The Washington PostDemocracy Dies in Darkness

Now some schools are testing kids for their ‘grit’ and ‘joy’ levels. Really.

Tony Roddy talks to members of the Joliet Boys and Girls Club about his walk across the United States during a presentation in Joliet, Ill. (Lathan Goumas/The Herald-News via AP)

In 2012, I published a post with this headline, “Sick of grit already.” Here it is 2016, and not only has “grit” become one of the watchwords in the education reform debate, but now some powers that be think they can teach it, measure it and test it.

Despite the fact that there is no consensus definition on how to do any of those things, the U.S. government, in 2017, is going to start collecting collecting data from students about their individual “grit” levels by asking those who take the National Assessment of Educational Progress to rate their own level of grit. Also their “desire for learning.”

And now, as this New York Times story says, schools are trying to find ways to measure and test grit, joy and other non-academic attributes. It says in part:

…starting this year, several California school districts will test students on how well they have learned the kind of skills like self-control and conscientiousness that the games aim to cultivate — ones that might be described as everything you should have learned in kindergarten but are still reading self-help books to master in middle age.
A recent update to federal education law requires states to include at least one nonacademic measure in judging school performance. So other states are watching these districts as a potential model. But the race to test for so-called social-emotional skills has raised alarms even among the biggest proponents of teaching them, who warn that the definitions are unclear and the tests faulty.

The “grit” fad is part of the broad debate about social-emotional learning, and how to provide holistic education to students, but even some proponents of SEL are concerned about where things are headed. The Times story notes that Angela Duckworth, the MacArthur fellow who has been at the forefront of the social-emotional learning movement and is publishing a book titled “Grit” in May, has withdrawn her involvement in the California project. “It is a bad idea,” she told the Times.

There are some educators who don’t agree, including Courtney Smith, lower school principal at Boston Collegiate Charter School in Massachusetts, who wrote in the Times that her school already tests students for social-emotional skills and she finds the admittedly flawed data useful.

For each student we know the following details within the first month of school: their level of trust, if they feel connected to others, if they are hopeful, if they feel like they can achieve. We receive data — however flawed — that typically takes years of relationship building to get from individual students during daily class time.

Other educators aren’t buying that argument. Nancy Carlsson-Paige, an early childhood education expert who is senior adviser to the nonprofit Defending the Early Years, said in a statement:

Testing children’s social and emotional skills is a bad idea.  These skills are crucial to school success and life long happiness—we’ve seen this through many research studies. But skills such as self and social awareness, managing emotions, developing empathy, forming positive relationships, and learning conflict resolution skills grow over time in children and from the inside out.  They develop in children as the result of interactions with others in classrooms that foster these skills through the curriculum, relationships, and activities specifically designed to encourage social and emotional skill building.
Research shows that reward systems can influence social and emotional behavior, but the learning does not last once the rewards are removed.  We want children to be kind and feel empathy for others even when the teacher isn’t looking or the promise of earning points isn’t there.  Research has also shown that self reporting does not match up with actual behavior.  Most importantly, we learn from moral development theory that the more we try to control children from the outside, the less they learn to regulate themselves from within.
Building skills for social and emotional awareness and skill should permeate every classroom and be encouraged in every child.  It’s  essential for their success in school and in life.  But testing these skills will only undermine that vital goal.

That’s just one reaction from the Defending the Early Years website; you can find the others here. Education historian and activist Diane Ravitch wrote in the Times as part of a discussion about testing grit:

Teaching children to persevere, to complete their work on time, to act courteously toward others, to accept success and defeat with equanimity is part of the everyday life of teaching and learning. It is not a separate subject. Testing whether children are experiencing joy or learning “grit” is sheer nonsense. If by “grit” we mean resilience, that is best taught informally, in the classroom, at home, on the playing fields, in the hallways, in the lunch room.

And here’s a post by two University of Pennsylvania professors of education — Joan Goodman and Sigal Ben-Porath — about the serious problems with attempting to teach and measure grit and other non-cognitive attributes.

By Joan Goodman and Sigal Ben-Porath

While educators are rejoicing over the lighter test burden for public schools in the re-enacted Elementary and Secondary Education Act (ESSA), they may soon be facing a new requirement: the learning and testing of non-cognitive (also sometimes called social-emotional) attributes. This domain is an ascending educational priority having already found its way into the Program for International Student Assessment tests (PISA) and the National Assessment of Educational Progress.

