MAIN ARTICLE
Three primary-school numbers tell different stories
The mathematics basic-competency rate for Primary Three was 84.7% this year, a decline of 0.7 percentage points from 2025. It had reached 89.9% in 2016 and has trended downward over the longer period. That 5.2-point difference warrants attention precisely because it cannot easily be dismissed as one unusual test day. Yet the measure describes the proportion crossing a defined threshold; it is neither the average mathematics score nor the proportion of children promoted to the next grade.
English moved in the opposite direction. Its basic-competency share increased by 3.1 percentage points to 86.3%, the highest recorded for the test. Chinese reached 81.7%, leaving English ahead of Chinese for a second consecutive year among Primary Three pupils. Such divergent trajectories do not, on their own, prove that any particular textbook, teaching method or school policy succeeded or failed. Rather, they point to the need to examine how pupils acquire different skills.
The Territory-wide System Assessment, or TSA, is designed to give schools information about basic competencies and help teachers adjust instruction. It uses a sample of roughly one in ten pupils at the assessed levels across the territory; in 2026 the exercise covered Primary Three and Secondary Three, not Primary Six. These data can illuminate broad trends. They should not be turned into labels of aptitude attached to individual children or used as a simplistic league table of schools.















