This lively collection of essays examines in witty detail the history of some of the concepts involved in bringing statistical argument "to the table," and some of the pitfalls that have been encountered. The topics range from seventeenth-century medicine and the circulation of blood, to the cause of the Great Depression and the effect of the California gold discoveries of 1848 upon price levels, to the determinations of the shape of the Earth and the speed of light, to the meter of Virgil's poetry and the prediction of the Second Coming of Christ. The title essay tells how the statistician Karl Pearson came to issue the challenge to put "statistics on the table" to the economists Marshall, Keynes, and Pigou in 1911. The 1911 dispute involved the effect of parental alcoholism upon children, but the challenge is general and timeless: important arguments require evidence, and quantitative evidence requires statistical evaluation. Some essays examine deep and subtle statistical ideas such as the aggregation and regression paradoxes; others tell of the origin of the Average Man and the evaluation of fingerprints as a forerunner of the use of DNA in forensic science. Several of the essays are entirely nontechnical; all examine statistical ideas with an ironic eye for their essence and what their history can tell us about current disputes.
It would be hard to overstate the importance of statistical knowledge in the present; almost all of science relies on statistical models, and technologies like machine learning and artificial intelligence are profoundly statistical in nature. There are few better guides through its incredibly strange history than Stigler, whose writing captures both elusive interpretive issues and technical mathematics—without ever oversimplifying. While readers might be most familiar with Stigler's encyclopedic History of Statistics (which remains a standard reference in the field), Statistics on the Tables shines as a lively, topic-driven exploration of statistical ideas, their creators—John Maynard Keynes makes a memorable appearance—and the controversies that sparked them. It also shows how statistics moved across a vast set of disciplines, from economics to theology. Do you want to really understand phenomena like regression to the mean or techniques like probabilistic sampling? Then read this.
lord acton is again right. if you wanna learn a science start with its history. in case of statistics, you will follow this little gem which is clearly carved from modern statistical conception by its own story.
This is an excellent book on varied topics in the history of statistics -- assuming you already know something about statistics. It is a compilation of historical essays Stigler wrote for various academic journals, so he assumes the reader has some knowledge of distributions, expectations, random variables, and so on. (But there is nothing someone with one or two courses in statistics shouldn't be able to work out.)
For essays published in academic journals, these are surprisingly readable. Stigler is an excellent historian and any statistician interested in the history of the field would be served well by this book.
The writing is good but it turns out this is a collection of independent essays, each about a very specific point in the history of statistics. For example the essay on Quetelet, a 19th-century statistician I've never heard of, assumes that you already know why he's important. Other essays seem to demand quite a bit of technical knowledge. I'll put this aside for now until I've read something more general - perhaps Stigler's own The History of Statistics: The Measurement of Uncertainty before 1900.
The last two books I"ve read have been about the lives of statisticians. This one is more detailed, goes into the math, goes back farther into history, and gives lots of references. Salsburg's book (the lady tasting tea) is more of a popular science book. It's a little easier to read, perhaps to whet your apppetite for this one.