Believing in experts

[Dill]

(07-03-2023, 01:14 PM)Belsnickel Wrote: I consider myself a social scientist because I do policy analysis and assessments which involves political science, sociology, psychology, economics, etc. It is an interdisciplinary field primarily in the realm of social sciences and poli sci specifically. Anyway, my rule of thumb is consensus. What is the majority of the literature saying? And when I say literature, I am talking about peer reviewed, replicable (preferably replicated), research. One article does not convince me. I will look for meta-analyses on a subject. If there isn't one, I will look at several articles on the topic. I will look at the statistical results in the article. Do they draw the same conclusions I would have? There are times where articles will draw conclusions that I would say "eh, the statistical power of that is low, I wouldn't say that for sure." I look at their methods to see if there is anything amiss in how they did their research. Can they actually draw causal conclusions? If the majority of research in the field says the same or similar things then that is what I will go with until there is something that disproves their research.

I care less about who is saying something than I do about what is behind their statement. Now, people who are well respected in their fields receive a degree of deference. Sure. Makes sense. But they can say stupid shit just as much as the first-year grad student can. The problem is that most people aren't literate in these things. They don't understand what to look for in a methods section that can give away bad design. They don't know what p-hacking looks like or even what statistical power is. This is why experts, and "experts," are relied on so heavily. People aren't able to understand this information or they are too lazy to really dig into it.

Well said, though I would say lack of science/statistical* literacy has long been with us, 

but it is a special problem now as some politicians have learned to focus on that illiteracy to 

discredit science-based policy, as that is now associated more with one party than the other.

They target the four corners of "deceit": government, academia, SCIENCE, and media.

Who is left to "trust" then? They have some suggestions that don't involve no fancy "scientific consensus." 


*e.g., I consider myself statistically CAREFUL, but not statistically literate.