MIT Robust System Design Session # 7 Analysis of Variance ANOVA.
-
Upload
laurel-newman -
Category
Documents
-
view
219 -
download
1
Transcript of MIT Robust System Design Session # 7 Analysis of Variance ANOVA.
![Page 1: MIT Robust System Design Session # 7 Analysis of Variance ANOVA.](https://reader037.fdocuments.net/reader037/viewer/2022103100/56649ef05503460f94c009e1/html5/thumbnails/1.jpg)
MIT
Analysis of VarianceANOVA
![Page 2: MIT Robust System Design Session # 7 Analysis of Variance ANOVA.](https://reader037.fdocuments.net/reader037/viewer/2022103100/56649ef05503460f94c009e1/html5/thumbnails/2.jpg)
MIT
Proposed Schedule Changes
•Switch lecture •No quiz – Informal (ungraded) presentation of term proj
ect ideas • Read Phadke ch. 7 --Construction Orthogonal
Arrays – Quiz on ANOVA – Noise experiment due
![Page 3: MIT Robust System Design Session # 7 Analysis of Variance ANOVA.](https://reader037.fdocuments.net/reader037/viewer/2022103100/56649ef05503460f94c009e1/html5/thumbnails/3.jpg)
MIT
Learning Objectives
• Introduce hypothesis testing
• Introduce ANOVA in statistic practice
• Introduce ANOVA as practiced in RD
• Compare to ANOM
• Get some practice applying ANOVA in RD
• Discuss / compare / contrast
![Page 4: MIT Robust System Design Session # 7 Analysis of Variance ANOVA.](https://reader037.fdocuments.net/reader037/viewer/2022103100/56649ef05503460f94c009e1/html5/thumbnails/4.jpg)
MIT
Hypothesis Testing
A technique that uses sample data from a population to come to reasonable conclusions with a certain degree of confidence
![Page 5: MIT Robust System Design Session # 7 Analysis of Variance ANOVA.](https://reader037.fdocuments.net/reader037/viewer/2022103100/56649ef05503460f94c009e1/html5/thumbnails/5.jpg)
MIT
Hypothesis Testing Terms
• Null Hypothesis (Ho) --The hypothesis to be
tested (accept/reject)
• Test statistic --A function of the parameters of the experiment on which you base the test
• Critical region --The set of values of the test statistic that lead to rejection of Ho
![Page 6: MIT Robust System Design Session # 7 Analysis of Variance ANOVA.](https://reader037.fdocuments.net/reader037/viewer/2022103100/56649ef05503460f94c009e1/html5/thumbnails/6.jpg)
MIT
Hypothesis Testing Terms (cont.)
• Level of significance (α) --A measure of confidence that can be placed in a result not merely being a matter of chance
• p value --The smallest level of significance at which you would reject Ho
p-value in a right-talied test
p-value in a two-talied test
![Page 7: MIT Robust System Design Session # 7 Analysis of Variance ANOVA.](https://reader037.fdocuments.net/reader037/viewer/2022103100/56649ef05503460f94c009e1/html5/thumbnails/7.jpg)
MIT
Comparing the Variance of Two Samples
• Null Hypothesis –
• Test Statistic –
• Acceptance criteria –
• Assumes independence & normal dist.
![Page 8: MIT Robust System Design Session # 7 Analysis of Variance ANOVA.](https://reader037.fdocuments.net/reader037/viewer/2022103100/56649ef05503460f94c009e1/html5/thumbnails/8.jpg)
MIT
F Distribution
• Three arguments
– d1 (numerator DOF)
– d2 (denominator DOF)
– x (cutoff)
![Page 9: MIT Robust System Design Session # 7 Analysis of Variance ANOVA.](https://reader037.fdocuments.net/reader037/viewer/2022103100/56649ef05503460f94c009e1/html5/thumbnails/9.jpg)
MIT
Rolling Dice
• Population 1 --Roll one die
• Population 2 --Roll two die
• Go to excel sheet “dice_f_test.xls”
![Page 10: MIT Robust System Design Session # 7 Analysis of Variance ANOVA.](https://reader037.fdocuments.net/reader037/viewer/2022103100/56649ef05503460f94c009e1/html5/thumbnails/10.jpg)
MIT
One-way ANOVA
• Null Hypothesis –
• Test Statistic –
• Acceptance criteria –
• Assumes independence & normal dist.
