Professor; Ph.D., Penn State, 1994.
In many experiments, one of the driving forces to implement a statistically sound sampling design is the cost of acquiring measurement from sampling units. In most situations, researcher has a choice between making costly, time-consuming, precise measurements and quick, inexpensive, imprecise measurements. One successful strategy is to use inexpensive measurements on a large number of individuals to determine which expensive measurements should be collected. Ranked set sampling provides a collection of techniques to provide detailed plans how to collect and analyze these expensive and inexpensive measurements. My current research involves drawing parametric and nonparametric inference based on these kinds of nonstandard sampling procedures such as ranked set sampling and its modifications. For example one of my most current research (through collaboration with Professor Steve MacEachern) involves developing estimation and testing procedures for the design of experiments for settings where the cost of each experimental unit is limited or expensive. Another joint work with Doug Wolfe and a graduate student, Roxana Alexandridis, deals with multi-sample inference for simple-tree alternatives. I am also collaborating with visitors, Omer Bilgin and J. V. Deshpande. My work with Omer Bilgin and Doug Wolfe involves application of ranked set sample methodology to animal breeding experiments. Another current work with Deshpande deals with developing a nonparametric test for the equality of scales in two-sample ranked set sampling models. I am also working on sub-sample statistics (sub-sample without replacement) to develop nonparametric inference for certain characteristic of the population. My most recent work in this area involves finding an optimal sub-sample design to improve the efficiency of the Mann-Whitney-Wilcoxon test. My third area of research concentrates on rank based inference. A joint work with Young Hun Choi in this area deals with score generating functions in rank regression.