Exam performance comes under
THE EXAM system has been at times an explosive subject, but when the managing director of the Edexcel exam board likens incoming changes to nuclear power, teachers ought to sit up.
Jerry Jarvis is talking about the sophisticated ways exams data can now be used. He wants to start a discussion in The TES about the potential of technology to transform the learning experience for pupils.
He understands there will be fears it could be misused and that once a decision is taken to provide information, it is difficult to retract.
From this year, pupils taking Edexcel exams will be able to receive online, if their school agrees to it information on their grades in particular papers and how close they were to grade boundaries.
However, this is just the start for Mr Jarvis, Edexcel and, possibly, other examination boards.
Heads, heads of department and individual teachers are to be handed detailed information on pupil performance, right down to how they performed on each question. At the click of a mouse, a school will be able to identify areas of pupil strengths and weaknesses. Comparisons will be possible between teaching groups and, by implication, between teachers.
In future, the same data could be offered to pupils and their parents.
Those who have sat exams could be told their likely chance of winning an appeal, which might help them in challenging their A-level results, or being successful at an exam resit, which would help after AS grades are announced.
And schools are likely to be offered the chance to analyse the data almost any way they choose, with the ability to compare pupil performance against what might be expected given the pupils' gender, ethnicity and the type of school they attend, for example.
All of this information has been made possible by Edexcel's move to examiners marking GCSEs and A-levels on computers. The system was launched in 2003 and now 90 per cent of its papers are handled in this way.
Mr Jarvis views the potential to hand over more information to parents as the most controversial move.
One fear, he concedes, is encouraging an exam resit and appeal culture. But he says it is only right that pupils are given the best chance of success.
Another worry is the potential for litigation. Pupils will be able to compare how their group fared against similar pupils taught by a different member of staff, right down to how specific questions were answered. This could provide ammunition to prove that a poor grade was caused by poor teaching.
"There is this notion of ambulance-chasing lawyers putting pressure on the system," says Mr Jarvis. "I do not want schools damaged by lawyers."
It is this fear, and the question of how the information might be used by heads, that is of most concern to the National Union of Teachers. John Bangs, the head of education, wants a protocol drawn up on how the information is to be used.
"In a sense, the more fine-grained information and feedback that is available to teachers and students, the better," he said. "But the problem is that the way it is couched could mean the results will automatically provide a picture of the quality and effectiveness of teaching."
Other factors, he warned, such as events in pupils' personal lives, could have an effect on results but would not be factored in. The protocol should say that information provided by the board should be used to inform in-school debate on future improvements, rather than punitively against staff who might be seen to be underperforming. It should not be used in school self-evaluation forms for Ofsted inspections, he said.
"Teachers will be nervous that the information that comes out will lead to false assumptions about the quality and effectiveness of teaching," said Mr Bangs.
Mr Jarvis concedes that a poor set of exam results does not necessarily mean bad teaching; other factors, including staff turnover or teacher illnesses, could explain dips in grades.
Professor Margaret Brown, of the Advisory Committee on Mathematics Education, is concerned that the information will lead to even greater teaching to the test. "This will provide a more efficient way of coaching children through exams." she said. "But I'm not sure it will do much for their general education. It provides a more scientific way of teaching to the test."
Heads are likely to welcome the extra data. Martin Ward, the deputy general secretary of the Association of School and College Leaders, said: "In terms of being able to see which questions your pupils answered well and which they did not, I think that's potentially useful to schools."
Edexcel highlights on its website testimony from schools that took part in a 1,500 pupil trial last year, of the scheme's potential to help schools looking to raise results.
Mr Jarvis believes that information should be shared widely, because it has the power to transform teaching and because pupils' futures hang on the results.
"I think organisations such as mine should be as open and transparent as possible," he says. "This stuff is too damn important to pupils. If you have been badly taught, you should know about it. Why should you be in a lottery, for God's sake?"
Jerry Jarvis, page 12
SCREEN GRABS: EXAMPLES OF THE INFORMATION EDEXCEL'S ONLINE SERVICE WILL PROVIDE
Edexcel's Results Plus service provides information sorted by school or college, by teaching group and by student.
For the school or college
Screen 1 (left) shows its overall performance in exams this year, GCSEs in this example, and links to other analysis screens.
For the teacher
Screen 2 (right) shows the performance of pupils in his or her class on a particular paper. Teachers are told the average score of their pupils for each question and then how this compares to the national average on that question. Headteachers and heads of department will also have access to this information.
For the pupil
Screen 3 (far right) shows how a pupil performed in a particular paper. The mark for each question is set out on the left, with the national average listed on the right. In two GCSE subjects (maths, as here, and science), this screen also identifies which area of the syllabus is being tested by each question.
WHAT THE NEW COMPUTERISED SYSTEM OFFERS SCHOOLS
This month, secondary headteachers will receive an overall summary of their exam results, including the percentage of GCSE A*-A grades and the larger percentage of A*-C grades, how these compare to last year's figures and the national averages for this year and last.
They will be able to compare pupils' grades with those in "similar" schools, although the categories used are relatively crude. State secondaries, for example, are listed simply as comprehensive, selective or modern.
Schools will be given analyses of results by gender and the chance to analyse the grades of each teaching group.
Each candidate's grades in each subject will be listed, with a "gradeometer" showing how close they were to achieving the next highest mark.
The feature that is potentially most powerful is information on how each class performed on each question, including each pupil's marks as well as the average score.
In two GCSE subjects, maths and science, Edexcel has mapped each question against the syllabus. This will allow schools to compare how they fared against the national average in each section, which will help them to focus on areas of weakness next year.
The board says eventually this analysis will be available across all subjects.
In future, it may also make available a digitised version of pupils' exam scripts, possibly for a fee.
All of this information has been made possible by Edexcel's move to computerised marking.
Potentially just as revolutionary, in terms of pupils' learning experiences, is the board's foray into computerised on-demand testing. From February, candidates for 650 papers will be able to take a specially written multiple choice exam designed to seek out strengths and weaknesses in their understanding of the syllabus. The marks, given immediately by computer, will not count towards their grade but may inform further teaching.
Edexcel is likely to charge for test entries, but says the fee would be insignificant to schools.
AQA, England's biggest exam board and an Edexcel rival, says it already makes similar tests available for its new range of science GCSEs.