| Roger Lee Berger | |
| Birth Date: | August 21, 1951 |
| Birth Place: | Kansas |
| Nationality: | American |
| Alma Mater: | Purdue University |
| Occupation: | statistician and professor |
Roger Lee Berger is an American statistician and professor, co-author of Statistical Inference, first published in 1990 with collaborator George Casella.[1]
Roger Lee Berger was born August 21, 1951, in rural northwest Kansas. His early interest in mathematics ripened as he attended the University of Kansas, switching from physics to mathematics with emphasis on real analysis as an undergraduate. Graduating in 1973, he studied graduate statistics at Purdue University, earning a master's degree in 1975 and a doctorate in 1977.[2] Building on the work of his advisor Shanti Swarup Gupta in subset selection,[3] Berger wrote Minimax, Admissible, and Gamma-Minimax Multiple Decision Rules to complete his doctoral studies.[4]
Berger's first professor appointment was at Florida State University in 1977. Later, he served at North Carolina State University and then Arizona State University, where he retired in 2017.[5] Throughout his career, he supervised eleven students in statistics on topics ranging from theory, such as exact hypothesis testing, more powerful tests with linear and elliptical constraints, to applied biostatistics. In 2003, Purdue deemed Berger a Distinguished Alumnus.[6]
Berger and George Casella met at Purdue and became fast friends, studying statistics together. In 1983 at the Eastern North American Region (ENAR) conference of the International Biometric Society, Casella asked Berger to co-author a new master's level introductory text to statistical inference, hoping to revise and improve Hogg and Craig's Introduction to Mathematical Statistics. This led to the publication of Statistical Inference in 1990. They continued to publish papers together, touching on topics such as generalized means and the reconciliation of Bayesian and frequentist testing, until Casella's death in 2012.