Samples and Populations

The Times Higher Educational Supplement published the chart shown here, demonstrating the results of their poll of academics' intentions about voting. (Not who they were going to vote for, but whether they would vote.
It seems that for some grades of individuals working in UK universities, the percentage of turnout would be surprisingly low. (To put those figures into perspective, the UK turnout for the 2005 general election was 61.3%.)
But there's something rather unusual, and perhaps important about academics that makes them rather a different group. Many are not British citizens, and if you're not a British citizen, you can't vote in a general election (my wife being an example of a non-British citizen academic working in the UK. She can, and does, vote in local elections and European elections, but not general elections - UK general elections, that is).
I don't know how the proportion of non-British staff varies across either grade or faculty, but I'll bet it's not random. It's my (extremely anecdotal) experience that Health Sciences academics (where I work) are most likely to be British, psychologists slightly less likely, and economists less likely again. I have no idea about engineers, who score lowest.
If it is the case that Britishness is not distributed randomly with respect to grade and faculty, the comparisons are, of course, meaningless.

(Just as an aside, I always feel that describing myself as "British" is a bit anachronistic, and if asked to give my nationality on a form, I tend to write "UK" but British is the techically correct thing to write. So if you live in, say, Northern Ireland, you can be British without ever having visited Britain. At least that's my understanding.)