Bad Statistics

I am sure that many of you will have seen Kellog’s advert proudly claiming that their cereals can give children a much better start to the day and help them perform well in school.

I don’t know how many of you have been able to read the small print which states that the children’s alertness is measured by the parents’ opinions and compared to when the children do not have anything for breakfast!

I can’t wait for a similar advert from petrol companies – Our petrol makes your car run better! (at least better than when you don’t put any fuel in!)

There was another example earlier this year when a cosmetic company based its advertising on 86% of women finding that the product worked. The total sample was 46 women!

As marketing campaigns and claims get more sophisticated it is important for us to maintain a sceptical attitude to these claims.



3 comments


    Kevin

    September 14, 2005

    I find this kind of thing even more pronounced in medical ‘research’ claims. Check out some of the MMR stuff, with sample sizes as small as eleven..

    markbrake

    September 14, 2005

    Makes you wonder if such bad stats shouldn’t be outlawed in some
    way.

    And one would have thought that the “ASA”:http://www.asa.org.uk/asa/ would have had guidelines on the use of statistical data.

    Clearly not. Or, at least, clearly not ones that have an ounce of efficacy

    markbanerji

    September 14, 2005

    “Lies, damn lies and statistics.”:http://www.bbc.co.uk/dna/h2g2/A1091350

Leave a comment


Name

Email(will not be published)

Website

Your comment

Designed by Forte Web Solutions