College administrators typically rely on standardized tests to decide which students should proceed directly to college-level classes and which students should start in remedial courses. But a new Alaska study adds more weight to a growing body of research showing that standardized tests are a lousy way to make this decision, and that it would be much better to look at students’ high school grades instead.
The Alaska study, conducted by a regional research laboratory funded by the U.S. Department of Education, found that SATs, ACTs and the placement tests used by the University of Alaska were all poor predictors of how a student might do in a college-level math or English class. Many students who did well on these exams bombed their college classes, and vice versa. Instead, the researchers found that if college administrators had simply looked at the students’ high school GPAs, they would have done a much better job at figuring out who needs to relearn high school material and who doesn’t.
“We definitely should be including GPAs when assessing college readiness,” said Michelle Hodara, the lead author of the study and a senior researcher at Education Northwest. “We found the same thing that community college researchers and practitioners are finding, that high school GPA is a really powerful measure of college readiness, even for students who want to earn a four-year degree.”
At the University of Alaska, half of all first-time freshmen pursuing a four-year bachelor’s degree are assigned to developmental math. About a third are told they need developmental English. Statistics like these aren’t uncommon across the United States. All this remedial education, during which students aren’t earning college credit but are still paying tuition, is one of the big reasons that student debt is rising, college graduation is taking longer and students are getting discouraged and dropping out entirely.
Still, most colleges still rely on tests for remedial course placement. And it’s easy to sympathize with college administrators who want to use an objective test. After all, students attend different high schools and take different classes. Some are rigorous. Some aren’t. Some teachers give easy A’s. Others are tough graders. Why would it be fair to let the student who took easy classes waltz into a college course, while the student who struggled under a demanding teacher is dispatched to a remedial class? Wouldn’t be better to figure out exactly what students know?
In theory, yes. But in practice, often students haven’t taken a math class in more than a year, and so the material isn’t fresh in their minds when they take the placement test. Or they haven’t reviewed the material just before the exam. But once they’re immersed in math classes again, it comes back to them, and they don’t really need to repeat an entire year of algebra.
Hodara argues that what students know, or “content knowledge,” isn’t the most important thing anyway. She says that GPAs capture important non-cognitive skills that tests don’t. “It’s likely that if you have a high GPA, even if you’re in an ‘easy’ class, you likely showed up and turned your homework in, and did things that are important for college readiness and success,” said Hodara.
In the study, Hodara and her colleagues looked at the records of all full-time students at the University of Alaska from 2008 to 2012. She found that students with a 3.0 high school GPA, or a B average, were at least 25 percentage points more likely to pass college-level classes with a C or higher than students who had a 2.0, or C average, in high school. By contrast, Hodara said there was only a “trivial” association between college entrance exam scores, or the placement exam scores, and students’ college course grades.
The Alaska data also revealed a group of savvy students who were originally assigned to take “developmental” classes but somehow managed to work the system and bypass them. They went straight into credit-bearing college courses, and 60 percent of them passed.
“They should never have been placed into developmental education in the first place,” Hodara said, citing this as evidence that many students are being mis-assigned.
Unfortunately, Hodara’s study didn’t calculate exactly how many students had high enough GPAs and yet were put into developmental classes. But the American Institutes for Research, a nonprofit research company, previously found that a 3.0 or a B average in high school is predictive of college completion. Likely, that’s also a good threshold to assume that a student is ready for college courses.
Some community colleges have begun to take notice of the research and changed their placement procedures. For example, a group of a dozen California community colleges began including GPA in their decision-making process in 2011. They’re placing more students into college courses from the get-go, and more students are passing them.