Improve - 7 - Fractional Factorial Experiments - v12-1

Improve Phase
Fractional Factorial Experiments

Welcome to Improve
Process Modeling: Regression
Advanced Process Modeling: MLR
Designing Experiments
Experimental Methods
Designs
Full Factorial Experiments
Creation
Generators
Wrap Up & Action Items
Confounding & Resolution
LSS Black Belt v12 MT - Improve Phase 2 © Open Source Six Sigma, LLC
Why Use Fractional Factorial Designs
Fractional Factorial Designs are used to:

• Analyze factors to find cause/effect relationships if the Analyze Phase was unable
to sufficiently narrow the number of factors impacting the output(s).
– Fractional Factorials are often referenced as “ screening experiments” -- fewer runs with
larger number of factors.
– Fractional Factorials are usually done in early stages of the improvement process.
Fractional Factorial Design Full Factorial Design

StdOrder A B C D StdOrder A B C D
1 -1 -1 -1 -1 1 -1 -1 -1 -1
2 1 -1 -1 1 2 1 -1 -1 -1
3 -1 1 -1 1 3 -1 1 -1 -1
4 1 1 -1 -1 4 1 1 -1 -1
5 -1 -1 1 -1
5 -1 -1 1 1
6 1 -1 1 -1
6 1 -1 1 -1 7 -1 1 1 -1
7 -1 1 1 -1 8 1 1 1 -1
8 1 1 1 1 9 -1 -1 -1 1
10 1 -1 -1 1
11 -1 1 -1 1
12 1 1 -1 1
13 -1 -1 1 1
14 1 -1 1 1
15 -1 1 1 1
16 1 1 1 1
Why Use Fractional Factorial Designs
Fractional Factorial Designs are also used to:

• Reduce time and cost of experiments because the number of runs have been
lowered.
– As the number of factors increases the number of runs required to run a full 2 k factorial
experiment also increases (even without repeats or replicates)
• 3 factors: 2x2x2 = 8 runs
• 4 factors: 2x2x2x2 = 16 runs
• 5 factors: 2x2x2x2x2 = 32 runs etc….
• Be an initial experiment that can be augmented with another fraction to reduce

confounding and estimate factors of interest.
The answer is in there

somewhere!!
Nomenclature for Fractional Factorials
The general notation for k-p

Fractional Factorials is:
R
– k = number of factors to be investigated
– p = number of factors assigned to an interaction column (also called
“ degree of fractionating” with 1=1/2, 2=1/4,3=1/8, etc.)
– R = design resolution (III, IV, V, etc.). It details amount of confounding to
compare design alternatives
– 2k-p = the number of experimental runs
5-1
V
The example clarifies how to use the nomenclature.
• How many factors in the experiment?
• How many runs if no repeats or replicates?
• What fractional design is this (1/8, 1/4 or 1/2)?
Half-Fractional Experiment Creation
Recall the 2x2x2 full 3-factor, 2-level Factorial Design. Suppose we needed to investigate a fourth
factor but we could NOT increase the number of runs because of time or cost. Select the highest
order interaction to represent the levels of the fourth factor. The ABC interaction will determine
the levels for factor D.
When we replace the ABC interaction with factor D we say the ABC 3-way interaction was
aliased or confounded with D. This experiment maintains balance and orthogonality.
– The first experimental run in the first row indicates the experiment is executed with factor D at the
low level while running all the 3 other factors at the low level.
Factor D
A B C AxB AxC BxC AxBxC
-1 -1 -1 1 1 1 -1
1 -1 -1 -1 -1 1 1
-1 1 -1 -1 1 -1 1
1 1 -1 1 -1 -1 -1
-1 -1 1 1 -1 -1 1
1 -1 1 -1 1 -1 -1
-1 1 1 -1 -1 1 -1
1 1 1 1 1 1 1
This is a half-fraction 24-1 design - a Resolution IV design with only 8 runs.
Half-Fractional Experiment Creation
Why is the design, shown as orange rows, called a “half” fraction? This is the design just
created on the previous slide. This is a half fraction since a full 2x2x2x2 factorial would take
16 runs. With the half fraction we can estimate the effects of 4 factors in 8 runs. What is the
cost? We lose the ability to study the higher order interaction independently!
A B C D AxB AxC BxC AxBxC

