CHAPTER 5- GENDER COMPARISONS IN COGNITIVE ABILITIES AND ATTITUDES ABOUT
- Article titled “He Thinks, She Thinks,” by Marsa (2007), in Discover magazine. The article
emphasizes that societal factors cannot explain gender differences. Instead, Marsa claims, “Our
brains are hardwired differently, and these anatomical variations in architecture and function
illuminate some of the reasons why men and women seem to come from different planets”
- From Theme 1 of textbook, when people who are not experts discuss gender comparisons in
thinking, they almost always emphasize gender differences and ignore the substantial evidence
for gender similarities. Furthermore, people who are not experts typically highlight biological
explanations for the small number of comparisons that reveal significant gender differences
- cultural explanations play a very important role in accounting for gender differences.
Background on Gender Comparisons
Cautions about research in gender comparisons:
1. Biased samples can influence results.
- Almost all the research on cognitive abilities focuses on college students, so this research is not
representative of the general population. We know almost nothing about adults who have not
attended college. In addition, most of the research on gender comparisons examines White men
and women in the United States and Canada. Conclusions about gender comparisons might be
different if these studies had included people of color.
2. People’s expectations can influence results.
- Biases can interfere at every stage of the research process. Example: researchers who expect to
find gender differences will tend to find them. The participants also have expectations about
cognitive gender differences. Recall: stereotype threat.
3. If we measure some ability, and then we create one graph for the scores of males and another
graph for the scores of females, the two distributions of scores will overlap substantially.
- Frequency distribution: tells us how many people in a sample receive each score. X axis is the
score, and y axis is the # of people receiving each score.
I. Example: give a vocabulary test to a group of women and men. Then we use their scores to
construct a hypothetical frequency distribution for each gender. If there is a tiny section of
overlap of the male and female frequency distributions then this means that males and females
received the same scores only in that one small region (Fig 5.1: roughly between 54 and 66).
The small overlap pattern of the distribution indicates that the two distributions are very
different (Fig 5.1: average man received a score of 40, average woman received a score of 80).
HOWEVER in real life, distributions of female and male characteristics rarely show the large
separation and the small overlap. They are much more likely to show a small separation and a
large overlap (Fig 5.2 most males and most females earn scores in the large region that extends
roughly between 35 and 85). Therefore, theme 1: males and females are reasonably similar. As
a result, their scores will overlap considerably (Figure 5.2: avg man received score of 57, avg
woman score 63).
4. Researchers seldom find gender differences in all situations.
- Cannot make general statements about gender differences. Instead, the gender differences often
disappear when we test certain kinds of people or when we look at particular situations. Thus,
gender differences can be modified; they are not inevitable. In short, many males and females
have remarkably similar psychological characteristics in many situations.
5. The cognitive gender differences are not large enough to have a major influence on a person’s
- Example: engineering as a career choice. At present, only about 12% of U.S. engineers are
women. Engineering clearly requires spatial skills, and the research shows that men are somewhat
more likely than women to earn higher scores on tests of spatial ability. - Can the gender difference in spatial skills account for the very small percentage of women in
engineering? Assume a career in engineering would require that a person must have spatial skills
in the top 5% of the general population. 7% of males and 3% of females typically place in the top
5% of the population. Thus, about 30% of the people with superior spatial abilities are female.
However, 30% is much greater than 12%. CONCLUSION: gender difference in spatial skills
might partially explain the relative absence of women in engineering. However, we need to look
for other factors that could explain that gap of 18%.
The Meta-Analysis Approach to Summarizing Multiple Studies:
- When psychologists want to obtain an overview of a specific topic, they typically review the
research by examining all the studies on that topic.
- Psychologists used to use the box-score approach (or counting approach) to reviewing
research. Researchers read through all the appropriate studies on a given topic and draw
conclusions based on a tally of their outcomes.
- Unfortunately, the box-score approach often produces ambiguous tallies.
I. Example: how many studies show no gender differences, how many show higher scores
for women, and how many show higher scores for men? Suppose that researchers locate
16 relevant studies; 8of these studies find no gender differences, 2 show higher scores for
women, and 6 show higher scores for men. One researcher might conclude that no gender
differences exist, whereas another might conclude that men score somewhat higher. The
box-score approach does not provide a systematic method for combining individual
- A more useful alternative, called “meta-analysis” (recall: the technique used for gender
comparisons in self esteem).
- Meta-analysis: a statistical method for combining numerous studies on a single topic.
Researchers first try to locate all appropriate studies on the topic. Then they perform a statistical
analysis that combines the results from all these studies, taking into account the variability of the
scores for both females and males. This analysis calculates the size of the overall difference
between two groups of people, such as females and males.
