Professional CornerIt is a common belief that boys and girls are equally smart. We raise our daughters to believe they can compete in the business world with any man, and succeed. We teach them to believe they can attend any college, competing for grades and clubs, and scholarships. Yet on average, more boys score high on college entrance tests. Here may some of the reasons why.
Gender and brains
By Dan Seligman
A fortnight ago I seized on some material in a new book - The g Factor, by psychologist Arthur Jensen - to demonstrate that men and women are, on average, equally brainy.
The new line, fortified by another recent book, is that while the sexes are indeed equal on average, this is not exactly the whole story. Our text this time is Gender and Fair Assessment by Warren Willingham and Nancy S. Cole, both eminences at the Educational Testing Service: Cole is President of ETS. Their prose is sometimes klutzy and jargonish, and a fellow still trying to get used to "parenting" could have done without any references to "studenting".
Still, the book bears an important message. When I saw its title, I assumed it would be an exercise in political correctness. I was wrong.
ETS creates the exams that get you in or keep you out of the colleges and various graduate schools you would like to attend. They are, therefore, endlessly controversial.
Two years ago ETS was taking heavy flak from the Office for Civil Rights in the U.S. Department of Education for its Preliminary Scholastic Assessment Test. The PSAT is used to identify prospects for National Merit Scholarships. Alas, the darned thing was turning up more boys than girls. Instead of vigorously defending the fairness of the test, ETS meekly, weakly and maybe also prudentially - after all, it is a federal contractor - agreed to change the PSAT.
The new version includes some tests of writing skills on which girls are more or less guaranteed to do better than boys. The education department says it hopes that with the changes the boy-girl ratio, which had been around 56-44 among the merit scholars, will move to something around parity. Unless it does, the Feds warn, additional changes will be required.
But here is the fascinating part. Having caved in to the education department, ETS produced an exhaustive report that is, in effect, the defense not made two yours ago.
Based on more than 1,500 studies of 400 different tests, Gender and Fair Assessment explains why it is normal and natural when boys are over represented among the top scorers in "high-stakes tests" (an ETS phrase) like the PSAT. Anybody interested in an excellent summary of the argument can get it in The ETS Gender Study at http://www.ets.org/.
How can it be that men and women are on average equally intelligent but more men score at the highest levels? The ETS answer: Male scores are more variable. That is, males are over represented among the gifted - and among the retarded.
This phenomenon of male variability is discernible on just about all measurers of cognitive ability. Whether the differences are a biological or socially - driven phenomenon, or some mixture of the two, is not clear, but there is no doubt that the differences exist, and powerfully affect test scores.
The book includes a striking table showing results among high school seniors for 154 different kinds of ability, from "verbal-writing" to "mechanical/electronics". In 10 of the 15 categories, and especially in those dominated by verbal skills, girls on average do better than boys. But in all 15 the boys' scores are more variable than the girls'. This guarantees that as you approach higher cutoff levels, males will increasingly tend to dominate the totals.
The ETS study shows that in the population as a whole, boys would constitute about 55% of those in both the highest and lowest 10%. Among college-bound kids, the effect is, inevitably, more striking. Boys in this group are 10% more likely than girls to end up at the highest (above 750) verbal level. On the math test, boys are almost three times as likely as girls to be at that level.
The proposition that guys are more variable is not a recent discovery. The phenomenon was resoundingly demonstrated in Scotland in 1932, when the government gave IQ tests to every child who, on the designated dates, was between 10 2 and 11 1/2-87,000 kids in all. No exceptions were allowed, except for the blind or deaf. The results, analyzed over many decades, show (a) no average ability difference between lads and lasses, and (b) greater variability among the lads. Not much different from what the ETS discovered six decades later.
The ETS turnaround on boy-girl testing is intriguing on several counts: First, it obviously took a lot of guts. To squarely state that males will be over represented in gifted groups is as politically incorrect as you can get. The woods are full of psychologists anxiously averting their gaze from the fact of greater male variability.
Finally, it seems possible-barely-that in going on record about the variability data, ETS will inhibit the Department of Education from pushing for additional sex quotas in test scores. As I write, final data on the new, revised 1997 PSATS are not yet available, and it is unclear whether the sex ratios at high levels will be close enough to modify the Office for Civil Rights.
I asked Nancy Cole what would happen if the new writing test did not push more girls into the higher scores. She said she wasn't sure, but that ETS had made no commitments about quotas. She then added, in a mixed metaphor that would hurt her in some writing tests: "We're stepping on so many different toes, we just decided to let the chips fall."
Forbes Magazine, April 20, 1998
CONTENTS (except as noted) ©2003-8 by Pediatric Services
Corporate Office in Morro Bay, California (San Luis Obispo County)
Click here to ask a question.
DESIGN ©2003 by William Blinn Communications
Worthington, Ohio 43085
Articles written by Pediatric
Services staff are copyright by Pediatric Services.