THE MYTH THAT SCHOOLS SHORTCHANGED GIRLS

  by Judith Kleinfeld  © 1998
Professor of Psychology
University of Alaska
Fairbanks, Alaska 99712
907-474-5266

Women's advocacy groups have waged an intense media campaign to promote the idea that the "schools shortchange girls. " Their goal is to intensify the image of women as "victims" deserving special treatment and policy attention. Their sophisticated public relations campaign has succeeded. The idea that girls are victimized by the schools has become the common wisdom, what educated people just assume to be true.

But the idea that the "schools shortchange girls" is wrong and dangerously wrong. It is girls who get higher grades in school, who do better than boys on standardized tests of reading and writing, and who get higher class rank and more school honors. It is young women who enter and graduate from college far more frequently than young men. It is women who have made dramatic progress in obtaining professional, business, and doctoral degrees. The great gender gap of the 1960s in advanced degrees has almost closed, especially in the professional fields to which ambitious women aspire. In the view of elementary and high school students, the young people who sit in the classroom year after year and observe what is going on, both boys and girls agree: Schools favor girls. Teacher think girls are smarter, like being around them more, and hold higher expectations for them.

This does not mean that males and females are equal on every educational outcome. In some areas, females do better than males, and in other areas, males do better than females. Females lag behind in two academic areas: mathematics and science achievement. Females also lag slightly behind males in attaining professional, business, and doctoral degrees. But males lag behind females in two other academic areas and by far wider margins: reading achievement and writing skills. Males are far more apt to end up at the bottom of the barrel in school, placed in special classes for students with learning disabilities. Males are also more apt than females to believe that the school climate is hostile to them, that teachers do not expect as much from them and give them less encouragement to do their best.

The myth that the schools shortchange girls is dangerously wrong because it has diverted policy attention from the group at genuine educational risk‹African-American boys. This is the group that scores lowest on virtually every educational measure. This is the group where an enormous gap does exist between males and females. But the African American gender gap favors females, who are pulling far ahead of males in college graduation rates and in obtaining professional degrees.

Where did the notion that the schools shortchange girls come from? And how do advocacy groups manage to convince people that it is girls who are victimized in the schools? What data do they use and what data do they ignore?

In this paper, I examine the charges made in a highly publicized report, How Schools Shortchange Girls, published by the American Association of University Women (1992). I show how the findings in this report are based on a selective review of the research and how findings contrary to the report's message were suppressed. These contrary findings indeed appear in studies the AAUW itself commissioned, but the AAUW not only did not include these findings in their media kits but made the data difficult to obtain.

To find out what is actually going on, how boys and girls do fare in the schools, I review the best available information on a wide variety of strong measures: school grades, class rank, honors and prizes in academic competitions, scores on standardized achievement tests, college entrance and graduation rates, and attainment of professional and doctoral degrees. To locate this information, I often had to do new analyses of government reports, which also emphasize the "women as victims" viewpoint‹showcasing the problems but not the progress. I examine as well charges that the schools shortchange girls based on weak measures, the view that girls are silenced in the classroom and suffer a dramatic loss of self-confidence at adolescence. I show that the research on which these charges are based have in some instances disappeared and in other instances have been distorted to make a political point. Research on gender differences in class participation, school climate, and self confidence provides a welter of conflicting findings, sometimes favoring girls, sometimes favoring boys, and sometimes showing no gender differences at all.

The AAUW Report: How Schools Shortchange Girls

The American Association of University Women (AAUW) put itself on the political map through its highly publicized 1992 report: How Schools Shortchange Girls. The media trumpeted the message around the world: In the schools, as in so many other areas of life, females are victims. Girls are silenced in the classroom, suffer a decline in self-esteem at adolescence, and fall far behind boys in such crucial subjects as science and mathematics. As the AAUW Executive Summary declares:

The educational system is not meeting girls' needs. Girls and boys enter school roughly equal in measured ability. Twelve years later, girls have fallen behind their male classmates in key areas such as higher-level mathematics and measures of self-esteem. Yet gender equity is still not a part of the national debate on educational reform. (p. 1)

The AAUW provides a glossy order form for this report. The form features a photograph of a classroom peopled with attractive girls and boys from many groups‹an African-American girl, an African-American boy, an Asian girl, a Caucasian boy. The irony is that there is one child in this photograph whom the schools are shortchanging, but this child is not a girl. This child is the African-American boy. This is the group in need of creative policy initiatives.