-1 -1 -1 -1 1 1 1 -1
1 -1 -1 -1 -1 -1 1 1
-1 1 -1 -1 -1 1 -1 1
Half Fraction: 1 1 -1 -1 1 -1 -1 -1
-1 -1 1 -1 1 -1 -1 1
Alias Structure: 1 -1 1 -1 -1 1 -1 -1
D = ABC -1 1 1 -1 -1 -1 1 -1
Note D settings 1 1 1 -1 1 1 1 1
-1 -1 -1 1 1 1 1 -1
are the same as 1 -1 -1 1 -1 -1 1 1
the ABC -1 1 -1 1 -1 1 -1 1
1 1 -1 1 1 -1 -1 -1
interaction -1 -1 1 1 1 -1 -1 1
1 -1 1 1 -1 1 -1 -1
-1 1 1 1 -1 -1 1 -1
1 1 1 1 1 1 1 1
Could we create a quarter fraction experiment out of the above matrix

and still study four factors at once?
Why or why not?
Graphical Representation of Half-Fraction
We have discussed half-fractional Experimental Designs for 4 factors:

The graphical representation shows the 8 runs we created on the previous 2
slides.
- A + - A +
Top line of previous slide
-
- C
+
B
-
+ C
+
- D +
Remember that D is confounded with the ABC interaction in this half-
fractional design.
Design Generators
Design Generators are an easier technique to use than to generate the

Fractional Factorial Designs by hand as done in the previous slides.
Design Generators help us EASILY find the confounding within the

Fractional Design.
A Design Generator is the mathematical definition for how to begin

aliasing a Full Factorial to create a Fractional Factorial.
Example of a Design Generator:
Design Generator D = ABC
This means the D column is the same as the ABC

interaction column; they cannot be distinguished from
each other so are called “confounded”.
Design Generators
Design Generator D = ABC

• Because of the Design Generator we can now fill out the D column
– For each row of D, multiply the values in the columns of A, B and C
together and create the column
• You may correctly suspect some 2-factor interactions are confounded
• Create contrast columns for AD, BD, CD using a similar technique used to
create the column for D
A B C AB AC BC D AD BD CD
-1 -1 -1 1 1 1
1 -1 -1 -1 -1 1
-1 1 -1 -1 1 -1
1 1 -1 1 -1 -1
-1 -1 1 1 -1 -1
1 -1 1 -1 1 -1
-1 1 1 -1 -1 1
1 1 1 1 1 1
MINITABTM Session Window
What does this mean?
Fractional Factorial Design
Factors: 4 Base Design: 4, 8 Resolution: IV

Runs: 8 Replicates: 1 Fraction: 1/2
Blocks: none Center pts (total): 0
Design Generators: D = ABC

Alias Structure
I + ABCD
A + BCD
B + ACD
C + ABD
D + ABC
AB + CD
AC + BD
AD + BC
So What is “Confounding”?
When using the “ Confounding” or “ Alias” pattern we assume the higher order
interactions in a Confounded effect are not significant.
– Sparsity of effects principle indicates that higher order interactions are
very rare.
• “While interactions are important they do not abound…, interactions that
are more complex than those involving two factors are rare.” Thomas B.
Barker
In the past example the D factor was Confounded with the ABC 3-way
interaction. When the effect is assigned to D which is Confounded with ABC we
assume because of the sparsity of effects principle the effect is entirely because
of the D factor.
Remember when two items such as an interaction with a Main Effect are
Confounded one cannot distinguish if the statistical significance is a result of the
Main Effect or the interaction or a combination.
Aliasing is another term for “Confounding”.
Confounded Effects With Fractionals
MINITABTM will automatically generate the Same levels

