Within minutes of walking into her first high-school physics class Amelia Unufe felt like an outsider.
"Are you sure you're in the right place?", the teacher said. The 16-year-old paused, and looked around at the rest of the students, hoping to find a friend.
Instead, what she saw would leave her confidence shaken. In that moment, Amelia realised she was the only brown kid in the room.
Last year, only 10 per cent of Pasifika students took NCEA Level 2 physics, compared to 17 per cent of all students across New Zealand secondary schools. In disadvantaged areas, those entry rates dropped even further. The figures were worst for Maori. At decile 1, the poorest schools, only 3 per cent of Maori were taking Level 2 physics, a total of 68 kids. Less than half scraped through to pass.
A Herald investigation has found the low uptake among poorer, Maori and Pasifika students wasn't confined to physics. It happened in chemistry, in maths, in history, in English.
It was a clear trend, long suspected by educators, but hidden in a murky backwater of previously unpublished data, and obscured by the simultaneous rise in achievement against Government targets.
Pass rates for the National Certificate in Educational Achievement (NCEA) this year hit an unprecedented high. The proportion of 18-year-olds who passed NCEA Level 2 grew by more than 10 per cent in just five years, to a staggering 83.5. The Government aims to reach 85 per cent by the end of 2016.
Achievement for Maori and Pasifika students climbed fastest. Maori pass rates - while still lower than all other ethnicities - surged 3 per cent in the last year alone. The results were celebrated in a media release by education minister Hekia Parata, who paid tribute to teachers, parents and students for the "excellent rise".
However, Herald analysis shows that beneath that rising tide was a deep sea of disparity. More students were getting NCEA, but at every step Maori, Pasifika and low-decile students were getting a different type of NCEA to their Pakeha, Asian and high-decile peers.
Official data on all standards sat in every subject for 2015 shows that Maori, Pasifika and low-decile students were less likely to take academic subjects than Pakeha, Asian and high-decile students.
When they did take academic subjects, Maori, Pasifika and low-decile students were less likely to sit exams.
When they sat exams, Maori, Pasifika and low-decile students were less likely to pass.
When they passed, it was less likely to be with "Merit" or "Excellence" grades, meaning a less competitive grade on application to limited-entry university courses, such as medicine, or engineering.
At the most severe end, the data showed less than half of Level 2 entries for the poorest Maori students - those from decile 1 schools - were in academic subjects, just 45 per cent.
Instead, Maori, Pasifika and low-decile students were more likely to be enrolled in "unit standards", gaining skills-based credits in more vocational subjects, most which are not university-approved.
For example, decile-1 Maori students were four times more likely than decile-10 Pakeha to take subjects in the "services sector" field - an area including hospitality, tourism and retail.
Disproportionate numbers of low-decile students were also enrolled in a miscellaneous field called "core generic", encompassing a mishmash of standards covering employment, communications and life skills.
A coffee-making course had low-decile students enrolled at almost twice the rate of students at more affluent schools.
"The data doesn't show us anything that we don't already know," said education minister Hekia Parata.
"It's a dilemma," said The University of Auckland's curriculum expert, Dr Aaron Wilson.
In a paper co-authored with the Ministry of Education's chief education science adviser Professor Stuart McNaughton, Wilson argued the divisions needed to be addressed.
He said while there was an important place for vocational pathways at secondary school, and while not all students would go to university, it was not right that low participation rates in more academic courses were more common for poor, Maori, or Pasifika students.
"In a society committed to equity for all its citizens, education cannot allow itself to
perpetuate divisions based on socio-economic status or ethnicity."
A flexible system
NCEA was introduced in 2002, replacing School Certificate, Sixth Form Certificate and Bursary with NCEA Levels 1, 2, and 3. Level 2 quickly became the benchmark - described as the minimum needed for "further education, employment, health outcomes and a better quality of life".
NCEA differed widely from the old system, firstly in that it wasn't scaled, so technically everyone could pass. Secondly, it was much more flexible. Subjects were broken down into standards, with each standard appointed a number of credits.
Standards could be internally or externally assessed, and both practical skills and academic knowledge counted towards the qualification in the same way. Each level required 80 credits to pass, but those credits could come from any mix of standards imaginable, as long as the numeracy and literacy requirements were met.
Level 2 could therefore either lead via Level 3 to university; or to a polytechnic course; or a trade. But if subjects and standards were chosen poorly, or at random, it also ran the risk of leading to very little at all.
"Choosing the wrong options too early closes doors," says Ant Backhouse, the chief executive of the I Have a Dream trust, which mentors students from disadvantaged backgrounds. "And it does happen really early. Subject choice for NCEA Level 1 is made at age 14. That's a pretty young age to be narrowing pathways."
For example, if a student dropped science at Level 1, and instead enrolled in tourism, it became almost impossible to change back. And because university entrance requires a certain number of credits from "approved" courses - the more traditional curriculum subjects - that could limit their future options, he said.
"Kids look at subjects like travel and think it's an easy way to get NCEA Level 2. But what happens when they get to Year 13 and decide they want to go to university and they can't?"
