February 22, 2014

Internal and external

There’s an interesting story in the Herald with interactive graphics comparing internal and external NCEA assessments for different subjects, levels, and decile of schools, over time.  The main thing I might change about the graphic is to display over deciles rather than over years, since that’s where the action is.

The general picture is fairly consistent: in low-decile schools, the students get substantially better grades on internal assessment than external. The difference is progressively smaller as you move up the decile scale, in some cases vanishing.  Interpreting the results is more difficult.

The lead says that students do better away from the pressure of exams, which is one explanation. Another, given by Professor Carnegie from VUW, is that the internal assessment is not very reliable. There are many alternatives views given in the story, and even some who says the differences over decile are reasonable and appropriate.

 

avatar

Thomas Lumley (@tslumley) is Professor of Biostatistics at the University of Auckland. His research interests include semiparametric models, survey sampling, statistical computing, foundations of statistics, and whatever methodological problems his medical collaborators come up with. He also blogs at Biased and Inefficient See all posts by Thomas Lumley »

Comments

  • avatar
    megan pledger

    I suspect there is also a problem in that the kids sitting the externals in decile 1 schools are different to the kids sitting externals in decile 10 schools.

    Decile 10 kids start doing L1 credits in year 10 (or earlier) so that by the time the externals come around in year 11 they already have enough credits and can flag the externals. The kids who take them are kids who like academics or kids who need the particular credits to advance e.g. science maths.

    Decile 1 kids don’t start L1 credits until year 11 which means their credit hunt is more time crunched so they pretty much have to sit the externals to get all the credits they need. The time crunch also effects their ability to do as well in the exams as the kids who have generated extra study time by starting L1 earlier.

    10 years ago

  • avatar
    Ash

    In general, it is much easier to get an Achieved (minimal to pass) in an internal assessment than in an external. For one thing, resits/resubmissions are often possible. This is especially the case for science and maths, where assessments only require one class period.

    It is not then surprising that the disparity between low and high decile schools in pass-rates is much smaller for the internals than in the externals.

    10 years ago

  • avatar

    I did the interactive and I did make a version which displayed deciles over years. Though it made the decile comparison more dramatic, it obscured the changes over time in deciles from external to internal which we felt was the more important thing to show story-wise. Also, though the decile achievement comparison is dramatic, it is perhaps more relevant to see what has happened within deciles over the last five year as NCEA has undergone a re-alignment in standards in 2010-11.

    10 years ago

  • avatar
    Claire Laverty

    High internal results do not necessarily mean they are not reliable. When you have a lower ability cohort, you generally design a course where they attempt less standards and are given a long time on each one.

    10 years ago

    • avatar
      Thomas Lumley

      Claire,

      The story does make those sorts of points. The lead is a bit dubious but the story is good.

      10 years ago

  • avatar

    Graham Jenson has done a better visualisation comparing deciles which might of interest to people commenting on this article.

    Here’s the link –

    http://maori.geek.nz/post/the_difference_between_rich_and_poor_schools_in_new_zealand

    10 years ago

    • avatar
      megan pledger

      He got it a bit wrong. The schools are put into 10 deciles so there are equal numbers of schools in each band *but* there are not equal number of kids in each band.

      Decile 10 states schools are very popular and Decile 1 schools are not so popular – so there are many more kids at Decile 10 state schools than at Decile 1 schools.

      The aveage dec10 state school in Wellington has 1300 students and the average dec 1/2 state school has 600 (based on 2008 roll though).

      But on the other side – some decile 10 schools have kids who IB or CE exams instead of NCEA.

      10 years ago

      • avatar

        Hey, I am the one who made the visualisation, and this error was pointed out to me by a friend as soon as I published it (D’oh).

        Although the Visualisation is technically correct, in that the total children at NCEA 3 in Decile 1 is lower than in Decile 10, it is misleading because relatively there are more children in Decile 10.

        I found some great data over at http://www.educationcounts.govt.nz/statistics/maori_education/schooling/6028. I tried too normalize the totals in my visualisation, but I get weird results where Decile 10 has less students than Decile 6 (still more exellences) and Decile 9 has the most students in NCEA 3.

        I am not sure if I will create a better version, probably not as creating it, although fun, was very time consuming. But, given it is an open dataset, you could give it a try, and tell me the results :)

        10 years ago