I recently encountered this article (shared by Gwern). The following is a riff off of that, which appeared partially as a buzz comment as well.
Education is very important.
Culture is very important.
The culture of not admitting mistakes which is described is quite different from here, but at the same time, an uncomfortably familiar facet of human nature. Look how bad it can be!
In America, it is more often simply keeping silent rather than admitting something which might make you sound stupid. That is also a mistake, _especially_ when it's a matter of being afraid to ask stupid questions. We've all been there, right? No one wants to sound uninformed. The problem is that the best way to _get_ informed is to ask these questions.
While we may understand this on a conceptual level, it is hard to get that "small amount of courage" needed to ask the critical questions immediately when we have the chance. Perhaps it is because it's an immediate cost for a delayed reward, and we are not very good at thinking of the delayed reward. Here is a trick which may help. So you are worried about making a good impression, right? By keeping silent when you don't understand, you don't make any impression on that issue. By asking questions, you give the impression that you are really interested. If you ask enough questions to really understand, you make the impression of being able to pick things up quickly. By practicing the art of asking good questions, you stand a better chance of making an impression as a good critical thinker.
What we need is a culture which encourages admitting mistakes and inadequacies as soon as possible. Specifically, we need to encourage the idea that the best way to *be* that competent, informed person is to seek feedback and correct yourself whenever you fall short. This also gets into the space of what we do when we argue. We bring in all the support we can to make our side right. Instead, we should be looking at the evidence to choose the right side. That is discussed in this article, also shared recently by Gwern. (It's also discussed extensively on Less Wrong.) How do we engineer a culture in which real discussion, from which both sides can learn, is encouraged? Here are a few tips (this time shared by Luke Palmer). These come from a system of government called Sociocracy.
I'm not buying into the whole of Sociocracy (and neither is Luke), but these tips for how to run small groups seem immediately useful. The basic idea is that our win/lose argument mindset is partly due to the pervasive culture of democracy. What do we do to resolve a dispute in a small group? We vote, and the winning side is considered correct. Sociocracy suggests an alternative: ask for everyone's objections, and don't make a decision until all the objections have been dealt with. This builds a better decision. The phrase "science is not a democracy" comes to mind. Science instead relies on the merit of arguments to convince individuals of the correctness of views. This is the idea behind sociocracy. (I do feel voting is still a good way of making decisions when unified decisions are needed and consensus can't be had or would be too expensive, though.)
So how do we build a more effective culture? By seeking out good ideas like this, and sharing them. (Ultimately, Luke's idea of experimenting with new cultures to see what works is also good.)