Wednesday, 29 October 2008

The fallacy of citing Google hits

My old friend David Hughes, onetime Daily Mail political editor, has a rather uncharitable piece about John Prescott in today's Daily Telegraph, following his BBC documentary on class on Monday night. I have a higher opinion of Prescott than Hughes, though I also have no doubt that he has a chip on his shoulder about class, but I am more concerned with the error he makes at the start of his piece - it is one increasingly made by columnists seeking to give greater authority to their pieces while appearing in tune with the times. Hughes tells us:
If you Google "John Prescott" and "class", you will get 199,000 results. That's quite an obsession. Showing a hithero unimagined flair for enterprise, Prescott has now decided to turn it into an industry.
In fact, I was told that there were 110,000 hits when I did so. But never mind. When I tried "David Hughes" and "class", I got 51,000 hits. When I did the same for "David Cameron", the total soared to 373,000.

But that is less important than the fact that many of the so-called hits are either repetitious or of little relevance. Indeed, after 360-odd hits in the Prescott list, you are told that the rest have been omitted to avoid repetition. Even then, the hits include details about an American football enthusiast called John Prescott. It used to be said there were lies, damned lies and statistics. Now it seems we must add 'Google results' to the list. Columnists beware.

No comments: