Often I see people using an effective mass model to describe electrons in the bottom of the conduction band. Spin orbit is then included as a perturbation in this effective mass model. But what is the justification for using this sort of model? Would the correct way not be to start from the full k dot p equation with spin orbit included and then derive an effective mass model from this? My concern is due to the fact that I would expect the effective mass to be dependent on the spin orbit interaction, but maybe that is not true. Also, what about external magnetic fields. Do these affect the effective mass?