High school math teachers' ratings of their students reveal a gender bias, according to an analysis of national survey data.
Teachers tend to rate white girls' math abilities lower than those of white male students, even when the girls' grades and test scores are comparable to boys, according to a team of researchers who analyzed data collected as part of the national Educational Longitudinal Study of 2002.
"We find evidence of a consistent bias against white females, which although relatively small in magnitude, suggests that teachers hold the belief that math is just easier for white males than it is for white females," write Catherine Riegle-Crumb and Melissa Humphries of the University of Texas at Austin in the April issue of the journal Gender & Society.
Starting in 2002, the ELS began following a nationally representative cohort of high-school students, collecting information from students, parents, librarians and teachers. More data will be collected from the cohort this year, according to the U.S. National Center for Education Statistics.
Math teachers indicated whether or not the students were enrolled in classes that fit their abilities, were too easy or too difficult.
"The bias teachers reveal against white female students may very well be something they are not consciously aware of, but it's usually subtle," said Riegle-Crumb in a news release issued by the university. "But it's definitely present, per our research findings."
ELS data showed most teachers rated the math abilities of male and female minorities lower when their test scores and grades were low, but this does not constitute a bias because reasonable data support the teachers' evaluations, according to the researchers.
However, this does not mean that minority students are free from negative stereotypes, according to Riegle-Crumb. Previous research suggests race plays into teachers' expectations of students.
You can follow senior writer Wynne Parry on Twitter . Follow LiveScience for the latest in science news and discoveries on Twitter and on .