Karl Pearson and Accounting Research

The great statistician Karl Pearson once took a bunch of skeletons, mixed up the bones randomly to recreate new skeletons. Then he asked anthropologists to tell whether the bones from these mixed-up skeletons were all from the  same persons; ie., that the bones had not been mixed up.

The anthropologists measured the length of the bones and the length of the skeletons they came from and based on the study of those ratios concluded that the mixed up skeletons had not indeed been mixed up.

From this rather bizzarre experiment (bless those skeletons whose identities had been stolen) led him to come up with the idea of  "spurious correlations". It has widely been known that that spurious correlations almost always exist when your variables in regressions are ratios.

The use of ratio variables is rampant in financial accounting so called "archival" research. I don't know if they have considered this aspect. If they have not, their results are probably spurious too.