Updated on
September 30, 2026
Dialogic Teaching: Classroom Examples, Mini App and PDF
Use dialogic teaching to connect ideas and examine reasons. Try the classroom talk classifier and download a practical prompt card for your next lesson.

What is dialogic teaching?
Dialogic teaching is an approach that uses purposeful classroom talk to develop understanding. Learners explain ideas, consider evidence and respond to each other's reasoning. The teacher chooses questions, teaches the knowledge needed for discussion and checks what each learner understands afterwards.
A discussion becomes dialogic when contributions affect what happens next. A teacher or learner takes up an idea, asks for its basis, connects it to another contribution or invites revision. A sequence of separate answers may be useful for checking knowledge. It offers less opportunity for shared reasoning.

Surface participation
The teacher asks several questions, accepts the expected answers and moves on.
Reasoning to explore
“Are you saying the character had no choice?” The learner corrects this and points to the text.
Follow-up check. Ask what the exchange changed in the learner’s interpretation, then check a short response.
Put both exchanges in front of the class. Ask what the teacher’s question changed in the learner’s reading of the text.
For example, a teacher asks why a puddle disappears. One learner says the water soaks into the ground. Another points out that a puddle on a hard surface also disappears. The teacher asks what that second example changes. Learners now have to test an explanation against a different case. The purpose is the scientific distinction, not the number of turns taken.
Dialogic teaching does not remove explanation, modelling or factual checking. Learners need knowledge to reason with. If the class does not understand a key term, teach it. If an answer conflicts with established evidence, address it. Keep discussion open where there are genuine alternatives, while keeping the subject content accurate.
The Exploratory Talk guide focuses on learners testing ideas together. Accountable Talk offers a related emphasis on reasons, shared knowledge and responsibility to the discussion.
Alexander's earlier framework describes talk as collective, reciprocal, supportive, cumulative and purposeful. These five principles are documented in his account of the Dialogic Teaching programme (Alexander, 2018). His 2020 book revised the framework. The list below explains the earlier version rather than presenting it as the complete current framework.
Use the principles to make professional decisions. They are not a requirement to use the same format in every lesson. A short pair rehearsal, a whole-class explanation and a longer evidence discussion serve different purposes. Alexander's current account emphasises teacher agency rather than an all-purpose formula.
Thinking Together research examines explicit teaching of talk ground rules. Learners practise offering reasons, considering alternatives and reaching a decision through discussion. Simply assigning a group does not teach these habits. A task can generate social interaction without generating accurate reasoning.
Start with one useful move. Model “I think this because…” or “What evidence would help us decide?” using content the class knows. Show a weak response and improve it. The class can notice the difference between repeating a view and giving a reason.
Sentence stems are temporary support. A learner who already reasons clearly may not need one. A learner who lacks the subject vocabulary may need a word bank or model before the stem becomes useful. Listen to the meaning of the response. Completing the sentence is not the learning goal.
Revoicing can make an idea available to the group. It needs a check. If a learner says, “The character was scared because of the dark,” the teacher might ask, “Do you mean the darkness itself frightened her, or something she thought was there?” Wait for the learner to confirm or correct that interpretation.
Once the meaning is clear, invite the group to examine the text. “Which line supports that?” keeps the discussion tied to evidence. “Does anyone have a different reading of that line?” makes room for alternatives. Avoid improving an answer so much that the learner's original thinking disappears.
In science, one learner identifies a fox eating a rabbit. Another identifies the rabbit eating grass. Ask how those ideas connect. Then ask what could happen if the number of rabbits changed. A shared diagram helps the class follow the relationships.
In English, ask a learner to name the idea they are responding to before giving their own view. This helps listeners trace the argument. It is one of the habits of discussion worth rehearsing on easy content before a demanding task.
The mini app below offers eight short exchanges. Use it to discuss Mercer's disputational, cumulative and exploratory talk categories with colleagues. Ask which features justify a choice. The scenarios are illustrations, not recordings of learners or a validated assessment of a teacher's classroom.
123Try one numbered response, such as asking for evidence, and note whose explanations change when you use it.
