ACT Scores and Achievement Gaps

The Washington Post has a story on ACT scores:

New results from the nation’s most widely used college admission test highlight in detailed fashion the persistent achievement gaps between students who face disadvantages and those who don’t.

Scores from the ACT show that just 9 percent of students in the class of 2017 who came from low-income families, whose parents did not go to college, and who identify as black, Hispanic, American Indian or Pacific Islander are strongly ready for college.

But the readiness rate for students with none of those demographic characteristics was six times as high, 54 percent, according to data released Thursday.

“That kind of shocked us,” ACT chief executive Marten Roorda said. “We knew it was bad, but we didn’t know it was this bad.”

The analysis of “underserved learners” was a first for the ACT, which is one of two major tests students can take to apply to college. The other is the College Board’s SAT.

In recent years, both tests have found major disparities in college readiness among students in the Washington region and around the country. Roorda lamented that these gaps have persisted despite efforts to improve schools under the banners of No Child Left Behind, Race to the Top and other national initiatives.

“You could argue that those investments should have made a clearer difference,” he said, “and that’s not what we’re seeing.”

Actually, there have been a lot more initiatives, local and national, than those Roorda mentioned, and they go back decades. I remember, for example, when busing was expected to reduce gaps.

More detail appears to be available in this report by ACT..

Since the Washington Post only looks at “underserved learners,” for completeness, it helps to know the entire distribution. Table 2.4 of the ACT report indicates that that Asian students get the top scores, on average, followed by white students, followed by those who decline to state their race, followed by those of two or more races.

I also notice a bit of a gap between males and females – males do better on math and science, females on English and reading. So this report is chock full of the same racist stereotypes we have seen for decades. How do we get rid of this persistent gap in outcomes? And I think we can agree we want to do it in a non-harmful way. For instance, we don’t want to reduce the achievement gap by harming the performance of Asian students. Anyone have any realistic suggestions? A realistic answer will, of course, be one that is implementable, and which doesn’t contradict data that has come up in the decades in which society has been trying to deal with the issue.