academic self-concept, which has been identified as a factor related to academic performance.

The mechanisms proposed by the Deficit Theory are long-term, with the detrimental effect of poor performance on anxiety levels occurring over years. This may be why the Deficit Theory is often supported by

longitudinal studies. On the other hand, the Debilitating Anxiety Model, particularly cognitive interference, proposes some immediate mechanisms for anxiety’s interference with performance (e.g., taxing working memory resources). This could explain why the Debilitating Anxiety Model is best supported by experimental studies such as those investigating stereotype threats.

**The Reciprocal Theory
**The tension between the deficit and debilitating anxiety theories is indicative of the very nature of the mathematics anxiety-mathematics performance relationship. Whilst poor performance may trigger mathematics anxiety in certain individuals, mathematics anxiety lowers or further reduces the mathematics performance in others.

This relationship suggests a model in which mathematics anxiety can develop either from non-performance factors such as social, emotional, biological predisposition or from performance deficits. Mathematics anxiety may then cause further performance deficits, via avoidance and working-memory disruption, suggesting the bidirectional relationship of the ** Reciprocal Theory**. The question of whether the mathematics anxiety-mathematics performance relationship is in fact reciprocal is likely to be best answered by longitudinal studies across childhood and adolescence since only longitudinal data can determine whether mathematics anxiety or weak performance is first to develop.

Some data suggest that previous achievement may affect a student’s mathematics levels of performance and that mathematics anxiety in turn affects future performance, and further proposes indirect feedback loops from performance to appraisals and emotions.

Mathematics anxiety in adults may result from a deficit in basic numerical processing (poor number concept, poor numbersense, and lack of mastery in numeracy skills), which would be more in line with the Deficit Theory. For instance, adults with high mathematics anxiety have numerical processing deficits compared to adults with low mathematics anxiety. Mathematics anxiety may result from a basic low-level deficit in numerical processing that compromises the development of higher-level mathematical skills. Highly mathematics anxious adults’ basic numerical abilities are impaired because they have avoided mathematical tasks throughout their education and in adulthood due to their high levels of mathematics anxiety, supporting the Debilitating Anxiety Model.** **

**Genetics, Environment and Mathematics Performance
**Genetic studies may help to elucidate whether mathematics performance deficits do in fact emerge first and cause math anxiety to develop. One such study suggests that some (9%) of the total variance in mathematics performance stems from genes related to general anxiety, and 12% from genes related to mathematics cognition. This may indicate that for some, mathematics anxiety is caused by a genetic predisposition to deficits in mathematics cognition. However, this does not preclude the possibility that the relationship between mathematics anxiety and performance is reciprocal.

Parental (and other authority figures in a child’s life) math anxiety could be transmitted to children; in other words, parents likely play an important role, either positive or negative. In that case, it is more of social transmission of attitudes towards mathematics rather than genetic.

Sometimes, some of the genetic factors are translated into or affected by stereotypical reactions. Stereotype threats also elevate anxiety levels, thereby affecting participation in and processing of math activities. Stereotype threat is the situation in which members of a group are, or feel themselves to be, at risk of confirming a negative stereotype about their group. Under stereotype threat, individuals are seen to perform more poorly in a task than they do when not under this threat. It is posited that this is due to anxiety elicited by the potential to confirm or disconfirm a negative stereotype about one’s group. This particularly applies to some minority and women’s groups.

The effect of increasing anxiety by stereotype threat can be seen in adults as well as in children. For example, research shows that 6–7 year-old girls showed a performance decrement on a mathematics task after they completed a task that elicited stereotype threat. Similarly, it has been observed that presenting women with a female role model who doubted her own mathematics ability reduced their performance in mathematics problems compared with a control group who were presented with a confident female role model.

Deficits in mathematics performance in women under mathematics stereotype threat appear because math anxiety coupled with the stereotype affect the working memory. This phenomenon supports the idea that mathematics anxiety taxes the working-memory resources and that reduces mathematics performance. The same phenomenon is active when mathematics anxiety affects mathematics performance as the compounding of stereotypes based on race, income level, and gender.

[1] See *Non-negotiable skills in mathematics learning* in previous posts of this blog.

**[2]*** Visit earlier posts on SMP on this blog. *

**[3]*** Visit earlier posts on CCSS-M.*

[4] See *The Math Notebook* on *Mathematics Learning Personality* by Sharma (1989).