Bayesian Statistics Blog

Bayesian Statistics Blog http://bayesianstats.com Most recent posts at Bayesian Statistics Blog posterous.com Sun, 12 Jun 2011 19:58:04 -0700 My theorem becomes ever more famous http://feedproxy.google.com/~r/BayesianStatisticsBlog/~3/ZNL_U3bU6SY/my-theorem-becomes-ever-more-famous http://bayesianstats.com/my-theorem-becomes-ever-more-famous <p> After only about 250 years, it seems the mainstream and the laypeople <br />are at last hearing about the eponymous Bayes' Theorem. (<a href="http://bayesianstats.com/bayes-theorem-is-independently-exchangeably-p">What a shame that some professional mathematicians have not managed to do the same</a>). <a href="http://www.boston.com/ae/books/articles/2011/06/05/after_centuries_of_dispute_a_theory_rooted_in_common_sense_wins_out/">Here's a review from the Boston Globe</a> (the Boston in the New World, that <br />is) about the "first-ever account of Bayes’ rule for general readers", <br />according to its <a href="http://yalebooks.wordpress.com/2011/05/19/the-theory-that-would-not-die-yale-publish-new-book-on-bayes-theorem/">publishers</a>: <br />'The Theory That Would Not Die: How Bayes' Rule Cracked the Enigma <br />Code, Hunted Down Russian Submarines, and Emerged Triumphant from Two <br />Centuries of Controversy' by <a href="http://www.mcgrayne.com/">Sharon Bertsch McGrayne</a>. Read quite a lot of it for free at <a href="http://books.google.com/books?id=_Kx5xVGuLRIC&lpg=PP1&pg=PP1#v=onepage&q&f=false">Google Books</a>, or --- like me --- <a href="http://www.amazon.co.uk/gp/product/0300169698/ref=as_li_ss_tl?ie=UTF8&tag=bayestatblog-21&linkCode=as2&camp=1634&creative=19450&creativeASIN=0300169698">buy a copy</a><img src="http://www.assoc-amazon.co.uk/e/ir?t=&l=as2&o=2&a=0300169698" border="0" height="1" alt="" style="border: none !important; margin: 0px !important;" width="1" /> pour encourager les autres. </p> <p><a href="http://bayesianstats.com/my-theorem-becomes-ever-more-famous">Permalink</a> | <a href="http://bayesianstats.com/my-theorem-becomes-ever-more-famous#comment">Leave a comment  »</a> </p><img src="http://feeds.feedburner.com/~r/BayesianStatisticsBlog/~4/ZNL_U3bU6SY" height="1" width="1"/> http://files.posterous.com/user_profile_pics/757812/bayes.jpg http://posterous.com/users/4wEYiML1RY6R Thomas Bayes mrbayes Thomas Bayes http://bayesianstats.com/my-theorem-becomes-ever-more-famous Mon, 06 Jun 2011 22:36:03 -0700 Bayes Theorem is independently (exchangeably?) proven again! http://feedproxy.google.com/~r/BayesianStatisticsBlog/~3/okYf8VpV1zc/bayes-theorem-is-independently-exchangeably-p http://bayesianstats.com/bayes-theorem-is-independently-exchangeably-p <p> Back in 1763 the World Wide Web wasn't yet fully developed [we only <br />had gophers for communication back then], so I can accept that not <br />everybody has heard of my Theorem, nowadays kindly called Bayes' <br />Theorem by those in the know. <p /> But even now, it seems, some poor souls have only got up to 17th <br />century mathematics, such as calculus and the Bernoulli distribution, <br />but ultra-modern 18th century stuff like my Theorem. <a href="http://arxiv.org/abs/1105.1486">And so these antediluvians have had to prove it all over again</a>. <p /> Seriously. And then they go on to call it the "exact method" (as <br />opposed to the frequentist "method" which assumes too much and answers <br />too little, but is clearly much more famous than my apparently <br />old-fashioned Theorem), wondering <p /> <blockquote class="posterous_medium_quote">It is not clear why the exact method isn’t mentioned in <br />most textbooks or, indeed, why it isn’t universally used instead of <br />the standard method. Apparently the exact method is not well <br />known.</blockquote> <p /> Indeed not. <p /> [A tip of the hat to <a href="http://xianblog.wordpress.com/2011/06/05/bayes-redux/">Xi'an</a> <br />(a.k.a. Christian P. Robert) for the generous promulgation of this <br />delightful finding] </p> <p><a href="http://bayesianstats.com/bayes-theorem-is-independently-exchangeably-p">Permalink</a> | <a href="http://bayesianstats.com/bayes-theorem-is-independently-exchangeably-p#comment">Leave a comment  »</a> </p><img src="http://feeds.feedburner.com/~r/BayesianStatisticsBlog/~4/okYf8VpV1zc" height="1" width="1"/> http://files.posterous.com/user_profile_pics/757812/bayes.jpg http://posterous.com/users/4wEYiML1RY6R Thomas Bayes mrbayes Thomas Bayes http://bayesianstats.com/bayes-theorem-is-independently-exchangeably-p Tue, 15 Feb 2011 22:53:37 -0800 ... http://feedproxy.google.com/~r/BayesianStatisticsBlog/~3/e4AZa_pQt_o/43207158 http://bayesianstats.com/43207158 <p> I just used the "..." argument in an R function successfully. I am <br />therefore great. This is especially impressive for a dead 18th-century <br />proto-statistician, no? </p> <p><a href="http://bayesianstats.com/43207158">Permalink</a> | <a href="http://bayesianstats.com/43207158#comment">Leave a comment  »</a> </p><img src="http://feeds.feedburner.com/~r/BayesianStatisticsBlog/~4/e4AZa_pQt_o" height="1" width="1"/> http://files.posterous.com/user_profile_pics/757812/bayes.jpg http://posterous.com/users/4wEYiML1RY6R Thomas Bayes mrbayes Thomas Bayes http://bayesianstats.com/43207158 Wed, 29 Dec 2010 17:13:32 -0800 Most unusual Bayesian application of the day [or decade]: The Bayesian-Moroni Prayer Analysis Calculator http://feedproxy.google.com/~r/BayesianStatisticsBlog/~3/ASDvHo76xR8/most-unusual-bayesian-application-of-the-day http://bayesianstats.com/most-unusual-bayesian-application-of-the-day <p> <a href="http://www.lds4u.com/lesson1/bayesian.htm">http://www.lds4u.com/lesson1/bayesian.htm</a> <p /> From the site: <p /> The last chapter of the Book of Mormon contains a promise to everyone who reads it. The promise states that if you read the book and ask God about it with a sincere heart, "real intent" and faith in Christ, then he will tell you the book is true by the gift of the Holy Ghost (See Moroni 10:3-5). The problem with this promise is that if the book isn't true, then God really didn't make the promise in the first place. The reader finds himself in a loop of circular reasoning where the way to find out if the book is true is based upon the premise that it is in fact true. <p /> Is this a sincere effort or not? Only a Bayesian analysis can help with that one... <p /> I'm not actually sure it is a valid calculator, actually, but it's too ridiculous a premise to look into too much anyway. Although, on reconsideration, Pascal's Wager is a similar-ish sort of argument what with trying to use probability theory to help with struggles of faith in God. The article about that at <a href="http://plato.stanford.edu/entries/pascal-wager/">http://plato.stanford.edu/entries/pascal-wager/</a> is well worth working through, by candlelight if necessary. <p /> -- </p> <p><a href="http://bayesianstats.com/most-unusual-bayesian-application-of-the-day">Permalink</a> | <a href="http://bayesianstats.com/most-unusual-bayesian-application-of-the-day#comment">Leave a comment  »</a> </p><img