How best to improve student performance? A recent study of the various factors that we generally expect to be effective (such as teacher-pupil ratio, school culture and so on) concluded that actually, the thing that has the greatest impact on performance is the quality of the feedback we give on a student’s work.
Improving the quality of feedback is the most cost-effective strategy for lifting our students’ game, and the best use of teachers’ time.
Of course, we have long understood the importance of feedback, and we know that we should criticise lightly and lay on praise with a large trowel. But it turns out that there is more to feedback than meets the eye.
Criticism, delivered in the best way, can be a vital spur to new effort; praise delivered in certain ways can do more harm than good. So what is the current thinking on optimising feedback?
Where do we set our threshold of expectation, the expectation that colours the extent to which we see a performance as adequate or otherwise? In the context of exams, that threshold is set by parameters of the task in hand, coupled with set notions of the standards the educational system aspires for its pupils to achieve.
The focus is on whether an answer is/isn’t right, or whether an essay is/isn’t adequately researched or written. That approach is necessary in judging the outcome of education, and it is tempting to transpose to the classroom when we are preparing students for those exams. But it is far from optimal in a pedagogic context.
Feedback works best when used tactically, in relation to the individual student’s pedagogic needs. As Vygotsky pointed out, skills develop in stages. At any one time, there are aspects of the task that have been mastered so well that they are virtually automatic now.
We do not need feedback in these areas to drive our learning forward. Equally, there are aspects of the task which are so far beyond our current capabilities that (however essential to the task per se) we cannot learn from feedback pitched at that level, it goes straight over our heads.
In between these is what Vygotsky called “the zone of proximal development”: the area of skill just beyond our present capabilities and which represents the next layer of skill to be mastered, but which is not so far beyond present skill to be incomprehensible. Optimal feedback focuses tactically, aiming precisely at this zone of proximal development.
Of course, good teachers naturally aim, more or less, at the zone of proximal development. But often, what we identify as being in that zone is driven as much by general aspirations as to where students in a given class ought to be as by individual pedagogic needs.
A sharper focus on the individual delivers more effective feedback – though research also shows that this approach has its own hazards...
Focus on individual needs, but not traits
Is intelligence a trait more or less fixed by our genes? What about mathematical or musical ability, or a flair for languages? Psychology once taught that all these abilities were fairly fixed and immutable traits, and many people still make those assumptions.
That perception often guides the feedback we give: in kindness, we go easy on those who seem not to have much ability for this or that (or anything); we even try to offer comfort, with comments such as “we can’t all be good at everything”. But research is showing that this way of tailoring feedback tactically to the individual is counter-productive.
Whether indirectly (by setting low standards for those we see as less able) or directly (by remarks suggesting, even obliquely and intending to help, that some people just don’t have certain abilities) feedback influenced by a teacher’s belief that the student lacks some underlying ability subtly conveys a disempowering and demotivating message, and depresses performance. How could I feel otherwise, if I take on board the message that I just don’t have the ability to do better?
And this is particularly unfortunate, since psychology would now not argue that the core skills are immutable! Almost every skill is in fact far more malleable than we once thought. Take general intelligence, for example: where once we thought we could test and pigeonhole IQs, we now know that it is quite normal for an individual’s IQ to vary over 20 or even 40 points. Certain circumstances (and expectations) can depress or elevate even this highly heritable trait. And if nothing is fixed, anything is possible.
Optimal feedback is tailored to the individual’s present pedagogic needs, but assumes – and conveys – no limitations on future possibilities.
Focus criticism on the detail
Received wisdom is that criticism is a necessary evil, to be kept to a minimum. But in fact, constructive criticism can be one of the more effective drivers of new development, a valuable source of information on how to build a better performance.
What research is showing is that it is general criticism that creates problems (“a poor piece of work” and so forth). Criticism at this level is demoralising, since its only message is one of inadequacy. And such general criticism is unproductive: it offers no clue as to specifically what was wrong with the work, or what specifically needs to be improved (still less how). In sum, general criticism carries little or no pedagogic value. Constructive criticism always focuses on detail. A focus at this level usually makes it possible to identify some detail to praise, generating a positive overall feel to the feedback.
Very specific comment on the elements of the task that have not worked out so well, feedback which not only explains the problem but offers suggestions for improvement can be highly motivating. Not only does such feedback have an intrinsic pedagogic value, but by entering the task in this detailed way, the teacher can be perceived as more of a partner in building that improvement – and such external support can play a bigger role than one might think.
Praise the specific
Not all forms of praise are productive. As with criticism, so with praise: general comments (“an excellent piece of work” or “you’re good at this”) can cause problems.
General praise is dangerous because it distracts attention from what really matters, which is specific skills that actually made the good performance. Instead, it encourages students to believe that their success rests on a trait ability – a perception that creates confidence, but a confidence which can be profoundly shaken when setbacks are (inevitably) encountered: has the trait vanished? And since it has no focus on skills, general praise offers no pedagogic support for overcoming setbacks. As with criticism, so with praise: a focus on the specifics points the student in the right direction: what matters, when mastering any task, is not a hypothetical “ability” but a very concrete focus on the skills needed for the task.
Detailed criticism and praise offers optimal feedback for shaping specific task-related skills, but what when the problem you want your feedback to address is primarily a lack of confidence?
Confidence is vital. Specific pedagogic feedback can’t have its optimal effect if learning is undermined by poor confidence. But how to give feedback that builds task-related confidence?
Praise that exaggerates current ability or success doesn’t work. You can not build confidence by over-egging the pudding in that way – teenagers see through that kind of thing at once, and it can come over as counter-productively patronising. But research suggests a different strategy: the cognitive placebo.
Performance improves when students think they have taken a mind-enhancing drug, even when they haven’t.
Pills and potions aren’t necessary for a cognitive placebo effect: simply convincing students that there is external support for their efforts (such as subliminal on-screen messages giving the answer) enhances performance, even when there is no such support.
It seems that these placebos work by providing a confidence (in an external support) that the student can not feel in his or her own resources. Simply telling students that they can do better on their own resources yields no parallel effect.
Obviously, classroom teachers cannot go around handing out sugar pills disguised as mind-enhancing drugs, nor pretend to students that the answers are being subliminally provided! But a study by Robert Hartley suggests a practical (and ethical) way to incorporate a placebo effect into feedback.
His study took a group of students who were typically performing poorly on maths tests and simply asked them to take the tests not as themselves, but “pretending to be someone really clever”.
Astonishingly, these students then performed at a markedly better level – though their performance fell back when they reverted to taking the tests just as themselves.
Pretending to be someone clever seems to recruit “external” support which builds confidence and releases better performance. Since the “external support” is wholly imaginary, this is a pure placebo effect – and it is a strong one.
This placebo is, in fact, well known to writers, performers and the like (try it yourself). It fits well with psycho-therapeutic advice for those wishing to change: fake it until it’s real – and sooner or later, the better performance does become one’s own. So why not recruit this placebo and incorporate the idea of pretending to be really clever into your feedback?
Dr Stephanie Thornton is a chartered psychologist and a former lecturer in psychology and child development.