More than a numbers game

The Royal Statistical Society is calling for its subject to be given greater prominence
Click to follow
The Independent Online

Statistics are a powerful force in all of our lives, as Sally Clarke knows only too well. In 1999 Mrs Clarke, a Cheshire solicitor, was convicted of killing her two baby sons on the basis of a statistical falsehood, an error challenged in court, but one so convincing and horrifying in its simplicity that the jury accepted it. Professor Sir Roy Meadow, it emerged, was plain wrong to assert as an expert witness that the chance of two children from one family dying in accidental cot deaths was one in 73 million.

Leading the complaints was the country's academy for professional statisticians, the Royal Statistical Society. Appalled at the misuse of numbers, it even wrote to the Lord Chancellor saying the conviction had no statistical basis, and it is thanks in part to the Society that Mrs Clarke was released earlier this year.

It is an extreme example, perhaps, but according to the Society, the misuse of figures at the core of it is sadly typical. Now the society is demanding changes to the school curriculum to foster a wider public understanding of the way that data are presented. The Government is running a national inquiry into the state of mathematics under the chairmanship of Professor Adrian Smith, principal of Queen Mary College, London and the Society has urged him to recommend a core, compulsory dose of statistical studies.

The use of statistical data is standard in both business and government, but according to the Society, a failure to understand them is equally widespread. Statistical justifications are everywhere, from pension calculations to hospital waiting lists and school league tables. No national newspaper can operate without batteries of bar charts, pie charts and tables. Yet it is the visual presentation of figures in particular that concerns the professional statisticians. The data are over-simplified, out-of-context, made to carry inappropriate significance - and generally misinterpreted. There is already ample evidence that many shoppers struggle to calculate their change, let alone the percentage reductions in sales.

"One of the problems that we see in society in general is that people are not sufficiently numerate," says Ivor Goddard, director general of the Society. "That really doesn't enable them to make evidence- based decisions. That's why part of our role is to try to encourage statistical literacy at all levels, to get more effective decision making. Our concerns are partly about how statistics are reported. People are being encouraged to make judgments that the data just won't bear."

The use of statistical measures and targets in public policy are a particular concern for the Society because they seem to have been introduced without taking into account the wider effect they would have on behaviour, whether in hospitals, schools or on the railways. Professor Harvey Goldstein, a fellow of the Society and a statistician at the University of London's Institute of Education, uses a critique of school league tables as a standard introduction to the subject for his students.

As a first step towards better public understanding, the Society would like to see children at the upper end of primary school familiarised with charts and graphs, and taught to question their significance. Eventually, says Gerald Goodall, the Society director of education, students should be able to read across a table of data and see different trends in different columns.

In one sense statistics has been a triumph in recent years. The serious study of statistics began in the early 19th century as a form of sociology, an attempt to map society in numbers. Now, the field has expanded so rapidly it finds itself at the heart of how we address the world. Higher level statistics has been essential to our defence against the spread of SARS, bird flu and BSE. It is a key component of weather forecasting. Even at school, the presentation of data has gained importance in apparently unrelated subjects such as biology and geography.

Part of this is down to advances in computing and, in particular, the advent of the pocket calculator which liberated the subject from log books or, at the higher end of things, the mainframe. Maths, though, is struggling as a subject, both at school and university or college, which is why the Government commissioned the Smith inquiry. In common with other "hard" subjects such as physics, it appears to be losing out to the humanities at A-level, while university departments are scaling down or closing altogether.

"We like to think it's a very exciting discipline and a very practical one with lots of opportunities. The problem is that we're not recruiting sufficient people to maths overall," says Ivor Goddard. A shortage of graduates meanwhile means a shortage of talented maths teachers.

Yet even with mathematics, statistics is losing out. When backs are to the wall, it seems that university maths departments will defend the theoretical, "pure" section of the field at the expense of related "applied" branches like statistics and mechanics. So university statistics departments are more vulnerable than others. University maths departments also seem to be behind a change to A-level maths which, from this September, sees the temporary abolition of a GCE statistics qualification, and a shift in emphasis which, in the view of the Society, favours pure rather than applied maths.

This, say the statisticians, is a disservice to A-level maths students because most will never darken the doors of a university maths department. It is also to ignore the potential of subject which often represents maths in its most practical form - a view also held by other learned societies.

"The Royal Physical Society in its presentation to Adrian Smith's inquiry pointed out that statistics could be a very motivating subject for people," says Professor Goldstein, who is also chair of the Society education strategy committee. "We would make out a case for a statistics core curriculum and the introduction of statistics to as many people as possible. This is partly for its own sake, and partly as a way into mathematics."

There has been a well-documented boom in the popular understanding of science but, so far, statistics has not managed to make itself a part of it. That, says the Society, needs to change.

Comments