A study out today suggests that California's high school exit-exam policy may be doing more harm than good for the state's lowest-performing students—especially those who are young women and students of color.
Implemented with the graduating class of 2006, California's exit test—known as the California Hlgh School Exit Exam or CAHSEE—has been controversial from the start. Proponents hoped the test would spur students to study harder, but opponents, in lawsuit after lawsuit, worried that an unintended consequence of the exams might be a drop in graduation rates among some of the state's most disadvantaged students.
The report posted online today by Stanford University's Institute for Research on Education Policy and Practice confirms some of the critics' worst fears. It shows that the exit exam led to an overall decline in graduation rates of 3.6 to 4.5 percentage points in the years after the policy took effect, yet without producing a strong effect on student achievement on other state tests.
Among females in the bottom achievement quartile, graduation rates fell by 19 percentage points after the high-stakes exam policy was put in place. That compares with a drop of 12 percentage points over the same period for male students with similar academic profiles.
Likewise, the poorest-performing black, Hispanic, and Asian-American students saw their graduation rates decline by 15 to 19 percentage points following the enactment of the exit-exam policy. The comparable graduation-rate decrease found among white students, in comparison, was a mere 1 percentage point.
"These are clearly troubling, and no one can be happy with a policy that is having such disproportionate effects," said Sean F. Reardon, the Stanford scholar who led the study. (For a look at what California's chief state school officer has to say about the results, see here.)
Just as intriguing, though, is the researchers' explanation for why the effects hit some groups of students harder than others: They chalk it up to "stereotype threat."
Stereotype threat, you may recall, is the idea that people's test performance can be artificially depressed if they are afraid they will confirm an unflattering stereotype about their racial or gender group by doing poorly. For example, women and African-Americans have been found to do worse on math exams after being asked to write their race or gender on their papers or after being told that their group typically scores low on an exam.
The Stanford researchers said they reluctantly fingered stereotype threat as a culprit in the post-CAHSEE graduation-rate declines after ruling out most other possibilities. Thinking that minority students might be attending schools with poorer resources, for example, they analyzed data for subsets of students in the same schools. The patterns stayed the same. They also eliminated possible bias in the tests themselves as an explanation after reviewing studies on that topic.
But, when the research team examined students' previous scores on other state tests, they turned up some evidence that minority students and women had underperformed on particular sections of the state exit exam. Women fared worse than their earlier performance might have predicted, for example, on the math portion. Asian students did worse-than-expected on English-language arts.
"It's a very specific pattern, so it's hard to explain it based on effort," Reardon says. "That's what persuaded me that there was a stereotype-threat story going on, that we have this other set of tests to compare it to, and they don't show the same pattern."
Look for more on this interesting study in next week's print edition of Education Week. I'm doing a story wrapping up the new findings from California with those from a handful of the nearly two dozen other states that now require students to pass an exit exam in order to graduate from high school.