Word verification is a cool way to prevent automated sign-ups and spam. The principle is that only a human being can “read” the “word” (which is actually a string of characters and not necessarily a word) and thus it prevent automated systems (like spambots) from doing damage. It’s almost wiped out the comment spam on my blogs.
There are problems with word verification though. One problem that I’ve faced is that in certain cases, it is extremely hard to figure out the string of characters in the image, because of the slanting and curvature on the letters. (I’ve encountered this several times at Yahoo and once or twice while commenting on blogs.)
In these situations, you are reduced to guess-work and though I’ve been right most of the time I’ve got the “word” wrong a couple of times. (No, my eye sight is fine, thank you very much.)
Also, for people who cannot see, this method is of no use. W3C has an interesting article about the problems with word verification technology and Wikipedia has an explanation of a captcha (an acronym for “completely automated public Turing test to tell computers and humans apart”).
PS: I found through these links via this blog post (Ben Boyle Lives Here), which I found through Google.