Are we failing to fully understand failure? - RSA

Are we failing to fully understand failure?

Blog 9 Comments

  • Education
  • Behaviour change
  • Social brain

Wow. What fortnight it has been!

The press coverage from our most recent publication about using behavioural insight in the classroom, Everyone Starts with an A, has been exciting, interesting, and somewhat exhausting.

We have been featured in the Telegraph, the Times, the TES, the Daily Mail, GovToday, and others.  Jonathan appeared on BBC Breakfast on Sunday morning, and Louise and Nathalie did various radio interviews over the past few days, including Drive Time, Radio 5 Live, and Voice of Russia.  And Jonathan has been engaging in lively twitter conversations on the topic. This is all in addition to the very welcome reception we got in Germany early last week, with articles in Die Welt, Die Sueddeutscher Allgemeine, and on the front page of Der Spiegel (with over 3,600 Facebook likes).

We have received many supportive comments from people who see the merit in trying out the approaches suggested in the report. It’s also fair to say, though, that not all of the comments have been positive and in fact many of them are from critics who seem to think that use of these approaches only serve to molly-coddle our pupils and fail to set them up for the harsh realities of the adult world.

But to think that these recommendations molly-coddle our youth is to misunderstand them or the science behind them. It seems that the people commenting like this are commenting on what they know from the newspaper articles, not from our report itself. So, in an attempt to clear up the confusion, let’s review the two most contentious recommendations from the report: everyone starts with an A, and give a ‘not yet’ grade instead of a ‘fail’. 


Everyone starts with an A.

Yes, yes, I know, we chose this as our title, so of course people will ask about it! But “Everyone starts with an A” is a lot catchier than the full explanation of “try giving your students an A at the beginning of the year, so that instead of working up to it they have to work hard to avoid losing it, because a well-known insight from behavioural science is that we tend to be ‘loss averse’ and thus are more motivated to avoid giving something up than we are to try to gain it”. It just doesn’t have the same ring to it. When months have been invested into a project such as this, it is not surprising that we want our ideas to broadcast and shared widely. So we did pick one of the most provocative recommendations from the paper in our title, in the hopes of generating interest in the report. And generate interest it did. But the flip side of picking a provocative title is that understandably all of the nuances and background behind it get lost throughout the chain. With each additional press release or article, the message gets somewhat warped, and “everyone starts with an A” can turn into “give students something they don’t have to work for” (for the record, that is NOT what we are suggesting!).

One misconception from critics is that this makes the A meaningless. Perhaps the first thing to say here is that, in our approach, just because you start out with an A doesn’t mean that you keep it. So at the end of the term (project, year, etc) it will not be the case that the entire class has achieved an A (but wouldn’t that be great?); some students will get an A and others won’t. The criteria for maintaining the A and how readily it is lost will be up to the teacher. So the A is no less meaningless or meaningful than it is in a traditional method of assessment. The key here is simply reframing the starting point so that instead of starting from the bottom, the pupils are starting from the top.

This reframing serves two purposes.

First, it is likely to improve effort levels. We know from behavioural science that people feel the pain of a loss more so than the pleasure of a similar gain. Decades of research by Nobel Prize winning psychologist Daniel Kahneman (who is speaking in London tonight) and many others in the field support this. And this concept has been tested in the classroom by Levitt et al, who offered students a reward if they improved on a test. For some students, they were told: take the test, we’ll grade it, if you’ve improved then you’ll get the reward. For others, they were told: here’s the reward (given upfront), take the test, we’ll grade it, if you’ve improved then you’ll keep the reward but if you haven’t improved then we’ll take it back. It’s worth repeating that both groups had the same  basic set-up of “be rewarded for showing improvement” and the only difference was whether the reward was given after the fact or before (with the threat of losing it). The researchers found that the second group performed better on the test, illustrating that this type of reframing may influence pupils effort levels.

Second, when a pupil starts out with an A, this plants a seed in their mind that they too are capable of achieving this grade. It is not reserved for the “brainy” or other select few; with the right effort, persistence, and support they too have the potential to keep this grade. Shifting pupils’ expectations can influence their own performance in meaningful ways.

Critics have also claimed that if you start with an A, the only way is down. This is responded to in more detail in this previous blog post, but the quick reply to this is that throughout the year a pupil’s grade might rise and fall in either approach – whether you start from the bottom or the top – and the point of our method is to change the starting point. (An even quicker reply is in this blog post, possibly the shortest ever on the RSA website).


Give a ‘not yet’ grade instead of a ‘fail’.

This is NOT, as some people have interpreted it, about “banning negative words”. This approach is about communicating the setback in such a way as to focus on the process of learning rather than on personal traits. As explained by Jonathan in this previous post, “Failure will always be an important part of any learning process, but there is a huge difference between communicating failure in a way that signifies you are failing because you are not good enough, and failing because what you have done is not yet good enough.”

