Value-added analysis

We've been following the debate of value-added analysis as a measure of teacher effectiveness. Can test scores really be an accurate measure of how well we teach? Where do the other "important" factors fall - poverty, literacy, access to books and materials, etc. - Teacher evaluation is too important and way more complicated than compiling a list of test scores for us to allow a once-a-year "event" dictate who is doing a good job in the classroom. I welcome accountability. All professionals should expect to be accountable for doing a good job. But is this the way to do it? Any thoughts?