When Does Giving Children a Choice Help Them Learn?
Give someone a bit of control over how they learn something, and they tend to remember it better. That finding turns up again and again, across ages and subjects.
A new paper published this year complicates things a bit. The researchers separated two things we usually lump together: having a choice, and being able to work out what that choice will lead to. The benefit of choosing showed up either way, but it shrank noticeably when people could not predict the outcome.
That is more useful than the general finding. It shifts the question from whether to offer choice to what kind of choice is worth offering.
It also lines up with a few other papers we have collected over the years.
Feedback that leaves the next move open
There is a similar point buried in the research on feedback. Susanne Narciss's work on elaborated feedback argues that when a child gets stuck, the most useful feedback often avoids handing over the answer. Instead, it gives them enough to understand what went wrong and work out the next step themselves (productive struggle).
Put next to the choice research, that makes sense. Good feedback keeps the child in the position of choosing, while making the consequences of that choice legible. It closes the gap between deciding something and understanding what your decision did.
A pause before the choice
Stop and Think, which came out of the UnLocke project and which I had a small hand in building, goes at this from another angle. It deliberately makes children wait before they can answer. No option to fire off the first thing that comes to mind.
On the face of it, that is the opposite of giving someone choice. In practice, it might be what makes a real choice possible. Without the pause, a child mostly reacts to whichever answer arrived first. The constraint creates the space in which a genuine decision can happen.
Choosing how to engage, not just what to answer
Elaborative interrogation is the technique of asking children why something is true rather than simply telling them. Peer consolidation research points somewhere similar: children who explain material to someone else tend to learn a good deal from doing it. Part of the reason seems to be that when a tutee asks something unexpected, the tutor has to build a fresh explanation on the spot, and that construction work is where a lot of the benefit sits.
Both of those give a child something to choose about beyond the answer itself. How to explain it. Which way in to take. That feels like a richer kind of choice than picking from a menu.
The finding that points the other way
Professor Paul Howard-Jones has done work on uncertain reward in learning games, and the finding runs in a different direction entirely. When the reward for getting something right is unpredictable, a spin of a wheel rather than a fixed number of points, engagement goes up rather than down. The uncertainty is doing the work. It is the same mechanism that makes fruit machines compelling, borrowed for better ends.
We built that into oodlü years ago as a wheel of chance after a correct answer, and it does exactly what the research suggests it should.
So does predictability help or hinder? Both, and the two findings are describing different parts of the same sequence. The choice research is about the consequences of your decisions being understandable. You need to know what pressing this button rather than that one will do. The reward research is about what happens after you have already done the work. There, a bit of surprise adds energy.
Predictable consequences. Unpredictable rewards. Those two can happily coexist, provided you are deliberate about which one you are using and where.
That last part matters most. Uncertainty dropped in at the point a child is trying to work out what to do next is just confusion. The same uncertainty, arriving after they have committed to an answer, is a reward. What separates the two is the structure around it: how the question is framed, when the choice is offered, and what happens in the moment after. Get the sequence right and both mechanisms pull in the same direction.
The practical bit
None of this tells a teacher or a parent what to actually do, and I am wary of pretending otherwise. I build software rather than teach, and the gap between a finding in a paper and a room full of thirty children is not one I am qualified to close.
But it has changed how I think about the choices we put in front of children in our own work. More options are easy to add and feel generous. Options where a child can see what each one will do are harder to design and probably matter more.
If you work with children, I would be interested in your view. When does giving a child a choice help, and when does it just become one more thing to decide?
We'd love to hear your thoughts on this. Find us on the social channels linked at the top of the page.