What's New
SIMULATOR 2.2 RELEASED – 10/18/2022
This is free to all registered users of 2.1 and 2.0.
Fixes
- Updated copy to clipboard of graphic windows to accommodate high-resolution screens with a scale set higher than 100%. This ensures the bounding box is the same size as the graphic so the graphic does not have to be cropped.
- Added references for simulations to the help system under What Does Simulator Do?
OPTIMIZATION AND VARIATION REDUCTION – SECOND EDITION (PDF and Hardcopy) – 10/1/2022
Covers how to optimize the average and reduce variation during product design, process design and process improvement. Provides a unified approach to engineering product and process variation.
This landmark book integrates statistical process control, design of experiments, robust design methods, Shainan’s methods and other lesser-known practices into a single system for engineering variation and optimizing performance. Professionals engaged in product design, process development, manufacturing and Six Sigma improvement activities will find this common framework has vast practical implications, not least of which is immensely improved communication between the various disciplines in the product lifecycle.
STATISTICAL PROCEDURE FOR THE MEDICAL DEVICE INDUSTRY – VALIDATIONS OF BOOK’S SPREADSHEETS COMPLETED – 9/1/2018
The validations of spreadsheets accompanying the book Statistical Procedures for the Medical Device Industry are complete and can be downloaded. There are two versions of each spreadsheet. It is the protected version that was validated. The protected version does not allow changes to be made other than data entered into the designated fields. The unprotected version allows you to access the formulas and make modifications, so cannot be validated.
REDONE WEB SITE – 9/1/2018
The website has been completely redone. All content has been retained. The former technical library is now the Blog. A Forum has been added to ask questions and for comments and suggestions. The store is now integrated with the website. The magnifying glass on the main menu can be used to search the entire website, making it easier to find items of interest. A good place to start is the methods page. There everything available on a given topic is listed including books, software, courses, blog topics and forum topics.
STATISTICAL PROCEDURE FOR THE MEDICAL DEVICE INDUSTRY – 9/25/17
Provides statistical procedures designed to ensure that all statistical regulatory requirements for medical devices are met. These regulatory requirements can be found in:
- Code of Federal Regulation 21CFR, Part 820 Quality System Regulations
- Medical devices – Quality management systems – Requirements for regulatory purposes
Available as both a hardcopy and pdf.
This book includes step-by-step instruction for executing 43 statistical procedures including confidence intervals, equivalency studies, regression, ANOVA, designed experiments, capability studies, gauge R&R studies, normality testing, sampling plans, control charts, change-point analysis, tolerance analysis and robust design (complete list). It also provides all the needed rationales, justifications and references.
3 NEW TECHNICAL BRIEFS ON TRENDING IN BLOG – 9/25/17
These technical briefs support STAT-10, Statistical Techniques for Trending Data, from the book Statistical Procedures for the Medical Device Industry.
Normalized IN charts solve the issue of not being able to normalize Individuals (I) charts. It is particularly useful for handling count data like complaints when counts are above 10. While similar to the Laney U’ and P’ charts, it can be used in other situations not involving count data. Average () and IN charts may be the only 2 types of charts you ever need.
The exception to the above statement “Average () and IN charts may be the only 2 types of charts you ever need,” is for nonnormal data. One common case of this is count data were the counts average 10 or less, such as complaint data. The U and P charts are commonly used in this case. However, for low counts both have increased rates of false signals, possibly as high as 1 in 11. Adjusted control limits are provided in these two briefs fixing this issue.