Changes to moderation finalised

6 May, 2008

NZQA has finalised details for the introduction of random sampling of internally assessed student work for external moderation.

This will replace the current self-selection of samples by schools.

Previously schools have been asked to select samples at the borderline of grade boundaries - the most difficult judgements. From July a system of random sampling will provide the basis for selecting moderation samples.

Over and above random sampling, teachers will also be able to request specific moderator feedback on student work that has raised issues needing clarification, particularly at the borderline between grades. Deputy Chief Executive, Qualifications, Bali Haque said the feedback teachers received on these issues would be qualitative. "It is about discussing the process involved in making an assessment judgement rather than confirming a grade decision," he said. "It will not appear on the moderation report for that standard."

Moderation is a process whereby internal assessment judgements are checked to ensure they are at the national standard. It is not a process of re-marking student work.

Improvements to NCEA announced last year included increasing the amount of moderation from 3% to 10% of student work. The increase will be achieved by increasing both the number of standards and the number of samples of work for each standard that are moderated. NZQA has appointed 33 moderators for the larger subjects, and these moderators started work in January 2008. Contractors are continuing to moderate in smaller subject areas.

A pilot of random sampling was carried out last year, involving Level 3 standards in 89 schools.

Mr Haque said the pilot had highlighted areas of improvement in the random sampling system.

Selection must now be undertaken by a person other than the assessor of the standard and should follow an alphabetical or numerical system. The selection method used by each school will be documented and filed for each standard and will be available for checking by NZQA.

Moderation results in previous years have indicated a 70% agreement by moderators with teacher assessment. The random sampling pilot indicated similar agreement.

"After looking at the results of the pilot, we have clarified the procedures for random selection of samples, and introduced an additional option, over and above the random sampling, which allows teachers to send further evidence close to the grade boundary to moderators and ask specific questions."

For more information contact:

Steve Rendle
Senior Media Advisor
NZQA
DDI: 04-463-4240
Mobile : 0274-575783

Page updated: 06 May 2008