The big eye-opener for me has been virtual posters. I wasn’t sure about them when I first heard of them, but for this they are superb. A very cursory scan through has already pointed me to some marvellous work at stats4stem.org; told me about “Courseload” (an enhanced e-book reader I think) and what looks like a very portable app-like approach to guided learning using something called Vizi and currently in beta. I’ve also been very encouraged by some work on PRS (personal response systems – audience response clickers or whatever we should call them). There is a poster describing a nice (non-randomised control) study comparing PRS taught students with non-PRS taught students which allows the authors to compare attainment and attitudes over two semesters. I’m still making a mess of attending the main presentations (due to my inabilities to convert time zones) but this really is a marvellous way to keep up to date and even get inspired.

According to Arthur Benjamin, current maths education aims to deliver pupils in a suitably advanced class where they can learn calculus. He suggests that in today’s world we need to better understand risk and reward. We also need to be able to understand data. And given that all these are subject to randomness, we need to be able to understand that concept. Hence all our maths education should be restructured in order to deliver them into that class with all the necessary skills and prerequisites. To my way of thinking, rather than being the pinnacle (if indeed it does occupy that position), some amount of calculus is needed on the way to the top of maths mountain.

Given that I agree with Bruno de Finetti – “there is no such thing as probability” – that might even be quite an enticing pinnacle for those who decide they don’t like maths (“it can’t be that difficult a class, it doesn’t exist”). Fun and jollity aside, there is a serious point lurking in this. If understanding uncertainty is at the pinnacle of maths education, it ceases to be important to cram loads of statistical “stuff” all through the curriculum. Maybe at elementary levels it becomes more important to fully understand number systems and operations than to meet “data” with all the extra conceptual processing required.

There is though a tremendous amount to take from this talk. For me, the main impression was the problem of silos in education, very powerfully illustrated with the “school as factory” illustration – indeed the very fact that an obviously gifted artist was bored by the one aspect of maths that should have most engaged them perhaps in itself highlights the problem. IMHO, one key point for statistical education is that I reckon rather a lot of the goals of statistical literacy are addressed in non mathematical subjects. History pupils are taught for example to criticise sources who asked the question, whom did they ask, what exactly did they ask, why did they ask it and so on). The very skill, with a small supplement of entirely data/statistical thinking, we need applied to the divergent thinking study cited by Sir Ken Robinson in his talk.