In The Invisible Gorilla, Chris and I discuss many aspects of the illusion of knowledge, the tendency to think we have a better understanding than we actually do. One aspect of this illusion is that we easily mistake surface understanding for deep understanding, what Leon Rozenblit and Frank Keil called the “illusion of explanatory depth.” That aspect of the illusion of knowledge leads us to think we have a deep understanding when all we really have is knowledge of the surface properties. In a recent op-ed in the LA Times, we argued that this illusion of knowledge, when coupled with technology that presents information in short, surface-level bursts, can lead to a mistaken belief that we actually understand more than we do.
One practical consequence of the illusion of knowledge is the planning fallacy – we almost always assume that new projects will take less time and resources than they actually do. In part, the planning fallacy arises when we fail to take into account all the unpredicted complications that can arise and we assume the simplest possible scenario. A new paper by Darron Billeter, Ajay Kalra, and George Loewenstein presents an interesting twist on the typical course of the illusion of knowledge. In most cases, the illusion of knowledge leads us to think that we’re more skilled or knowledgeable than we actually are and it leads us to underestimate how long it will take us to accomplish our goals. Billeter et al looked at predictions for how long people would take to learn a new skill such as typing on a Dvorak keyboard.
Before using it, they were overconfident in estimating how long it would take them to learn, just as we would expect from overconfidence in our own knowledge. However, as soon as they tried it out and realized that they couldn’t succeed without practice, they over-corrected their expectations and assumed it would take them longer to learn the skill than it actually did. Apparently, people lack insights into how quickly they can acquire new skills even though they initially think they won’t need new skills at all. I do wonder, though, whether these underestimates apply only to people who are capable of learning new skills fairly readily. I wonder if the estimates of old dogs would be better once they try the new product — that is, would their corrections be better calibrated to their actual ability to learn.
hat tip to Cynthia Graber at Scientific American
Billeter, D., Kalra, A., & Loewenstein, G. (2011). Underpredicting Learning after Initial Experience with a Product Journal of Consumer Research, 37 : 10.1086/655862