Great data project

Stuff reports:

New Zealand has one of the most unequal education systems in the world, according to Unicef, and OECD research indicates students from low-decile communities perform better than their higher-decile peers if they attend advantaged schools.

To counteract such biases, some Western systems have turned to value-added (VA) models of assessment that measure students’ improvement throughout a year, rather than simply recording an end grade, like NCEA.

Yep the VA model is the most useful – how much gain has there been.

While VA models established in New South Wales and Tennessee provide a blueprint for efforts here, critics say they promote a one-size-fits-all approach that could narrow the curriculum and institute performance-based pay for teachers.

What’s more, without complex data on students’ backgrounds, they may not highlight the effect social advantage has.

Joel Hernandez thinks he can fix that.

The researcher at the New Zealand Institute, a public policy think tank, is about half way through a year-long project to build New Zealand’s first “contextualised” VA model.

Using integrated data from the ministries of education and social development, corrections and immigration, and weighted NCEA results from the country’s 500 secondary schools, the model aims to adjust for socio-economic factors and determine what effect individual schools have on students’ achievement – if any at all.

The unreleased preliminary results, Hernandez says, are significant.

By Christmas, the model should be able to profile students’ likelihood of success based on older students with similar backgrounds; how well they do in NCEA, whether they’re likely to take on tertiary study, “whether they end up in a Corrections facility, [or] their potential to go on a benefit”.

That looks like fantastic research. Can’t wait for the results.

The data available here is more sophisticated than that on which other jurisdictions base their VA models.

 Statistics New Zealand’s Integrated Data Infrastructure includes household income, internet access, parents’ education, and children’s health when they start school.

“Other countries are just using things like a free lunch as a proxy for lower socio-economic [status]”, Hernandez says.

Yeah the IDI rocks.

However, privacy laws limit what the New Zealand Initiative can do with that information. In accessing the data, it has agreed not to identify which schools are the most and least effective.

“At the moment, we’re arguing it’s impossible to identify a single school from a single dot [on a graph],” Hernandez says.

Crampton will instead offer the results to the Ministry of Education in the hope it will relay them to schools – especially if it turns out the variability between and within schools is higher than expected. 

Hopefully the Government will use this info.

Comments (28)

Login to comment or vote

Add a Comment

%d bloggers like this: