How frequent testing and individual targets can raise attainment.
Excellent teaching, short modules, interesting activity-based lessons and frequent tests followed by the setting of attainable targets are one recipe for improved attainment in science. Huxlow School in Northamptonshire, where the A to C grade percentage has improved over four years from 27 to 56 (putting it high among last year's 50 most improved schools) is an excellent example. At this 600-pupil comprehensive the science department has been in the vanguard of a drive to increase attainment levels at key stage 3 - which is where the spadework for improved external examination performance has to be done.
Late last year, the Office for Standards in Education said of Huxlow's KS3 National Test results: "Attainment in science is well above that achieved in similar schools and above the national average for all schools." The target setting regime which has been crucial to this has been devised and run by Jon Davis, head of science recently promoted to deputy headship.
"We started from the premiss that we had to add significant value," says Mr Davis. "Our children are below national average on entry, so it was not going to be enough to just do a standard job."
The baseline is the National Foundation for Educational Research Cognitive Abilities Test given to the pupils as they leave primary school. Then, through KS3, the science syllabus is divided into two to three-week modules, all activity based, all incorporating good Science 1 work. At the end of each there is a test similar in structure to the KS3 national tests.
"The first year we ran it," says Jon Davis, "pupils had to get 50 per cent to pass - and they retook it until they did." Even this fairly basic approach resulted in improvement. But it soon became clear that some refinement was needed. "It was unfair to the brightest, who only had to get half marks to pass, and also to the less able, some of whom had trouble hitting 50 per cent," says Jon Davis.
It was a problem that some tests were harder than others - the average of the scores recorded by all the pupils varied from 45 per cent on one test to 75 per cent on another. At this point, it might have been easy to become bogged down in trying to refine the tests. However, says Jon Davis: "We are not test writers. To rewrite them all would have been difficult if not impossible."
The answer was to leave the tests as they were and to base notional "pass marks" on the average performance of pupils who had taken the test previously. For example, the Year 7 "Life Processes" test has turned out to be quite easy, and the average score of all pupils taking it is 71 per cent. In the "Forces" test, however, the average is 55 per cent. There are more than 30 tests in the KS3 science programme, and the average mark for each is known and recorded. This mark is used for target setting.
The data collected so far shows that a pupil who hits the average for every test throughout the key stage has a high chance of achieving level five in the KS3 national tests. When a pupil is being urged to improve on this they are told how much they need to exceed the test average by - the very brightest may be asked to add 25 per cent each time - so that on some tests, a pupil may be expected to get full marks. A less able pupil, aiming for less than level five, could be given a target of perhaps 15 per cent below the average. Each child is tracked and the performance closely monitored. The running average is recorded, because this irons out any exceptional performances in individual tests. Scores are discussed, new targets are agreed with the pupil and written down. A high-performing pupil's targets will be revised upwards but, says Jon Davis, "Targets are never lowered." There is a multi-layered system of merits and certificates to reward pupils who consistently hit or exceed their targets.
There is good correlation between module test results and KS3 levels because the department now has enough data to underpin the statistics and overcome anomalies. Every child in the year group (sizes of year groups vary, but typically there might be 90-100) will do about 30 tests in KS3. As more data is collected, the whole system becomes more reliable.
There is clear demonstration here of how an individual department can influence whole school improvement. Science, with other content-rich subjects, is particularly suited to a system of short modules with practical work and frequent tests.
The principle of close targeting is used in other departments at Huxlow, however, and Jon Davis is helping to refine the process. Headteacher Christine Staley says: "Target setting is the key to all improvement along with the quality of teaching and learning."
In September, of course, with a new curriculum and presumably new modules, the whole thing will have to start again. As Jon Davis says, though:
"Within the life of a national curriculum, enough data is gathered to give you the information you need."