Maestro De Ingles

Testing for competence rather than for "intelligence."
By McClelland, David C.
American Psychologist, Vol 28(1), Jan 1973, 1-14.
Abstract
Argues that while traditional intelligence tests have been validated almost entirely against school performance, the evidence that they measure abilities which are essential to performing well in various life outcomes is weak. Most of the validity studiesare correlational in nature and fail to control for the fact that social class might be a 3rd variable accounting for positive correlations between test scores and occupational success, and between level of schooling achieved and occupational success. It is suggested that better measures of competence might be derived by analysis of successful life outcomes and the competencies involved in them,criterion sampling, and assessment of communication skills. (41 ref.) (PsycINFO Database Record (c) 2012 APA, all rights reserved)

It is not often that a newspaper article annoys me to the point of wanting to set the record straight. But Conor Ryan's defence of the Government's testing regime (Comment, 12 February) was a masterclass in the selective use of evidence to make dubious points. No onewho has witnessed the workings of the test system at close quarters for several years, considering the evidence in the round, could look at it objectively and believe his central argument is right.
Last week's primary review from Cambridge University said that children's learning was distorted by tests. Ryan's view is that, although Sats have been under pressure, we should not be panicked intoscrapping externally-marked national testing, for the notion that results can be used to hold teachers to account for pupils' performance – to everyone's benefit.
It is a seductive point, which readers might buy were they unfamiliar with the extent of the contrary evidence. In fact, each of Ryan's arguments is less than watertight. Sats results have indeed improved under Labour. But even if oneaccepts that the gains are genuine – and there have been independently-verified studies suggesting results in English are substantially overstated – there is no evidence that the testing regime itself has driven this. Other factors include the introduction of dedicated teaching strategies for English and maths and the recruitment of an army of classroom support staff over the past 10 years.
Thetest results are also of questionable reliability for gauging the overall quality of a child's understanding. The best evidence of this is the fact that most secondary schools set exams for new Year 7 pupils, because they do not trust key stage 2 test data.
The most misleading point is saved for last. Ryan says: "By all means change the tests and reduce the marking burden, but let's not abandonaccountability."
But no one who is arguing for reform wants to ditch accountability. Personally, I would advocate school-by-school accountability to parents by, for example, more detailed, rounded Ofsted reports, but there are other models..
By "accountability", Ryan means accountability through the results of a single set of narrow national tests. This encapsulates the belief that the onlyeffective way to teach is to set pupils a centrally-designed test at the end of their course, with rewards for schools with good results and penalties for schools with bad. This is exceedingly pessimistic, given the investment the Government has made in a generation of teachers it says is the best ever. But it is a view.
What is really dishonest is not to be open about the mounting evidence on thedownsides.
In last year's Select Committee inquiry into exams, the Government was the only one of 52 respondents to defend the current system. Ranged against it in recent years have been the select committee itself, the Children's Society, the Royal Society and a host of other scientific organisations, the Cambridge University-based Primary Review, at least one major exam board, teachers' unions and...