376 – Contributed Oral Poster Presentations: Section on Statistical Computing
Small Sample Properties of JR Estimators
John Kloke
University of Wisconsin
Joseph McKean
Western Michigan University
Rank-based estimation in linear models has recently been extended to include cluster correlated data; see Kloke, McKean, and Rashid (2009). These JR estimators have been shown to be asymptotically normal under general mixed models. Asymptotic theory for inferential procedures, including confidence intervals and tests of general linear hypothesis, have also been developed. We conducted a series of Monte Carlo simulations to examine the empirical level and power of JR estimates in small samples. We are also developing an R package for these estimates. In this paper we summarize the results of our simulation study, highlight the software, and make recommendations on use.