HOW RELIABLY CAN THE HIGGS-BOSON MASS BE PREDICTED FROM ELECTROWEAK PRECISION DATA

From the LEP precision data and the measurement of the W-boson mass, u pon excluding the observables R(b), R(c) in a combined fit of the top- quark mass, m(t), and the Higgs-boson mass, M(H), within the Standard Model, we find the weak 1 sigma bound of M(H) less than or similar to 900 GeV. Stronger upper bounds on M(H), sometimes presented in the lit erature, rely heavily on the inclusion of R(b) in the data sample, Upo n including R(b), the quality of the fit drastically decreases, and by carefully analyzing the dependence of the fit results on the set of e xperimental input data we conclude that these stronger bounds are not reliable. Moreover, the stronger bounds on M(H) are lost if the deviat ion between theory and experiment in R(b) is ascribed to contributions of new physics. Replacing s(W)(-2)(LEP) by the combined value s(W)(-2 )(LEP+SLD) in the data sample leads to a bound of M(H) less than or si milar to 430 GeV at the 1 sigma level. The value of S-W(-2)(SLD) taken alone, however, gives rise to fit results for M(H) which are in confl ict with M(H) greater than or similar to 65.2 GeV from direct searches .