2.0 EXPERIMENTS COMPARING TWO VALUES
(updated Spring 2005)
A new car engine has been developed. It is desired to determine
if the new engine (B) is superior to the engine currently in use
(A) in terms of MPG.
A wealth of information has been developed ove

14.5 ORTHOGONAL ARRAYS
(Updated Spring 2003)
Orthogonal Arrays (often referred to Taguchi Methods) are often employed
in industrial experiments to study the effect of several control factors.
Popularized by G. Taguchi. Other Taguchi contributions include:

12.0 FRACTIONAL FACTORIAL DESIGNS
(Updated Spring,2001)
We now need to look at a class of designs that are good for
studying many variables in a relatively limited number of tests,
fractional factorial designs.
Previously we saw that if we carefully selec

13.0 BICYCLE EXAMPLE
(Updated Spring 2005)
Interested in time it takes to pedal up a hill. 7 variables of interest:
Variable
Low
High
1-Set
Up
Down
2-Dynamo
Off
On
3-Handlebars
Up
Down
4-Gear
Low
Medium
5-Raincoat
On
Off
6-Breakfast
Yes
No
7-Tires
Hard
So

8.0 TWO-LEVEL FACTORIAL (2K) DESIGNS
(UPDATED SPRING, 2005)
Surface finish of a part produced by a turning process is of interest (Ra value in IN)
How is the surface finish affected, if at all, by the feed rate and
the presence/absence of coolant?
Examine

11.0 MODEL BUILDING
(Updated Spring 2005)
Model Building Steps:
1. Postulation of a tentative model form:
y = b 0 + b 1 x 1 + b 2 x 2 + b 12 x 1 x 2 +
This is a linear model with interaction term. If we add second order terms,
then a full second order mo

9.0 TAGUCHIS CONTRIBUTIONS
(Updated Spring 2001)
Model of the Engineering Design Process
Decomposition of the factors that influence product/process
performance
Emphasis on variation reduction-use of the loss function and signalto-noise ratio
Robust d

4.0 Two-Way ANOVA
(Updated Spring 2005)
Design of Experiment techniques are widely applicable within
engineering and the sciences. We will now consider an example
that comes from the Food Science area.
Hot Dog Example
Recent Federal Regulations have relax

7.0 MODEL DEVELOPMENT
(Updated Spring 2005)
A response dependent variable is truly related to a set of
independent variables by a mathematical function
y = f ( x 1, x 2, x 3, , x n : 1, 2, )
x 1 x n independent or input variables
1 , 2 set of parameters.

3.0 ANOVA - ANALYSIS OF VARIANCE
(updated Spring 2005)
In our previous discussions, we considered two levels for the
variable under study (i.e., engine type A versus engine type B).
We will now examine a technique that can be used for
examining multiple (

5.0 TWO-WAY ANOVA WITH NO REPLICATION
(Updated Spring, 2001)
The effect of different medical treatments on the survival times of animals
given different poisons is under investigation. 12 animals were randomly
assigned to the 3 Poisons and 4 Treatments. T

DESIGN OF EXPERIMENTS
1.0 STATISTICS REVIEW
(updated Spring 2005)
Random Variable
Discrete random variable: Number of up spots on a throw die; Exam
score; etc.
Continuous random variable: Time between car arrivals at a spot light
(11.3, 51.2 etc.); Diamet

14.0 RESPONSE SURFACE METHODOLOGY (RSM)
(Updated Spring 2001)
So far, weve focused on experiments that:
Identify a few important variables from a large set of candidate variables, i.e., a screening experiment.
Ascertain how a few variables impact the resp