Statistical Software for Resampling Methods

Medical data often require complex models. for example, clinical monitoring produces time induced cand relationships among variables change due to medical intervention. Many popularly used biostatistdepend on approximations made for mathematical tractability. Resampling methods are computationallytechniques which approximate the distribution of a statistic using only the observed data. The two-fresampling methods are that (1) they are conceptually simple and (2) they often apply in complex prothrough other techniques. Phase I research will focus on designing software in the S-PLUS environmenfacilitates use of resampling methods. We plan to achieve this through judicious use of presentationflexible software design. This design will support the different needs of (1) data analysts and (2)who want to modify and extend resampling capabilities. In Phase I we will prototype the design, focumodel. Resampling methods offer the opportunity to revolutionize statistical practice: making it easunderstand, yet applicable to complex problems.