Stat664  Class Notes  Spring 2016


This page is under construction and is subject to changes. Click on the links for available classnotes. Please visit often for new materials.


R codes used to generate graphics for Class Notes A to D are archived in Stat6640NotesFigures.zip.


  1. Design of Experiments
    1. Controlled Experiments Versus Observational Studies
    2. Principles of Planning Experiments
  2. Review: Theory Review and Preparation
    1. Univariate/Multivariate Normal Distributions
    2. Central/Noncentral Chi-square Distributions
    3. Central/Noncentral F Distributions
    4. Some Important Results from Sampling Normal Distributions
    Also see Moment Generating Functions.
  3. Review: One-way ANOVA Models C1.pdf (§ 1 to § 3), C2.pdf (§ 4 to § 6), C3.pdf (§ 7 to § 8), C4.pdf (§ 9 to § 10), C5.pdf (§ 11 to § 12.5.10)
    1. Completely Randomized Experiments
    2. Models and Assumptions
    3. Estimations
    4. Diagnosis of Model Assumptions
    5. ANOVA Table
    6. Randomization and Hypothesis Tests
    7. Sample Size Determination for One-factor Design
    8. Inferences for Linear Combinations of Means
    9. Follow-up Study
    10. Robustness of ANOVA Analysis
    11. Welch t Statistic
    12. Simultaneous Inferences, Dunnett's upper fractiles ( dtable.05 (5%), dtable.01 (1%) )
  4. Two-way ANOVA Models D1.pdf (§ 1 to 2.12), D2.pdf (§ 3), D3.pdf (§ 4)
    1. Completely Cross Designs Versus Nested Designs
    2. Balanced Two-Way Fixed-Effects ANOVA Model
    3. Unreplicated Two-Way Fixed-Effect ANOVA Models
    4. The Need for Transformation and How to Do It
  5. Multi-way ANOVA Models (E.pdf)
  6. Variance Component Models (F.pdf). Also see a special discussion About Expected Mean Squares (ems.pdf) for unrestricted/restricted models.
  7. Nested Designs (G.pdf)
  8. Designs and Analysis of Experiments with One Or More Blocking Factors H.pdf (§ 1) H2.pdf (§ 2 to § 4) (updated)
    1. Latin Square Designs
    2. Graeco-Latin and Hyper-Latin Square Designs
    3. Balanced Incomplete Block Designs
    4. Youden Designs
  9. Two-Level Factorial Experiments. R Codes & Data Sets for 2-Level Designs.
  10. Response Surface Methodology, R Codes for Response Surface Designs.
  11. Experiments with Mixtures, R Codes for Experiments with Mixtures.

2016-01-11