Academic 'stereotype threat' is real -- Willingham
By Daniel Willingham
Boys outscore girls in standardized tests of science. For example, in the 2005 National Assessment of Educational Progress, boys did slightly better than girls at fourth grade, eighth grade, and twelfth grade. Quick, what’s the reason for the difference?
Boys are somehow innately more predisposed to technical subjects?
Boys are more often steered toward science because it’s seen as a male field?
How about stereotype threat?
Stereotype threat was first described by psychologist Claude Steele in 1995. The idea is that some of the observed differences on tests don’t reflect real differences in knowledge, nor test bias. Rather, they are due to test anxiety. When a member of a stereotyped group believes she is being evaluated on something relevant to the stereotype, she becomes anxious about her performance.
In Steele’s original experiment he gave different groups of African-Americans the same test. The group told that the test was diagnostic of intellectual ability did worse on the test than another group told that it was *not* diagnostic of intellectual ability.
In another methodology, researchers don’t manipulate subjects’ perception of the test, they manipulate subjects’ perception of themselves—they simply remind subjects of their status as a member of the group. That too makes a difference on test performance.
A recent study on girls and science showed how insidious stereotype threat can be. Researchers had high school students read a three-page passage from a standard high chemistry textbook. Three accompanying photographs showed chemists at work. In one all three were male, in another, all three were female, and in a third condition, there was a mix.
When students tested on their comprehension of the textbook passage, the results showed a reliable effect of the photographs. When all men were depicted, boys outscored girls. When all women were depicted, girls outscored boys. When men and women were depicted, boys and girls scored the same.
What are the practical implications of such findings? Should we make sure that every textbook shows both genders and every ethnic group in equal proportion?
To the extent possible, that seems a wise course, but it probably won’t be enough—the important scholars in various disciplines will not be evenly divided along these lines.
I take two messages from studies like this one. First, at least part of the difference observed among groups on tests has nothing to do with ability, interest, or the validity of the test. It’s due to subtle, unintended messages about identity sent just before the test.
Second, even if perfectly balancing gender and ethnicity in textbooks proves a heavy lift, there are other ways parents and teachers can send the message—explicitly and implicitly—that *all* work is appropriate for *all* students.
The differences observed in this study were not huge—about 1.5 questions out of 12. But if we’re concerned about social inequities, and if we’re concerned that all latent talent be fully expressed, this may be one of the easier fixes.
Follow my blog all day, every day by bookmarking washingtonpost.com/answersheet. And for admissions advice, college news and links to campus papers, please check out our new Higher Education page at washingtonpost.com/higher-ed Bookmark it!
| May 31, 2010; 12:17 PM ET
Categories: Daniel Willingham, Guest Bloggers, Science | Tags: gender gap, gender stereotypes, science, stereotype threat, study on gender gap
Save & Share: Previous: Kids spend Memorial Day in school
Next: The schooling of Buddy Valastro, the Cake Boss
Posted by: itkonlyyou94 | June 1, 2010 8:56 AM | Report abuse
The comments to this entry are closed.