Much work by W G Cochran might apply, and by Sir Ronald Fisher, as well as other historical developments, such as the Youden Square. I'm partial to work by Ken Brewer which applies with regard to the coefficient of heteroscedasticity as shown in 261596397_Ken_Brewer_and_the_coefficient_of_heteroscedasticity_as_used_in_sample_survey_inference, and traced by Ken, and in Cochran, W.G(1977), Sampling Techniques, 3rd ed., John Wiley & Sons, to work in Fairfield Smith, H.(1938), "An empirical law describing heterogeneity in the yields of agricultural crops," The J. Agri. Sci., 28, 1-23. The coefficient of heteroscedasticity, and methods to estimate it, are important to weighted least squares regression (of which OLS regression is an overused special case).
I know personally that Ken Brewer is a very good person, have heard that William Cochran was too. It also seems clear that many may have found Fisher to be cantankerous. But this question is not about personalities, but rather historical uses of modeling and statistical design for the benefit of agricultural research. Notes on the chronological order of events and discoveries are of interest. One discovery leads to another, but sometimes we may forget lessons learned, or need to know which ones to revisit. At the very least, I thought a question on the history of agricultural statistics could be entertaining.
Dear James! I hope the paper in the attachment will be Ok for you. Vladimir
E. E. Gbur et al. (2012, "Analysis of Generalized Linear Mixed Models", American Soc. of Agronomy) provide a brief historical description of the development of some statistical models, and cite Seal, H.L. 1967. The historical development of the Gauss linear model. Biometrika 54:1–24. with regard to the development of least squares
Thank you both for the references.
(Raul - Might either Gbur or Seal have anything available on ResearchGate?)
- Badges
- Science topic
- Similar topics
- Agriculture