Setting targets is no way to improve educational standards, argues former headteacher and now legal consultant to the Association of School and College Leaders (ASCL), Richard Bird
Every sensible businessman knows that when you make an indicator a target it ceases to be of any value as an indicator. For example, it may be a useful indicator for a car sales business to know how many cars it sold in a new registration month. However, its value goes if salespeople are told that the target this year is to increase sales in that month: particularly when the judgement of the success of an outlet is based on hitting the target. Sales are delayed from the previous month; they are brought forward from the month following. Now sales may indicate the ingenuity of the sales force; but not much else. This piece of folk wisdom unfortunately appears to have passed by government. Once it is stated, educational examples multiply. In the 1950s businesses used to advertise ‘Smart Lad Wanted’. (There never seemed to be the same demand for smart lasses.) Then opportunities opened up. Suddenly the lads answering the ads didn’t seem smart enough. So, as firms caught on, they began to advertise for youngsters with five O-levels or, if they weren’t too fussy about the degree of smartness, five Grade 1 CSEs. Then came GCSE and the indicator of smartness was five A-Cs at GCSE and, finally five A*-Cs.
Fixed percentages
Since this seemed to be a de facto standard of employability, the government picked it up as an indicator of the relative success of secondary schools. The problem was, of course, that grades at O-level had been statistically determined: only the smart lads could pass because only a fixed percentage were allowed to. You might have a dazzling selection of skills but if you were not in the top cohort you did not pass. As an indicator of smart lads it was useful; but not as indicator of standards.
But GCSE was referenced to attainment. If you made the standard, you made the grade. The number of pupils getting five A*-Cs went up. As a result it might be of use to an employer who was interested in a set of skills. It could tell the government how many youngsters were meeting a national standard. However, if the employer was still interested in the smart lad (or lass by now) then it was useless, because more and more children were meeting the standard. (A similar problem now faces the Russell Group universities.) What was happening was what Matthew Arnold, government inspector and poet, had prophesied long ago: ‘The school examinations… are, as I have said, a game of mechanical contrivance in which the teachers will and must more and more learn how to beat us.’
Beating examination systems
For Matthew Arnold the question was how teachers could ‘by ingenious preparation’ contrive to get children through examinations ‘in reading writing and ciphering, without their really knowing how to read, write and cipher.’
The skills of beating examination systems have not been lost to teachers today. Question-spotting; framework-providing; technique-coaching are all alive and well and producing their misleading results. And all this was before governments determined to use the indicator as a target. Then, of course, the inevitable happened. Techniques were sharpened and scores rose. But this was not all. The problem is, though, that as the stakes rise, the temptation is to adjust and alter fundamentals. For example, the kind of learning tested, however inadequately, by O-levels and essentially by GCSE, is in a sense dissociated. That is, the nearer it is to independent generalisation and abstraction, the more it fulfils its purpose. But many British youngsters feel that this dissociated learning has no connection with their very practical aspirations for their careers. We do not in Britain have the German meister tradition, where a senior technician is expected to be a master of theory as well as practice and to lay the foundations of that mastery of theory in vocational school. There is not the cultural assumption that practice must rely on theory. So to give youngsters the sense that their aspirations had a place in school there have always been examination subjects that at least give the impression of being linked directly to work. To say this is not to devalue such work but simply to point out that it is different from the O-level/GCSE sort of work and intentionally so. GNVQs have been the manifestation of this in England and Wales over recent years, and the majority of the new diplomas will be another. As the stakes rose for headteachers to improve their five A-C pass rate, those GNVQs which gave four GCSE equivalent became a weapon of choice.
Changing meanings
As a result five A-C or equivalent has changed its meaning. This does not mean that the work the pupils have done has been useless or that they have not learned something valuable. It is just that five A-C no longer means ‘smart lad wanted’ nor that the overall standard of education in generalised or abstract learning has risen. It is not necessarily better or worse in itself. It may be better. It is simply useless as an indicator of the kind of improvement it was originally intended to measure.
