August 4, 2016

Much Ado About Grit? Interview with a Leading Psych Researcher

By Bellwether

Share this article

What is grit? Can it be measured accurately, and is it different from other personality traits? If so, how well does an individual’s level of “grit” predict how successful that person will be in the future? And is grit an innate characteristic, or can it be improved with practice?
The answers to these questions suddenly matter a great deal for schools. As states begin to implement the Every Student Succeeds Act, there’s widespread interest in incorporating “non-academic” factors such as grit into the way states define what it means to be a successful school.

Marcus Crede

Marcus Crede photo via Iowa State University


To learn more about grit and the research behind it, I reached out to Marcus Crede, an Assistant Professor in the Department of Psychology at Iowa State University and the author of a provocative new study called “Much Ado about Grit: A Meta-Analytic Synthesis of the Grit Literature.” After reviewing the full academic literature on grit, Crede challenges much of the popular narrative. For example, his study finds that grit is barely distinct from other personality traits and that standardized test scores, attendance, and study habits are much better predictors of long-term success than grit.
Below is a lightly edited transcript of our conversation.
Aldeman: Can you start by saying what made you want to pursue this work on grit?
Crede: Well, I have over the last 10-12 years now done a fair amount of work looking at predictors of educational success. I’m trained as an industrial psychologist, but education is similar to the workplace, and there’s a big literature on all sorts of constructs. Because many of these fields have enormous literatures, I’ve used meta-analysis techniques to pull everything together in a specific domain, figure out what works and what doesn’t work, and try to write about it so other researchers can get some greater clarity. My co-authors and I did one where we read around 2,000 articles in our attempt to pull a large body of work together in one place.
Our paper on grit was another example of that. About 10 years had passed since grit was first proposed and I figured there might be enough empirical literature to conduct a meta-analysis.
Can you summarize what you found?
We found data from 88 different samples, covering over 66,000 people in total. We had four big findings.
The first one is that grit does not really correlate with or predict success particularly well, certainly not as strongly as some grit proponents have described. In comparison to other predictors of academic success, grit performs relatively poorly.
The second big finding was that grit is very strongly related to conscientiousness, a personality trait psychologists have been studying for a very long time. This suggests grit is merely a re-packaging of conscientiousness, a new label slapped on an existing variable. That’s obviously very troublesome, because if it’s nothing new then we’re not likely to get much out of it.
The third finding was a big surprise to me. A lot of what’s been written by grit, particularly by Dr. Angela Duckworth, seems to have been based on some pretty severe statistical misunderstandings. There are two major errors that she’s made. One, she’s over-stated the size of some of these effects, in one case more than thirty-fold. And two, she’s also incorrectly argued that grit is composed of these two pieces that we can simply put together, the consistency of interest and the perseverance of effort components. That was troublesome to us, that nobody has picked this up before. (Editor’s note: You can read Duckworth’s response on these findings here.)
The fourth one, and it’s one that we don’t actually write about that much in the paper, is that the way grit has been described to the public in interviews and popular writings doesn’t match up well with the actual empirical findings. This is true not only in general across the entire field, but even Dr. Duckworth’s work really doesn’t support what she says about grit in public interviews.
So those are probably the four biggest takeaways from our work.
Before moving on, can you say more about this last point, that the way grit has been portrayed doesn’t really match up with the academic literature?
Sure, let me take two examples of what Dr. Duckworth has said in the public eye. There was a recent New York Times interview a few weeks ago, in which she was being interviewed about her new book. In response to the first question, “Why is grit so important?” her response was that grit, “beats the pants off I.Q., SAT scores, physical fitness and a bazillion other measures to help us know in advance which individuals will be successful in some situations.”
In her TED talk, she made a similar statement. But if you go to her research findings, they really don’t support those claims at all. One of my favorite papers of hers in terms of research approach and rigor was a 2014 paper in which she looked at four different, very large samples to see how well grit predicts success. And there, grit just looks terrible in comparison to these other variables that she looks at. I’ll give you two quick examples.
