How learning happens.
By Anya Kamenetz
March 3, 2016
"The enthusiasm is getting ahead of the science."
-- Angela Duckworth, Ph.D.
In just a few short weeks, students in California will be taking high-stakes tests. But the tests won't just cover math, reading and science. Students will also be responding to survey statements like "I usually finish what I start," or "I can do anything if I try."
A group of big-city districts there is among the first to try to measure students' self-control, empathy and other social and emotional skills — and to hold schools accountable for the answers.
The new federal education law requires states to include at least one non-academic outcome in their accountability formulas, so these kinds of tests are likely to become more common nationally. The Nation's Report Card will be asking questions like this next year; so will the international PISA test.
- To Measure What Tests Can't, Some Schools Turn To Surveys
- Nonacademic Skills Are Key To Success. But What Should We Call Them?
- A Key Researcher Says 'Grit' Isn't Ready For High-Stakes Measures
But researchers I've talked to describe a moment of painful irony, where ideas they've advocated for years are catching on and being taken mainstream — before they're ready.
"The enthusiasm is getting ahead of the science," Angela Duckworth, the psychologist and MacArthur Fellow most associated with the concept of grit, told me in May. She has since resigned from the advisory board of the group that's working to test these qualities in California.
The science shows they're important things to teach and learn. Students who believe they can do better with more effort, who try harder, who can delay instant gratification and control their impulses, who take feedback well and know how to work on teams, are likely to become happier, healthier, more successful adults.
What's missing right now, though, is a consensus on how best to cultivate those qualities.
(The problem is inherent in the slippery language we use to describe them. If social-emotional qualities are "traits," then they might be fixed, even genetic. If they're "skills" or "habits" then, it stands to reason, they can be coached or taught.)
And researchers definitely disagree on the best way to measure them. Asking students to take surveys, or complete certain tasks, are powerful tools in an experimental, research context. But they're not as robust when applied on a wide scale to measure the performance of hundreds of thousands of students, or as a driver of high-stakes decisions like whether a student should move on to the next grade.
"We feel like people will read about something designed for research, and immediately assume that you can start grading 4th-grade teachers on [their students' answers]," Duckworth said. As examples, Duckworth cited the famous "marshmallow test" and her own "grit scale."
The marshmallow test is Stanford psychologist Walter Mischel's famous experiment, where 4-year-olds were told they could have one piece of candy right now, or they could have two pieces in 15 minutes. The ones who managed to wait turned out to be much more successful adults.
The grit scale is a 12-item checklist that Duckworth developed that features self-ratings on items like like, "I usually finish what I start."
The grit scale is powerfully predictive in experiments. But if schools and even teachers are graded on their students' replies, it would be extremely easy to nudge students to give the right answer on a survey.
Same problem with the marshmallow task: you can coach children on successful strategies for waiting (like singing a song to distract yourself.) But will that really improve their long-term outcomes?
Chris Gabrieli, leader of the group behind these changes in California, told me that measurements, with consequences attached, are the only way to move the needle on nonacademic skills.
"Until the education system begins to value these attributes, they won't focus on them," he says. "That's why we started with measurements."
He points out that in the case of California, schools where students score low on these attributes won't be punished. Instead, they'll be paired with higher-performing schools for mentoring and support.
The researchers who are raising the alarm have one negative historical analogy in mind: Self esteem.
In the 1970s and 1980s, insights into the importance of high self-worth led to a widespread school culture that handed out praise willy-nilly and gave awards for participation.
This misapplication of research did little to raise achievement. Experts like Jean Twenge have argued that it also led to skyrocketing narcissism among younger Americans.
"When people start thinking, 'I'll make the kids feel good and they'll learn,' that's how the self-esteem movement gains its traction," Stanford's Carol Dweck, most associated with the concept of growth mindset, told me. "People think any intuition, whether tested or untested, is valid."