November 2017
« May    

97.3% of all “97.3% of all statistics are made up on the spot” quips are tired old clichés

Note: I began writing this entry some time ago and didn’t manage to get round to finishing it – until now. So although the Times article in question is a little out of date, I still feel that the point is worth making.

Another newspaper column appears to have joined with Ben Goldacre’s Guardian column in the worthy cause of systematically and entertainingly (but not overly patronisingly) critiquing some of the dodgy survey data, statistical analyses and scientific reporting found in mainstream newspapers and magazines. In principle, this is what I regard as ostensibly A Good Thing. Long may this practice be spread further into more national newspapers, hopefully covering a much broader audience.

David Aaronovitch’s latest column in the Times covers dodgy data on a variety of issues: firstly the Turin shroud, then on perceptions of aggressive children (AKA da yoof of today), and finally a “thin-air” estimate of the number of sex workers active in the UK that the originator never intended to be definitive, which has become exactly that by being bandied about the media with reckless abandon. It’s really good stuff. Brief, concise, solid critique – entertaining, and never sounding like a crazy conspiracy theorist. I approve. However, one observation that gives me cause for concern is the feedback from readers. Allow me to explain.

Read the rest of this entry