A category is useful when it helps you decide a next move. If learners agree without reasons, ask for the evidence. If they dispute a point, clarify the shared question. If they compare explanations, help them check the subject knowledge and reach a supported conclusion.
A Turn and Talk routine can make the preparation short and predictable. Everybody Writes gives a written route into discussion. Adapt either routine when its recording or speaking demands create a barrier.
Teacher voices · Say It
Would you use this approach every time for every question?
@yourbiztutor
Use a short exchange when it can reveal a useful reason. Make space for a longer discussion when evidence or interpretations need weighing. Direct explanation still has a place.
In a lesson on plant growth, ask learners to explain why a plant kept without light changes. Provide observations from the lesson's experiment. Let them prepare an explanation before the group discusses it. A useful prompt is, “Which observation supports that claim?”
If a learner says soil is the plant's food, a partner's agreement does not make it correct. Teach the distinction between nutrients, water and the sugars made through photosynthesis at the appropriate level. Use a diagram to support the explanation. Then ask learners to explain a different case independently.
Discussion reveals a misconception. It does not replace the teaching needed to correct it. Keep the experiment's limits visible. A test of one plant under one condition cannot settle every claim about growth.
Present two contrasting accounts of an event. Ask who produced each source, when, for whom and for what purpose. Learners identify the claim and evidence before deciding what each source can help them establish.
An official report is not automatically more reliable than a personal letter. It may omit inconvenient details or serve a particular audience. A letter may offer a first-hand perspective but have limited knowledge. Ask learners to corroborate the accounts with other evidence and distinguish what they can conclude from what remains uncertain.
A learner might say, “This report gives numbers, but we need to know how they were collected.” Another might answer, “The letter describes what one person saw, so it helps with that person's experience.” Those contributions offer reasons for using the sources differently. They move beyond selecting the source that sounds most authoritative.
Ask learners to compare two interpretations of a poem's ending. Each finds a line that supports one reading and a line that might complicate it. Begin with preparation before asking for public contributions.
During discussion, ask, “How does that respond to the previous idea?” and “What in the language makes you think that?” A learner may revise an interpretation or retain it with a stronger reason. Different answers can remain plausible when the text supports them.
Close with a brief individual paragraph. Look for the connection between the claim and the selected line. That gives a clearer check of understanding than counting who spoke most.
Participation means having a route into the shared reasoning. It does not mean every learner must speak publicly on cue. Plan for AAC, written responses, a partner sharing an agreed idea or another established communication route. Follow individual support plans and consult the relevant staff when adapting demands.
Arrange the space so learners can hear or access contributions and see shared material. Do not require eye contact as evidence of listening. A learner's accent, dialect or fluent presentation is not a measure of the quality of their reasoning.
Give a question before requesting an answer. Thinking time and pair rehearsal can help some learners, but do not remove every source of anxiety. Avoid making a reluctant speaker the focus of a public correction. The NHS guidance on selective mutism advises reducing speaking pressure. Classroom routines should support an agreed plan, not act as treatment.
Keep key vocabulary, evidence and the developing argument visible. A short note or diagram allows learners to revisit an idea instead of remembering every spoken contribution. Offer an accessible alternative where writing or vision presents a barrier.
The Voice 21 framework describes physical, linguistic, cognitive, and social and emotional aspects of oracy. Use it as a broad planning language. A teacher-created reflection aid informed by that framework is original practice guidance. It is not a validated assessment scale and should not diagnose a learner or grade a teacher from one visit.
Contribution frequency shows who is heard. It does not establish who understands most. A learner can speak confidently while repeating an error. A quiet learner may have a sound explanation but need a different communication route.
During discussion, note the accuracy of a claim, the reason offered and the response it receives. Ask a follow-up that distinguishes a remembered phrase from understanding. Then use an individual check, such as a fresh problem, a short explanation or a labelled diagram.
During a learning walk, record the task and a brief exchange. Ask whose ideas were taken up and whether someone changed an answer for a reason. Do not infer a whole class's understanding from the most fluent speaker. Compare repeated samples and use them for coaching.
