Further methods of estimating the parameters of the weibull distribution

Probability-weighted moments, an order statistic technique, generalization of the total Q on test, maximum likelihood, generalized least squares, and ordinary least squares are all evaluated as methods for estimating the parameters of the Weibull distribution. The root mean squared error in the estimated 95 percent and 99 percent quantiles is used to compare (a) the effectiveness of parameter estimations and (b) the root mean squared error in the estimated 95 percent and 99 percent quantiles. Simulations are used to look into these requirements for sample sizes of 10 and 25, as well as a wide range of shape parameter values.