In the first study in that paper, she looks at retention in the Army special operations forces selection course. She has data on 677 soldiers who are going through this very rigorous process, and she’s looking to see whether grit predicts who gets through the course. It turns out that physical fitness predicts five times as much variance as grit, and IQ predicts three times as much variance as grit. So to me, when I look at that, that means that IQ and physical fitness beat the pants off of grit, not the other way around.
Another study in that same paper looks at retention for almost 5,000 Chicago public schools students. And again, standardized admissions test scores do a much better job; they explain about twice as much variance in retention as grit does. And when she looks at marital success, grit doesn’t explain any variance of success, “success” being whether someone stayed married or not.
So when looking at her findings, it strikes me that the rhetoric often doesn’t match what her own work indicates, and certainly not what the work of other researchers suggests.
Can you walk us through how grit is measured? Are there different measures of grit and how do they work?
The primary way in which this is done — there are some other approaches — but most of the literature relies on self-reports. Students, or soldiers, or spelling bee contestants are asked to rate themselves on a grit scale. There are two versions, a longer one and a shorter one, and pretty much everyone uses one of those two scales. These are self-report measures, which is one of the reasons why this doesn’t work so well. You’re asking people not only to report on something that’s socially desirable — we all want to think of ourselves as determined and persistent. It also relies on self-awareness that, certainly for some students and even for some adults, may not be there. Will we really find people who say, “Listen, I do give up really easily” or “I change my interests all the time.” It’s such an undesirable response that even if people were aware of it, they may not admit to it.
The reliance on self-report measures may be problematic. I could see more value potentially in ratings by other people, teachers, co-workers, supervisors, parents perhaps, who have more accurate assessment of a particular individual’s level of grit. If this topic is going to move forward, that may be a more fruitful area of inquiry.
There’s a new education law that allows states flexibility in defining what a “quality school” looks like, not just academic measures but potentially other non-academic or school quality measures. There’s interest in exploring grit or other things to assess how schools are doing. How would you respond to that interest?
It’s an interesting question to figure out what it means to be a “successful student,” and I completely agree that success is more than scoring well on tests. I’m a big fan of expanding what we mean by “success” and what we mean by “achievement.”
However, I haven’t seen good work done on figuring out what that would mean. I may just not be aware of it, since I read less in the K-12 literature than I do on higher education. But I’m certainly open to that.
I think it would be a very useful thing to do to redefine what we mean by success beyond simply grades. Once we do that, there may a whole host of other predictors and variables that come into play. My own work has identified a bunch of other factors that are highly relevant, and certainly more predictive of success, than grit. Many of them get close to what we see for IQ, ACT, and SAT scores, so there’s a lot of potential there.
Can you elaborate on what those factors might be and how they might be measured?
Sure, my co-authors and I have been working on this over the last decade. Some of the most promising ones include study habits and study skills, the way that students practice for exams, how they revise material, whether they try to cram it in at the last minute, do they practice on a regular basis. The effect size for these behaviors is about twice the size of what it is for grit. We see very good prediction, in higher education, for variables that reflect a student’s ability to adjust to a new environment. College students arrive on campus, and are often away from their family for the first time, and they have to manage their own finances, do their own laundry, and figure out classes that are structured quite differently to what they are used to. Some students really struggle with that. How they handle that transition seems to be a good predictor of whether they stay in college.
Again in higher education, simply going to class is twice as good a predictor of grades as grit is. Making people come to class reduces failure rates and poor grades very substantially.
And then there’s work done on decision-making capacity and procrastination and a whole host of other factors. Meta-cognition, for example, has shown some positive effects. That measures whether someone is aware of how they’re doing and what’s working and whether they know when to ask for help. There’s a well-developed literature that suggest there are lots of variables to target.
