gender cues and test scores
With Subtle Reminders, Stereotypes Can Become Self-Fulfilling
By Shankar Vedantam
Monday, December 11, 2006; A02
The questions seemed innocuous. One group of female students was
asked about their telephone service, while another group was asked
their views on co-ed housing.
Both groups were then asked how pleasant it would be to listen to
music for a class assignment, to analyze a poem, to solve an
equation or to take a calculus exam -- questions that tested their
preference for math or the arts. Students were assigned to the
groups at random, so you would expect similar answers from both
groups, right?
Wrong.
As tens of thousands of high school, college and university
students receive grades for their work in the fall semester this
week, there might be more to the grades than whether the students
spent the semester partying or studying. A number of experiments
show that subtle cues can alter academic preferences -- and test
performance -- without anyone being aware of it.
The women in the group who were asked about co-ed housing
expressed a greater preference for the arts compared with the
women who were asked questions about their telephone
service. Researchers believe that reminding the students about
their sex -- as a question about co-ed housing would do -- subtly
activated an association with the sexual stereotype that the arts
are feminine, and math is masculine.
Before we go any further into these troubling waters, let us halt
for an important caveat: This phenomenon does not explain all
differences in test scores or academic preferences. People clearly
do have different temperaments, and talent and application matter
greatly. If you spent the last semester tuned out, that provides a
simpler explanation for poor grades than the activation of subtle
stereotypes.
But the research does raise troubling questions for academicians
and parents, not least because the cues we are talking about are
so commonplace. Indeed, the research study on the arts-math
preferences, conducted by psychologists Jennifer Steele at York
University in Toronto and Nalini Ambady of Tufts University in
Boston, proved that cues do not have to involve explicit
questions.
In another part of the study they recently published in the
Journal of Experimental Social Psychology, two groups were told to
focus on a plus sign on a computer screen. Periodically, they
would see flashes on the right or left.
What the students did not know is that the flashes displayed a
word for less than one-tenth of a second, followed by a string of
"X's." The word was presented too quickly for the undergraduate
women volunteers to be aware of what it said -- but it was enough
to make a difference at a subliminal level.
When words with feminine associations such as "doll" or "lipstick"
or "skirt" were flashed, students were more likely to express a
preference for the arts over math compared with those who were
flashed the words "hammer," "suit" or "cigar."
"It is disturbing to think I can show you words outside your
awareness and that can influence your preference," Steele said.
Women are not the only ones affected in this way, of
course. Reminding white men of the stereotype that Asians are
better at math can lower the performance of the white men in math
tests.
Again, no one is saying such cues turn brilliant students into
dullards, but the cues do cause measurable differences in
scores. "Sometimes it is a couple of questions, but when you are
talking of acceptance into top universities, one or two questions
can make a huge difference," Steele said.
Ambady is currently studying how to address these issues, which
obviously have implications for fairness. It gets very
complicated: When 5- to 7-year-old and 11- to 13-year-old Asian
girls are subtly reminded of their Asian identity, they do better
at math tests; when subtly reminded about their sex, they do
worse. In other words, both positive and negative stereotypes have
effects, and both can be activated simultaneously-- meaning that
people seeking to fight stereotypes are well-advised to be
cautious.
Ambady said that drawing attention to the girls' individuality --
by asking about their favorite book or movie, for example, or
asking them to list a few things about themselves that they liked
and disliked -- caused them to do much better on math tests
compared with girls primed with a negative gender stereotype that
subtly reminded the girls of their group identity.
In another intriguing study, David Butz, a psychology graduate
student at Florida State University, found that displaying the
American flag in a room when students are asked to solve math
problems or anagrams can influence performance. As with other
experiments, the students themselves were not aware that the
subtle cue made a difference -- in fact, most said they did not
even notice the flag.
Butz designed the study after Florida law mandated that an
American flag be hung in public classrooms. He found that the flag
boosted the performance of white students but not
minorities. White students given a math test in a room without a
flag solved 44 percent of the problems. Those shown the flag
solved 51 percent. Minorities solved 42 percent of the problems
without the flag and 41 percent with it -- no difference.
