I've just finished reading this social sciences book about how interventions do or don't work. It was fairly hard going, but interesting. The key message is that you can't just trust your own common sense. The author discusses things like 'Sure Start' schemes for children which, whilst well-meaning, are seldom tested to see if they actually work. Even more shocking is that when they are tested they can often be proved to not work. Surprisingly, some tiny interventions do work. The book also includes discussions about negative stereotypes and ways to lessen the effects.
This has strong connections with User Experience and measurability. It also makes me think about engagement at work.