Mathematics anxiety and mathematics performance

Executive summary

• Not all mathematics difficulties stem from cognitive problems.
• Many pupils and adults have a debilitating emotional reaction to any situation involving mathematics: mathematics anxiety.
• Mathematics anxiety is distinct from other forms of anxiety.
• Mathematics anxiety negatively correlates with mathematics achievement.
• Mathematics anxiety likely disrupts working memory processes crucial for mathematics.
• Mathematics anxiety may be a component of a vicious circle, anxiety leading to worse performance and performance decrement justifying anxiety.

 Mathematics anxiety: Overview

Mathematics is often perceived as a difficult subject by students, parents, and teachers alike. Difficulties are most often attributed to cognitive reasons (lack of ability, preparedness, practice, and knowledge). In contrast, emotional factors are rarely considered or are easily written off as potentially persistent causes of mathematical difficulties. However, it is increasingly recognized in psychology and education that many students have serious emotional problems with mathematics that inhibit their performance and progress in this domain. In particular, many students have a debilitating emotional reaction to mathematics termed ‘mathematics anxiety’ (MA).

Anxiety can be defined as the “feeling of uncertainty and helplessness in the face of danger”¹. MA has been defined as “a feeling of tension and anxiety that interferes with the manipulation of numbers and the solving of mathematical problems in…ordinary life and academic situations”^2-3. MA ranges from the feeling of mild tension to experiencing strong fear of mathematics. MA can manifest itself in different ways, for example, as feelings of apprehension, dislike, tension, worry, frustration, and fear^4-6.

MA as a specific anxiety is likely to be an acquired, socially constructed anxiety that is not restricted to test situations or classroom settings. Hence, while MA is typically detected in school environments, it can also appear in any situations involving mathematics (e.g., when handling numbers in shops or needing to solve calculations in everyday life). That is, MA can generalize to various situations. An important consequence of MA may be that otherwise perfectly intelligent and mathematically capable persons develop a severe avoidance of situations involving any kind of mathematics and do not choose careers involving the application of mathematics⁷.

MA disproportionally affects girls and, therefore, may contribute to the worldwide observable gender gap in science, technology, engineering, and mathematics (STEM) subjects. In addition, MA may also contribute to girls’ stereotypical career choices and career aspirations^8-9, contrary to the fact that they often show equal or better mathematical performance than boys¹⁰.

Measurement of MA: Questionnaires

Academic anxieties such as MA are not considered clinical anxiety disorders. Similarly, academic anxieties are currently not recognized in the Diagnostic Statistical Manual of Mental Disorders (DSM–5; American Psychiatric Association, 2013), nor in the International Classification of Diseases (ICD–10; World Health Organisation, 1992). In fact, in clinical practice, questionnaires alone cannot be used to diagnose specific phobias (American Psychiatric Association, 2013; World Health Organisation, 1992). Nevertheless, questionnaires are used extensively in educational and psychological research for identifying MA and test anxiety in children and adults.

The Maths Anxiety Rating Scale (MARS) is a frequently used adult MA questionnaire¹¹. The so-called Abbreviated Maths Anxiety Scale, abbreviated AMAS, a 9-item questionnaire, has been developed based on MARS¹². A thorough overview⁴ suggested that AMAS could be a major tool for adult MA research as it is brief and delivers results consistent with the much longer MARS. The AMAS has been translated into various languages. There have been reliable and valid Persian, Italian, and Polish translations of the AMAS^13-15.

Importantly, the AMAS contains questions that are not suitable for primary school children. For example, some questions refer to looking up information in tables in the back of mathematics books. However, such tables are not typical in primary school mathematics books. Yet, some other alternative questionnaires¹⁶ may include too specific questions (e.g., about specific addition results such as 13 + 6). However, such questions are again not very suitable for reliably determining MA. First, younger children may find them more difficult than older children. Second, mathematically underperforming children may find such concrete questions more anxiety-inducing than more advanced students. However, some of the more advanced students may actually feel substantial MA when encountering more difficult problems, whereas other advanced students may not feel much MA. Taking into account these considerations, it is imperative to develop a reliable (consistently performing from time to time) and valid (measuring a well-defined construct) instrument for measuring MA in young children¹⁶.

