Charles on… anything that comes along

I recall a Tory election press conference - it must have been the 1992 election, because the fabulous political hack Tony Bevins was alive and raising hell - in which the Tories tried to demonstrate something or other using two pie charts, one ‘before’ and the other ‘after’ to show how wonderfully things would grow under their beneficient rule.

However, the charts were misleading, because they’d used the radius of the chart to show how things would increase. This, of course, meant that the ‘after’ chart was larger by a factor of r-squared, nicely misleading everyone. Bevins, and others, pointed this out, which rather derailed the Tories. (If anyone can remember what this occasion was, please put us all out of my meanderings.)

Anyhow, CNN has now done something nicely similar in a poll it ran to show how opinions about Ms Schiavo and whether she should be left to die ran across political parties.

Here’s the broad stats: Democrats in favour: 62 per cent. Republicans: 54 per cent. Independents: 54 per cent. (The question was “Based on what you have heard or read about the case, do you agree with the court’s decision to have the feeding tube removed?”)

But as Mediamatters points out, by using a false origin (starting at 53 per cent), CNN made it look as though those durn Democrats were wildly in favour, while Republicans and independents were, like, whatever.

And to add insult to injury, it’s a poll with a sampling error of +/- 7 per cent. So the 62 per cent of Democrats might be 55; the 54 per cent of Republicans might be 61.

Remind me, what was the point of carrying out the survey, then?

These posts might be related (the database thinks..):
• How ATM fraud almost brought down the British banking system (26 October 2005; score: 27.32%)
• "Tails are cool" (26 October 2005; score: 24.51%)
• In The Guardian: why Google bought into AOL; and review of the year (22 December 2005; score: 22.71%)