JSM Proceedings, Risk Analysis Section
American Statistical Association
Weibull, Maximum Likelihood, Reliability
Maximum Likelihood estimation of the two-parameter Weibull distribution is straightforward; however, there are multiple methods for maximum likelihood estimation of the three-parameter Weibull. This paper presents an evaluation of these methods using four data sets including oil spill data from the Gulf of Mexico. Highlighted are fairly major differences in the estimated parameters between nine statistical packages. A VBA routine has been developed allowing practitioners to implement three-parameter Weibull maximum likelihood estimate within Excel. The code and support documentation are available free at http:faculty.otterbein.edu/WHarper.
Harper, William V.; James, Thomas R.; Eschenbach, Ted G.; and Slauson, Leigh, "Maximum Likelihood Estimation Methodology Comparison for the Three-Parameter Weibull Distribution with Applications to Offshore Oil Spills in the Gulf of Mexico" (2008). Mathematics Faculty Scholarship. Paper 20.
Harper, William V., Thomas R. James, Ted G. Eschenbach, Leigh Slauson (2008), "Maximum Likelihood Estimation Methodology Comparison for the Three-Parameter Weibull Distribution with Applications to Offshore Oil Spills in the Gulf of Mexico", JSM Proceedings, Risk Analysis Section, pp. 1231-1238, American Statistical Association, Denver, CO.