Why are people so careless of statistics? Understand that I think people should not smoke because it is clearly bad for you and it is disgusting. So is misusing statistics.
In this report by BBC News there are 101,000 nurses in a study. Some have sudden cardiac deaths. Of these 75 are smokers, 148 are past smokers and 128 are non-smokers. While the article says that 75 deaths + 148 deaths + 128 deaths is 315 total deaths, it is in fact 351. So BBC needs to employ more proofreaders and the overall incidence of sudden cardiac deaths is 0.003 or about 3 in 1000. Not really a lot.
The article reports that smoking doubles the chance of sudden cardiac deaths. However it is impossible to know what this means since we are not told how many smokers, past smokers and non-smokers there were altogether.
The article also says, “For every five years of continued smoking, the risk went up by 8%.” What could they possibly mean by this? The total incidence is only about 0.3%. Eight percent is almost 27 times that.
…and this vagary helps to undermine the valid message that smoking is a bad idea.
When someone tells you that x doubles the chance of y they have told you nothing. There is a huge difference between odds doubling from 1 in 100 to 2 in 100, and 1 in 2 to 2 in 2, or 3 in 1000 to we cannot know what.