Carey et al.¹⁶ have developed the modified AMAS (mAMAS) questionnaire suitable for primary and secondary school children. The study used data from 1,746 British children and found that the mAMAS is a good measure of MA in children. The mAMAS has two subscales, Learning MA and Evaluation MA¹². Both scales can be considered to be part of the same MA construct. The mAMAS is a freely available open access resource published in the Supplementary Material of Ref. ¹⁶.

Notably, while the AMAS and the mAMAS differentiate between the above-noted Learning and Evaluation subscales, other MA scales may define different subcomponents. For example, some researchers separate MA elicited by testing situations from different types of MA (e.g., anxiety for doing arithmetic, using mathematics in everyday life, or manipulating numbers)¹⁷. Others¹⁸ supposed that MA may have two dimensions: (1) a cognitive component (usually labelled as “worry”) that concerns worries about performance, and (2) an affective component (“emotionality”), which refers to tension, nervousness, and related physiological reactions felt in evaluative settings¹⁹.

Measurement of MA: Neuroscience and implicit measures

A general problem with all kinds of questionnaires is that they are based on self-report. Hence, questionnaire results may be affected by metacognitive processes, self-knowledge, and cultural biases. It is an important goal of educational neuroscience to develop objective measures of MA. Such measures could be based on physiological parameters such as cortisol levels, functional magnetic resonance imaging (fMRI) data, electro-physiological or electro-myographic indices, and eye-tracking parameters. Alternatively, implicit psychological measures (that do not require overt, conscious deliberation on behalf of pupils) could also be used in principle, such as implicit priming measures²⁰. At the moment, the practical applicability of physiological and priming measures is not yet clear. More research is needed to determine their validity and reliability.

Prevalence of MA

Prevalence estimates of MA depend on the definition of high MA levels based on questionnaire measurements. Inevitably, all such definitions are subjective and are based on researcher judgement. Importantly, MA researchers use discrepant definitions of what constitutes high MA. Some have defined high MA as achieving MA scores higher than +1 standard deviation (SD) above the mean²¹. This definition of high MA assumes that MA scores are normally distributed. Hence, a cutoff of +1 SD above the mean would imply that about 17% of children would have high MA. However, the distribution of MA scores is typically not reported. In fact, recent large population investigations of MA suggest that MA scores are not normally distributed¹⁶. Therefore, the +1 SD definition of high MA is questionable.

A recent large investigation¹⁶ determined the distribution of MA scores in the United Kingdom in more than 1,700 children aged 8 to 9 years (primary school) and 12 to 13 years (secondary school). The study used the above-discussed mAMAS questionnaire. The distribution was non-normal (see Figure 1 in Ref. ¹⁶). Hence, arbitrarily, researchers considered children to have high MA if they had a score higher than the 90^th percentile of the score distribution. In other words, this means that the top 10% most math-anxious children were considered to have high MA. This definition meant that in order to be classified as highly math-anxious, children had to achieve a score of 30 or higher out of a maximum score of 45 on the mAMAS. Notably, this 30 cutoff score for high MA definition entirely depends on the empirically observed data distribution. Hence, this cutoff score can be considered empirically validated.

MA is distinct from other forms of anxiety

MA is positively correlated with so-called test anxiety, elicited by testing situations ([22]; r = 0.52). However, a major review found that 63% variance in MA cannot be explained by test anxiety²² and MA is typically considered a distinct construct from test anxiety⁴. Moreover, as MA is one of the most prevalent academic anxieties in childhood, many test anxiety questionnaires may actually measure MA primarily. This is because when asked about their own academic anxieties, many children may primarily think about their negative experiences and anxiety in mathematics classes. Hence, a correlation between MA and test anxiety does not allow us to conclude which anxiety form is “primary.”

Another anxiety form that has been shown to relate to MA is general anxiety. General anxiety refers to an individual’s tendency to feel anxious about everyday situations. General anxiety often has factors of physiological anxiety, worry, and social anxiety (see, for example, the Revised Children’s Manifest Anxiety Scale²³). A large study of 800 Italian children found substantial shared variance between general anxiety and MA²⁴. Nevertheless, this correlation was less strong than the correlation between test anxiety and MA.

