When women are told that they are bad at math, they will perform worse on math tests than women who are told that's a stereotype they can overcome, a new study shows.

Researchers at the University of British Columbia have discovered that when women are told their gender is irrelevant to their mathematical abilities, they do much better on math tests.

"The interesting thing -- the new thing (our study) brings into the picture," said Ilan Dar-Nimrod, a doctoral student and the lead author of the UBC study, "is that women process information differently if they believe the cause is genetic.

"But they can overcome a stereotype if they think the cause is environmental," he told CTV.ca.

The research builds on what is known about a concept called stereotype threat, in which being reminded of a commonly accepted negative belief about your group can elicit behaviour that mimics the stereotype.

The phenomenon of stereotype threat exacts a huge cost, said the study's co-author, Steven Heine, an associate professor of social psychology at UBC, because "negative stereotypes hurt people's performance.

"But stereotypes lose their teeth and are less harmful," Heine said, when attributed to a cause such as societal expectations.

The UBC researchers wanted to find out if it would have an impact on test scores if women were given different explanations for the stereotype that women aren't good at math.

Over a three-year period, 135 women were asked to take tests that were deliberately designed to be very difficult. The tests were similar to those given as graduate school entrance exams.

Before completing the math section of the tests, the women were given one of four essays to read. Three of the essays discussed gender differences regarding math.

One essay asserted that there were no gender differences in math performance, the second blamed genetic differences for poorer performance by women, and the third maintained that the way girls were taught in elementary school had an impact on math scores.

The fourth essay didn't mention math. Instead, it discussed the subject of women in art. Previous research has indicted that a seemingly neutral reminder of her gender is enough to have a negative impact on a woman's test performance.

In this UBC study:

• Women who were told that their previous experience with math would be the deciding factor in test scores gave twice as many correct answers as women who read the essay blaming their genes.
• Women who were merely reminded of their gender also performed worse than those told there was no difference in male and female abilities.

The tests would have made most people struggle, Heine said. When the women found the math section challenging, this created stress that caused them to credit their difficulty to the stereotype -- and they did more poorly than those told the stereotype didn't apply.

The UBC researchers said the study may indicate a need for scientists to look carefully at how they report discoveries of genetic links to intellectual and physical attributes.

There is a risk of genetically-linked causation being adopted as a negative stereotype, they said.

Stereotypes by themselves aren't necessarily bad, because they can be useful as cognitive shortcuts, said Dar-Nimrod. But "being a stereotype doesn't mean it's true or it's false," he said.

Stereotypes are not true of everyone within a group and when people believe genetics will determine an outcome, the belief can become a self-fulfilling prophecy.

"For example, if you're having problem with your weight and find out there's a gene that causes that, you may lose motivation to try to lose weight because you think it's in your DNA (and believe) 'thanks to my parents I will be heavy set,'" he explained.

Since motivation can be a strong a factor in success, believing genetics predetermine an outcome can sabotage what could have become a successful weight reduction program, he said.

Gender differences in performance have long been a controversial topic. Harvard president Lawrence Summers was forced to resign this past summer after he suggested that there may be fewer women in the sciences because they don't have the same natural abilities as men for scientific endeavour.

Barbara Keyfitz, president of the Association for Women in Mathematics, told the Canadian Press that the issue can be resolved by providing strong role models for women who want to enter fields involving high level mathematics. That will break the stereotypical expectation, she said.

With files from the Canadian Press