Re: Time to be pedantic
@RM Myers: "It would be nice if all studies could include a professional statistician, since many of the issues are not at all obvious to non statisticians."
Many studies would love to have a statitician on board, but there aren't enough statisticians to go around - there aren't even enough to go round for research ethics committees! My last REC was fortunate enough to have a really good statistician, and she (and, to a much lesser extent, I) used to get VERY frustrated with really fairly basic errors which meant that the project wasn't testing what it thought it was. All thus because no-one was adequately teaching stats to researchers, and there was no-one to run a project past.