In an opinion piece published by the New York Times, Mindset Scholar Angela Duckworth argues that social-emotional competencies and habits of mind are critically important, but measures of these skills should not be used for low- or high-stakes accountability in education.
Evidence shows that skills such as self-control and social intelligence contribute to student outcomes above and beyond cognitive ability. Moreover, research shows they are learned: the extent to which students develop social-emotional competencies and habits of mind depends on their experiences at both home and school. And we know certain strategies can help students build these skills.
Finding ways for parents and teachers to provide feedback to students on their social-emotional development is critical, Angela argues. She and her colleagues are still in the process of developing tools that can be used for this purpose. But there are certain limitations to the measures themselves—and what happens when measures are used as metrics in accountability systems—that raise serious concerns about efforts to use measures of social-emotional skills to compare, reward, or punish schools.
One issue is ‘reference bias.’ The responses of students, parents, or teachers to questions that ask them to rate students’ behavior are shaped by the norms in that particular family, classroom, or school. In other words, the same ‘displayed’ level of self-control can be rated as higher or lower by different observers in different contexts. This is an issue even when the measures are used for ‘low-stakes’ accountability, when higher- and lower-scoring schools are paired up to learn from each other. Scientists have developed performance tasks for research purposes that avoid this type of bias, but they can be time-intensive and are subject to ‘practice effects’ if students take them repeatedly.
Second, tying measures to external rewards and punishments—such as in ‘high-stakes’ accountability—creates incentives for cheating. As Angela writes, “while carrots and sticks can bring about short-term changes in behavior, they often undermine interest in and responsibility for the behavior itself.” Surveys used in research on social-emotional skills do not have these same incentives attached to them.
The Mindset Scholars Network issued a brief last year that reflects these concerns. Moreover, evidence suggests that measures other than surveys developed for social-emotional research can aid policymakers and practitioners who want to focus attention on building students’ social-emotional skills. These include more objective measures that directly predict student outcomes: student attendance, discipline referrals, and course failures. School climate surveys may be able to provide actionable data to complement these types of measures that are already collected by schools and districts.
As Angela concludes, social-emotional skills are key to students’ long-term success and well-being, as is the practice of giving them feedback and coaching as they build these skills. But measures of these skills used for scientific research or student feedback should not be used as metrics in school accountability.
tags: measurement & assessment