Me, 2012: Growth mindset is amazing! Now I know why my students aren’t trying, and how to fix it. I’m going to encourage growth mindset in my classroom.
Me, 2017: I liked growth mindset…. but the evidence doesn’t stack up. I’ve just read a big study which found no link between growth mindset and student success (cited in Responsive Teaching).
Me, 2022: I have no idea if growth mindset is real. I’ve read robust studies showing it has no effect – and robust studies showing it does.
This post untangles what I’ve learned and unlearned about growth mindset. It tries to make sense of the contradictory studies. And it tries to clarify whether we should be encouraging or ignoring growth mindset.
2012: Growth mindset is amazing!
We’ve all seen the posters.
- Fixed-mindset Phil believes you’re either smart, or you’re not. When he fails, he gives up: ‘I’m Number 71,000,000, so why try harder?’
- Growth-mindset Glenda believes she can get smarter. When she fails, she knows it’s temporary, she tries harder, or switches strategies.
Can we persuade Phil to be a bit more like Gina? A classic study helped launch growth mindset, by saying ‘yes’. Some students were taught memory strategies, other students were taught growth mindset: you can get smarter (Blackwell et al., 2007). (Inevitably), the growth mindset group got higher grades…
My working assumption in 2012: growth mindset is amazing! I should teach my students to use it. I do.
2017: I liked growth mindset, but the evidence doesn’t stack up
We can pick holes in the original study: one hundred students in just one school. But we don’t need to: recent, robust studies have carefully undermined these claims.
- Czech Republic: high school students with a growth mindset got slightly lower test scores (than students with a fixed mindset). They didn’t choose to retake tests more. If they did retake, they didn’t improve more (Bahník and Vranka, 2017).
- Scotland: students with a growth mindset didn’t have higher grades before entering university. Or in their first year of university. Or in any of the following three years (Li and Bates, 2020).
- USA: Researchers took six claims about growth mindset: 1) People with a growth mindset care more about learning, 2) They persist for longer, 3) They bounce back better from setbacks, 4) Their hair is more lustrous and their skin looks great… They ran a whole load of tests on undergraduates, for example, if a student said they had a growth mindset, did they also care more about learning. They found: i) No evidence for three claims, ii) Weak evidence for two claims and iii) Strong evidence against one claim: students with a fixed mindset respond better to negative feedback (Burgoyne et al., 2020).
Reading academic writing sometimes feels like wading through treacle. Recently, I had to read this sentence about a dozen times – “We failed to reject the null hypothesis that the means are equal” – just to work out that the means were about equal. A crystal clear sentence of criticism sets of alarm bells Li and Bates (2020) argued they had presented:
“Compelling evidence calling into question core predictions and assumptions of mindset theory.”
If growth mindset advocates claim it only works for certain students – they went on – at certain ages, and in certain interventions, they’re moving the goalposts – and no longer behaving scientifically.
Burgoyne et al. (2020) went further:
The foundations of mind-set theory are not firm… bold claims about mind-set appear to be overstated…. Time and money spent on mind-set-related programs diverts resources from other programs with potentially greater effects and stronger theoretical underpinnings.
My working assumption (2017-2022): growth mindset isn’t a valid construct – don’t bother.
I have no idea what’s going on with growth mindset
This may not come as a surprise. David Didau’s doubt about growth mindset grew gradually from early 2015, while Tom Chivers highlighted methodological concerns in early 2017. I was contentedly ignoring growth mindset.
It all went wrong for me when I read this article title: “A national experiment reveals where a growth mindset improves achievement (Yeager et al., 2019).” The researchers gave a bunch of students 25-minute online activities designed to persuade them they can get smarter – that growth mindset is real. And it worked.
Recently, lots of psychologists have been caught out sneakily moving goalposts in their research. Lots of other psychologists have been having conferences about how to make better goalposts. The designers of this study had clearly been going to those conferences, because their goalposts were made of titanium and anchored in granite. Like most studies, students were randomly allocated to either learn about growth mindset, or something else (a control group). But the researchers also:
- Preregistered, publishing their plans in advance (instead of collecting data, finding a significant result, then claiming this was what they were looking for all along).
- ‘Blinded’ teachers, so they didn’t know which students had learned about growth mindset (they might then treat those students differently). The intervention was online: teachers didn’t know which students had been directed to the pages about mindset, and which had been directed to the control material.
- Blinded analysts, sending the data to a third party to analyse. These analysts didn’t know they hypothesis or which group of students had learned about growth mindset. (Study designers have a vested interest in the study succeeding, which may influence the decisions they make in analysing the data.)
