The thesis analyzes statistical software for probability proportional to size (PPS) sampling without replacement using Monte Carlo simulations. It tests if software like SAS, SPSS, and R correctly implement the Hanurav-Vijayan and Rao-Sampford algorithms by comparing estimated joint selection probabilities from simulations to the true probabilities. The results show R estimates match the true probabilities, but SAS and SPSS estimates differ significantly, suggesting their implementations are incorrect or use a suboptimal pseudorandom number generator.