I'm not much given to envy, but when the key stage 3 English tests come round, I look over at my colleagues in the independent sector and go green. They do not have to do them. They can do better things instead.
So what is wrong with the tests? Well, their dependence on descriptors for a start. There is a place for descriptors in English: they can detect a given range of skills in answers candidates will all formulate differently - such as explaining a point of comprehension or writing a letter in a prescribed way. But the downside is that descriptors work on a "best fit" basis. That means that no element of the tests is identified in the candidate's mind with a strict right-or-wrong culture. "Best fit" also admits discretion in marking that can lead to inequities.
A strict right-or-wrong culture is indispensable if the highest educational standards are to be achieved and English should not settle for being a "get as much right as you can" subject when it can do better. In the best English, everything is right but as long as KS3 tests rely unduly on descriptors, they will foster neither perfectionism nor watertight marking. The solution is a separate section to test spelling, punctuation and grammar on a right-or-wrong basis. All of these were highlighted by the Qualifications and Curriculum Authority's Standards At Key Stage Three English (1999) as needing attention anyway, so isolating and concentrating on them can do nothing but good. Key stage 3 English should also include formal verbal reasoning tests. Precise methods of thinking are as much a part of good English as precise usage.
Another weakness of the tests is the complete Shakespeare play - prescribed by authority with minimal choice, entailing unnecessarily daunting reading skills for many and tested terminally. Classroom experience shows tha the complete Shakespeare play is not the best way to develop a love of Shakespeare in 14-year-olds.
The teaching of Shakespeare requires a multi-faceted approach, powered by the enthusiasm and expertise of individual teachers. Shakespeare's life is interesting and reliably documented and should be studied, along with his times. A lot is also known about Shakespeare's venues, stagecraft and actors and all that should be drawn on, too. There is absolutely no need to study a complete play before GCSE. Far better to work on a selection of scenes that can really mean something to individual classes.
As for terminal testing, forget it. Criterion-referenced, externally moderated coursework is the way forward. That has been true since the complete Shakespeare play was first legislated on to teachers in state schools. It is even truer now: word-processing makes redrafting easier, desktop publishing encourages pupils to take pride in their work, and the Internet facilitates research. That includes The TES's www.learnfree.co.uk, though I do wish that, when the Bard answers, he could include his sources. All teachers have an example to set, including electronic ones.
Another faulty aspect of the English tests is the appeals procedure, not least because the tests are descriptor-based; making appeals is a time-consuming and often unsatisfactory business unjustified by the value of the tests themselves.
The appeals procedure can best be improved if the tests are improved. That means primarily reducing the quantity of "best fit" marking in the timed components and letting schools develop and assess the Shakespeare component themselves. If implemented, this last point would also meet the common complaint of teachers that today's examinations culture inhibits innovation and demeans their professional judgement.
Dr Colin Butler is senior English master at Borden grammar school, Sittingbourne, Kent