Don't eat the marshmallow yet!

One of the things that has typically gotten ignored is that relatively small, and seemingly insignificant things can be a strong predictor for future, more significant behaviour. The long-term followup to one such experiment was covered by the New York Times in an article entitled Marshmallows and Public Policy. (The New Yorker has a similar article with a whole lot more detail). ... and for the ADD crowd, here's the corresponding talk from TED:

On a similar note, I've come across references to John Gottman in a couple books, neither of which was focused primarily on marriage. Based upon the research he's done, he's able to predict long-term success of marriage with high accuracy based on watching a few minutes of a couple conversing about a subject not directly related to their marriage. (What's crazier is that an analysis of a urine sample from the couple's children also seems to work quite well).