From wxs@stat.washington.edu Mon Mar 3 10:46:13 1997 Received: from mx3.u.washington.edu by lists.u.washington.edu (5.65+UW96.08/UW-NDC Revision: 2.33 ) id AA38802; Mon, 3 Mar 97 10:46:12 -0800 Received: from calypso.stat.washington.edu (calypso.stat.washington.edu [128.95.17.161]) by mx3.u.washington.edu (8.8.4+UW96.12/8.8.4+UW96.12) with SMTP id KAA10181 for ; Mon, 3 Mar 1997 10:46:11 -0800 From: wxs@stat.washington.edu Received: by calypso.stat.washington.edu (950511.SGI.8.6.12.PATCH526/UW-NDC Revision: 2.26 ) id KAA28046; Mon, 3 Mar 1997 10:46:09 -0800 Date: Mon, 3 Mar 1997 10:46:09 -0800 Message-Id: <199703031846.KAA28046@calypso.stat.washington.edu> To: statgrads@stat.washington.edu, biostatgrads@biostat.washington.edu, students@math.washington.edu, cegrads@u.washington.edu, Cc: denis@redwood.rt.cs.boeing.com, chair@stat.washington.edu, peter@stat.washington.edu, kristin@stat.washington.edu Subject: Industrial Strength Course on Experimental Design Course Announcement Spring 1997 DESIGN AND ANALYSIS OF EXPERIMENTS Biostat/Stat 577 Instructor ---------- Denis G. Janky, Senior Principal Scientist, Applied Statistics Group, Boeing Information & Support Services, 865-3577, denis@redwood.rt.cs.boeing.com Time W/F 4:00 - 5:20 PM (a change to Tu/Th at this time is possible ---- if all students agree) Place Padelford C-301 ----- Prerequisites ------------- I will assume familiarity with basic statistics concepts, such as distributions, parameters, t-tests, etc. Some review material will be provided. Students with one or two introductory statistics courses should be adequately prepared. Overview -------- Experimentation is a central activity in science and engineering, and a working knowledge of modern methods for designing and analyzing experiments is essential. This course offers training in the technical aspects of experiment design and analysis, but always with the "real-life" application in mind. I will draw extensively on my 18 years of industrial experience (6 years in clinical trials research, and the remainder in engineering applications), and on material prepared for courses I taught to engineers at Boeing. I will assign weekly homework, giving students hands-on experience. Use of statistical software will be encouraged; I will provide temporary copies of the JMP package. JMP has a point-and-click interface and is particularly easy to use. I will conduct in-class software demonstrations. I also have substantial experience with SAS, and many examples will be illustrated using this software. Target Audience --------------- Graduate and advanced undergraduate students in science and engineering. Statistics or Biostatistics students are also welcome; for these students I hope to provide a useful conceptual link between linear models theory and applications. Content ------- This course is an introduction to the philosophy and mechanics of designing experiments and formulating conclusions from the resulting data. I will stress the link between design and statistical models. The primary designs covered will include completely randomized designs (one-factor, two-factor, etc.), randomized blocks designs, and the two-level factorial (2**k series) and fractional factorial (2**k-p series) designs. Other topics will include sample size determination and sequential augmentation of 2**k-p structures. The analysis methods covered will include analysis of variance (ANOVA), graphical analysis of residuals, transformations, and Daniel plots. Textbook and Course Materials ----------------------------- Box, Hunter and Hunter (1978). Statistics for Experimenters: An Introduction to Design, Data Analysis and Model Building. John Wiley & Sons (recommended). A course notebook containing copies of overhead viewfoils, and notes prepared by the instructor will be provided at minimal cost. The lectures are based on the course notebook. The Box, Hunter, and Hunter text provides similar material and is a time-honored reference. It is strongly recommended, but not required. .