The four strands of oracy: what the evidence really shows

The gist. Cambridge researchers built a four-part way of describing good talk, physical, linguistic, cognitive, and social and emotional, by narrowing eight starting categories down through expert discussion rather than by testing which split was scientifically correct.

What the study found

Funded by the Education Endowment Foundation, a University of Cambridge team worked with School 21 in London over roughly two years to build an oracy assessment toolkit for Year 7 learners. They reviewed existing frameworks, then consulted an eight-person expert panel covering drama, English, sociolinguistics, applied linguistics and assessment, meeting three times across the project. Teachers, speech therapists and bodies including Cambridge Assessment and the National Literacy Trust were consulted too.

The panel's first version of the framework had eight categories. Discussion and feedback narrowed it to four: physical, linguistic, cognitive, and social and emotional, each holding several named skills such as voice projection, vocabulary choice, structuring talk and turn-taking. The one classroom comparison reported in the paper, between two schools, was explicitly too small for the researchers to call any difference statistically significant.

What it might mean for your classroom

The four strands came from expert consensus and a review of existing frameworks, not from a test of which categories are correct. That still makes them a useful shared vocabulary for planning and noticing talk in any subject, including exploratory talk in discussion-heavy lessons.

• Move. Pick one strand, physical, linguistic, cognitive, or social and emotional, and give feedback on a single discussion task using only that lens.
• Move. Match the strand to the learner: build social and emotional skills such as turn-taking before pushing a quiet learner for polished content, or ask a talkative learner for cognitive structure and reasons instead of more words.
• Move. Differentiate your feedback by subject: lean on linguistic register and rhetorical technique in English, or physical voice and pace in drama, rather than treating every subject's talk the same way.

Treat it as an experiment, not a recipe: notice which strand a learner is strongest in and which one holds them back, then treat that as your own evidence, not the researchers'.

How solid is this?

Promising. This rates how solid the evidence is, not how useful the framework is. A funded, two-year research programme built the four strands by structured consensus, not by testing what oracy is scientifically made of. The researchers say so themselves: the framework is written to be usable by teachers, not to satisfy the rigorous criteria of an academic linguist. The one classroom trial in the paper, 12 learners in each of two schools, was too small for the researchers to call any difference statistically significant, so it cannot be used to prove the framework works.

The source

Mercer, N., Warwick, P. & Ahmed, A. (2017). 'An Oracy Assessment Toolkit: linking research and development in the assessment of students' spoken language skills at age 11-12.' Learning and Instruction, 48, 51-60.

Read the original study