Control charting normal variance – reflections, curiosities, and recommendations
Publication date
2010-01-01
Document type
Konferenzbeitrag
Author
Organisational unit
Advanced Mask Technology Center (AMTC), Dresden
Scopus ID
Conference
9th International Workshop on Intelligent Statistical Quality Control ; Beijing, China ; September 2007
Publisher
Physica-Verlag
Series or journal
Periodical volume
9
Book title
Frontiers in statistical quality control 9
First page
3
Last page
18
Part of the university bibliography
Nein
Language
English
Abstract
Following an idea of Box, Hunter & Hunter (1978), the consideration of the log of the sample variance S² became quite popular in SPC literature concerned with variance monitoring. The sample standard deviation S and the range R are the most common statistics in daily SPC practice. SPC software packages that are used in semiconductor industry offer exclusively R and S control charts. With Castagliola (2005) one new log based transformation started in 2005. Again, the search for symmetry and quasi-normality served as reason to look for a new chart statistic. Symmetry of the chart statistic could help in setting up two-sided control charts. Here, a comparison study is done that looks especially to the two-sided setup, straightens out the view of the available set of competing statistics used for variance monitoring and, eventually, leads to recommendations that could be given in order to choose the right statistic.
Version
Published version
Access right on openHSU
Metadata only access