A British study looking at possible gender differences in the effects of math anxiety involved 433 secondary school children (11-16 years old) completing customized (year appropriate) mental mathematics tests as well as questionnaires designed to assess math anxiety and (separately) test anxiety. These sources of anxiety are often confounded in research studies (and in real life!), and while they are indeed related, reported correlations are moderate, ranging from .30 to .50.
Previous research has been inconsistent as regards gender differences in math anxiety. While many studies have found significantly greater levels of math anxiety in females, many studies have found no difference, and some have even found higher levels in males. These inconsistencies may stem from differences in how math anxiety is defined or measured.
The present study looked at a rather more subtle question: does the connection between math anxiety and math performance differ by gender? Again, previous research has produced inconsistent findings.
Findings in this study were very clear: while there was no difference between boys and girls in math performance, there were marked differences in both math and test anxiety. Girls showed significantly greater levels of both. Both boys and girls showed a positive correlation between math anxiety and test anxiety, and a negative correlation between math anxiety and math performance, and test anxiety and performance. However, these relationships between anxiety and performance were stronger for girls than boys, with the correlation between test anxiety and performance being only marginally significant for boys (p<0.07), and the correlation between math anxiety and performance disappearing once test anxiety was controlled for.
In other words, greater math anxiety was linked to poorer math performance, but it was significant only for girls. Moreover, anxiety experienced by boys may simply reflect test anxiety, rather than specific math anxiety.
It is worth emphasizing that there was no gender difference in performance — that is, despite laboring under the burden of greater levels of anxiety, the girls did just as well as boys. This suggests that girls might do better than boys if they were free of anxiety. It is possible, however, that levels of anxiety didn’t actually differ between boys and girls — that the apparent difference stems from girls feeling more free to express their anxiety.
However, the finding that anxiety is greater in girls than boys is in line with evidence that anxiety (and worry in particular) is twice as prevalent in women as men, and more support for the idea that the girls are under-performing because of their anxiety comes from another recent study.
In this study, 149 college students performed a relatively simple task while their brain activity was measured. Specifically, they had to identify the middle letter in a series of five-letter groups. Sometimes the middle letter was the same as the other four ("FFFFF") while sometimes it was different ("EEFEE"). Afterward the students completed questionnaires about their anxiety and how much they worry (Penn State Worry Questionnaire and the Anxious Arousal subscale of the Mood and Anxiety Symptom Questionnaire).
Anxiety scores were significantly negatively correlated with accuracy on the task; worry scores were unrelated to performance.
Only girls who identified themselves as particularly anxious or big worriers recorded high brain activity when they made mistakes during the task (reflecting greater performance-monitoring). Although these women performed about the same as others on simple portions of the task, their brains had to work harder at it. Then, as the test became more difficult, the anxious females performed worse, suggesting worrying got in the way of completing the task.
Greater performance monitoring was not evident among anxious men.
[A reminder: these are group differences, and don't mean that all men or all women react in these ways.]
 . Gender differences in mathematics anxiety and the relation to mathematics performance while controlling for test anxiety. Behavioral and Brain Functions [Internet]. 2012 ;8(1):33 - 33. Available from: http://www.behavioralandbrainfunctions.com/content/8/1/33/abstract
 . Sex moderates the relationship between worry and performance monitoring brain activity in undergraduates. International Journal of Psychophysiology [Internet]. Submitted . Available from: http://www.sciencedirect.com/science/article/pii/S0167876012001444