More Bad Science

Posted by on Sep 6, 2008 in Uncategorized | One Comment

51NPP8ZDXZL._SL500_AA240_

I've long been a fan of Dr. Ben Goldacre whose Bad Science blog and column in the Guardian continues to shed some light on some of the worst examples of contemporary…er, bad science.

Back in April 2003, he outlined his targets in a manifesto

  

"First, of course, we shall take on duff reporting: ill-informed,
credulous journalists, taking
        their favourite loonies far too
seriously, or misrepresenting good science, for the sake of a
        headline.
They are the first against the wall.

    Next we'll move on the
quacks: the creationists, the new-age healers, the fad diets. They're
        sad and they're lonely. I know that. But still they must learn.
Advertisers, with their wily ways,
    and their preposterous diagrams of
molecules in little white coats: I'll pull the trigger.

    And the same goes for the quantum spin on government science. I'm watching you all"

Today's column is a must read for marketing and marketing research professionals and anyone involved in creating or interpreting data.The story of how the media willingly distorted a study of happiness in across the UK (by focussing on the small differences between the small samples taken in different cities.

But let's be honest, it's not just newspaper reports that do this – all kinds of business folk misread survey and other datasets every single day, in every company, consultancy and media and creative agency, everywhere in the world. We squeeze out differences and significance where there is none, hide and deny the parts of the datasets that don't fit or flatter our opinions and mistake correlation for
causation.

However many times the folk who collect the data say (as in the featured study) that "the variations …are not statistically significant" – and in my experience market researchers and data folk are really not that good at ensuring their users understand the limits of the data – we users insist on misreading data for our own ends and purposes.

So do yourself a favour: buy your favourite data user a copy of Ben's book. And one for yourself. Read it. Note and mark, as my old headmaster used to say.

My copy just arrived and it is just as brilliant as I hoped.

And remember, Ben is watching you

And now so am I…

1 Comment

  1. Tom
    September 8, 2008

    You might like this service (that Goldacre was tangentially involved in setting up) called Behind The Headlines.
    It takes two stories from the daily papers about healthcare issues, then quickly analyses them to see if what the researchers/journalists said can actually be backed up by the research itself.
    Some of it is very elucidating/informative.
    http://www.nhs.uk/news/Pages/NewsIndex.aspx