Wednesday, 14 September 2011

The truth about statistics


A recent episode of the BBC science programme Bang Goes the Theory included a feature which appeared to be, at first glance, yet another of the increasingly frequent media attacks on red meat consumption because of alleged health risks.

One of the presenters was handing out bacon sandwiches on a market stall, accompanied by a sign reading ‘bacon increases risk of bowel cancer by 20%’.

He then produced a second plate of bacon sandwiches and another sign which read ‘bacon increases the risk of bowel cancer from 5% to 6%’.

As he tried to give away the sandwiches to passers by, the vast majority opted to take them from the second plate, afraid of their massively increased risk of bowel cancer if they picked one from the first plate.

Of course, as he eventually pointed out, the bacon sandwiches on both plates were exactly the same, as was the increased risk of bowel cancer associated with eating them. Both signs gave the same message, but expressing it in a different way completely changed the public perception.

Whether the bowel cancer statistic is correct is a completely separate debate which we have already addressed in the facts about red meat and health, however the programme is a great illustration of how statistics can be used to influence public opinion.

For the meat industry, which is often targeted by single issue groups who use statistics as ammunition to make people question their meat eating habit, TV programmes like this which help make the public more statistically-savvy can be no bad thing.