Russ Whitehurst, one of the most cool-headed education researchers, throws cold water on the grit craze seizing some quarters of the ed reform world. As I warned in my recent review of the Angela Duckworth and Paul Tough books, “This new attention to character skills has many of the markings of previous failed fads…. In short, school and educator practice with respect to character skills is running far ahead of knowledge.” I wanted to cool grit fever, but in his recent piece Russ throws icy cold water on it. He raises excellent points but I wonder if Russ is too cold on the importance of character skills.
Russ makes three arguments: 1) A recent study that compared grit scores among fraternal and identical twins suggests that grit may be heritable to a large degree, which would make it unrealistic to expect schools or others to be able to alter it; 2) The twin study as well as a meta-analysis of grit research found that grit only explains about 2-3% of the variance in achievement scores, which Russ thinks makes it a poor predictor of other outcomes; and 3) The meta-analysis suggests that grit may be highly correlated with conscientiousness, one of the Big 5 personality traits that psychologists have been studying for a long time.
I think Russ is most persuasive on the last point. Grit may be more of an effective marketing brand than a new contribution to the field. But whether it is really distinct from conscientiousness or not, this does not establish whether grit and other character skills are important for education reform.
Russ is much less persuasive with his second argument. The fact that grit or other character skills may not be strongly predictive of achievement test results is not surprising if these non-cog measures capture something that is important independently of cognitive ability. That is, the true test of the predictive power of “noncog” measures is not whether they are correlated with cognitive measures (like achievement scores), but whether they are correlated with later life outcomes. As it turns out out, they are. As this recent piece in Economics of Education Review by Collin Hitt, Julie Trivitt, and Albert Cheng shows, their character skill measure collected in middle or early high school is predictive of later educational attainment, employment, and earnings in 5 different longitudinal panel data sets, even after cognitive ability and other factors are controlled.
Russ’ first argument is the most important for educational reform. If grit or other character skills are not malleable, then why bother devoting a lot of energy to trying to address them in schools? Russ is correct to point out that about 37% of the perseverance component of grit is heritable, but that does not establish that educational policy and practice are unable to alter the non-heritable factors that form grit and other character skills.
There is a growing body of evidence that suggests character skills are malleable and that education plays an important role in shaping and altering character. Albert Cheng and Gema Zamarro have a new study that shows students randomly assigned to teachers who possess stronger character skills experience an increase in their own character skills. And Albert Cheng has another study using a student fixed-effect research design that has the same finding. This would not be possible if character skills were not malleable.
In addition we have an entire literature on pre-school and school choice suggesting that educational interventions can produce long-term success without improving short term achievement test scores (and vice versa). It’s not well-understood exactly how these benefits are being produced, but a reasonable explanation is that early childhood and effective school choice programs may improve character skills without also increasing test scores. And in the long-run the improvement in character skills may be more important for success.
Russ Whitehurst is right to warn us about the irrational exuberance some have about grit, but he shouldn’t throw out the baby with ice cold bath water. There seems to be something very important about character skills in education even if we do not fully understand how to define, measure, or alter them.
—Jay P. Greene