Statistics 590

Fall 2005

Steve Buyske

buyske@stat.rutgers.edu

Grades are posted on the FAS Gradebook (as well as with the registrar).

• Syllabus
• Errata for textbook
• Office Hours
• Homework
• SAS code from 9/26 class
• First Take Home Quiz Due Oct 10
• Solutions for First Take Home Quiz
• SAS code for Quiz 1
• SAS code from 10/3 class
• SAS code from 10/10 class (factorial example)
• Review problems for Mid-term Exam
• Solutions for Midterm Exam
• Quiz and exam grades can be found at the FAS Gradebook site
• SAS code from 10/31 class
• SAS code for mixed effects designs
• SAS code for a nested design
• SAS code for randomized complete block designs
• SAS code for latin square designs
• Second Take Home Quiz Due Dec 5 Use the restricted model for all problems
• SAS example in an unbalanced design
• SAS example of a balanced incomplete block design
• SAS example of a split plot design
• SAS example of analysis of covariance
• SAS example using proc mixed for inhomogeneity of variance
• Quiz 2 Solutions
• Final Exam chapter coverage and problems for review
• Other links of interest:
  • Downloadable Datasets from the text
  • SAS stuff:
    • A rough guide to using SAS in the CCF computer labs.
    • It's possible to purchase your own license for SAS
    • There's also the SAS Learning Edition. Be sure to get version 2.
    • SAS on PCs in Rutgers labs
    • SAS on eden (batch mode) Note on high resolution graphics for batch mode: adding the line: goptions rotate = landscape device=ps; near the beginning will generate a postscript file with the graphics. Writing device=gif instead will generate a gif file.
    • RU general documentation on using Unix
    • General SAS documentation (from RU addresses only)
    • General SAS V8 documentation
    • General SAS V9 documentation
  • QQ plot examples of normal data. Enter qqnorm(rnorm(20)) in the window and hit submit. After that you can just reload the resulting web page.
  • F (central and noncentral) calculations For a p-value, enter 1 - pf(f, df1, df2) in the window (where you put in the actual numbers for f, df1, and df2, and hit submit. If you want the noncentral F, include the term ", ncp = xx" where you put the value of the noncentrality parameter in for xx. For a critical value, use qf(0.95, df1, df2)
  • On line calculator for central F distribution.
  • On line calculator for noncentral F distribution.
  • On line calculator for F, Chi-squared, and other distributions.
  • Russ Lenth's power calculators
  • MIchael Friendly's ANOVA power calculator

To read PDF files (such as the quiz) you need Adobe's Acrobat Reader: