My new book, Intelligent Accountability: Creating the conditions for teachers to thrive is out now.

The argument I make is that while accountability is wholly necessary for teachers to thrive it is too often applied unintelligently and so backfires. I discuss a set of principles designed to get the best out of teachers, thereby getting the best from your students. And when I say ‘best’, I categorically do not mean piling stress onto teachers in the hope of gaming exam results. By creating the conditions for teachers to thrive, we are likely to get much more of what we want: better exam outcomes; happy, rounded students; and teachers who are satisfied with their professional environment.

Here’s what Laura McInerney has to say on it:

For years education needed someone to explain what accountability actually is and how it can be more than a horrible punishment. This book delivers on that need, and then some. A magisterial sweep of research from psychology, sociology, and economics all with the aim of helping school leaders support genuine improvements in teaching as opposed to simply generating feel-good moments. Not just about accountability, this book looks deeply at what it means to be a moral, good and intelligent leader – an extraordinary read, thoroughly recommended.

The book is divided in seven chapters and, over the next few weeks, I’ll be releasing a series of posts which will make clear how to get accountability right (or at least, better.) Here is a brief overview of the ideas explored in each chapter:

Chapter 1: Why we need to embrace ignorance and learn to love uncertainty

  • We always make decisions with incomplete knowledge; no amount of data or information will allow us to make perfect decisions.
  • We are all prone to predictable cognitive biases; decisions made quickly or in isolation are more likely to be suboptimal.
  • Certainty makes us incurious and close-minded. We are likely to make better decisions if we can maintain uncertainty.
  • Experience often leads to overconfidence. but success is as likely to be due to luck as expertise.
  • Seeking collective knowledge and striving to align decision- making to a collective purpose can reduce some sources of uncertainty.
  • Sufficiency is more desirable than unrealistic perfectionism.
  • Overconfidence can lead to the best case and ‘this works’ fallacies; make sure you have planned for the worst case scenario and have a plan for checking your biases.
  • Being honest about our mistakes is likely to build humility and help us to avoid overconfidence.
  • Professional scepticism – being open to ideas but scrutinising them – is an important part of decision-making.
  • Teaching is a complex domain and therefore top-down decisions made about teaching should be made with particular caution.

Chapter 2: The surplus model of school improvement

  • The quality of teaching is one of the few factors within our control that is a good bet for improving the quality of a school.
  • Teachers are a scarce and valuable resource that needs to be nurtured and protected.
  • Teachers are more likely to increase their effectiveness in schools with supportive learning environments, concrete learning processes and leadership that reinforces learning.
  • School leaders need to believe that all teachers can improve, despite the fact that they may not know how best to work with certain individuals.
  • The deficit model of school improvement assumes that problems are due to lack of effort or expertise and seeks to apportion blame.
  • The surplus model of school improvement assumes that all teachers are well intentioned and working at capacity, and seeks to uncover reasons for problems.
  • The deficit model will tend to result in tighter accountability, whereas the surplus model is more likely to result in greater trust.
  • Deficit models create a culture of suspicion and offer incentives for people to behave perversely.
  • Schools can learn the value of distributed cognition from aviation.

Chapter 3: Trust

  • Trust is an essential component for the efficient working of schools. The less teachers are trusted, the greater the need for monitoring and accountability.
  • A policy of tit-for-tat – assume other people are trustworthy and willing to cooperate until they prove otherwise – is not only the most successful evolutionary adaptation for the smooth running of social groups, it’s also a sensible way to treat teachers.
  • Teachers thrive when they feel trusted; when we distrust others, their performance is more likely to decline than improve.
  • Trust is a crucial factor in teachers’ well-being.
  • Pascal’s wager suggests that the benefits of assuming teachers are trustworthy outweighs the costs of assuming they are not.
  • If we want to be trusted we must also be trustworthy; the two most important factors indicating our trustworthiness are honesty and expertise.
  • Honesty requires more than simply avoiding lies. Not only does dishonesty erode trust, it also enables us to ignore our mistakes and pretend that the negative consequences we experience don’t exist or aren’t our fault.
  • To be worthy of trust we must prove that we are knowledgeable and interested in evidence.

