Many thanks to those who responded to "Does anyone know who in the 1940s / 1950s whilst not challengeing Shewhart's thinking on limits proved that R-bar from range data can be used to calculate the standard deviation used in SPC charts please?"
The responses I received are below:
a.. try following this thread http://www.itl.nist.gov/div898/handbook/pmc/section3/pmc321.htm seems to point towards Patnaik, 1946
b.. In "Modern Methods for Quality Control and Improvement", 1986 by Harrison M. Wadsworth, Kenneth S. Stephens and A. Blanton Godfrey on page 147 they write under Ranges 'Early interest in the range and its properties including quality control applications was show by a series of papers by:
a.. Tippet (1925) On the extreme individuals and the range of samples taken from a Normal population, Biometrika, Vol 17, p 364
b.. Pearson (1926) A further note on the distribution of range in samples taken from a Normal distribution, Biometrika, Vol 18 p 173
c.. McKay and Pearson (1935) A note on the distribution of range in sample sizes of n, Biometrika, Vol 18. p 173
d.. Pearson and Hartley (1942) The probability integral of the range in samples of n observations from a Normal population, Biometrika, Vol 32, p 301
e.. Hartley (1942). This work has generated tables giving the mean value and standard deviation of the range from random samples drawn from the normal distribution as well as tables of the probability integral of the range. Since the range is a measure of the universe dispersion, two significant relationships have evolved from the referenced works as follows.
f.. (1) the standard deviation of individuals equals the true average range divided by constant of proportionality factor (d2)
g.. (2) the standard deviation of individuals equals the standard deviation of the range divided by the constant of proportionality factor (d3) where d2 and d3 are functions of the sample size."
c.. Have a look at Davies, O. and Pearson,E. JRSS Supl 1934. Owen Davies and Egon Pearson looked at in the 1930s as far as I am aware.
d.. I have J.N.R. Jeffers, "Use of Range/Standard Deviation Tables", Forestry 1952 25(1):66-68, see
http://forestry.oxfordjournals.org/cgi/content/abstract/25/1/66
Note that Tippett was responsible for the tables.
The link to this article is "Tables of Range and Studentized Range" by H. Leon Harter, Annals of Mathematical Statistics, v31, p.1122-1147, 1960
e.. http://www.jstor.org/pss/2983598 and is titled, "Methods of Estimating from Samples the Population Standard Deviation."
f.. I looked up my copy (one of my favourite stats books) of Pearson and Hartley's 'Biometrika Tables for Statisticians, Vol 1'. (1970, but first edition 1954)
Page 45, section 12.5, gives expressions for the expectation and variance of the range of samples of size n, and these are attributed to Pearson (1902) and Tippett (1925). Tippett (1925) published tables for n = 2(1) 1000.
Shewhart (1931) quotes the Tippett paper in 'Economic Control of Quality', page 202.
If any of this is part of what you are looking for, let me know if you'd like me to forward a scan of the relevant pages.
Tippett (1925) is 'On the Extreme Individuals and the Range of Samples taken from a Normal Population', Biometrika, XVII, 364-387.
g.. I think that L. H. C. Tippett, " On the extreme individuals and the range of the samples taken from a normal population," Biometrika, 17 (1925), 364-387may be an important early paper. Other papers worth looking at might be: -
H. O. Hartley, The range in random samples, Biometrika vol. 32 (1942) pp. 334-348
E. S. Pearson and H. O. Hartley, The probability integral of the range in samples of n observations from a normal population, Biometrika vol. 32 (1942) pp. 301-310. If Tippett did the early work why is d2 referred to as Hartley's constant?
|