In the latest edition of the British Psychological Society’s Research Digest, Bradley Busch writes about a new study which compared the effects on memory of reading in silence to those of reading out loud. Noah Forrin and Colin MacLeod’s paper, This time it’s personal: the memory benefit of hearing oneself, explores what’s been termed the ‘production effect’ – a neat name for the memory advantage of saying words aloud over simply reading them silently. The speculation is that the effort of saying something out loud appears to make information more cognitively ‘sticky’, creating stronger schematic connections in long-term memory. This advantage appears to be strongest when we say things out loud ourselves, but also holds true when listening to someone else read aloud. 
This has, I think, some important implications for two areas of classroom practice. Earlier this year I wrote about why there’s good reason to believe that asking students to listen to a text being read aloud whilst simultaneously expecting them to ‘follow along’ in silence is likely to over burden working memory and lead to poorer comprehension.  I suggested that this was a form of multitasking which human beings are simply incapable of. Instead, we ‘task switch’, diverting attention from reading to listening, with a lag of a few seconds with every switch. My conclusion was as follows:

To be absolutely clear: reading aloud to students, especially weaker readers, is a good thing to do. Prosody – the sound and rhythm the words make – really aids comprehension, especially with difficult texts. The problem comes when children are expected to follow along at the same speed. Because they lack the fluency to do this their working memory overloads and derails comprehension.

Reading aloud to students – or even better, having them read aloud – is likely to also boost their memory. There are obvious classroom disadvantages to having every child read aloud, but asking the most skilled reader (probably the teacher) in the room to read aloud is likely to benefit everyone.
The second implication is around improving students’ writing. It’s been some years now since I first proposed my ‘simple theory of writing‘, but essentially what I suggest to teachers is that intervening at the point of writing is an ineffective way to improve writing. Intervening at the point of speech seems to be far more effective. Many students are unable to write down their thoughts because they simply don’t possess the knowledge of how to express their ideas in written language. If we scaffold students’ spoken language by making them ‘speak like an essay’ then my observation is that the act of speaking in academic language seems to strengthen the schema of how to communicate in academic language and thus writing becomes possible. My anecdotal finding has been that this is far more effective than using writing frames, so it’s pleasing to find some empirical support for the idea.
When I’m training teachers in effective instruction, one of the points I make is that if you want students to remember something, you have to make them say it aloud. If you just get one or two students to repeat something, then only one or two students will remember it. If one students finds a way to avoid having to speak, that’s the student who will be least likely to remember it. The answer is to use fast paced call and response whenever you want students to pay special attention to memorising a particular idea or phrase.
Of course, none of this is definitive proof that saying things out loud supports memory – the current research is based on participants studying word lists, so it will be important to see if the findings can be replicated with longer texts – but there’s a clear and plausible mechanism to explain why we should expect the production effect to be an effective way to teach and study.