CountCamp

The Regressive Backbone — statistics, made clear. By Benjamin Telkamp.

The book

The Regressive Backbone is a statistics book with the working notes left in: at the bottom of every page you can see where the author is still weighing things up, and where he still has his doubts. The first chapters are in English now; the rest, so far, only in the Dutch edition. To practise as you read, three exercise books in Dutch go with it chapter by chapter, in R, JASP and SPSS; an English R one is planned.

Exercise books

Exercise books in SPSS, R and JASP — most useful if you study at Leiden University, Hogeschool Leiden or Erasmus University, and for one-to-one help. Regression, ANOVA, psychometrics — step by step, with animals that behave badly. In Dutch for now; the first English ones are on their way.

Read a p-value

A test statistic from a paper — t(38) = 2.24, z = 1.96 — and you see the p-value straight away, as an area under the curve, with the z, t, F and χ² tables beside it. A tool, not a toy — though it is in the toy box too. (in Dutch)

The toy box

Small things that run in your browser: shuffle a correlation and see whether it survives chance, read a p-value off as an area, take a change apart into four pieces, or guess against an elephant who remembers everything. Nothing to install.

The judgement of Solomon

Four groups, one therapy, and a pretest that does something of its own. Step by step you take the first group’s improvement apart into four pieces, until only one is left: what the pretest and the therapy do together. Then, as an encore, the same four pieces come out of a single regression. (in Dutch)

Services

Everything here is free. Sometimes you want more than a book: someone who sits down beside you until it clicks. That is the part you pay for — one-to-one help, a second pair of eyes on your research or your thesis, exam training, guest lectures.

About

Who’s behind this, and why. And how to reach him.

The doubt is left in

Statistics is the one subject that takes uncertainty seriously; its textbooks rarely sound like it. Most of them speak with the certainty of someone who has never had to think about anything. The Regressive Backbone doesn’t: it practises what it preaches about uncertainty. Uncertainty is not a problem. It is contact with reality.

That is why every page of the book ends with a button that opens the working notes. There are two kinds. The first is the workbench — wordings still to be chosen between, a figure still to come. The second is the doubt, and that one is about you: does the reader see here that this is about the population, or does it need to be sharper? Under a question like that, the extra explanation is right there. If you’re following, you simply read on; if you stall, you open the note and find exactly that piece. It is not a peek behind the scenes. It is a second way of reading.

And it does something else. Where the material is hard, the page simply says so, in the words of someone who is still weighing up how to explain it. If you get stuck there for a moment, that is not down to you — that is where everyone’s thinking stalls, the writer’s included. You stand beside the author instead of beneath him. That’s exactly where this book wants you to be.