Hey Seth — this piece is super helpful, thanks for taking the time.
How should the ~66% of papers that have no position be interpreted? Is it misleading to suggest that 97% of the literature supports the hypothesis, if it’s really 97% of ~33%?
*edit: You clearly state that the 97% is of those papers that take a position, so I have no issue with your use of the number here. But it’s used elsewhere without that caveat and I don’t know what to make of those papers that don’t take a position.