alias structure which lists all the Confounded
Effects. Alias Structure
Note: For this case ~ AA BC
BC ABC
ABC
+1
+1 +1
+1 +1
+1
– A is Confounded with BC I + ABC -1
-1 -1
-1 +1
+1
– B is Confounded with AC -1
-1 -1
-1 +1
+1
– C is Confounded with AB A + BC
The Confounding means any effect noted B + AC B AC ABC
cannot be specifically assigned to either of the -1 -1 +1
Confounded factors. C + AB
+1 +1 +1
– Remember we will use the sparsity -1 -1 +1
principle. +1 +1 +1
Note: This is a level III design and is C AB ABC

NOT recommended since Confounding -1 -1 +1
exists between Main Effects and 2- -1 -1 +1
factor interactions. +1 +1 +1
+1 +1 +1
I’m certainly confounded!!
Experimental Resolution
Remember R in the nomenclature referenced the Resolution. k-p

This useful visual aid remembers definitions of the Confounding
designated by the Resolution. 2R
Resolution III Fully Saturated Design
Hold up Three Fingers, One on one
hand and Two on the other. This
illustrates the Confounding of main
Main Effects Two Way Interactions effects with two way interactions.
Resolution IV
Next hold up four fingers
The Confounding is main effects with
three way interactions or…
Main Effects Three Way Interactions
Two way interactions Confounded with

other two way interactions.
Two Way Interactions Two Way Interactions

Experimental Resolution
k-p
R
The visual aid is shown through Resolution V.
Resolution V
Hold up Five Fingers, One on one hand and
Four on the other. This illustrates the
Confounding of main effects with four way
Main Effects Four Way Interactions interactions or …
Two way interactions Confounded with

three way interactions.
Two Way Interactions Three Way Interactions
MINITABTM Fractional Factorial Design Creation
Fortunately, MINITABTM creates the designs for us to prevent having to create a

Fractional Factorial by hand. This output, found in the MINITAB TM Session Window
after creating a Fractional Factorial design, should be understood because it also
informs us of the Resolution of the design.
Stat>DOE>Factorial>Create Factorial Design … 4 factors, Designs, ½ fraction
Fractional Factorial Design
Factors: 4 Base Design: 4, 8 Resolution: IV

Runs: 8 Replicates: 1 Fraction: 1/2
Blocks: none Center pts (total): 0
Design Generators: D = ABC

Alias Structure
I + ABCD
A + BCD
B + ACD
C + ABD
D + ABC
AB + CD
AC + BD
AD + BC
(5 -1)
2V Fractional Design Resolution V
Example of a very useful Fractional Design often used for screening designs.
Run A B C D E
1 -1 -1 -1 -1 1
2 1 -1 -1 -1 -1
3 -1 1 -1 -1 -1
E
4 1 1 -1 -1 1
5 -1 -1 1 -1 -1
6 1 -1 1 -1 1
7 -1 1 1 -1 1
B
8 1 1 1 -1 -1
C
A 9 -1 -1 -1 1 -1
D 10 1 -1 -1 1 1
11 -1 1 -1 1 1
Pros Cons 12 1 1 -1 1 -1
13 -1 -1 1 1 1
5 factors (Main Effects) 16 trials to get 5 Main Effects 14 1 -1 1 1 -1
10 2-way interactions 2nd order interactions are 15 -1 1 1 1 -1
Main Effects only Confounded Confounded with 3rd order 16 1 1 1 1 1
with rare 4-way interactions
MINITABTM’s Display of Available Designs
Fractional Designs
are colored boxes
without “Full”.
Note: Since we discourage Design Resolution III or IV, MINITABTM has shaded
these as RED and YELLOW for cautionary. GREEN is acceptable because Main
Effects are not Confounded with lower level interactions.
DOE Methodology
1. Define the Practical Problem

