Shoddy survey work

Posted by on Dec 22, 2008 in Uncategorized | One Comment


Ben "Bad Science" is at it again

Doing that thing where he catches naughty folk who should (OK, do) know better, doing bad stats (this time with regard to uncontextualised survey work)

In my experience, this is not an unusual misuse of data.

You'll see similar stuff every day of the week in customer satisfaction work, brand or ad tracking studies and concept testing.

Whatever you think about the underlying validity of survey work (regular readers will know my pov all too well), this headline-grabbing-one-data-point-uncontextualised-squeezing-out-significance-because-what-else-can-we-do-the-client's-paid-for-it-and-needs-something research practice is just poor.

Whether you sit at the ARF or the MRS, you should really be denouncing the kind of thing this piece highlights. It's p****** in the water-supply, isn't it?

There's no real excuse for this kind of stuff going unchallenged, is there?

1 Comment

  1. Simon Kendrick
    December 22, 2008

    I completely agree, particularly with the trade bodies being more active in calling the miscreants out.
    Most examples I see belong to the “cheap pr” school of research – with ad tracking and cust sat, the miconstructions and spin are kept internal only 🙂
    I flagged up a previous example from Canon at but one could look at Brand Republic pretty much any day and find a new example in a couple of minutes