William Volk in his Preface pointed out that his take of applied statistics is traceable to a course he had taught in 1951. The McGraw-Hill Book Company published the first print in 1958. Its frontispiece reads: “for Dorothy whose confidence is without limits”. What a touching view on confidence without limits! I bought my first copy in the 1960s when I was working in the Port of Rotterdam. I have placed Jan Visman’s 1947 PhD thesis on coal sampling and William Volk’s Applied Statistics for Engineers side-by-side on the same bookshelf. I have tried to find out more about Volk after we had come to Canada in 1969 but to no avail. I wanted to write a Wiki page about Volk, his textbook, and his grasp of variances as displayed in Section 7.1.4 Variance of a General Function and in Section 7.3 Confidence Range of Variances. I still wonder whether or not Volk was of Dutch decent.
It was fortunate to have met Jan Visman and Greg Gould on ASTM Coal and Coke. I found out the hard way that TUDelft did not teach Visman’s take on coal sampling. I have learned most about Visman and his sampling theory after he went from Ottawa to Edmonton. I wrote a Wiki page about Dr Jan Visman and his work. His sampling theory underpins the interleaved sampling protocol. It was readily accepted for mineral concentrates simply because it gives a single degree of freedom for each sampling unit. ASTM recognized me for 25 years of continuous membership in 1995. Good grief! All I really did was do what I like to do! The more so since Matheron’s curse of his novel science of geostatistics had not yet impacted my work. On a trip to Australia I lost a copy of Volk’s Applied Statistics for Engineers. What’s more, Australian customs confiscated my bag with red and white beans. I knew a bit about toads and rabbits but never thought red and white beans were as bad. I used my beans to prove that large increments give a higher degree of precision than small increments. Simple comme bonjour! Testing for spatial dependence between measured values in ordered sets was straightforward with on-stream analyzers. When Volk wrote about the power of Student’s t-test to detect a bias, he showed how to derive Type I errors and Type II errors. I have taken to talking about Type I risks and Type II risks.
Gy’s 1979 Sampling of Particulate Materials and Volk’s 1980 Applied Statistics for Engineers still stand side-by-side on the same shelf. Gy had mailed me a copy with his compliments and his invoice on Christmas 1979. Elsevier Scientific Publishing Company has released it as Part 4 in Developments in Geomathematics. Part 1 is Agterberg’s 1974 Geomathematics and Part 4 is David’s 1977 Geostatistical Ore Reserve Estimation. Missing in Gy’s textbook between degenerate splitting processes and degree of representativeness are degrees of freedom. Gy points in Section 31.3 on page 381 to “a Student-Fisher’s t distribution with ν=N-1 degrees of freedom (DF)”. Gy was almost on the mark. His wordy gems struck me as Gy’ologisms! His references to Matheron’s new science of geostatistics are beyond the pale. Next to Gy’s 1979 Sampling of Particulate Materials, Theory and Practice on the same shelf stands David’s 1977 Geostatistical Ore Reserve Estimation, Journel & Huijbregts 1978 Mining Geostatistics, Clark’s 1979s Practical Geostatistics, and Mandel’s 1964 The Statistical Analysis of Experimental Data.
What Matheron and his disciples cooked up at a geostatistics colloquium on campus of The University of Kansas, Lawrence on 7-9, 1970 will not stand the test of time. Matheron conjured up Brownian motion along a straight line. His disciples stripped the variance off the distance-weighted average and called what was left a kriged estimate. Infinite sets of variance-deprived kriged estimates and zero kriging variances added up to the genuine scientific fraud of geostatistics.