Skip to content

World Statistics Day

16th Oct 2015

It’s ‘World Statistics Day’ on October 20th!

OK, it’s not quite as exciting as Christmas, but it does merit a moment of reflection – at least to encourage a next generation to marvel and pursue the wonder of statistics.

As a young lecturer at Cambridge, my then Faculty made the decision to make undergraduate statistics voluntary instead of a compulsory option, which it had been up until then. At the time I thought this a shame, but later concluded that it was a disaster. The students often complained about having to do the stats, and most of the Faculty didn’t care that much about them either, except for my wonderful colleague Brendan Burchell who taught them. Almost overnight, the number of students who took statistics fell from around 120 a year to around 20. It was a pattern that was not unique to Cambridge, and did a great disservice both to the students and society.

For those who don’t love anything better than to curl up with their laptop and do regressions, let me try and explain the 101 of stats and why they are so cool. The most basic distinction in stats is between ‘descriptive’ and ‘inferential’. It is a key conceptual idea, and one missed by everyday references to statistics. A descriptive stat is essentially a statement of fact, such as three, or 50 percent, of the apples in my fruit bowl are green, and the other three are red. A population census is like this, telling us how many people live in a city, how many are female, or how many are aged under 18. Pretty useful if you need to know how many schools to build. But most stats are actually ‘inferential’: they enable us to make inferences about what is happening in the world, or how one thing relates to another. For example, a well-constructed random sample of a 1,000 people might allow us to infer how 50 million people might vote or feel about an issue. Or an analysis of a set of variables might tell us, within a certain range of confidence, that if person X thinks or does Y, then they will also probably think or do Z. That is really cool. It’s the key to unlocking a whole world of secrets and understanding.

One of the most exciting things happening in the world today is that there’s a powerful convergence of massive growth in data; governments and organisations applying experimental approaches (see for example the fantastic work of the Educational Endowment Foundation) and a growing understanding of the fuzzy, probabilistic nature of the world we live in. In the UK, we’re encouraging our civil servants to sharpen their statistical and methodological skills, not just to leave this extraordinary power to the analysts in the basement. Statistics are not just a noun, a passive thing, but a verb – a powerful process to peer into the world around us.

My son, starting at university, just told me (unprompted!) that he’s decided to do a stats module. I couldn’t be more proud, but also think it’s the right call. A lack of statistical skills will look to the next generation rather like not being able to read looked to previous generations, but often our universities themselves have let these skills atrophy. The Nuffield Foundation have recently co-funded 50 much-needed university posts in quantitative methods. Similarly, the Royal Statistical Society has been running courses to help sharpen MPs stats skills. We should do more. Whether you’re 18 or 80 – don’t leave stats to the statisticians – they’re the keys to the world! Have a look inside!