Does Reading for Pleasure Improve Exam Results?

What the study actually measured

The research most often quoted here comes from the Institute of Education at UCL. Alice Sullivan and Matt Brown used the 1970 British Cohort Study, which has followed the same people since birth, and analysed around 6,000 of them: how often they read at 10 and at 16, set against test results in maths, vocabulary and spelling at 5, 10 and 16.

Children who read books often at 10 and more than once a week at 16 did better — and, importantly, made more progress between the two ages rather than simply starting higher. Vocabulary showed the largest effect; spelling and maths were smaller but present. The combined effect of reading books often, using a library and reading newspapers at 16 was reported as around four times the advantage of having a parent with a degree.

The maths result is the interesting one

Nobody is surprised that reading helps vocabulary. Maths is the finding that makes the study worth citing, and there is a mechanism that makes it ordinary rather than mysterious: almost every mark on a maths paper sits on the far side of a sentence. Word problems, multi-step questions and the command words that decide what is being asked are all reading tasks before they are arithmetic ones.

There is a second and quieter mechanism. Sustained reading is practice at holding a long chain of information in mind while working out what it adds up to, which is also what a five-mark structured question demands. That is a plausible explanation rather than a demonstrated one, and it is worth marking as such.

What the study does not prove

This is observational research, and the fair description is that it is very good observational research rather than an experiment. Sullivan and Brown compared children from similar social backgrounds who had reached similar test scores at the earlier age, which removes the two most obvious rival explanations. What it cannot remove is a factor nobody measured — a household habit, a temperament, an appetite for sitting still — that produces both the reading and the progress.

So "reading improves grades" is a reasonable reading of the evidence and not a proven fact. There is a specific trap in treating it as one. The study observed children who chose to read. It did not test what happens when a fifteen-year-old is required to read for half an hour every night, and it offers no basis for expecting the same result.

What actually follows for a family

If the effect is real, the mechanism is volume — a great deal of text, read fluently, over years. Anything that increases volume is working with the finding and anything that decreases it is working against it, including well-meant rules about which books are worthy.

The timescale deserves saying plainly too. The study measured six years of childhood. Starting a reading habit in Year 11 is a good thing to do for its own reasons, but it is not what was measured, and nobody should promise a grade for it.

• Let the child choose, including the series, the comic and the book below their reading age — fluent reading is where the gains sit
• Keep a book running alongside school reading, so that reading has a form which is not homework
• Widen the range once the habit is established: journalism and non-fiction carry vocabulary fiction does not
• Audiobooks build vocabulary and comprehension, but not decoding or spelling — a fair substitute for part of it, not all
• Reading aloud to an older child still counts; a ten-year-old's listening vocabulary runs well ahead of what they can read alone