English and Japanese titles on red cover


Hyojyun Beizu Toukeigaku (A First Course in Bayesian Statistical Methods)


Peter D. Hoff (author), IRIE Kaoru, SUGASAWA Shonosuke, HASHIMOTO Shintaro (translators)


304 pages, A5 format




June 01, 2021



Published by

Asakura Publishing Co., Ltd.

See Book Availability at Library

Hyojyun Beizu Toukeigaku

Japanese Page

view japanese page

The Japanese translation of the well-received graduate-level textbook on Bayesian statistics. There are several book reviews published in statistical journals; see the author's website. Students of economics might find the review by econometricians useful (Koop, 2010).
My first encounter with Bayesian statistics was in 2008, a year before the publication of the original textbook. Then, the focus of Japanese textbooks was further on the computational aspect of Bayesian statistics. Things that I really wanted to know as a Bayesian beginner, such as the foundation of Bayesianism and real data examples, were often avoided. That hindered my study, forcing me to alternate between multiple textbooks, including some classics (e.g. Zellner, 1986). Ironically to my year-long struggle in getting to the point of the basics of Bayesian statistics, my graduate friends at Duke read Hoff's textbook and got ready for the research-level topics in just a semester, which made me extremely jealous. Then, when Shintaro Hashimoto (Hiroshima U) and Shonosuke Sugasawa (UTokyo) talked to me about their translation project, I immediately decided to join them.
Although we wrote that there had not been a Japanese textbook on Bayesian statistics like Hoff's, that does not mean all Japanese textbooks on Bayesian statistics are bad. Shigematsu (1985) and Nakatsuma (2007) are both well-known for their detailed explanations of the important concepts, models, and examples in Bayesian statistics. In comparison with those books, Hoff (2009) and our translated version stand out in covering a range of topics, such as hierarchical and generalized linear models with real statistical problems, that are not limited to economic and financial data. Providing R-codes as this textbook did has become a norm in the present statistical education, which is also true in Japan (e.g., Kakamu, 2022).
When we first presented our translation project, the publisher raised an honest concern whether the textbook was already outdated and if there was any alternative available in Japanese. To our relief (and surprise), the first 1200 copies were sold out in just a week. This good news shaped our posterior belief that the textbook would continue to be the very first course of Bayesian statistics for the next few (or tens of) years in graduate-level education in Japan, as well as would help the study and research of the UTokyo members.

(Written by IRIE Kaoru, Assistant Professor, Graduate School of Economics / 2022)

Related Info

Original book:
Peter D. Hoff, A First Course in Bayesian Statistical Methods, Springer, 2009.
Gary Koop  (Econometrics Journal, volume 13, issue 1, p. B1-B5, 11 February, 2010)
James Cheshire  (The Journal of the Royal Statistical Society, Series A (Statistics in Society), Volume 173, Issue 3, p. 694-695, 17 June, 2010)
Lasse Koskinen  (International Statistical Review, volume 78, issue 1, p. 150, 1 April, 2010)

Try these read-alike books: