How not to prove-by-Google
PZ Myers points to a right-wing buffoon who claims to have proved that Evil Socialist indoctrination is going on in universities to get students to do Barack Obama-style community organizing. He completely misuses Google to prove this. I wrote a long comment about it, but it’s so long, I may as well put it up as a blog post instead.
If we are going to do proof-by-Google for academia, it’s quite simple. Academia has it’s own set of domains - .ac.xx (where xx is the country code, .edu.xx and just .edu (in the US). So, if you want to prove-by-Google something about American academia, a far more sensible way to do it is through relative counts of results with site:.edu as an operator.
Let’s calibrate then: the word ‘biology’ on .edu returns 27.9m results. Business returns 63.3m. Philosophy returns 21.7m results. Economics returns 22.6m results. Mathematics returns 18.2m - [mathematics OR math] returns 33.5m. [theology OR “religious studies”] returns 2.04m. Darwin returns 1.3m.
[community AND organizing site:.edu] returns 1.7m results. If we accept proof-by-Google as a valid methodology, then we should perhaps investigate here. This doesn’t prove the right-wing argument: that just proves that there are a lot of web pages about community organizing in the .edu domain. That there exists lots of pages in the academic domain that discuss a subject doesn’t mean that the subject is being taught on a course to students. I just searched for [cheese site:.ac.uk] and Google informs me that Imperial College, London’s top science and technology college, has a Cheese Society - their webpage is in the .ac.uk domain though.
Then it hit me: the reason why I get lots of results is because of Community Colleges: any mention of organizing on a page on any community college website is bumping up the figures, even if it’s a page about organizing one’s trip to the local natural history museum as part of Intro to Geology or whatever.
So, let’s exclude mentions of Community Colleges. We can do this by putting “community organizing” in quotes. Run it again: [“community organizing” site:.edu]. What do we get? 54,000 results.
Just as a point of comparison, I put in [“intelligent design” site:.edu]. Surely, if one can conclude that community organizing is being taught on American college campuses on the basis of 54,000 Google hits, then we can see whether or not that evidence would suffice for something else? It returns 41,400 results ([“intelligent design” site:.edu]). Presumably, the right-wing can now accept that they aren’t being “expelled” from the academy - they are being discussed only a little bit less on the academic web than community organizing. Parity for insanity!
But, you might say: this instrument is a bit clumsy. All those pages about intelligent design are rather negative. They include that terrible statement by Lehigh University saying that ID is just some crazy idea that Behe accepts but nobody else does, and a nice article by the philosopher Robert Pennock talking about the wedge strategy and whatnot. They are discussing ID - but they are negative, thus proving their bias…! Which proves the point. Google can give you a rough idea about some things, but you need to look at those results honestly to get a flavour of what they are about. In the case of community organizing, the pages aren’t actually connected with academic disciplines or course catalogues. It’s not like you are going to go along to a class on Shakespearan poetry and find yourself forced to go out and help the poor organize politically. I did see a few pages in there listing community organizing as something that people might want to do after finishing college: having done something like politics or law, becoming involved with politics might be a fairly reasonable thing to do.
Another point of comparison: [chapel site:.edu] returns over two million results. If Google proves that people are being indoctrinated by the left into socialist-Obama-community-organizing (and, my, isn’t helping poor people get politically active wholly against God’s word), then surely the two million pages about college chapels is more than enough counterbalance…?
Proof by Google is not really very good proof at all. It’s interesting, and sometimes a good way of testing to see if a sociological hypothesis has weight. But you need a point of comparison. For instance, I did some very amateur research with Google and found that BBC News grossly over-represent the views of the Catholic Church as compared to the views of academics in a way that other UK news publishers did not. But without comparing it to other similar news outlets, it’s just a bunch of useless numbers. Lies, damn lies, and Google search results…
