Some Recent Results in Lognormal Parameter Estimation Using Grouped and Ungrouped Data

Abstract

Two and three parameter lognormal distributions were fitted to the distributions of certain types of automobile insurance claim payments. Maximum likelihood estimates of the parameters were obtained using both grouped and ungrouped data; for the three parameter cases and the two parameter grouped data cases, the method of scoring worked well. Comparison of the efficiencies of the estimates using grouped data showed that logarithmic grouping was substantially better than equal interval grouping.