Students do much better when they are internally assessed than when they are put under the pressure of an exam, a comprehensive Weekend Herald analysis of NCEA entries reveals.
Internal assessments are becoming increasingly important to secondary school students as the place of old-fashioned exams fades - and they generally result in better achievement.
The difference in achievement rates between the two types of assessment can be nearly 50 per cent, although the gap differs according to subject, level and school decile.
From today, readers can use an interactive graphic on nzherald.co.nz to see exactly how the results differ.
To take one example: students in decile one schools who were studying maths with calculus achieved 83 per cent of internal assessments at Level 3 in 2012. The achievement rate for external assessments was much lower, with only 34 per cent of entries achieved.
One person to have raised concerns about internal assessments is Professor Dale Carnegie, head of engineering and computer science at Victoria University.
A vocal critic of NCEA who advocates a return to percentage scores, he said there were well-founded concerns over the moderation of internal assessment.
As a result, engineering depart-ments at Victoria University and the University of Auckland had moved to insist students completed specific external assessments.
"In essence, we do not trust the internal assessment anywhere near as much as the external," Professor Carnegie said.
However, the Ministry of Education and other experts say the differences are to be expected and show the system is working.
Rowena Phair, the ministry's deputy secretary, viewed the data and said it reflected NCEA's flexibility and relative complexity as opposed to the old exam-based system.
There were several factors that contributed to students achieving better results in internal assessment, Ms Phair said:
• Students could be assessed at a time when they were ready for assessment, rather than months later at the end of the year,
• A reassessment opportunity might be available following further study,
• A wider sample of student evidence could be used in making the final judgment on student achievement.
"The difference in achievement rates between internally and externally assessed standards can be seen across other subjects and has not changed substantially over time."
Internal assessments are set and marked by teachers, with grades checked by other teachers and samples in turn checked by NZQA.
That process and the overall integrity of NCEA is overseen by an independent advisory group, the Technical Overview Group Assessment.
Group chairman, Emeritus Professor Gary Hawke, said there was "certainly no cause for concern" about what the Weekend Herald data showed, and the public could have confidence in NCEA and its assessment methods.
The focus was on monitoring the validity of assessment against the relevant standard, and "the equivalence of internal and external assessment is an incidental outcome of those processes".
Professor Hawke said internal and external assessments were intentionally used to assess different learning, and a gap in achievement was therefore not necessarily alarming. Any disparities were always investigated, but could have many reasons.
Internally assessed standards allow teachers to give students much more explicit guidance, which was one logical explanation for the general pattern that internal results were higher. "Remembering that we are concerned with recognising achievement and not selecting an elite, it should be understood as a better directed assessment process rather than any reduction of rigour."
Education Minister Hekia Parata said the moderation process was thorough, and various independent reviews had backed the integrity of NCEA assessment.
"We take it seriously. I was criticised for putting a commissioner in to Moerewa School in the Far North, and that was because we found their awarding of NCEA to their senior students did not meet the moderation standards of NZQA."
The Weekend Herald analysis also shows that in many subjects the achievement difference between internal and external assessment lessens in higher decile schools - those that are in wealthier areas.
Taking the maths with calculus example, at decile 10 schools, 95 per cent of internals were achieved at Level 3, compared to 74 per cent of externals. The gap was much bigger at decile one schools (83 per cent to 34 per cent).
Asked about this difference, education researcher Professor John Hattie said it was not surprising nor concerning.
The former University of Auckland academic, who now works in Melbourne, said too many higher decile schools had internal assessment that too closely mimicked exams. "The internals were meant to be different in nature and form from the externals and certainly not expected to be so highly correlated as measuring different attributes within a subject."
The flexibility of NCEA meant schools could choose which standards and units to include in a subject, Professor Hattie said. This would likely create differences between subject structure at different deciles.
Allan Vester, chairman of the NZ Secondary Principals' Council and head of Edgewater College in Pakuranga, agreed, saying another likely influence was student literacy.
"I think there has been an increasing link between results in the externals and results in the English standard that assesses ability to handle an unfamiliar text. A trend in NCEA externals has been to increase the degree of literacy required. This means that students in lower decile schools where the achievement rates in the unfamiliar text standard is lower will also have difficulty in other externally assessed standards."
Dr Stuart Middleton, of Manukau Institute of Technology, said students at lower-decile schools would typically be less secure in their knowledge of a subject, and the environment of internal assessments would be more supportive.
"There is a view in developmental psychology that a real indication of what someone can do is what they can do in a setting that's helpful."
Pressure to perform
Schools are favouring the use of internal assessments in part because of Government pressure to raise student achievement, some of those involved in education say.
