turns out to be something discovered at least a decade ago
The first reason for this is that students and Academics largely don't read the literature. The second is that for decades now the literature is garbage. In my tiny area of expertise, I can't tell you how often I don't make it past the abstract before chucking a paper into the round filing cabinet. My theory is that statistics ruined the Academy; they allow people to conduct "research" by floating in the wind and shoving their garbage data from their garbage "experiments" through garbage "models" and claiming they've done something. Prior to the middle of last century a researcher was expected to develop formulae and a mechanistic theory for how they thought shit works and for these to actually be useful; now its all garbage correlations.
Then they expect me to "listen to the experts." Its the Gell-Mann Amnesia Effect PhD Edition.
I'm convinced most people are too dumb or lazy to understand statistics or even counting numbers.
It's not even that correlations are useless, it's that people will fail to even try to isolate what they're trying to measure and try to use quantitative analysis for a sample of 15 data points over 30 years.
The first reason for this is that students and Academics largely don't read the literature. The second is that for decades now the literature is garbage. In my tiny area of expertise, I can't tell you how often I don't make it past the abstract before chucking a paper into the round filing cabinet. My theory is that statistics ruined the Academy; they allow people to conduct "research" by floating in the wind and shoving their garbage data from their garbage "experiments" through garbage "models" and claiming they've done something. Prior to the middle of last century a researcher was expected to develop formulae and a mechanistic theory for how they thought shit works and for these to actually be useful; now its all garbage correlations.
Then they expect me to "listen to the experts." Its the Gell-Mann Amnesia Effect PhD Edition.
I'm convinced most people are too dumb or lazy to understand statistics or even counting numbers.
It's not even that correlations are useless, it's that people will fail to even try to isolate what they're trying to measure and try to use quantitative analysis for a sample of 15 data points over 30 years.