In yesterday‘s battle of the religions, the strongest argument against Mother Teresa was given by Paul, who related that she was friends with all sorts of nasty politicians and that she’s been accused of spending money that came from questionable sources. But if that’s all you can say about her, it won’t cut much ice with Sun Myung Moon, who also was friends with various unsavory characters and scammed all sorts of money. So Moon is more of a badass. But this isn’t a contest about who’s the toughest guy—we’re looking for a seminar speaker here, not a rogue sociologist.

I’ll have to with Mother Teresa. I doubt we’ll see any faith healings but I’m persuaded by Ken’s negative case against Moon:

Moon’s wikipedia page describes him as a “religious leader, businessperson, political activist, and media mogul.” In my experience people with “rock star” status like this make for bad seminar speakers because they tend to be full of anecdotes and fluff, and light on rigorous empirical evidence.

Good point. The last thing we need here is a goddam Ted talk.

And today we have a contest between two artists! Vincent is more famous and would certainly be the bigger draw, but I wouldn’t be surprised if Grandma could give a more coherent lecture. On the other hand, according to wikipedia, “she was a Society of Mayflower Descendants and Daughters of the American Revolution member.” And that sounds pretty duuuulllllllll. It’s up to all of you to make the strongest and wittiest arguments on both sides.

P.S. As always, here’s the background, and here are the rules.

The post Vincent van Gogh (3) vs. Grandma Moses appeared first on Statistical Modeling, Causal Inference, and Social Science.

Please comment on the article here: Statistical Modeling, Causal Inference, and Social Science

The post Vincent van Gogh (3) vs. Grandma Moses appeared first on All About Statistics.

Ben O’Neill writes:

I am a fellow Bayesian statistician at the University of New South Wales (Australia).  I have enjoyed reading your various books and articles, and enjoyed reading your recent article on The Perceived Absurdity of Bayesian Inference.  However, I disagree with your assertion that the “doomsday argument” is non-Bayesian; I think if you read how it is presented by Leslie you will see that it is at least an attempt at a Bayesian argument.  In any case, although it has enough prima facie plausibility to trick people, the argument is badly flawed, and not a correct application of Bayesian reasoning.  I don’t think it is a noose around the Bayesian neck.

Anyway, I’m just writing because I thought you might be interested in a recent paper on this topic in the Journal of Philosophy.  The paper is essentially a Bayesian refutation of the doomsday argument, pointing out how it goes wrong, and how it is an incorrect application of Bayesian inference.  (And also how a correct application of Bayesian inference leads to sensible conclusions.)  Essentially, the argument confuses total series length with remaining series length, and sneaks information from the data into the prior in a way which is invalid.  Once this is corrected the absurd conclusions of the doomsday argument evaporate.

I don’t really have anything more to say on this topic (here’s my argument from 2005 as to why I think the doomsday argument is clearly frequentist and not particularly Bayesian) but I thought some of you might be interested, hence the pointer.

The post Bayes and doomsday appeared first on Statistical Modeling, Causal Inference, and Social Science.

Please comment on the article here: Statistical Modeling, Causal Inference, and Social Science

The post Bayes and doomsday appeared first on All About Statistics.

Thanks to the Excel stylings of Paul Davidson:

Our competition is (approximately) 1/4 done!

And I’ve been thinking about possible categories for next year’s tourney:

New Jersey politicians
Articulate athletes
Plagiarists
People named Greg or Gregg
Vladimir Nabokov and people connected to him
. . .

Ummm, we need 3 more categories. Any suggestions? Real people only, please. In some future year we can have an all-fictional category.

The post The bracket so far appeared first on Statistical Modeling, Causal Inference, and Social Science.

Please comment on the article here: Statistical Modeling, Causal Inference, and Social Science

The post The bracket so far appeared first on All About Statistics.

For yesterday, I’ll have to go with Gandhi, the original badass of nonviolence. Zbicyclist found this quote, “He propagated that . . . we should take only that which is required, in minimum quantity. We should not eat to appease our taste buds,” which implies that Gandhi shouldn’t pick the caterer—but that’s not an issue, we never have any special food with our seminars anyway.

Jonathan argued in favor of Kubrick based on this quote: “Well, you don’t make it easy on viewers or critics. You’ve said you want an audience to react emotionally. You create strong feelings, but you won’t give us any easy answers. That’s because I don’t have any easy answers.”

But, as a Bayesian, I do want easy answers, so this line of argument doesn’t work for me.

And now for today we have two people who’ve done a lot, but might not be very articulate seminar speakers. The nun is seeded #4 in the Religious Leaders category; Sun Myung Moon is listed under Cult Figures. For sheer spectacle, ya gotta go with the Moonies. On the other hand, with Mother Teresa we might see a documented miracle. According to Wikipedia, “In 2002, the Vatican recognised as a miracle the healing of a tumor in the abdomen of an Indian woman, Monica Besra, after the application of a locket containing Mother Teresa’s picture.”

It’s been 13 years, so maybe Teresa is up for another miraculous cure?

P.S. As always, here’s the background, and here are the rules.

The post Mother Teresa (4) vs. Sun Myung Moon appeared first on Statistical Modeling, Causal Inference, and Social Science.

Please comment on the article here: Statistical Modeling, Causal Inference, and Social Science

The post Mother Teresa (4) vs. Sun Myung Moon appeared first on All About Statistics.

John Carlin, who’s collaborated on some of my recent work on Type S and Type M errors, prepared this presentation for a clinical audience. It might be of interest to some of you. The ideas and some of the examples should be familiar to regular readers of this blog, but it could be useful to see how this material is presented to a different sort of group than I usually encounter.

The post Statistical Significance – Significant Problem? appeared first on Statistical Modeling, Causal Inference, and Social Science.

Please comment on the article here: Statistical Modeling, Causal Inference, and Social Science

The post Statistical Significance – Significant Problem? appeared first on All About Statistics.

Please comment on the article here: Econometrics Beat: Dave Giles' Blog

The post Applied Nonparametric Econometrics appeared first on All About Statistics.

.

As part of the week of recognizing R.A.Fisher (February 17, 1890 – July 29, 1962), I reblog Senn from 3 years ago.  

‘Fisher’s alternative to the alternative’

By: Stephen Senn

[2012 marked] the 50th anniversary of RA Fisher’s death. It is a good excuse, I think, to draw attention to an aspect of his philosophy of significance testing. In his extremely interesting essay on Fisher, Jimmie Savage drew attention to a problem in Fisher’s approach to testing. In describing Fisher’s aversion to power functions Savage writes, ‘Fisher says that some tests are more sensitive than others, and I cannot help suspecting that that comes to very much the same thing as thinking about the power function.’ (Savage 1976) (P473).

The modern statistician, however, has an advantage here denied to Savage. Savage’s essay was published posthumously in 1976 and the lecture on which it was based was given in Detroit on 29 December 1971 (P441). At that time Fisher’s scientific correspondence did not form part of his available oeuvre but in1990 Henry Bennett’s magnificent edition of Fisher’s statistical correspondence (Bennett 1990) was published and this throws light on many aspects of Fisher’s thought including on significance tests.

.

The key letter here is Fisher’s reply of 6 October 1938 to Chester Bliss’s letter of 13 September. Bliss himself had reported an issue that had been raised with him by Snedecor on 6 September. Snedecor had pointed out that an analysis using inverse sine transformations of some data that Bliss had worked on gave a different result to an analysis of the original values. Bliss had defended his (transformed) analysis on the grounds that a) if a transformation always gave the same result as an analysis of the original data there would be no point and b) an analysis on inverse sines was a sort of weighted analysis of percentages with the transformation more appropriately reflecting the weight of information in each sample. Bliss wanted to know what Fisher thought of his reply.

Fisher replies with a ‘shorter catechism’ on transformations which ends as follows:

A…Have not Neyman and Pearson developed a general mathematical theory for deciding what tests of significance to apply?

B…Their method only leads to definite results when mathematical postulates are introduced, which could only be justifiably believed as a result of extensive experience….the introduction of hidden postulates only disguises the tentative nature of the process by which real knowledge is built up. (Bennett 1990) (p246)

It seems clear that by hidden postulates Fisher means alternative hypotheses and I would sum up Fisher’s argument like this. Null hypotheses are more primitive than statistics: to state a null hypothesis immediately carries an implication about an infinity of test
statistics. You have to choose one, however. To say that you should choose the one with the greatest power gets you nowhere. This power depends on the alternative hypothesis but how will you choose your alternative hypothesis? If you knew that under all circumstances in which the null hypothesis was true you would know which alternative was false you would already know more than the experiment was designed to find out. All that you can do is apply your experience to use statistics, which when employed in valid tests, reject the null hypothesis most often. Hence statistics are more primitive than alternative hypotheses and the latter cannot be made the justification of the former.

I think that this is an important criticism of Fisher’s but not entirely fair. The experience of any statistician rarely amounts to so much that this can be made the (sure) basis for the choice of test. I think that (s)he uses a mixture of experience and argument. I can give an example from my own practice. In carrying out meta-analyses of binary data I have theoretical grounds (I believe) for a prejudice against the risk difference scale and in favour of odds ratios. I think that this prejudice was originally analytic. To that extent I was being rather Neyman-Pearson. However some extensive empirical studies of large collections of meta-analyses have shown that there is less heterogeneity on the odds ratio scale compared to the risk-difference scale. To that extent my preference is Fisherian. However, there are some circumstances (for example where it was reasonably believed that only a small proportion of patients would respond) under which I could be persuaded that the odds ratio was not a good scale. This strikes me as veering towards the N-P.

Nevertheless, I have a lot of sympathy with Fisher’s criticism. It seems to me that what the practicing scientist wants to know is what is a good test in practice rather than what would be a good test in theory if this or that could be believed about the world.

References: 

J. H. Bennett (1990) Statistical Inference and Analysis Selected Correspondence of R.A. Fisher, Oxford: Oxford University Press.

