Hembree on Math Anxiety

Reviewing a 1990 paper by Ray Hembree on math anxiety; a meta-study of approximately 150 papers, with a combined total of about 25,000 subjects. (Note the high sample size makes almost all findings significant at the p < 0.01 level). Math anxiety is known to be negatively correlated with performance in math (tests, etc.), and more common among women than men.

Math anxiety is somewhat correlated with a constellation of other general anxieties (r² = 0.12 to 0.27). Work to enhance math competence did not reduce anxiety.

Whole-group interventions are not effective (curricular changes, classroom pedagogy structure, in-class psychological treatments). The only thing that is effective is out-of-classroom, one-on-one treatments (behavioral systematic desensitization; cognitive restructuring); these have a marked effect at both lowering anxiety and boosting actual math-test performance.

In short: Addressing math anxiety is largely out of the hands of the classroom teacher. Unless the student has access, or the institution provides access, to one-on-one behavioral desensitization therapy, no group-level interventions are found to be effective.

Also recall that elementary education majors have the highest math anxiety, and the lowest math performance, of all U.S. college majors. (It seems possible that some entrants choose elementary education as a career path precisely because they are bad at math and see that as one of their limited options; I know I've had at least one such student say something to that effect to me.) This clearly dovetails with Sian Beilock's 2009 finding that math-anxious female elementary teachers model math-anxiety particularly to their female students, who imitate the same and wind up with worse math performance and attitudes by the end of the year (link). And this general trend of weak education majors has been the case in the U.S. for at least a century now (link).

So we might hypothesize: A feedback loop exists between poor early math education, heightened math anxiety among female students, and those same students returning to early childhood education as a career.

See below for Hembree's table of math anxiety by class and major (p. 41); note that elementary education majors, and those taking the standard "math for elementary teachers" (frequently the only math class such teachers take), are significantly worse off than anyone else:

Hembree, Ray. "The nature, effects, and relief of mathematics anxiety." Journal for research in mathematics education (1990): 33-46. (Link)