When carrying out a two factor experiments using split plot design. The conventional method of reporting is to use main and interaction effects. My question is that can interaction of two factors be considered as an effect because i have come across researchers saying interaction is not an effect.