One large recent investigation in the United Kingdom aimed to differentiate between various anxiety forms and their role in relation to academic performance²⁵. The study found that younger (8- to 9-year-old) children’s MA, test anxiety, and general anxiety levels were very similar. That is, MA tended to appear strongly linked to other forms of anxiety. Put otherwise, children who were generally anxious were also likely to experience high levels of test anxiety and MA. In contrast, older (11- to 13-year-old) children showed more specific forms of anxiety with more specific impact on academic performance. That is, some children may have been generally anxious, but others now showed only specific MA or anxiety to testing situations. Moreover, the generally more anxious children had better math performance than the children with specific MA. Hence, there seems to be a developmental change in the pattern of various anxieties during the school system: It seems that specific academic anxieties become more and more pronounced during schooling and that specific academic anxieties become linked to poorer academic performance more than general, nonacademic anxieties.

MA has a negative correlation with mathematics performance

Meta-analyses of many studies have found that MA has a moderate correlation of about r = -0.27 to r=-0.34 with mathematics performance^22,5. Note that the negative sign indicates that high MA scores are associated with lower mathematics performance. These estimates are in perfect agreement with a recent large population study that measured r = -0.3 correlation (95% confidence interval = -0.34 to -0.25) between MA and standardized math achievement scores in more than 1,700 primary and secondary school pupils in the United Kingdom²⁶. These correlations are relatively low, but they do indicate a consistent relationship.

There is evidence that MA can already negatively affect performance in 5- to 7-year-olds²⁷. However, a recent study with more than 800 Italian children observed lower effect sizes of r = -0.13 and r = -0.07 correlations between MA and math performance in primary school girls and boys respectively²⁴. In contrast, correlations were comparable to the above-noted effect sizes (r = -0.34 and r = -0.28) in secondary school girls and boys. These observations appear to be in line with the above-discussed proposal that MA and academic anxieties become more linked to academic performance during schooling. That is, at least in some countries, MA and negative attitudes toward mathematics may increase during schooling, may reach adult levels at the secondary school level, and then persist into higher education and adulthood [see Ref. ²⁵ for extended discussion].

While girls show overall higher levels of MA than boys (see author’s brief, “The origins of math anxiety and interventions“), the amount of correlation is the same in both genders (that is, the slopes of bivariate MA versus math performance fitted lines are the same in both genders).

Theories of the negative impact of math anxiety on math performance

According to the so-called deficit theory, MA is the consequence of poor math performance²⁸. That is, children would expect themselves to perform poorly in mathematics tests and would therefore predict/expect poor performance on these tests. This prediction would then elicit anxiety and act as a self-fulfilling prophecy. Importantly, according to this view, only children with poor math performance would develop MA.

According to the so-called interference theory of test anxiety²⁹, people in a situation would experience both task-directed drives and anxiety drives. Task-directed drives would help in fulfilling task requirements and completing required jobs. In contrast, anxiety drives may result in two different behaviours. First, if they are not overwhelming, they can trigger task-relevant efforts that can result in reduced anxiety levels and trigger task-relevant drives/behaviour. Second, if anxiety drives become overwhelming, they can trigger self-directed, task-irrelevant behaviour that interferes with task performance. Pupils can experience heightened heartbeat and may anticipate punishment, loss of status, or loss of self-esteem. They may also experience a strong desire to escape the negative situation. If the anxiety drives become very strong, pupils may develop a habit of reenacting task-irrelevant behaviours, which may result in long-term performance decrement.

The debilitating MA model builds on the above interference model and assumes that relative low performance in mathematics is a consequence of high MA rather than vice versa³⁰. Basically, pupils would experience anxiety which would trigger task-irrelevant behaviours consuming working memory resources, which in turn would decrease mathematics performance. Put simply, children would think about many task-irrelevant potential negative outcomes (e.g., failing the test, being unable to solve the test, others performing much better than them, being punished at home for bad marks) and these irrelevant thoughts would then heavily compromise working memory performance. As discussed in Brief 1, working memory is a vital mental space where mathematics operations can be carried out. Hence, it is feasible to assume that anxiety consuming working memory resources may be able to cause performance decrements in mathematics^31-32.

Evidence suggests that MA negatively affects performance primarily on more difficult than easier mathematics problems ³². For example, Ashcraft and Krause gave mathematics problems to adults with varying levels of MA. People with the highest levels of MA had lower performance on medium-hard and hard problems than people with low levels of MA. The findings are consistent with the working memory interference explanation of MA because higher levels of working memory performance would be needed to solve more difficult than easier problems.

Importantly, the deficit theory and the debilitating anxiety theory of MA are not mutually exclusive²⁸. They may both be valid factors: 1) being valid in different groups of pupils, or 2) being valid at the same time as mutually contributing factors to amplifying MA.

