Why is it so difficult to improve teaching and learning across a school and what can we do about it?
Isaac Newton is a legend of science. He stands out even amongst the greats of the Enlightenment era, a period of history that brought us the modern world, built upon the scientific method, industrial technology, political philosophy and economics. As an early adopter of the scientific method, Newton made great strides in understanding the motion of objects in our universe. His laws of motion could predict the exact location of planets in the night sky hundreds of years into the future. They are so good that they still form the backbone of high school physics today.
During the Enlightenment era, many believed that the ‘Age of Reason’ would eventually answer all the questions of the day. To some extent, they were right. The impact of the scientific method on our knowledge of physical universe (and resulting technological advancements) is staggering, as is the improvement in life expectancy, wealth, and the consequences on our environment.
Physical and social systems are very different beasts
However, Newton himself was much less able to predict the social events that would unfold, even one day in the future. After losing a fortune on stock market speculation, he said: "I can calculate the movement of the stars, but not the madness of men." He did not understand social systems. Nobody really does! We find ourselves in 2022, over 300 years later, still unable to predict the movement of stock markets, in the grip of inflation, and unable to answer a very important question: “How do we improve teaching and learning across a school?” - Yes, this article is about teaching and learning!
We do know something about this question, but nowhere near as much as many would have hoped 300 years ago.
Why is it so difficult to know what needs improving and how to do it?
One of the greatest barriers to school improvement is knowing what the current state of a school and its teachers really is. In 2015, Robin Hogarth & Emre Soyer described a framework for thinking about systems, which can be thought of as learning environments. Each learning environment can be placed on a continuum between ‘Wicked’ and ‘Kind’. A kind learning environment is predictable and has constrained conditions where patterns regularly repeat; whereas a wicked learning environment is unpredictable and has unconstrained conditions where patterns rarely repeat, making them difficult to identify.
The following table compares the main features of kind and wicked learning environments:
Source: Randy Fath – Unsplash
Kind Learning Environments
Predictable - closed system
Constrained conditions - clear rules
Feedback is unambiguous - clear links between cause and effect
Discrete skills that are suitable for Deliberate Practice
The right lessons often learned from experience
Best practice - one best way to reach a desired outcome
Examples: Chess, Accountant, Pilot
Source: CDC - Unsplash
Wicked Learning Environments
Unpredictable - complex/open system
Unconstrained conditions - vague/no rules
Feedback is ambiguous - debatable links between cause and effect
Complex skills that are unsuitable for Deliberate Practice
The wrong lessons can easily be learned from experience
Emergent practice - Many routes to a desired outcome (equifinality)
Examples: Teaching, Politics, Pandemics
The classic example of a kind learning environment is the game of chess, where there are many constraints and a finite number of options. In a game of chess, feedback loops are quick and clear, allowing for reliable lessons to be learned from experience. As a result, chess appears to lend itself towards deliberate practice, a concept coined by the late Anders Ericsson whereby a discrete set of skills and routines can be practiced in order to become an expert. This is also why computers can now beat the best humans at chess and other games.
Now think about a typical classroom from the teacher’s perspective. Instead of 32 highly constrained chess pieces, you have 32 unique, intelligent and dynamic children. What are the constraints? How regularly do patterns repeat? A classroom is a classic example of a wicked learning environment where teachers struggle to identify patterns because the feedback is often vague, slow, and ambiguous (learning is hard to measure!). In these environments, the wrong lessons can be learned from experience, and deliberate practice is less applicable because the discrete skills and routines are not obvious. This is particularly true for experienced teachers who have already learned the more discrete skills that most of us learn in the first few years of our career. It is also why you don’t see any computers replacing classroom teachers, yet.
No silver bullets
It’s ironic that classrooms are such difficult places for teachers to learn. For a concrete example, look no further than the initial stages of the pandemic when schools were suddenly forced online with little or no time to prepare. The terms ‘synchronous’ and ‘asynchronous’ learning were almost unknown to the profession in 2019 but ended up at the centre of heated debates about ‘what works’ for remote/online teaching in 2020. As a profession, we made great progress in working out when synchronous or asynchronous methods might be more effective, but there’s so many factors that must be considered when making the choice - there is no simple answer.
Interestingly, a common feature of social systems, like education, medicine and politics, is a desire to find simple solutions to resolve knotty problems. There’s always plenty of people willing to offer an elixir but, unfortunately, they are rarely in possession of it. This isn’t a new phenomenon. The word ‘charlatan’ comes from 16th century France where tricksters would attract crowds with music before trying to sell cure-all medicines. Does this remind you of one or two keynote speakers you've seen at conferences over the years? Brain Gym, anyone?
Whilst the complexity of teaching leaves it exposed to tricksters offering simple solutions, I believe that the uncertainty about best practices is what makes teaching such an amazing profession to be a part of. If the solutions to improving education were simple, they would have been resolved long ago and we’d have superstar teachers competing in teaching championships, just like we have grandmaster chess players. Personally, I relish the challenge of improving learning under these wicked conditions, and embracing the complexity is an important place to start.
It’s not rocket science; it’s discipline
Famous scientists like Newton and Einstein, have a popular image of making great leaps forward based on moments of inspiration, like an apple falling on their head. But the truth is that useful knowledge emerges over time from many sources. Newton himself said, in a rare moment of humility:
“If I have seen further, it is by standing on the shoulders of giants.”
It’s a wonderful phrase but I see a single giant: Scientific inquiry, guided by a method of making predictions, rigorously testing them, and updating predictions based on the results. It is the discipline of the method that has delivered these moments of inspiration.
