Andreas Schleicher finds the criticisms in TES levelled against Pisa tests to be "completely unjustified". He declares Pisa to be a "robust" measure of performance across countries (and uses this word a number of times) but also admits to a large measure of uncertainty in the rankings. He says that this is no secret - but it does undermine the claim of robustness, which implies dependability; something that Pisa does not have.
He also admits considerable variations in topics taught across a number of countries. But this further undermines the reliability of the tests. If tests reflected only things taught in exactly the same way, he says, we would end up with a useless lowest common denominator of knowledge. Therefore, to construct viable and wide-ranging tests at a high level across 50 countries seems an impossible task. It has always seemed to me, as a mere teacher of English, that to construct "robust" (ie, valid) language tests across just two or three European languages would be nigh on impossible, let alone 50.
It does not follow that in the interests of "accountability" (but whose?) we must have these Pisa results, even though they are flawed. This calls into question the whole concept of accountability.
Nigel Probert, Porthmadog, Wales.