Pure criterion-referencing failed
The introduction of NCEA was from one extreme to the other extreme on the norm-criterion referencing scale. The old exams were scaled so that the top student always got 96% and exactly half the country passed and half the country failed. The mark was, in effect, a ranking, not an indication of the candidate’s ability. Then NCEA went to the other end: there was to be no scaling whatsoever, only the comparison of a student’s work with set criteria.
The change was clearly a little bit too much. Any good system uses both norms and criteria. Either without the other will be meaningless.
The recommendation of the SSC is a fundamental shift in the assessment of NCEA. It is good that they allow tolerances and explanations, though — it is not such a drastic change in the sense that it is still designed to have standards remain consistent over time. But to allow scaling, the "artificial technique of ensuring a sufficient number of passes", as the NZ Herald editors put it, makes the results just that: artificial.
If all goes well, they shouldn’t have to use these safety nets to scale results. But this is another of NCEA’s reality checks. In the end, as politically correct and philosophically correct as it may seem, pure criterion-referencing without normative moderation won’t do.