OK, we get it, titanium goalposts. So what happened? Well, after these two 25-minute sessions:
- Students were more likely to believe they could get smarter
- Lower-attaining students got higher grades
- Students were more likely to choose challenging maths courses the next year
Wait, so do we have to believe in growth mindset again? On Twitter, sceptics accepted the titanium goalposts, but pointed out that Carol Dweck, doyenne of growth mindset studies was one of the authors. But what no one could tell me was why this was a problem. She’s not a magician – do we think she whispered to the teachers which students were which? She didn’t know either! And in any case, should we really be surprised that the world expert in growth mindset… is researching growth mindset again?
Now I don’t have a working assumption. “The foundations of mindset theory are not firm (Burgoyne et al., 2020)” but “A national experiment reveals where a growth mindset improves achievement (Yeager et al., 2019).”
Synthesis: making sense of conflicting studies
How can we make sense of this? How can robust studies show having a growth mindset makes no difference – and then a robust study show that encouraging a growth mindset makes a difference?
I think we can reconcile it by looking at how the growth mindset intervention worked. To tell students about growth mindset, the intervention:
Featured stories from both older students and admired adults about a growth mindset, and interactive sections in which students reflected on their own learning in school and how a growth mindset could help a struggling ninth-grade student next year (Yeager et al., 2019).”
Is this really about growth mindset? I mean, yes, it is. But we know, for example, that role models have a powerful influence on students’ beliefs and actions. As Li and Bates (the researchers who found no link between growth mindset and achievement) put it, research on mindset has “confounded other motivational factors.”
Attempting to encourage a growth mindset may have helped – not because a growth mindset is ‘real’, but because the attempts influenced motivation in other ways.
This resolves the contradiction:
- If you ask students whether they have a growth mindset, then see if they try harder or get better grades, you won’t see an impact (what Bahnik and Vranka, Burgoyne and co, and Li and Bates all did).
- But if you encourage a growth mindset, you may say something, do something, give some example, which motivates students and leads them to get better grades (even if growth mindset isn’t real).
I think the thing may be self-efficacy. When students believe they can do things, they are more likely to do them. And role models and persuasion can help to change their belief. But maybe it’s just role models.
Working assumption: growth mindset isn’t real – as such. But when we try to persuade students they can do better, they should persist, and bounce back, and use feedback… we may help students see the value of their efforts.
Conclusions (epistemic status: tentative)
1. Growth mindset probably isn’t real.
2. Motivation is real, more motivation can help (duh), and encouraging messages about how we learn can help students try harder.
Also:
3. There are no miracle substances in education. Any time we hear that a recently-made-up-construct (growth mindset, grit, emotional intelligence, resilience) is more important than knowledge and skill in predicting outcomes, they’re wrong.
4. Sometimes we do the right things for the wrong reasons (for example, check out this, on how much learning styles helped improve one class).
5. But the closer we get to understanding how and why something is (supposed to be) working, the more likely we are to be able to identify the active ingredients, and use them effectively.
If you enjoyed this, you might like
This post, describing a range of interventions to encourage student effort – fitting them to the classroom.
This post, looking at self-efficacy, and how we encourage it.
This post, offering an overview of ways to convince students to learn
References
Bahník, Š. and Vranka, M. (2017). Growth mindset is not associated with scholastic aptitude in a large sample of university applicants. Personality and Individual Differences, 117, 139-143.
Li, Y., & Bates, T. C. (2020). Testing the association of growth mindset and grades across a challenging transition: Is growth mindset associated with grades?. Intelligence, 81, 101471.
It all comes around.
Also important to remember that in the Blackwell et al study, the impact on achievement was not statistically significant at the 5% level. The authors stated “This effect is likely only marginally significant because of our small sample size and should be replicated with a larger sample with more power to detect the interaction” (p. 257). This, of course, assumes that the additional cases included in a larger sample would be similar to those in the existing sample, but there is no reason to assume that this would be the case. Indeed, assuming that a larger sample would produce a statistically significant result shows (a) a lack of understanding of null hypothesis significance testing and (b) is a good example of “hypothesizing after results are known” or “HARK-ing”. If you decide that you need to collect more data and leave your threshold for significance unchanged, you have just increased the chances of making a Type I error.
Interesting article with a lot of food for thought.
“When students believe they can do things, they are more likely to do them” particularly stood out.
I suspect it has more to do with how we, as teachers, approach it in our classroom. If we come to it with enthusiasm, it will be more impactful than if we aren’t. However, I enjoyed this article.