Daniel Kahneman

Under the stress of the event, we felt, the soldiers' true nature would reveal itself, and we would be able to tell who would be a good leader and who would not. But the trouble was that, in fact, we could not tell. Every month or so we had a "statistics day," during which we would get feedback from the officer-training school, indicating the accuracy of our ratings of candidates' potential. The story was always the same: our ability to predict performance at the school was negligible. But the next day, there would be another batch of candidates to be taken to the obstacle field, where we would face them with the wall and see their true natures revealed. I was so impressed by the complete lack of connection between the statistical information and the compelling experience of insight that I coined a term for it: "the illusion of validity."

Amos and I shared the wonder of together owning a goose that could lay golden eggs – a joint mind that was better than our separate minds. … Some of the greatest joys of our collaboration-and probably much of its success – came from our ability to elaborate each other's nascent thoughts: if I expressed a half-formed idea, I knew that Amos would be there to understand it, probably more clearly than I did, and that if it had merit he would see it. Like most people, I am somewhat cautious about exposing tentative thoughts to others – I must first make sure that they are not idiotic. In the best years of the collaboration, this caution was completely absent.

I got quite nice results in my one-question studies, but never wrote up any of the work, because I had set myself impossible standards: in order not to pollute the literature, I wanted to report only findings that I had replicated in detail at least once, and the replications were never quite perfect. I realized only gradually that my aspirations demanded more statistical power and therefore much larger samples than I was intuitively inclined to run. This observation also came in handy some time later.