To make matters more complicated, participation in specific standards could also affect a student's future. Take maths - at most schools students weren't allowed to progress from Level 1 to more advanced courses if they hadn't passed certain algebra standards. But because algebra was hard, many dropped it and instead ended up taking alternative maths courses made up of unit standards, which didn't count when it came time to apply to university.
"That's what happened to me," said Anthony Vainikolo, an 18-year-old nursing student in his first year at AUT. "NCEA was confusing, and then they told me at the end I didn't get University Entrance because I'd done the wrong maths."
Anthony had to spend six months taking a Level 4 paper with the help of a Uni Prep course to gain entrance to his degree. That meant more money on his student loan, and more time at university, and therefore more pressure on his family. He already works through the night at Countdown two days a week to help out, as well as studying full time.
"If I'd known I would have focused on doing the right papers. But I was young, and going with the flow and wasn't really worried until the end."
International studies have found channelling students into non-academic pathways too early can cause problems.
The issues run deeper than simply gaining entrance to university - although data shows that as an option, tertiary study remains attractive given the higher pay packets of its graduates.
Some courses were "dead ends" and didn't allow for transitioning to any kind of further study. Some added little value in terms of numeracy and literacy. Some simply didn't extend students in the way academic subjects would.
The OECD's "Equity and Quality" report found students who left school without the relevant skills also cost society, with poorly-educated people limiting economies' capacity to "produce, grow and innovate".
Victoria University's Michael Johnston, the co-author of the newly-released book NCEA in Context, said the skills taught in traditional subjects were important for "active citizenship" in the modern age.
"You've got to be careful not to say that some kinds of knowledge are more valuable," he said.
"There's nothing wrong with subjects like hospitality or tourism or the trades, but they don't tend to give students the deeper educational benefits of having to evaluate and think critically," he said.
"Sometimes I didn't know where to turn"
Behind closed doors, principals have been studying trends in NCEA data for years. Unlike the public, they have access to masses of information about pass rates - and unlike the Herald, they have breakdowns by school.
Many will compare their marks with their neighbours, trying to work out who is doing what to get their kids to pass.
Allegations have swirled about schools "gaming the system" to bolster pass rates, reaching their peak when official statistics are published each year.
The most controversial practices are well known: Schools might push students into "easier" pathways - such as those with more internal assessments or where lower literacy levels are required.
Students who fail too many tests during the year might be withdrawn from that subject, and placed in a "more suitable" course.
Similarly, kids on the cusp of failing could be pulled out of their regular programme and encouraged to focus only on the credits they're missing out on; or allowed to re-sit the same assessment again at a later date.
Teachers might also hunt for standards to match to skills - one teacher told how she had failed a student sitting an English achievement standard in giving a speech, but was able to find a unit standard in "communications skills" to award instead. That way the student was still rewarded for her work, and gained vital credits towards NCEA.
While the Ministry of Education said it aimed to discourage schools from picking and choosing so-called "easy" standards, none of the above practices are outside the rules. In fact, the rationale behind them often makes sense.
If a student is two credits off gaining a qualification, getting them back in January to do some extra work and pass seems the right thing to do. They will be able to move to the next level, and they'll feel a sense of accomplishment.
As one teacher put it: "Say you've got students from families with endemic unemployment, or generations failing at school, and suddenly they're passing. That's huge. Before, the split used to look like upper-decile schools passing, lower-decile schools failing. Now at least lower-decile schools are helping kids get something."
Ant Backhouse believes, however, that too often schools are not acting in the students' best interests, but their own. Pass rates are used in marketing to attract out-of-zone pupils, and the effect of the 85 per cent government target is strong.
"Some schools are obviously precious about their own results. They're not prepared to take a risk because it's going to show up negatively. So students are manipulated into subjects where they schools know they can pass."
In low deciles the effect was exacerbated, he said, as many families didn't understand the system and weren't able to push back against the lingering low expectations of Maori and Pasifika kids. Instead groups like I Have A Dream were left to fill the gaps, Backhouse said.
Amelia Unufe was one of his. To get to that first physics class, Backhouse first had to help mentor her from the bottom maths stream to the top. It then took a series of meetings with the dean to argue she was capable of taking high-level science.
Even then, the teachers' assumptions: "Are you sure you're in the right place? Did you meet the criteria?" left a lasting sting.
"I think because my older brothers dropped out or kicked out of school the people expected the same from me. They thought I was just another brown kid trying to get into Uni," Amelia said.
"Being placed in the 'bottom class' at the start as well, I felt dumb."
At the time, Amelia was set on a career in health, so she knew she had to take maths and science. She also chose accounting, English and technology fabric. Eventually, Amelia changed her mind, but due to her broad choice of subjects she was able to meet the entry requirements for a fashion design degree.
"Lots of people think if you're going to study fashion you won't need university entrance, but you do," she said. "And now I'm about to start my own business, so I'm grateful I took maths and accounting."
Amelia, of Tongan heritage, was the first in her family to go to university. On arrival, she was again disappointed to find she was the only Pasifika student in her course.
"It was really tough. I had my culture backing me up but at the same time I didn't have anyone like me. Sometimes I didn't know where to turn."
