How to Lift Attainment Across a Large Online Cohort

The bigger the cohort, the easier it is for averages to hide the pupils going backwards. Large-scale provision tends to report activity — lessons delivered — rather than learning, because learning is harder to measure at volume.

Martin High measures the learning. Over two years it has run close to a thousand lessons across all four Purple Ruler programmes — in-school alternative provision, the online academy, one-to-one and stretch support — for more than seventy pupils, with around thirty lessons live at any time, and tracked the entry-to-exit quiz throughout.

Breadth is Martin High's defining feature: four programmes running at once, for a cohort large enough that most schools would settle for delivering it, never mind improving on it. The test of such a partnership is whether the numbers move.

They do — and not at the margins.

Across the lessons sampled, the cohort's entry-to-exit quiz average rose from 47% to 88% — a forty-one-point gain, the largest in this set. Behind the number is ordinary good teaching: one pupil “did an outstanding job… arrived in good spirits, engaged fully in the lesson, and demonstrated excellent participation throughout”.

At seventy-plus pupils, a gain that size is not a fluke of small numbers; it is a system working.

The lesson for trusts scaling provision: size is no excuse for measuring activity instead of learning. Track the quiz across the whole cohort, and a large partnership can post gains a small one would envy.