Earlier today, I published a blog post in which I aggregated a BuzzFeed article by Tom Chivers about so-called mindset theory. Mindset theory, which is mostly the work of Stanford University psychologist Carol Dweck, postulates an important distinction between “growth” or “fixed” mindsets. If you have a growth mindset, you respond to failure with an attitude something like, Well, I’m going to have to get better at this so I can succeed next time (I’m cribbing from my original post). If you have a fixed mindset, on the other hand, you might say, I’m clearly not good enough to do this — I should give up and move on.
Mindset theory’s proponents claim that this simple concept can go a surprisingly long way toward explaining why some people succeed and others fail, and also that people — schoolkids, most excitingly — can be induced to adopt a growth mindset, which leads to better performance. As a result, the concept has spread far and wide in education circles.
Chivers’s article raised a number of troubling points about the quality of evidence undergirding these ideas. “[S]ome statisticians and psychologists are increasingly worried that mindset theory is not all it claims to be,” he wrote. “The findings of Dweck’s key study have never been replicated in a published paper, which is noteworthy in so high-profile a work. One scientist told BuzzFeed News that his attempt to reproduce the findings has so far failed. An investigation found several small but revealing errors in the study that may require a correction.” Chivers went on to lay out evidence suggesting people should be more skeptical of the claims made by Dweck and other mindset theory proponents.
In my original post, I more or less endorsed Chivers’s argument, encouraging people to read his article. After it went up, though, I received an email from David Yeager, a psychology professor at the University of Texas at Austin who has conducted mindset-theory research. Yeager argued, convincingly, that Chivers’s article failed to mention a bunch of research that would have complicated his story line. I looked back at my own post and realized that I hadn’t been skeptical enough in my aggregation of someone else’s work, especially given that this is a subject area where I don’t have a great deal of familiarity.
Which brings us to this post. My original blog post on Chivers’s article used to live at this URL, but I realized that it would have required significant enough updates and corrections that it made more sense to fully rewrite it and explain my errors. For transparency’s sake, the old post is here in its original form, and I’ll spend the rest of this post explaining what happened.
The short version is that since Chivers’s article focused primarily, in its rundown of the mindset-theory research, on that single study by Dweck that has failed to be replicated, I in turn leaned too heavily on that study. Specifically, I wrote that “Much of [the excitement over mindset theory] is based on a single article Dweck published in 1998 with Claudia Mueller.” This was an overblown claim on my part. In fact, as Yeager pointed out, enough research has been conducted on mindset theory that one meta-analysis examining the link between mindset and performance looked at 113 studies, finding that people with mindsets “characterized by the belief that human attributes are malleable rather than fixed” were significantly more likely to engage in a number of goal-related behaviors and strategies linked to enhanced performance. (It’s worth noting, as always, that correlation doesn’t imply causation.) Another meta-analysis, in the context of a broader examination of 92 educational interventions, looked at six studies of mindset interventions pertaining to Dweck’s ideas and found that they appeared to be fairly effective at improving educational performance.
Now, it is also the case, as Chivers pointed out, that Dweck has made some overheated-seeming claims about mindset theory: “Her website claims that a fixed mindset caused the Enron scandal, while a growth mindset can encourage cooperation between Israelis and Palestinians,” he wrote. “‘Almost every truly great athlete — Michael Jordan, Jackie Joyner-Kersee, Tiger Woods, Mia Hamm, Pete Sampras — has had a growth mindset,’ she believes.” In addition, Dweck’s foundational study, which offered up eye-popping results, has failed to replicate and contained some fairly important statistical errors (which she has since acknowledged). Chivers also noted that some other papers in the mindset-theory space appear to have statistical problems of their own, and quoted Columbia statistician Andrew Gelman, an extremely credible voice on these matters, as saying Dweck and her colleagues are “using statistical methods that are known to be biased” (Gelman has also expressed skepticism of some of Dweck’s work on his blog). And he isn’t the only statistically sophisticated researcher who has questioned the the quality of this research – Chivers quoted others, too.
So it could be argued that mindset theory, like so many other ideas in pop psychology, has sometimes been oversimplified and overhyped, and/or that there appear to be some holes in it here or there. But it’s not fair to level such critiques, or to amplify them as I did, without also addressing the other side of the ledger: There is a decent quantity of published evidence which at least partially supports Dweck’s ideas. I do think Chivers should have mentioned those meta-analyses, if only to explain why he didn’t buy their conclusions (there’s one fleeting mention of meta-analyses that comes in a quote, but nothing from Chivers himself), but that’s not an excuse: My blog post had my name on it, not his, and it was sloppy of me to aggregate his story without first seeking to better inform myself about the mindset-theory research landscape.
Anyway, I’ll do better in the future. In the meantime, here’s a blog post just published by Dweck defending her work. If you read Chivers’s article, read Dweck’s post too.