The Weibull family with survival function exp{-(y/sigma)(alpha)}, for
alpha > 0 and y greater than or equal to 0, is generalized by introduc
ing an additional shape parameter theta. The space of shape parameters
alpha > 0 and theta > 0 can be divided by boundary line alpha = 1 and
curve alpha theta = 1 into four regions over which the hazard functio
n is, respectively, increasing, bathtub-shaped, decreasing, and unimod
al. The new family is suitable for modeling data that indicate nonmono
tone hazard rates and can be adopted for testing goodness of fit of We
ibull as a submodel. The usefulness and flexibility of the family is i
llustrated by reanalyzing five classical data sets on bus-motor failur
es from Davis that are typical of data in repair-reuse situations and
Efron's data pertaining to a head-and-neck-cancer clinical trial. Thes
e illustrative data involve censoring and indicate bathtub, unimodal,
and increasing but possibly non-Weibull hazard-shape models.