Readers can use an interactive graphic on nzherald.co.nz to see exactly how the results differ.
Murray Black, deputy principal at Auckland's Lynfield College, said he did not believe internal assessments were necessarily easier, "but there is an element of the known".
Click here to see a graphic showing NCEA results.
"And there's an element of being able to balance the teaching focus ... a little more aligned to what the assessment is going to be that the teachers have made.
"I don't think it's anything deliberate, but there is a subtle difference between the externals and internals in that respect."
Mr Black's role is NZQA exam management at Lynfield, a decile 7 school with 1852 students, and he said increasing internal assessments could make learning more intensive for both student and teacher.
A realignment of the standards available to schools by NZQA over the past three years had replaced some external and unit standards with internal assessments.
"We are looking at it more from the fact it's more pressure on students and staff to manage a year's worth of study where there are more internals.
"We're not necessarily keen on having more internals. But the structure has changed."
Internal assessments are set and marked by teachers, with grades checked by other teachers and samples in turn checked by NZQA. They can often be re-sat by students, unlike exams.
The Herald analysis also shows that in many subjects the achievement difference between internal and external assessment is less in high-decile schools - those that are in wealthier areas.
Students in decile 10 schools who were studying maths with calculus achieved 95 per cent of internals at level 3, compared with 74 per cent of externals. The gap was much bigger at decile 1 schools (83 per cent to 34 per cent).
Some university engineering departments have expressed concern that students' school achievement does not match their knowledge, and pointed the finger at internal assessment.
Other experts say independent reports have shown that marks given under NCEA can be trusted and internal and external assessments assess different learning, and therefore differing achievement rates should be expected.
Meeting standards and students' needs
When Kirsty Farrant set about designing a new science course for her students she shied away from standards that involve them sitting exams.
The physics, chemistry and biology courses at Wellington's Newlands College are still relatively traditional in structure, but the new science issues course is different.
Ms Farrant, head of science at the school, said she wanted the course to reflect an aim set out in the New Zealand Curriculum, which was that students would become "responsible citizens in a society in which science plays a significant part".
Even if students would not go on to be scientists, Ms Farrant wanted them to finish with competency and skills that would play a part in their future life - for the scientific content learned in the course to be a means to other important ends.
She said only internal standards gave her the flexibility and control over content such a goal required.
"We could pick something that's topical and is of interest to the students, and use that for our teaching and assessment, rather than relying on external assessments which are far more prescribed in terms of the context and content."
About 80 per cent of the science issues course is internally assessed, including standards on the impact of dairy farming on the environment and how humans might survive on Mars.
On the perception by some that internal-heavy courses were easier, Ms Farrant said the more accurate observation was that they were different.
"In the science issues course, there are some standards that we do that require a huge amount of work, they're no easier than sitting an external.
"The difference is you can assess them in a way that meets the needs of the student but also meets the requirements of the standard. I honestly think often it's about difference."
