Comments on: Google Cliche

By: Scottish Roundup » Blog Archive » Labour’s by-election woes and the SNP’s SFT

Sun, 25 May 2008 00:23:26 +0000

[...] Robert Sharp is fed up with journalists who use the hoary old “I found so many results on Google for x” cliché. The number of results is, of course, meaningless. [...]

By: Robert

Robert — Thu, 22 May 2008 15:42:23 +0000

Quite right Jacq. I imagine there would be a correlation on similar principles to that mentioned by Matt and Clarice, above.

By: Jacq

Jacq — Thu, 22 May 2008 14:55:10 +0000

What about the people who quote the number of Facebook Groups?

By: Alex Burford

Alex Burford — Wed, 21 May 2008 20:45:28 +0000

Well I love Gordon Brown, and that brings up 329 results with quotation marks. Still, I wouldn’t bet my mortgage on him winning the next general election. Market research can be so misleading…

A x

By: Matt Munro

Matt Munro — Wed, 21 May 2008 12:09:35 +0000

I think this is the problem of trying to apply quantiative measurement to qualitative information.

When I was an undergrad google (scholar) hits were taken as an indication of which areas of research are the most active. It is obviously a crude indicator, but was, at least in that context , consistent with other data.

How good it is at measuring the more qualititative aspects of the zeitgeist is questionable.

I still hate Gordon Brown though, if that helps.

By: Clarice

Clarice — Tue, 20 May 2008 14:24:57 +0000

Well, I’m not sure this is strictly true. It tells you plenty, it just doesn’t tell you what a lot of people say or imply that it tells you. As this post points out, you just have to take into account various features of the search, and contrast the results within some sort of context to make them meaningful. If you do that, where’s the problem? To say this cliche tells you nothing makes me think of the words “baby” and “bathwater”.

It’s akin to saying null hypothesis significance testing should be done away with on the grounds that it is so often misapplied or misinterpreted.

Or, as Nelson Goodman said (1977) that similarity doesn’t explain representation. It may not be sufficient, but that doesn’t mean that it’s not necessary.