In the nearly 100 years since Ronald Fisher first published his book on The Design of Experiments, statisticians and engineers have developed innovative applications and new ways to integrate planned experimentation into their work. In this panel, five experts will discuss modern DOE methods that are spreading through the research & development communities. What's been helpful, and what's on the horizon for the future of DOE?
Speaker bios available by clicking the link above.
This presentation makes the case for modeling both mean and standard deviation to achieve on target results with minimal variation. It demonstrates benefits via examples where experimenters took advantage of making multiple measurements for every run in their design. Newcomers to statistical design of experiments (DOE) often overlook this opportunity to achieve more robust operating conditions. Attend this webinar to master DOEs aimed at meeting specifications and doing so with utmost reliability.
Can’t make this time? Register anyway so that you are notified when the recording is ready.
Date: Wednesday, September 25, 2024
Time: 10:00am Central US Time
Optimize your products and processes with accurate prediction models. Learn how to get the most out of your RSM design by following a few key analysis steps. See how automated model-reduction tools build simpler models that predict more precisely. Then discover how diagnostics confirm your model’s validity. Finally, learn how key statistics like lack of fit and various R-squared measures characterize the polynomial model.
Discover what you need to know about the diagnostic plots used to validate an analysis of variance (ANOVA). Learn a little about how these plots are made and more importantly, how to interpret the signals that indicate problems with the analysis. In this webinar, we will explore how careful residual analysis can be key to a successful DOE.
This webinar details incredibly useful assessments provided by Stat-Ease software for evaluation of any set of input data, whether existing (unplanned) or from a ‘proper’ design of experiments (DOE). Learn how to watch for issues that degrade the information that you hope to extract and strengthen your ability to assess your data quality!
Discover methods for creating experiment designs progressively so that knowledge can be gained steadily via iterative steps. Learn how to augment completed designs that fall short of adequately modeling the critical response(s). This might salvage a great deal of experimental work that would otherwise go for naught.
This webinar provides valuable insights on Stat-Ease® 360 software’s special modeling tools for binary data, counts, and deterministic results (such as those collected from computer simulations). The focus will be on the practical aspects, with minimal emphasis on theory and technical details.
Learn how Python has been integrated into Stat-Ease 360. This tutorial walks through connecting Python, extracting data from SE360, and some other more complex examples.
Discover DOE tools aimed at developing systems that hold up when transferred to the field. It features factorials geared for testing many variables in a minimum number of runs—just enough to reveal effects that may lead to failure.
Pat Whitcomb, Stat-Ease founder, illustrates how to take best advantage of designs geared for hard-to-change process settings. While running through a number of case studies with Design-Expert® software, he provides statistical details and practical advice on the pluses and minuses created by the split-plot factor layout.
This talk deals with thorny issues that confront every experimenter: How to handle results that fit badly with your chosen model. Design-Expert software provides graphical tools that make it easy to diagnose what is wrong—damaging outliers and/or a need for transformation.
Pat Whitcomb reveals some tricks for making the most of your DOE.
- Using std error to constrain optimization
- Using Cpk to optimize your DOE
- Combining categoric factors
- Using diagnostics to uncover analysis problems
How to use automatic model selection tools to build on appropriate models. Pros and cons of the methods are discussed.
- Forward/Backward/Stepwise
- p-value, AICc, BIC, All Subsets