Praise effort, not ability


From chapter 9 of “59 Seconds” by Richard Wiseman (emphasis mine):

In the late 1990s, Claudia Mueller and Carol Dweck from Columbia University conducted a large-scale programme of research into the psychology of praise. Their experiments involved more than 400 children, aged between ten and twelve, who were drawn from a variety of ethnic and socio-economic backgrounds. In a typical study, the children were presented with an intelligence test in which they were asked to look at rows of shapes and, based on logic alone, work out which shape should come next in each series. After they had worked through the problems, the experimenters took away their workbooks, calculated the scores but provided each child with false feedback. They explained that each child had done really well and solved 80 per cent of the problems correctly.

In addition to this feedback, one group of the children were told they must be really bright to have solved so many puzzles, while another group were greeted with stony silence. According to the self-help gurus who promote the positive power of praise, just spending a few seconds complimenting a child’s ability can have a dramatic effect. The results revealed that they are right, but perhaps not in quite the way they had anticipated.

In the next stage of the experiment, the researchers told the children that they could choose one of two tasks. They were told one of the tasks was quite difficult and so they might not succeed, but they would be challenged and learn even if they failed. In contrast, the other task was much easier, so they were likely to do well but learn little. Around 65 per cent of the children who had been told they were intelligent opted for the easy task, compared to just 45 per cent of those who had not been praised. The children who had been told they were intelligent were far more likely to avoid challenging situations, and instead stick to the easy stuff. This is not exactly good news for the ‘praise be’ approach to parenting. However, worse was still to come.

In the next phase of the experiment, the researchers gave the children some more puzzles. This time, the puzzles were much harder than the first set and so, as a result, most of the children did not perform especially well. Afterwards, all the children were asked how much they had enjoyed the puzzles and whether they would continue working on them at home. Dramatic differences between the groups emerged. The children who had received just a single sentence praising their intelligence found the difficult puzzles far less enjoyable than their classmates, and were far less likely to work on them in their own time.

Even more bad news for the advocates of praise emerged in the third and final part of the study. After struggling with the difficult puzzles, the experimenters asked the children to try one final test. This last set of puzzles was just as easy as the one the children had encountered at the start of the study. Even though the two groups of children had obtained roughly the same scores at the beginning of the experiment, their performance on the final test was very different. The pattern of results was exactly the opposite of that predicted by many self-help gurus. The children who had been told they were intelligent obtained far lower scores than the others.

Why should praise have counter-intuitive and counter-productive effects? According to Mueller and Dweck, there are several factors at work. Telling a child they are intelligent might make them feel good, but can also induce a fear of failure, causing the child to avoid challenging situations because they might look bad if they are not successful. In addition, telling a child they are intelligent suggests they do not need to work hard to perform well. Because of this, children may be less motivated to make the required effort and be more likely to fail. Unfortunately, if they subsequently obtain a low mark, it is also more likely that their motivation will collapse, and a sense of helplessness will set in. After all, low marks suggest they are not as bright as they were told and that there is nothing they can do about it. The psychological impact of poor results should not be underestimated. At one point in the Mueller and Dweck study, all the children were asked to tell their classmates how well they had performed on the test involving the difficult puzzles. Almost 40 per cent of the children who had been praised lied about their grade, compared to around 10 per cent of those who had not been praised.

Does this mean that all praise is bad praise? So far, I have only described the results from two of the three groups of children involved in the Mueller and Dweck experiment. After getting their initial ‘well done, you obtained 80 per cent’ feedback, a third group also received a single sentence of praise. However, this time the experimenters praised effort not ability, noting that they must have tried really hard to have achieved such a high mark. These children behaved very differently from those in the two other groups. When it came to choosing between a challenging or easy task, only about 10 per cent of them opted for the easy option. Compared to the children who had been told that they were intelligent, or received no praise at all, those in the ‘you must have tried very hard’ group found the hard problems more enjoyable and were more likely to try to solve them in their own time. Finally, when given another set of easy problems at the end of the experiment, those in this third group solved significantly more than they did the first time round.

The results clearly show that being praised for effort is very different to being praised for ability. According to Mueller and Dweck, the children praised for effort were encouraged to try regardless of the consequences, therefore sidestepping any fear of failure. As a result, the possibility of learning outweighs the fear of obtaining a low mark, and they prefer taking the challenging task to the easy option. Also, by definition, these children are more motivated to try hard in future in tests and are more likely to succeed. And, even if they do fail in the future, they can easily attribute their low marks to not trying hard enough, which avoids the sense of helplessness that can set in when poor results are seen as an indication of an innate inability to think.


7 responses »

  1. Excellent article. I’ll keep it mind when praising my grandkids, who are at an age to understand that their efforts can produce certain results.

  2. A good treatment of this and other talent/effort praising strategies appears in Bounce, by Matthew Syed. He develops a related theme of “the talent myth” – showing that almost all success at the top level correlates not with any genetic or environmental advantage per se*, but with 10,000 hours of deliberate/purposeful practice. Very high levels of practice (say 6,000 hours) might appear to be plenty, so we assume that the top musicians – for instance – have that extra “something”: but analysis of the data shows that the real top performers have those extra hours behind them. So Liberace had it right: “See what you get if you practise?”

    * Of course, family and environment may facilitate the practice.

  3. There is an enormous flaw in this study, and its results need a bit of sceptical thought. The topic and conclusion concerns situations where you have regular ongoing praise from people whose worth you have experience of and where you have some chance of assessing the validity and consistency of that praise by interactions partly under your control over a lengthy period of time. The experiment uses just one dose of praise (or not) from strangers in an artificial situation controlled entirely by those same strangers.

    People, kids and grown-ups, are continually monitoring sources of feedback for long-term quality as well as immediate meaning. Only feedback sources of consistent quality really change behaviour. All of that is simply adaptive social behaviour which allows us to distinguish chance and whim from consistency and meaning in our world.

    So, the experiment gives children feedback from unknown and therefore uncertain sources; the feedback is unrelated to their behaviour (actually as well as in the experience of the participating children). So, the children play a safe short-term game and maximise immediate benefit to themselves. Fairly rational, but not supporting what Wiseman and others claim.

    I’m not even sure that the study is all that ethical actually.

    • Don’t you think kids would usually give credence to the opinion of a researcher in an experiment like this? I am inclined to think they would. And as far as I have understood, we are rather good at reading meaning and correlation into random input – hence why we believe in gods and believe animal behaviour can predict earthquakes and so on.

      Even if the kids don’t fully believe what the researcher tells them, it could still be that they fear the researcher changing his/her mind about their intelligence if they perform badly. There are many situations where we don’t believe the positive pronouncements made on us, especially in cultures where flattery and polite compliments abound; I think this effect has real-world relevance quite possibly at all stages of life. Especially since doing worse at the third set of tasks, which were as easy as the first, is a pretty serious consequence and to my mind is self-sabotage – something I know a fair bit about. 😉

Leave a Reply

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

You are commenting using your 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