Schools names starting with I work!

The New Zealand Herald published an article today with the following headline:

National Standards shock: Big classes work

In it, they claim
[box]The Herald on Sunday has conducted a comprehensive survey of schools’ national standards results, before the Ministry of Education publishes them this week.[/box]
A comprehensive survey. Crikey, it must be true then.

I wonder what other gems are buried in this data? I’ve done my own comprehensive study and unearthed a pretty shocking revelation. The first letter of the school’s name has some bearing on performance. In the graph below you can clearly see that if you send your child to a school starting with I, they will get the best outcomes.

I expect to see schools changing their names pretty quickly. Oh, and also increasing class sizes because the Herald says that works too.

If you don’t understand what I’m saying, I’ll spell it out. The claim the Herald is making makes no comparison based on location, socio-economic situation, timing, teacher ability, or any other possible influence. It simply compares class sizes across the country and says “big is good”. Now, if the Herald was reporting a controlled, well designed study (like the Tennessee study for example), I’d be amazed by the outcome and it would change my (admittedly minimal) understanding of eductation.

Think about it. In New Zealand small classes are likely to skew toward country schools. These schools have different access to resources (including teachers), and will have mixed year levels. Now, if the Herald article had compared class sizes in the same locations (with similar teachers and kids) and found bigger is better, or even said “small country schools do worse than big city schools”, that would be interesting.

But it’s not interesting research. It’s a lazy article reporting a statistically irrelevant fact for shock value.

8 Replies to “Schools names starting with I work!”

  1. Thanks Ben. This is exactly the issue we were afraid of, and one that the press wouldn’t ever acknowledge before the fact.

    The other issue is our naturally compatible human responses to simple patterns and drama. People will make judgements on poor data. As a result, schools now have so much more to manage than just education.
    Expect it to change class practice, and that won’t always be good.

  2. The most worrying thing isn’t the research in and of itself (although you’re right that it connects draws conclusions which are WAY too spurious), but that the bulk of readers lack the good judgement to put the effort into thinking for themselves and verifying the interpretation of “hard data”.

    Much easier to just accept the results as shown because, well – it’s the Herald and they’re a respectable news source, right?

  3. To then see Hekia Parata use ratty data and erroneously-drawn conclusions as a base for “vindication” on her “cut-funding-enlarge-classes” disaster is so enraging.

    Hair’s gonna start falling out all over the statistically-minded nation, yo.

  4. Yes, that article made me mad, not so much because of the ‘correlation = causation’ analysis, newspapers do that kind of lazy reporting all the time, but because it was also angled toward a particular political viewpoint. If they did one or the other of those things, then I could live with it, but to do both smells like reprinting of propaganda to me. Lazy, lazy, lazy.

  5. Popular schools top their schools up to achieve their maximum class sizes, through a ballot if necessary. In fact popular schools can become overcrowded if their zone contains more students than they can really accommodate

    Popular schools are also overwhelmingly high decile, not low decile.

    We therefore end up with a correlation between class size and decile, but it is DECILE that should be seen as the link to achievement.

    You can’t investigate a variable like class size without keeping other significant variables constant! Statistics 101!

    I would think that if you analyse by Decile, you will find smaller class sizes do better than large – but even then we must take care. Very small classes may be mixed in years, and may have insufficient cohort to attract specialised teaching or good synergy among students.

    But it beggars belief that in an era where PERSONALISED teaching for students is being recognised as a success factor, that the NZ Herald would prominently tell Hekia Parata what she wants to be told in this fashion. If the article were to be believed, not only would we say 40 in a class was better than 25, but also 200 is better than 40! The article is absurd.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.