Does he agree with the statement of Bali Haque, deputy chief executive of the New Zealand Qualifications Authority, on the decision to start randomly checking schools’ internal NCEA marking: “The positive is that it’s happening now.”; if so, why?
Yes; I think the decision to pilot the random sampling of moderation of internal assessment as announced by the New Zealand Qualifications Authority on 12 March is a positive move. The addition of random sampling as a further quality assurance check for current moderation systems was a response to the “post-assessment moderation” part of the 2005 review.
Why was random sampling not made part of the National Certificate of Educational Achievement (NCEA) moderation model 6 years ago, when the New Zealand Qualifications Authority failed to act on official advice from the Rhoades report in 2001, which noted that failing to have a decent sample of randomly selected school work would undermine the validity and reliability of the whole NCEA moderation model?
Because people did not agree with the recommendation made in the Rhoades report at that time. But the same issue was picked up in 2005, and that is what is being implemented now.
That is a good question, because it is important to put each of these changes into the context of the fact that there was a major review of NCEA in 2005, and over the past 14 months I as the current Minister have been engaged in working with the New Zealand Qualifications Authority and the Ministry of Education through the issues raised by that review. We have addressed about 168 recommendations to date. There are 14 that are ongoing in nature. We have made a number of changes, obviously, in the operation of NCEA, which means that it is working extremely well now, and we have been making changes in design—for example, the record of learning and results notes are clear that the internal assessment results are on line earlier, the grey-point average is no longer recorded, the profiles of expected performance have been refined, and I am about to announce in the next few weeks the last of the recommendations and how they will be addressed.
Would the Minister agree that whilst it is important in any national assessment system to have high levels of reliability, which random checking will assist, it is just as important to achieve high levels of validity, which standards-based assessment caters for somewhat better than norm-referenced assessment?
The member is exactly right. This standards-based assessment, of course, does provide for a more valid form of assessment than one that simply takes an aged cohort and strings it out along a bell curve.
Why did the Minister and the New Zealand Qualifications Authority fail to act on official advice, again, in 2004 to implement random sampling for the moderation of NCEA internal assessment—advice given by the State Services Commission to “improve the validity of moderation”; why is he only just now announcing a whole raft of changes to NCEA that totally contradict everything he said about NCEA since he has been the Minister?
If I can just put these comments back into context for the member, who is also new to the role, as I am—that is, there was a major review in 2005 and we have been steadily working through 168 recommendations to date. Fourteen are ongoing, with the last few to be addressed in the next few weeks, and I am sure she is looking forward to those.
How come the Minister and the New Zealand Qualifications Authority have been scurrying about, implementing a pilot for random sampling to check internally assessed work for NCEA, when only 5 months ago both of them were saying that the assessment model was “robust and reliable” and there was no problem?
If I can put these comments back in the context that I have to keep putting it for the member, we have been unrolling a range of recommendations since 2005. I think in fact the use of the 3 percent sample of the scripts that were coming in from borderline decisions was a robust and very useful model, but the recommendation in 2005 is that we should carry on and address random sampling, and that, of course, is what we are doing.
What context can one put his statements in so they do not look like a flip-flop, when only 5 months ago he was saying that the assessment model was “robust and reliable” and now he is implementing a pilot to demonstrate the exact opposite?
It is a bit strange for a member from that party to talk about flip-flops. I think that every single major policy that National has stood against over the last little while has now been flip-flopped on. So I am sure they do not want attention drawn to that. I just go back to the fact that we have been unrolling a range of recommendations since 2005. We are just about at the end of them. I know that when they are finished the member will welcome them and begin to support the NCEA. After all, Mr Bill English was one of the people who helped to introduce it.
Does the Minister accept, when he is discussing the unrolling of a raft of changes to NCEA, that those changes completely contradict everything he has been saying about NCEA since he has become Minister, in particular that random sampling of marked work for internal assessment is now required, when only 5 months ago he said there was no problem and that everything was robust and reliable?
Since I have had this job I have been firmly in favour of standards-based assessment—
—as my predecessor was, and as was Dr Nick Smith when he was in Government. I understand that the National Party has always consistently supported the notion of standards-based assessment. I am sure the current spokesperson does too, and that is why she is looking forward to the finalisation of the few remaining recommendations.