March 1908
William Sealy Gosset Publishes “The Probable Error of a Mean” as “Student”
In March 1908, William Sealy Gosset published “The Probable Error of a Mean” in Biometrika under the name “Student.” Gosset worked for Guinness in Dublin, where costly brewery and barley experiments often produced only small sets of observations. His paper supplied a way to reason about uncertainty when a population’s variability had to be estimated from a small sample. The work grew from practical statistical problems at Guinness rather than from an academic post. It became the foundation of what is now called Student’s t-distribution and helped make carefully designed small-sample experiments useful in brewing, agriculture, and many other fields. Guinness permitted technical publication while requiring employees to use pen names; the surviving evidence supports the corporate policy, but not a more specific story about why Gosset chose the word “Student.”
Explore the Storylines
Tags
Sources
These sources support the dating and historical claims in this entry.
Gosset, William Sealy (“Student”). “The Probable Error of a Mean.” Biometrika 6, no. 1 (March 1908): 1–25.
Ziliak, Stephen T. “Guinnessometrics: The Economic Foundation of ‘Student’s’ t.” Journal of Economic Perspectives 22, no. 4 (2008): 199–216.
Physiological Society. “The Strange Origins of the Student’s t-Test.”
Continue Exploring
Related entries based on shared topics.