Quote:
Originally Posted by mike alexander
Never liked the idea of linear regression on log-transformed data. Too many people assume linearity with no good reason except it's easy do do and understand.
|
I do that.
Thing is (as with their small animals), it works quite well for the "next few points". I've even extrapolated much further; but I'm not an academic. I'm surprised that it took the field so long to test an alternate model- if that indeed is the case here. I wish I could read the whole paper...