Ask HN: Why can't we prove the Blub Hypothesis?
"Surely if lisp makes a programmer N times more efficient then it would be easy to study scientifically. Have there been any studies comparing productivity?"
(http://news.ycombinator.com/item?id=1016583)
I've never personally seen one and can almost immediately rattle off a thousand reasons why it would be impossible, but that doesn't mean it is.
How would you structure a study to make it practical to gather real, objective, and interpretable data on programming efficiency? It'll be necessary to consider controls, sample size (Mech. Turk, perhaps), and, largely, the definition of "programming efficiency", but with all the talk about how Java is Blub I'd love to see it actively, quantitatively hamper someone's work.