School Equity Index 2023

Forums › Off topic › School Equity Index 2023

freitasm

BDFL - Memuneh
80652 posts

Uber Geek
+1 received by user: 41037

Administrator

ID Verified

Trusted

Geekzone

Lifetime subscriber

#312000 6-Mar-2024 16:58

I was reading about school funding and how the Equity Index replaced the Decile Index.

That's fine. The Education website links to a spreadsheet listing the amount given to each school depending on Equity Index.

But they don't link to a school list anywhere.

I found an OIA request that was denied because "the information is public" with a link to a Sharepoint-hosted spreadsheet. But that spreadsheet is not linked from anywhere else except the OIA, so not very "public".

For the record, here's the link to 2023 School Equity Index (New Zealand).

Referral links: Quic Broadband (free setup code: R587125ERQ6VE) | Samsung | AliExpress | Wise | Sharesies

Support Geekzone by subscribing (browse ads-free), or making a one-off or recurring donation through PressPatron.

jonherries

1434 posts

Uber Geek
+1 received by user: 316

Trusted

Subscriber

#3203841 6-Mar-2024 17:36

The interesting thing to me is the song and dance about how much better it is, but there is no documentation (or there wasnt when i last checked - good reminder to check again) about how it is actually calculated.

There is a bit of pseudo-math but nothing that actually confirms how it works, how peer review was done or why it should be trusted.

Edit: update still nothing that shows anything remotely like statistical analysis.

Jon

Basil12

124 posts

Master Geek
+1 received by user: 74

#3203848 6-Mar-2024 18:11

It is a bit odd. The Please Read page refers to a table below of the 37 variable but then doesn't list them. And despite having an area covering Taranaki / Wanganui / Manawatu the list suggests Linden, Foxton, Shannon and Manawatu schools are in Wellington.

Quik internet referral link

cddt

1965 posts

Uber Geek
+1 received by user: 1904

#3203896 6-Mar-2024 20:24

Incredibly frustrating that statistical models are used but not made public.

I reckon it's because they know that if all statistical models developed by the government were made available (e.g. from Education, Justice, Treasury, MBIE, etc.) then some ~~pesky~~enterprising statistics postgrads will go through and find flaws in half of them.

A lack of transparency is a lack of accountability.

My referral links: BigPipe | Mercury

Goosey

3016 posts

Uber Geek
+1 received by user: 870

Subscriber

#3203961 7-Mar-2024 06:36

cddt:

Incredibly frustrating that statistical models are used but not made public.

I reckon it's because they know that if all statistical models developed by the government were made available (e.g. from Education, Justice, Treasury, MBIE, etc.) then some peskyenterprising statistics postgrads will go through and find flaws in half of them.

A lack of transparency is a lack of accountability.

Isn’t it the post grads usually working for these crown departments?
and the been a grad for a few decade types?
(no offence to anyone), but in a world where bad actors like stealing data from health systems, I’d say there’s bad actors wanting to interfere with the ever popular education zone and decile arguements.

wellygary

8810 posts

Uber Geek
+1 received by user: 5287

#3203992 7-Mar-2024 09:04

Its on the web, but again it is an orphan link, there appear to be no pages the include the URL

https://assets.education.govt.nz/public/Documents/our-work/changes-in-education/2023-EQI-numbers-All-schools-and-kura.xlsx

ockel

2031 posts

Uber Geek
+1 received by user: 545

#3204016 7-Mar-2024 10:20

Basil12:

It is a bit odd. The Please Read page refers to a table below of the 37 variable but then doesn't list them. And despite having an area covering Taranaki / Wanganui / Manawatu the list suggests Linden, Foxton, Shannon and Manawatu schools are in Wellington.

Here is a list of the variables https://assets.education.govt.nz/public/Documents/our-work/changes-in-education/Equity-Index-variables-fact-sheet-Aug-2022.pdf

While I'm not a huge fan of the EQI it cant be worse than the overly simplistic decile system it replaced. However it has swung to the complete other end of the spectrum with 37 variables per child for 517,000 primary school children and 305,000 secondary school children. Thats a huge amount of data with zero visibility of the integrity of the data. This is now a data analysts dream for collecting, collating and analysing. And managements worst nightmare for data quality.

Like everyone else I hate black boxes but to see inside the black box will probably require tradeoffs in terms of privacy and data security.

Personally I think that the EQI is poorly designed and a whopping waste of money for school funding. Any OIA is never going to get past privacy concerns so the questions should be more about:

what tools are being used for database management (is it simplistic like Excel or is there a dedicated DB)
how many FTE's are being employed for EQI (is this now an industry in itself and too much administrative time wasted on the process rather than the outcome)
what is the audit process for the data, how often is it audited and what measures are employed to ensure data quality, data integrity and process integrity (is it GIGO or more robust)
what level of autocorrelation occurs between the 37 variables (can the same outcome be derived with fewer variables with little compromise)

The current system for funding for enrolments isnt great but neither is the system that Act proposes (funding for attendance) nor punishing students/parents for non-attendance. Neither is particularly well thought out. I'd propose a mixed-funding model - with a baseline of funding for enrolment and and at-risk funding component for attendance (whereby the school becomes incentivised to ensure the children attend or there are funding consequences). Theres still flex in the system for sickness, akin to sick leave in employment, but multiple days absence with no doctors certificate means a reduction in per-student attendance funding.

