Dweck as in Wreck

Carol Dweck has won a juicy prize for work on ‘growth mindset’ – the notion that belief that your intelligence is a fixed trait cripples you, while a ‘growth mindset’ allows progress.

Her studies do not replicate, of course, because this idea is false. It’s obvious bullshit. As it turns out, ants can’t move rubber tree plants.

What do ‘grit’, ‘growth mindset’, ‘power posing’, ‘priming’, and ‘stereotype threat’ have in common? As far as I can tell, the more obvious the bullshit, the better the Fools at the Top like it.

Advertisements
This entry was posted in Uncategorized. Bookmark the permalink.

58 Responses to Dweck as in Wreck

  1. Distrust but verify says:

    On the one hand, the growth mindset and stereotype threat people made their names with noisy underpowered studies that don’t seem very replicable.

    But I was struck to see that a group of them have come up with an educational intervention that met its pre-registered hypothesis of boosting GPA for low-performing students in a sample of thousands:

    http://psycnet.apa.org/record/2016-15978-005

    There’s plenty to gripe about (the intervention doesn’t show overall improvement in grades, just in lower-performing students, but I saw them pre-register the hypothesis after finding the result in a previous sizable study. I would be very interested to hear your take on what’s happening there.

    • Toddy Cat says:

      Let’s see if it replicates. Color me suspicious…

    • AppSocRes says:

      “…the intervention doesn’t show overall improvement in grades, just in lower-performing students…”

      Regression towards the mean?

    • Distrust but verify says:

      This is the pre-registered larger replication of a previous study. And regression to the mean doesn’t cause a difference between control and treatment group.

      • spottedtoad says:

        Her group did a bunch of shady stuff in their analysis of other studies. (http://slatestarcodex.com/2015/04/22/growth-mindset-3-a-pox-on-growth-your-houses/ – Scott is overly gracious to the author when he responds.) The basic problem is that if you let me slice and dice my sample however I want, I can construct a subgroup where the treatment group goes up by a significant amount relative to the control group. I’ve seen education studies where the raw values for the intervention group go down relative to the control group by a large and statistically significant amount, and then they apply some absurd growth curve or structural equation functional form to it to make it look like there’s a positive effect. (Here’s an example https://spottedtoad.wordpress.com/2016/01/21/but-what-is-the-intervention/) I think that’s absurd, and kind of defeats the whole point of what RA Fisher was thinking of for randomized experiments in the first place, but it’s actually the ticket to the big time in education schools these days.

      • Distrust but verify says:

        That’s what I thought the first time they found it. But then they replicated the slice-and-dice in a large pre-registered sample with strong p-values. That’s what’s so striking and what I would like to see explained by analysts like yourself.

  2. karl william liebhardt says:

    This woman’s site is AMaaaazing. https://mindsetonline.com/index.html
    What sort of academic with any dignity would put up a slick site seemingly meant to entice a purchase rather than illuminate?

  3. Randy McGregor says:

    I read on the internet (Don’t remember where. It seemed legit…) that men out-achieve women in the workplace because they are over-confident. If true, doesn’t this lend some support to the idea that attitudes are influential?

  4. Burgerhampton says:

    Psychology has been taken over by women, that’s the problem.

    • Ursiform says:

      Can you support that assertion? It’s not like male psychologists are noted rationalists.

    • Meagain says:

      Yep. Since men and women have different interests, we should expect difference in selecting research topics. For example, the fact that men are more interested in “things” and mechanisms and women are more interested in “people”, interpersonal relationships and feelings (“People-things” dimension) will bias their search for causes of behavior in different directions. Female psychologists, more than male, will be prone to explaining behavior in terms of social influence, mentalistic causes and seeing talk as a cure for behavioral problems. Care and empathy being more pronounced in women makes female psychologists more obsessed with the idea that unpleasant experiences in childhood are such a big deal for later personality and increase of women in psychology explains why psychology today is so overly protective to kids and so concerned that if you hurt their feelings they might grow up one day being Ted Bundy or something.

      • Garr says:

        “care and empathy being more pronounced in women” — I don’t know why people believe this. Men are much more affectionate with and attentive to their kids than women are, for example — at least in NYC, where this is noticeably the case with parents from every social class and every ethnic/racial group. As for who’s more caring and empathetic in marriages and romantic “relationships” — it’s so obviously the case that men are more caring and empathetic that the point seems hardly worth arguing about.

  5. JerryC says:

    The belief that ants can too move rubber tree plants, as soon we build a socially just utopia, is what lets wannabe social engineers “dream big”.

  6. Call it a “fools-to-the-top” mindset.

  7. pyrrhus says:

    Fostering the belief in comforting nonsense is a feature, not a bug, for our Elites….

  8. Rhetocrates says:

    “What do ‘grit’, ‘growth mindset’, ‘power posing’, ‘priming’, and ‘stereotype threat’ have in common? As far as I can tell, the more obvious the bullshit, the better the Fools at the Top like it.”

    No, it’s obvious but morally-flattering bullshit. They’re at the top of the heap, so if these theories are true they obviously have the gumption, which is a medal they can pin on their chests. If it’s just innate, then all that hard work they feel like they did doesn’t matter, right?

    • JayMan says:

      People don’t like that prospect, do they?

      • Cavalier says:

        People used to boast about the benefits of good breeding. It doesn’t much lend itself to democratic pretensions, though; triply so if every stray comment is public and on record and travels at the speed of light.

      • Zenit says:

        For thousands of years, people rather liked the prospect of nobles born to rule, and peasants born to serve (at least the nobles liked it).

    • ...and this applies to group differences says:

      As Greg states, the growth mindset and grit are probably rubbish – https://twitter.com/degenrolf/status/776473421299511297

      Steve Hsu, who is normally sensible, loves the growth mindset idea – http://infoproc.blogspot.co.uk/2017/09/feynman-schwinger-and-psychometrics.html

      My guess to square these:

      Achievement beyond IQ obviously exists within Asian migrant populations. Check out university graduation rates to cognitive ability tests in Pakistanis in the UK, or Chinese-American representation in the Ivy League relative to their % of the US population and an assumption of a 7-10 point gap. It’s well beyond expectation, and beyond anything explained by observed IQ within universities at large and the Ivy League. It’s so not IQ alone it’s not even funny.

      Now growth mindset is mostly a flattering explanation, in that it proposes a strong cultural trait of will to power, determination, ambition and humility. In theory these are strong individuals with self belief.

      The simple realities that these groups are positioning themselves in a particular way, at a group level, and don’t really have any particular internal positive individual trait is far less flattering.

    • Matthew Mckenzie says:

      Conscientiousness, which is what ‘grit’ and ‘growth mindset’ to me seems to be all about, is the second, distantly, largest indicator of all sorts of positive life outcomes after IQ. From what I’ve heard this finding is also pretty robust. So I see where the interventionists are coming from. A high conscientious IQ 100 person is going to do better than one who is low conscientiousness. Which seems to me to at least partly explain East Asian schooling. Unfortunately for the interventionist crowd, it’s looking like it is just as heritable as IQ and it is pretty well correlated with IQ anyway. So the whole thing is looking like a wash.

  9. j says:

    Carol S. Dweck is the latest re-incarnation of Horatio Alger, the classic American motivation myth-maker, following the footsteps of gurus like Benjamin Franklin (“Autobiography”) and Napoleon Hill (“Think and Get Rich”). In my miserable youth, these popular American self-help books helped me. Dweck is a respectable Halabi name and refers to long-necked bottle.

    • MawBTS says:

      In my miserable youth, these popular American self-help books helped me.

      I read neuro-linguistic programming books as a child (“Six simple steps to REFRAME YOUR ENTIRE REALITY!”). They made me feel good. I don’t regret them.

      People derive inspiration from the stupidest things. A song. A picture. A fortune cookie. A crooked televangelist. That’s to be expected, we’re emotional and irrational. Somewhere there’s a young woman who’s just read Testosterone Rex and now has the confidence to break through the glass ceiling or something.

