American Statistical Association New York City Metropolitan Area Chapter Memorial Sloan-Kettering Cancer Center Biostatistics Seminar

Many, if not most statistical methods implicitly or explicitly depend on assumptions about the underlying distribution of the observations, assume asymptotic or approximate distributions, or are optimal (most accurate, most efficient, etc.) for particular types of distributions. Through heuristics, empirics and graphical display, we will explore the what, why and how of using transformations to bring data, real and simulated, closer to those assumptions. We can thus reduce the mathematics needed for mathematical statistics.