2. Establish the Experimental Objective
3. Select the Output (response) Variables
4. Select the Input (independent) Variables
5. Choose the Levels for the input variables
6. Select the Experimental Design
7. Execute the Experiment and collect data
8. Analyze the Data from the designed experiment and draw
Statistical Conclusions
9. Draw Practical Solutions
10. Replicate or Validate the Experimental Results
11. Implement Solutions
Just follow these simple steps…..
Fractional Factorial Example
1. Define the Practical Problem

• 8 factors are of interest in increasing the output but process knowledge is
limited because of a previously poor gage for the output
2. Establish the Experimental Objective
• The output is to be maximized
3. Select the Output Variables
• The output is labeled Y and has a Gage R&R % study variation of less than
5%
4. Select the Input Variables
• The Input Variables are simply labeled A through H
5. Choose the Levels for the Input Variables
• For simplicity sake of this exercise the Levels can be expected to be
appropriately set and we will only work with coded levels
Fractional Factorial Example (cont.)
6. Select the Experimental Design

 Select the appropriate design in MINITAB TM and create this exact worksheet in columns
C1 through C12.
 We have no reason to believe curvature exists and are satisfied no replicates are
required.
 For ease of this exercise be sure NOT to have randomized the experiment.
7. Execute the Experiment and Collect the Data

 Select the appropriate design in MINITAB TM and create this exact worksheet in
columns C1 through C12.
 We have no reason to believe curvature exists and are satisfied no replicates are
required.
 The resources and time allow us to only run the experiment with 16 treatment
combinations or experimental runs.
ile
Wf
t o . MT
er ix
f – Ref E Matr
f O
l e cuto ll FF D
Ta b f u
For
8. Analyze the Data and draw Statistical Conclusions

 Before doing any analysis, let’s review what Confounding exists in this highly
fractionated Factorial Design
 The Main Effects are Confounded with numerous 3-way interactions
 The 2-way interactions are Confounded with numerous 2-way interactions
 This is important and must be remembered in our analysis.
f ile
. M TW
o
er t ix
f – Ref E Matr
f O
l e cuto ll FF D
Ta b f u
For
We want 95% confidence in our Statistical Conclusions for this example.
We have generated the initial Pareto of effects.
 The AC 2 factor interaction is Confounded with other 2-way interactions but we will
assume for now using the Confounding table from MINITAB TM the 2-way AC interaction
is actually the EH 2 factor interaction because both factors E and H are significant.
 The second highest effect for a 2 factor interaction AF. We will look at the Confounding
table and assume it is the BE 2-way interaction since the B and E factors are significant.
 The 2-way interaction AE also is significant with the alpha above 0.1. We cannot find
another 2-way interaction that might be significant using just the B, E and H factors.
 If the AE interaction is kept in the model then to maintain “hierarchical order” factors A
and E must be kept in the model.
 We will now reduce the model to see its impact.
The Reduced Model is shown here and we want 95% confidence to include terms.
Notice the AE 2-way interaction has the smallest effect of the statistically significant terms and
factor A kept in the model to maintain the “ hierarchical order” also has a small term and is
statistically insignificant. We choose to reduce the model and remove those terms. R-sq
should not be severely impacted. If it was impacted severely we would reconsider this choice.
Factorial Fit: Y versus A, B, E, H

Estimated Effects and Coefficients for Y (coded units)
Term Effect Coef SE Coef T P

Constant 22.001 0.04381 502.21 0.000
A 0.144 0.072 0.04381 1.64 0.139
B 4.939 2.469 0.04381 56.37 0.000
E 12.921 6.461 0.04381 147.48 0.000
H -6.246 -3.123 0.04381 -71.29 0.000
A*E -0.351 -0.176 0.04381 -4.01 0.004
B*E -3.836 -1.918 0.04381 -43.78 0.000
E*H 8.244 4.122 0.04381 94.09 0.000
S = 0.175232 R-Sq = 99.98% R-Sq(adj) = 99.96%