Makes you wonder, doesn't it, about the hidden power that lies in
the ordinary things around you?
By Shankar Vedantam
Monday, December 11, 2006; A02
The questions seemed innocuous. One group of female students was
asked about their telephone service, while another group was asked
their views on co-ed housing.
Both groups were then asked how pleasant it would be to listen to
music for a class assignment, to analyze a poem, to solve an
equation or to take a calculus exam -- questions that tested their
preference for math or the arts. Students were assigned to the
groups at random, so you would expect similar answers from both
groups, right?
Wrong.
As tens of thousands of high school, college and university
students receive grades for their work in the fall semester this
week, there might be more to the grades than whether the students
spent the semester partying or studying. A number of experiments
show that subtle cues can alter academic preferences -- and test
performance -- without anyone being aware of it.
The women in the group who were asked about co-ed housing
expressed a greater preference for the arts compared with the
women who were asked questions about their telephone
service. Researchers believe that reminding the students about
their sex -- as a question about co-ed housing would do -- subtly
activated an association with the sexual stereotype that the arts
are feminine, and math is masculine.
Before we go any further into these troubling waters, let us halt
for an important caveat: This phenomenon does not explain all
differences in test scores or academic preferences. People clearly
do have different temperaments, and talent and application matter
greatly. If you spent the last semester tuned out, that provides a
simpler explanation for poor grades than the activation of subtle
stereotypes.
But the research does raise troubling questions for academicians
and parents, not least because the cues we are talking about are
so commonplace. Indeed, the research study on the arts-math
preferences, conducted by psychologists Jennifer Steele at York
University in Toronto and Nalini Ambady of Tufts University in
Boston, proved that cues do not have to involve explicit
questions.
In another part of the study they recently published in the
Journal of Experimental Social Psychology, two groups were told to
focus on a plus sign on a computer screen. Periodically, they
would see flashes on the right or left.
What the students did not know is that the flashes displayed a
word for less than one-tenth of a second, followed by a string of
"X's." The word was presented too quickly for the undergraduate
women volunteers to be aware of what it said -- but it was enough
to make a difference at a subliminal level.
When words with feminine associations such as "doll" or "lipstick"
or "skirt" were flashed, students were more likely to express a
preference for the arts over math compared with those who were
flashed the words "hammer," "suit" or "cigar."
"It is disturbing to think I can show you words outside your
awareness and that can influence your preference," Steele said.
Women are not the only ones affected in this way, of
course. Reminding white men of the stereotype that Asians are
better at math can lower the performance of the white men in math
tests.
Again, no one is saying such cues turn brilliant students into
dullards, but the cues do cause measurable differences in
scores. "Sometimes it is a couple of questions, but when you are
talking of acceptance into top universities, one or two questions
can make a huge difference," Steele said.
Ambady is currently studying how to address these issues, which
obviously have implications for fairness. It gets very
complicated: When 5- to 7-year-old and 11- to 13-year-old Asian
girls are subtly reminded of their Asian identity, they do better
at math tests; when subtly reminded about their sex, they do
worse. In other words, both positive and negative stereotypes have
effects, and both can be activated simultaneously-- meaning that
people seeking to fight stereotypes are well-advised to be
cautious.
Ambady said that drawing attention to the girls' individuality --
by asking about their favorite book or movie, for example, or
asking them to list a few things about themselves that they liked
and disliked -- caused them to do much better on math tests
compared with girls primed with a negative gender stereotype that
subtly reminded the girls of their group identity.
In another intriguing study, David Butz, a psychology graduate
student at Florida State University, found that displaying the
American flag in a room when students are asked to solve math
problems or anagrams can influence performance. As with other
experiments, the students themselves were not aware that the
subtle cue made a difference -- in fact, most said they did not
even notice the flag.
Butz designed the study after Florida law mandated that an
American flag be hung in public classrooms. He found that the flag
boosted the performance of white students but not
minorities. White students given a math test in a room without a
flag solved 44 percent of the problems. Those shown the flag
solved 51 percent. Minorities solved 42 percent of the problems
without the flag and 41 percent with it -- no difference.
Makes you wonder, doesn't it, about the hidden power that lies in
the ordinary things around you?