[mandalar12]

Here is an article [1] trying to estimate that kind of numbers (wage, number of active workers, etc.) from various sources and questionning the ethics of using MTurk for science purposes.

The bottom line is: one third of tasks are achieved by people relying on MTurk (as primary or secondary money input) having working conditions that would not be considered acceptable in any first-world country (hourly wages around $2, working long hours with, obviously, no kind of worker advantage or protection).

[1] https://hal.archives-ouvertes.fr/file/index/docid/569450/fil...

[jbarrios]

Harvard recently published a paper on the mturk demographics, you can find it here: http://scholar.harvard.edu/files/dtingley/files/whoarethesep...

If you're looking for something less academic, try out this great blog: http://experimentalturk.wordpress.com/