On this page
Many students read a physics or biology chapter twice, highlight half of it, and still cannot answer a question about it the next day. Reading the words is not the same as understanding them, and rereading rarely closes the gap. Ann Brown spent her career on that problem. With Annemarie Palincsar she built a method called reciprocal teaching, which turns reading into four small, checkable moves. This post covers who she was, what the studies found, and how to use the four moves on a textbook chapter.
Who was Ann Brown?
A UC Berkeley news release from June 1999 records that she was born in Portsmouth, England, during the Second World War, graduated from the University of London in 1964 and received her PhD in psychology there in 1967. She was a professor at Berkeley's Graduate School of Education from 1988, holding the Evelyn Lois Corey Chair, and served as president of the National Academy of Education. She died on 4 June 1999, at 56. The same release notes that she had just co-edited How People Learn: Brain, Mind, Experience and School, a major synthesis of learning research from the National Research Council.
Her later work, called Fostering Communities of Learners, had students design research projects, divide up tasks and act as experts for each other. The thread running through all of it was that good learners actively check their own understanding, and that this can be taught.
What is reciprocal teaching?
In Palincsar and Brown (1984), the method was built around four activities that strong readers use without being told:
Summarising: say in a sentence or two what the passage was about.
Questioning: ask a question about the main idea, the kind a teacher might ask.
Clarifying: spot what is confusing, such as a term, a step or a reference, and fix it.
Predicting: say what the next part is likely to cover.
The "reciprocal" part is how it is taught. The teacher and students take turns leading a dialogue about each section of text. The adult acts as a model, guiding students to work with the text in more sophisticated ways, and students practise leading the discussion themselves.
What did the research find?
The 1984 paper reported two studies with seventh-grade students who were poor comprehenders. In the first, reciprocal teaching was compared with an intervention modelled on typical classroom practice. The reciprocal group made greater gains and kept them over time. The quality of their summaries and questions improved, and the gains showed up on classroom comprehension tests, on new tasks using the same skills, and on standardised comprehension scores. The second study found many of the same results when ordinary volunteer teachers ran the method with their own reading groups.
A decade later, Rosenshine and Meister (1994) reviewed sixteen studies. With standardised comprehension tests the median effect size, favouring reciprocal teaching, was 0.32. With comprehension tests written by the experimenters it was 0.88. The gap is worth noticing: tests matched to what was taught show bigger effects than broad tests, which is true of most interventions.
How do you use it on a science chapter?
The original studies used everyday reading passages with school students, so treat this as an adaptation. It suits dense textbook prose, such as an NCERT section on the photoelectric effect or on the cell cycle.
Chunk the text. Work one subsection at a time, roughly half a page.
Summarise in two sentences, book closed. For a section on the photoelectric effect: "Light below a threshold frequency ejects no electrons, however bright it is. Above it, the maximum kinetic energy rises with frequency, not intensity."
Ask one exam-style question. "Why does doubling the intensity not change the stopping potential?" Our guide to writing your own practice questions has templates.
Clarify one thing. Mark the sentence you could not explain to a friend, then look it up or work it through. Often it is a term like "work function" that the paragraph assumes you know.
Predict the next part. "Next it will probably give Einstein's equation and a graph of stopping potential against frequency." Checking your prediction makes you read the next part with a purpose.
With a partner or study group, rotate the leader role for each chunk, exactly as in the original method. The leader runs the four moves; the others answer, challenge and add. Two people with one chapter is enough.
Alone, write the four lines in the margin of your notes. It is slower than highlighting, and that is the point. Our post on reading a textbook actively gives more ways to slow down usefully.
What can a student take from how she worked?
Study what good learners already do. Brown did not invent the four moves; she noticed them in strong readers and made them teachable.
Make thinking visible. In reciprocal teaching every move is said out loud, so mistakes can be caught. Saying your reasoning aloud works the same way for a numerical.
Learn from a model, then lead. In the studies an adult modelled the moves and students took turns leading. When you learn a new problem type, follow a worked example first and then try to lead yourself.
The clarifying step is where most students get stuck, because they cannot tell what they do not know. Fermi's AI tutor notices where a student is stuck and decides whether to explain, ask or show, which suits that step once you have tried it yourself.
Take the next section you have to read this week and write the four lines, summary, question, clarification and prediction, before you move on.
