651-303-0229

651-303-0229

  • Home
  • DOE Dispersion Simulator
  • Example Projects
  • FDA Validation Automation
  • References
  • Contact Us
  • More
    • Home
    • DOE Dispersion Simulator
    • Example Projects
    • FDA Validation Automation
    • References
    • Contact Us
  • Home
  • DOE Dispersion Simulator
  • Example Projects
  • FDA Validation Automation
  • References
  • Contact Us
Mercer Quality Consulting

Quality Improvement enabled by Statistical Thinking

Quality Improvement enabled by Statistical Thinking Quality Improvement enabled by Statistical Thinking

Analyzing a DOE with a Dispersion response and many Reps

 

  •  Most designed experiments (DOEs) seek to shift a mean response, but a few aim to minimize the dispersion of a response, using the standard deviation (STD) of replicates at each condition. Obtaining sufficient statistical power is often a challenge; dozens of replicates may be needed to detect a substantial reduction in dispersion.


  • In the problem which motivated this example many replicates were feasible.  A two-level factorial DOE studied a highly automated gage for measuring the axial strength of a mass-produced metal object. After 30 replicates per condition had been collected, the question arose whether to calculate one standard deviation estimate per condition (based on 30 raw data), two standard deviation estimates per condition (15 raw data each), three standard deviation estimates (10 each) etc.? 


  • A Monte Carlo simulation showed that the best arrangement of the raw data are often one that approximately balances the number of dispersion estimates and the dispersion subgroup size per each condition (e.g., five standard deviation estimates of subgroup size 6) for all but the most severely skewed raw data distributions. 


  • This website provides free access to a simulator which may be used by the reader to explore a variety of two-level factorial DOEs, as needed. This enables experimenters in any subject domain to design such DOEs with sufficient power, while making best use of the raw data.


  • This Monte Carlo simulator estimates the power to detect a stated difference between standard deviations. The Simulator allows you to investigate 2 to 7 factors from a variety of data distributions including Normal, Beta, Exponential, Gamma, Largest Extreme Value, Logistic, Smallest Extreme Value, and Weibull.


  • This is an application designed to work in Minitab versions 20, 21 and above. It uses a custom dialog box to select the simulation parameters and then run a Minitab Macro with the selected parameters.


1. Follow the first video below 

to see how to download and extract files.


2. Follow the Second video to see 

how to run the Minitab Application.


3. Watch the third video to see an example.



Download Minitab Macro and Dialog box files (zip)Download
CONTACT

Demo Videos not available yet!

Mike@MercerQualityConsulting.com



Contact Mike for direct help in using the simulator

Download and Extract

Download the "Download Minitab Macro and Dialog box files,zip" file and  extract the files to the C: drive

How to Run the application

0. Go to the C:\MtbMacros\DispDOESim\ folder.

1. Double click the Shortcut 

2. A Dialog box will open to allow you to select the parameters used in the simulation. 

3. Press the Create Script button 

4. Paste the script into Minitab's Command Line Pane 

5. Press Run.

6. Results appear in a few seconds

The Simulator in Action!

Examine how to determine the total number of data points allocated to a designed experiment to obtain at least an 80% Power to estimate the value of the Standard Deviation.

Simulator Dialog Box

You can choose DOEs with 4, 8, or 16 runs. 2^2, 2^3-1, 2^3, 2^4-1, 2^5-2, 2^4, 2^5-1, 2^6-2, 2^7-3.

You can choose from eight data distributions. Beta, Exponential, Gamma, Largest Extreme Value, Logistic, Normal, Smallest Extreme Value. Weibull.

Distributions parameters

How to compute 1st and 2nd parameters from Mean and Stdev

Download PDF

Copyright © 2023 Mercer Quality Consulting, LLC, - All Rights Reserved.


Powered by GoDaddy