Know someone trying to make sense of this? Send them the guide.
Share on WhatsAppWhy we use peer comparisons instead of rankings
Single-number scores and letter grades primarily measure community wealth, not school quality. A school's overall score is often predictable from household income alone, which means rating systems can reinforce existing inequalities rather than reveal anything about teaching or learning. SchoolSeek uses peer comparisons instead: how does this school compare to similar schools in similar circumstances?
What is a peer group?
How a peer group is defined
Example: A Quintile 2 primary school in the Western Cape is only compared to other Quintile 2 primary schools in the Western Cape — not Quintile 2 schools in Gauteng, and not Quintile 3 schools in any province.
Independent schools: An independent school in the Western Cape is compared to other independent schools in the Western Cape. Some carry a quintile in the government's records even though the classification does not apply to them, so we group on sector rather than on that leftover value — comparing a fee-funded school to a quintile-funded one would not be a like-for-like comparison.
A peer group consists of schools that share the same quintile, phase (primary, secondary, or combined), and province. A Quintile 2 primary school in the Western Cape is only compared to other Quintile 2 primary schools in the Western Cape. This way, schools are measured against others facing similar resource constraints and serving similar communities. Independent schools form their own peer group rather than sitting inside a quintile band, because the quintile classification only applies to public schools and describes government funding, not fees. For pass rate comparisons, phase is not used as a dimension (since only secondary and combined schools write matric) — pass rates are compared within the same sector and province, and within the same quintile for public schools.
How percentiles work
Within each peer group, we calculate where a school falls for measurable indicators like the learner-to-educator ratio. If a school has a "lower L:E ratio than 80% of similar schools", it means 80% of schools in the same peer group have a higher (less favourable) ratio. The L:E ratio includes all staff classified as educators, including principals, HODs, and deputy principals. It is not the same as actual class size. The percentile is displayed as a visual bar showing the school's position among its peers.
What we measure — and what we cannot
With current EMIS data, we can compare learner-to-educator ratios within peer groups. This is a resource indicator: it reflects staffing levels, not teaching quality or actual class sizes. The ratio counts all staff classified as educators, including those in management who may not teach classes directly. We also display NSC matric pass rates for secondary and combined schools, compared within quintile-based peer groups.
What happens when a peer group is small?
Some peer groups have fewer than 5 schools. When this happens, the percentile comparison is less statistically meaningful, and we flag it with a note. A school being "in the top 20%" of a 3-school group tells you less than the same position in a 50-school group.
Sources
- Department of Basic Education — EMIS Q3 2025 national extract
- DBE School Performance Report 2025 (NSC results) — education.gov.za
- South African Schools Act 84 of 1996 (quintile framework)
More Guides
Data sourced from the Department of Basic Education EMIS database. Read our full methodology