Trust in testers

It is time to replace the examination régime at 16 and 18 by something more appropriate. The coalition government has been solidifying its place by its Baccalaureate reforms at both ages, but this is a move in quite the wrong direction. Whatever the wider purposes that the examination system may serve – competition for higher studies, comparisons between schools, or whatever – its core aim is to find out how well students are faring in their learning. Most people take it for granted that public examinations are the most appropriate vehicle for this. They do not always ask whether there are better alternatives. The examination régime has many faults, among them financial, psychological, sociological and ethical. It is a multi-million pounds drain on school budgets; it causes anxiety in many examinees; it disproportionately benefits students from families and schools who know how important examination success is in getting ahead in life and do what they can to guarantee it; it encourages undesirable attitudes like teaching to the test, an instrumental attitude to learning, and even, sometimes, cheating. In addition, its inclusion of norm-referencing as well as criterion-referencing features can obstruct the core aim just mentioned. We saw an example of that in 2012 in the use of five-year-old data from SATs results at 11 to predict GCSE performance at 16. The examination regime also has epistemological deficiencies. These are more serious in that they strike at the heart of what examining is supposed to be. The more public examinations diverge from testing facts or routines, the more problematic their results can be. Following a philosophical line of argument that Davis (1995, 2008, chap. 1) has opened up, if we are looking for evidence of how far students have a deeper grasp of a subject – the English Civil War, say – this means taking into account the way in which a wider background of ideas bears on the connections they make between one consideration and another. What links do they make between notions – both past and present – like economic development, social class, religion, kingship, democracy, power? These connections, and this background, differ from individual to individual. Unlike the recording of an isolated fact, they are not present on an examination script. They belong to its hinterland. This raises a conceptual difficulty for examiners looking at the work of multiple candidates all of whom are strangers to them. How well can they know what lies off-the-page? If they make a judgment about what this is, how far will this judgment depend on their own background of ideas, including their value judgments, and so is likely to differ from that of other examiners marking the same question?