Quality Magazine

Gage R and R Studies Online

May 19, 2003
Identifying and reducing measurement variation is the whole reason for doing gage repeatability and reproducibility (R and R) studies. To help novices and experts, this month's Web review looks at sites related to Gage R and R studies. These sites explain how to conduct a study and provide step-by-step solutions to common Gage R and R problems.

www.itl.nist.gov/div898/handbook/mpc/section4/mpc4.htm. This page comes from the Engineering Statistics Handbook, section 2.4, which discusses Gage R and R studies. The page is set up in an outline format with links to other sections in the handbook. For example, clicking on the first question: What are the issues for a Gage R and R study?, a page labeled 2.4.1 comes up. Each topic, such as Repeatability, Reproducibility, Stability and Bias all link to other pages or sections in the handbook that describe what each term is and how it relates to Gage R and R. Click Stability on the Gage R and R main page, then click Resistivity probes under the term case study. What comes up is the purpose and an outline of linked terms for that study. If a visitor scrolls down the Gage R and R main page, there are linked goals that explain how to test resolution of instruments or estimate the differences between gages. Or, there are other questions to click on, such as what are the design considerations for the study? How do we collect data? How do we quantify variability of measurements? There is also a link to Remedies and strategies. A visitor will find good information regarding Gage R and R studies. This is a good site to bookmark for a reference.

www.engineeredsoftware.com. This Web site is sponsored by Engineered Software Inc. and gives step-by-step solutions to many common problems relating to Gage R and R. On the home page, click Reliability Problems in the sidebar. A page comes up displaying links to problems and white papers, such as determining the optimum burn-in duration or how the Weibull distribution can be used to model populations with multiple failure modes. Or view white papers that demonstrate how to reduce testing requirements by over 50%, or explain why exponential distribution is often misused. A visitor can even type in a problem in the space provided to be examined and addressed by one of the company's engineers. On the home page, click on Reliability Problems or Gage Capability Problems to get to this feature. Either submit a problem or click on one of the problems listed. Links on the Gage Capability Problems page cover bias studies, determining repeatability and reproducibility, designing a repeatability and reproducibility study, and determining linearity. Each of these links to a step-by-step example. There is also a software demo download available.

www.uwstout.edu/ctem/technology/beed/mfge320/Gage%20R&R%20Present_files/frame.htm. This Web site is a Gage R and R measurement study. It briefly describes the concept and method used by some Detroit automobile manufacturers. On the sidebar is a list of the topics covered in this study, such as what is a measurement study, additivity of variances, gage error, operator error, part error, gage and operator error and total measurement system error. If a visitor clicks on the index card at the bottom left-hand side of the page, it will show the key points under each topic. For example, under gage error, it explains how gage error is used interchangeably with repeatability or equipment variation. It goes on to explain other key points as well. The same information is covered in the slide presentation, except that the slides have graphs. To advance through the slides, click the arrows at the bottom of the page. There are a total of nine slides. This type of site allows visitors to leisurely look through this method. For more detailed information, look up the book Analysis of Gage R and R Studies, A Review of Industry Methods, by P. Prond, Manufacturing Systems Engineering, University of Wisconsin-Madison 1991.

www.qualityamerica.com/knowledgecente/articles/PAKRepeatAnalysis.htm. This site is part of Quality America Inc.'s Web site. There is a monthly column on Repeatability Analysis, Gage R and R Studies and Calibration Studies. The author is Paul A. Keller and for his brief biography and e-mail address click his name. In a recent column, Keller gives an example using three operators measuring 10 pieces three Arial. Within the document, there is a link to the data used for the example. Clicking on another link, QA-Calibrate, brings up a page that provides an overview of software that can be used to do Gage R and R. Not quite sure how to do a Gage R and R study? Click the link titled how to conduct an R and R study. On this page, Keller gives the definition of repeatability and reproducibility, and explains how to do a study. Within this document is a link to calibration studies. By clicking this link, visitors can read about how to maintain an effective calibration control system. To find out what is meant by tampering, click the word. A page comes up explaining what tampering is and what it does to the process.

Site to Hide
Did you know that the first Peanuts comic strip appeared in seven U.S. daily newspapers on October 2, 1950? For more Peanuts trivia, archived comic strips, collector information, games and other fun features, check out www.snoopy.com.

Need Some Help?
Quality Online visitor, Halton Reginald needs some help. Here's the query: "How can I implement an effective corrective and preventive action program?" Can someone help Halton or any of our Quality Online visitors? Go to the Reader Forum, www.i-boards.com/bnp/qm/default.asp, and post your quality and manufacturing problems, or try to help a fellow visitor.

Feedback
From "Reader Forum" visitor Ven Batchu.

Q: Here is my question. Company "A" manufactures, assembles and supplies to Company "B", and Company "B" also manufactures and assembles a similar product. Company "A" and Company "B" have the same component suppliers. Company "B" asked Company "A" to supply one component from Company "A"'s inventory, along with a product part approval process (PPAP) submission. Company "A" supplied the component along with PPAP documents except for a part submission warrant. The part submission warrant could not be submitted because there is a column to fill in entitled "Suppliers Manufacturing Location." Company "B" has rejected Company "A"'s shipment and asked for a corrective action. What is Company "A" supposed to do? How can Company "A" submit a part warrant when he is not a manufacturer of the component? I am looking for a solution.

An answer came from Quality Online visitor Raymundo Herrera.

A: According to QS-9000 4.2.4.2, Subcontractors Requirements, you should utilize PPAP with your supplier, so just ask for the part submission warrant and all paperwork related to your component supplier. Then forward this information to Company "B".

This comment is in response to a question and answer that was published in the June 2001 Feedback section (p. 14). The question and answer have been reprinted below for reference.

Here is the question from "Reader Forum" visitor Tom.

Q: I am looking for a software package that generates certificate of analysis or conformance sheets. Most products collect data but are not designed to print sheets to send to customers.

Here is the answer from Quality Online visitor Joe Perito.

A: I'm sure there is software available to do this, but I fail to see the reason to buy any. It's simple enough to design your own certificate of analysis in any spreadsheet, such as Lotus 1-2-3. Keep a copy of the blank form and fill it in for each new lot. Fill in a line with the lot number and save the file under that lot number, customer name or date. With the automatic sorting features of today's spreadsheets, you can recall the fill names, sort by lot numbers, find lot numbers, sort by customers, dates or lot numbers in ascending or descending order. You probably already have a spreadsheet program. With a little bit of time, you'll have a database, plus a variety of sorting options, automatic calculations, macros and reporting options.

Here is a comment to this question and answer from Quality Online visitor Paul Heckman.

A: I agree with Joe Perito that there isn't a good reason to buy a software package to print certificates. However, I don't agree with using spreadsheet software to do this. This is really a database function. Most database programs, such as Microsoft Access, allow for easy setup of forms and reports, as well as the data tables. This format is much more searchable for data to do other analysis with than a spreadsheet format. Prior to switching to Access, we had over 3,000 spreadsheet certificates that were virtually impossible to link together for other data manipulation.