![Page 11: MIT Robust System Design Session # 7 Analysis of Variance ANOVA.](https://reader037.fdocuments.net/reader037/viewer/2022103100/56649ef05503460f94c009e1/html5/thumbnails/11.jpg)
MIT
ANOVA & Robust Design
Product / Process
Noise Factors
Control Factors
Signal Factor Response
H: Factor setting A1 is more robust than factor setting A2
Optimize robustness
H: This noise factor affects the mean
![Page 12: MIT Robust System Design Session # 7 Analysis of Variance ANOVA.](https://reader037.fdocuments.net/reader037/viewer/2022103100/56649ef05503460f94c009e1/html5/thumbnails/12.jpg)
MIT
ANOVA and the Noise Experiment
• Did the noise factors we experimented with really have an effect on mean?
• Switch to Excel sheet “catapult_L4_static_anova.xls”
![Page 13: MIT Robust System Design Session # 7 Analysis of Variance ANOVA.](https://reader037.fdocuments.net/reader037/viewer/2022103100/56649ef05503460f94c009e1/html5/thumbnails/13.jpg)
MIT
Why Test This Hypothesis?
• Factor setting PP3 is more robust than setting PP1 • Phadke --“In Robust Design, we are not concerned wit
h such probability statements, we use the F ratio for only qualitative understanding of the relative factor effects”
![Page 14: MIT Robust System Design Session # 7 Analysis of Variance ANOVA.](https://reader037.fdocuments.net/reader037/viewer/2022103100/56649ef05503460f94c009e1/html5/thumbnails/14.jpg)
MIT
Analysis of Variance (ANOVA)
• ANOVA helps to resolve the relative magnitude of the factor effects compared to the error variance
• Are the factor effects real or just noise?
• I will cover it in Lecture 7.
• You may want to try the Mathcad “resource center” under the help menu
![Page 15: MIT Robust System Design Session # 7 Analysis of Variance ANOVA.](https://reader037.fdocuments.net/reader037/viewer/2022103100/56649ef05503460f94c009e1/html5/thumbnails/15.jpg)
MIT
Additive Model
• Assume each parameter affects the response independently of the others
![Page 16: MIT Robust System Design Session # 7 Analysis of Variance ANOVA.](https://reader037.fdocuments.net/reader037/viewer/2022103100/56649ef05503460f94c009e1/html5/thumbnails/16.jpg)
MIT
Analysis of Means (ANOM)
• Analyze the data to discover mA1 , ai ...
![Page 17: MIT Robust System Design Session # 7 Analysis of Variance ANOVA.](https://reader037.fdocuments.net/reader037/viewer/2022103100/56649ef05503460f94c009e1/html5/thumbnails/17.jpg)
MIT
Analysis of Variance (ANOVA)
• Analyze data to understand contribution of control “error variance”
![Page 18: MIT Robust System Design Session # 7 Analysis of Variance ANOVA.](https://reader037.fdocuments.net/reader037/viewer/2022103100/56649ef05503460f94c009e1/html5/thumbnails/18.jpg)
MIT
Breakdown of Sum Squares
SS due to mean
GTSS
SS due to factor A
Total SS
SS due to factor B
SS due to error
etc.
