Corrected: In an earlier version of this story, a quote from ETS researcher Jan M. Alegre was incorrectly attributed. Chris Gabrieli鈥檚 work on developing new measures of noncognitive skills was incorrectly described. He is working with a group of seven California school districts on new measures.
The nation鈥檚 premiere federal testing program is poised to provide a critical window into how students鈥� motivation, mindset, and grit can affect their learning.
Evidence has been building for years that these so-called noncognitive factors play a role in whether children succeed both academically and socially. Now, the National Assessment of Educational Progress, often dubbed the 鈥渘ation鈥檚 report card,鈥� is working to include measures of these factors in the background information collected with the tests beginning in 2017.
鈥淭eachers self-report spending 10 percent of their teaching time on noncognitive skills. That鈥檚 more time than students spend on any subject other than English and math鈥攎ore than they spend on arts, for example,鈥� said Chris Gabrieli, an adjunct lecturer with the Transforming Education project at Harvard Graduate School of Education and a co-founder of the National Center on Time & Learning in Boston.
鈥淚t鈥檚 not a question of whether schools are going to do more working on noncognitive factors,鈥� he said, 鈥渋t鈥檚 of whether we are going to have any instrumentation at all that lets us know which things are working and which things are not.鈥�
Researchers from the Educational Testing Service described the project at a symposium here last month at the annual conference of the Association for Psychological Science. The background survey will include five core areas鈥攇rit, desire for learning, school climate, technology use, and socioeconomic status鈥攐f which the first two focus on a student鈥檚 noncognitive skills, and the third looks at noncognitive factors in the school. These core areas would be part of the background survey for all NAEP test-takers. In addition, questions about other noncognitive factors, such as self-efficacy and personal achievement goals, may be included on questionnaires for specific subjects to create content-area measures, according to Jonas P. Bertling, ETS director for NAEP survey questionnaires.
Careful Wording
Researchers tested different variations of the questions with 140 students in grades 4, 8, and 12 from a representative sample of ethnic and socioeconomic backgrounds in a three-state area around the District of Columbia. Students were led through a 90-minute interview, in which they answered the survey questions and then discussed their individual thought process in responding.
For example, students across all grades showed no difference in how they rated themselves on questions that asked them about their mindset in different contexts, and they reported preferring general questions rather than those specifically about school, said Jan M. Alegre, an ETS researcher.
鈥淎 majority of students preferred questions that went beyond a simple yes or no, whether they did something or not,鈥� she said.
Small changes in phrasing make a difference in how well students respond to the questions. For example, 鈥�4th graders didn鈥檛 know what 鈥榯hinking abstractly鈥� meant. ... Students had difficulty describing what experiencing failure meant and what it meant to be committed,鈥� Ms. Alegre said. As a result, the researchers changed survey questions asking whether students had 鈥渆xperienced failure鈥� to 鈥渕aking mistakes鈥� and changed 鈥渃ommitted to goals鈥� to 鈥渃ontinue to work toward my goals.鈥�
The background questions will go through a third and final round of review in spring 2016, before the questions are administered beginning with tests in 2017.
Not for Accountability
Schools will not be judged based on the NAEP noncognitive measures of their students, but other such tests for accountability purposes may be on the horizon.
A coalition of seven California districts that have received waivers from some federal accountability requirements are developing a new accountability system, in which 40 percent of a school鈥檚 evaluation will take into account school culture and students鈥� social and emotional learning. Within the latter section, researchers are completing the field testing of growth mindset, self-efficacy and self-management, and social awareness measures with 9,000 students and 1,000 teachers. Mr. Gabrieli said the new measures are expected to be in place next year. He is helping to develop the new measures.
However, while poor performance on academic accountability measures can lead to sanctions in many districts, coalition schools with poor ratings for noncognitive skills will simply be paired with a higher-scoring mentor school.
As measures of noncognitive skills become more ubiquitous, Mr. Gabrieli said, it will be important to track disparities between students鈥� reports of their mindset and tenacity, and teachers鈥� observations of them.
鈥淚 have often seen in data collected in smaller samples this tendency for teachers to rate students on separate [questions] basically the same, as if they have one view of Johnny as a good kid or less-good kid,鈥� Mr. Gabrieli said. 鈥淚t鈥檚 hard to get teachers to follow the rules of 鈥榯his construct is about self-regulation and that one is about interpersonal skills.鈥� 鈥�
