19  Afterword

This text was written for Exercises in Psychological Statistics, a course I have been teaching since 2024 at the University of Tokyo’s College of Arts and Sciences. Its contents range from how to use R to Bayesian modelling, and I have rather freely packed in whatever I felt like covering. The selection reflects my own predilections; there is an outline of sorts, but the chapters were written without much concern for length. This is not a textbook in the “one chapter per class meeting” sense, and may feel inconvenient as such. In the 2025 edition of the course I in fact presented the chapter titles and let the students choose which to cover, treating the rest as material to read on their own.

Although the course is hands-on with R, in the age of generative AI it is not necessarily true that students must write code by themselves. I chose to publish this text in Quarto and as a Web page partly so that code can be copied easily. I am not opposed to using generative AI — this book itself has had AI assistance (proofreading, drafting answer checks for exercises). What matters is to use generative AI rather than to be used by it. The point is not to earn course credit but to grow; the joy of learning and discovery should not be ceded to machines.

I chose an online format also because R and Stan are free software that keep evolving. Versions change, links break; an online document can be updated. I still felt the pull of the printed page, so PDF and EPUB versions also exist.

In the 2024 edition not all chapters were available; in particular, the Bayesian chapters were written alongside the 2025 course. I had planned to finish everything in 2024, but it took me until the end of the summer holidays in 2025. The delay has had its compensations: multivariate analysis and Bayesian modelling could be included with room to breathe.

Much of the material overlaps with what I have written elsewhere — the same author, after all. I am not sure how cleanly the niches should be carved; this volume has perhaps a slightly more exercise- and practice-oriented flavour. I do, however, think it useful to have a single resource that treats psychological-statistics exercises — including Bayesian methods — in a unified way. If it proves useful to anyone, I shall be glad.

Since this is my own course material, it has not been formally peer-reviewed. Any errors are mine alone. If you notice anything, please let me know; I will fix it immediately.

20 September 2025 Koji Kosugi