Measuring and teaching the non-cognitive has strong attractions. Advocates believe it will boost academic success, since some identified non-cognitive traits have been linked to better test scores. Further, it identifies qualities of potential importance to a well-lived life usually ignored by schools. It validates students who are strong in such areas but weaker in academic accomplishments. Best of all is the anticipated bonus of off-loading discipline to students’ inner controls, thereby alleviating a constant stressor for teachers and students.

Yet the intrusion of non-cognitive teaching and testing into schools is not a prospect we welcome. Our opposition is based on two claims: teaching and testing this loose amalgam of traits is impossible to do well and would be undesirable even if feasible.

It cannot realistically be done.

Initially problematic is the compound concept termed non-cognitive. It can include personality (usually understood as underlying dispositions), social-emotional attributes (usually referenced in behavioral terms), “character” (traits such as grit and self-control), and “21st century skills” (whatever is required for success in the information age). Nothing but confusion can come from lumping together this non-specific array of qualities that exempts only strictly academic performance.

Second, the qualities in question are unstable even in adults, and more so in children. Children are both developing and erratic. Their behavior and motives differ from one setting to another, one classroom to another, one year to the next. A child may be sullen and reserved in English but ebullient and joyous in Science. This known inconsistency suggests that the best way to change a child is to place her in a setting that will elicit or suppress qualities of concern. If you want to encourage leadership, give the child an opportunity to lead. Changing a setting, as opposed to more didactic teaching, would require offering multiple interest-based activities, not a direction encouraged by advocates of non-cognitive teaching and testing.

Third, assuming both an unlikely agreement on the selection of non-cognitive aptitudes and the stability of these aptitudes, there remain technical measurement problems. Currently, evaluators rely on self- or teacher-reports, both notoriously unreliable and subject to various distortions (halo effects, attempts to please the tester, etc.). Given the numbers of children to be tested, setting up experiments, instead of administering questionnaires, to observe if Johnny displays more or less self-control across various situations would be both phenomenally costly and inaccurate.

Fourth, teaching these traits is misaligned with the current preparation of teachers, and outstrips their expertise. Teachers can admonish and praise, model conscientiousness and kindness, perhaps even extinguish or encourage particular actions; they have long done so. But the broad dispositions included in the personality and social-emotional domains are often rooted in the home and larger social environment, in genetics and family history. They do not change easily. Given the difficulties therapists face in altering the attributes of children referred for treatment, how do we expect teachers to accomplish the task?

It shouldn’t be done.

Teaching the non-cognitive overshoots the educational mandate as well as teacher know-how. Since when were schools charged with redoing personalities? Parents and communities have not collectively given educators’ permission to take on this personal restructuring. Different parents value different qualities, their preferences reflecting cultural as well as individual leanings. There is hubris in claiming schools know what qualities universally make for a worthwhile life and reaching a consensus is unlikely. Is it more advantageous, in terms of ultimate life success or, better, life satisfaction and usefulness to be outgoing rather than shy; to be enthusiastic rather than indifferent towards school; to be docile rather than excitable; to be curious about computers rather than literature; even to be social rather than solitary?

Selecting particular attributes from this under-defined concept opens a Pandora’s Box of intrusive teaching and evaluation with no agreed-upon criteria for selection. We fear that as schools attempt to isolate critical non-cognitive traits, it is their own normative biases that will surface, without strong theoretical or empirical justification.

It is predictable that the identified qualities will be those most convenient for the school culture, that schools would prefer conforming, obedient, cooperative students who can be counted on to accept and follow daily routines, comply with assignments, be gentle towards others, and have an easy-going extroverted disposition that accepts school demands with minimal grumbling.

When the measurement of success includes “time-on-task,” number of assignments submitted on time, and self-reports, the opportunity to cultivate or even recognize critical thought as a desirable trait is limited. It would be surprising indeed if independence, originality, thoughtful dissent, assertiveness, or risk-taking were selected for cultivation and measurement, yet these very qualities are likely to characterize the successful 21st century entrepreneur or political leader.

Self-control and perseverance sound admirable, but when such attributes serve goals chosen by the school, they become thinly disguised terms for doing what you are told.