Was spurious correlation of ratios inadvertently explained by Howe?

33 Views Asked by At

Roger Howe pointed out that vector sums tend toward convexity. Could this explain the spurious correlation of ratios? The idea that I have is that division of one variable by another essentially turns the answer into a vector-like entity, and the resulting relatively-high positive correlation is simply a result of the geometry that is tending towards convexity. Is there anything to this?