So, if this all snark, does that mean there's no "strong correlation between the undergraduate enrollment of an institution and its Google Trends index"? Because that would be a pretty interesting correlation.
I'm also curious why Princeton's search volume seems to have declined. Obviously it doesn't mean Princeton is going to disappear, but what does it mean? Could be statistically insignificant, perhaps.
>I'm also curious why Princeton's search volume seems to have declined. Obviously it doesn't mean Princeton is going to disappear, but what does it mean?
Google trends represents share of search volume, not absolute volume [1]. As the higher education bubble continues to grow, ivy league schools have not grown as fast as other schools, and there are new schools being created as well. All of this activity contributes to Princeton's (for example) share of search decrease within their sector, and overall.
It is really easy to find something that correlates with your data, when your data follows a simple trend. Now if princeton enrollment was going up and down a lot, and the google trends followed the same shape, that would be something.
md224 wrote: "I'm also curious why Princeton's search volume seems to have declined. Obviously it doesn't mean Princeton is going to disappear, but what does it mean? Could be statistically insignificant, perhaps."
I have seen this in most nerdy topics associated with early internet adoption in Western English speaking countries -- or at least that is my theory. Basically as other countries come online and late adopters come online, they do not search for the same types of topics earlier adopters looked for.
This trend for academic or technical topics decreasing as a percentage of Google trends is universal, see here:
I'm also curious why Princeton's search volume seems to have declined. Obviously it doesn't mean Princeton is going to disappear, but what does it mean? Could be statistically insignificant, perhaps.
EDIT: Here are some comparisons:
http://www.google.com/trends/explore#q=%2Fm%2F05zl0%2C%20%2F...
Interesting how the initial spread seems to narrow.