A Deep Dive into Desirable Difficulties
Jonathan Firth's Memory & Metacognition Updates #99
Hello, and welcome back to Memory & Metacognition Updates. I hope you had a wonderful break.
There is a lot to share this year! This week, as promised, a post that focuses on the fascinating and important concept of desirable difficulties.
What are desirable difficulties?
As a reminder, or if you’re new (in which case, thanks for joining us!), certain difficulties in learning are beneficial. Difficulties such as spacing out practice (the spacing effect – see update #39), and retrieval (see update #36) are called ‘desirable’ because they lead to better learning (in terms of retention and transfer), in comparison with easier strategies such as re-reading.
The latter are associated with fewer errors during practice, and students tend to choose them when studying. But they lead to worse outcomes.
Which difficulties are desirable?
Researchers Robert and Elizabeth Bjork, who coined the term ‘desirable difficulties’, have expressed some concern about it being misinterpreted. They say (Bjork & Bjork, 2023):
“…from an instructor’s standpoint, it may lead to the assumption that simply introducing difficulty into one’s lectures or lessons is a way to achieve better learning by your students. We have had to emphasize that the word desirable is important, and that many difficulties are undesirable during instruction and forever after.”
In other words, this is not about making education as challenging as possible, and certainly should not involve making explanations needlessly obscure or high-level. A suitable level of challenge will depend on the students and their stage.
What makes a difficulty desirable?
This leads to the question of what makes some difficulties desirable while others are not.
The answer is simpler than you might expect:
Difficulties are desirable when they boost learning, not performance.
To do so, we need to mimic the conditions under which the material will later be retrieved and used.
For example, when learning to drive, it would be easier to practice by driving round the same block multiple times, with an instructor sitting beside you and telling you exactly what to do. As a learner under such conditions, you’d make very few errors, if any.
However, once our lessons are over, we have to drive without an instructor telling us what to do, on complex and sometimes unfamiliar roads.
The desirable difficulties framework would suggest, therefore, that practice should resemble that realistic situation, with a variety of road conditions to deal with, and reduced guidance or feedback.
Impact of desirable difficulties
The impact of desirable difficulties is again relatively simple to summarise. As they involve an increased level of challenge, applying these in learning situations will lead to:
Slower progress,
More errors,
or both.
These are short- or medium-term (performance) problems, however. Over the long term, the techniques lead to better recall and better transfer, as mentioned above. In other words, they improve learning (see update #89).
Major examples
These considerations lead directly to the following techniques, many of which I have discussed on these updates before:
1) Variation of the conditions of practice
One difficulty discussed throughout the literature is variation in the conditions of practice (e.g. Bjork, 1994). Variation makes things more difficult because the context of each practice opportunity is less recognisable, and is less likely to cue (remind us of) the correct response.
In the driving example mentioned above, such variation might include practising on multiple roads. For a student practising research skills, they could do multiple types of tasks, work in varied groups and contexts, and so on.
“Critical information needs to be multiply encoded, not bound to single sets of semantic or situational cues” (Bjork, 1994, p. 188).
Practice which is varied helps with application to new contexts, and hence with creativity. Competence and expertise gained through repetition leads to a lack of flexibility in thinking, while varied practice leads to the ability to transfer and innovate (Gube & Lajoie, 2020).
2) Spacing
Spacing out practice leads to more difficulty and worse performance because it is harder to complete a task if there has been a delay since you last practised. This sounds like a bad thing! But as with other desirable difficulties, the short-term cost of waiting before you practise is better for learning in the long term.
Practical examples might include practising a skill after a few day’s delay, or reviewing new factual material a month or so after you first engaged with it.
It’s worth noting that the initial practice session needs to be effective, in the sense that it leads to high levels of performance. It’s usually not enough for the teacher to briefly mention something, for example. The students need to have really got the hang of the new task/concept.
3) Reduced feedback and guidance
Another strategy that the Bjorks and their colleagues discuss very widely is reducing feedback. This may be the difficulty that feels most counterintuitive to teachers – we tend to see it as our role to correct our students and give detailed comments on their work.
However, feedback also makes things easier. There is less incentive for the learner to think deeply and figure things out. It is also unlike most real-world situations – typically we don’t get expert feedback while we do tasks (teaching being a good example), but rather need to work out how things are going from subtle context clues.
Closely related to the above, could we reduce the level of guidance and help to students?
A ‘sat-nav’ approach, where step-by-step guidance is given, makes performance easier, but reduces the demand of a task. It is, perhaps, easy to see why this could undermine learning.
Of course, there are limits. The degree to which it is sensible or feasible to reduce guidance will depend on the task and on the students (see ‘level of difficulty’, below).
4) Retrieval
As already mentioned, retrieval practice can be seen as a desirable difficulty. The process of retrieving material from long-term memory is more error-prone than re-reading the same material, but better for learning.
We can therefore contrast active retrieval from memory with alternatives that apply to the classroom. A teacher may decide to summarise a concept to their whole class, with the class sitting and listening. A more effective version of that task would involve asking the students to write the summary, perhaps followed by peer assessment to check that they incorporated the key details.
In independent study, retrieval practice involves moving away from re-reading of notes or slides and towards self-testing and brain dumps.
5) Interleaving
Interleaving has some similarities to variation (see technique 1, above), but is less about the context or location of practice, and more about the content. In particular, interleaving involves varying the order of examples and problems, contrasting different types.
Interleaving tends to feel more challenging to students compared to looking at multiple examples of the same type, but as with the other strategies mentioned, it leads to better learning and transfer.
I won’t say too much more about interleaving today, as I wrote two posts about it very recently. Check out the first of these below.
Further questions
The techniques described above are powerful and important, and can, when fully understood, be applied to an enormous range of learning situations.
All the same, many questions remain. How difficult should each one be? In what types of tasks do they work best? Are there any other desirable difficulties?
As this has already been quite a long post, I will tackle these issues in a couple of weeks’ time – and also share some follow-up reading.
Take care, thanks for reading, and happy New Year! 😎
Jonathan
I hope you are enjoying Memory and Metacognition Updates. Subscribe for free to receive new posts and support my work – and consider sharing it with colleagues!! Thank you!
Please note that my slides and similar materials are shared under a CC BY-NC 4.0 license. This means you can use or adapt them with attribution for non-commercial purposes. If you wish to use my materials for other purposes, feel free to get in touch.