Chapter 4: Accountability

  • Accountability processes can lead to unsustainable workload burdens that make it harder for teachers to be effective.
  • If we are not held accountable for our behaviour, we are less likely to behave morally.
  • Poor accountability processes result in teachers trying to look their best rather than trying to be their best.
  • We are prone to thinking in false dichotomies. Instead of looking for weak compromises between competing extremes, we should be working out if polarised positions can be held in creative tension.
  • Using the process of thesis, antithesis and synthesis can help us to make better decisions and arrive at stronger conclusions.
  • Intelligent accountability makes teachers accountable for the trust we place in them.
  • Accountability is intelligent if:
    • Teachers know how they will be held accountable before judgements are made.
    • The views of the people holding teachers to account are unknown.
    • Teachers believe that those holding them to account are well informed and interested in accuracy.

Chapter 5: Equality, fairness and autonomy

  • Although the aims of equality are well intentioned they inevitably result in people being treated unfairly.
  • Treating teachers fairly means giving them the support and freedoms they need to be most effective.
  • The more teachers are trusted, the greater autonomy they ought to be allowed.
  • Autonomy must be earned; it should be clear to all teachers how it can be earned.
  • Teachers can be loosely categorised according to perceptions about their effort and expertise: high effort and high expertise, low effort and high expertise, high effort and low expertise, low effort and low expertise.
  • There is no sufficiently reliable process to confidently identify the most or least effective teachers: remember that you should cultivate uncertainty and embrace your ignorance.
  • Rights and responsibilities must be balanced; teachers and school leaders must be subject to same principles of accountability.
  • Consistency can be overrated; negotiate your non-negotiables.
  • It is more reasonable to make some things non-negotiable than others: anything that requires teachers’ judgement and expertise is best left negotiable.

Chapter 6: Improving teaching

  • The quality of teaching is a more important factor in school improvement than the quality of teachers.
  • Although we know being taught by the most effective teachers makes a huge difference to students, we don’t have the ability to reliably identify the best individual teachers.
  • Judging teachers by students’ outcomes is a flawed way to identify the most effective teachers; teaching more able students makes a teacher twice as likely to be labelled as effective.
  • Students’ results are as much down to natural volatility as they are to any school or teacher factors.
  • Lesson observation is even more flawed: on average, 63% of observational judgements will be wrong.
  • Teachers don’t simply improve with experience: effectiveness begins to plateau after the first three years.
  • We only continue improving in ‘kind’ domains – that is, environments where we get reliable feedback.
  • Teaching is, to some extent, a ‘wicked’ domain, in that teachers don’t receive reliable feedback on the quality of their teaching.
  • Instructional coaching is a good bet for nudging teacher improvement.

Chapter 7: Intelligent leadership

  • Too much training in leadership is based on faulty intuition and wishful thinking; better leadership development requires a greater understanding of, and engagement with, research and evidence.
  • Leadership is a domain specific not a domain general skill: school leaders need to be more expert in education.
  • The primary role of school leadership is to remove extraneous demands on teachers so they can focus on planning and teaching the very best curriculum possible.
  • The traits that we would like leaders to possess are often at odds with the traits that get leaders promoted and make them successful.
  • What seems to make leaders successful is often illusory. Successful leaders have been lucky in the past but will eventually regress to the mean.
  • Senior leaders are likely to occupy a bubble where they only see and hear what those around them think they want to see and hear.
  • The constant drive to make schools outstanding can easily undermine efforts to be effective.
  • Leaders need to be particularly aware of the limitations and effects of using metrics to measure performance.
  • We are likely to make decisions on the data we have available, whether or not it is reliable or valid.
  • Predicting individual students’ outcomes is a very poor use of data.
  • Remember Goodhart’s law: when a measure becomes a target, it ceases to be a good measure.