Is there a need for ten people to spend a year developing a new language test? I may be ignorant of the process, but it strikes me as a large figure. I say this in light of your concern about the integrity of the test. This is a challenge, because it costs so much to develop a new test. My thrust was this: does it really need to cost $1 million? Do we really need ten people to spend a whole year developing a language test?
On June 18th, 2009. See this statement in context.