Uncertainty: The Soul of Modeling, Probability & Statistics

Description

cites [Ed Feser](https://isidore.co/calibre/#panel=book_list&search=authors:%22%3DFeser%2C%20Edward%22), [St. Thomas Aquinas](https://isidore.co/aquinas/), [William A. Wallace, O.P.](https://isidore.co/calibre/#panel=book_list&search=authors:%22%3DWallace%2C%20William%20A.%2C%20O.P.%22), and [Stanley Jaki, O.S.B.](https://isidore.co/calibre/#panel=book_list&search=authors:%22%3DJaki%2C%20Stanley%20L.%2C%20O.S.B.%22)!See [his lecture](https://youtu.be/-t5lHXAvuLQ) for the [Broken Science Initiative](https://brokenscience.org/), wherein he "explains why academic science is critically failing to do its intended purpose: produce results that replicate. (October 07, 2022)"Briggs: "AI" = "[Another name for probability models](https://www.wmbriggs.com/post/6465/)" *vel* "[Curve fitting](https://www.wmbriggs.com/post/27654/)".* * *This book presents a philosophical approach to probability and probabilistic thinking, considering the underpinnings of probabilistic reasoning and modeling, which effectively underlie everything in data science. The ultimate goal is to call into question many standard tenets and lay the philosophical and probabilistic groundwork and infrastructure for statistical modeling. It is the first book devoted to the philosophy of data aimed at working scientists and calls for a new consideration in the practice of probability and statistics to eliminate what has been referred to as the "Cult of Statistical Significance." The book explains the philosophy of these ideas and not the mathematics, though there are a handful of mathematical examples. The topics are logically laid out, starting with basic philosophy as related to probability, statistics, and science, and stepping through the key probabilistic ideas and concepts, and ending with statistical models.Its jargon-free approach asserts that standard methods, such as out-of-the-box regression, cannot help in discovering cause. This new way of looking at uncertainty ties together disparate fields — probability, physics, biology, the “soft” sciences, computer science — because each aims at discovering cause (of effects). It broadens the understanding beyond frequentist and Bayesian methods to propose a Third Way of modeling.“Briggs, an adjunct professor of statistics at Cornell University, cautions his readers to carefully examine the uncertain reliability of such conclusions when these tools are used. His challenging premises are thoroughly supported by philosophical explanations as to why these traditional approaches need to be questioned. … Briggs provides fully fleshed out reasoning, impressive support, precisely worded insight, and graphical illustrations, as appropriate, to justify his stand. … Summing Up: Recommended. Upper-division undergraduates and above; faculty and professionals.” (N. W. Schillow, Choice, Vol. 54 (6), February, 2017)“This is a book about probability and probabilistic reasoning. It is more philosophy than mathematics, but it does have mathematical content and it relies in some measure on mathematical reasoning. … This book is worth a look by anyone who teaches probability and statistics.” (William J. Satzer, MAA Reviews, August, 2016)“[This book] is not for sissies, true, but its clear-headed (i.e., Aristotelian) approach to the subject of truth (which, in the end, is what exercises in probability and statistical analysis are all about, notwithstanding what they tell you in school) is refreshing: a long, cool drink of plain speaking about intellectual topics that, in these hot and humid days, is as enlivening as it is enlightening.” (Roger Kimball, The New Criterion's Critic's Notebook, newcriterion.com, August, 2016)“This book has the potential to turn the world of evidence-based medicine upside down. It boldly asserts that with regard to everything having to do with evidence, we’re doing it all wrong: probability, statistics, causality, modeling, deciding, communicating—everything. … the book is full of humor and a delight to read and re-read.” (Jane M. Orient, Journal of American Physicians and Surgeons, Vol. 21 (3), 2016)**William M. Briggs, PhD,** is Adjunct Professor of Statistics at Cornell University. Having earned both his PhD in Statistics and MSc in Atmospheric Physics from Cornell University, he served as the editor of the American Meteorological Society journal and has published over 60 papers. He studies the philosophy of science, the use and misuses of uncertainty - from truth to modeling. Early in life, he began his career as a cryptologist for the Air Force, then slipped into weather and climate forecasting, and later matured into an epistemologist. Currently, he has a popular, long-running blog on the subjects written about here, with about 70,000 - 90,000 monthly readers.* * *The 18th century English poet William Cowper wrote `And diff'ring judgements serve but to declare/That truth lies somewhere, if we knew but where'. The ancients suggested that truth lived at the bottom of a well: Briggs's view is that she lives in the mind, and is aimed at by probability. Uncertainty, and thus probability, is entirely epistemological, and is properly a branch of philosophy with mathematical aspects rather than one simply of mathematics [p. 244]. The book of common prayer declares that `There is none that doeth good, no not one', and this, in Briggs's view, is applicable to (almost) all statistical practitioners. In fact, whenever there is evidence involved, we probabilists, statisticians, decision makers, etc., are doing something wrong. The important thing to note is that all probability is conditional on evidence, and the reader will see here similarities to older work by Keynes, Cox and (more recently) Jaynes. In Briggs's view probability is not necessarily a number, though it might be. It is not physical, subjective, nor a limiting relative frequency. And this view of course has considerable impact on the statistical methods used and the conclusions reached in any scientific investigation. Not only are common statistical phrases like `due to chance', `random effects' and `these results are significant and not due to chance' [p. 87] verboten, but our customary ideas about regression, time series and hypothesis testing come in for severe criticism. Briggs's harshest criticism is perhaps reserved for the $p$-value, the pertinent section being given the heading `Die, $p$-value, die die die'. Briggs stresses the importance of understanding cause, and here models hold a fundamental position: `Models should be used to make probabilistic predictions of observable entities. These predictions can, in turn, be used to make decisions' [p. xii]. Uncertainty is not addressed just to the conveyors and purveyors of statistical methods but also to the users. The reader will find himself being ineluctably led from an opinion about which there can be little, if any, doubt to one that seems contrary to all accepted, and acceptable, statistical practice. If Briggs's arguments are accepted by those whose business is the teaching of statistics, then we are in for a major upheaval. Briggs's earnest prayer is that more and more use will be made of predictive methods, with the corresponding eventual replacement of `parametric-centric, hypothesis testing decision-is-probability classical procedures' [p. 243]. The latter, though perhaps much easier to handle because of the software available for their use and because of the convenience of the `significance' that may be achieved, tend to provide over-certainty (the logical approach presented here unfortunately leads to a decrease in certainty). Honesty in science perhaps demands more, and to quote Francis Quarles, another early English poet, `the road to resolution lies by doubt'. There is no substitute for thought.[Reviewed by A. I. Dale](https://mathscinet.ams.org/mathscinet-getitem?mr=3497108)