Statistical Theater of the Absurd: “Stat on a Hot Tin Roof”? (Rejected Post Feb 20)

Dear Reader: Not having been at this very long, I don’t know if it’s common for bloggers to collect a pile of rejected posts that one thinks better of before posting. Well, here’s one that belongs up in a “rejected post” page (and will be tucked away soon enough), but since we have so recently posted the Fisher-Neyman-Pearson “triad”, the blog-elders of Elba have twisted my elbow (repeatedly) to share this post, from back in the fall of 2011, London. Sincerely, D. Mayo

Egon Pearson on a Gate (by D. Mayo)

Did you ever consider how some of the colorful exchanges among better-known names in statistical foundations could be the basis for high literary drama in the form of one-act plays (even if appreciated by only 3-7 people in the world)? (Think of the expressionist exchange between Bohr and Heisenberg in Michael Frayn’s play Copenhagen, except here there would be no attempt at all to popularize—only published quotes and closely remembered conversations would be included, with no attempt to create a “story line”.)  Somehow I didn’t think so. But rereading some of Savage’s high-flown praise of Birnbaum’s “breakthrough” argument (for the Likelihood Principle) today, I was swept into a “(statistical) theater of the absurd” mindset.

The first one came to me in autumn 2008 while I was giving a series of seminars on philosophy of statistics at the LSE. Modeled on a disappointing (to me) performance of The Woman in Black, “A Funny Thing Happened at the [1959] Savage Forum” relates Savage’s horror at George Barnard’s announcement of having rejected the Likelihood Principle!

The current piece taking shape (in my mind) also features George Barnard and recalls our first meeting in London in 1986. I’d sent him a draft of my paper “Why Pearson Rejected the Neyman-Pearson Theory of Statistics” (later adapted as chapter 11 of EGEK) to see whether I’d gotten Pearson right. He’d traveled quite a ways, from Colchester, I think. It was June and hot, and we were up on some kind of a semi-enclosed rooftop. Barnard was sitting across from me, very erectly, looking rather bemused.

The curtain opens with Barnard and Mayo on the roof, lit by a spot mid-stage. He’s drinking (hot) tea; she, a Diet Coke. The dialogue (is what I recall from the time[i]):

 Barnard: I read your paper. I think it is quite good.  Did you know that it was I who told Fisher that Neyman-Pearson statistics had turned his significance tests into little more than acceptance procedures?

Mayo:  Thank you so much for reading my paper.  I recall a reference to you in Pearson’s response to Fisher, but I didn’t know the full extent.

Barnard: I was the one who told Fisher that Neyman was largely to blame. He shouldn’t be too hard on Egon.  His statistical philosophy, you are aware, was different from Neyman’s.

Mayo:  That’s interesting.  I did quote Pearson, at the end of his response to Fisher, as saying that inductive behavior was “Neyman’s field, not mine”.  I didn’t know your role in his laying the blame on Neyman!

Fade to black. The lights go up on Fisher, stage left, flashing back some 30 years earlier . . . ….

Fisher: Now, acceptance procedures are of great importance in the modern world.  When a large concern like the Royal Navy receives material from an engineering firm it is, I suppose, subjected to sufficiently careful inspection and testing to reduce the frequency of the acceptance of faulty or defective consignments. . . . I am casting no contempt on acceptance procedures, and I am thankful, whenever I travel by air, that the high level of precision and reliability required can really be achieved by such means.  But the logical differences between such an operation and the work of scientific discovery by physical or biological experimentation seem to me so wide that the analogy between them is not helpful . . . . [Advocates of behavioristic statistics are like]

Russians [who] are made familiar with the ideal that research in pure science can and should be geared to technological performance, in the comprehensive organized effort of a five-year plan for the nation. . . .

In the U.S. also the great importance of organized technology has I think made it easy to confuse the process appropriate for drawing correct conclusions, with those aimed rather at, let us say, speeding production, or saving money. (Fisher 1955, 69-70)

Fade to black.  The lights go up on Egon Pearson stage right (who looks like he does in my sketch here, from EGEK 1996, a bit like a young C. S. Peirce):

Pearson: There was no sudden descent upon British soil of Russian ideas regarding the function of science in relation to technology and to five-year plans. . . . Indeed, to dispel the picture of the Russian technological bogey, I might recall how certain early ideas came into my head as I sat on a gate overlooking an experimental blackcurrant plot . . . . To the best of my ability I was searching for a way of expressing in mathematical terms what appeared to me to be the requirements of the scientist in applying statistical tests to his data.  (Pearson 1955, 204)

Fade to black. The spotlight returns to Barnard and Mayo, but brighter. It looks as if it’s gotten hotter.  Barnard wipes his brow with a white handkerchief.  Mayo drinks her Diet Coke.

Barnard (ever so slightly angry): You have made one blunder in your paper. Fisher would never have made that remark about Russia.

There is a tense silence.

Mayo: But—it was a quote.

End of Act 1.

Given this was pre-internet, we couldn’t go to the source then and there, so we agreed to search for the paper in the library. Well, you get the idea. Maybe I could call the piece “Stat on a Hot Tin Roof.”

If you go see it, don’t say I didn’t warn you.


[i] We had also discussed this many years later, in 1999.

Categories: Statistics | Tags: , , , , | 13 Comments

Post navigation

13 thoughts on “Statistical Theater of the Absurd: “Stat on a Hot Tin Roof”? (Rejected Post Feb 20)

  1. I have two questions, only the first of which undoubtedly has an answer, even if I don’t know it.
    Why did Barnard find Fisher’s “Russian” statement so out of character as to doubt Fisher could have said it? (He more than doubted it, he thought it impossible, until of course, we looked it up.)

    Why did Barnard think I’d invent a quote like that?

    • Corey

      Answer to the first question: the halo effect. Answer to the second question: a valid deduction from false premises.

      • Corey: Explain how the halo effect enters, halo around whom? And where does Russia enter? .
        Also, false premises need not need to a valid deduction, but never mind that.

  2. Corey

    I hypothesize that Barnard was subject to the halo effect regarding Fisher — Barnard’s respect for Fisher;s statistical acumen skewed his judgment as to whether Fisher would have written that curious passage you quoted.

    Jaynes suggests that Russia — and the U.S. — enter as a swipe at Neyman, who (although of Polish extraction) was born in Imperial Russia and emigrated to America.

    It was valid (but not sound) for Barnard to conclude that you had blundered somehow given his premise that Fisher wouldn’t have made the remark about Russia.

  3. I think Corey’s hypothesis is interesting. Barnard was a communist in his youth and for years could not get into the USA. Neyman was dismissed from the UN observer group in Greece after the war for fraternising with the left and (I have been told – hearsay evidence) dismayed Polish statisticians on a visit to Poland by informing them they were living in a paradise. On the positive side Neyman appointed David Blackwell at a time when racial prejudice made this a controversial move. Barnard, whom I corresponded with a few times (I treasure the letters) was a great admirer of Fisher’s approach to inference. So my take is that he had sympathy for Neyman’s politics but was an enthusiast for Fisher’s approach to inference.

    Still, Barnard’s error is curious.

    As regards Fisher’s dislike of communism I recommend the reading of CP 229 ‘What sort of Man is Lysenko’.

  4. I’m pondering the remarks from the two of you, but not really getting clearer: Neyman was Russian, and Fisher is criticizing Neyman so why would Barnard doubt that Fisher could cast aspersions on Russian-style 5-year plans? Not grasping the relevance of hiring Blackwell either, but I think there’s good evidence that Neyman was supportive of women, and I’ve always been intrigued by his acknowledgments of Janina Hossiason (may have the spelling wrong)–will explain why another time. I too have letters from Barnard, and a few other funny stories I may tell some time…

    But, back to Barnard, what shocked me at the time, was that here he was complimenting me on my paper and taking me seriously (I was like 28) and then to come out with the suggestion that I was inventing (or erroneously stating) a quote? Wouldn’t he have checked if he had doubts, and why bother meeting me if I was inclined to invent lines to attribute to R.A. Fisher? I mean it doesn’t matter, we looked it up, and we had several subsequent conversations over the years and he was always supportive. I wish I had talked with him more.

    • Stephen Senn

      I don’t think that Fisher would have regarded Neyman as Russian. I am sure that he would have thought of him as Polish. (After all, we think of Michael Collins as being Irish and not British.) However Fisher may have believed (correctly I think) that Neyman had some Soviet sympathies.

      Despite being an admirer of Fisher the scientist I agree with Corey, however that this is not Fisher’s finest moment and I think that Barnard accusing Deborah of making things up is not his finest moment either.

      The reference to Blackwell has no relevance at all except that I was implicitly suggesting that Neyman was blind to the crimes of Stalin and Stalinism and felt it appropriate to mention that there were, however, many aspects of his personal intervention in ‘politics’ (in the widest sense)that were positive. Deborah has mentioned another and that he was supportive of women is something I have heard also.

  5. Corey

    Fisher is basically stating that the mindset of Russians and Americans is so focused on engineering that they’ve forgotten or never learned how to do “the process appropriate for drawing correct conclusions,” that is, science, properly — perhaps even implying that it should therefore be left to Englishmen. In any event, it’s a huge and unsupported overgeneralization, and I’m inclined to think that Barnard considered that kind of argument beneath Fisher’s dignity.

  6. Thanks so much to Stephen and Corey, it’s rare that others are interested in reflecting on such idiosyncratic features of personality in relation to these statisticians. Of course it’s clar enough what Fisher’s point is, but if it’s wrong-headed to allude to Russian 5-year plans, why was there no objection to “the U.S. also”.

    I’m a little surprised, though, at the suggestion (I think by both of you) that Fisher was being undignified or extreme, considering that this is the line we hear quite often in criticism of the NP (behavioristic) approach. Fisher, by contrast, is depicted as the “evidential” one.

    • Stephen Senn

      I think it is one thing to claim that the methods of science ought to be fundamentally different to those of manufacturing ( a matter of interesting debate in my opinion) and then adding a claim that persons who may mistakenly think so have been misled by the nature of the societies in which they live. I have no problem with the first half of this; it is the second I think is ‘unhelpful’. The reason I picked up on the Russian rather than the US connection is that, ( I quote Deborah Mayo quoting George Barnard), “You have made one blunder in your paper. Fisher would never have made that remark about Russia.”

      In other words it was Barnard who particularly picked up on the Russian connection and it is Barnard’s reaction that was being commented on. I have speculated above as to why Barnard objected to this in particular,
      Stephen

  7. Stephen: Thanks for this. But surely the nature of one’s society and the kind of knowledge that is of value do influence the conception of the relevant goals for inquiry. Anyway, putting aside the Barnard comment, what is of most relevance is the continuation of Pearson’s view of the role of hypothetical sampling distributions to causal knowledge in the case at hand. In the paper that Barnard had read, I traced out Egon’s philosophy, so far as I could make it out, from this and other papers, and Barnard agreed with it! This is something that is still not adequately understood—but I’ll keep trying to illuminate it (as I increasingly get clearer on it).
    By the way, Pearson had long before written a book on acceptance sampling (2 actually, including the one whose copies were all burned)*—something I didn’t know until Aris Spanos brought to light the one that still exists.

    *Noted in the Cox-Mayo conversation in the RMM volume.

  8. Stephen Senn

    Deborah: Thanks I agree with your point about societies but I still don’t think that makes Fisher’s reference to the USA and Russia helpful. Once an argument is made the argument is on the table so if Neyman is wrong to impose acceptance sampling methods on scientific inference (which is, as I said, an interesting subject for debate) he is wrong whether this is because of or despite any Russian or American influence. So I think that Fisher raises an interesting argument but dragging Russia and USA into it is not so good and in any case Pearson was British.

    Having said that, I am being rather hypocritical since I have an addiction to rhetoric.

    As regards sampling I think that Fisher and Pearson disagree on how the hypothetical samples are to be generated. Fisher ascribed (rightly or wrongly) to Neyman and Pearson the view that it was the probabilistic sampling mechanism generating the actual samples that also generated the inferential population of samples used for analysis. Fisher, however, stressed the necessity of considering relevant subsets.

    This appears to be at variance with his stress on randomisation, but I think it can be justified it one takes the view that the intended analysis should guide the randomisation rather than that the actual randomisation should impose the analysis.

    See Fisher’s letter to Yates 2 November 1955 in Henry Bennett’s edited correspondence of Fisher.

  9. I recall being unable to track down Bennett’s book in the past, I don’t know if it is now available.
    It seems to me, by the way, that all three of the “triad” papers are rather combative, perhaps surprisingly so (compared to current trends)—doesn’t it? Maybe I expect it from Neyman and Fisher, but Pearson’s opening remarks seem a bit belligerent for him, especially as he’s referring to his “heresy” (of introducing the alternative). But I’m not clear on what Pearson is denying wrt Barnard’s penetrating observation….but all of this is really just of human interest, I think.

I welcome constructive comments for 14-21 days

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Blog at WordPress.com. The Adventure Journal Theme.

Follow

Get every new post delivered to your Inbox.

Join 410 other followers