src="http://feeds.feedburner.com/~r/BayesianStatisticsBlog/~4/ASDvHo76xR8" height="1" width="1"/> http://files.posterous.com/user_profile_pics/757812/bayes.jpg http://posterous.com/users/4wEYiML1RY6R Thomas Bayes mrbayes Thomas Bayes http://bayesianstats.com/most-unusual-bayesian-application-of-the-day Mon, 20 Dec 2010 18:13:00 -0800 A supposed Bayesian analysis of whether the Swedish police actions against Julian Assange are politically motivated http://feedproxy.google.com/~r/BayesianStatisticsBlog/~3/O4fbIDH2-dA/a-supposed-bayesian-analysis-of-whether-the-s http://bayesianstats.com/a-supposed-bayesian-analysis-of-whether-the-s <p> <p>Bayesian methods are perfect for forensic and other police <br />investigations. A couple of books that show how in very different ways <br />are <a href="http://www.amazon.co.uk/gp/product/0471141828?ie=UTF8&tag=bayestatblog-21&linkCode=as2&camp=1634&creative=19450&creativeASIN=0471141828">A Probabilistic Analysis of the Sacco and Vanzetti Evidence (Wiley Series in Probability and Statistics)</a><img src="http://www.assoc-amazon.co.uk/e/ir?t=bayestatblog-21&l=as2&o=2&a=0471141828" border="0" height="1" alt="" style="border: none !important; margin: 0px !important;" width="1" /> and <a href="http://www.amazon.co.uk/gp/product/0470091738?ie=UTF8&tag=bayestatblog-21&linkCode=as2&camp=1634&creative=19450&creativeASIN=0470091738">Bayesian Networks and Probabilistic Inference in Forensic Science</a><img src="http://www.assoc-amazon.co.uk/e/ir?t=bayestatblog-21&l=as2&o=2&a=0470091738" border="0" height="1" alt="" style="border: none !important; margin: 0px !important;" width="1" />. I'm sure there are many, many others. <p /> The New York Times in-house stats-blogger Nate Silver has <a href="http://fivethirtyeight.blogs.nytimes.com/2010/12/15/a-bayesian-take-on-julian-assange">tried to write a model analysis using Bayesian thinking</a> of whether the ongoing Swedish sex crimes case against Julian Assange, the personage most widely associated with the Wikileaks website (and whose OKCupid profile you can pruriently find at <a href="http://fivethirtyeight.blogs.nytimes.com/2010/12/15/a-bayesian-take-on-julian-assange">http://www.okcupid.com/profile/HarryHarrison</a>), is politically motivated. After starting with a frankly facile argument about whether his fellow passenger on the bullet train in Japan is Japanese, Caucasian, or a combination of the two, Silver goes on to provide not <br />light, but conspiracy theory heat, about that important question. It <br />boils down to saying that because he's a high-profile figure who has <br />annoyed many powerful people, including governments around the world, <br />that therefore it increases the probability that the current <br />shenanigans are politically motivated. Therefore, Mr. Silver <br />concludes, we should be more sceptical that the investigations are <br />done purely for crime-prevention purposes. The crucial two paragraphs: <p /> What is less ambiguous here, however — as in the case of my bullet <br />train analogy — is the underlying context. The handling of the charges <br />suggests that the motivation for bringing them against Mr. Assange is <br />political. If the motivation is political, then the merits of the <br />charges might matter less. Even if they fail to result in a <br />conviction, the authorities might nevertheless succeed in, in essence, <br />incapacitating Mr. Assange for several months, and preventing him from <br />releasing further documents through WikiLeaks. They might also injure <br />Mr. Assange’s reputation among the public: certainly I have learned <br />more about details Mr. Assange’s personal life in recent days than I <br />would care to know. <p /> Under these circumstances, then, it becomes more likely that the <br />charges are indeed weak (or false) ones made to seem as though they <br />are strong. Conversely, if there were no political motivation, then <br />the merits of the charges would be more closely related to <br />authorities’ zealousness in pursing them, and we could take them more <br />at face value. <p /> Two quick points: Bayesianism doesn't have one right answer; it is <br />subjectivist. This analysis can be argued over by reasonable people. <br />I, for one, consider it simplistic. (For example, Silver points out <br />that the only other Interpol Red Alert against a sex offender issued <br />by Sweden this year was for a man accused of multiple sexual assaults <br />against children; a different kettle of pervert-fish indeed. But <br />what's the denominator? In other words, how many suspects of Swedish <br />sex crimes were abroad this year at the moment when the Swedes wanted <br />to talk to them? This is a crucial question without which a useful <br />likelihood function cannot really be constructed, hence no useful <br />Bayesian analysis). <p /> But a more important point to make is: so what? Simply because of <br />Assange's notoriety, according to this analysis, we must conclude that <br />political motivations are more likely factors for what is going on. <br />But only a fool would leave it at that and decide the whole case is <br />suspect, before making a placard and protesting outside the English <br />courthouses whether Assange will be extradited. Most people would <br />continue to do what they did before --- i.e. let the legal process <br />take its course --- and hopefully realise, if they hadn't up until <br />this point, how incredibly unjust the whole system of European Arrest <br />Warrants are. <p /> The promotion of Bayesian methods amongst the general public is a joy <br />to see, but it has to be done rigorously, otherwise laypeople will <br />justifiably see it as just another way to reinforce one's own <br />prejudices as opposed to exposing them.</p> </p> <p><a href="http://bayesianstats.com/a-supposed-bayesian-analysis-of-whether-the-s">Permalink</a> | <a href="http://bayesianstats.com/a-supposed-bayesian-analysis-of-whether-the-s#comment">Leave a comment  »</a> </p><img src="http://feeds.feedburner.com/~r/BayesianStatisticsBlog/~4/O4fbIDH2-dA" height="1" width="1"/> http://files.posterous.com/user_profile_pics/757812/bayes.jpg http://posterous.com/users/4wEYiML1RY6R Thomas Bayes mrbayes Thomas Bayes http://bayesianstats.com/a-supposed-bayesian-analysis-of-whether-the-s Thu, 09 Dec 2010 14:52:00 -0800 Statistics is rapidly going mainstream http://feedproxy.google.com/~r/BayesianStatisticsBlog/~3/ZKuqEiJJPHY/statistics-is-rapidly-going-mainstream http://bayesianstats.com/statistics-is-rapidly-going-mainstream <p> <p>UPDATE: <a href="http://www.youtube.com/watch?v=oOOmqHzkkOo">The programme is now available in full on Youtube</a>. I found out from <a href="http://blog.revolutionanalytics.com/2010/12/the-complete-joy-of-stats.html">the always-enjoyable Revolution blog</a>.