Tim Harford’s (2021) book, Data Detective: Ten Easy Rules to Make Sense of Statistics, isn’t really about statistical methods. It has a much broader scope, and it is less technical than book on methods. It deals with knowledge in general, our relationship with knowledge, and the factors that determine that relationship for individuals and collectives. The book is very well-written. You get the sense that Harford has written the text slowly and patiently, based on how well the elements within each chapter are synthesized into a coherent product. The quality of writing makes the book engaging, entertaining and light (you can even read a chapter or two at the end of a tiring work day), and the light and entertaining writing almost makes us forget that the author’s intention is to educate us about the value of statistical reasoning.
Each chapter combines the conceptual-statistical principles (the outer view, or the bird’s-eye view) with personal-anecdotal elements (the inner view, or the worm’s-eye view). The ten principles in the book, each examined in a chapter, are brought to life with interesting examples, which is why going through the ten principles in the form a short list (Table of Content) would not provide a fair sample of the full text. Nevertheless, here is the list of chapter titles:
- Search Your Feelings
- Ponder Your Personal Experience
- Avoid Premature Enumeration
- Step Back and Enjoy the View
- Get the Backstory
- Ask Who Is Missing
- Demand Transparency When the Computer Says No
- Don’t Take Statistical Bedrock for Granted
- Remember That Misinformation Can Be Beautiful, too
- Keep an Open Mind
The ten chapters are followed by a final chapter, titled “The Golden Rule: Be Curious.”
Harford’s general stance regarding the value of statistics is described with reference to the position of Darrell Huff, the author of the 1954 book, How to Lie with Statistics. The Huff-Harford polarity is a theme we continue thinking about until the end of the book, because the theme goes through different variations throughout the chapters. In Chapter 1, that central theme takes the form of a polarity between (unexamined) feelings and careful reasoning; in Chapters 2-3, we turn to the tension between naïve realism and scientific thinking; in Chapters 4-6, we distinguish two types of evidence: (i) what is readily available and (ii) what is initially hidden from view; in Chapters 7-8, we read about the opposition between the aim of knowing and the forces that want to instrumentalize and overwhelm that aim (i.e., blind technological forces or political forces); finally, Chapters 9-10, discuss the possible disharmony between our commitments to knowledge and our tendency to be swayed by aesthetic-driven and identity-driven dispositions.
Perhaps the most important feature of Harford’s writing is his optimism. His optimism isn’t annoying or out of touch, because it is conditional, careful, and responsible. It is an optimism that knows it must be backed by sustained collective effort. He invites us to share his optimism, as well as his responsibility, to become more attentive to what we believe, and how we believe, to value curiosity and transparency, and to not take for granted what has already been achieved by others. Even if you disagree with his optimism, you might still benefit from the book, since his treatment of the ten principles can be easily disentangled from this message of optimism and hope.
Noting, once again, that the book isn’t a technical treatment of statistics, I would recommend it as a general, accessible treatment of the foundations of statistics and the place of statistical-scientific reasoning in larger social-historical contexts.