Psychological Statistics by Design: A Free Quick-Start Guide

A free quick-start guide to Psychological Statistics by Design: From Research Question to Defensible APA Results with Python, R, and PyStatsV1 (NEKpress Research: Proof-First Guides, Book 3; paperback ISBN 979-8190396775). It shows what the book does and for whom, traces the argument about baseline adjustment that the book's design-first approach answers (Lord's paradox, its causal analysis by Holland and Rubin, ANCOVA in randomized trials and outside them, and covariates as a researcher degree of freedom), and explains its central idea: decide the claim, the unit, and the estimand before choosing a model, and keep a record that ties every printed number to its rows, code, and independent check.

The worked example follows the book's Chapter 13 study, SWL-S07, a synthetic two-group intervention study with a baseline score: from the design record and data layout, through three correct but different answers (follow-up scores alone, change scores, and the registered analysis of covariance), the model's assumptions, and a rerun in Python and an independent rerun in base R, to an APA-style sentence with its matched limitation. The guide then maps all 20 chapters, gives five exercises with answers, and points to the free resources of the series. Appendix A runs the study on Windows 11, macOS, or Ubuntu with the book's companion, PyStatsV1 0.26.0, without make. The book itself works ten such studies, from a flawed first case to a reviewer packet. A one-page handout is included.

Every number the guide prints from the study was recomputed from the companion's CSV file by new code, in exact rational arithmetic; the study is synthetic. The published works in the background section are cited to their sources. This DOI identifies the guide, not the book; the book itself has no DOI.