Stat 441 Assignments.

• Homework 1. (Due 01/29).

• Homework 2. (Due 02/07)

• Homework 3. (Due 02/14)

• Homework 4.

• Homework 5.

• Homework 6.

• Homework 7.

• Homework 8.

• Homework 9.

• Homework 10.

Stat 541 Assignments.

• Homework 1. (Due 01/29)

• Homework 2. (Due 02/07)

• Homework 3. (Due 02/14)

• Homework 4.

• Homework 5.

• Homework 6.

• Homework 7.

• Homework 8.

• Homework 9.

• Homework 10.

Course notes.

• (1) Introduction

• (2a) One-factor CRD, ANOVA, estimation of effects.

• (2b) Inferences: one mean, differences in means, contrasts.

• (2c) Multiple comparison procedures.

• (2d) Random effects model and variance component estimation.

• (2e) HOV Tests, transformations, weighted least squares.

• (2f) Simulations to study ANOVA HOV assumption.

• Linear models and matrix examples.

• (3a) Randomized complete block designs (RCBD).

• (3b) Simple repeated measures designs (SRMD).

• (3c) Latin square designs and balanced incomplete block designs (RCBDs).

• (4a) Two-factor factorial experiments.

• (4b) Multiple comparison tests for two-factor factorial experiments.

• (4c) 2 x 3 factorial experiment.

• (4d) Tukey's Test for Nonadditivity, three-factor factorial, factorial with block, exact and approximate F-tests.

• (4e) Two-factor factorial example using weighted least squares and with a transformation.

• (5a) Designs with nested effects.

• (5b) Replicated latin square designs.

• Designs with split plotting.

Examples using SAS.

• Sleep deprivation example with contrasts.

• Multiple comparisons.

• ANOVA for random batches of raw materials and variance component estimation using Proc Varcomp.

• Loom Data. Bartlett's, Brown-Forsythe's, and Levene's HOV tests.

• Weighted least squares example

• Determining a transformation using the empirical method and using a Box-Cox procedure.

• Cotton fiber breaking strength randomized complete block design.

• RCBD analyses with and without blocks.

• RCBD with missing observation. Type I vs Type III analyses.

• Simple repeated measures design (Winer data).

• Latin square design example.

• Balanced incomplete block design (Box, Hunter, and Hunter data).

• Two-factor factorial design (2x2).

• Two-factor factorial with multiple comparisons.

• Example revisited. HOV test and Bonferroni MCT using Proc Glm.

• Two-factor factorial design (3x2).

• Two-factor factorial designs -- random effects model and mixed effects model.

• Tukey's Test for Nonadditivity.

• Three-factor factorial design.

• Two-factor factorial design with blocks.

• Using SAS to generate expected mean squares table for three factor factorial designs -- (i) all factors random (ii) two factors fixed, one factor random.

• Two-factor factorial example using weighted least squares.

• Two-factor factorial example with a transformation.

• Two-stage nested design.

• Three-stage nested design.

• Design with nested and crossed effects.

• Design with nested and crossed effects.

• Replicated latin square design (RLSD-1).

• Replicated latin square design (RLSD-3).

• Replicated latin square design (RLSD-2) (Fleiss data).

• Split plot design.

• Split-split plot design.

• Machines with operators example.

• Example Revisited: Time trend adjusted multiple comparisons.

Examples using R.

• Multiple comparison tests.

• ANOVA for random batches of raw materials and variance component estimation.

• R code for HOV tests.

• R code for Box-Cox transformations.

• R code to perform simulations to study the ANOVA HOV assumption.

• R code for RCBD.

• R code for RCBD with missing observation.

• R code for Model without run order.

• R code for model with run order.

• Data set used in R code for models with and without run order.

Exam I Data Sets.

• Problem 1.