I. Example, for verbal ability, a meta-analysis can combine numerous previous studies into
one enormous super-study that can provide a general picture of whether gender has an
overall effect on verbal ability.
- A meta-analysis yields a number known as effect size, or d.
I. Small overlap, large separation → large d
Example: d for the gender difference in height; here, the d is 2.0. This is a huge difference! In
fact, the overlap between the male and female distributions for height is only 11%
II. Large overlap, small separation → small d
Example: the d values for psychological gender comparisons are relatively small. Study:
Hyde (2005a) examined 128 different meta-analysis measures that focused on gender
comparisons in cognitive skills. RESULTS: 30% of gender comparisons were in the “close-
to-zero” range (d less than 0.11), 48% had a small effect size (d = 0.11 to 0.35), 15% had a
moderate effect size (d = 0.36 to 0.65), and only 8% had a large effect size (d greater than
0.65). CONCLUSION: majority of these comparisons of cognitive abilities showed either no
gender difference or a small gender difference.
III. Same overlap, no separation → zero d
Example: meta-analysis of numerous studies shows that males and females received exactly
the same overall score, the d would be zero
Cognitive Abilities that Show NO Consistent Gender Differences:
1. General Intelligence
- Females and males are similar in general intelligence, as measured by total scores on an IQ test.
People who construct intelligence tests often eliminate test items that show a gender difference. As a result, the final versions of the intelligence tests usually reveal gender similarities. However,
IQ scores for males show greater variability than IQ scores for females .
- Other research also shows gender similarities in general knowledge about history, geography, and
other basic information.
- Gender similarities also found in “multitasking”
2. Complex Cognitive Tasks
- males and females are equally competent when they form concepts and when they solve a variety
of complex problems
- Males and females are also similar in their performance on a variety of creativity tasks.
- It was assumed that for “learning style,” girls learn best in a cooperative environment and boys
learn best in a competitive environment. However, researchers have not discovered gender
differences in learning style.
Four kinds of cognitive abilities with some evidence of gender differences: (1) memory, (2) verbal ability,
(3) mathematics ability, and (4) spatial ability.
1. Memory Ability:
- Women tend to score higher on a variety of memory tasks.
- In one kind of memory task, people see a list of words. After a delay, they are asked to remember
the words. In general, women are somewhat more accurate on this kind of memory skill.
However, the nature of the items on the list may influence the results. Example: Colley and
colleagues (2002) gave women and men a list of items to remember. The list was labeled either
“Grocery store” or “Hardware store.” The items on the list were equally likely for both kinds of
stores (for example, nuts, salt, and disinfectant). RESULTS: women recalled many more items
than men from the “grocery” list, but women and men recalled a similar number of items from the
- Women tend to be more accurate than men in remembering events from their own lives. Recall:
mothers are more likely to discuss emotional topics with their daughters, rather than their sons.
As a result, girls have more opportunities to practice remembering these personal events. The
gender differences in memory for life events is therefore consistent with the research in cognitive
psychology, which shows that people with practice and expertise in a specific area remember this
material more accurately than non-experts.
- Women tend to be more accurate than men in recognizing faces, as well as faces from a different
ethnic group. Example: Swedish women performed better than Swedish men in recognizing the
faces of people from the South Asian country of Bangladesh.
- Women are also more accurate than men in recalling details about a person’s hair and clothing
- Women are also better than men in remembering objects that they have seen at an earlier time and
also in remembering where they have seen these objects, according to a meta-analysis by Voyer
and his colleagues (2007).
- However, men and women are similar in remembering abstract shapes
Summary: women perform somewhat better than men on a variety of memory tasks. Should wait for a
meta-analysis before drawing firm conclusions. However, women generally earn somewhat higher scores
on memory tests for words, life events, faces, and objects.
2. Verbal Ability:
- Females score somewhat higher than males on a small number of verbal tasks, although the
overall gender similarities are more striking.
General Verbal Ability
- Young School- age children: more similarities than differences in verbal ability.
- Adolescents and adults: research shows gender similarities in language skills such as spelling,
vocabulary, word associations, reading comprehension, and learning a second language. - However, females seem to be somewhat better at verbal fluency, or naming objects that meet
certain criteria, such as beginning with the letter S
- Females have scored higher on tests of writing ability recall: meta-analysis is the ideal statistical
tool for combining the results of a number of studies on a specific topic. Study: Hyde and Linn
(1988) conducted a meta-analysis on overall gender comparisons in verbal ability. RESULT: The
average effect size (d) was only 0.11, just slightly favoring females. This value is very close to
zero. CONCLUSION: overall gender differences do not exist.
