If you remember back a month or two ago, there was a lot of fuss about Australia’s results in the latest results published by the Program for International Student Assessment (PISA). Don’t think that this is a new topic, though – the organisation publishes a new set of figures at the end of every year, so the story comes up again and again. It’s an easy piece to keep writing, really – if the results are positive, cue nationalistic pride in our education system along with a few articles about how this shouldn’t fool us into forgetting the disadvantaged students who are doing it tough, and if the results are negative, bash the Australian education model vigorously.
In all the articles I’ve come across on the topic, though, none has mentioned the first thing that occurred to me – everything can be statistical noise and mean reversion. Noise -> confounding factors influence the results, so the change in score can’t be fully attributable to a decline in education quality. Mean reversion -> even with a long-term upwards trend in educational quality, sometimes there are going to be particularly good and particularly bad years for test results. And mean reversion predicts that a really good year will be followed by a less-good year – same for bad years.
So any dip in national test results, rather than indicating a sudden erosion of educational standards or practices, is really just a natural consequence of our string of good years in the past. Instead of fretting about the difference between this year’s result and previous year’s, the level-headed commentator should take heart from Australia’s above-average results (we are easily above the OECD mean for each category) and concentrate on more important challenges such as how technology is transforming students’ future workplaces.