Some time ago I read the slides from Greg Wilsons excellent presentation Bits of Evidence.
Today I found out that it's on YouTube (see comment).
I find it fascinating that the scientific studies of software development apparently seems to be mostly ignored or unknown to our community.