The effects of these indicator/targets have run wider. Teachers and smart lads and lasses have been smart enough to spot that some subjects are intrinsically more difficult than others. As a result, a new generation of English children is being educated in the belief that they are genetically incapable of conversing with foreigners in their own languages. At this point the indicator/target becomes not just useless but a definite source of harm. What has happened at Key Stage 4 has happened at Key Stage 2 with a vengeance. The level for average children has been turned by fiat into a standard that has to be reached to access the secondary curriculum (though without research into what actually is necessary to access the secondary curriculum). In this wonderland schools must aspire to make all children average or above. Academic studies show that reading standards, after falling during the second world war, improved fairly rapidly until the 1960s (essentially getting back to their pre-war level). With the widening of the primary school curriculum under the National Curriculum reading scores probably fell back. There has been a recovery since about 1995 but real progress has been slow and painful. Nevertheless, key stage test results have soared. The only explanation for this apparent paradox is that teachers have improved their coaching techniques, rather than children improving their performance. The ‘pressure and support’ provided by government has speeded up the evolution of coaching.
An old problem
Though test results have risen an old problem has re-emerged: ‘That which was fixed as the minimum for gaining the grant on every child becomes the maximum of the teacher’s aims and efforts, and everything else is, not perhaps intentionally, but certainly practically, discouraged by the enormous value attached to the required subjects. To make the essential support of the schools depend on reading, writing, and arithmetic has, I know, struck the death knell in many a school to that higher teaching out of which intellectual stimulus is well nigh exclusively drawn.’ (Inspectors’ Annual Reports, 1866-67, p261)
At Key Stage 2 in 2007 half of all curriculum time is spent on maths and English. The same danger now faces the secondary curriculum as Key Stage 3 ‘loosens up’ at the very same time that schools are being put under pressure to achieve a two-level improvement over the Key Stage. This is serious. A ‘loosened up’ Key Stage 4 is no longer intended to provide the balanced general education for school leavers. Key Stage 3 has had to shoulder that task. If Key Stage 3 goes the same way as Key Stage 2, we shall be in serious danger of producing the highest scoring and – in so far as their formal education has anything to do with it – the most ignorant generation since the abolition of the Revised Code. There are none so blind as those that cannot see and those who are hopelessly committed to targets feel that ever smarter and more ingenious targets will solve all these problems. This underestimates the teaching profession and the pressures of teaching. The search for ever-smarter targets is futile in purpose and dangerous in effect. The potential damage that targets have done, however, is as nothing to what they will do if they are brought down to the level of the individual teacher and made to stick by being linked to pay. Fortunately, this is being resisted by trade unions.
Finding the answers
It is easy to criticise of course. What should we be doing? Part of the answer has got to be to turn what are now targets back to indicators and to use those indicators with sufficient care for them to be useful. But the key focus has to be elsewhere.
If we turn to management gurus from outside education we find that they leant heavily on statistical knowledge. However, they did not rely on statistics to set targets. On the contrary, W Edwards Deming, for example, saw targets as disastrous for precisely the reasons we have seen. The true use of statistics was to define accurately when a low performance was the result of individual lack of capacity and when the result of faults in the system. The way to improvement was to improve process. The way to make true progress is to focus on improvement in process, not to improve outcome scores. But it must be a process that is connected to the purpose of the work, not focused on a target. Processes in a school can be as general and as undiscriminating as how homework is set. They may be something as focused as assessment for learning. These key success factors if they come together, will result in genuine achievement. For a teacher to be working scientifically on process and for this work to be managed leads to genuine educational improvement. Some of the improvement may be fairly accurately indicated by some tests when these are not an end in themselves. However, even they are no longer targets, test results in themselves are not wholly adequate indicators. For example, is the success of ICT in schools to be judged by ICT tests or by the use young people make of IT and the internet? Outcomes as measures are legitimate in a research project. In a research project the validity of outcomes is part of the debate: are they the correct measures; do they measure accurately the right things? But if outcomes are placed beyond debate, as they have to be to set targets, then they are worse than useless and as we have seen, nothing but distortion can be anticipated.
References
Arnold, M ‘The Twice Revised Code’, in Arnold on Education, London, Penguin 1973