L. J. Savage (1976) On rereading R A Fisher. The Annals of Statistics, 441-500.

Related articles

• JERZY NEYMAN: Note on an Article by Sir Ronald Fisher (errorstatistics.com)
• E.S. PEARSON: Statistical Concepts in Their Relation to Reality (errorstatistics.com)
• Fisher, Statistical Methods and Scientific Inference (errorstatistics.com)

Please comment on the article here: Error Statistics Philosophy » Statistics

The post Stephen Senn: Fisher’s Alternative to the Alternative appeared first on All About Statistics.

Yesterday‘s competition is a toughie. If it were up to me, I think I’d have to go with Jesus. Here’s why: I’d come into the seminar with lots of resistance, like, c’mon, Jesus, I totally don’t believe the hype. As Hernan put it in comments:

Jesus spoke in parables to avoid committing to a specific point, like he was saying things without really saying them. In other words, he was deliberately ambiguous. If he were a modern day researcher, he would propose all this ambiguous theories, that can always come true with a bit of data and multiple comparisons. Jesus would publish in Psych Science.

So I’d resist. But, then, Jesus being who he is, I expect he’d charm the hell out of me. And, by the end, I’d probably be worshipping the guy. It sounds like a truly transformative experience that nothing secular could match. I mean, Jesus is the ultimate nice guy, right? But with a backbone of steel. Muscular Christianity and all that.

And his material is good, too. As Tom wrote in comments:

Jesus has by far the best encore of all time, so we know he must be a bit of a showman. Also – five loaves and two fishes will take care of the catering for the seminar. Dinner and a show – what more do you want?

All that and parables too. Take that, Larry David.

Buuuuuut—and this is a bit “but”—check out this comment from Dalton:

I don’t know. People keep trying to schedule a seminar with the dude, and he never shows:

http://en.wikipedia.org/wiki/List_of_dates_predicted_for_apocalyptic_events

That’s a big problem. It would be pretty embarrassing for us to announce this big seminar and then have the speaker not show up.

Who does he think he is? Lauryn Hill???

So, sorry, it’ll be Tolstoy who advances this time. Thanks, as always, for your thoughtful and humorous exchanges in the comments. This is a lot more fun than working, right?

Next, we have a truly titanic struggle. In one corner we have Mohandas Gandhi, seeded as the #1 Religious Leader of all time. (Remember, we have Founders of Religion in a different category.) And he’s up against Stanley Kubrick, unseeded in the Cult Figures category and also with a strong anti-war message. I can only imagine that Kubrick would have a much more entertaining series of stories. But Gandhi . . . what can you say? Talk about an impressive person. Maybe just being in the same room with the guy would be enough. What do you think?

P.S. As always, here’s the background, and here are the rules.

The post Mohandas Gandhi (1) vs. Stanley Kubrick appeared first on Statistical Modeling, Causal Inference, and Social Science.

Please comment on the article here: Statistical Modeling, Causal Inference, and Social Science

The post Mohandas Gandhi (1) vs. Stanley Kubrick appeared first on All About Statistics.

Andrew Dolman writes:

Just in case you need another example of why it is important to consider what the intercepts in a model represent, here is a short comment I [Dolman] just got published correcting a misinterpretation of a simple linear model, that would not have happened if they had centered their predictor around a sensible value.

The actual topic here is “Macrophyte effects on algal turbidity in subtropical versus temperate lakes.” That is so cool! I google-imaged “algal turbidity” and found the above pretty picture, which is on a webpage with a full discussion of turbidity. Surprisingly interesting stuff.

Back to Dolman and interpreting the intercept: Here’s his graph:

You can perhaps already see what happened. The other researchers noticed that the two lines had essentially the same intercept, and they concluded that the two groups differed only in the slopes, not the level of the lines. But this was a mistake in the context of the data. It’s similar to the example that Jennifer and I give in our book with the regressions of earnings on height. The intercept of such a regression doesn’t have much meaning, as it corresponds to the earnings of someone with a height of zero inches.

The post Another example of why centering predictors can be good idea appeared first on Statistical Modeling, Causal Inference, and Social Science.

Please comment on the article here: Statistical Modeling, Causal Inference, and Social Science

The post Another example of why centering predictors can be good idea appeared first on All About Statistics.

When playing with Peter Rossi’s bayesm R package during a visit of Jean-Michel Marin to Paris, last week, we came up with the above Gibbs outcome. The setting is a Gaussian mixture model with three components in dimension 5 and the prior distributions are standard conjugate. In this case, with 500 observations and 5000 Gibbs iterations, the Markov chain (for one component of one mean of the mixture) has two highly distinct regimes: one that revolves around the true value of the parameter, 2.5, and one that explores a much broader area (which is associated with a much smaller value of the component weight). What we found amazing is the Gibbs ability to entertain both regimes, simultaneously.

Please comment on the article here: Xi'an's Og » R

The post amazing Gibbs sampler appeared first on All About Statistics.