Gauge Coupling Unification: Strings versus SUSY-GUTs

Standard (level one) heterotic string models with standard model gauge group predict the unification of $SU(3)$ and $SU(2)$ gauge couplings whereas the $U(1)$ factor is unified modulo an unknown normalization factor $k_1$. On the other hand the unification mass is known. I argue that this situation is quite analogous (though opposite) to that in SUSY-GUTs in which the $U(1)$ normalization is known ($k_1=5/3$) but the unification mass $M_X$ is unknown. I emphasize that $k_1$ should be taken as a free parameter in the string approach (quite in the same way as $M_X$ is taken as a free parameter in SUSY-GUTs). If this is done, the success of the string approach concerning gauge coupling unification is comparable to that in SUSY-GUTs