Action, despite the complexity: Learning to Improve
Complexity causes uncertainty, which in turn can cause us to feel stuck or demotivated. But the opposite is true. Because we can’t easily predict the consequences of our actions in schools, we must act and learn from it. As Teddy Roosevelt once said about leadership “Do what you can, with what you have, where you are.”
Anthony Bryk and others, from the Carnegie Foundation for the Advancement of Teaching, distilled decades of research and experience from the field of improvement science into their book Learning to Improve. They offer a powerful account of how a different method, the Agile Learning Loop, has transformed improvement efforts in social systems like schools and hospitals. There’s a lot more to improvement science than the Agile Learning Loop but, in effect, it provides the discipline of a method, like a map guiding us through the complexity.
An outline of the method can be seen in the image below:
The Agile Learning Loop is practical:
Unlike scientific research, the purpose of the Agile Learning Loop is much more practical and is stated as follows in Learning to Improve:
Focus on improving actual day-to-day work in classrooms
Focus on multiple quick tests of change and iterative improvement
Recognise that variation in organisational contexts is a core design and development challenge, rather than just some externality to be ignored.
Orchestrate diverse expertise between researchers, practitioners, designers and developers
Reframe the overall mission as accelerating how a field learns to improve its core work.
Let’s unpack this a little more with direct reference to the 3 parts of the learning loop:
Standard work process is the target of our improvement efforts. It should be grounded in commonly accepted standards (e.g., coaching, formative assessment tools, safe-guarding, behaviour/relationship etc), whilst also taking account of the unique context of the school. One way of thinking about it is the phrase “What we share in common [in how we learn and behave] is far greater than what divides us”. However, our differences matter too much to ignore – you can’t just dump an instructional coaching program in a school and expect it to work like it did elsewhere, or even expect it to be the right thing for that school at all.
The Improvement hypothesis is a lot like a scientific hypothesis, but it relies more on the experiences of those making the predictions because the context is so important and the knowledge base in education is far weaker. This is something that education could do so much more about if only there were better connections between academic researchers and schools. Just imagine if all university education departments teamed up with networks of schools to work out how to implement ideas at scale.
Practical measures: Bryk claims that we cannot improve at scale what we cannot measure, and I have been convinced by his argument. Essentially, the purpose of practical measures is not to control every variable and try to prove that something works. Instead, it is intended to get just enough useful information to judge whether a change is making the intended improvements. It’s about judgement, not proof. Also, by attempting to measure something, we learn more about the effectiveness of the measure itself. This means we improve the measures over time, not just the target of our change.
A method is only as useful as the spirit in which it is applied.
The Agile Learning Loop emerged from the continuous improvement movement, known for its impact on the Japanese Car Industry, The British Olympic Cycling Team, and Silicon Valley. In the case of Toyota, their Kaizen approach to improvement wiped the floor with their American competitors, who eventually fought back decades later by applying the same method. It took Ford and GM a long time to respond because Kaizen is not an easy thing to implement. A method is only as useful as the spirit in which it is applied. The managers of Toyota’s factories did not just go to the factory floor to solve problems. Instead, they systematically visited the floor, talked to front-line workers and created a culture where useful feedback emerged over time. These small, iterative improvements accumulated over time to deliver massive results.
The spirit of this method is very similar to the one that scientists like Isaac Newton follow: “Let's focus on being a little less wrong tomorrow than we were today”. The key phrase here is ‘less wrong’. It’s the humility of recognising our own ignorance, and the willingness to become less so over time.
The Great Task of our generation:
The consequence of a Learning to Improve approach is a cultural shift within these schools: As Bryk says: “It means recognizing that today's problems cannot be solved through isolated individual actions. Each participant holds expertise that is valuable in solving a given problem, but each also recognises that he or she must join with others to solve it.” In my view, this is a profound shift away from the atomisation of teacher and school development. Only a team with a culture of continuous improvement can achieve the changes we seek in education.
This approach challenges other cultural norms found in many schools such as the tendency to blame individuals when things are not working as they should, instead of the system or processes in place. The evidence is quite clear on this and is captured nicely in Learning to Improve: “Individual personnel account for a tiny percentage of an organisation's overall performance. The predominant causes of failure lie in how we organise the work that we ask people to carry out.”
Another common feature of schools is the self-sacrificing, ‘heroic’ teacher/leader, which causes misery for so many educators and puts a huge number off even applying for school leadership. On this point, it links to Diversity Equity Inclusion and Justice (DEIJ). DEIJ is a systemic issue that goes far beyond the scope of this blog. However, too often, school improvement initiatives are concocted by well-intentioned senior leaders without enough consultation with key stakeholders. To be clear, as evidenced in the previous paragraph, this is a systemic problem, not the fault of individual leaders. But the impact is captured well by computer scientist Telle Whitney “When we limit who can contribute, we in turn limit what problems we can solve.” I believe that many schools would focus on more impactful improvement initiatives if there was a culture of continuous improvement, supported by a method like the Agile Learning Loop, to ensure broader input from the whole community.
Whilst Isaac Newton was a great thinker about physical systems, I think Atul Gawande is a great thinker about social systems. In a Ted Talk about how to improve health care, I think he has thrown down the gauntlet for us all in education:
"Making systems work is the great task of my generation of physicians and scientists. But I would go further and say that making [social] systems work – whether in healthcare, education, climate change, making a pathway out of poverty – is the great task of our generation as-a-whole.”
Next time: But what should we Improve?
The Agile Learning Loop tell us how to implement improvement, but it doesn’t tell us what to focus on. That requires a solid foundation of knowledge. If Isaac Newton was standing on the shoulders of giants, who’s shoulders are we standing on when deciding what to improve in our schools?
If you can’t wait for that, check out Harry Fletcher-Wood's blog where he dissects the Education Endowment Foundation’s report on effective professional development:
