Reciprocal Teaching

Reciprocal Teaching was developed by Annemarie Sullivan Palincsar and Ann Brown at the University of Illinois and published in their landmark 1984 study, which remains one of the most empirically validated reading-comprehension methodologies in education. Their study took students with weak comprehension scores, ran them through 20 sessions of the four-move routine (predict, question, clarify, summarise), and measured transfer to independent reading on new texts. The results were striking: roughly two grade levels of gain on transfer measures, with effects holding 8 weeks post-intervention. Rosenshine and Meister's 1994 meta-review confirmed median effect sizes of 0.32 on standardised tests and 0.88 on experimenter-developed comprehension measures across 16 replications.

What makes the methodology work is that the four moves correspond to what skilled readers already do silently. Skilled readers predict what comes next based on the title, the genre conventions, and the paragraph just read. They author internal questions while reading ('wait, why did the character do that?'). They notice when a word or phrase confuses them and seek clarification. They build running summaries that let them pick up where they left off after an interruption. Reciprocal Teaching makes these invisible processes visible, names them, and rotates them among students so every student practises every move.

The role rotation is the structural mechanism. In a group of four, each student takes one of the four moves for the first paragraph: the predictor names what the next paragraph should be about, the questioner authors a question the text answers, the clarifier resolves a confusing word or phrase, and the summariser states the gist. Then roles shift one position. By the end of a four-paragraph text, every student has practised every move. Without rotation, students settle into roles by strength: the strong reader summarises every time, the emerging reader predicts every time, and the routine produces specialisation rather than comprehension growth.

The teacher's role is to gradually fade. On day one, the teacher models all four moves on the chosen text, making the reasoning audible ('I'm noticing that the word 'precarious' in this sentence sounds like it might mean uncertain, because it's set against 'stable' in the next clause'). On day two, the teacher shares the moves with students. By session 5 or 6, the teacher sits outside the group and listens, intervening only when the routine itself breaks down (a student silenced by a peer, a procedural confusion), not when comprehension wavers. The group will repair its own comprehension; that is the pedagogy.

Implementation requires patience. Sessions 1-3 always look awkward. Students are learning a new vocabulary of comprehension and the routine feels stilted. Comprehension gains start showing up around session 8, and transfer to independent reading shows up around session 12. Teachers who quit at session 3 because 'this isn't working' miss the actual payoff. The methodology demands a 12-16 session commitment, which is why it works best as a sustained reading routine rather than a one-off lesson plan.

Text selection matters more than most teachers expect. The text must be 200-400 words for Classes 3-5, up to 600 for Classes 6-8 (the routine breaks down beyond a single class period). It must be dense enough to genuinely need clarification: a text with no confusing word and no hidden inference starves the clarifier role and the routine goes through the motions. Short informational passages, narrative excerpts with subtext, primary-source documents, and dense textbook sections all work. NCERT chapters often include reciprocal-teaching-friendly passages explicitly, and ICSE prescribed texts hold up well too.

The methodology has a domain restriction worth respecting: it works for reading-dense content (English, dense informational text in science and social studies, primary sources in history) and does not work for math procedures, art critique, or PE. The four moves are reading-comprehension strategies; they have no leverage when there is no text to predict against. This is not a limitation of the methodology; it is a clean scope, and Reciprocal Teaching pays back the investment richly within that scope.

Heterogeneous grouping is the final design choice. Mixed-level groups of four produce stronger gains than homogeneous groups, because strong readers benefit from articulating their process (which deepens their own understanding) and emerging readers benefit from the structured turn-taking (which keeps them engaged when they would otherwise hide). This is the cooperative-learning insight applied to comprehension instruction: structure the routine so that every member is necessary, and the routine itself does much of the differentiation work.