Centre stage paper – Days 1 & 2: The challenge

An extract from
A new shape for schooling? Deep Learning – 1

 

Our working definition of deep learning is this:
‘Deep learning is secured when, through personalisation, the conditions of student learning are transformed’.
In terms of our specification of the learner in a school where personalisation is embedded (see A new shape for schooling?), deep learning refers to the section in bold:

An articulate, autonomous but collaborative learner, with high meta-cognitive control and the generic skills of learning, gained through engaging educational experiences with enriched opportunities and challenges, and supported by various people, materials and ICT linked to general wellbeing but crucially focused on learning, in schools whose culture and structures sustain the continuous co-construction of education through shared leadership.

Over the past decade there has been a shift in emphasis in many schools from teaching towards learning. Excellent teaching must be complemented by excellent learning. This leaves the profession with the issue: how do we secure the move from teaching to learning? And then from learning to the deep learning that will equip students for the 21st century world of work?

It should be stressed at the outset that we are not seeking to establish a new theory of learning. There are many theories of learning available and the complexity of the issue is such that it could not be covered in a short article. Instead, we are concentrating on the ways in which the gateways interact and overlap to change the conditions of learning and to create opportunities for deeper learning experiences. The term ‘deep learning’ first came into common usage in the 1980s when Noel Entwistle and colleagues published research that distinguished between deep and surface learning. Their distinction is outlined below.

Deep approach

Intention: to understand ideas for yourself

• relating idea to previous knowledge and experience
• looking for patterns and underlying principles
• checking evidence and relating it to conclusions
• examining logic and argument cautiously and critically
• becoming actively interested in course content

Surface approach

Intention: to cope with course requirements

• studying without reflecting on purpose or strategy
• treating the course as unrelated bits of knowledge
• memorising facts and procedures routinely
• finding difficulty in making sense of new ideas presented
• feeling undue pressure and worry about work

The recent trend in schooling appears to have been towards surface learning. In part, this is due to the current examination system (at key stage 3 and key stage 4 in the UK). This puts teachers under greater pressure to teach to the test and students to cram relevant facts that will be regurgitated – and then often promptly forgotten. The system and the teaching methods used to enable students to succeed foster student dependence on the teacher and lead to surface rather than deep learning. However, through working across the gateways in the deep learning cluster, schools are finding that they can rectify this issue and help students to become deeper learners.

The deep learning cluster contains student voice, assessment for learning and learning to learn (undoubtedly other gateways also have a role to play in the development of deep learning). These gateways are clearly focused on the learner and the development of learning. They require students and teachers to work in partnership, while placing emphasis on students taking more responsibility for their own learning and progress.

Over the last two years, the SSAT’s personalising learning team has encountered many schools carrying out groundbreaking work in one or two of these three gateways. Some schools have made excellent progress through working in two gateways at the same time. What has been less common is to find schools working across all three gateways in a coherent manner. Only now are a small number of examples starting to emerge. It is clear that where schools are working on two or more of the deep learning gateways, the effects are more profound. Work across at least two of these gateways demonstrates what David Hargreaves describes as high leverage innovation: it requires teachers to work smarter rather than harder, removing the danger of their becoming exhausted or burnt out. Better still; such work leads to a greater impact on students, as it is likely to be long-term and embedded.

The three gateways that are clustered to form deep learning interact in highly complex ways. Critically, all three gateways create conditions and opportunities for staff and students to engage in co-construction. Student voice research projects can quickly lead to student suggestions for improvement to systems and structures. In assessment for learning, close partnerships between students and staff can lead to the co-construction of assessment criteria and lesson planning. In learning to learn, students are encouraged to reflect on their own learning; at its best this enables them to develop individual learning plans with staff with a high degree of confidence. Working on any one of these three gateways makes it easier to progress on one or both of the others.

Clearly, the interactions between these gateways are fundamental to the development of deep learning. All three help to develop specific learning needs, and place engagement at the heart of the process. Together they help to develop confidence, independence and responsibility, along with meta-cognitive control. In turn, these are indicators of the student for whom learning has been successfully personalised. They contribute to the capacity for co-constructed approaches to teaching and learning. The three emphasise that learning is not something that is done to you; it is an active process in which you participate with your teachers. More crucially, when put together, all three help to develop what we are calling ‘the conditions for learning.’

The conditions for deep learning

1. Learning conversations

Conversations between students and teachers take place every day. However, it is critical that more of these conversations have a deeper focus on learning so that they can drive the personalisation of student learning. The most striking feature of the deep learning gateways is that they encourage teachers and students to enter into conversations that are crucially focused on learning. All teacher-initiated talk should be an invitation to student voice. We should actively encourage students to talk to us about teaching and learning issues, involving them as active participants in the learning process. Personalising learning depends heavily on this.

More understanding is needed on how to structure these conversations (both inside and outside the classroom). How do we know we are having a learning conversation? How do we ensure that the focus is on the development of a child as a learner? There is clear overlap here between the work of the personalising learning team at the Trust and the extensive research of Professor Robin Alexander (University of Cambridge) around dialogic teaching. Alexander’s research points to the fact that the most commonly used types of classroom talk (rote, recitation and instruction/exposition) often fail to provide the cognitive challenge that will help to develop students as learners.

What is needed instead is dialogue, which can be between teacher and class or group, teacher and individual or pupil and pupil. As Alexander says, ‘it is aimed at achieving common understanding through structured, cumulative questioning and discussion which guide and prompt, reduce choices, minimise error, and expedite handover of concepts and principles’. The difficulty is that, although dialogue may be one of the best ways to develop a student’s cognitive ability, it is also the most demanding in terms of teacher skills. The work of Alexander is helpful here, as he has developed five indicators of dialogic teaching. These indicators state that dialogic teaching is collective, reciprocal, supportive, cumulative and purposeful.

