I should teach the world that science is not about sweeping statements from authority figures in white coats, but is instead about fallible data. The importance of that data is understood to varying degrees, by individual people, who are variably capable of interpreting data. These people come to a verdict as individuals, about how useful that information is, in supporting or refuting a hypothesis. Anyone could and should do this.

I also wish that everyone understood the British epidemiologist Austin Bradford Hill's criteria for causation. People are narcissistic enough that the science they are most interested in is the science that relates to themselves and their health. If they are going to worry about their health, then they should understand enough to worry about the right things.

See Ben Goldacre's website.