![Page 19: MIT Robust System Design Session # 7 Analysis of Variance ANOVA.](https://reader037.fdocuments.net/reader037/viewer/2022103100/56649ef05503460f94c009e1/html5/thumbnails/19.jpg)
MIT
Breakdown of DOF
1 SS due to mean
n
(# levels) -1 factor A
n-1
(# levels) -1 factor B
DOF for error
etc.
n = Number of ηvalues
![Page 20: MIT Robust System Design Session # 7 Analysis of Variance ANOVA.](https://reader037.fdocuments.net/reader037/viewer/2022103100/56649ef05503460f94c009e1/html5/thumbnails/20.jpg)
MIT
Computation of Sum of Squares • Grand total sum of squares
• Sum of squares due to mean = nμ2
• Total sum of squares
• Sum of squares due to a factor
• Sum of squares due to error – Zero with no replicates – Estimated by “pooling”
![Page 21: MIT Robust System Design Session # 7 Analysis of Variance ANOVA.](https://reader037.fdocuments.net/reader037/viewer/2022103100/56649ef05503460f94c009e1/html5/thumbnails/21.jpg)
MIT
Pooling
• Provides an estimate of error without empty columns or replicates
• Procedure
– Select the bottom half of the factors (in terms of contribution to Total SS)
![Page 22: MIT Robust System Design Session # 7 Analysis of Variance ANOVA.](https://reader037.fdocuments.net/reader037/viewer/2022103100/56649ef05503460f94c009e1/html5/thumbnails/22.jpg)
MIT
F-statistic
factor for DOF
factor for SS factor for squaremean
variance Error
factor for squaremeanF
error for freedom of degrees
error to due squares variance Error
– F=1 Factor effect is on par with the error – F=2 The factor effect is marginal – F>4 The factor effect is substantial
![Page 23: MIT Robust System Design Session # 7 Analysis of Variance ANOVA.](https://reader037.fdocuments.net/reader037/viewer/2022103100/56649ef05503460f94c009e1/html5/thumbnails/23.jpg)
MIT
Confidence Intervals for Factor Effects
• Phadke
– Variance in ai is error variance / replication #
– 95% confidence interval for factor effects is two standard deviations in ai
• How does one interpret this value?
![Page 24: MIT Robust System Design Session # 7 Analysis of Variance ANOVA.](https://reader037.fdocuments.net/reader037/viewer/2022103100/56649ef05503460f94c009e1/html5/thumbnails/24.jpg)
MIT
ExampleCatapult Experiment
• Switch to Excel “Catapult_L9_2.xls”
![Page 25: MIT Robust System Design Session # 7 Analysis of Variance ANOVA.](https://reader037.fdocuments.net/reader037/viewer/2022103100/56649ef05503460f94c009e1/html5/thumbnails/25.jpg)
MIT
Homework
• Grades are exceptionally high
• Some are spending vast amounts of time
• This represents 20% of the final grade
Mean 94.2 101.0 96.5 95.4
Standard deviation 2.7 4.4 6.3 1.5
Maximum 98 109 101 97.3
![Page 26: MIT Robust System Design Session # 7 Analysis of Variance ANOVA.](https://reader037.fdocuments.net/reader037/viewer/2022103100/56649ef05503460f94c009e1/html5/thumbnails/26.jpg)
MIT
Quizzes
• Some consistently score high
• Others struggling, but learning
• Remember, this is only 10%
Quiz #1
Quiz #2
Quiz #3
Quiz #4
Mean 74.3 83.4 77.5 82.8
Standard deviation
19.3 10.4 22.3 17.6
Maximum 100 100 100 110
![Page 27: MIT Robust System Design Session # 7 Analysis of Variance ANOVA.](https://reader037.fdocuments.net/reader037/viewer/2022103100/56649ef05503460f94c009e1/html5/thumbnails/27.jpg)
MIT
Next Steps
• Hand in homework #5
• Homework #7 due on Lecture 10.
• Next session tomorrow
– Present your ideas for a term project
• Following session
– Quiz on ANOVA
– Homework #6 (Noise Exp.) due
– Constructing orthogonal arrays (read ch. 7)