What is worth remembering is that boys used to be the group considered shortchanged by the schools. The idea that the schools shortchanged boys was part of the common wisdom through the 1970s. As Brophy (1985) reminds us:

Claims that one sex or the other is not being taught effectively in our schools have been frequent and often impassioned. From early in the century (Ayres, 1909) through about 1970 (Sexton, 1969; Austin, Clark, & Fitchett, 1971), criticism was usually focused on the treatment of boys, especially at the elementary level. Critics noted that boys received lower grades in all subjects and lower achievement test scores in reading and language arts. They insisted that these sex differences occurred because the schools were "too feminine" or the "overwhelmingly female" teachers were unable to meet boys' learning needs effectively. (pp. 115-116)

As this paper documents, girls surpass boys in some academic areas and boys surpass girls in other areas. Indeed, a far stronger case could be made for the view that "the schools shortchange boys" than the other way around. After all, it is boys who get consistently lower grades in school even though they score just as high or higher than girls on many standardized tests of achievement. This is strong evidence of bias against boys. It is boys who end up far more often than girls in special education classes for students with serious learning problems. It is males who are falling behind in college attendance. As recent survey research shows, it is boys , especially minority boys, who believe that teachers are not as apt to encourage them to achieve their goals or do their best (Harris, 1997, pp. 10, 13).

The AAUW has done women and the nation a service in drawing attention to the gender gap in science and mathematics and in encouraging an array of policies and programs designed to boost female performance in these fields. But the schools need to be equally concerned about the problems of boys. Boys mature more slowly than girls, for example, in areas like verbal skills. Late-maturing boys can be stigmatized as poor learners and assigned to "low-ability groups in the primary grades, especially in reading" (Halpern, 1997, p. 1098). Boys are also more active than girls and more difficult for teachers to handle. "Bright, bored, and rambunctious boys" have been diagnosed with attention deficit disorder and placed on drugs like Ritalin (Zachary, 1997, A1).

Neither girls nor boys nor the nation itself are served by politicized research and "noble lies." Major assertions in the AAUW report are based on research by David and Myra Sadker that has mysteriously disappeared. Evidence which contradicts their thesis that the schools shortchange girls is buried in supplemental tables obtainable only at great difficulty and expense. Such shady practices undermine public confidence in social science research. This damage done by the AAUW report will have repercussions that last far beyond the immediate issue of whether either girls or boys are shortchanged in the school.

Gender Differences in School Grades, Rank in Class, and Honors

If schools as an institution were shortchanging females, such gender discrimination should be easy to spot. Schools give clear and measurable rewards: grades, class rank, and honors. These rewards are valuable in gaining admission to a selective college or graduate school and in gaining a desirable job. Which group‹males or females‹receives a disproportionate share of the school's rewards?

From grade school through college, females receive higher grades and obtain higher class ranks. They also receive more honors in every field except science and sports.

Grades: That females receive higher grades in virtually every subject is undisputed. In reviewing the literature on gender differences in cognitive tests, for the flagship journal of the field, American Psychologist, Halpern (1997, p. 1102) points out that "higher grades in school, all or most subjects" is an area of unquestioned female advantage. Another recent, comprehensive review of the research literature on gender differences in school performance comes to the same conclusion:

Data from a wide variety of sources and educational settings show that females in all ethnic groups tend to earn higher grades in school than do males, across different ages and eras, and across different subject matter disciplines. Many researchers in past times and today consider this to be such an obvious fact that they treat it as axiomatic....Modern reviews of the subject are unanimous in their finding of higher grades for females (Dwyer & Johnson, 1997, pp. 128-129).

The female advantage in grades, while consistent, is not necessarily large. Among high school students who took the ACT in 1992, for example, the overall female GPA was 3.00; the overall male GPA was 2.89 (Willingham & Johnson, 1997, Table S-14). Even in mathematics and in science, female high school students who took the ACT got higher grades than males.

In college, females also receive higher grades than males, a pattern evident in national samples from the 1970s that continues into the 1990s . Table 1 shows the pattern.

Table 1: Women Get Higher College Grades But Differences Are Small
 

Source: From Women at Thirtysomething (p.114), by C. Adelman, 1991, Washington, DC: U.S. Department of Education.

Class Rank and Honors: Since girls receive higher grades in school, they should also surpass boys in class rank. This is exactly what happens. Examining gender differences in high school class rank and honors in a nationally representative sample from the 1970s, Adelman (1991, p. 3) makes this point, "No matter how one slices the high school class of 1972, women's mean class rank exceeded that of men by a minimum of 10 points." Caucasian women attained, on the average, the highest class rank (67th percentile), while African-American men attained, on the average the lowest class rank (44th percentile). African-American women ranked far higher (56th percentile) than African-American men.

The same pattern of female advantage in grades and honors shows up in the 1990s, in a nationally representative longitudinal study of the high school class of 1992 (NELS Second Follow-up, cited in Dwyer & Johnson, 1997, p. 139). In the academic arena, high school girls outdistanced boys in making the honor roll, in getting elected to a class office, and in receiving writing awards and other academic honors. In the academic arena, boys outdistanced women in vocational-technical honors and in awards in science and mathematics competitions.