A focus on quality not quantity
Each year when NCEA results are released, schools with significant improvements in achievement get a letter of congratulations from the education minister.
At Heretaunga College in Upper Hutt, Principal Bruce Hart isn't holding his breath.
In 2014, his was one of the few schools in the country which saw their Level 2 results drop, after a change in the way it runs its NCEA programme - one that does not include aiming for 85 per cent.
"I don't resist the target I just don't pay attention to it," Hart says. "The number means nothing. It doesn't tell you anything about the meaningfulness of that qualification to a student's future."
Heretaunga does its best not to employ what Hart calls "dodgy" practices to get students through. Instead it has re-written all its courses, broken down subjects further to provide more contextual, engaging programmes to allow students to really personalise their learning.
This includes subjects like "dystopian fiction" - made up of English standards; or "beyond the field" - looking at the social impact of physical education, which combines standards from different disciplines. Instead of the usual 24 credits, courses have just 14, and students will do seven or eight a year, instead of the traditional five.
It also has a hospitality academy that goes to Level 3, and a relationship with an agricultural college. All its students work towards "meaningful pathways" - be that trades, or services, or academic.
It's been hard work, but Hart is determined students will leave with everything they need for their next step, even if it means his school languishes in the middle of the league tables.
"If I wanted, I could lift my Level 2 pass rate straight away by getting a bunch of kids to sit something like 'retailing'," he said. "But for a large number it's not relevant. Equally, if you're doing an employment skills course, then CV writing is great. But to take kids out of a maths course and put them in that to lift their credits, that is a concern."
He says he fears there's been a switch from developing knowledge and skills to a focus on passing at all costs.
"I have some real cynicism about it. If you have a meaningless target and everyone strives to get to that target then that's what becomes important and not necessarily the quality of qualification," he said.
"It's a conundrum. Are we producing independent, life-long learners, by doing all these things? We haven't really debated the impacts of what we are doing."
Last year, a Herald survey asked all secondary schools in Auckland for their students' reading levels upon entry at Year 9. The results were stark.
While all schools had a huge variation,at the lower deciles, which are largely populated by Maori and Pasifika kids, huge chunks of the student body were arriving, at age 13, with the reading levels of nine-year-olds.
Chair of the Secondary Principals Council, Allan Vester, said those schools faced a "Catch-22".
"You want to give students the best chance to go on to something useful post-school but don't want to put them into courses where they have little chance of success," he said.
"But if school is going to be the portal to further study and a career, and fewer students are choosing subjects that are academic than that's reducing their opportunities."
Vester said if schools were trying to catch kids up, the real argument needed to be around which kids should get the help, and how much help they should get. School funding was currently under review, and he was hopeful meaningful changes would be made.
A fix to the issue may be more complicated than funding and resourcing, however. Despite the crucial role of schools to NCEA, most experts don't believe the blame for the inequality should fall to teachers and principals.
Chief education science adviser to the Ministry of Education, Professor Stuart McNaughton, said, rather, the inequality was an unintended consequence of the design and delivery of the education system as a whole.
"I don't see this as a matter of an isolated fault. Assessment itself is just one part of a very complex system. We need to identify those parts of the system that contribute to this particular outcome and on which we can work."
McNaughton highlighted early maths and science education; guidance counselling; and school resourcing as areas for further attention.
He said a shared understanding of "desirable outcomes" for secondary school graduates was also an area for work.
In other words, we need to discuss what we want our kids to aim for.
"We're in the business of education"
The NCEA target was due to expire this year. However, education Minister Hekia Parata has no plans to get rid of it, and it will continue until at least 2017.
Asked if the target had contributed to the inequalities, the minister did not respond.
Instead Parata said more young Maori and Pasifika people were leaving school better qualified and with brighter prospects. NCEA was enabling students to study what they were passionate about, she said.
"After all, not everyone is interested in learning how to conjugate French verbs or perform complex quadratic equations but that doesn't mean they are not learning."
She said the Herald analysis of subject-level data showing the inequality for poor, Maori and Pasifika students didn't tell her anything new. In fact, it was the reason the Government had initiated a number of programmes aimed at raising achievement within those groups.
"We recognise that there is still more to be done to continue improving outcomes for students from lower decile schools and Maori and Pasifika students, and we are focused on that," Parata said.
"But that shouldn't mean that we don't celebrate the achievements already taking place."
Both the minister and the ministry - which says it was also aware of the inequalities at subject level - were keen to emphasise the quality of NCEA. Previously, debate has centred on whether it stood up internationally, or whether practices like internal assessment were robust.
More than 10 years on, experts say it is time for the conversation to shift away from quality, and instead to practice - and away from devolving into a "two-tier" system.
Victoria University's Michael Johnston says it's about questioning what we value. For him, the fix lies in a return to a focus on deep, meaningful learning, rather than ticking boxes and gaining credits.
"The thing about NCEA is that it's not a dreadful or a wonderful system in and of itself. It requires good practice to get the best out of it."
"You need integrity to use it well. You have to say, bugger the political pressure. We're in the business of education."