Statistics gifts?

Posted on by

The American Statistical Association has an annual recommended gift list. (I think they had Red State, Blue State on the list a couple years ago.) They need some more suggestions in the next couple of days. Does anybody have any ideas?

22 thoughts on “Statistics gifts?”

  1. Amazon's Kindle DX because it's excellent for reading PDFs downloaded from academic journals. I got it a couple of weeks ago and have been in love with it ever since.

    David Salsburg's The Lady Tasting Tea would make a nice gift to a statistician. XKCD: Volume 0 is a nice choice for any math/science/computer person.

    The first season of Sherlock just came out on DVD and Blu-Ray and I think BBC did an amazing job modernizing Sherlock Holmes. There are only 3 episodes and each one is 1.5 hours long. Sherlock places a lot of emphasis on data and making inferences, so that's sort of related to statistics.

    P.S. Holmes would have been a Bayesian, had he been real and a statistician instead of a consulting detective.

  6. Well, there's this company that sells stats-themed T-Shirts ("Don't be mean"); google them up yourself. But your post is actually a bit vague: Are we talking gifts for which statisticians are the recipients?

  8. Mikhail Popov, for Sherlock Holmes Peircian might be a better guess.

    Even a whole book written on just that
    Sebeok, T. (1981) "You Know My Method." In Sebeok, T. "The Play of Musement." Indiana. Bloomington, IA.

    Sherlock Holmes being an especially acute abductor thereby expediting and economizing inquiry.

    Very nice wiki entry http://en.wikipedia.org/wiki/Abductive_reasoning

    Many think Peirce was not a Bayesian but my guess is that they have not read carefully or extensively enough.

    Some points on that in the wiki entry and well as suggested break with Karl Popper.

    Also guess if we bring full model criticism into Bayesian in a fully explicit credible manner – we will end up where Peirce was in 1914.

    Maybe then it could be called Bayesianicism or Bayesician.

    K?

  9. Jason – with references such as Stigler and Fisher-Box in that comic book – it looks very promising.

    It would be nice if anyone who has read it could provide some quick comments.

    K?

  11. K – the comic book is an introductory statistics book with a lot more pictures of people saying stuff. It's not all that different from other intro stats books – but the comic format might help people to get past their initial fear. There's a review on Google books that says: "If the only cartoons you can think of to illustrate statistical concepts are drawings of people being driven insane by the same statistical concepts, then maybe you have written the wrong cartoon guide."

    Back to the original question. Thinkgeek.com has lots of math related goodies. Personally, I want a pi-zza cutter. http://www.thinkgeek.com/homeoffice/kitchen/e616/ .

    Also a good t-shirt for recent PhDs: http://www.thinkgeek.com/tshirts-apparel/unisex/g…

  12. Mike thanks for mentioning my wife's Etsy store, she was thrilled that somebody would suggest Nausicaa Distribution!

    When I was looking for a birthday gift for her I found that ThinkGeek is currently the only company still producing slide rules:

    http://www.thinkgeek.com/gadgets/tools/be12

    She was very happy with this math/retro gift. (And it also made us very apprecative of modern calculators.)

  13. "The first season of Sherlock just came out".

    Not so; it's been out for months … in Britain.

    Why assume the US is everyone's reference?

    Less pedantically: I do agree. It is good.

  15. Upon more reading – the pursuitworthiness of the wiki entry I suggested seemed less clear …

    But lead to this (in one of the links)

    "Peirce takes abductive reasoning to lead to judgments about the relative pursuitworthiness of theories; conclusions that can be thoroughly disconnected from assessments of truth-value."

    which if I add "e.g posterior probabilities" at the end might be worth posting somewhere on this blog.

    K?

  16. Bill that is amazing! I haven't ever seen a circular slide rule before. The design makes sense from a use standpoint, so I wonder why this was not the standard way to make them…

    This will go on my Christmas list. :)

  18. Peirce couldn't have been more blatant (and surprisingly contemporary-sounding) in rejecting the "conceptualists"—the Bayesians of his day. Quotes from Peirce are strikingly identical to those one finds in Neyman and Pearson, by contrast. It isn't just the subjectivism he denounced, he was actually one of the few people in the world to understand that ampliative inference is or ought to be a matter of error-correction and testing. I can give numerous citations…There's quite a treasure trove in his work…

    forgot all my info for signing in so this probably won't work…

  19. Andrew, I remember that you had a circular slide rule, but I can neither confirm nor deny the brand name, nor whether you used it to do your homework.

  21. D. Mayo: I would be willing to assist and repost as a separate more appropriate blog post – Andrew likely would too.

    There are challenges for statisticians to "get Peirce" – reading him is difficult and as you know better than I, Peirce was wrong about everything, as Peirce knew and fully acknowledged.

    Also I believe, especially regarding "error-correction and testing" the Bayesian approach has evolved considerably since 1914 and even since my attempt to clarify Frequency versus Bayes in clinical research: Two Cheers for Bayes. Controlled Clinical Trials 1996.

    And this is primarily due to computing capabilities. Today by the time one dots the i,s and crosses the t,s of a frequency approach (to even the partial standards of say David Cox) one could have carried out dozens of (“purposeful”) Bayesian analyses thereby accelerating "error-correction and testing" or “pursuitworthiness” of the modelling involved.
    (And in many – the so called credible intervals will be very similar to confidence intervals, having good coverage properties and sometimes even better coverage properties, in some sense)

Comments are closed.