While males are still ahead in gaining mathematics and science honors, females are making strong gains. From 1995-1998, close to 40 percent of the winners of the most prestigious science competition, the Westinghouse Science Talent Search, were female (Science Service, 1998). The Westinghouse Science Talent Search requires high school students to complete a project in science, mathematics, and engineering and submit a report communicating the results. The work goes on over many months, often with the assistance of a parent, teacher, or other researcher. The contest is notable for producing winners who later go on to win a Nobel Prize. Westinghouse finalists from the 1940s through the 1970s were overwhelmingly male. The number of females among the top 40 finalists has increased since the 1980s and is approaching parity (Table 2).

Table 2: Females Are Increasing Among Westinghouse Science Finalists

Source: Science Service, Westinghouse Foundation, 1998.

Gender Differences in Standardized Tests of School Achievement

Even though girls surpass boys in school grades, the schools might still be shortchanging girls if they are getting good grades but not learning as much as boys. Grades, after all, are based not only on how much students know but also on conformity to institutional demands, such as whether students follow the teacher's directions and turn in their assignments on time. Scores on standardized achievement tests provide a measure of school achievement less influenced by such extraneous influences as willingness to obey the teacher's directives.

On standardized achievement tests, females typically surpass males in writing ability, reading achievement, and certain other verbal skills while males surpass females in science and mathematics. In the general population of males and females, however, sex differences in achievement tests are typically small‹except for the big female advantage in writing.

The research literature on sex differences in achievement test scores is voluminous. Various studies use various standardized tests, for example, the California Achievement Test, the Iowa Test of Basic Skills, and the tests developed by the National Assessment of Educational Progress. Standardized test information is available in many different years in many different locations.

To make sense of this mass of information, contemporary researchers use a statistical technique called a "meta-analysis." (For those readers who prefer, a simpler version of the same basic pattern discussed in this section begins on page 15.) Essentially this technique offers a simple way to combine the findings from many different standardized tests, given to different samples in different years, and using different scoring systems. A statistical measure called the standard mean difference (D) summarizes and communicates the results across these studies. This statistic is easy to understand. Basically, "D " is the average difference between the test scores females receive and the test scores males receive in an area like mathematics achievement. "D" is calculated simply by subtracting the male mean across all these tests from the female mean across all these tests, which yields the average difference in test scores between females and males. This difference is then divided by a measure of the variability (average standard deviation) in the test scores of females and males.

Using D allows researchers to combine studies and to come up with a strong estimate of the average difference between males and females. If females and males do not differ on the measure of intellectual performance, then D is zero. A positive D indicates a difference in favor of females. A negative D indicates a difference in favor of males. By convention, a D of .20 to .49 is considered a "small" difference; a D of .50 to .79 is considered to be a "medium" difference; and a D of .80 or higher is considered to be a "large" difference.

In a comprehensive review of the literature on gender differences on standardized test scores, Willingham, Cole, Lewis, & Leung (1997) bring order to this complex and disputed mass of studies. They have created a data set focusing on the performance of large national samples of 12th grade students on standardized tests, with emphasis on the National Assessment of Educational Progress. The core of their data base consists of about 60 achievement tests grouped into 10 different academic categories, such as 1) Verbal-writing, 2) Verbal-reading, 3) Math concepts, and 4) Natural science.

In most academic areas, sex differences in achievement, where they exist at all, are "small" (Table 3). Females surpass males in writing skills, language use, reading, and study skills. Males surpass females in mathematics, science, and geopolitics, but the differences are too slight to reach the accepted criterion of a "small" difference except in geopolitics. The only gender difference approaching "medium size" occurred in writing skills, which favored females.
Table 3: Standardized Achievement Test Scores Are More Apt to Favor Females But Most Differences Are Small : National Samples of Students at Grade 12
 

Source: Adapted from Supplement to Gender and Fair Assessment (pp. 58-59) by W. W. Willingham and L. M. Johnson, 1997, Mahwah, NJ: Lawrence Erlbaum Associates.

In short, on these comparisons, across many different tests, not much difference occurs between males and females in achievement in standardized tests in the general population. Most gender differences are small and favor females more often than males. The only gender difference of medium size, writing abilities, favors females.

A simpler way of looking at the same, basic pattern is to examine male and female scores on the National Assessment of Educational Progress, tests given to a nationally representative sample of students to examine American students' performance in the four basic skill areas: reading, writing, mathematics, and science. At the end of high school, females vastly surpass males in writing abilities and reading abilities. Males surpass females in science and mathematics, but the male advantage in these subjects is far smaller than the female advantage in reading and writing.

