The residual (Swedish: avvikelse) is the difference between a school's actual merit value and the model-predicted value:
Residual = actual merit value − model-predicted merit value
Interpretation
- Positive residual — the school performs better than the model predicts. Students achieve higher results than what the background variables would suggest.
- Negative residual — the school performs worse than expected. Results fall below what is typical for schools with a similar student composition.
- Zero — the school performs exactly as expected.
Why is the residual useful?
The residual makes it possible to compare schools with entirely different student populations. A school in a socially disadvantaged area with a positive residual is doing a better job (relative to its conditions) than a school in an affluent area with a negative residual — even though the former school has a lower absolute merit value.
Limitations
The residual should not be interpreted in isolation as a definitive measure of school quality. The SALSA model captures approximately 53% of the variation — factors such as the school's resources, leadership, socioeconomic nuances and local conditions are not included. A single year's residual can also be affected by random variation in a small student cohort.