Friday, May 19, 2006

The Fine art of Googlometry

Published on www.crikey.com.au May 19th, 2006

In the pages of Crikey, Google seems to be morphing from a search engine which finds information into a pseudo-scientific survey tool. It's being used as a measure of the importance, popularity and cultural significance of various phenomena. Such misuse is no doubt tempting as Google is free, instantaneous and its sheer ubiquity gives it a certain superficial authority. But it's misleading to use Google in this way, and it's certainly no substitute for statistical analysis.

In the recent past Crikey has published a number of articles in which the author uses the number of hits in Google to somehow substantiate his case. On April 27, Crikey published this sentence. “If you type the words ‘military incompetence' into Google, you'll find 46,500 entries.” If you type in the phrase today you get 36,700 hits. That doesn't mean that militaries world wide have become more competent. Also if you Google "military victory" you get over 600,000 hits. Does that mean that militaries are more victorious than they are incompetent? The message is, 46,500 hits on Google for “military incompetence” tells us nothing. On 11 May Crikey claimed “There are 29,800 Google entries when you type in the words ‘mateship' and ‘John Howard'”. This fact was used to support the claim that the Australian Prime Minister John Howard uses the word “mateship” a lot. What this really tells us is that there are 29,800 webpages in which feature both the word “mateship” and the phrase “John Howard”. It tells us nothing about how often PM John Howard uses the word, whether or not he uses it more or less than he used to or whether or not he uses it more or less than other politicians. Basically, it tells us nothing. It's a meaningless phrase.

We've also seen instances of Googlometry in articles on the media coverage of the Beaconsfield Mine disaster (11 May 2006) and the parochial nature of Melbourne (11 July 2005). Crikey should drop this pseudo-research and concentrate on what it does best, informed opinion.

No comments: