Aaron Childs
Associate Professor, Mathematics and Statistics

I am interested in the diverse applications of order statistics to statistical inference and outlier accomodation. Some of my research involves order statistics from independent and non-identically distributed (I.NI.D) random variables. I.NI.D. random variables are a generalization of the multiple-outlier slippage model, and therefore have applications to outlier accomodation. Order statistics also have applications in classical inference problems; for example I have developed some hypothesis tests based on order statistics and inverse sampling which provide a more efficient alternative to the standard chi-square test for homogeneity. I have also developed a set of Maple programs that allow one to compute series approximations for moments of order statistics. Another area of interest to me is waiting time problems. My contribution to this area has been to develop new methods, some of which are based on uniform random variables and generating functions, which enable one to solve a wide variety of very complex waiting time problems. I have also been involved in some consulting projects, one of which requires testing for independence between virus and respiratory disease data using time series analysis.