</p> <p>Statistics is going mainstream, and I know this from just one data </p> <p>point: Hans Rosling's new documentary for the BBC, "The Joy of <br />Stats". It's an hour-long propaganda piece on why statistics is not <br />the dry, boring subject you were taught at school, but an exciting and <br />actually indispensable tool for understanding the world. <p /> Statistics is science. Statistics is philosophy. Statistics is <br />knowledge and power. With cheap computation and storage, the golden <br />age of statistics is upon us. Watch the start of this new age at <br /><a href="http://www.bbc.co.uk/iplayer/episode/b00wgq0l/The_Joy_of_Stats/">http://www.bbc.co.uk/iplayer/episode/b00wgq0l/The_Joy_of_Stats/</a> <p /> If you're not in the UK, or the expiry time of 4:29AM GMT Wed, 15 Dec <br />2010 has passed, you can instead see some clips at <br /><a href="http://www.open.ac.uk/openlearn/whats-on/the-joy-stats">http://www.open.ac.uk/openlearn/whats-on/the-joy-stats</a> <p /> And the film might end up on Rosling's Gapminder website at some <br />point, according to one ccomment I saw online, a drop in the <br />universe-ocean of data. He's so passionate I can believe he'll do it. <p /> If I have one criticism --- and I have to have at least one to make it <br />worth your while to read the above sycophancy --- is that there wasn't <br />enough emphasis on how ordinary people can use statistics in their <br />lives. There was a section on citizens of San Francisco bringing <br />online maps of crime report statistics to meetings with <br />policepeople. All very well; but how about using statistics as we live <br />our lives and need to make decisions? A lively discussion of <br />rationality versus intuition could have been really fascinating and a <br />little more engaging than yet another story about big telescopes. <p /> But I'm being picky. Please watch "The Joy of Stats" now if you <br />haven't already, and hug your nearest statistician (or "statso", if <br />you prefer) at your leisure. We'll really appreciate it. <p /> That link again: <a href="http://www.bbc.co.uk/iplayer/episode/b00wgq0l/The_Joy_of_Stats/">http://www.bbc.co.uk/iplayer/episode/b00wgq0l/The_Joy_of_Stats/</a></p> </p> <p><a href="http://bayesianstats.com/statistics-is-rapidly-going-mainstream">Permalink</a> | <a href="http://bayesianstats.com/statistics-is-rapidly-going-mainstream#comment">Leave a comment  »</a> </p><img src="http://feeds.feedburner.com/~r/BayesianStatisticsBlog/~4/ZKuqEiJJPHY" height="1" width="1"/> http://files.posterous.com/user_profile_pics/757812/bayes.jpg http://posterous.com/users/4wEYiML1RY6R Thomas Bayes mrbayes Thomas Bayes http://bayesianstats.com/statistics-is-rapidly-going-mainstream Wed, 08 Dec 2010 16:33:45 -0800 The apogee of art, philosophy and life is here: http://feedproxy.google.com/~r/BayesianStatisticsBlog/~3/n1U_7VkMmOk/the-apogee-of-art-philosophy-and-life-is-here http://bayesianstats.com/the-apogee-of-art-philosophy-and-life-is-here <p> <div class='p_embed p_image_embed'> <a href="http://posterous.com/getfile/files.posterous.com/bayesianstats/X9m5AD33N9tLSm9900qUod44VXYcl3EUnVTpVj0QN54GtDHUpLuowGyRGjjh/74317_10150301349610316_862480.jpg"><img alt="74317_10150301349610316_862480" height="267" src="http://posterous.com/getfile/files.posterous.com/bayesianstats/Q65RxcE9cNDFIUOynGDMSoDZeKY6XeDBZYQwhzdjqYzHkQa9ZWNfBjhtvXAp/74317_10150301349610316_862480.jpg.scaled.500.jpg" width="500" /></a> </div> <p>I would love to find out who came up with this. I have my hunches (geddit?!?).</p> </p> <p><a href="http://bayesianstats.com/the-apogee-of-art-philosophy-and-life-is-here">Permalink</a> | <a href="http://bayesianstats.com/the-apogee-of-art-philosophy-and-life-is-here#comment">Leave a comment  »</a> </p><img src="http://feeds.feedburner.com/~r/BayesianStatisticsBlog/~4/n1U_7VkMmOk" height="1" width="1"/> http://files.posterous.com/user_profile_pics/757812/bayes.jpg http://posterous.com/users/4wEYiML1RY6R Thomas Bayes mrbayes Thomas Bayes http://bayesianstats.com/the-apogee-of-art-philosophy-and-life-is-here Sun, 26 Sep 2010 04:24:00 -0700 Sunday R trick http://feedproxy.google.com/~r/BayesianStatisticsBlog/~3/F8CmkgV8OMo/sunday-r-trick http://bayesianstats.com/sunday-r-trick <p> <div><span style="font-family: arial, sans-serif; font-size: 13px; border-collapse: collapse;">This is not particularly Bayesian... you might think. But actually, computer graphics is a favourite hobby of mine, despite my having lived during the 18th century. Prince of Persia was good enough for us back then. I still remember drawing a Mandelbrot set on a graphical calculator.</span></div> <p /> <div><span style="font-family: arial, sans-serif; font-size: 13px; border-collapse: collapse;">Anyway, Bill Venables of CSIRO on the other side of the world from the Motherland sent the following delightful parlour trick to the <a href="https://stat.ethz.ch/mailman/listinfo/r-help">R-help mailing list</a>. Running the jif function with no parameters (i.e. by typing in </span><span style="font-size: 13px; border-collapse: collapse;"><span style="font-family: courier new, monospace;">jif()</span></span><span style="font-family: arial, sans-serif; font-size: 13px; border-collapse: collapse;"> at the R command line) displays the image further down. I don't think complex numbers are required for this to be implemented, but at least now I know how </span><span style="font-size: 13px; border-collapse: collapse;"><span style="font-family: courier new, monospace;">plot</span><span style="border-collapse: separate; font-size: small;"> </span></span><span style="font-family: arial, sans-serif; font-size: 13px; border-collapse: collapse;">works with complex numbers, which I guess was the point of the exercise.</span></div> <p /> <div><span style="font-family: arial, sans-serif; font-size: 13px; border-collapse: collapse;">Original posting follows:</span></div> <p /> <blockquote class="gmail_quote" style="margin-top: 0px; margin-right: 0px; margin-bottom: 0px; margin-left: 0.8ex; border-left-width: 1px; border-left-color: #cccccc; border-left-style: solid; padding-left: 1ex;">I was looking for an example of complex variables in R.  This one is trivial, but rather cute (though World War II aficionados may 'come over all funny').</blockquote> <blockquote class="gmail_quote" style="margin-top: 0px; margin-right: 0px; margin-bottom: 0px; margin-left: 0.8ex; border-left-width: 1px; border-left-color: #cccccc; border-left-style: solid; padding-left: 1ex;"><br /></blockquote> <blockquote class="gmail_quote" style="margin-top: 0px; margin-right: 0px; margin-bottom: 0px; margin-left: 0.8ex; border-left-width: 1px; border-left-color: #cccccc; border-left-style: solid; padding-left: 1ex;">See if you can guess the image before you try the function.  It's not difficult.</blockquote> <blockquote class="gmail_quote" style="margin-top: 0px; margin-right: 0px; margin-bottom: 0px; margin-left: 0.8ex; border-left-width: 1px; border-left-color: #cccccc; border-left-style: solid; padding-left: 1ex;"><br /></blockquote> <blockquote class="gmail_quote" style="margin-top: 0px; margin-right: 0px; margin-bottom: 0px; margin-left: 0.8ex; border-left-width: 1px; border-left-color: #cccccc; border-left-style: solid; padding-left: 1ex;"><div class="data type-r"> <table class="lines" cellspacing="0" cellpadding="0"> <tr> <td> <pre class="line_numbers"><span rel="#L1" id="L1">1</span> <span rel="#L2" id="L2">2</span> <span rel="#L3" id="L3">3</span> <span rel="#L4" id="L4">4</span> <span rel="#L5" id="L5">5</span> <span rel="#L6" id="L6">6</span> <span rel="#L7" id="L7">7</span> <span rel="#L8" id="L8">8</span> <span rel="#L9" id="L9">9</span> <span rel="#L10" id="L10">10</span> <span rel="#L11" id="L11">11</span> </pre> </td> <td width="100%"> <div class="highlight"><pre /><div class="line" id="LC1">jif <span class="o"><-</span> <span class="kr">function</span><span class="p">(</span>res <span class="o">=</span> <span class="m">100</span><span class="p">)</span> <span class="p">{</span></div><div class="line" id="LC2"> z <span class="o"><-</span> sample<span class="p">(</span>do.call<span class="p">(</span>complex<span class="p">,</span> subset<span class="p">(</span>expand.grid<span class="p">(</span>real <span class="o">=</span></div><div class="line" id="LC3">                                                 seq<span class="p">(</span><span class="m">-3</span><span class="p">,</span> <span class="m">4</span><span class="p">,</span> len <span class="o">=</span> <span class="m">7</span><span class="o">*</span>res <span class="o">+</span> <span class="m">1</span><span class="p">),</span></div><div class="line" id="LC4">                                                 imaginary <span class="o">=</span></div><div class="line" id="LC5">                                                 seq<span class="p">(</span><span class="m">-2</span><span class="p">,</span> <span class="m">2</span><span class="p">,</span> len <span class="o">=</span> <span class="m">4</span><span class="o">*</span>res <span class="o">+</span> <span class="m">1</span><span class="p">)),</span></div><div class="line" id="LC6">                                     real <span class="o">></span> <span class="m">-2.439</span> <span class="o">&</span> real <span class="o"><</span> <span class="m">3.717</span><span class="p">)))</span></div><div class="line" id="LC7"> del <span class="o"><-</span> <span class="m">2</span><span class="o">*</span>base<span class="p">::</span>pi<span class="o">/</span><span class="m">32</span></div><div class="line" id="LC8"> plot<span class="p">(</span>z<span class="p">,</span> type <span class="o">=</span> <span class="s">"n"</span><span class="p">,</span> asp <span class="o">=</span> <span class="m">1</span><span class="p">,</span> ann <span class="o">=</span> <span class="kc">FALSE</span><span class="p">,</span> axes <span class="o">=</span> <span class="kc">FALSE</span><span class="p">)</span></div><div class="line" id="LC9"> points<span class="p">(</span>z<span class="p">[((</span>Arg<span class="p">(</span>z<span class="p">)</span> <span class="o">+</span> del<span class="o">/</span><span class="m">2</span><span class="p">)</span> <span class="o">%/%</span> del<span class="p">)</span> <span class="o">%%</span> <span class="m">2</span> <span class="o">==</span> <span class="m">0</span></div><div class="line" id="LC10">          <span class="o">|</span> Mod<span class="p">(</span>z<span class="p">)</span> <span class="o"><</span> <span class="m">1.15</span><span class="p">],</span> col <span class="o">=</span> <span class="s">"red"</span><span class="p">,</span> pch <span class="o">=</span> <span class="s">"."</span><span class="p">)</span></div><div class="line" id="LC11"><span class="p">}</span></div></pre></div> </td> </tr> </table> </div><br /></blockquote> <p /> <div>This leads to the following image being drawn. Trying to work out how (or why) this happens is quite fun. Can you draw other Axis flags? (They lost WW2, so the least we can do in return is to use their flags as programming exercises).</div> <p /> <div><div class='p_embed p_image_embed'> <img alt="Jif" height="480" src="http://posterous.com/getfile/files.posterous.com/bayesianstats/8Wq1FZ7d2YqHxPwYP7Q37rbCe4wcXRMLN4yiJ0Tn967ZN7IIYAsEaYcLfE8n/jif.png.scaled500.png" width="480" /> </div> </div> </p> <p><a href="http://bayesianstats.com/sunday-r-trick">Permalink</a> | <a href="http://bayesianstats.com/sunday-r-trick#comment">Leave a comment  »</a> </p><img src="http://feeds.feedburner.com/~r/BayesianStatisticsBlog/~4/F8CmkgV8OMo" height="1" width="1"/> http://files.posterous.com/user_profile_pics/757812/bayes.jpg http://posterous.com/users/4wEYiML1RY6R Thomas Bayes mrbayes Thomas Bayes http://bayesianstats.com/sunday-r-trick Mon, 20 Sep 2010 08:07:36 -0700 He's even closer http://feedproxy.google.com/~r/BayesianStatisticsBlog/~3/HjIQN8Z3YXs/hes-even-closer http://bayesianstats.com/hes-even-closer <p> <a href="http://xkcd.com/795/">He's done it</a> <a href="http://bayesianstats.com/xkcd-is-getting-close">again</a>! Is there a funnier comic artist alive right now? (Notwithstanding that apparently Bayesians don't have a sense of humour, as alleged by a <a href="goog_2047252233">professor of </a><i><a href="http://bayesianstats.com/glymour-vs-dawid-fight-fight-fight#pcomment_commentunit_5414484">philosophy</a></i>).<p /><div><img title="conditional_risk.png" src="http://imgs.xkcd.com/comics/conditional_risk.png" alt="conditional_risk.png" /><br /></div> </p> <p><a href="http://bayesianstats.com/hes-even-closer">Permalink</a> | <a href="http://bayesianstats.com/hes-even-closer#comment">Leave a comment  »</a> </p><img src="http://feeds.feedburner.com/~r/BayesianStatisticsBlog/~4/HjIQN8Z3YXs" height="1" width="1"/> http://files.posterous.com/user_profile_pics/757812/bayes.jpg http://posterous.com/users/4wEYiML1RY6R Thomas Bayes mrbayes Thomas Bayes http://bayesianstats.com/hes-even-closer Thu, 15 Jul 2010 14:12:24 -0700 Illegal statistics http://feedproxy.google.com/~r/BayesianStatisticsBlog/~3/FDJQvfwjCSY/illegal-statistics http://bayesianstats.com/illegal-statistics <p> <a href="http://magazine.amstat.org/blog/2010/07/01/statreasoning710/2/">Even the Supreme Court of the United States of America can't define a p-value properly</a>:<p /><div><blockquote class="gmail_quote" style="margin-top: 0px; margin-right: 0px; margin-bottom: 0px; margin-left: 0.8ex; border-left-width: 1px; border-left-color: rgb(204, 204, 204); border-left-style: solid; padding-left: 1ex;"> Our experience with this case also suggests the judiciary would benefit from a better understanding of fundamental concepts of hypothesis testing. Both the U.S. and Michigan Supreme Court opinions state, “Standard deviation analysis seeks to determine the probability that the disparity between a group’s jury-eligible population and the group’s percentage in the qualified jury pool is attributable to random chance.” The probability referred to, of course, is the <span style="margin-top: 0px; margin-right: 0px; margin-bottom: 0px; margin-left: 0px; padding-top: 0px; padding-right: 0px; padding-bottom: 0px; padding-left: 0px;">p</span>-value, which is calculated assuming random (chance) selection of the jury pool from the eligible population.</blockquote> <p /><div>But then, should we be surprised? <a href="http://www.conceptstew.co.uk/PAGES/prosecutors_fallacy.html">Lawyers have a whole statistical misconception named after them, after all</a>. And they have a habit of <a href="http://www.rss.org.uk/PDF/RSS%20Statement%20regarding%20statistical%20issues%20in%20the%20Sally%20Clark%20case,%20October%2023rd%202001.pdf">condemning innocent people based on dodgy statistics</a>...</div> </div> </p> <p><a href="http://bayesianstats.com/illegal-statistics">Permalink</a> | <a href="http://bayesianstats.com/illegal-statistics#comment">Leave a comment  »</a> </p><img src="http://feeds.feedburner.com/~r/BayesianStatisticsBlog/~4/FDJQvfwjCSY" height="1" width="1"/> http://files.posterous.com/user_profile_pics/757812/bayes.jpg http://posterous.com/users/4wEYiML1RY6R Thomas Bayes mrbayes Thomas Bayes http://bayesianstats.com/illegal-statistics Sun, 27 Jun 2010 13:10:00 -0700 Poorly calibrated football pundit http://feedproxy.google.com/~r/BayesianStatisticsBlog/~3/y5655b3qMjQ/poorly-calibrated-football-pundit http://bayesianstats.com/poorly-calibrated-football-pundit <p> <p><div class='p_embed p_image_embed'> <a href="http://posterous.com/getfile/files.posterous.com/bayesianstats/RYvBBRsd10BuWFsETC7xiEzF9zhQwW0j1kF49HE4rRtBDgqtjpcGa1nza9uc/snapshot17.png"><img alt="Snapshot17" height="283" src="http://posterous.com/getfile/files.posterous.com/bayesianstats/Cx8gVwVKliO5ChKH0BSCt8YikQHCBRJiDVUgVGHQ6FYKW88c8KujBoqVDII4/snapshot17.png.scaled.500.jpg" width="500" /></a> </div> </p> <p>From the <a href="http://news.bbc.co.uk/sport1/hi/football/world_cup_2010/matches/match_51">BBC's coverage of the Germany-England World Cup match</a> (where Germany convincingly and utterly unsurprisingly thrashed England), we had this unfortunate prior distribution from the pundit Mark Lawrenson:</p> <div> <div> <blockquote class="posterous_short_quote">It is England v Germany so it is 50-50 but England have better players, or more better players, so I expect them to win.</blockquote> </div> <div>Oh dear. Lampard had the right reaction to that, as you can see above!</div> <div> </div> <div> </div> </div> </p> <p><a href="http://bayesianstats.com/poorly-calibrated-football-pundit">Permalink</a> | <a href="http://bayesianstats.com/poorly-calibrated-football-pundit#comment">Leave a comment  »</a> </p><img src="http://feeds.feedburner.com/~r/BayesianStatisticsBlog/~4/y5655b3qMjQ" height="1" width="1"/> http://files.posterous.com/user_profile_pics/757812/bayes.jpg http://posterous.com/users/4wEYiML1RY6R Thomas Bayes mrbayes Thomas Bayes http://bayesianstats.com/poorly-calibrated-football-pundit Wed, 24 Mar 2010 11:58:49 -0700 Bad Bayes still bad http://feedproxy.google.com/~r/BayesianStatisticsBlog/~3/d_1RFhqvunI/bad-bayes-still-bad http://bayesianstats.com/bad-bayes-still-bad <p> <a href="http://tamino.wordpress.com/">Tamino</a>, a notorious "climate change" blogger, is alleged to also be a statistician. <a href="http://tamino.wordpress.com/2010/03/11/not-a-random-walk/">He certainly seems to know something about time series</a>. (Thanks to <a href="http://motls.blogspot.com/2008/09/who-is-tamino-grant-foster-identity.html">this investigation</a>, we know that Tamino is <a href="http://www.aavso.org/news/foster.shtml">Grant Foster</a>, writer of <a href="http://www.eastangliaemails.com/emails.php?eid=1019&filename=1254163518.txt">"blog diatribe"-style</a> climate <a href="http://ccsm.ucar.edu/cas/Trenberth/trenberth.papers/Foster_et%20alJGR09_formatted.pdf">papers</a>. His affiliation in the linked paper is "Tempo Analytics, Westbrook, Maine", but I can't find any other reference to it online). Unfortunately he might be somewhat off-base when it comes to other statistical principles. His <a href="http://tamino.wordpress.com/2010/03/22/good-bayes-gone-bad/">discussion of Bayesian analysis</a> is so confused that I'll leave it to Andrew Gelman, professor of statistics at Columbia University, <a href="http://scienceblogs.com/appliedstatistics/2010/03/hey_statistics_is_easy.php">to summarise it for us</a>: <blockquote>Kent Holsinger sends along <a href="http://tamino.wordpress.com/2010/03/22/good-bayes-gone-bad/">this</a> statistics discussion from a climate scientist. I don't really feel like going into the details on this one, except to note that this appears to be a discussion between two physicists about statistics. The blog in question appears to be pretty influential, with about 70 comments on most of its entries. When it comes to blogging, I suppose it's good to have strong opinions even (especially?) when you don't know what you're talking about.</blockquote> Update: Gelman repeated himself on his academic blog, where he elaborates on his opinion in the comments. It's strange that when I tried commenting (twice) on "Tamino"'s blog to refer him to Gelman's comments, I didn't succeed; but when someone else did the same but with the qualifier that "[Gelman] comes around to Tamino’s side" [which not actually true] in his later comments the link appears. At the time of writing the comment thread ends with "Tamino" abusing a commenter trying to correct one of his calculations until he eventually admits he was indeed wrong. Oh dear. </p> <p><a href="http://bayesianstats.com/bad-bayes-still-bad">Permalink</a> | <a href="http://bayesianstats.com/bad-bayes-still-bad#comment">Leave a comment  »</a> </p><img src="http://feeds.feedburner.com/~r/BayesianStatisticsBlog/~4/d_1RFhqvunI" height="1" width="1"/> http://files.posterous.com/user_profile_pics/279821/5250362.gif http://posterous.com/users/36zyxeYjLc2Z Guy Freeman Guy Guy Freeman http://bayesianstats.com/bad-bayes-still-bad Mon, 15 Mar 2010 15:20:56 -0700 Another sampling from the great frequentist malpractice genre in the sky http://feedproxy.google.com/~r/BayesianStatisticsBlog/~3/7ElNrsWllJc/another-sampling-from-the-great-frequentist-m http://bayesianstats.com/another-sampling-from-the-great-frequentist-m <p> That this isn't well-known amongst the general public is a disgrace, but the "scientific method" as carried out by academic careerists has long been only a poor substitute for real science: <blockquote class="posterous_medium_quote">It’s science’s dirtiest secret: The “scientific method” of testing hypotheses by statistical analysis stands on a flimsy foundation. Statistical tests are supposed to guide scientists in judging whether an experimental result reflects some real effect or is merely a random fluke, but the standard methods mix mutually inconsistent philosophies and offer no meaningful basis for making such decisions. Even when performed correctly, statistical tests are widely misunderstood and frequently misinterpreted. As a result, countless conclusions in the scientific literature are erroneous, and tests of medical dangers or treatments are often contradictory and confusing.</blockquote> From <a href="http://www.sciencenews.org/view/feature/id/57091/title/Odds_are,_its_wrong">sciencenews.org</a>. Then follows the usual errors relating to interpretation of hypothesis tests and other applied frequentist gunk. There is an interesting point made about how randomisation isn't all that (although what the alternative should be is anyone's guess), before... behold! <blockquote class="posterous_short_quote">Such sad statistical situations suggest that the marriage of science and math may be desperately in need of counseling. Perhaps it could be provided by the Rev. Thomas Bayes.</blockquote> A lovely line. Whether this latest example of the litany against the standard operating procedure of too many scientists from all disciplines will change anything more than the previous attempts to do so is moot. </p> <p><a href="http://bayesianstats.com/another-sampling-from-the-great-frequentist-m">Permalink</a> | <a href="http://bayesianstats.com/another-sampling-from-the-great-frequentist-m#comment">Leave a comment  »</a> </p><img src="http://feeds.feedburner.com/~r/BayesianStatisticsBlog/~4/7ElNrsWllJc" height="1" width="1"/> http://files.posterous.com/user_profile_pics/279821/5250362.gif http://posterous.com/users/36zyxeYjLc2Z Guy Freeman Guy Guy Freeman http://bayesianstats.com/another-sampling-from-the-great-frequentist-m Sun, 14 Feb 2010 15:28:00 -0800 Glymour vs Dawid: fight! fight! fight! http://feedproxy.google.com/~r/BayesianStatisticsBlog/~3/KgndjHovCxY/glymour-vs-dawid-fight-fight-fight http://bayesianstats.com/glymour-vs-dawid-fight-fight-fight <p> <p>This delightful example of an academic hissy fit is from almost a year ago, but still has the power to shock in its lack of professionalism. I hope it shatters the all-too-common stereotype amongst some people of academics as calm intellectuals who want nothing more than to hear each other's honestly held views. Phooey to that.</p> <p>The topic of discussion (if it can be called that when the level of aggression is this high) is Phil Dawid's article <a href="http://clopinet.com/isabelle/Projects/reading/Dawid_NIPS08_causality_preprint.pdf">"Beware of the DAG"</a>, which makes very reasonable points about what possible causal inferences can be undertaken for different levels of "causal" assumptions on a Directed Acyclic Graph (which is more commonly known as a Bayesian Network when the DAG encodes conditional independence statements, as is the case here), and discusses how reasonable and testable these "causal" assumptions are. In particular, he dismisses the popular but mostly deluded endeavour of "causal discovery", which is at once ill-posed and, so far, in my opinion, ill-answered. This hits right at what <a href="http://www.hss.cmu.edu/philosophy/faculty-glymour.php">Clark Glymour</a> is trying to sell, though. But that still doesn't quite explain what <em>caused</em> this reaction...</p> <p>NB. No-one from the mailing list replied to this directly, as far as I can tell. The archive for that month is available <a href="http://mail.encours.org/pipermail/causality-ml/2009-April.txt">here</a>.</p> <p>And now, enjoy, typos included:</p> <p>From: <strong>Clark Glymour</strong> <cg09></p> <p>Date: Thu, Apr 23, 2009 at 01:12</p> <p>Subject: [Causality-ML] Professor Dawid's paper</p> <p>To: causality-ml</p> <p>Professor Dawid ‘s worry is announced in his abstract:</p> <p>“My fundamental concern is the relationship between, on the one hand, properties or concepts relating to an external reality, such as probabilistic independence or causality,which we wish to elucidate or manipulate; and, on the other hand, formal representations of such properties by means of mathematical or logical structures, such as graphs. It is important to avoid confusing the picture with the reality.”</p> <p>Be at ease. Absolutely not to worry. I have never once, not once, seen someone draw a graph or write a formula when they actually thought they were manipulating what the symbols were supposed to denote. Not once. Word and object, we are ace at distinguishing those. So I thought, having solved Professor Dawid’s concern, I should stop, but I read on a little ways.</p> <p>To a really important announcement:</p> <p>“it is … worthy of continual repetition and emphasis, that there is absolutely no logical reason for there to be any connexion whatsoever between observations made under the different regimes of seeing and doing: a system may very well behave entirely differently when it is kicked than when it is left alone”</p> <p>Good point that. I checked my logic books, no proofs of that connection. Also no proofs that the past ever was, no proofs that the future will come to be, no proofs that Professor Dawid has mental states, no proofs that an external world exists, no proofs that the so-called laws of nature will hold next week. Not much use, those logic books, unless you assume or hypothesize stuff and then want to know the consequences.</p> <p>On the other hand, there is this funny literature—I wonder if Professor Dawid has read it—where people investigate when, under what various assumptions about the world, other things follow. Like, for example, there is this subject called Euclidean geometry where assumptions are make about space, and then all kinds of interesting other things are proved about space, really amazing stuff—you could use it to design buildings even. But I read that the assumptions are not always true. Pity. Also, there was this guy Newton who had these three assumptions, and then some “rules of reasoning” at the back. He got these amazing consequences, which mostly turned out to be correct, although I hear that his assumptions don’t always hold, and I sure could not find his rules of reasoning in my logic book.</p> <p>I guess it can’t be the same with observing and doing. There just couldn’t be any assumptions about the connections and proofs from assumptions that you can make the kind of inferences Professor Dawid is talking about—inferences from observations to effects of actions. Or proofs that under other assumptions you can’t. Couldn’t be. So, nah…</p> <p>In fact, Professor Dawid is really helpful about this. He tells us that if we make the wrong assumptions, or not enough of the right ones, we won’t get that logical connection between seeing and doing. Not a chance of it.</p> <p>“We say that a DAG D with node-set V, a set of variables, represents a collection C of CIproperties over V if the relation (bunch of symbols here, way over my head) is in C if and only if S and T are d-separated by U in D. This relationship between a D and a collection of CI properties will constitute our semantic interpretation of a DAG.”</p> <p>Well sure enough, one thing talks about causality, the other talks just about probability. Different terms. My logic book tells me there have to be terms in common between the premises and the conclusion—unless the conclusion is logically true. Kind of like “force” and “acceleration” or like “probability” and “unbiased,” or “perpetual” and “motion”--no logical connection. So Professor Dawid really nailed that one.</p> <p>Well, I should go on reading this stuff, like how we should just talk about probability because graph theory isn’t mathematics (so many silly people who thought they were doing mathematics) and how science is all about conditional independence not causation (I knew those physicists and chemists and epidemiologists had to be crazy talking about what does and doesn’t cause what), and I am sure he has discovered a lot more stuff than those crazy causal guys who think they have methods that have discovered errors in a mass spectrometer aboard a satellite (imagine—they weren’t ever there), and how to tell what rocks are made of from the radiation bouncing off them, and how to reduce the rate of college dropouts in a college, and that acid rain caused plant die offs in an estuary, and the processes that go on in the brain in an experiment (something about fmri), and even global climate teleconnections—those guys are so crazy. But since I solved the problem Professor Dawid had at t!</p> <p>he beginning, I will just have a martini.</p> <p>Bye,</p> <p>Clark Glymour</p> </p> <p><a href="http://bayesianstats.com/glymour-vs-dawid-fight-fight-fight">Permalink</a> | <a href="http://bayesianstats.com/glymour-vs-dawid-fight-fight-fight#comment">Leave a comment  »</a> </p><img src="http://feeds.feedburner.com/~r/BayesianStatisticsBlog/~4/KgndjHovCxY" height="1" width="1"/> http://files.posterous.com/user_profile_pics/279821/5250362.gif http://posterous.com/users/36zyxeYjLc2Z Guy Freeman Guy Guy Freeman http://bayesianstats.com/glymour-vs-dawid-fight-fight-fight Sun, 31 Jan 2010 11:01:03 -0800 Oh dear indeed http://feedproxy.google.com/~r/BayesianStatisticsBlog/~3/aCcultZp0hA/oh-dear-indeed http://bayesianstats.com/oh-dear-indeed <p> From <a href="http://timworstall.com/2010/01/27/oh-dear-27/">one of Tim Worstall's "Oh dear" posts</a>: <blockquote class="posterous_short_quote"> Maybe inequality and poverty in modern Britain are important and maybe they’re not. It’s entirely possible to argue it either way and to a large extent depends upon your Bayesian priors. </blockquote> So naturally I had to <a href="http://timworstall.com/2010/01/27/oh-dear-27/#comment-40082">reply</a>, <blockquote class="posterous_medium_quote"> Actually this is a value judgment and would be expressed through a utility function. Bayesian priors (and posteriors) are probability distributions expressing subjective degrees of certainty over parameters of interest. </blockquote> No indication thus far that the blog author has taken this on board... Is it better that Bayesian concepts are invoked incorrectly rather than not at all? I believe so, but we must continue to strive towards fuller understanding of them amongst non-statisticians. Because otherwise we have to deal with things like <a href="http://bit.ly/2Qml6j">this</a>: <blockquote class="posterous_medium_quote"> First, what did you think was the probability of success in Afghanistan before the mission began? This is the prior probability, which we’ll call Ps. The probability of failure, Pf, is one minus this. </blockquote> <blockquote class="posterous_medium_quote"> Second, what is the probability that we’d see the number of deaths we have, if the mission were succeeding? Call this Pd|s. <strong>One minus this gives us Pd|f</strong>. [emphasis mine, calculation thankfully not] </blockquote> Again, comments to the contrary had no effect on deflecting the author in his enthusiasm on this occasion. Oh dear. </p> <p><a href="http://bayesianstats.com/oh-dear-indeed">Permalink</a> | <a href="http://bayesianstats.com/oh-dear-indeed#comment">Leave a comment  »</a> </p><img src="http://feeds.feedburner.com/~r/BayesianStatisticsBlog/~4/aCcultZp0hA" height="1" width="1"/> http://files.posterous.com/user_profile_pics/279821/5250362.gif http://posterous.com/users/36zyxeYjLc2Z Guy Freeman Guy Guy Freeman http://bayesianstats.com/oh-dear-indeed Thu, 17 Sep 2009 10:40:55 -0700 Book review: "Bad Science" by Ben Goldacre http://feedproxy.google.com/~r/BayesianStatisticsBlog/~3/vaDDKpGpl6k/book-review-bad-science-by-ben-goldacre http://bayesianstats.com/book-review-bad-science-by-ben-goldacre <p> "Bad Science" by Ben Goldacre, £12.99, 338pp, Fourth Estate Bought as a present for someone else, borrowed temporarily to see what the fuss is about Defends "science" against its enemies No-nonsense style Focuses on medical matters Political comments mar it Slight over-simplifications p-value mistake HG Wells quote mistake? Is "perfect science" possible? Bayes is the only option we have? </p> <p><a href="http://bayesianstats.com/book-review-bad-science-by-ben-goldacre">Permalink</a> | <a href="http://bayesianstats.com/book-review-bad-science-by-ben-goldacre#comment">Leave a comment  »</a> </p><img src="http://feeds.feedburner.com/~r/BayesianStatisticsBlog/~4/vaDDKpGpl6k" height="1" width="1"/> http://files.posterous.com/user_profile_pics/279821/5250362.gif http://posterous.com/users/36zyxeYjLc2Z Guy Freeman Guy Guy Freeman http://bayesianstats.com/book-review-bad-science-by-ben-goldacre Wed, 16 Sep 2009 11:06:03 -0700 "Is risk management too complicated and subtle for InfoSec?" --- I think just mathematics is too complicated and subtle for some people http://feedproxy.google.com/~r/BayesianStatisticsBlog/~3/AlSQj4SNMX8/is-risk-management-too-complicated-and-subtle http://bayesianstats.com/is-risk-management-too-complicated-and-subtle <p> It's interesting to see how knowledge of Bayesian methods exists in certain fields while ignorance of the details leads to weird conclusions concerning their usage. A good example of this phenomenon is<a href="http://superconductor.voltage.com/2009/09/the-twoenvelope-problem-in-risk-management.html"> this mangling of the two-envelope problem</a> --- supposedly a "paradox" that Bayesian decision analysis fails at --- which is then used to argue that therefore Bayesian analysis of risks is actually useless and that instead <blockquote class="posterous_medium_quote">In the absence of reliable risk information, a similar approach to information security may be the best that we can do – just try different things and see which works the best. You might call this approach “experimental security.” There may be no better approach.</blockquote> Yeah, just experimenting without any inferential tools makes sense... Funny how it allows the analyst to believe anything he wants without anything to back it up. The takedown is painstakingly given <a href="http://newschoolsecurity.com/2009/09/is-risk-management-too-complicated-and-subtle-for-infosec/">here</a>, but the only comment to it at the time of writing should make it clear just how entrenched the forces of "irrational pragmatism" are: <blockquote class="posterous_medium_quote">They Bayesian approach has many beautiful mathematical properties, but it fails to make contact with reality — it has no pragmatics. Worse, it fails to recognize that there is more than one person in the world. In the Bayesian world there is only one subjective probability, “mine”. The fact that you exist and have your own subjectivity that just might have something to do with our agreed-upon response to any particular problem is totally irrelevant. All the technical mathematical results in the world can’t get past these foundational problems.