In 1975 I spoke at a local high school's Career Day. I was invited because of a study my husband and I conducted under a grant from the National Science Foundation (NSF). The purpose of the study was to determine why there are relatively few women in mathematics. It received a modest amount of press and inspired the school's invitation.

According to our study, I told the students, lack of encouragement is a salient cause of women failing to pursue mathematics either academically or professionally. I appealed to their teachers and advisors to especially encourage girls who displayed a talent for mathematical reasoning.

"I'll be honest with you; I don't encourage girls to go into mathematics," a guidance counselor told me afterwards. "They wouldn't be good at it, and in any case what would they do with it?"

My memory of this incident surfaced in response to a recently issued study, that concludes that girls are born with less mathematical ability than boys. The authors, Camilla Benbow and Julian Stanley of Johns Hopkins University, claim that the difference in abilities is so great that it is hard to imagine that it is due entirely to socialization. They also "favor the hypothesis" that the difference in abilities, as exhibited in performance on the College board's Scholastic Aptitude Test (SAT), "results from superior male mathematical ability , which may in turn be related to greater male ability in spatial tasks;" (Science, Dec. 12, 1981).

"Do Males Have a Math Gene?" (Newsweek, Dec. 15, 1981) and "The Gender Factor In Math," (Time, Dec. 15, 1981) are just two of the memorable headlines generated by this study, which received nationwide media attention. I'm very concerned about the possible negative impact this may have on education. Educators may not jump to the conclusions as offered, but the Benbow-Stanley study could be used to justify differential treatment of boys and girls. It's a mischievous hypothesis to which I take exception.

Between 1972 and 1979 Benbow and Stanley conducted six "talent" searches to find "mathematically precocious youth." They chose seventh and eight graders who scored in the upper two to five percent on standardized mathematics achievement tests (math SAT). They found 10,000 children, 43 percent of whom were girls, and invited them to take the SAT. Stanley claims that the math SAT --generally given to high school juniors and seniors - serves as an aptitude test when given to seventh and eight graders.

But what does it mean when you give a seventh or eighth grader a college-level exam? How are we to interpret the results? We scarcely know what the test scores mean when they are made by high school students. Many educators now question the ability of the SAT to test mathematical reasoning, problem solving, and spatial ability. In recent years, this question has triggered a major controversy.

Another fault I find with the interpretation is that they equate "training" with "number and types of courses." The boys and girls in the study presumably had the same number and types of math courses. Does that mean that the boys and girls had the same "training?" I think not. There can be extraneous training and conditioning which has immeasurable effects.

For example, it's very much a part of our society for boys -- but not girls -- to be made to feel that they must prepare for a livelihood. That sort of "training" has its effects. Boys may not do well in math, but they know they must keep trying. Girls, on the other hand, may do well in math and then drop it for a very superficial reason. Sometimes, a girl interested and talented in mathematics will be discouraged.

A study of mathematically gifted students by Linda Brody and Lynn Fox of Johns Hopkins University showed that parents of gifted girls were less likely to view math or science as appropriate careers for their children. In some cases, girls consider it unfeminine to go into mathematics; it's considered a masculine domain. Unless we discourage such thinking, what is to stop it -- particularly in the publicity wake of the Benbow-Stanley study?

When I got my Ph.D. in 1957, there were 11 women in the entire United States who recieved doctorates in mathematics. That was five percent of the Ph.D.s in mathematics that year. As recently as 1975, when we did the NSF study, women earned 8 or 9 percent of the math Ph.D.s. By 1979, the figure rose to 13 to 15 percent. So there has been an increase in the number of women entering mathematics and attaining doctorates, and also in the number of undergraduate women majoring in mathematics. It would be very disappointing if the publicity received by the Benbow-Stanley study reversed that progress due to a preconception that women have less ability.

Benbow and Stanley shift the cause of observed differences in math scores to alleged male superiority in spatial tasks. But studies of spatial abilities have yielded ambiguous results, with girls sometimes scoring higher than boys. In addition, recent research does not support the genetic hypothesis that spatial ability is determined by an X-linked gene.

If we are to accept gender as a determining factor in math ability, is it any more ludicrous to extend that conclusion to other abilities? What if it becomes a trend? How far can such conclusions be extended?

Girls generally do better on verbal ability tests than do boys. Do we therefore presuppose that girls have a verbal gene? And since it is said that girls cry more easily than boys, should we therefore assume that girls have blue genes?

Let's consider what sort of action might be appropriate according to such conclusions. On the basis of girls' proven verbal superiority, we might discourage boys from taking English; it's such a feminine domain, like the kitchen.

Cheekiness aside, the point I'm making is quite serious. As I mentioned, I am greatly concerned about the impact the Benbow-Stanley study might have on education. Certainly anything that impacts education is going to mushroom in society. I don't think we can afford to discourage young people with talent in mathematics.

We need to improve the methods of teaching mathematics -- not only to produce mathematicians, but to increase the level of mathematical thinking, computation skills, problem solving, and spatial ability in a broader segment of the population.

Right now there is a tremendous need for high school mathematics teachers. A few years ago there was a glut; now we can't find them. Why? Because mathematically trained people are in great demand in many other, higher paying fields, such as computer science. This is a most practical reason to encourage girls with obvious ability in mathematics: it is applicable and intrinsic in many careers.

A 1973 study by sociologist Lucy Sells compared young men and women in the incoming freshman class at the University of California at Berkeley by their previous course-load in mathematics. Fifty-seven percent of the male students, but only eight percent of the female students, had completed four years of high school mathematics. Four years of mathematics is a prerequisite for freshman calculus. Fifteen out of 20 courses offered to freshmen required calculus and were, as such, closed to females in the study. The absence of calculus severly limited the careers open to these women.

After carefully reading and rereading the study, I find no conclusive evidence that mathematical or spatial ability is an innate trait, or that girls are born with less ability. Their findings do not support such conclusions.

In the April 10, 1981 issue of Science, seven letters disputed the interpretation offered by Benbow and Stanley. The letters are representative of a dissenting body of professionals --that objects to the Benbow-Stanley hypothesis. The Benbow-Stanley study, and the controversy it has generated, is reminiscent of the fruitless nature/nurture controversy over I.Q. It has the same ideological ramifications, except that now we compare the difference in scores between sexes rather than races. If we harness some of the energy now being expended in arguments over the nature and source of mathematical ability, we might improve mathematical learning for both sexes.