According to Professor Lindsay, such freedom may come at a high price. The weaker schemes will produce invalid assessments of children’s current stage of development, he predicts. Too many children may therefore be misclassified, as they were when this form of assessment was used experimentally in the 1960s. In other words, some assessment schemes will not be useful - as either a diagnostic tool or a base for value-added measures.

Inevitably, the Qualifications and Curriculum Authority has argued that Professor Lindsay doth protest too much. All the accredited schemes, it says, have satisfied the QCA criteria and all will be field-trialled before they are introduced.

Professor Lindsay is not, however, the only doubting Thomas. Dr Peter Tymms of Durham University also questions whether some schemes are “reliable” - in the sense that two teachers using the same checklist would reach the same conclusions about any one child. And there are other doubts. Will there be a knock-on effect on nurseries, forcing them to prepare children for the assessment? Is there a danger that in mapping the abilities of very young children, some negative labels will be attached?

Most early-years teachers are, however, familiar with this form of assessment, and are sanguine about the September launch of the national system. In Surrey, Birmingham, Sheffield, Wandsworth and many other authorities, baseline assessment is well-regarded and long-established. In Birmingham, where assessment is based on teachers’ observations of children over a period of weeks, the authority feels that it not only provides invaluable information for schools but stimulates staffroom discussion on teaching and learning.

Such assessment can also provoke the useful dialogue with parents that the QCA is seeking. It may even make an important contribution to racial equity. The confirmation that black children in Birmingham are just as able as their white classmates when they start school has refocused attention on the reasons for their relatively poor performance later on in the system.

Even so, baseline assessment will have to stay on trial until a thorough evaluation can be carried out in two or three years. One thing is, however, already clear. The multiplicity of schemes will not deliver the straightforward value-added data on children’s progress and school performance many politicians would like. In any case, recent research has confirmed that some infant schools teach maths more effectively than English, and vice versa; equally, some “bring on” children with low baseline scores but do little for pupils of average ability.

It would be convenient for the compilers of performance-tables if the truth were more simple. Unfortunately, it isn’t.