Late last year, David Koretz published a book that exposes the fallacies behind using standardised test data as a means of evaluating teachers and schools[1]. The Testing Charade is a brilliant and well-researched critique. It focuses on the US where education policy has yoked teachers’ pay, position and progression to success in meeting test-based targets, but it resonates with our own situation in England.

Koretz adapts Campbell’s law, to the effect that when test scores are used for accountability, they are subject to corruption pressures that distort the education they are trying to measure. He compares this with the distorting effects of “targets” in health care - and suggests that Campbell’s law is at the heart of the VW emissions scandal. In education, he sketches a continuum of responses that lead from outright cheating - including altering student answers, but encompassing the pressure to remove students from the testing cohort by whatever means - through questionable practices in coaching students to recognise patterns in questions, using techniques that get high marks without any real understanding. He wonders where we draw the line between cheating and “gaming” the system: all such exercises constitute a corruption of good teaching.

Problems with testing

His main point is that measurement-driven policy leads to measurement-driven instruction and that this does not lead to improved standards. The focus on test-prep leads to distortions with which Brits are all too familiar: reallocation of resources to subjects covered by the key assessments, and reallocation within subjects to the topics that appear in the tests. Koretz inveighs against Lemov and Farr - in his view, their manuals for effective teaching posit assessment not as the end of the teaching and learning process, but as the starting point - putting the cart before the horse.

Evidence is emerging that teachers’ success in raising test scores is not necessarily correlated with success in raising student engagement[2]. Measures of teacher effectiveness based on test scores leave out important dimensions of what makes a good teacher. 

In this country, teachers’ pay and prospects are less directly linked to standardised tests than they are in the USA. But high-stakes public exams are used, through league tables, to invite comparisons between schools. This creates pressure on teachers to optimise short-term outcomes, at the expense of long-term learning goals.

The distorting (corrupting?) effect of exams is nowhere more egregious than at GCSE. This battery of high-stakes tests distorts education for the two - increasingly three - years preceding it, for what end? Since it no longer marks the terminal point of compulsory education, it cannot be justified in terms of leaving certification. Instead, it has become a means of establishing accountability at system-level of measuring school success. Schools and teachers should be accountable to students, parents and society. But surely this is possible without imposing the kind of testing that, as Koretz shows, bends out of shape the education that we all want to improve.