费希尔《实验设计》统计学思想探析
An Exploration of the Statistical Thought in R. A. Fisher’s The Design of Experiments
DOI: 10.12677/sa.2026.158180, PDF,   
作者: 胡 勇:西北大学科学史高等研究院,陕西 西安;陕西省文化遗产数字人文重点实验室,陕西 西安
关键词: 费希尔实验设计女士品茶统计学R. A. Fisher Experimental Design The Lady Tasting Tea Statistics
摘要: 费希尔1935年出版的《实验设计》被公认为现代实验统计学的奠基之作。本文以“女士品茶”这一经典案例为切入点,从“实验作为认识工具”的视角重新审视费希尔的统计学思想。研究表明,“女士品茶”不仅是随机化检验与显著性检验的直观示范,更是费希尔对其“归纳推断逻辑”的理论演绎。在这一案例中,随机化程序、零假设检验与确切的概率计算三者构成一个自洽的认识论闭环,揭示了费希尔实验设计思想的深层结构:以随机化消除系统性偏倚,以重复估计误差,以显著性检验实现“从数据到推断”的归纳逻辑。这一思想途径奠定了当代科学实验的方法论基础,并深刻影响了波普尔以来的科学哲学发展。
Abstract: R. A. Fisher’s The Design of Experiments, published in 1935, is universally acknowledged as a foundational work of modern experimental statistics. Taking the classic case of “the lady tasting tea” as the entry point, this paper re-examines Fisher’s statistical thought from the perspective of “experiment as an epistemic tool”. The study demonstrates that “the lady tasting tea” serves not only as an intuitive demonstration of randomization tests and significance tests, but also as a theoretical deduction of Fisher’s “logic of inductive inference”. In this case, the randomization procedure, null hypothesis testing and exact probability calculation form a self-consistent epistemological closed loop, revealing the deep structure of Fisher’s thought on experimental design: eliminating systematic bias through randomization, estimating error through replication, and realizing the inductive logic of “from data to inference” through significance testing. This intellectual approach laid the methodological foundation for contemporary scientific experiments and exerted a profound influence on the development of the philosophy of science since Karl Popper.
文章引用:胡勇. 费希尔《实验设计》统计学思想探析[J]. 统计学与应用, 2026, 15(8): 67-75. https://doi.org/10.12677/sa.2026.158180

参考文献

[1] Fisher, R.A. (1935) The Design of Experiments. Oliver & Boyd.
[2] Lew, R. (2014) Experimental Design. In: Wiley StatsRef: Statistics Reference Online, Wiley.
https://doi.org/10.1002/9781118445112.stat05892
[3] Kempthorne, O. (1980) The Teaching of Statistics: Content versus Form. The American Statistician, 34, 17-21.
https://doi.org/10.1080/00031305.1980.10482704
[4] Salsburg, D. (2001) The Lady Tasting Tea: How Statistics Revolutionized Science in the Twentieth Century. W.H. Freeman and Company.
[5] 陈希孺. 数理统计学简史M. 长沙: 湖南教育出版社, 2002.
[6] Rosenberger, W.F. (2026) The 41st Fisher Memorial Lecture from Fisher to CARA: The Evolution of Randomization and Randomization-Based Inference. Journal of the Royal Statistical Society Series A: Statistics in Society, 189, 497-511.
https://doi.org/10.1093/jrsssa/qnaf002
[7] Hacking, I. (1988) Telepathy: Origins of Randomization in Experimental Design. Isis, 79, 427-451.
https://doi.org/10.1086/354775
[8] Lehmann, E.L. (2011) Fisher, Neyman, and the Creation of Classical Statistics. Springer.
[9] Box, J.F. (1978) R.A. Fisher: The Life of a Scientist. Wiley.
[10] van Doorn, J., Matzke, D. and Wagenmakers, E.J. (2020) An In-Class Demonstration of Bayesian Inference. Psychology Learning & Teaching, 19, 36-45.
https://doi.org/10.1177/1475725719848574
[11] Mugnier, M. (2024) R.A. Fisher’s Exact Test Revisited.
https://doi.org/10.48550/arXiv.2407.07251
[12] Hemerik, J. and Goeman, J.J. (2021) Another Look at the Lady Tasting Tea and Differences between Permutation Tests and Randomisation Tests. International Statistical Review, 89, 367-381.
https://doi.org/10.1111/insr.12431