William M. Briggs

Statistician to the Stars!

Uncertainty: Book Update

r3

First, note the New & Improved title! Uncertainty, which was suggested by the book’s Foreword author Steven Goldberg, about whom more in a moment. The publisher has the full, complete manuscript (and LaTeX source files) and so the copy editing begins, and later the page proof edits. Average time to publish is about five months from the manuscript submission, I’m told. So we’re looking at June to July. Ish. This would be in time for the summer statistical meetings August in Chicago.

To the colleagues to whom I gave draft copies of the book, I’d say that since then (depending on when you received a copy) the book has changed about 10-20%, all additions and clarifications. The final copy is thus not like what you have. I also removed a large number of typos placed by my enemies—though I’m sure these evil agents have inserted new ones when I wasn’t looking.

To my readers, I once again remind that there is in the book no global warming, no religion, no cultural matters at all, except in a very few instances where science and culture intersect. The book is entirely on the philosophy and practical nature of uncertainty. Even NPR listeners won’t be able to find anything objectionable. No: strike that. NPR listeners can find anything objectionable. But in my book they’d have to work at it.

What’s the book’s main message? In the sciences, we’re too damn sure of ourselves. Why? Because our methods, particularly our probability and statistical methods, don’t do what people think they do. It’s time for a complete, wholesale change in practice. Toss hypothesis test, p-values, parameter estimates onto the flames. (I’m channelling Hume.) Return to the only sane goals: understanding cause and making verifiable predictions.

Strong stuff, no? It takes a full book to prove it. But proofs there are, and plenty of them. Incidentally, I remind readers that I am happy to come and talk on these matters—free! (As long as expenses are paid. But I am a cheap date.) I’ll be in the San Francisco Bay area next Thursday (the 28th). Sign up fast before the parousia!

Steven Goldberg, Professor Emeritus of Sociology, City College, City University of New York, very graciously agreed to write the book’s Foreword. When it gets closer to publication, I’ll post that Foreword, which is a better advertisement than I could ever write.

If you aren’t acquainted with Goldberg’s work, you should be. In particular, there are three works every reader of this site must have (the descriptions are culled from Amazon):

  • When Wish Replaces Thought: In Part One – “Why We Behave as We Do” – Goldberg examines the death penalty; the questions of “normality”; the meaning of behavioral cause; the theory of patriarchy; myths (and truths) about black athletic superiority; and the value of standardized tests. In Part Two – “Why We View the World as We Do” – he examines the truths in stereotypes; the logical structure of Freudian theory; the “correct” use of language; the abortion issue; and science, social science, and bad social science.
  • Fads and Fallacies in the Social Sciences: Steven Goldberg has devoted his career to exposing fallacious reasoning, misrepresented fact, and ideological agendas in the social sciences. His scholarly critiques offer alternative, and sometimes controversial, explanations that are notable for their logical integrity and loyalty to empirical reality. Best known for his work in the physiological roots of sex differences, he has also written on a myriad of other subjects, which, as he bluntly states, “are as fallaciously reasoned in professional journals as in the cocktail party conversations that naively repeat the errors first propounded in those journals.” Among the subjects addressed are the validity of intelligence tests, group differences, the death penalty, sex differences in aggression and cognition, the family, abortion, and the nature of modern society.
  • Why Men Rule: A Theory of Male Dominance: The first edition of this book was lavishly praised by many authorities as the most formidable demonstration of an unpopular truth: males rule in all societies known to history or anthropology, for reasons arising from innate physiology, a brute fact that can never be conjured away by tinkering with social institutions. This new edition has been completely rewritten in the light of two decades of scholarship and debate, taking account of all published criticisms of earlier editions.

Goldberg also wrote a nice book on math, Mathematical Elegance: An Approachable Guide to Understanding Basic Concepts, perfect for those who have difficulty understanding why math is so important. Also, a few years back I did a complete review of Fads and Fallacies in the Social Sciences, which can be found here: I, II, III, IV.

22 Comments

  1. You’re probably safe on NPR. Listeners may not actually work at anything. I think they have a preprinted list of objections and what to express as outrage.

  2. If they haven’t heard it on NPR then it doesn’t exist.

  3. William–

    When will you be accepting advance orders? Sell via Amazon? How about having Nassim Nicholas Taleb (The Black Swan) comment on your book?

  4. Before my children could drive themselves to school, despite all the noise they made, I usually arrived at work catching a daily poem here and listened to an author or a celebrity revealing the story behind his/her book and life here or here on our way home. Now I can listen to those shows with two clicks. Love NPR! It brightens my day.

  5. I would suggest tweaking the new title a bit. People generally don’t handle uncertainty very well so this title is priming them not to like the book. It implies danger lurking around the corner. Smart people avoid danger. Cover-judging being an important reality, a title should be attractive, not off-putting. The negative vibe needs to be balanced with a slant that’s more reassuring. Judith Curry achieves this by introducing some whimsy into her blog entries about “The Uncertainty Monster.” Even something along the lines of “Disarming Uncertainty” or “The Over-Certainty Trap” seem less likely to dissuade potential readers. Maybe others can offer better suggestions?

    http://judithcurry.com/2015/04/23/stalking-the-uncertainty-monster/

  6. Gary—Judith Curry’s use of “the uncertainty monster” might be less disarming, but when I buy a book, I expect the inside to match the cover (uncertainty is not a monster to me). In a way, “Uncertainty” might actually arose curiousity. After all, people don’t like a lot of things, but their curiousity leads them to stay around and investigate anyway. It’s hard to find a title that appearls to a large group but is still accurate.

