Editorial: Confidence in exam must be restored

The news that the inquiry into the new scholarship exam has been widened to the entire national certificate of education will have shaken educational circles in this country. It suggests a fundamental re-assessment of the newly introduced system, the work of more than a decade of careful and contentious design and preparation. All along, voices have been warning it was fundamentally flawed. It sounds like the Prime Minister, for one, agrees with them.

Education Minister David Benson-Pope, reflecting the mood in the ministry no doubt, is not ready to concede that much. He did not formally announce the widening of the scholarship inquiry, preferring to mention it in passing during question time in Parliament. When that news attracted more attention than he wanted for it, he sought to downplay it, suggesting a wider inquiry was always implicit in the original terms of reference. Those terms made no mention of the NCEA.

The last thing anyone wants to do is undermine the confidence of students in the new assessment system early in a new school year. But they have been well enough briefed on the system they face to know that the variations in results of the scholarship exams offered for the first time last year are likely also to be a characteristic of the standard NCEA. Greater variation in results is only to be expected from a system that has abolished the insidious practice of scaling students' marks.

Helen Clark complains that the Qualifications Authority, which devised and manages the new system, does not acknowledge there is a problem. Whether there is a problem depends on the view one takes of scaling. If you believe, as the Qualifications Authority does, that it is unfair and discouraging to students who do not achieve as well as others, you will accept an alternative with a less predictable range of results. But if you believe, as many academics do, that scaling is a perfectly valid and proper way to iron out differences in the difficulty of subjects, the setting of exam papers and the expectations of markers, you will worry about the alternative.

The new system rests on the assumption that objective standards of achievement can be defined and exams set accordingly with reasonable consistency across subjects, between markers and from one year to the next. "Reasonable consistency" might not mean the unwavering, predetermined range of results that was engineered across all subjects every year by means of scaling to the bell curve of intelligence. But reasonable consistency probably should mean results more equitable between subjects than the scholarship has produced.

If those who set test papers and mark them cannot produce more consistent results than the first scholarship has produced, what confidence can we have in the rest of the Qualifications Authority's procedures for ensuring exam setters and markers are working to similar standards? This question goes to the heart of the idea of "standards-based achievement" on which the whole NCEA is based. No wonder a survey of school principals, as reported in the Weekend Herald, suggests they too have lost confidence in the system.

It is very early for this to happen. The NCEA has been phased in to three senior levels of secondary school over three years and last year was the first run for level 3 NCEA and scholarship. Teachers, even the many who remain sceptical of its practicality, appear to have made a conscientious attempt to follow its procedures for internal assessment and "moderation" of standards between schools. Their confidence and commitment will have been shaken by the latest results.

It is up to the Qualifications Authority, the Education Ministry and its minister to restore the confidence of teachers and students quickly, not by denying there is a fundamental problem but by demonstrating they can get to grips with it, and fix it.