I was listening to a podcast of a BBC interview with Ian “Supercrunchers” Ayres. Supercrunchers is a book which illustrates all the ways in which the ‘new econometrics’ or ‘social stats’ is revolutionising – well lets not get carried away – improving the judgement of all sorts of people in making important decisions in many walks of life.

Anyway, and I’m sorry I can’t find the podcast – it was late 2007 on the BBC world service – he observed that stats should be taught in schools. Generally probability is taught in schools but very little stats. As Ayres said, stats is hugely useful in life. A good exposure to stats in school would be useful throughout life. Calculus is a wonderful thing. Its elegance still excites me. But it’s of much less general use in later life – unless you turn out to be a calculus using adult.

I’m not suggesting doing away with the rest of the curriculum, or skewing it entirely towards direct usefulness in later life, but there’s nothing wrong with pushing it towards usefulness. I remember doing oodles of stuff that was not particularly elegant, not a particularly good education in mathematics, when I was doing it. Trigonometry, more calculus than we should have. We should have done some stats. And if we wanted to be ‘purer’ about it, we should have done more set theory and less trigonometry.

Reminds me of some of my views on the use (or rather, very tepid use) of spreadsheets in schools.

I agree. Understanding stats just a little might put an end, at least to some degree, to the bogus stats that advertising agencies make up. Like “more doctors prefer lux than any other soap”. Which means exactly what? That doctors are soap? Or “95 fat-free” actually means “contains a full 5 fat”. What does it actually mean to say “three times less fat than other brands”?

Teaching some stats in school would ensure that people could at least ask intelligent questions about the baloney that is paraded as “statistics” in an attempt to fool them into buying things and politicians.

The trouble is that a lot of people are frightened of statistics as a subject. They think it is confusing. I know people who have decided not to do a uni course just because it included stats as one of the subjects in first year.

But we all use statistics all the time in everyday life. We do correlations in our heads intuitively so it shouldn’t be too hard to make it a relevant and even enjoyable subject at school.

Btw, what is “stastistics”? 10% more than “statistics”.

I agree, wholeheartedly, on both stats and spreadsheets. Only the mathematically inclined and ambitious learn calculus any more these days, since it is largely optional (but boosts your score, since it is hard, after all).

I think you are too optimistic — its already hard enough to teach stats to university students who arn’t in maths-style courses, and those that are forced to learn it often complain constantly about it, hence it has been scaled back in many majors in universities where it used to be more predomoninant (students don’t see the need to learn such things). It is also being scaled back by default in high schools in some states where some very simple stuff is taught, as its generally in maths courses which students take less and less.

Unfortunately statistics and probability are areas that the human brain is just not well-wired to comprehend – or, at least, it’s well wired to perform intuitive approximations that might have made sense in our evolutionary past, but are often hopelessly wrong. The stats class that was part of VCE back in the early 90’s was actually pretty good, but to this day I still find myself making careless assumptions about statistical information where I should know better.

A good grasp of basic statistics and probability, and at the very least, a thorough recognition that intuition is often wrong, seems to me an essential part of basic critical thinking skills. And it’s critical thinking skills in general that often get short shrift in educational programs, which is surely one reason that advertisers and politicians and various other demagogues get away with sounding convincing while uttering various half-truths and logical fallacies.

I agree with the principle – people should do more stats. Picking up on Conrad and NPOVs point, it is very difficult to teach stats, and even harder to do it well.

(I was very good at trig and geometry which boosted my high school maths marks but have never used those skills and am totally useless at pool.)

Doing first year psych at Uni was a laugh. It was for both science and arts students. I’d had plenty of exposure to stats (I think we did about twelve weeks of it) in Maths B in form 6, and most of the science undergrads dealt with some very basic stats quite easily. But the arts students struggled so hard, it was painful to watch.

Nicholas is correct on the over-emphasis on Calculus and de-emphasis of Stats at school. Sinclair is right that it is very difficult to teach Stats. Even more, Roger has nailed a long-time bugbear of mine and that is the abuse of stats by innumerate journalists, who thus are easy prey for PR releases. Follow any MSM reporting (outside the business pages) on issues relating to health, education, etc. and marvel at the innumerate nonsense they print. For instance, the whole debate over state funding of private schools never ceases to amaze

I took two second year university Stats courses. The first a standard 2nd Year Psychology “Research Methods and Statistics” course, the second “Theory of Statistics” in the Maths department. In Maths, by Week 5 we were elbow deep in double integrals and hypergeometric distributions. In Psychology by Week 5 we were up to means, medians, and modes.

The Maths course was even more deathly boring than the Psych course. If you are a Maths major (as I was) Multivariate Calculus, Complex Analysis, Algebra, etc. are extremely stimulating and creative, but Stats is very hard to make attractive. And in order to get to a level of competence so that the non-Math student has enough confidence to actually enjoy critiquing arguments/research that are quantitatively/statistically based takes at the very least a one semester uni. stats course. Trying telling that to the “Misogyny in Indonesian washing powder commercials” set!

Trouble is we all need to become much more numerate and trained in Statistics as data production and analysis have exploded over the past decade and not going to slow down soon. Anybody who claims they are taking part in “Critical Thinking” and yet has no training in Stats is kidding themselves.

Stats is basically boring. What is interesting is what you do with it. Personally I thought stats to be deadly dull until I had some data that I needed to describe.

Project based learning?

It’s easy to understand why stats have a reputation for being boring. On the other hand, I have rarely been so excited and worried as I was by the stats dealt up by Possum and the Poll Bludger in the lead-up to the election last year. Every point, every slight narrowing, every reassurance, every nuance of meaning and every projection – everything was read and devoured with both hope and trepidation. Possum even made a little roller-coaster video about it that was almost as exciting as the real thing. So project-based makes sense, or at least “something-that actually-matters-to-you”-based.

Arthur

As I said, the problem is to get to that point where you can really appreciate and use the power of stats you have to take a one semester stats course at uni. If you don’t it’s hard to get beyond first base.

I did ‘statistics and probability’ for a semester in high school. The subject was ‘Maths I’. Does this still exist? Or has it vanished? I did no calculus at school and have a sneaking feeling it was possible to choose between statistics and calculus, and I chose the former. This is in Qld in the late 80s… I don’t recall ever doing calculus at school, to be honest.

I then did more statistics at university – I remember one of the subjects being pretty easy (along the lines John describes above), but the follow-up being quite, ahem, challenging.

I then had to do a year’s worth of financial accounting and financial management at law school. I have no idea why, and suspect UQ law school may have scrapped the requirement. I’m glad I did it though – I now know when I’m being sold a pup, at least. Do any other law school around the country do this?

My recollection of university level statistics is that it required a reasonably sophisticated understanding of calculus (particularly when dealing with continuous data), so I don’t believe you can teach one without the other. Anyway, let’s face it: mathematics is hard and (without intending to denigrate them) most people just aren’t bright enough to do it.

I remember at high school being told by the bloke who taught us Euclidean geometry that it had no practical use (which is really a bit of an overstatement – it’s at least a useful approximation to the shape of the world). I finally understood what he meant when I encountered other formal systems, taught as such, at university. I think too much is made of whether or not something is “useful”. Some things are worth learning for their own sakes, and a lot of mathematics falls into that category, as does English literature.