This is a critical study on the McWhorter model, taking into account the conditions for additivity of GR spectra. A major finding is that the model does indeed give 1/f spectra. The calculated a-values agree with experimental values of MOSFETs in strong inversion. A simple comparison of experimental a-values with values calculated from a ¿µ-model or a McWhorter ¿n-model does not allow us to decide which the correct noise source is. However, the dependence of a on the gate voltage has often been used as a criterion for the correct source. Nevertheless, the VGT-dependence does not lead to unambiguous conclusions either.