Analysis of Variance for Y (coded units)
Source DF Seq SS Adj SS Adj MS F P
Main Effects 4 921.55 921.545 230.386 7502.91 0.000
2-Way Interactions 3 331.20 331.198 110.399 3595.34 0.000
Residual Error 8 0.25 0.246 0.031
Total 15 1252.99
The further refit model shows an adequate model because:

 Simplicity of terms; which is desired but NOT required
 R-sq is quite high (overly unusual for practical experiments)
 No or few unusual observations which would be noted below the
ANOVA in MINITABTM’s Session Window
 Residuals are appropriate
Factorial Fit: Y versus B, E, H

Constant 22.001 0.07167 306.98 0.000
B 4.939 2.469 0.07167 34.46 0.000
E 12.921 6.461 0.07167 90.15 0.000
H -6.246 -3.123 0.07167 -43.58 0.000
B*E -3.836 -1.918 0.07167 -26.76 0.000
E*H 8.244 4.122 0.07167 57.51 0.000
S = 0.286673 R-Sq = 99.93% R-Sq(adj) = 99.90%

Analysis of Variance for Y (coded units)
Source DF Seq SS Adj SS Adj MS F P

Main Effects 3 921.46 921.462 307.154 3737.52 0.000
2-Way Interactions 2 330.70 330.705 165.352 2012.05 0.000
Residual Error 10 0.82 0.822 0.082
Lack of Fit 2 0.10 0.099 0.050 0.55 0.597
Pure Error 8 0.72 0.722 0.090
Total 15 1252.99
The Residuals Analysis is adequate and appropriate because:

 The residuals are concluded to be Normally distributed
 No pattern for Residuals in the Versus Fits or Versus Order analysis
Statistical Conclusions to maintain terms in the model must consider:

 Maintaining hierarchical order
 A 2-way interaction must have the involved factors in the model also
 High statistical confidence with the P-value less than your alpha risk
 A higher R-sq or model explanation of the process changes is desired
 Proper residuals and few to no unusual observations
No, no unusual
observations here…
9. Draw Practical Solutions

 We must remember our Experimental Objective is to increase the output Y.
 Looking at the positive coefficient for B and E we know if we put those factors
at the high level or value of +1 the output increases
 Looking at the negative coefficient for H we would think we should operate at
the low level or value of -1. However the 2-way interaction of EH shows a
coefficient that is larger and would result in a net decrease in the output of Y so
we must set H to a +1 or the high level.
 A big reminder is we have ASSUMED the 2-way interactions involved the
factors we left in the model.
Factorial Fit: Y versus B, E, H

Constant 22.001 0.07167 306.98 0.000
B 4.939 2.469 0.07167 34.46 0.000
E 12.921 6.461 0.07167 90.15 0.000
H -6.246 -3.123 0.07167 -43.58 0.000
B*E -3.836 -1.918 0.07167 -26.76 0.000
E*H 8.244 4.122 0.07167 57.51 0.000
S = 0.286673 R-Sq = 99.93% R-Sq(adj) = 99.90%
It can be difficult to optimize the solutions to get the Practical Solution desired.
Using Response Optimizer in MINITABTM helps us find the Practical Solution of setting
the factors left in the model all at the high level or +1.
Stat>DOE>Factorial>Response Optimizer…..set goal to maximize
Practical Conclusions to keep in the model include:

 Simple models can be useful depending on the project or process
requirements
 Terms with practically large enough significance even if statistically
significant
 Impact of R-sq by removing a term with low effects
 Ability to set and control the controllable inputs in the model may decide
on the use of terms
 Robust designs or minimal variation requirements may require close
inspection of interactions’ effects on the Y
 If multiple outputs are involved in the process requirements, balancing of
requirements will be necessary
That’s a lot of juggling….
10. Replicate or Validate the Experimental Results

