Studying out loud vs in your head: which works?
In short
- Items read aloud are remembered better than items read silently — the production effect.
- The advantage is relative. Reading everything aloud removes most of it.
- Speaking from memory is a different and larger effect than reading aloud from the page.
- Practical rule: read silently, recall aloud.
Somebody in the library is muttering their way through a pathology chapter. Somebody else is reading the same chapter in silence, faster, and finding the muttering annoying.
The mutterer is doing something defensible. They are also probably doing it wrong.
The finding
Read a list of words, some aloud and some silently. Later, the words you spoke are recognised and recalled better than the words you read in your head. This is the production effect, and it has been replicated a great deal since MacLeod and colleagues characterised it in 2010.
The usual explanation is distinctiveness. Within a list of otherwise identical experiences, the spoken items carry extra properties — you heard your own voice, you moved your mouth, the item has more attached to it. At retrieval, those extra properties are extra routes back.
It works for words, non-words, and sentences. It is not large per item, and it costs almost nothing. That combination is unusual enough to take seriously.
The catch that gets left out
The distinctiveness account makes a prediction, and the prediction holds: the advantage is mostly relative.
When a study has some items spoken and some silent, the spoken ones win. When everything is spoken, the within-list distinctiveness has nowhere to come from, and the benefit shrinks substantially. The spoken items were not helped by being spoken in the absolute — they were helped by being different from their neighbours.
The practical consequence is blunt, and it is the opposite of what the person muttering through the whole chapter is doing:
Reading the entire chapter aloud is close to reading the entire chapter silently. It just takes longer.
If you are going to use production, use it selectively — on the definitions, the mechanisms, the numbers you keep losing. The items you speak have to stand out from the ones you do not.
The bigger effect is not reading aloud at all
There is a much larger effect standing next to this one, and it is easy to confuse them.
Reading aloud from the page: you can see the answer. The only work is articulation.
Saying it from memory with the page closed: you have to retrieve it first. Retrieval practice is one of the most robust effects in the memory literature, and it is far larger than production — in Roediger and Karpicke’s comparison, repeated testing beat repeated studying by a wide margin at a one-week delay, and the ordering reversed entirely at short delays, which is exactly why re-reading feels better than it is.
So the ranking, for the same twenty minutes:
- Recall it aloud, page closed — retrieval plus production.
- Recall it silently, page closed — retrieval.
- Read it aloud from the page — production only.
- Read it silently — neither.
Most people who “study out loud” are doing (3) and comparing it to (4). The available upgrade was (1), and it was free.
Where speaking earns its keep
Three situations where saying it aloud does more than the laboratory effect suggests:
You are studying in a second language. Retrieval, sentence assembly and articulation are separate loads. Practising them silently trains two of the three, and the exam asks for all three. We wrote about this in Studying medicine in English as a second language.
You will be assessed by speaking. Whatever your programme calls it, the format rewards structure and fluency, and both are motor skills that do not transfer from writing.
You are not at a desk. This is the underrated one. Recall-aloud is the only serious form of revision you can do while walking, cooking or on a bus. A silent method needs your eyes; a spoken one does not.
The rule
Read silently. Recall aloud.
That is the whole finding, compressed. The reading is input and there is no advantage to performing it. The recall is where the learning happens, and speaking it adds a second, cheap, independent benefit on top of the retrieval you were already getting.
Live Talk is built on exactly this line: the page is closed, you say the answer, and something asks you why.
Sources
- MacLeod, C. M., Gopie, N., Hourihan, K. L., Neary, K. R., & Ozubko, J. D. (2010). The production effect: Delineation of a phenomenon. Journal of Experimental Psychology: Learning, Memory, and Cognition, 36(3), 671–685.
- Roediger, H. L., & Karpicke, J. D. (2006). Test-enhanced learning: Taking memory tests improves long-term retention. Psychological Science, 17(3), 249–255.