  7. Sheri,

    Judith Curry’s “ClIMATE SCIENCE AND THE UNCERTAINTY MONSTER”

    Is not about uncertainty in the abstract, it is about how uncertainty is used and abused in science in general and climate science in particular.

    http://journals.ametsoc.org/doi/pdf/10.1175/2011BAMS3139.1

    Monster hiding.
    Uncertainty hiding or the “never admit error” strategy can be motivated by a polit ical agenda or because of fear that uncertain science will be judged as poor science by the outside world. Apart from the ethical issues of monster hiding, the monster may be too big to hide and uncertainty hiding enrages the monster.

    Monster exorcism.
    The uncer ta int y monster exorcist focuses on reducing the uncertainty through advocating for more research. In the 1990s, a growing sense of the infeasibility of reducing uncertainties in global climate modeling emerged in response to the continued emergence of unforeseen complexities and sources of uncertainties. Van der Sluijs (2005, p. 88) states that “monster-
    theory predicts that [reducing uncertainty] will prove to be vain in the long run: for each head of the uncertainty monster that science chops off,
    several new monster heads tend to pop up due to unforeseen complexities,” analogous to the Hydra beast of Greek mythology

  8. Thanks for the Goldberg reference. I hope your book does well.

  9. MattS: I know that. I read the post.

  10. You’ll need a snappy subtitle to distinguish the title from all the other books called “uncertainty”, such as “the wit and wisdom of Briggs the bodacious.”

  11. Sheri, that’s just my gut reaction. Certainly focus groups and surveys with wee-pees will determine what’s best. 😉

  12. Gary—But of course. How else can these things be determined!

  13. Sheri,

    Uncertainty may not be a monster to you, but many scientists working in fields with potential public policy implications treat uncertainty as if it is a monster.

    That is the point of Judith Curry’s article.

  14. Yes, MattS, but I really cannot see anyone working in a field with potential public policy implications buying Briggs book anyway. They live by keeping their heads in the sand. Anything in the title that indicates there might be wiggle room on a position will more than likely drive them away. Thus, I am not certain one is alienating potential buyers with the title. (Marketing is a tough area, especially in products that have limited potential buyers. If the goal was to make money only, writing about Certainty would have been the way to go. Uncertainty will only appeal to smaller group who are still interested in science and scientific method devoid of politics. I am sure Briggs knows this. To be honest, there probably is no title that would get true believers to buy the book. As noted, they want Certainty, so the title would have to be a lie to get them to buy the book.)

  15. Looking forward to your book Briggs!

  16. For the record – Scotian started it!

    Contest:
    How to blame Brigg’s spelling enemas for a new title to his book.
    “Certainty requires we be uncertain.”
    “Uncertaintantly”
    ” what me error”
    “We have met the enemy and they be — oops — already used
    ‘ how to fake an error bar like you mean it”

  17. William,
    It’s been a while since I looked in on your blog, so I’m wondering what is the target audience for your new book and what the prerequisites are? What level of math is required?

  18. Briggs

    January 22, 2016 at 2:30 am

    Jay,

    First, my apologies for whatever painful event in your life which precluded you receiving your daily dose of joy from this site. I trust it has resolved itself? Second, excellent question.

    Any working or would-be scientist, hard and soft, can and should read the book. It’s too much for the general public. Philosophers of science interested in probability and cause should also read. Anybody who ever uses any kind of model, especially statistical, must read it. I command it. Math? Hardly any, and that which is there can be skipped.

    At one point I develop the origin of parameters in one popular model (Binomial). Parameters are those dismal creations that live inside models and which receive attention all out of proportion to their actual importance, but I have to use some math to show this. The proof is only there to show how it’s done. I leave it to others to show the origin of parameters in other models. I also need math to show how finite exchangeability fails, despite its many attempts at proof. Math lurks behind much else, and I have the odd schematic equation, the occasional sketching of this or that, but math is always beside the point in understanding uncertainty.

    No, I focus on understanding. I tried hard to eliminate jargon and wrote as plainly as I could, so you don’t need to be a professional statistician or quantum physicist to follow. (This will be seen as a detriment by some, who will complain that the writing might be appropriate for a blog post, it scarcely befits the learned blah blah blah…)

  19. William,
    Thanks for the detailed reply. I will certainly be buying a copy and I hope it does well. Death to the p-value!

  20. Matt,

    I’m looking forward to reading this. In my field (business database programming and “business intelligence”) there is a huge push by industry leaders to embrace ‘machine learning’ and ‘big data’, but virtually no actual understanding of what the models are actually doing and how accurate they really are (Microsoft is pushing Azure ML and “internet of things” everywhere). Of course the end goal is to allow leadership to abdicate responsibility to a ‘predictive’ model (“I was just doing what the model said I should”). Naturally there are some processes which tend to be simpler and with a good understanding of the causes, we can be reasonably certain at predicting outcomes – but it seems the more abstract we get, the more leadership wants computerized predictions for ridiculous measures.

    Perhaps even if only one business leader will read your book, it will save someone millions of wasted dollars.

  21. Briggs

    January 22, 2016 at 6:30 am

    Nate,

    Reminds me I haven’t updated the Machine Learning Big Data Deep Learning Data Mining Statistics Decision & Risk Analysis Probability Fuzzy Logic FAQ in a while.

  22. Bill S: I must admit “How to fake an error bar like you mean it” has a certain appeal!

Leave a Reply

Your email address will not be published.

*

© 2016 William M. Briggs

Theme by Anders NorenUp ↑