Abstract :
We will study some possible non universal corrections that could affect single measurements of the Weinberg angle sin2θW (leaving the full average essentially unchanged). We will concentrate on the class of the models that introduces a new light gauge boson and an extended Higgs sector, but we restrict the Higgs charges choice in order to automatically achieve zero mixing at the tree level between the Z0 and the Z′. This will more naturally keep small the universal effects with a relatively light Z′, but it will imply only vector-like coupling between this boson and the standard fermions. A SO(10) gauge boson which distinguishes between familiies can have such properties and could explain a deviation in the AFBτ without affecting the τ polarization measurements at LEP.