A common measure is what lets an outsider be recognised at all

The argument

Defenders of standardised testing argue that its purpose is to make achievement legible to people who do not know the student. Before common examinations, places at selective institutions were allocated largely through recommendation, interview, family connection and the reputation of the school a candidate came from, and these channels systematically favoured those already inside them. A test that everyone sits and that is marked without knowing who wrote it gives a student with no connections, from an unregarded school, a way of demonstrating what they can do that does not depend on anyone vouching for them. Defenders argue that the alternatives proposed all reintroduce the judgements the test was designed to bypass. Teacher assessment varies with how well a student is liked, how confident they appear and how resourced their school is. Portfolios and projects reward the students with help at home. Interviews reward manner, accent and social familiarity. References reward knowing someone worth asking. Each of these is a channel through which existing advantage operates, and each is harder to audit than a test score, because there is no comparable record to inspect when a decision looks wrong. Defenders also argue that comparability has uses beyond selection. A common measure is what makes it possible to see that one school's students are being taught far less than another's, that a particular group is falling behind, or that a change in method has helped. Without it, each school's account of its own performance is the only account available. Defenders argue that the inequalities critics of testing care about are known largely because tests revealed them, and that abolishing the measurement would remove the evidence while leaving the inequality in place.

Premises

Counter-arguments

Critics argue that the meritocratic account describes the intention rather than the result, since preparation for these tests is now an industry and the students who do best are those whose families can buy coaching, practice and time. They argue that anonymity at the point of marking is irrelevant when the inequality enters long before the paper is written. On detection, critics argue that inequality can be measured by sampling a population for research purposes without attaching consequences to individual students, so the diagnostic use does not require a high-stakes system. And they note that the historical comparison flatters testing by setting it against overt patronage rather than against the better alternatives now available.

Rejecting the premises

[Rejecting P1] Anonymity at marking is beside the point when the inequality enters through years of unequal preparation before the paper is written. [Rejecting P2] The alternatives can be moderated, sampled and audited, and the comparison is being made against overt patronage rather than against better modern methods. [Rejecting P3] Inequality can be measured by sampling a population for research without attaching consequences to individual students. [Rejecting C] A mechanism whose outcomes can be bought through coaching does not in fact recognise outsiders, whatever its design intends.