Educational psychologist Carol Dweck explains that the way we think about intelligence, ability, and performance sits along a spectrum between two very different mindsets: the fixed mindset and the growth mindset. A fixed mindset is the belief that ability is a relatively unchangeable trait; in contrast, a growth mindset is the belief that ability can be improved with practice and persistence, just like a muscle can be strengthened through lifting weights. Dweck’s research shows that a fixed mindset student will be proud when they get an answer correct, but when they come up against a set-back, they lose motivation because to them it signifies that they have reached the capacity of their (fixed) ability. Growth mindset students also feel proud when they do well, but they attribute it not to a personal trait but instead to the steps that they took to get there. When growth mindset students encounter a setback, they are more likely to stay resilient and motivated, because they understand that with more practice and support they can improve next time. Our suggestion to give a “not yet” grade instead of a “fail” is one way in which a growth mindset can be encouraged in the classroom.

There are other ways of fostering in pupils a growth mindset, too, beyond this ‘not yet’ approach. For example the way that teachers and parents give feedback can have affect mindsets. Praising for ability or intelligence (“Great! You’re so clever!”) fosters a fixed mindset, whereas praising for the process (“Great! You really focussed while working on that problem set!”) helps to build a growth mindset.  Additionally, teaching students about the brain and how connections are made while we are learning serves to both explain that indeed we can continue to learn and improve throughout our lives, and also highlights the universality of learning and thus might especially help those pupils who self-identify as being part of a group around which there are certain stigmas or pre-conceptions (based e.g. on gender, ethnicity, or socioeconomic status).

In a recent email to us in response to the report, one reader, the head of psychology in a large college, wrote in with a wonderful quote which I think sums up the spirit of this chapter: “My experience over many years has lead me to a mantra which I bore my students with, ‘there is no such thing as brainy, there is just hard worky’.”


Free download available

Hopefully these two explanations will go some way towards clarifying these two particularly attention-grabbing recommendations. We encourage people to read the report here and to download the accompanying poster.

I don’t want to leave the impression that all the responses to the report or the recommendations therein have been critical. We have received many, many kind comments and supportive emails; thank you to those who have written in. The report has been described as “brave”, “hopeful”, and “insightful” and many recognise that the behavioural science underpinning these recommendations has a lot to offer to educational policy and practice.

Join the discussion


Please login to post a comment or reply

Don't have an account? Click here to register.

  • If talent/genetics does seem to play some role – as research
    seems to suggest, and we both seem to agree – then neat, but entirely
    one-sided, quotes like ‘there is no such thing as brainy, there is just hard
    worky’ don’t really have any useful place, do they?

  • Hi Matthew. You and I have had this discussion (several
    times!) before. So I know that you already know my answer, but for the benefit of anyone else reading the comments I'll add it here. "Talent" or genetics play a role, but what we're learning is that these are less fixed and deterministic than previously thought. Fields such as Neuroscience and epigenetics show that 'we' (including our brains and gene expression) are malleable. Understanding the constraints and enablers of this malleability may help us to change in constructive ways. And, yes, please read the paper, I think you'll enjoy it! :o)

  • Hi Nathalie,

    ‘There is no such thing as brainy, there is just hard worky' might well be a powerful and effective thing to say to students.

    But whether it is also true scientifically is another matter - I don't believe it is true and I don't believe it really reflects what Dweck thinks, if you look closely. (You have to look very closely, as she perhaps downplays this aspect of her work significantly).

    I've already highlighted a number of times (eg on Matthew Taylor's blog) that it's a Gladwellian myth that 10,000 hours of 'hard worky' makes you an expert. Recent research has shown that this isn't the case. A more balanced - nature and nurture - explanation is needed which, of course, is not liked by advocates of a totally malleable 'blank slate'.

    Dweck herself says "Certainly, people may have different genetic endowments and aptitudes to begin with" though "intelligence is greatly effected by nongenetic factors and is not static".

    I personally much prefer this kind of both/and - integrative - approach: it's a mix of brainy and hard worky.

    This is what the research that undermined Gladwell's 10,000 hour rule found. It's what Dweck says, as I've quoted here.

    Despite being viewed as another advocate of unlimited malleability and a 'blank slate' Guy Claxton says that there "do seem to be… genetic differences in people's intelligence" but there is "a wide envelope of variation around that 'base point' that depends on experience, encouragement and self-belief".

    My feeling is that we can either keep playing into dichotomies and end up caught up in the to and fro of ideological games, or else we can deliberately go for a more both/and, deliberately integrative approach. Maybe you've done that, I don't know. (NB Guilty admission: I've not had time to read your report yet, too much pressing stuff on my plate right now. I'm looking forward to reading it though).

  • Bob, Thanks for your email and i look forward to speaking in person. Nathalie

  • Hi Natalie,
    Thanks for the reply - if I could just focus in on one aspect ... The value of formative assessment does seem to have been embraced at schools I have worked with. The negative (according to Black) impact of summative assessment, less so. I wondered whether, suggesting 'starting with an A' (summative grade) implied that the Black work had been discredited. I think it would be fair to say that Black suggests that, regardless of what grade is given, the impact is worse than no feedback.