AERA: Measuring Persistence and Self-Control Through Tasks, Not Tests
San Antonio, Texas
Much research on social-emotional learning relies on surveys to gauge students' self-control or academic persistence, but the skills only matter if they make a difference in what students actually do.
That's why researchers at the annual conference of the American Educational Research Association here are exploring ways to replace surveys of social-emotional skills with performance tasks that may better predict both short-term goals like learning in the classroom and, in the long run, even whether they complete college.
The problem with, for example, asking students to report whether they have good self-control or grit is, the students with the most stringent standards are likely to judge themselves as worse than they are, and more lax students will judge themselves as better than they are. When Joseph O'Brien of the University of Texas at Austin and David Yaeger of Stanford University looked at nearly 1,300 high school seniors in 15 schools that specifically focused on grit and self-control, they found the schools where students reported the lowest scores for self-control and grit were actually the ones where students later had the highest rates of college attendance and persistence.
It's the Dunning-Kruger effect in action; if you are surrounded by goof-offs and you at least turn your homework in on time most of the time, you probably feel pretty diligent. But if you are a super-hard-working kid in a school chock full of overachievers, you often feel more like a slouch by comparison. When looking just within each school, where students were comparing themselves against the same standard, Yaeger and O'Brien found that students' self-reports did predict their later college persistence—but comparing between schools gave the wrong result. Asking students in a later study to compare themselves specifically to the best or worst students in their school—or the world—helped somewhat, but not enough to significantly improve the results.
The Proof Is in the Performance
So how can educators make sure they aren't measuring students' own biases (or those of their teachers or parents)? By giving them the opportunity to live up to their own expectations. Marissa Hartwig and her colleagues at Washington State University-Pullman and the University of California-Davis asked about 660 high school students to complete a set of challenging math problems after tutorials in algebra, geometry, and trigonometry, telling them it would help them practice and cement their understanding of the material—the typical reason teachers tell students to finish their homework. But the researchers also made sure the students had plenty of distractions available, in the form of video games and videos that students were allowed to use whenever they wanted.
"We're interested in what students are actually doing during their study time," Hartwing said.
They found that students who initially reported having a higher growth mindset—the belief that skills are not inborn, but can be improved over time with practice—did better on the subsequent tests. But that effect was mediated by how long students stuck it out during the practice sessions.
"Pesistence is a key reason why mindsets matter," Hartwig said. "Having growth mindset is associated with higher sustained learning, as is longer persistence. Students' time on task can provide some insight into students' willingness to persist."
In a separate study, Maria Cutumisu of the University of Alberta, Canada, similarly found that high-performing middle school students in Chicago and New York were more likely to ask for critical feedback on a poster project, and to revise it when they had the opportunity, than students who were lower-performing.
Students' academic persistence can depend on the subject, though, cautioned Brian Galla of the University of Pittsburgh and his colleagues. "Without exception, students recognized that academic work was more important to their long-term goals than anything else they could be doing ... and yet at the same time they were almost desperate to be doing almost anything else," Galla said. "Students were less happy, less confident, and they were far less intrinsically motivated when they were doing academic work than anything else ... even sleeping."
Like Hartwig's group, Galla and his colleagues asked 900 high school students to complete a series of "extremely boring" but skill-building math problems "designed to simulate the real-world choices students must make in their day-to-day lives" around academic diligence.
How long students persisted in the practice problems, they found, predicted not only how likely they were to graduate high school six months later, but students with longer persistence also had increased odds of being enrolled in a full-time degree program after high school.
Interestingly, the researchers found in a follow-up study of more than 300 8th graders, persistence varied depending on whether the practice questions were math, spatia,l or verbal tasks, with students able to persist significantly longer in the verbal questions. Taken all together, the three time-on-task scores predicted the students' end-of-year grade point average, but Galla noted that, "Doing math homework and English homework were about as important to the students, but students found English homework more enjoyable than math homework. Math was reported as more mentally effortful than English homework. This can help to explain this difference between math and reading grit tasks."
For more on using performance tasks, see www.noncog.org.