</blockquote> Wouldn't it be better to admit ignorance of the issues at hand and then give your opinions on that basis rather than just spout nonsense? There is clearly much education about Bayesian analysis to be done, starting with demolishing incorrect preconceptions that are already out there. </p> <p><a href="http://bayesianstats.com/is-risk-management-too-complicated-and-subtle">Permalink</a> | <a href="http://bayesianstats.com/is-risk-management-too-complicated-and-subtle#comment">Leave a comment  »</a> </p><img src="http://feeds.feedburner.com/~r/BayesianStatisticsBlog/~4/AlSQj4SNMX8" height="1" width="1"/> http://files.posterous.com/user_profile_pics/279821/5250362.gif http://posterous.com/users/36zyxeYjLc2Z Guy Freeman Guy Guy Freeman http://bayesianstats.com/is-risk-management-too-complicated-and-subtle Sun, 08 Mar 2009 12:44:47 -0700 xkcd is getting close http://feedproxy.google.com/~r/BayesianStatisticsBlog/~3/u-Fge_tFfvA/xkcd-is-getting-close http://bayesianstats.com/xkcd-is-getting-close <p> <div class='p_embed p_image_embed'> <img alt="Media_httpimgsxkcdcom_uypsi" height="185" src="http://posterous.com/getfile/files.posterous.com/import-cfwf/ezwxGtxIfcefGucbqJGlkxpvlCEHvhbGghdghgdsuccjgIDDtGmpfJJfoGCw/media_httpimgsxkcdcom_uypsi.png.scaled500.png" width="459" /> </div> After an episode about <a href="http://xkcd.com/539/">significance</a>, xkcd has now reached the intellectual stage of <a href="http://xkcd.com/552/">joking about correlation and causation</a> and how they might -- or might not -- relate to each other. So the question is: when is he [= MAP gender of comic creator] going to mention Bayes in some way? It's only a matter of time, surely, and I'll take that as a sign that it's as mainstream (in geek circles, anyway -- although geeks seem pretty mainstream these days) as the frequentist idea of significance. And then I'll sue for patent infringement. I might be joking. </p> <p><a href="http://bayesianstats.com/xkcd-is-getting-close">Permalink</a> | <a href="http://bayesianstats.com/xkcd-is-getting-close#comment">Leave a comment  »</a> </p><img src="http://feeds.feedburner.com/~r/BayesianStatisticsBlog/~4/u-Fge_tFfvA" height="1" width="1"/> http://files.posterous.com/user_profile_pics/279821/5250362.gif http://posterous.com/users/36zyxeYjLc2Z Guy Freeman Guy Guy Freeman http://bayesianstats.com/xkcd-is-getting-close Thu, 12 Feb 2009 16:34:23 -0800 How many meanings can Bayesian statistics have? or When can I get off this ship? http://feedproxy.google.com/~r/BayesianStatisticsBlog/~3/ncsBzKzrTzo/how-many-meanings-can-bayesian-statistics-hav http://bayesianstats.com/how-many-meanings-can-bayesian-statistics-hav <p> Andrew Gelman, whose <a href="http://www.stat.columbia.edu/~cook/movabletype/mlm/">blog</a> is always a good read (and is also updated much more often than this one!), provoked a discussion about <a href="http://www.stat.columbia.edu/~cook/movabletype/archives/2009/02/different-meani.html">the different "meanings" of Bayesian statistics</a>. You might find the comments there interesting; I admit I found the whole thing a little hair-splitting for my taste. <a href="http://xianblog.wordpress.com/">Christian Robert</a> -- who wrote the superb book <a href="http://www.amazon.com/gp/product/0387715983?ie=UTF8&tag=chrprobboo-20&linkCode=as2&camp=1789&creative=9325&creativeASIN=0387715983">"The Bayesian Choice"</a> -- started the whole thing off by describing his bemusement over how much <a href="http://xianblog.wordpress.com/2009/01/22/bayes-theorem/">"fascination for Bayes’ Theorem [there] seems to be outside Statistics"</a>. After all, it's just a theorem, right? He contends that the theorem itself mustn't be confused with the interpretation of the axioms of probability, which is the contentious and "interesting" aspect of the whole endeavour. But I feel this is pedantic. A "Bayesian" is almost always someone who believes in the subjective interpretation of probability statements, so that Bayes' Theorem can be used as an means to update one's beliefs about quantities, hypotheses, and so on. Frequentists don't reject Bayes' Theorem itself -- they can't, as it's just a consequence of the probability axioms, and they even use it uncontentiously for calculating diagnostic test properties such as predictive value -- but they do reject its use for "updating beliefs". Christian also seems bemused by <a href="http://yudkowsky.net/rational/bayes">a rather long but entertaining "justification"/explanation for Bayes' Theorem</a> that I've been meaning to link to for a long time. [It is one of the top results that come up when googling for the term "Bayesian"]. Again, if you accept the axioms of probability, the theorem is just a consequence, so you don't need to justify it any other way, he maintains. (As he put it, <blockquote class="posterous_medium_quote">The theorem per se offers no difficulty, so this may be due to the counter-intuitive inversion of probabilities as the one found in the example of the first blog. But the fact that people often confuse probabilities of causes and probabilities of effects—i.e. the right order of conditioning—does not require a deeper explanation for Bayes’ theorem, rather a pointer at causal reasoning!</blockquote> ) But just being told something is true, and even being convinced it is true mathematically, doesn't help most people. To get them to understand it intuitively sometimes requires something more, like a story, or an example. Not everyone is a mathematician after all... [I'm sorry if I misrepresented anyone's views here. Please tell me if I've got something wrong and I'll try and put it right.] </p> <p><a href="http://bayesianstats.com/how-many-meanings-can-bayesian-statistics-hav">Permalink</a> | <a href="http://bayesianstats.com/how-many-meanings-can-bayesian-statistics-hav#comment">Leave a comment  »</a> </p><img src="http://feeds.feedburner.com/~r/BayesianStatisticsBlog/~4/ncsBzKzrTzo" height="1" width="1"/> http://files.posterous.com/user_profile_pics/279821/5250362.gif http://posterous.com/users/36zyxeYjLc2Z Guy Freeman Guy Guy Freeman http://bayesianstats.com/how-many-meanings-can-bayesian-statistics-hav Sun, 01 Feb 2009 17:20:13 -0800 A new look; a new start too? http://feedproxy.google.com/~r/BayesianStatisticsBlog/~3/mRDC6Jry4Ls/a-new-look-a-new-start-too http://bayesianstats.com/a-new-look-a-new-start-too <p> I've upgraded the Wordpress backend to version 2.7, which you probably didn't notice. More obviously, I've changed the aesthetics around here to something more... professional. I think the layout is cleaner and the font is clearer and bigger. Tell me what you think of it. It's evident that there is still a need for more advocacy of Bayesian methods for all sorts of quantitative analyses, let alone for Bayesianism as a lifestyle choice! I want to use this site, including the only part of it that has any substance -- this blog -- to further the goal of Bayesianist awareness. If you have any ideas for how I can help, e.g. by including links to tutorials explaining what Bayesian statistics is, or by setting up a forum where Bayesianists can discuss anything from philosophical issues to how to get their colleagues thinking in a Bayesian way, then share them with me. </p> <p><a href="http://bayesianstats.com/a-new-look-a-new-start-too">Permalink</a> | <a href="http://bayesianstats.com/a-new-look-a-new-start-too#comment">Leave a comment  »</a> </p><img src="http://feeds.feedburner.com/~r/BayesianStatisticsBlog/~4/mRDC6Jry4Ls" height="1" width="1"/> http://files.posterous.com/user_profile_pics/279821/5250362.gif http://posterous.com/users/36zyxeYjLc2Z Guy Freeman Guy Guy Freeman http://bayesianstats.com/a-new-look-a-new-start-too