Quote:
Originally Posted by Discus
I think most people tend to shy away from designing or analysing multivariate experiments for fairly obvious reasons (perhaps not least of which may be math-o-phobia).
|
It's not so much math-o-phobia as ignorance and laziness. You really do need to know what you're doing with multivariate experiments, and the discipline required to carry one out is unattainable for most people. And yet the effects of interactions between factors are often more important than the main effects.
There is no doubt in my mind that the amount of N in a fertilizer will affect the way plants respond to different amounts of K. Same with Ca and Mg. But I doubt anyone will carry out an experiment that will study these interactions. Experiments like the absurd one listed earlier (on Phals) are designed by agricultural chemists, not statisticians - and that's really too bad.