Provisional NCEA results released by the Government this week showed significant improvement over five years.
But there are questions about what such a drive to push up results means for students, schools, and the integrity of NCEA itself.
The Government has set a target for 85 per cent of 18-year-olds to have NCEA Level 2 or an equivalent qualification in 2017.
Allan Vester, chairman of the NZ Secondary Principals' Council and head of Edgewater College in Pakuranga, said internal assessments allowed different skills to be assessed than in an exam and generally had lower rates of non-achievement.
That meant it was likely that they would be seen by schools as a way of meeting the 85 per cent target.
Another factor was a desire to perform better in "league table"-style comparisons with other schools. That was not necessarily negative if handled correctly, he said, and safeguards included a rigorous and improved moderation process.
However, Mr Vester said, if the push for achievement went to unrealistic levels there was risk that public perception of NCEA would suffer.
In total, 76.8 per cent of students left school last year with at least NCEA Level 2, compared with just over 74.3 per cent in 2012. The achievement rate has increased by more than 10 per cent since 2008.
Though results are improving, others - such as the Programme for International Study Assessment (Pisa) international rankings - have not, and some university departments are concerned that students' school achievement does not match their knowledge.
Another worry is students being steered into "soft" subjects that lead nowhere but reflect well on school pass rates, with alleged examples documented by the Auckland-based I Have a Dream trust.
Education Minister Hekia Parata dismissed the notion that achievement targets could be harmful and cited the professionalism of teachers as one reason that marks were robust.
"If we thought their job was to go out there and just give kids qualifications because it's going to meet a target set by a government, that's impugning a whole profession," she said.
Asked if the 85 per cent target might be pushed up if it is met by 2017, Ms Parata said it could well be.
"I'm already saying to schools now, go to 100 per cent and knock yourselves out ... we are pretty serious about this target, but we're serious about it for meaningful qualifications."
A Weekend Herald analysis of NZQA data shows a large increase in the number of internal assessments being used - up from 1,517,360 entries in 2008 to 2,061,721 in 2012.
One cause has been the deliberate replacement of unit standards with internally assessed achievement standards by NZQA. Unit standards can only be assessed as Achieved or Not Achieved and are normally used by less academically able students. Though the overall number of internal assessment entries has rocketed upwards, figures provided by the Ministry of Education do not show significant changes when credits are taken at a per student level.
Ms Parata said the balance of external and internal credits used to achieve NCEA levels had remained largely the same.
However, students were now doing more assessment than necessary and if "extra" credits were taken into account, it did change the comparison.
Mr Vester said schools were "safeguarding" the ability of students to succeed by letting them pick up more internal credits during the year. Such an approach means that some students now sit end-of-year exams needing fewer marks to achieve their NCEA level.
That would effectively reduce the importance of exams. However, Ms Parata said, comparison between different assessment methods missed the point - more key was the act of gaining credits itself.
"The point of an education system isn't to try and catch kids out. The point is to try and find their learning style ... external examinations just don't suit everyone."
NCEA: a guide
• Since being introduced in 2002 NCEA has become New Zealand's main qualification system for secondary school students.
• It is much more complex than traditional exam-based qualifications.
• NCEA is offered at three levels: Level 1 (usually Year 11/15-year-olds), Level 2 (Year 12/16-year-olds) and Level 3 (Year 13/17-year-olds).
• Students' learning is assessed both during and at the end of the school year.
• Assessment is "standards- based" - if a student achieves a set standard for a particular section of a subject or course, they are awarded a pass, regardless of how others have performed.
• Each standard carries a certain number of points or credits that are awarded when a student satisfies the assessment for that standard.
• NCEA allows schools to select which standards (components) to include in a subject, and they can adjust that subject to suit the needs of their students.
• The same flexibility means schools can offer different versions of core subjects like English - some designed for "academic" students, and others for students thought to need a more "practical" knowledge.
• In place of traditional A, B, C, D grades, NCEA credits are awarded as Not Achieved (N), Achieved (A), Merit (M) or Excellence (E).
• Some assessments are done internally and during the school year, by those teaching the subjects.
• Other assessments are done externally through exams and portfolios and at the end of the school year, overseen by specially appointed examiners.
• Subjects are often made up of a combination of internal and externally assessed standards, which means students earn some credits during the year and more through end-of-year exams.
• To achieve a course endorsement (recognition for students who perform exceptionally well) both external and internal assessments must be used.
• If a student fails to achieve a standard or set of standards that are internally assessed, there is normally one more opportunity to resubmit written work or be reassessed for it. No such opportunity exists for externally assessed standards.
Source: Understanding NCEA, by Irena Madjar and Elizabeth McKinley, Starpath Project for Tertiary Participation and Success.