‘Feedback’ is used to describe the high-pitched whine that emanates from a speaker when the audio waves are picked up by a microphone, or other audio-input device, and amplified back through the speaker in a continuous, building, loop. It is piercing, irritating and unwelcome.
This form of feedback is a useful metaphor for another meaning of the term; the act of providing students with guidance on how to improve. Regurgitated and amplified through misguided school policies and CPD sessions, feedback has become a cacophonous racket which threatens to drown out the everyday melody of teaching and learning.
The term feedback was coined by Norbert Wiener during the Second World War to mean where the outputs of a system are routed back as inputs, creating a virtuous circuit or loop that leads to improvements in the system. Despite the aim being to achieve a ‘desired state’ where the system works effectively, the corrective feedback itself was described by Wiener as ‘negative’ as the output indicated error or insufficiency.
“Negative feedback loops minimize differences between the current situation and the desired situation by feeding the outcome of an action back into the system. Any discrepancy between the outcome and the desired situation leads to a corrective action whose intent is to reduce the gap.” Wiener
In teaching, the act of correcting students obviously predated Wiener’s work. However, the field of Cybernetics, developed by Wiener and his contemporaries, subsequently influenced how we thought about dynamic, complex systems in industry, society and education.
The concept of negative feedback (identifying the error in an output and routing this back in to the system to reach the desired state) has taken hold in education and in many schools has become the most elevated aspect of the teaching process. We have placed increasing importance on examining the output of a sequence of lessons (i.e. the students’ work) to assess where it is deficient against out ‘desired state’. Corrective information is fed back in an attempt to create a virtuous cycle of improvement.
As the feedback frenzy took hold, the hours spent by teachers looping through this endless cycle amplified until we reached the current workload crisis.
[As an aside, you can see a similar slavish commitment to the feedback loop, over a longer time-period, with the regular data-collection and intervention cycles which have also driven up workload for teachers – “that student is two grades below target; there is an error in the system!”]
But I believe there is a fatal flaw in our application of Wiener’s ideas. It derives from what is often identified as the ‘desired state’ and how the negative feedback is routed back in to the system.
When we examine a piece of student’s work we seek to identify how it falls short of the piece of work we would really like to see. The model we hold in our minds (the ‘desired state’) enables us to identify the errors and feed these back. This feedback goes to the student, after all it is their work we are marking and if they know how the work is deficient they will be able to move it towards the desired state.
What we are forgetting is that the piece of work is the output of a sequence of learning, not the output of the process the student went through to create the piece of work. If the output falls short of the desired state, the fault might be at any point along the learning process. Why would the power to correct this shortfall lie solely in the hands of the student? If this were true, we would not have designed, delivered and guided students through the sequence of learning in the first place. We would have left them to get on with it.
This error arises when we see the student’s work as a product of their ‘skill’ and not as the outcome of the knowledge which they have, or have not, acquired through the learning process. We provide feedback to the student such as ‘add more detail to your explanation’ or ‘analyse the causes of this’, as if the problem lies with the fact that they’ve just forgotten that that is what they should do. The problem is that they haven’t acquired sufficient knowledge to produce the work. The ‘desired state’ is not the model piece of work we hold in mind, it is the schema of knowledge we want the student to acquire which enables them to produce the piece of work.
Remember that Wiener advocated feeding information about deficiencies back in to the system to close the gap between the actual and the desired state. In this case, the system is teaching and learning (not the production of the final piece of work by the student). The deficiencies in students’ work should therefore inform the teaching process; the system which produced the outcomes.
This interpretation of Wiener’s concept of feedback, as applied to an educational setting, profoundly changes what we, as teachers, do with the information we glean from examining students’ work. Rather than turn this information into feedback to the students themselves, we should use this data to inform our teaching.
In other words, we interpret insufficiency in students’ work as an imperfection in the learning process we designed and delivered, rather than a deficiency in the student.
This re-thinking of the feedback loop brings benefits over current practice:
- We stop the hours spent providing personalised, negative feedback to each student.
- The error-information informs the most powerful part of the process; the teaching.
- Teachers begin to see error as a result of the teaching and learning process rather than as a symptom of inadequacies in the student.
If we are to reduce the gap between what students know and what we would have them know, our energies should be directed at refining the teaching process following our assessments of students, rather than waste time telling the student what they did wrong.
One thought on “The best form of feedback is more teaching”
Fascinating idea that is causing me to think. These are my immediate thoughts.
1 ‘We’ didn’t design and deliver the ‘learning process’, we designed and delivered the ‘teaching process’
2 “If this were true, we would not have designed, delivered and guided students through the sequence of learning in the first place. We would have left them to get on with it.”
Seeing as students do indeed design their own learning processes, impacted by those processes that we designed for teaching then they can quite easily do most of the work them themselves.
3 It depends where the issue arose i.e. in the process or the content as to what the most appropriate feedback design might be
4 It depends on whether you see the process as that which you do as a teacher. If we talk ‘system’ then it depends on how you define the ‘system’ and consequently where and how feedback should be applied to influence the individual component.
5 The terms positive and negative feedback were ‘”coined” long before WW2. As a simple, Europe centric example we have windmills in the 16th century using centrifugal governors on windmills and then variously during the industrial revolution. We have amplifiers at the start of the 20th century and ignition timing to control motor car engines etc etc
6 We have both positive and negative feedback both often used to limit variation in output. As a constructivist, limiting variation in output is far from my objectve.
7 The best form of ‘feedback’ has always been more teaching and in a learning system that relies on autodidaxy when appropriate then this can easily be completed by the student.
8 Although it is useful to take a systems approach when describing/defining learning and/or teaching we need to be aware that the teaching system and the learning system are interlinked but also largely separate. One or indeed both is under the partial if not total control of the learner.
9 It is when we adopt a systems view os learning that we can start to see the benefits of AI to provide feedback and metacognition to avoid the teacher, because after all the teacher is probably the least informed component of the system and the one which should be replaced if at all possible.
All fascinating stuff and very relevant.
So yes I think it is difficult to argue that more teaching (instruction) is not the most effective form of feedback. However the examples you give above are not really designed (or if they are they are poorly designed) to put right errors in the process and or content. Generally they are just additional tasks which likely repeat the deficiencies of previous efforts.
I would applaud you attempt to try the link this idea of feedback to the skills/knowledge debate. I think it is inventive to suggest that current efforts provide feedback on the process when what is need is feedback on the knowledge.
Neither of these ideas can however be supported via rational debate, even if for some they exhibit face validity.
On the plus side, always good to adopt a systems approach in my view, among others of course.