      But it still should be mentioned that these books are just intellectual masturbation. They’re written to make you feel good, to press your brain’s reward buttons by flattering you and your preconceptions. If they’re misrepresenting themselves as science books, then that sucks.

    • j says:

      What is remarkable is that these old “Think hard and you can levitate” thing is selling itself as science and its gurus have Ph.D.s. To quote Greg, “but it is a lie”. Could be that Standford University is granting fake academic diplomas?

      • karl william liebhardt says:

        I couldn’t believe the schlock I was reading on her site. Like….a palm reader having an ad on some dismal cable network….and then I see she’s at stanford, yale, etc. holy shit! what happened to my country?

        • j says:

          The country is as always was. People always loved self help motivational books. It is the universities that changed. In India, you can have a Ph.D. in Astrology from a recognized university. Stanford is trying to reach that level.

    • gcochran9 says:

      Perhaps a re-incarnation, but certainly not a descendant.

  10. Jon says:

    … maybe the arbiters were emboldened by Cordelia Fine’s precedent in prizes for nonsense.

  11. Ivan says:

    Andrew Gelman wrote extensively about lack of replication in the “power pose” study as well as in other similar “research”, e.g. pizza-gate, “priming”, etc.:

    “Another failed replication of power pose” ( http://andrewgelman.com/2016/10/22/30294/ )

    On the other hand, I am quite mystified how an otherwise smart person could come up with the following statement:


    To me, Wade’s comments follow the dictionary definition of racism. The point is not whether he characterizes a racial group as superior “in all traits.” Nobody thinks that. Even David Duke, I’m sure, recognizes black superiority in sprinting and basketball. The point is that when Wade makes his comparisons, inherent superiority is implied when he writes, for example, that the Malays can’t emulate Chinese business success because of genetics, or when he makes other such comparisons throughout the book.

    http://andrewgelman.com/2014/05/09/nicholas-wade-paradox-racism/

    It’s as if a relatively smart person has two (maybe more) brains that function completely independently, a rational and religious parts that just are not connected. Truly puzzling.

    • Jim says:

      Obviously from a naturalistic perspective the probability of all the thousands of human ethnic groups averaging the same in any trait is at least as improbable as all the mountain ranges of the Earth having the same average altitude or all the lakes of the Earth having the same average depth.

      People understand this about traits like height. But assertions about intelligence have to most people an evaluative component. They do not think of these assertions in terms of the categories of “true” or “false” but instead in terms of the categories of “good” or “bad”. So empirical evidence on differences in average intelligence between ethnic groups are irrelevant to them. Such assertions are judged in a moral/normative framework not in an empirical/scientific framework.

      Regarding Gelman’s remarks. He seems to think that attributing greater Chinese commercial success as compared to Malays to genetic differences is morally wrong.
      I wonder does he not think that greater black success in sprinting and basketball is due to genetics?

      Of course if Gelman says that attributing Chinese success vs. Malays to genetics is “racist” I can’t really argue with him. Whether a statement is “racist” or not is a moral judgement devoid of any objective meaning. But whether a statement is true or false has nothing to do with whether the statement is “good” or “bad”.

      Gelman seems to be saying that it’s OK morally to believe that same groups are better at some specific activity like basketball but morally wrong to be believe that some groups are “better” in some absolute sense. But Wade was just saying that the Chinese are better at business than Malays not that the Chinese are superior to Malays in some absolute sense.

      Aristotle made the point that it makes no sense to say “X is better than Y”. Such an assertion simply leads to the reply “better at what?”. Similarly to the assertion “X is good” the reply would be “good for what?”.

      It’s difficult to discuss this stuff with people like Gelman because they are conceptualizing the issues in a moral\normative framework whereas Wade is conceptualizing the issues in an empirical/scientific framework.

      • gcochran9 says:

        For all you know he’s consciously lying. Give the man the benefit of the doubt.