For a small change in questioning, see Wait Time. For choosing a prompt, see questioning by purpose. The critical thinking and classroom talk guide offers a related reasoning focus.
Start with one curriculum question rather than rolling out a new format across every lesson. Agree what understanding the task should reveal. Teach it, sample a short exchange and review the learners' individual responses with a colleague.
Classroom practice · Say It
| Make the connections visible | ||
|---|---|---|
| Starting point | Try in your lesson | Check the boundary |
| The school describes a Year 12 / Grade 11 geography discussion about Amazon deforestation. A simple map can show which contribution responds to an earlier idea. | A map shows connections and participation. It does not show that every explanation is accurate or that every learner understands. Follow it with a subject check. |
Separate participation from learning in that review. Did more learners have access to the task? Did their explanations become more accurate? Did they respond to one another's reasons? A noisy room, a quiet room or a polished performance cannot answer those questions alone.
Choose one next move from the evidence. It might be a clearer model, a stronger prerequisite check or a different contribution route. Repeat the task's reasoning demand with new content. Look for improvement over time instead of grading a teacher from a single observed discussion.
The EEF intervention included training, mentoring, materials and video review. Those conditions matter when interpreting its results. A poster or a sentence stem alone does not reproduce the tested programme. If using recordings for staff development, follow the school's consent, privacy and safeguarding arrangements.
Before the next lesson, write the one question you want the class to reason about, and the response you will model if the first answers stay short.
Dialogic describes an exchange in which people respond to and develop ideas together. In teaching, that means purposeful questions, relevant knowledge, connected responses and checks of learning. It does not mean the teacher must withdraw or leave errors uncorrected.
In history, learners compare two accounts of an event and explain what each can support. A response challenges one interpretation using evidence. The teacher clarifies context, connects the contributions and checks each learner’s revised explanation.
A discussion may collect separate opinions. Dialogic teaching deliberately connects contributions to build and test understanding. The teacher plans the learning goal, prompts reasons and checks what learners know afterwards.
The supported EEF Year 5 programme reported positive short-term results in English and science. The maths estimate was uncertain. The KS2 follow-up did not find clear effects. These findings support careful use and evaluation, not a guaranteed gain from any talk activity.
Every learner needs access to the reasoning task. Public speech is one route. Plan alternatives around communication needs and support plans. Do not mistake quietness, accent or eye contact for understanding.
Choose familiar content and a question that calls for a reason. Model one response move. Give preparation time, listen to an exchange and finish with an individual check. Add complexity after the routine becomes useful.
Teaching learners to reason together has a research base. That does not mean any activity involving talk produces the same result. The task, subject knowledge, ground rules and teacher response all matter.
| Evidence | Useful finding | Limit for your classroom |
|---|---|---|
| Mercer, Wegerif and Dawes, 1999 | Teaching ground rules changed aspects of talk and reasoning in a small matched-control programme (Mercer et al., 1999). | A small matched-control intervention. It does not prove every sentence stem or discussion format works. |
| Mercer and Sams, 2006 | Explicit language teaching supported reasoning in small-group maths work (Mercer and Sams, 2006). | A programme of teaching, rather than simply placing learners in pairs. |
| Howe and colleagues, 2019 | In 72 primary classrooms, some attainment outcomes were associated with building on ideas when participation was widespread (Howe et al., 2019). | Observation cannot establish causation. Results differed by outcome. Widespread participation does not mean every learner must speak publicly. |
| EEF Dialogic Teaching trial and addendum | The supported Year 5 programme reported about two additional months in English and science and one in maths. | The maths estimate was uncertain. The later KS2 follow-up did not find clear effects. Training, mentoring and video review were part of the programme. |
The EEF's oral-language interventions headline is mainly based on reading comprehension outcomes. Its collaborative learning entry covers many approaches. Neither figure is a guaranteed gain from a particular classroom talk routine.
The PDF card lists five response moves, from asking for a reason to checking understanding. Adapt the moves to your class; the card supports planning and is not a validated intervention.
Choose one response move and use it every time a learner answers in the next lesson. Afterwards, check whether individual explanations have got stronger.