- Note: Ironically, researchers seldom study the two general areas in which females occasionally
have the advantage, memory and verbal abilities. In contrast, there is much more research about
mathematical and spatial abilities, areas in which males may have an advantage.
- Example: SAT test for college admission: The critical reading portion of this test includes reading
comprehension and sentence completion. Gender differences on this part of the SAT are minimal.
For example, in 2009, the average SAT critical reading score was 498 for women and 503 for
- Gender differences are also minimal for the Advanced Placement examinations in several related
areas, specifically, English language, English literature, and all foreign languages
- Males are more likely than females to have language problems.
- School systems report reading disabilities about four or five times as often for boys as for girls
- Study: Shaywitz and coauthors (1990): objective measure of the term reading disability: poor
reading skills that are not accounted for by the level of general intelligence. These researchers
used this operational definition to study children in Connecticut. RESULTS: roughly the same
number of boys and girls met the criterion of having reading disabilities. Specifically, boys were
about 1.2 times more likely than girls to have reading disabilities.
- Study: Rutter and coauthors (2004): performed a more recent analysis of children’s reading
disabilities in New Zealand. They used a similar definition of reading disability to Shaywitz and
colleagues (1990). They included general intelligence in their analysis. RESULTS: boys were
about twice as likely as girls to have reading disabilities.
- Why schools identify reading problems four to five times more often in boys than in girls:
Research shows that boys have more trouble focusing their attention, whereas girls are more
skilled at controlling their behavior. It’s likely that teachers target the more active, less attentive
boys as having reading disabilities. These boys may be referred to a reading clinic on the basis of
their behavior, rather than their poor reading skills. *An equally disturbing problem is that many
girls probably have genuine reading disabilities, but they sit quietly in their seats and hide their
disabilities. These well-behaved, neglected girls will miss out on the additional tutoring in
reading that could help them thrive in school.
Summary: boys are more likely than girls to have reading disabilities.
3. Mathematics Ability:
- Performance in mathematics is the cognitive ability that receives the most attention from both
researchers and the popular press. Media reports lead one to expect large gender differences in
math ability, favoring males.
- However, females and males in both the United States and Canada now complete the same
number of math courses during high school.
- most of the research shows gender similarities in math ability
- Females actually receive higher grades in math courses. The only measure on which males
perform substantially better than females is the mathematics section of the SAT
General Mathematics Ability
- Most comparisons of males’ and females’ ability on mathematics achievement tests show gender
similarities. - Study: Hyde and colleagues (1990): a meta-analysis of 100 studies, based on standardized-test
scores of more than 3 million students (NOT including math SAT scores). RESULTS: d of only
0.15. The two distributions are almost identical.
- National Center for Education Statistics (2004) reported the scores for eighth-grade students on a
standardized mathematics test. Average scores from 34 different countries throughout the world.
RESULTS: the boys’ average was higher than the girls’ average in 16 countries, the girls’
average was higher than boys’ average in 16 countries, and girls and boys had the same averages
in 2 countries.
- Hyde and colleagues (2008): provide additional evidence of gender similarities on standardized
math exams. These researchers analyzed test scores for 7.2 million students in 10 U.S. states.
They found consistent gender similarities for students of all ages, from 2nd grade through 11th
grade, even when the tests included complex math problems.
Summary: The results of the research on general math abilities show gender similarities in mathematics.
Grades in Mathematics Courses
- Females earn higher grades in fifth-, sixth-, eighth-, and tenth-grade mathematics as well as in
college math courses. Females also earn higher grades in related areas, such as high-school
science courses and college-level statistics
- Kimball (1989, 1995) proposed that females perform better when dealing with familiar situations,
such as exams on material covered in a mathematics course. In contrast, males perform better
when dealing with unfamiliar situations, especially the kinds of math problems included on the
SAT. Kimball points out that females’ high grades in math courses deserve wider publicity. This
publicity would encourage females, their parents, and their teachers to be more confident about
girls’ and women’s competence in mathematics.
The Mathematics SAT
- There is high media attention to performance on the math portion of the SAT. Example: 2009:
women received an average score of 499, in contrast to 534 for men.
- Is the math SAT test a valid index of ability in mathematics? A test has high validity if it
measures what it is supposed to measure. For example, the SAT is supposed to predict students’
grades in college courses. The SAT has high overall validity because people with higher SAT
scores generally do earn higher grades in college math courses. However, the math portion of the
SAT is not valid with respect to its prediction that women will earn lower grades in college math
courses than men do. The math SAT underestimates women’s actual math performance.
PROBLEM: colleges and universities reject female students who would be likely to earn higher
math grades than the accepted male students. Based on validity studies such as these, some
colleges and universities have stopped using the SAT or have modified