STAT/ME 424 (Spring 2009)
Statistical Design of Experiments
for Engineers
Instructor: Professor Wei-Yin Loh
Send email to: loh@stat.wisc.edu
Prerequisites
At least one previous statistics course
Text
Design and
Analysis of
Experiments, 7th edition, by
D. C. Montgomery, Wiley
Course
Objectives
- Understand the importance of well-designed
experiments
- Learn how to design factorial experiments
with and without blocking, and with and without
replication or fractionation
- Learn how to fit models to data from
designed experiments
- Learn how to estimate the parameters in
the fitted models
Syllabus
Chapter 3: Oneway ANOVA
-
- Analysis of fixed effects
- Model checking
- Simultaneous comparisons
Chapter 4: Randomized blocks and
Latin squares
-
- Randomized complete block design
- Latin square design
- Graeco-Latin square design
Chapter 5: Introduction to
factorial designs
-
- Two-factor factorial design
- General factorial design
- Blocking in a factorial design
Chapter 6: Two-level factorial designs
-
- 2^2 design
- 2^3 design
- Replicated and unreplicated 2^k designs
- Addition of center points
Chapter 7: Blocking and
confounding in 2^k designs
-
- Blocking a replicated 2^k
design
- Confounding in a 2^k design
- Partial confounding
Chapter 8: Two-level fractional factorials
-
- Half and quarter fractions
- Design resolution
- General 2^(k-p) fraction
- Blocking fractional
factorials
- Resolution III and Plackett-Burman
designs
- Resolution IV and V designs
Chapter 9: Fitting regression models
-
- Linear regression
- Parameter estimation
- Hypothesis testing in multiple regression
- Confidence interval estimation
- Regression diagnostics
Chapter 10: Response surface methods (if time permits)
-
- Method of steepest ascent
- Analysis of second-order surface
- Designs for fitting response surfaces
Computing
- MINITAB, DESIGN-EXPERT, or JMP statistical software for homework
assignments and project data analysis
- Only pocket calculators are allowed in
exams
Grading
Course grade will be based on a weighted average
of standardized scores from
- Weekly homework assignments
- Two fifty-minute midterm exams
- An individual project involving the design
and analysis of a two-level factorial experiment
- A two-hour final exam
Return to instructor's homepage.
This page last updated on October 7, 2008