Many of these variables are also easier to intervene with. My gut tells me it’s much easier to teach a 10th grader how to study more effectively, which is really a skill or a habit, than it is to change their personality, which is what grit is perceived as. I’ve talked about this before, but I would be somewhat unhappy as a parent if my daughter came home from school and said her teachers were trying to change her personality. But if it’s a skill that’s being taught, or helping students decide when to ask for help, we might look on those sorts of interventions more favorably.
We talked about the most common measures of grit, and you highlighted the challenges of relying on self-reports. It strikes me that something like attendance is easier to count, so we wouldn’t have to rely on self-reports. What about something like study skills? How are those measured?
There are many approaches. We did our big review, which took us about 3 years, and it’s been almost 10 years since we published it. There we identified a wide array of approaches. The funny thing we noted at the time is that there is one inventory which was developed in the 1950s that’s also a self-report inventory, but it worked beautifully as a predictor of how students would perform. A lot of the more modern approaches fall short. They’ve been developed without a careful attention to detail that was prevalent in the earlier work.
That earlier, really good inventory, called the Survey of Study Habits and Attitudes, has unfortunately fallen somewhat out of favor. Hopefully researchers and practitioners will look at our work and renew their attention to it. There’s one more modern one, called the LASSI (Learning and Study Strategies Inventory) that also works very well. Those two — to me at least— are the two gold standards for measuring study skills and study habits.
Are there contexts where these concepts may matter more than others, whether that’s grit or other types of behaviors? In particular, how do they play out in an academic setting? 
One thing we speculate on in our paper is that maybe the reason grit doesn’t predict so well is that we’re looking at a performance domain where grit may at times be counter-productive. If you have a well-defined task, you know exactly what you’re supposed to be doing and how you’re supposed to be doing it, but it’s going to take you a lot of persistence to finish it. That might be a situation where grit is valuable. Similarly, if you have a very steep learning curve, where you have to persist a long time to master some basic skills before you can get to a level of competence, like learning a musical instrument, grit may also be helpful.
But there are other domains where grit may be counter-productive. Creative tasks, for example, don’t lend themselves well to just banging your head against a wall. A good example of that was Albert Einstein. He was desperately trying to come up with mathematical equations to describe his theory of relativity, and for years he couldn’t do it. He had made an earlier mathematical error that he hadn’t picked up on, and he may not have had the mathematical skills to really get it right. He had to get help, so he went to some of his fellow scientists, Marcel Grossman and David Hilbert, who were better mathematicians. And once he did that, it wasn’t too long before he managed to figure it out.
I think we need a lot more work trying to figure out where these characteristics are important and where they’re less important. That applies to other characteristics as well. There’s a personality trait called “openness to experience” that might be important for creative tasks, but it might be unimportant or even counter-productive in many academic settings. Thinking about where it matters and where it doesn’t matter is going to be important for grit researchers and anyone who works in the performance domain.
It’s one thing to measure something and know whether it’s predictive of desired outcomes, but do we know enough about how to improve some of these behaviors? How would a school or teacher work with students to enhance these attitudes or behaviors?
I don’t know how to do it, and I haven’t see a whole lot of interventions on how to enhance grit. I know there’s been some work on a related concept on resiliency. There have been some resiliency studies, and Peter Harms at the University of Alabama has recently published a meta-analysis of the interventions related to that. My sense is that these interventions may work in the short term, but over the longer term the effects begin to fade away. I’m not aware of how these things could be shifted in a systematic way.
We know personality develops over the lifespan. There are natural events that happen, and as we get older and life experiences accumulate, personality does shift. But doing that in a planned manner through an intervention, I haven’t seen anything.
Finally, for a somewhat off-the-wall question: Where do you stand on the question of whether grits are a breakfast or dinner food? Sweet or savory?
I would definitely say a breakfast food. But I was raised in South Africa, so my answer may not be representative of what anyone in the U.S. thinks.

More from this topic

Processing...
Thank you! Your subscription has been confirmed. You'll hear from us soon.
ErrorHere