I normally groan when I hear about the establishment of yet another campaign group but Straight Statistics, launched last week, is a welcome exception.
Public confidence in official statistics is so low – and the presentation of statistics by companies, academics and the media frequently so poor – that we really need a body to protect the integrity of statistical information. Straight Statistics promises to play that role, with funding from the Nuffield Foundation.
My personal knowledge of some of the journalists and statisticians involved – particularly the director Nigel Hawkes, a former science editor and health editor of The Times – gives me confidence that Straight Statistics will succeed in its aim of restoring confidence by exposing bad and promoting good statistics. (I should add that Simon Briscoe, statistics editor of the Financial Times, is on the organisation’s board.)
The Straight Statistics website already contains good analysis of statistical deficiencies, for example of the Research Assessment Exercise (which determines university funding) and of the Home Office presentation of the national DNA database.
I suspect more of a challenge will be to find excellent statistical presentations, which others can follow. Sadly, it is easier to spot the failures than examples of really good practice.
Straight Statistics can play a public, campaigning role that is beyond the remit of the government’s Statistics Authority and the professional Royal Statistical Society.
As David Lipsey, the chairman of Straight Statistics, puts it, “the epidemic of statistical abuse is ultimately a threat to the integrity of our democracy. It has increased, is increasing, and ought to be diminished.