Random NCEA sampling too late for student guinea pigs

Today's announcement that Labour is finally adopting random sampling of NCEA marking is not only an admission of failure but comes at least seven years too late says National's Education spokeswoman, Anne Tolley.

"Labour was told in 2001 that there should be random sampling for all work. They only acted on that advice in May last year, and even then only introduced a pilot."

A report written in 2001 by government statistician David Rhoades, who was contracted to review the moderation system, compared it to the previous model, and considered whether NCEA was credible enough to be understood by the public and education sector. That review was damning.

"Doctor Rhoades made it clear that he was concerned about the whole process of marking selection. He insisted that marked work must be randomly selected."

His report said:

* Given the amount of effort that is going into moderation, the statistical value of the moderation sample is less than it could be.'
* The moderation system will be unable to reliably detect problems with marking a particular standard at a particular school in a particular year.'
* The proposed system has no statistical procedure for maintaining standards between years'.

"Only now, seven years later, has Labour got around to taking expert advice.

"Unfortunately, it is far too late for the 250,000 students who have been the guinea pigs in Labour's NCEA experiment."