4 ms·
I liked your example of scientists from history. One of the things I'm minorly obsessed with is the idea that we've gotten better at science itself recently.
by enkid 6y ago
I liked your example of scientists from history. One of the things I'm minorly obsessed with is the idea that we've gotten better at science itself recently. So not only did they not know they weren't eliminating the proper variables, they had fewer techniques to know which variables to eliminate. This is similar to the application of statistics. We're just now realizing issues with publication bias and how to properly due meta-analysis that can help identify those problems. Think about not knowing how to apply even simple statistics to science.
- kerkeslager 6y agoBut that's exactly my point: we still are groping around in the dark on a lot of issues. Can you explain to me why P=.05 is the ubiquitous standard for statistical significance? No? Neither can anyone else I've come across. It doesn't get much more basic than that--we still don't know how to apply even simple statistics to science. And what we as a species know, we as individuals don't know. It's true that the knowledge of the human race has moved forward with regard to applying statistics to science, but if you make a habit of reading scientific papers and checking their math, you'll quickly realize that there are critical errors in lots of papers. And even if the scientists get the math right, the journalists and marketing teams who publicize their results rarely do. Better tools don't do us much good if we don't use them.