A.C.T. Scores Rise, But Gender Gap In Peformance Worries Some
Average scores on the American College Testing Program entrance exam rose this year for the first time since 1989, the Iowa City-based firm announced last week.
But A.C.T. officials downplayed the signficance of the tiny rise--a tenth of a point on a scale of 36--and standardized-test critics voiced alarm at test results showing a slight increase in the "gender gap'' between male and female students.
The A.C.T. is a multiple-choice exam divided into four sections: English, mathematics, reading, and science reasoning. Approximately 875,000 students in the high school graduating class of 1993 took the exam.
Over all, composite scores rose to 20.7 for the class of 1993, after remaining constant at 20.6 for four years.
Nevertheless, said Kelley Hayden, a spokesman for the testing organization, "We didn't want to make an issue of [it], because it has some meaning, but it's not the main thing.''
Mr. Hayden also expressed concern at the results of the A.C.T.'s annual survey of students about whether they had taken a core curriculum in high school, which it defines as at least four years of English, three years of math, three years of social studies, and three years of science.
Among test-takers in the 1993 class, still only about half--52 percent--reported completing the core classes, although that level was up from 44 percent four years ago.
"What continues to surprise us a little is the number of students who indicate that they are intending to go on to college, to be an engineer or a doctor or a lawyer,'' Mr. Hayden observed, "and you look at what they've taken in high school, and they're not taking the kinds of courses they ought to be taking to prepare themselves for that.''
The spokesman saw brighter news, however, in gains charted by minority-group members, especially in conjunction with increases in the number of minority test-takers who had completed a core curriculum.
"Greater numbers of minority students are getting more of the kind of education they need in high school to prepare themselves for college,'' he said.
Scores for African-Americans, Mexicans, and Asian-Americans increased by 0.1 over last year, while scores for Native Americans increased by 0.3. Scores for Hispanics remained constant.
Test or Society Bias?
The A.C.T. came under fire from a group critical of standardized tests, which accused it of ignoring a slight increase in the gender gap.
For the class of 1992, the mean score for female test takers was 20.5, while the mean score for males was 20.9. This year, females' scores stayed the same, while males' scores increased by 0.1, to 21.0.
The A.C.T. "regularly underestimates the abilities of females,'' charged Cinthia Schuman, the executive director of the National Center for Fair and Open Testing, or FairTest.
"Females continue to do better in both high school and college,'' she argued, "yet perform less well on timed multiple-choice standardized tests such as the A.C.T. or S.A.T.''
This assessment format "works to the disadvantage of females who excel in classroom work where real thinking and writing and research is involved, but do not do as well in a quick-paced multiple-choice exam,'' Ms. Schuman said.
Mr. Hayden countered that the gap in composite scores is a result of girls' lower scores on the math and science-reasoning sections.
"Although the situation for women in science and math is improving, still the fact remains that not enough female students are taking all of the math courses and all of the science courses,'' he said. "The bias is not in the test, the bias is in the system, in society.''
Ms. Schuman responded that girls take as much basic algebra and geometry as boys, "which are generally all the things you need for those tests.''
She also pointed to research showing that girls who are enrolled in the same classes and earn the same grades as boys still do not perform as well on the tests.