This book is work in progress. I restarted essentially from scratch in 2025. It’s a strongly frequentist introduction to statistics without any friggin’ software package. In this book you get to use paper, a pen, and a pocket calculator. Why does “hands-on introduction to statistics” mean that we’re flooded with R commands starting on page 1? Why is it evil to use frequentist statistics by ramming commands into R without a deeper understanding, but it’s hip and totally “good” to do the same with Bayesian statistics? This book is for people without a deadline who would like to do some serious thinking (with me) about what inferential frequentist statistics is actually trying to do for us. Without the usual misinterpretations. No real-life data sets are used in this book, only intentionally simplified fictional examples from toy linguistics. Who cares about the examples anyway?
The GitHub links to this highly unfinished work are here…
Repository: https://github.com/rsling/Inferencelinguistics
PDF direct: https://github.com/rsling/Inferencelinguistics/blob/main/main.pdf
Since I’m not a statistician and above all not perfect, please don’t hesitate to look at whatever is there. If you find errors, tear the PDF apart, throw it into my face, and tell me about those errors as harshly as you can. Yell at me if necessary. I’m serious! As opposed to many of our colleagues, I find nothing more satisfying than being corrected and understanding why I was wrong. I’m not being sarcastic. Email me or submit a ticket on GitHub.