Thesis (BS Applied Mathematics) -- University of the Philippines Mindanao, 2002

The Miller and Rice Gaussian quadrature method, being known as the best discrete approximator of continuous distributions, was applied to contagious empirical distributions. This was done to further investigate its capabilities in approximating discrete distributions as well as its performance compared to some known discrete approximation methods namely, the means of equally likely intervals method and the interval-Miller and Rice method. Thirty-nine variations of contagious empirical distributions, each having one hundred replicates, were simulated based from the theoretical contaminated normal distribution. The evaluation of the performances of these methods was based on their accuracy (measured by the percentage errors) in estimating the moments, expected utility, and certainty equivalence. The results showed that the Miller and Rice method is still the best approximator compared to the other two methods. However, its accuracy greatly depended on the parameters of contagious empirical distributions namely the means, variances and relative assigned weights of the two densities