You're missing the point. The weakness isn't in data analysis, its the reasoning process itself that is superstitutious. I'm not suggesting there is a better process, but we shouldn't be blind to the problem. To attribute a hypothesis to an observation, then collect data and develop experiments to reinforce that line of thinking, is no different than superstitution development. The only difference is rigor.
The data analysis is the point. Your own statements show you're starting from assumptive logic and therefore faulty reasoning. You are literally using faulty reasoning to say scientific reasoning is faulty. My example: "To attribute a hypothesis to an observation, then collect data and develop experiments to reinforce that line of thinking," This is exactly what the methods are developed to avoid! This is not to address corruption, which does exist. In those cases they look to reinforce.
Oddbean new post about login | logout