The current alarm over the credibility of national testing has its roots in the revelation - nearly two years ago - that a school could perform well in the national curriculum tests but still receive a poor report from the Office for Standards in Education.Never one to mince his words, Chris Woodhead, the chief inspector, opined that: "Either OFSTED is rubbish or performance tables are rubbish." Not surprisingly, he chose in his well-publicised remarks before Christmas to emphasise the defects of the tests administered by the Qualifications and Curriculum Authority, rather than those of OFSTED.
Under pressure from ministers, anxious not to have their target-related policies discredited by a key official, the chief inspector swiftly backpedalled. But he had sparked a controversy which shows no sign of dying down, partly because his doubts are shared by a number of teachers.
Today in The TES, Nick Tate, chief executive of the QCA, defends the tests (letters, page 20); and Chris Whetton, of the National Foundation for Educational Research, explains the lengths to which the NFER goes to make them as reliable and valid as possible (Platform, page 17).
This rich brew has been intensified by this week's publication of three reports on OFSTED itself. There is much in the findings to please Mr Woodhead - but they tend to focus on attitudes to inspection rather than the central question: how valid and reliable are his own inspectors' verdicts?
Establishing such reliability is a complex task - especially since both our systems for evaluating schools have to perform too many functions. The SATs not only generate performance tables (which themselves have several purposes), but are supposed to inform parents and teachers about children's progress and chart national trends.
The main aim of inspection reports (which in any case draw heavily on test results) is to inform parents, but other objectives include identifying failing schools and inadequate teachers, school improvement, adding to a database of good practice, and providing a general picture of what's happening in our schools.
Much hard thinking is still required as to how these two forms of evaluation can be improved and further validated. But this does not mean that either of them is "rubbish".