Here's what's on my mind this week: Why do people settle on just one record to answer their research question? In other venues, I see people asking all the time how many records they need to prove the answer to a question -- is one enough? or two? or three? Maybe it comes from a lifelong int...