When pass rates at GCSE and A level steadily increased year-on-year, education secretary Michael Gove told us that this was bad. It was a result of dumbing down, teaching to the test, slack examining, easy coursework, too many retakes and so on.

However, when the proportion of schools judged to be good or outstanding leaped by 9 percentage points in a year (“Ofsted’s approach ‘is not backed by research’”, 13 September), Ofsted chief inspector Sir Michael Wilshaw (inset, below) told us this was good. It was a result of him having “galvanised the system”.

Does Sir Michael not know that correlation does not imply causation? That any change in a system designed to improve it often results in a temporary improvement simply because there is a change (the so-called Hawthorne effect)? And are we to understand that it is a good thing that this year exam pass rates have declined (so our students are ostensibly not as good) while Ofsted ratings have improved (so our schools, apparently, are better despite this)?

I think the two Michaels need to take a basic statistics class. I’ll be happy to teach it.

Matthew Handy, Director of mathematics, dotmaths.