AliExpress

Shop now on AliExpress (affiliate link).

deadlyllama

1283 posts

Uber Geek
+1 received by user: 476

Trusted

#3204040 7-Mar-2024 10:56

ockel:

Here is a list of the variables https://assets.education.govt.nz/public/Documents/our-work/changes-in-education/Equity-Index-variables-fact-sheet-Aug-2022.pdf

While I'm not a huge fan of the EQI it cant be worse than the overly simplistic decile system it replaced. However it has swung to the complete other end of the spectrum with 37 variables per child for 517,000 primary school children and 305,000 secondary school children. Thats a huge amount of data with zero visibility of the integrity of the data. This is now a data analysts dream for collecting, collating and analysing. And managements worst nightmare for data quality.

What legislation/policy allows the Ministry of Education to find out our family income over our childrens' lifetimes, in sufficient detail to split out wages & salary vs self employed income? Surely that's a massive privacy issue!

deadlyllama

1283 posts

Uber Geek
+1 received by user: 476

Trusted

#3204045 7-Mar-2024 11:06

deadlyllama:

What legislation/policy allows the Ministry of Education to find out our family income over our childrens' lifetimes, in sufficient detail to split out wages & salary vs self employed income? Surely that's a massive privacy issue!

Answering my own question

Statistics NZ does a final check to ensure all data being released meets IDI confidentiality rules before
releasing the data to the Ministry of Education. These final school-level EQI numbers are the only information
the Ministry sees and uses.

So it must be that MinEdu passes a list of parents * schools over to Stats NZ and they get back "at school x, n% of parents are in income band a, m% in income band b, ..."

I guess that's OK then.

MikeAqua

8024 posts

Uber Geek
+1 received by user: 3817

#3204102 7-Mar-2024 14:08

I'd be interested to see how it lines up against deciles. Looking at my old primary school ... it was in decile two. If I sort the from largest to smallest equity score and divide into deciles ... it's in decile two! (I know, I know n=1). I guess this reflects that you can change the scoring system, but that doesn't change community fundamentals.

We have gone from a 1 to 10 integer score to a 600 to 300 integer score. I'm making an assumption there about the possible range. I'm not sure we have gained any clarity, in fact we may have lost it. Perhaps that was the intent.

Mike

cruxis

512 posts

Ultimate Geek
+1 received by user: 346

#3204162 7-Mar-2024 15:46

deadlyllama:

Answering my own question

Statistics NZ does a final check to ensure all data being released meets IDI confidentiality rules before
releasing the data to the Ministry of Education. These final school-level EQI numbers are the only information
the Ministry sees and uses.

So it must be that MinEdu passes a list of parents * schools over to Stats NZ and they get back "at school x, n% of parents are in income band a, m% in income band b, ..."

I guess that's OK then.

That is why during the last census "some" might lied about household income and being part of certain ethnic group. To game the system so their local school gets more funding. Overheard it during school gate talk.😁

ockel

2031 posts

Uber Geek
+1 received by user: 545

#3204215 7-Mar-2024 18:41

MikeAqua:

I'd be interested to see how it lines up against deciles. Looking at my old primary school ... it was in decile two. If I sort the from largest to smallest equity score and divide into deciles ... it's in decile two! (I know, I know n=1). I guess this reflects that you can change the scoring system, but that doesn't change community fundamentals.

We have gone from a 1 to 10 integer score to a 600 to 300 integer score. I'm making an assumption there about the possible range. I'm not sure we have gained any clarity, in fact we may have lost it. Perhaps that was the intent.

But we've probably employed a shed load of people to collect, collate and analyse the data. For little apparent change. Yay for public service jobs that add zero value.

New World

Shop on-line at New World now for your groceries (affiliate link).

zenourn

281 posts

Ultimate Geek
+1 received by user: 168

ID Verified

Trusted

#3204337 8-Mar-2024 07:41

I work in the Integrated Data Infrastructure lab so have a bit of understanding of how this is calculated. See these reports:

https://www.educationcounts.govt.nz/__data/assets/pdf_file/0003/196005/Equity-Index-Technical-Report-Final.pdf

https://fyi.org.nz/request/21253/response/80708/attach/4/1301098%20Response%20and%20Appendix.pdf

Overall the statistical methodology for calculating this metric is reasonable. I however find this whole funding system frustrating. I'm the board chairperson of a low equity-index school and setting the budget can be a struggle with the reduced money we receive from the government. I get rather jealous when seeing what some other schools get.