Prior Data Conflict

My friend Gero (pictured here - and I still don't have my prize money, BTW) works in the field of prior data conflict.  To simplify extraordinarily, this deals with the situation in which you've begun collecting data under certain assumptions about what you're going to find, and your first few results look like that theory is going to be blown out of the water.

I mention this purely because I just heard a snippet of Front Row on Radio 4 in which I received my first information about Joss Whedon's The Avengers: it's an embarrassment to feminism with terrible dialogue and a strongly right-wing message.  So claims John Wilson, at least, about which I know nothing (save to say that anyone who precedes quoting MacBeth by pointing out he's about to use Shakespeare's words is either condescending, foolish or, on Shakespeare's "birthday", being entirely too cute).

I could phone Gero and ask him how to handle the exceptionally tricky task of figuring out exactly how to alter my preconceptions.  Or, I could just read more reviews.  It's almost as though all the stuff Gero (and I, and BigHead, and Ibb) work on turn out to not be useful in daily life.  How strange...