Labour social development spokeswoman Jacinda Ardern questioned what the cut would mean for the study.
"Obviously, when you are carrying out a longitudinal study, consistency is critical - but also the size of the cohort; that was a part of this study that allowed us, really for the first time, to get the scale we needed to build a better understanding of the different ethnicities that have been growing up in New Zealand."
Ardern said she'd been told that the cohort's reduction in size would impact the ability to build this understanding.
"This is a massive cut in the number of the people who are part of this study, so it really raises questions around to what degree is that going to undermine the study itself, and also the ethics of the study.
"There are families who are really invested in this now, who will be told of course that they're no longer a part of it."
The study's administrators have not yet responded to requests for comment and Superu referred questions to Goodhew's office.
Goodhew argued a smaller sample size did not reduce the value if the study, and statistical experts were consulted by both parties to finalise a sampling approach which enables the research and analysis that is needed to answer the critical questions.
"The study will continue to support quality research and analysis, while also being more sustainable for the long-term.
"It was always envisaged that the study would become self-sustainable. The study is already funded in part from other sources.
"Now that the study is well established, we believe it is time for it to be funded through a contestable research pool like similar studies.
"This will be the case from 2018/19."
Goodhew said that since the study started, more sources of longitudinal data had become available, and how we think about longitudinal studies has evolved.
But Associate Professor Craig Stevens, president of the NZ Association of Scientists, said a 70 per cent reduction in sample size less than half-way through any study was "a drastic re-shaping" of the original concept.
The cuts highlighted the difficulties in funding any long-term study, he said.
"Making a big splash at the launch and then expecting work that is essentially for the public good to 'become self-sustainable' essentially sets up a slow-motion train-wreck that compresses the initial vision, scope and quality of the evidence produced.
"If we want to have a society that makes decisions based on knowledge and data then we need to support this type of work, at a carefully though-out sample size, that gets us the understanding we need.
"All this would have been justified, evaluated and accountably reviewed in the initial proposal."
Science commentator and Auckland University physicist Professor Shaun Hendy said while administrative data collected by government agencies was a powerful source of new information, he believed it should be regarded as complementary to longitudinal studies rather than a replacement.
"It is clear that administrative data can be biased by the way it is collected and by the priorities of the agencies collecting that collect it.
To get full value from administrative data, other independent sources of data were needed to correct for these biases, he said.
"For this reason, I think that longitudinal studies will remain very important, maybe becoming even more so as we try to make the most of administrative data."
