BLOGS
StatsMadeEasy Blog
My wry look at all things statistical and/or scientific with an engineering perspective.
Also, see the Stat-Ease blog for tips on making DOE easy. For example, a recent posting provides my advice on “Strategy of Experiments for Formulations: Try Screening First!” Take a look!
|
|
New features released for Design-Expert® and
Stat-Ease® 360 software!
The Stat-Ease development team ushers in a new era with the newest version of Design-Expert (DX) and Stat-Ease 360 (SE360) software: the beginning of time-based releases per the “release early, release often” approach of the modern software industry. This first release (v22.0) features:
- Analysis Summary: expands the previous “Coefficients Table” with more model-fit statistics. Easily view p-values, R-squares, model equations and more, across all responses.
- Custom Graphs: node provides greatly expanded capability to the prior “Graph Columns.” You can now plot analysis data like predicted values and residuals—plus distinguish points by size and symbol.
- Shaded Intervals: Prediction, confidence, and tolerance intervals can now be set to “shaded” to color the bands, thus creating a far better view of the uncertainty around the fitted response.
- Hosted License Option: Run your software on multiple devices without requiring an on-premises license server, thus reducing costs for DIY software deployment and management.
- Direct Excel Import/Export (SE360 Exclusive): Avoid the hassle of copy and paste!
- Random Blocks (SE360 Exclusive): Blocks can now be treated as a random input. This facilitates precise statistical analyses for uncontrolled variables such as the day-by-day differences blocked out in this demonstration.
These are only a few of the many niceties provided in this latest iteration of DX/SE360. Contact our Sales team now to purchase a license or lay out the path for an upgrade to these leading-edge statistical tools developed by the world-class team of Stat-Ease programmers, statisticians, and engineers.
|
|
FAQ
Can factorial models be extrapolated?
Original question from a Researcher | Chemical R&D:
“We have recently purchased a Stat-Ease software license and need some help getting started. An Excel file with the relevant data is attached.”
Answer:
OK, I used the handy Import function to bring in this 23 factorial design with 2 center points. For maximum flexibility, the import feature applies polynomial models—apropos for response surface methods (RSM) but not two-level factorials.* I worked around that via Advanced Options on the Configure tab for each response—making them Factorial for modeling, thus taking advantage of DX's specialized tools for these designs. Then I analyzed each response. The first one, Yield, was most interesting—its Diagnostics indicating via the Box-Cox plot that an inverse square root transformation enhances the fit.
PS It is great you did center points for this characterization design, but next time do 3 or 4 to provide more power for the curvature test.
*(The Import tool assumes that data coming in stems from non-designed experiments that typically feature three or more levels for each factor, as discussed in my webinar on Evaluating and Exploiting Existing Data.)
Follow-up question:
"Thank you for your reply—it was very helpful. I was wondering whether it would be possible to do extrapolation to predict the yield (or any other response) using the model equation for factor ranges beyond what got tested in this DOE study?"
Follow-up answer:
The model, being strictly empirical, is not valid outside of the range tested. However, using it for speculative purposes may be merited—provided the predictions make some sense based on your subject matter knowledge. In that case, you may do well by doing a one-off run at the extrapolated conditions, and, if that works out, running a new two-level factorial with center points (but more of them!) in that previously unexplored region. Happy hunting for an even higher yield!
(Learn more about predictive modeling by attending a presentation of Modern DOE for Process Optimization.)
|
|
EVENT ALERT
Do you need a speaker on DOE for a 2023 learning session within your company or professional society at regional, national, or international levels? If so, please get back to me. – Mark
|
|
WEBINAR ALERT
Free webinars — Sign up now to take advantage
|
|
INFO ALERT
Import Data Mini-Tutorial
Importing data into Design-Expert and Stat-Ease 360 is very simple. Consultant Shari Kraber demonstrates how in this 5-minute YouTube video.
|
|
WORKSHOP ALERT
Sharpen up on DOE—Enroll before classes fill
You can do no better for quickly advancing your DOE skills than attending a Stat-Ease workshop. Our expert instructors provide you with a lively and extremely informative series of lectures interspersed by valuable hands-on exercises. Enroll early to ensure your spot!
|
|
“When a coincidence seems amazing, that's because the human mind isn't wired to naturally comprehend probability & statistics.”
—Neil DeGrasse Tyson
|
|
Stat-Ease, Design-Expert and Statistics Made Easy are registered trademarks of Stat-Ease, Inc.
Circulation: Over 4000 worldwide
|
|
|
|