Research Spotlight: Generating Learning

During lockdown, between the setting of remote lessons, dialogue with pupils and the endless search for shops still stocking pasta, my wife and I were asked to write a book about Generative Learning. This is part of a series of books started, and edited, by Tom Sherrington with his look at Rosenshine’s Principles of Instruction and that will continue with explorations of the research around Cognitive Load Theory, formative assessment and much more.

At first the timing seemed less than ideal. We were somehow going to have to find the time to research this book and write it, all whilst juggling the demands of being a teacher during unprecedented upheaval. However, after taking the time to revisit this paper on Generative Learning from Logan Fiorella and Richard Mayer we realised the timing couldn’t have been better. Writing this book was going to show us how to teach as effectively as possible during lock down and to help us prepare for any disturbances to learning that this next year may present.

Generative Learning in Action

The idea behind generative learning is to design learning activities in such a way that pupils can make sense of new material and integrate it into their schema ready to apply it in new situations. Generative learning doesn’t replace effective instruction, but it does help to answer the question, what happens next? How do pupils take what they have heard and read and turn it into something they understand? Importantly, for a remote learning world, it asks, how can they do this with a minimum of further guidance?

The research underpinning Fiorella and Mayer’s work suggests that learning activities are based around 3 stages that make up the SOI model.

  • Select – Pupils select information from whatever has been presented during instruction (from a teacher’s explanation, textbook page, graph, photograph, video etc.) that is necessary to complete the task set.
  • Organise – They take the information they have selected and turn it into a new form. This involves them thinking hard about it (using Willingham’s maxim that “Memory is the residue of thought”).
  • Integrate – Finally, the task requires them to reflect on how this new information links to what they already know about the topic. Linking new learning to their prior knowledge.

They suggest eight activities that have a strong research base to support their use as generative strategies. These are:

  • Summarizing
  • Mapping
  • Drawing
  • Imagining
  • Self-testing
  • Self-explanation
  • Teaching
  • Enacting

Many of these strategies are ones that I have used many times over the years but what this research allowed me to do was to see how I could use them within the SOI model to ensure that pupils were actually learning from them; not just doing them. When I reflected back on how I had used them in the past, many of the strategies had resulted in pupils largely copying material from place to place and not really carefully selecting it, organising it or linking it to their prior knowledge. For example, in the past a mapping task on generative learning might have resulted in something like this.

Here, key words relating to the theory have been copied out around a heading. If a map were to follow the SOI model it should look more like this:

Here we can see that information has been more carefully selected and organised around different sub-headings, requiring the student to think more carefully about what the information actually means. It could also be that they have had to link in their prior knowledge, such as the influence of Bjork’s desirable difficulties or the work of Ebbinghaus on the testing effect.

The same process applies to all the other generative activities. Pupils need to know how to summarise in a way that doesn’t mean copying out the sentences that look most important. Instead, strategies like Cornell Notes can be used to ensure it actually supports learning with pupils having to be more selective about how they summarise. The same is true for strategies like turning text into images or using self-explanation to check for, and deepen understanding. The problem is that these strategies, although highly effective, require more difficult work than taking the easy way out and transferring information from place to place and making it look pretty. For pupils to embrace these strategies they need explicit instruction in not only how to use them but why. Trying to achieve this during remote learning was very difficult and the results were mixed. The pupils who embraced it thrived but there will still those who would try and take the easy way out or who didn’t really understand the subtle differences I was asking them to make in the way they were working.

This term I am teaching pupils how to use some of these generative activities in class whilst we are together. I want my pupils to know how to use a concept map effectively, or how to summarise or test themselves, so that if we end up with more remote learning in the future, they are well prepared. Even without remote learning, these strategies are incredibly useful in the classroom and will come into their own during revision tasks. A little investment now in how to study effectively will pay dividends before long.

One of the most useful things about writing this book and researching the different strategies and their implications was looking at how they would work, and if they would work, in different contexts. Some, like enacting, work brilliantly with younger children who struggle to grasp abstract concepts, but will be of limited utility for our pupils. Doing a summary works great if you are dealing with large amounts of text but could be counter productive in science, or in some areas of geography, where material has already been summarised into diagrams or tables. As with any research it needs recontextualising by teachers on the ground who are experts in their own contexts and subjects.

I would urge people to consider generative learning and its implications for independent study. I think the last few months helped to reveal that gulf between those with the skills to study well and those without them. We need to help close that gap.

Quick take-away

Use the SOI model to ensure that the work we are asking pupils to do leads to meaningful learning. Are they simply transferring information from one place to another or are they having to carefully select and organise information into a new form and link it to what they already know?

Teach them the strategies that will be useful in the future so that when you ask them to summarise, map or draw something they know how you want them to approach it and why.

Mark Enser

Head of Geography and Research Lead

2 thoughts on “Research Spotlight: Generating Learning

  1. Pingback: Lessons in Revision | Heathfield Teach Share Blog

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s