Anthere said:
We do not know WHO read the encyclopedia.
When I was at the Grauniad we commissioned a company to come and attach to the site, for a period of some weeks, a pop-up survey that was seen by a randomly selected subset of users, who then would have the option of filling it in. You can also learn a certain amount by looking at web server logs. Country of origin, ISP, and some information about the software being used to browse. Also Referer (where Wikipedia is reached as a result of an external link on another website.) With tracker cookies (a form of spyware) you may also find out which other sites are being visited. A benign use of this latter technique would be an in-house domain-wide tracker that would record the Wikipedia sister sites that have been visited by a user. This latter technical information wouldn't tell us about our users so much as how Wikipedia was being used. How much was from academia, how much from school networks, how much from different countries and different ISPs, and so on.