Data Analysis Papers
Power and Sample Size Determination for Linear Models
John M. Castelloe, SAS Institute Inc. and Ralph G. O'Brien, Cleveland Clinic Foundation, SUGI Proceedings, 2001.
Abstract
This presentation describes the steps involved in performing sample size analyses for a variety of linear models, both univariate and multivariate. As an analyst you must gather and synthesize the information needed, but you should be able to rely on the analytical tools to accommodate the numerous ways in which you can characterize and solve problems. Examples illustrate these principles and review relevant methods. User-written, SAS software-based programs already handle a wide variety of problems in linear models. Now, SAS Institute itself is developing software that will handle a rich array of sample size analyses, including all those discussed in this paper.