Table 4: The Gender Gap Favoring Females in Reading and Writing Is More than Twice the Size of the Gender Gap Favoring Males in Science and Mathematics:
 

Source: From Digest of Education Statistics 1997 (Tables 107, 113, 118, and 126), National Center for Education Statistics, 1997, Washington, DC: U.S. Department of Education.

To put these sex differences in perspective, consider the difference in each subject between Whites and Blacks. At age 17, in reading, Whites surpass Blacks by 29 points; in writing by 32 points; in mathematics by 27 points and in science by 47 points (National Center for Education Statistics [NCES], 1997 b, Tables 107, 113, 118, 126). The enormous achievement gaps in America concern race, not sex.

Sex differences on achievement tests are small among high school males and females in general. But let us ask a different and equally significant question: Where is the talent? Do males or females dominate the top of a field? These are the conspicuous achievers who create cultural images of success. Among the top students in a subject area, a different picture emerges.

Among students at the top of the heap, gender differences in achievement test scores can be large and consequential even when only slight differences exist in the general population. In the top 10 percent of high school students, females surpass males in writing ability and reading achievement while males surpass females in mathematics, geopolitics, and science performance.

Using the same database combining the achievement test scores of 12th grade students, Willingham et al. (1997, pp. 80-83) examined the sex distribution among the top 10 percent of the students. The top 10 percent in a high school class is not a very select group, not the group that is apt to achieve national prominence in an area. Still, we can see the outlines of a gender problem emerging. In this top group, even in high school, males dominate the top group in science (7 out of 10); mathematics (almost 6 out of 10); history and civics (6 out of 10). Females dominate the top group in writing (7 out of 10) and in reading (6 out of 10). In other select groups, such as students who take the Scholastic Aptitude Test and the mathematics and science Advanced Placement Tests, men also score substantially higher than women, especially in areas like physics (Bae & Smith, 1997).

In short, differences in the performance of males and females in the general population are small, even in science and mathematics. But more males end up at the top in science and mathematics, among the most conspicuous achievers. Why?

One important reason has less to do with bias than with biology‹the greater variability of males on many human characteristics. Most of us have in our minds an image of the bell-shaped curve that comes from IQ tests‹a voluptuous bell curve with a generous middle and spreading extremes. We do not stop to consider that bell-shaped curves can take other forms. A bell-shaped curve with exactly the same average, for example, can be high and peaked. The bell-shaped curve in a male population tends to take the voluptuous shape with more males at the extremes; the bell-shaped curve in a female population tends to be high and narrow with fewer females at the extremes. The illustration below shows two such bell-shaped curves, exaggerated to make the point that two populations with the exactly the same averages can nonetheless have very different numbers of people at the extremes.

On many characteristics, the bell-shaped curve among males takes the voluptuous form: More males appear among the top talent and more males appear at the bottom of the barrel. As a consequence, males more often end up in the ranks of conspicuous achievers. As Willingham & Cole (1997) point out:

Greater male variability tends to work to the advantage of males at the top of the score distribution...More variable male scores exaggerate [emphasis added] any male advantage at the top....If male scores are more variable, there is less female advantage at the top than would ordinarily result from a higher female mean.
(p. 51)

In short, greater variability among males means that more academic stars, those at the extreme right end of the normal curve, are apt to be males. But this variability also means that more males will be at the extreme left of the normal curve, academic duds. This is exactly what happens.

Gender Differences in Special Education and Learning Disabilities

Males More Often Appear At the Bottom of the Barrel in Schools, Labeled as Impaired and Assigned to Special Education Classes

The over-representation of males in special education classes and in virtually every other category of emotional, behavioral, or neurological impairment is undisputed. In reviewing cognitive tests that typically show sex differences, Halpern (1997) summarizes this research:

Males are overrepresented at the low-ability end of many distributions, including the following examples: mental retardation (some types), majority of attention deficit disorders, delayed speech, dyslexia (even allowing for possible referral bias), stuttering, and learning disabilities and emotionally [sic] disturbances.
(p. 1102)

Far more boys than girls end up in special education programs. Even the AAUW report (1992, p. 19) underscores this point, "Boys outnumber girls in special education programs by startling percentages." Overall, twice as many boys as girls end up in special classes for the impaired (Table 5).

Table 5: More than Double the Number of Males Are Enrolled in Special Education Programs: Ratio of Males to Females
 

Source: Adapted fromThe Condition of Education 1997, (Table 46-2), National Center for Education Statistics, 1997, Washington, DC: U.S. Department of Education.

The AAUW report attributes this discrepancy to school bias: teachers discriminate against badly behaved boys. Mislabeling boys may indeed be part of the explanation. But many of these disabilities appear long before boys even enter school. Reviewing research on sex difference in learning disabilities (Nass, 1993) reports large differences in male-female ratios across many disorders, including such disorders as autism, which appear early in life (Table 6).

continued on page 2

Copyright 1998 Judith Kleinfeld , all rights reserved