Three of these indicators are helpful in describing the criteria for a successful learning conversation:

• reciprocal
• supportive
• cumulative.

Any school seeking to make progress in this area will have to consider how often these conversations currently take place and how often are such conversations needed? The more pressing issue, however, might be when do these conversations take place? And even more difficult are the issues of how these conversations should be recorded and how the time is found for them to take place. Two of the ways that schools are taking this forward is through making greater use of the new technologies and by employing a greater range of adults other than teachers to engage in these conversations with students.

2. Meta-cognitive control

The ultimate goal of learning conversations is the development of meta-cognitive control. All students must acquire the ability to think about their own thinking in order to develop the capacity to select from the learning tools they have at their disposal, and to monitor and evaluate their progress.

The deep learning gateways help to develop such control. Through them, students are enabled to direct their energies towards developing the learning skills and knowledge that they have jointly identified with the teacher through learning conversations. For deep learning to occur, students need to internalise the standards set by their teachers so they can identify for themselves what they need to do in order to improve. Their meta-cognitive control must be developed to such a degree that they know when they need help from others and when they can go it alone.

3. Growing Learner Autonomy

The final condition for deep learning is that students are enabled to develop autonomy as learners. Learner autonomy would seem to be the goal of personalising learning, for it is at this point that students are able to make decisions about their own learning, decisions that will be entirely personal to them.

The question of autonomy in learning has been raised sporadically over the last 20 years. It is often an area of controversy, as the very term ‘autonomy’ suggests an attack on the traditional power relationships in schools and classrooms. However, learner autonomy does not mean that the teacher becomes redundant – indeed, the teacher’s role is fundamental to its development. Candy, in Self-direction for lifelong learning, argues that this is because the process of growing autonomy is dynamic and relies on the intervention of teachers. In practice, for students to take greater control over their learning, they have to be aware of the tools and strategies at their disposal; and these skills and strategies must be taught. Hence the development of learner autonomy relies on the input and guidance of skilled teachers. What is clear, however, is that, as learner autonomy grows, the role of the teacher must change. Learners who have grown in autonomy will want to have more control over what they learn, how they learn and when they learn. The potential for co-construction in such a climate is huge.

The rich literature on learner autonomy fails to provide us with any single agreed definition. A useful start perhaps is: ‘the autonomous learner is a self-directed learner who has the freedom to act independently’. The attributes of learner autonomy are outlined in figure 1. Autonomous learners are self-directed. They have an understanding of what they need to do to successfully complete a task and are able to undertake that task independently. Typically, they are aware of their own strengths and weaknesses and this empowers them to make decisions, including when to ask for help. They are aware of which tasks they should complete alone, and which require collaboration with others. They are aware of the appropriate processes and behaviours to use in different contexts. They have highly developed skills of critical analysis and reflection. This reflection will centre on the processes of learning, as well as its content and outcomes. Finally, they are highly self-motivated to apply the clear goals and success criteria established at the outset of tasks. These attributes can, in part, be developed through the deep learning gateways, and they are critical to personalising learning.

Figure 1: Growing learner autonomy  

Conclusion

We began with the premise that deep learning is secured when, through personalisation, the conditions of student learning are transformed. Schools are already making progress towards this through their work across one or more of the gateways in the deep learning cluster. It is clear that there are significant overlaps across these gateways and, similarly, there are overlaps and interactions between the conditions for deep learning themselves. For example, meta-cognition is developed through learning conversations. However, in order for learning conversations to be effective, students must already have some meta-cognitive skills, most likely developed through the successful use of assessment for learning strategies and through a learning to learn curriculum.  Similarly, students who are growing in autonomy may do so through either of the other conditions, it is not necessarily a sequential pattern. So the conditions are perhaps best displayed as a cycle. Students will, at different times, focus on different conditions and the conditions themselves are as much outcomes as they are processes of deep learning. What is clear is that all three conditions will, in the school working towards personalisation, be co-constructed.

Figure 2: The deep learning cycle

If we accept the conditions of deep learning, then this has implications for the traditional structures of schools. There is an issue here around the training of student and existing teachers to develop in their students the skills they will need to be successful learners. As a pre-condition for this work, teachers must be trained more in the facilitation of learning; equipped as experts in assessment for learning strategies and be much more familiar with the learning to learn approaches that help to develop meta-cognitive control within their students. If conversations are to be successful, then they must be a genuine invitation to student voice. In this area there is room for many more schools to involve their students directly in the business of schooling, rather than in the organisational issues of uniform and canteen arrangements.

Many questions remain - there will be those students whose socioeconomic background and family circumstances will always be a barrier to deep learning. Some of these issues can be tackled in school through work on the Every Child Matters agenda and through the pastoral intervention.  However, some of these problems are so large that one school alone cannot address them and a wider group of agencies will need to support students (see Deep support 1). The area of mentoring and coaching is central to supporting students in becoming deeper learners. There are also practical issues to be addressed, such as how to create the time needed for detailed conversations? Can this be done through the traditional lesson structures of most schools?  If not, what implications does this have for other areas of the personalising learning agenda? Similarly, can schools working towards deep learning use traditional pastoral and reporting structures? Some of these concepts are explored further in the SSAT pamphlet series, A new shape for schooling?

ABOUT THE AUTHOR

Ms Emma Sims is Development and Research Coordinator, Leadership and Innovation Network at the Specialist Schools and Academies Trust, in London, UK.

 

Go to top of page     Go to online discussion