By the Numbers
Organizing Data for Quality Control: Manual 7A
What can we learn from test data? The answer is quite a bit, if the information is presented properly. Those charged with organizing test result data can find guidance about the process in the Manual on Presentation of Data and Control Chart Analysis, or Manual 7A, which takes the reader step by step through presenting test information “in a concise form more readily interpretable than the unorganized mass of original data.”
One example used throughout the manual is the transverse strength of brick tested according to C 67, Test Methods for Sampling and Testing Brick and Structural Clay Tile. The manual’s first part illustrates how the data can be organized to obtain quantitative information about the brick or other manufactured products. The second part shows how confidence limits can be determined for the mean and standard deviation of the brick’s transverse strength. With the third part, control charts related to creating the bricks can be established and evaluated to determine the adequacy of the manufacturing process for transverse strength.
Collectively, the manual provides assistance using the examples of brick transverse strength, iron sheet coating thickness and copper wire breaking strength so that manufacturers can control the quality of any product through connecting specifications, production and inspection in an ongoing process.
The quality of the manual is critical, just as understanding the quality of a product is important for manufacturers or others testing that quality. “While Manual 7A is not an ASTM standard, E11 members have meticulously edited each revision and treat this document as if it were to be processed as a standard. Knowing that other committees refer to Manual 7A, E11 believes it is very important to be sure that what is printed there is correct,” said Dean Neubauer, senior engineering associate, Corning Inc., and adjunct professor, Center for Quality and Applied Statistics at the Rochester Institute of Technology, who led the most recent revision to the work.
The work’s organization includes clearly delineated sections with summaries and references for additional information. The three parts consist of:
Today’s three-part manual grows out of the original work, ASTM Manual on Presentation of Data, published in 1933, as well as two supplements added in 1935. In 1937, a single volume first combined all three parts in one reference.
A special committee, organized in what was then Committee E01 on Methods of Testing, developed the manual. The group included such leaders in statistics and quality control of the time as H.F. Dodge, “the father of acceptance sampling,” and H.G. Romig, a key architect of modern day sampling plans who worked with Dodge on a number of seminal publications in the field.
Since that early work, E11 has revised the manual several times, but the core material holds true to the original. “The basic statistical methods for describing and illustrating a data distribution have remained largely intact over the many years,” Neubauer said. “Of course, as new methods were developed, the most useful ones were incorporated into the manual to keep it up to date.”
Overall, the volume encompasses easily understood graphics and statistical method explanations used in quality control. “Manual 7A is fairly inexpensive compared to the vast majority of textbooks containing similar information,” Neubauer said. “It should appeal to both ASTM members as well as nonmembers who want a well-written introduction to these basic statistical methods.”