## Boys vs. Girls on the SAT

Adding Up the Differences Between Boys and Girls

by Mark Perrry
Published at The American
December 10, 2010

The College Board recently released 2010 SAT test results for college-bound high school seniors, and here are some highlights.

1. Boys scored significantly higher on the 2010 SAT math test than their female counterparts, by a difference of 34 points. This 30-point-plus male advantage on the SAT math test follows a pattern that has persisted since at least 1972.

2. For all SAT math scores of 580 and above (70th percentile and higher), male students outnumbered female students. As test scores increased by 10-point intervals from 580 to 800, the male-female ratio steadily increased, reaching a peak of 2.08 males per female for perfect scores of 800 (8,072 males vs. 3,887 females).

3. More females (827,197) than males (720,793) took the test in 2010. Adjusting for those differences in sample sizes, 1.12 percent of males scored a perfect 800 compared to 0.47 percent of females who did so, for an adjusted male-female ratio of 2.38 to 1.

Despite the strong evidence that boys continue to outperform girls year after year on the SAT math test, we hear statements like this from Professor Janet Hyde at the University of Wisconsin: “There just aren’t gender differences anymore in math performance. So parents and teachers need to revise their thoughts about this. … Stereotypes are very, very resistant to change, but as a scientist I have to challenge them with data.”

The SAT data, however, clearly suggest otherwise. So Hyde and her colleagues fall back on a rather weak statistical explanation of lower math test scores for females: “sampling artifact.” Because more girls now take the SAT test than boys, Hyde says, we are “dipping farther down into the distribution of female talent, which brings down the average score. That may be the explanation, rather than girls aren’t as good as math.”

With that background information, watch the discussion below on gender differences in math, reading, and writing, including a critical analysis of the “sampling artifact.”