|
Abstract:
|
The new version of JMP (Version 6) provides data analysts with and easy and powerful sets of statistical and graphical platforms to handle most common data analysis situations. Fortunately for JMP users it includes a powerful scripting language to extend its capabilities. In this paper we will use JSL to program additional statistical tests that exists in the literature but are not usually included in the standard statistical methods texts and do not exists in most statistical software including JMP. We will limit this paper on comparisons of two statistics (involving means, medians, variances, counts, correlations, proportions) under a variety of conditions that include parametric assumptions, distribution free tests and when the two might be correlated. The JSL code for the test will be made available to the JMP community so that users will have more tests available to them.
|