First, it was suggested that some pupils may have generally high anxiety levels to start with and may consequently also be predisposed to develop high MA irrespective of their math performance²⁵. That is, in this group of pupils, the debilitating anxiety model may be more valid. In contrast, other pupils seemed to have only specific academic anxieties, and they also showed the lowest relative levels of mathematics performance. That is, in this group, the deficit theory may be more valid.

Figure 1. Reciprocal theory. A vicious circle between debilitating MA and deficit-induced anxieties. A major question is whether weak mathematics performance or anxiety may come first in starting the vicious circle. Different factors may be primary in different groups of children.

Second, the so-called reciprocal theory²⁸ considers the option of both debilitating anxiety and deficit-induced anxieties being in play at the same time. Figure 1 demonstrates a potentially vicious circle of debilitating anxiety and deficit-induced anxiety amplifying each other during development. On the one hand, lower ability children may experience anxiety due to their low perceived performance on math tasks. This perception may then induce debilitating anxiety in subsequent tests. This debilitating anxiety would then result in further performance decrements. On the other hand, normally achieving children would first experience debilitating anxiety. This anxiety could result in performance decrement in mathematics test that would induce (perceived) deficit-related anxieties. These could then further justify pupils’ debilitating anxieties as well.

In the medium term, such anxieties would keep children away from home-based, extracurricular, and optional math education opportunities that would cause further relative (to others) performance decrements. Hence, a vicious circle between debilitating and deficit-induced anxieties may be in play in a portion of children.

More MA studies are needed worldwide

Most studies of MA have been carried out in the USA and United Kingdom^22,4. Worldwide coverage of research is lacking, especially in developing countries. However, the existing studies suggest that MA is as important, if not a more important factor, in mathematics learning and gender discrepancies in learning in developing countries^33-34.

