The goal of this research is to investigate the feasibility of producing and marketing a robust statistical software package which will compute exact significance tests for ordered and unordered categorical data on microcomputers. There has been a long-standing need for such software but none of the current statistical packages on mainframe or microcomputers has filled this need. The feasibility of producing this product will be explored from the point of view of smaller memory, smaller word-size and slower CPU speeds on microcomputers as compared to mainframe computers. This investigation will thus yield a clear understanding of the limits of problem size that can be tackled by the newest generation of microcomputers. Marketing feasibility will be assessed mainly through contacts with firms marketing statistical software. Firms like MINITAB, STAPRO, SYSTAT, etc. will be offered the opportunity to add our software as an extension module to their existing systems. The feasibility of marketing the product as a stand-alone system will also be explored.National Cancer Institute (NCI)
