I was glad to see the issue of national testing getting some good exposure (TES Scotland Plus, May 19).

Current education practice is all at sea. It is based on the political compromise of the early 1990s, when teachers, parents and directors supported HMI to defend 5-14, resisting simplistic English models of testing.

To avoid league tables, testing happens “when the child is ready”. To reduce ranking, testing is within broad levels. But the levels are too broad: children spend several years working within a level. Also, since the test is used to “confirm teachers’ judgments” after a level hs been achieved, it adds nothing to what is known.

If we want to use tests to improve learning, our pupils need regular, accurate, normative information in a few significant performance areas. They, supported by their teachers and parents, can then measure their own progress and set personal targets. A commitment to the equal value of each child, supported by rules of confidentiality (as with medical records), would allay fears about information being wrongly used.

Publication of aggregated information at school level should involve not meaningless league tables but multi-factor statistical analysis which can be used to identify school trends within a quality assurance framework. The market, or a national system, can do this.