Course Information for Dr. Dan Voss, College of Science and Mathematics, Wright State University

STT 647, Winter 2007
 - Course Information
 - Syllabus
 - SAS example programs (subject to revision/updating/change): 
    ch19eg.txt, ch23ta01.txt, ch24eg.txt, ch25eg.txt, ch25pr15.txt, ch25pr17.txt, 
    ch25sec6.txt, ch25ta12.txt, ch27ta02.txt,  ch27ta12.txt,  nesting.txt,  ch27ta08.txt, 
    ch21eg.txt,  ch22eg1.txt,  ch22eg2.txt,  power.txt, 
 - doe checklist
 - Sample midterm exam and corresponding SAS program;
   Sample exam

 Homework Assignments: (To be determined as we go along)
 - ch19:  For your assigned data: (i) plan a FDA; (ii) do your planned FDA; and (iii) do an EDA.
    Provide a brief write-up of this, including SAS programs and output as an appendix.
 - ch24:  For your assigned data (i.e. the data of problem 24.6 or 24.9), do an exploratory data analysis.
    Check model assumptions for the full model, see what is significant in the analysis of variance, do
    whatever multiple comparisons are of interest based on the analysis of variance, and summarize the results.
 - ch25:  25.9a  (Use SAS proc glm to do the ANOVA)
              25.10bcde (For part e, just use Satterthwaite's approximation, doing this by hand using
                                 the output from proc glm to obtain a 95% (not 90%) confidence interval for the
                                 variance component, and also use proc mixed to verify the number of d.f. and resulting
                                 95% confidence interval obtained via Satterthwaite's approximation.
 - ch25:   Using the methods we have discuss for analysis of 2-factor mixed models in chapter 25, analyze the data
               for the wine-judging example (ch27, pg 1132, date in table 27.2).  Do the analysis using an unrestricted
               mixed model including interaction effects.  You will have zero error degrees of freedom, so there will be
               no residuals to plot.  Do the ANOVA, compare fixed effects of wine using Tukey's method with 95%
               confidence, and estimate and construct 95% confidence intervals for variance components if possible. 
               Try using both proc glm and proc mixed.
 - ch 27:  Conduct an exploratory data analysis of the data of exercise 17.11, using both the glm and mixed procs, and
               discuss the results

 - Data for problems by chapter:
    Chapter    Problems
        19        10, 12, 14, 16, 18, 20
        20          8
        22          7
        24          6, 9, 12
        25          9
        27          Table 27.2, 11 

-------------------------
Link to Dan's homepage

This page was lasted updated 02/20/2007.  Please send comments to Dan Voss.