Date Published: March 28, 2017
Publisher: Public Library of Science
Author(s): Emma Carey, Amy Devine, Francesca Hill, Dénes Szűcs, Joseph Najbauer.
Individuals with high levels of mathematics anxiety are more likely to have other forms of anxiety, such as general anxiety and test anxiety, and tend to have some math performance decrement compared to those with low math anxiety. However, it is unclear how the anxiety forms cluster in individuals, or how the presence of other anxiety forms influences the relationship between math anxiety and math performance.
We measured math anxiety, test anxiety, general anxiety and mathematics and reading performance in 1720 UK students (year 4, aged 8–9, and years 7 and 8, aged 11–13). We conducted latent profile analysis of students’ anxiety scores in order to examine the developmental change in anxiety profiles, the demographics of each anxiety profile and the relationship between profiles and academic performance.
Anxiety profiles appeared to change in specificity between the two age groups studied. Only in the older students did clusters emerge with specifically elevated general anxiety or academic anxiety (test and math anxiety). Our findings suggest that boys are slightly more likely than girls to have elevated academic anxieties relative to their general anxiety. Year 7/8 students with specifically academic anxiety show lower academic performance than those who also have elevated general anxiety.
There may be a developmental change in the specificity of anxiety and gender seems to play a strong role in determining one’s anxiety profile. The anxiety profiles present in our year 7/8 sample, and their relationships with math performance, suggest a bidirectional relationship between math anxiety and math performance.
Mathematics anxiety (MA) encompasses emotions of fear, tension and discomfort which are felt by some individuals in situations involving mathematics, and which may interfere with one’s performance of mathematical tasks . MA has been seen to relate to math performance in children as young as 5–7 years old [2–4], and this relationship remains in adolescence and adulthood, with two meta-analyses showing correlations of -0.27 and -0.34 between MA and math performance [5,6]. It seems most likely that this relationship is bidirectional, with poor performance contributing to some cases of MA, and MA causing a performance decrement in at least some affected individuals (see  for review). Our analysis here aims to investigate how MA is related to other forms of anxiety in a very large sample of 1720 UK children (aged 8–9 years) and adolescents (aged 11–13 years), and how each individual’s anxiety “profile” relates to their academic performance. Measurement of test anxiety and general anxiety as well as MA is novel in a very large sample study, spanning two age groups. Furthermore, we use a combined person- and variable-centered analysis (latent profile analysis), which uniquely enables us to investigate the complex relationship between anxiety forms and performance.