The Right tilts at class-size windmills

Written By: - Date published: 8:48 am, September 24th, 2012 - 10 comments
Categories: education, making shit up - Tags:

The Righties are wetting themselves because the national standards supposedly show higher results in larger classes.

Of course, you can’t compare data between schools with any validity – “they are not moderated, so one school’s “well below” may be another’s “at” or “above”” writes the newspaper that helpfully provided a tool to let you compare schools.

And Danyl at Dimpost has shown that a third of the supposed class size link is down to special schools. Not that it matters – because there’s no validity in comparing the data across schools.

The rest of the difference is probably down to deciles. Danyl has also shown that national standard results and decile are closely and unsurprisingly correlated. Not that it matters because making claims about national achievement based on unmoderated results from individual schools is nonsense.

So, what have we learned? Nothing. Certainly nothing about the impact of class sizes, anyway. But maybe a little more about the Rights’ willful blindness.

lprent: Added links. And while I’m on the post I’d have to say that Jonathon Milne at the Herald on Sunday needs to be sent back to school for both some basic remedial stats. Due to the lack of detail in his article to evaluate his useless maths, some remedial journalism looks in order as well.

10 comments on “The Right tilts at class-size windmills ”

  1. Bunji 1

    Danyl’s analysis is excellent.
    A good chunk of the correlation is special schools having small classes and unsurprisingly poor achievement. If you take out the decile correlation it’s almost all gone. And I’m sure that the last bit will be rural schools that are very small – you get 1 teacher teaching 6 age-groups and it’s going to be a struggle to aim at them all…
    [Edit: and as noted in the OP, that’s remembering the data is dodgy & unmoderated as well…]

    The point of view from England (where they’re slowly scrapping Standards as a failed experiment), is quite interesting (disappointed in our direction…)

  2. Ed 2

    I wonder how consistent other statistics are as well. Certainly many schools have ‘special needs” children in regular classes – often with a one on one assistant for say 3 hours a day. Are these classed as “staff”? Often such children will be disruptive, especially when no additional aid is given. Are private schools under the same requirement to accept such children? Some schools have many children for whom English is not their home language – that will have particular effect on (English) reading ability in the early years. that is not to say that such schools have poor results, in fact some parents will welcome the opportunity for their children to be taught in a school with such diversity, but such ‘education’ is perhaps not measured in the “National” standards. Transient children may also be a challenge for schools in some areas – when one child represents 3 children that have spent a third of a year each in the school which one do the statistics represent?

    National standards appear to be a simplistic ‘solution’ to a complex issue that is not a problem. Not one child will be better educated as a result of National’s imposition of this regime. Some schools will however have improved their reporting to parents; most appear to be saying it has made no difference to the quality of reporting, but they may now use slightly different words – and I am not aware of any survey that tests whether parents are better informed.

  3. Blue 3

    Isn’t it great how one of our major daily papers makes a totally unsubstantiated link between class size and achievement based on the raw data alone?

    Sadly, there’s probably not even one moron at that paper who realises how intellectually bankrupt that is.

    • Draco T Bastard 3.1

      There is now but I doubt if we’ll get an apology from the journalists for their misleading reporting.

  4. prism 4

    Attitudes to education by present pollies, the Terminators. Don’t know, don’t want to know. Class sizes are today’s smoke screens for wasted expenditure on useless education that doesn’t excite thought or prepare the young for non-existent jobs in our country so poorly run for its citizens. What pollies learned at school is how to selectively use education to advance their own desires and prejudices. The pollies education interest is in increasing the returns from education business from the payments of foreign students.

    But now because governments don’t value a wide and general education appealing to all children and moulded to advance their own potentials the pollies are just using the education system as a giant factory concentrating on ways to reduce public school spending by economic means with savings in a volume situation. That conforms to the sort of country government is directing with their policies which offers little employment to people who make things, but has many accountants, lawyers, financial and insurance and real estate salespeople which circulate present money, clipping the ticket as it passes through each group’s hands.

    New money is mostly being created by primary industry as it always has been. And now govmnt are trying to siphon money out of that golden goose by charging sector groups for much of the biosecurity controls and technical help that has become necessary with the incursions of pathological problems coming with the increased imports bringing stuff we should be making ourselves. Continually we find that governments are not running the country showing good husbandry of our NZ resources and encouraging and supporting new business that will provide employment for the graduating students from secondary schools.

    There was one good report this morning about vouchers encouraging new developments in technology, with two institutes in NZ taking these up and spending $4million out of $20 million budgeted. Labour commented. What they were reported as saying was just a negative about only $4 million spent. A chance for Labour to say that they would be pushing such bright and practical ideas and doing much more than playing around the edges when they were back in government. That’s not what came over.

  5. Trevor Mallard 5

    The related issue is the schools that have special classes or units. Their results will also be skewed. And they are harder to identify. Though I suppose that all that is just layering on another level of bullshit.

  6. Draco T Bastard 6

    And Keith_Ng takes the NZHerald to task over there inability to understand statistics as well.

    The closer the R^2 value is to 0, the poorer the fit. Your R^2 value is 0.137. If we take out the special schools, that goes down to 0.0738. This is what we would call a “poor” fit.

    • NickS 6.1

      Even an R^2 of 0.137 is utterly horrible, ffs if you tried to pass that off as a sign class size was a significant predictor of academic achievement you’d be laughed at. And that’s ignoring the underlying issues surrounding the raw data set vis controlling for the wide range of known variables.