Lecture 6.1 - University of South Carolina
Download
Report
Transcript Lecture 6.1 - University of South Carolina
The Essentials of 2-Level Design of Experiments
Part I: The Essentials of Full Factorial Designs
Developed by Don Edwards, John Grego and James Lynch
Center for Reliability and Quality Sciences
Department of Statistics
University of South Carolina
803-777-7800
Part I.3 The Essentials of 2-Cubed Designs
Methodology
– Cube Plots
– Estimating Main Effects
– Estimating Interactions (Interaction
Tables and Graphs)
Statistical Significance (Effects
Probability Plots)
Example With Interactions
A U-Do-It Case Study
Methodology
23 Designs
23
Means What?
– 3 Factors (Usually Labeled A, B, C)
– 2 Levels Lo (-) and Hi (+)
– Comparing 8= 23 Recipes
Methodology
23 Designs - TV with Three Adjustment Knobs
Lo Lo Hi
Picture
Knob Setting
Is At The Top
Lo
Lo
Hi
A
B
C
Hi
Hi
Lo
Methodology
Tabular and Graphical Methodology
Cube Plots To See Relationships
Between The Response and Factor
Effects
Sign Tables To Estimate Factor
Effects
Probability Plots To Determine
Statistically Significant Factor Effects
Interaction Graphs and Tables To
Interpret Interactions
ANOVA Tables
Methodology
Cube Plot
Methodology
Cube Plot
Y
Y
7
8
Y
Y
4
3
+
Y
Y
5
B
A
Lo
Hi
Lo
Hi
Lo
Hi
Lo
Hi
Factors
B
Lo
Lo
Hi
Hi
Lo
Lo
Hi
Hi
C
Lo
Lo
Lo
Lo
Hi
Hi
Hi
Hi
Response
y
Y1
Y2
Y3
Y4
Y5
Y6
Y7
Y8
6
+
C
_
Y
Y
1
2
_
A
+
_
Note How The Responses are
Entered into the Cube
(Lo = - and Hi =+)
Methodology
Cube Plot
Y
Y
7
8
Y
-+-
4
+
6
+
C
_
+--
--A
C
Lo
Lo
Lo
Lo
Hi
Hi
Hi
Hi
Response
y
Y1
Y2
Y3
Y4
Y5
Y6
Y7
Y8
Y
--+
B
_
A
Lo
Hi
Lo
Hi
Lo
Hi
Lo
Hi
Factors
B
Lo
Lo
Hi
Hi
Lo
Lo
Hi
Hi
+
_
Note How The Responses
are Entered into the Cube
(Lo = - and Hi =+)
– Y1 is the Response when all the
Factors are Lo (- - -)
– Y2 corresponds to (+ - -),
Y3 to (- + -) and Y5 to (- - +)
Methodology
Cube Plot
+++
-++
++-
-++
--+
+
C
_
+--
--A
C
Lo
Lo
Lo
Lo
Hi
Hi
Hi
Hi
Response
y
Y1
Y2
Y3
Y4
Y5
Y6
Y7
Y8
+-+
B
_
A
Lo
Hi
Lo
Hi
Lo
Hi
Lo
Hi
Factors
B
Lo
Lo
Hi
Hi
Lo
Lo
Hi
Hi
+
_
Note How The Responses are
Entered into the Cube
(Lo = - and Hi =+)
– Y8 is the Response when all the
Factors are Hi (+ + +)
– Y4 corresponds to (+ + -),
Y6 to (+ - +) and Y7 to (- + +)
Methodology
Example 1: Targeting a Process/Reducing Variation
Methodology
Example 1 - Accuracy versus Precision
Not Accu rate
bu t
Precise
Accurate
an d
Precise
x xx x
x xxxx x
x x
x
xxxxx xx
x
xxx xx
Not Accu rate
an d
Not Precise
Accurate
bu t
Not Precise
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
Methodology
Example 1 - Improving a Process
Which Factors Affect
– Accuracy?
– Precision?
Methodology
Example 1 - Targeting a Process/Reducing Variation
Various Types of Significance
Statistical
Engineering
Economic