ACTIVITY, an ACT publication. ACT homepage
 
AUTUMN 2003   Volume 41/Number 3 
 
 

Progressive Testing Programs Challenge Notion of ‘Gender Bias’

This is a test:
Who earns better scores on college admissions tests?

A. boys
B. girls
C. neither boys nor girls

The answer is C, neither boys nor girls. Surprised?

Don’t feel bad if you are. For years, national media stories have talked about a “gender bias” in college admissions testing. The stories noted higher scores for boys and suggested the cause to be a bias in the tests.

Now, thanks to progressive ACT Assessment® programs in Colorado and Illinois that administer the test to all public high schoolers, there is a new story to report. The differences are due not to a bias in the test, but to the composition of the groups who take it.

The students who traditionally took the ACT Assessment were “self-selected,” that is, just those interested in applying for college admission. For the last two years, Colorado and Illinois have administered the ACT Assessment to all public high school students. ACT researchers now can compare the scores of the self-selected group (that is, the national ACT Assessment sample) to homogeneous state samples in Colorado and Illinois (that is, samples that include virtually all public high school students in those states). 
“When all students take the ACT Assessment, the average Composite scores for males and females are virtually equivalent.”
—Nancy Petersen, vice president,
   ACT research

“Although males tend to earn higher average Composite scores on the ACT Assessment than do females, this difference is due to a self-selection factor that disappears when the test is administered to all students, as in Colorado and Illinois,” said Nancy Petersen, ACT’s vice president for research. “When all students take the ACT Assessment, the average Composite scores for males and females are virtually equivalent.”

The comparison shows that greater percentages of girls than boys choose to take the ACT Assessment. The group of girls, therefore, represents a broader spectrum of achievement, including more lower-achieving members than does the self-selected group of boys. When the average Composite scores of that larger group of girls are compared to the boys’ average Composite scores nationally, the average scores for the girls are lower.

The differences by gender disappear when all students are tested.

Previous Article « Autumn 2003 Index | Top of Page » Next Article

 

ACT Home | Contacting ACT | Site Index

© 2008 by ACT, Inc. Terms of Use and Privacy Policy.