This module builds on and focuses on more applied statistical tools needed if you are going to use the tools on your own in the real world. The classical textbook approach from module 1 is often not sufficient for real world challenges.
This is an advanced course and the second of three within the topic of data analysis and DoE. This module builds on Data-driven root cause analysis and design of experiments (DoE) 1 and focuses on more applied statistical tools needed for data analysis and designed experiments. The classical textbook approach from module 1 has its limitations and may not be sufficient when working with data heavy real world challenges.
You will learn to establish the mathematical relationships between critical quality attributes (CQA) and critical process indicators (CPI) so a process can be monitored online predicting final product quality. You will also learn to establish the relationships between CPI and critical process parameters (CPP) so you are able to act if quality prediction is bad. Knowing these relationships is a prerequisite for product design, process optimisation and process control.
This course is relevant for you if you are e.g. working with development, optimisation or process/product trouble shooting. You want to be able to use and understand the statistical tools behind hypothesis testing, regression analysis and Design of Experiments independently in real life challenges.
If you are participating in a Six Sigma Green Belt or Black Belt course, this module will be voluntary. The covered tools will become useful when working with data analysis on your DMAIC project.
After participating in this course, you will be able to:
- Validate your statistical model by performing residual analysis in multiple regressions and lack of fit
- Reduce your DoE model so only terms with predictive power for new observations are left
- Use logistic regression to analyse discrete data
- Minimise number of tests and optimising processes with Design of Experiment (DoE)
- Calculate sample sizes on mean to reach sufficient power of a designed experiment
- Establish the relationships between CQA, CPI and CPPs for the CPPs that survived the screening in a model experiment
- Use historical data to build, analyse and reduce mathematical models
- Un-equal variances
- Matched pairs
- Logistic regression
- Linearity studies
- Data transformation
- Model reduction
- Residual analysis in multiple regression
- Sample size on mean value – power calculation
- Lack of fit
- DoE - response surface model
- DoE catapult exercise – finding optimal solution
The preferred software for this course is JMP from SAS and the instructors will demonstrate all of the exercises in JMP. You are also welcome to use Minitab for the exercises where it is possible and will still be able to receive assistance from our instructors.
However, even if you are currently not a JMP user, we recommend you to install JMP on your computer before participating in the advanced courses, since some of the exercises will only be possible to perform with JMP.
This is an advanced course and the second of three within the topic of data analysis and DoE. The next level module in within this topic is the following.
The module Data-driven root cause analysis and design of experiments (DoE) 1 is the introductory module within this topic and is a prerequisite for this module.
If you want to devote more focus on the development of performance and processes, you can get inspiration from methodical courses such as Design for Six Sigma or Quality Function Deployment offered by Storm Management. These courses will help you map and operationalise customer demands and develop future processes.
Furthermore, you have the opportunity to continue your education as a Six Sigma Green Belt or Black Belt, and become more competent and able to optimise, lead and develop processes.
Venue: NNE's headquarters, Nybrovej 80, 2820 Gentofte, Denmark
Price: DKK 6,000 + VAT (includes course materials, refreshments and lunch)
Language of instruction: The course will be held in English unless all participants speak and understand Danish
Duration: 1 day
Course dates in 2018
- March 7
- May 23
- September 12
- November 21