TESTING EQUALITY OF THE SCALE PARAMETERS OF PARETO DISTRIBUTIONS; TWO COMPUTATIONAL APPROACHES

We consider testing the equality of two scale parameters in Pareto distributions. There is no exact/asymptotically test for this problem. Therefore, we propose two newly methods based on the concepts of generalized p-value (Tsui and Weerahandi [8]) and computational approach test (Pal et al. [7]) for this problem. These methods are feasible approaches to the statistical hypothesis test. Using simulation studies, we show that the actual sizes of proposed methods are close to nominal level and powers of these methods are satisfactory. We also illustrate these methods using real data.