 After we have determined with 95% statistical confidence we must replicate
the results to confirm our assumptions; such as which 2-way interactions
were significant among the Confounded ones
 If the results do not match the expected results OR the project goal further
experimentation may be needed
 In this case we were able to achieve 29.8 on average with the process setting
of E, B and H and so the results are considered successful in the project
We win, we win…!!
11. Implement Solutions

 Work with the process owners to develop the Control Plans to
sustain your success
Fractional Factorial Exercise
Exercise objective: Open file “ bhh379.mtw” and analyze

using the 11 Step methodology.
1. What kind of Factorial Design is this?
2. Generate Factorial Plots in MINITABTM.
3. Create the Statistical and Practical model.
Summary
At this point you should be able to:
• Explain why & how to use a Fractional Factorial Design
• Create a proper Fractional Factorial Design
• Analyze a proper model with aliased interactions
Not that kind of model!!

Improve - 7 - Fractional Factorial Experiments - v12-1

Uploaded by

Copyright:

Available Formats

Improve - 7 - Fractional Factorial Experiments - v12-1

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Improve - 7 - Fractional Factorial Experiments - v12-1

Uploaded by

Copyright:

Available Formats

Improve Phase

Fractional Factorial Experiments

Process Modeling: Regression

Advanced Process Modeling: MLR

Fractional Factorial Designs are used to:

Fractional Factorial Design Full Factorial Design

Fractional Factorial Designs are also used to:

• Be an initial experiment that can be augmented with another fraction to reduce

The answer is in there

The general notation for k-p

This is a half-fraction 24-1 design - a Resolution IV design with only 8 runs.

A B C D AxB AxC BxC AxBxC

Could we create a quarter fraction experiment out of the above matrix

We have discussed half-fractional Experimental Designs for 4 factors:

Design Generators are an easier technique to use than to generate the

Design Generators help us EASILY find the confounding within the

A Design Generator is the mathematical definition for how to begin

Example of a Design Generator:

Design Generator D = ABC

This means the D column is the same as the ABC

Design Generator D = ABC

What does this mean?

Fractional Factorial Design

Factors: 4 Base Design: 4, 8 Resolution: IV

Design Generators: D = ABC

Aliasing is another term for “Confounding”.

MINITABTM will automatically generate the Same levels

Note: This is a level III design and is C AB ABC

I’m certainly confounded!!

Remember R in the nomenclature referenced the Resolution. k-p

Two way interactions Confounded with

Two Way Interactions Two Way Interactions

Two way interactions Confounded with

Two Way Interactions Three Way Interactions

Fortunately, MINITABTM creates the designs for us to prevent having to create a

Fractional Factorial Design

Factors: 4 Base Design: 4, 8 Resolution: IV

Design Generators: D = ABC

1. Define the Practical Problem

Just follow these simple steps…..

1. Define the Practical Problem

6. Select the Experimental Design

7. Execute the Experiment and Collect the Data

8. Analyze the Data and draw Statistical Conclusions

We want 95% confidence in our Statistical Conclusions for this example.

We have generated the initial Pareto of effects.

 We will now reduce the model to see its impact.

Factorial Fit: Y versus A, B, E, H

Term Effect Coef SE Coef T P

S = 0.175232 R-Sq = 99.98% R-Sq(adj) = 99.96%

The further refit model shows an adequate model because:

Estimated Effects and Coefficients for Y (coded units)

Term Effect Coef SE Coef T P

S = 0.286673 R-Sq = 99.93% R-Sq(adj) = 99.90%

Source DF Seq SS Adj SS Adj MS F P

The Residuals Analysis is adequate and appropriate because:

Statistical Conclusions to maintain terms in the model must consider:

9. Draw Practical Solutions

Factorial Fit: Y versus B, E, H