200 K. The failure of this model in predicting Hall factor and Hall mobility at low temperatures and dopant densities is due to the omission of the proper anisotropy formulation in the model of the valence band of silicon. This weakness in the model would be remedied by use of the exact model in a manner similar to that of Nakagawa and Zukotynski [11,77].

     Further improvements to this model would include the exact formulation of the valence band of silicon on Kane's [27] model, as prescribed by Nakagawa and Zukotynski [11,77]. Thus the proper nonparabolicity and anisotropy would be included. For better comparison with experimental results, the variation of ionization energy with dopant density should be included in the calculation of hole density for the deeper impurities. Also the effects of compensation and the presence of other p-type impurities in silicon samples must be considered. It would also be of great benefit to extend this study into the heavy doped region.