        • Jim says:

          Yes he could be. I can’t read his mind.

        • Ivan says:

          I’d be happy with explanation, but he seems to an honest to God believer in the cause. I visit his site occasionally to check my Bayesian intuitions, and saw some other semi-religious claims (e.g. related to the recent elections). His Wade review is the most striking example of incongruity, though.

          I do not believe he consciously lying because he is not afraid to criticize dubious statistics based research produced by luminaries in various areas of sociology, neuro-science, medicine, etc, presumably to some personal detriment, too. E.g.:

          http://andrewgelman.com/2017/04/27/whole-fleet-wansinks-architecture-design-trillion-dollar-industry-founded-pseudoscience/

          Were he afraid to participate in the “sensitive” genetics discussions, he would have simply refrained from doing so. But, no, he published a whole negative review titled with the usual Pavlovian labels.

          To me, that sort of “multicameral” reasoning is truly mysterious. There seems to be a breaker switch in the brain that makes people like him (or some of my close relatives who I know are not lying) to just mutter tired cliches almost automatically, like a talking parakeet.

          • Jim says:

            I think it’s natural for many people to believe that the world has an inherent moral structure. This view of the world as run by moral agents viz. “gods” is deeply embedded in how people think about reality. Of course the world isn’t run by gods. It’s run by differential equations. The likelihood of reality exemplifying simple patterns like equality that happen to conform to our moral desires is zero.

          • Julian says:

            This is something Jon Haidt has written a lot about. Where there are sacred or taboo matters normally rational people can become quite irrational.

            • gcochran9 says:

              If I said that Eskimos were three feet tall, nobody would get upset about it, because it’s obviously untrue.

              • Jim says:

                Yes when people begin to suspect in the back of their minds that some statement which by their values is “bad” might actually be true they may become quite upset.

                Of course for a statement that has no moral significance or is obviously untrue there would be no concern.

              • fortaleza84 says:

                It’s not just that it’s obviously untrue, it’s because it’s obvious that nobody is going to believe it.

                There are plenty of statement which can be made which are obviously untrue, but still upsetting because some people are likely to believe them.

      • Ivan says:

        “It’s difficult to discuss this stuff with people like Gelman because they are conceptualizing the issues in a moral\normative framework whereas Wade is conceptualizing the issues in an empirical/scientific framework.”

        That’s an interesting perspective. The power pose “research” implications exists presumably in the same moral universe as his denial of intelligence differences among various groups of people (the “power pose” presumably helps women advancement). If so, why criticize the former as pseudo-science but support a different sort of pseudo-science ?

        http://www.slate.com/articles/health_and_science/science/2016/01/amy_cuddy_s_power_pose_research_is_the_latest_example_of_scientific_overreach.html
        Probably, people like him use a morality multiplier as it were when thinking about matters of that kind, so the “power pose” falsity overweighs the moral part and can be criticized, therefore, while admitting group differences in intelligence is so morally abhorrent that the veracity in this case is overridden by morality.

      • Julian says:

        Founding father of Singapore Lee Kuan Yew was quite open about average differences in work ethic and academic ability between groups, including Malay and Chinese. He even referenced the Bell Curve findings at one point. Ironically, although Nixon is seen as liar, he is one Western politician who appears to have been aware of psychometrics. He was well aware of Jensen’s work for example. Contrast with Bill Clinton who apparently avoided reading the Bell Curve. Yew on Nixon:

        “Tom Plate: Who was the greatest US President that you’ve seen?

        “LKY: But for the misfortune of Watergate, I would say Richard Nixon. He had a realistic view of the world. He was a great analyst, realistic, but also a tactician to get things done.”

        Giants of Asia: Conversations with Lee Kuan Yew: Citizen Singapore p88-89.

  12. For some reason these repeated attempts remind me of the Irishman in “Blood on the Coal.”
    Now an Irishman named Murphy said ‘I’ll stop that iron horse!’
    As he stood athwart its passage – and it crushed him dead of course.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s