Getting students to apply knowledge and skills in new contexts isn’t easy but some approaches are more likely to yield results, writes Harry Fletcher-Wood

Let’s say we’ve successfully taught students something new: they understand it, can apply it and retain it. We now face a harder challenge: helping students transfer their learning to new contexts. For example, if we’ve taught students to graph data in maths, we want them to do so unprompted in future lessons, and to think to use it in science, history and perhaps in a future career. This turns out to be very tricky.

To transfer knowledge to a new situation, a student must:

  • Recognise that their existing knowledge is relevant
  • Recall it correctly
  • Apply it successfully

Often, students fall at one of these hurdles. They treat graphing in science as a novel challenge, for example, or recall their original learning imperfectly. But if what we teach is to make a real difference to students, they must be able to apply it beyond our lessons – whether next week or in five years’ time. How can we encourage this?

A seminal review begins by emphasising that researchers disagree “about the nature of transfer, the extent to which it occurs and the nature of its underlying mechanisms”. The authors try to help by categorising transfer. We might ask students to transfer learning between places (from the classroom to the sports field), social situations (from individual work to group debate) and modalities (talking about something they have previously written about, for example). We may also ask them to transfer learning across more abstract boundaries: between knowledge domains (from maths to science) and functional contexts (applying academic skills in non-academic contexts).

The ‘nearer’ the transfer, the easier it is

The ‘nearer’ the transfer – the more similar the context – the easier it is. For example, it’s usually easier to transfer learning from English to history than English to PE; it’s easier to recall an idea in the same classroom than in a different one.

Most research examines near transfer: few look at whether a skill taught in science is used five years later in maths. So the first thing to take from the research is to be realistic. We may hope our lessons stick with students for life, but if they recall it with another teacher next week we’re doing well.

How can we promote transfer? A helpful review and meta-analysis looks at how the questions we ask and the quizzes we give can help. First, researchers found that asking similar questions helps students to recall what they know. For example, if I introduced similes by asking about the author “likening one thing to another”, the same form of words makes it more likely students will remember and apply this knowledge.

This may seem unsurprising (and limiting) but it’s useful. We can promote transfer by sharing the language and question stems we use across contexts. We can agree how we will talk about graphing across departments for example, or share question ideas with parents.

The review also emphasised the value of asking a variety of questions about a concept. The researchers termed this ‘elaborative encoding’. Sticking with similes, I might ask, “What technique is this? Why isn’t it a metaphor? What else could the author have likened this to?” Approaching the same idea from multiple directions and getting students thinking harder about it helps them to access it again in future. (I discussed this in a previous research review, here).

Finally, the authors found that quizzes can be a powerful way to create learning that “generalises to different situations and different test types”. We can also give feedback after quizzes, leading students to revisit, rethink and restudy what they have learned. The researchers have created a helpful practical guide to their findings for teachers.

We can’t guarantee students will recall what we teach them in future years, but we can make it more likely. First, we need to ensure students’ initial understanding. Then we can help students to transfer their understanding by preparing them (using elaborative encoding to help them access key ideas), prompting them (using similar language when asking them to recall those ideas) and challenging them (through retrieval practice and feedback).