References

1. May, R. The Meaning of Anxiety. Revised ed. (W. W. Norton & Company, 1977).
2. Ashcraft, M. H. Math anxiety: Personal, educational, and cognitive consequences. Current Directions in Psychological Science 11, 181–185. doi:10.1111/1467-8721.00196 (2002).
3. Richardson, F. C., & Suinn, R. M. The mathematics anxiety rating scale: psychometric data. Journal of Counseling Psychology 19, 551–554 (1972).
4. Ashcraft, M. H., & Ridley, K. S. Math anxiety and its cognitive consequences: A tutorial review. In Handbook of Mathematical Cognition (ed. Campbell, J. I. D.) 315– 327. (New York, NY: Psychology Press, 2005).
5. Ma, X., & Xu, J. The causal ordering of mathematics anxiety and mathematics achievement: a longitudinal panel analysis. Journal of Adolescence 27, 165–79 (2004).
6. Wigfield, A., & Meece, J. L. Math anxiety in elementary and secondary school students. Journal of Educational Psychology 80, 210–216 (1988).
7. Ma, X. A meta-analysis of the relationship between anxiety toward mathematics and achievement in mathematics. Journal for Research in Mathematics Education 30, 520– 540 (1999).
8. Arnot, M., David, M., & Weiner, G. Closing the Gender Gap Post-War Education and Social Change. (Wiley, 1999).
9. Collins, C., Kenway, J., & McLeod, J. Gender debates we still have to have. The Australian Educational Researcher 27(3), 37-48 (2000).
10. Devine, A., Fawcett, K., Szűcs, D., & Dowker, A. Gender differences in mathematics anxiety and the relation to mathematics performance while controlling for test anxiety. Behavioral and Brain Functions 8:33 (2012).
11. Suinn, R.M., Taylor, S., & Edwards, R.W. Suinn mathematics anxiety rating scale for elementary school students (MARS-E): psychometric and normative data. Educ Psychol Meas 48:979-986 (1988).
12. Hopko, D.R., Mahadevan, R., Bare, R.L., & Hunt, M.K. The abbreviated math anxiety scale (AMAS): construction, validity, and reliability. Assessment 10:178-182 (2003).
13. Vahedi, S., & Farrokhi, F. A confirmatory factor analysis of the structure of abbreviated math anxiety scale. Iran J Psychiatry. 6, 47-53 (2011).
14. Primi, C., Busdraghi, C., Tomasetto, C., Morsanyi, K., & Chiesi, F. Measuring math anxiety in Italian college and high school students: Validity, reliability and gender invariance of the Abbreviated Math Anxiety Scale (AMAS). Learning and Individual Differences 34, 51–56 (2014).
15. Cipora, K., Szczygiel, M., Willmes, K., & Nuerk, H.-C. Math anxiety assessment with the abbreviated math anxiety scale: Applicability and usefulness: Insights from the Polish adaptation. Frontiers in Psychology, 6. (2015).
16. Carey, E., Hill, F., Devine, A., & Szűcs, D. The modified abbreviated math anxiety scale: a valid and reliable instrument for use with children. Frontiers in Psychology 8, 11, 1–13 (2017).
17. Pletzer, B., Wood, G., Scherndl, T., Kerschbaum, H. H., & Nuerk, H.-C. Components of mathematics anxiety: factor modeling of the MARS30-Brief. Frontiers in Psychology 7, 1–14 (2016).
18. Wigfield, A., & Meece, J. L. Math anxiety in elementary and secondary school students. Journal of Educational Psychology 80, 210–216 (1988).
19. Dowker, A., Sarkar, A., & Looi, C. Y. Mathematics anxiety: what have we learned in 60 Years? Frontiers in Psychology 7, 508 (2016).
20. Rubinsten, O., Bialik, N., & Solar, Y. Exploring the relationship between math anxiety and gender through implicit measurement. Frontiers in Human Neuroscience 6, 279 (2012).
21. Ashcraft, M. H., Krause, J. A., & Hopko, D. R. Is math anxiety a mathematical learning disability? In Why Is Math So Hard for Some Children? The Nature and Origins of Mathematical Learning Difficulties and Disabilities (ed. Berch, D. B. & Mazzocco, M. M. M.) 329–348 (Baltimore, MD: Paul H. Brookes Publishing, 2007).
22. Hembree, R. The nature, effects, and relief of mathematics anxiety. Journal for Research in Mathematics Education 21, 33–46 (1990).
23. Reynolds C. R., & Richmond B.O. Revised Children’s Manifest Anxiety Scale: Second Edition. 2012.
24. Hill, F., Mammarella, I., Devine, A., Caviola, S., Passolunghi, M.C., & Szűcs, D. Math anxiety in primary and secondary school students: Gender differences, developmental changes and anxiety specificity. Learning and Individual Differences 48, 45-53 (2016).
25. Carey, E., Devine, A., Hill, F., & Szűcs, D. Differentiating anxiety forms and their role in academic performance from primary to secondary school. PLoS One. PLoS ONE 12(3): e0174418 (2017).
26. Devine, A., Carey, E., Hill, F., & Szűcs, D. Cognitive and affective math problems largely dissociate: Prevalence of developmental dyscalculia and mathematics anxiety. Journal of Educational Psychology 110(3), 431-444 (2018).
27. Krinzinger, H., Kaufmann, L., & Willmes, K. Math anxiety and math ability in early primary school years. Journal of Psychoeducational Assessment 27: 206-225 (2009).
28. Carey, E., Devine, A., Hill, F., & Szűcs, D. The chicken or the egg? The direction of the relationship between mathematics anxiety and mathematics performance. Frontiers in Psychology 6: 1987 (2015).
29. Mandler, G., & Sarason, I. G. A study of anxiety and learning. Journal of Abnormal and Social Psychology 87, 166-173 (1952).
30. Ashcraft, M. H., & Faust, M. Mathematics anxiety and mental arithmetic performance: An exploratory investigation. Cognition & Emotion 8, 97–125 (1994).
31. Ashcraft, M. H., & Kirk, E. P. The relationships among working memory, math anxiety, and performance. Journal of Experimental Psychology General 130, 224–237. doi:10.1037/0096-3445.130.2.224 (2001).
32. Ashcraft, M. H., & Krause, J. A. Working memory, math performance, and math anxiety. Psychonomic Bulletin & Review 14, 243–248 (2007).
33. Reali-Arcos, F., Jimenez-Real, W., Maldonado-Carreno, C., Devine, A., & Szűcs, D. Examining the link between math anxiety and math performance in Colombian students. Revista Colombiana de Psicología. 25(2), 369-379 (2016).
34. Ibrahim Burak Olmez, Serkan Ozel. Mathematics anxiety among sixth and seventh grade Turkish elementary school students. Procedia – Social and Behavioral Sciences 46, 4933-4937 (2012).