Wednesday, 4 May 2016

Allan McCutcheon RIP

I'm saddened today to learn of the untimely death of my friend and colleague Allan McCutcheon. I first met Allan in 1992 when we were  teaching at the Essex Summer School in Social Science Data Analysis. Allan taught a course on latent-class analysis and I taught a course on log-linear models so we immediately had a common interest. 

We were part of a little group of instructors, including, among others, Neal Beck, Harold Clarke, Jörg Blasius, Peter Schmidt and Steffen Kühnel who, after the work was done, would chew the cud    over a bottle of wine before cramming into my tiny Peugeot and heading off into the Essex countryside in search of a good dinner.

Everyone warmed to Allan. He was a true gentleman, generous with his time (especially with students) and unfailingly enthusiastic about whatever he was doing or currently reading ( which was just about everything). He always wanted to learn something new and he was one of those people who just had a lust for life. It was inspiring  to see him again every year and hear about his latest hobby or interest: flying, long-distance swimming, cookery, fine wines, sky-diving, fast cars...

Professionally I learned a lot from him, especially about the dark art of estimating latent class models. Allan was the first person to show me Jeroen Vermunt's Lem program for categorical data analysis. That must have been back in 1993. He had already mastered all the tricks  and to me it was real eye opener. I retired my copy of GLIM and became a convert. 

We contributed a jointly authored chapter on categorical data analysis to a volume celebrating the 30th anniversary of the Essex Summer School and he was a joy to write with. I found  money from somewhere to invite him over to London, got him to give a seminar and we spent a bit of time sorting out data for our examples and probably more time sampling dining opportunities. A couple of weeks later I received the draft of his part of the piece and soon we were done.

Reading all the Facebook tributes pouring in today, many from former students as well as from friends and family, you can get a sense of  just how many lives Allan touched and how much of a difference he made. I don't know how Allan would like to be remembered, but this is how I will remember him, having fun and treating life as an awfully big adventure.

Thursday, 21 April 2016

Songs for Swinging Lovers (of Bourdieu)

It had to be U - the SVD song by the multi-talented Michael Greenacre.

Tuesday, 19 April 2016

Why do institutions insist on stupidity?

I don't know what the answer to my question is, but it seems an important one to ask.

I submit an article to a well known sociology journal. The journal has a policy of double blind anonymity. Fair enough, it's their ball and if I want to play I have to play by their rules. I make sure that there are no obvious clues in the text  as to who has written it. Within a couple of  hours I get the text sent back to me repeating their anonymity policy and telling me that I have not adhered to it and therefore my submission has been unsubmitted. Strange.

My sin is that I have included a reference to my own work.

But wait a minute. If I were to write: " I have written previously (Mouse and Mills, 2014)" then clearly my authorship is revealed and it is correct to disallow it. But if I write , as I did, something like "... recent contributions to this literature are Spatchcock and Bogbrush, (2010), Mouse and Mills (2014)..." then  nothing at all is given away. In fact if Mouse and Mills (2014) is a standard reference then deleting it and substituting "Author Reference" doesn't make it less likely a referee will identify me as the author, it will, if they have any knowledge of the field, make it more likely! So the policy in this case will achieve the exact opposite of what it is intended to do.  D'oh!

Not that I  really care one way or the other. The claim that authors are anonymous is just bullshit anyway. It's trivially easy for referees to find out who the authors of journal submissions are and if you an expert in an area you know more or less who is doing what without having to go anywhere near Google. You've heard the seminar, seen the conference presentation, read and cited the working paper long before the thing is published.  I have a paper which has been outright rejected by two journals which is already being cited and appears in class reading lists (not my own!). 

The only people we are kidding by pretending that refereeing is double-blind is ourselves. Nobody believes it, so why do journals insist on maintaining the fiction? Academics really are very good at making themselves look absurd.

Wednesday, 13 April 2016

RIP Arnold Wesker

So Farewell
Sir Arnold Wesker.

Yes. In your kitchen
it was a wedding feast.

Chicken soup with barley,
Roots. Honey.
Chips with everything.

But now
You've gone
To your own and golden city.
Words on the wind.
Say goodbye.

E. J. Porter (15½)

Friday, 18 March 2016

Friday music

The weekend's here; time for some musical genius: Michael Hedges' cover of She Drives me Crazy.

Monday, 14 March 2016

The Changeling

Robin Jenkins' The Changeling is one of the best novels I've read for a long time. To my shame I'd never heard of Robin Jenkins and I had no real idea what I would find inside the covers. It seems that north of the border Jenkins is still well known and  his novels even feature in school examination syllabuses. Down here he gets zero name recognition. That's a real shame because he is, in my estimation, much more than a minor regional novelist.

The book's West Coast of Scotland setting  appealed to me partly because it brought back childhood memories of being taken on the boat to Rothesay  (a vicious gull stealing a banana from my five-year old hand is indelibly etched on my brain) and of going from Glasgow on family outings to Helensburgh, Largs, Garelochhead and Loch Long. One of the the oddities of that part of the world is that you can easily take a day trip from some of the worst urban deprivation in Western Europe, to an area of outstanding natural beauty where incongruously you can  see the sinister black shapes of submarines floating on the deep (and extremely cold) water.

The easiest way to sum the novel up is that it is about geographical, social and moral liminality. Boundaries are transgressed, certainties are disrupted,  home truths are brought to the surface and left unresolved. A good but somewhat sanctimonious school teacher tries to do a good thing, possibly for bad reasons and simply doesn't understand the damage he is going to cause both to the object of his generosity and to those around him that he loves.

The bleakness of the tale is relieved by an element  of grim humour  in the episodes that describe the collision of different social worlds. The ending is perhaps a little melodramatic but by this point the tension that has been created by so many misunderstandings, changes of mind and partial insights into self-delusion is so great that some kind of catharsis, albeit a tragic one, is needed to bring the book to an end. Perhaps if he had been writing today he would just have left things hanging in the air. 

The moral of the story: the Good Samaritan had it easy.

Friday, 11 March 2016

P. Values and all that

I feel I should write something about the  ASA pronouncement on p-values, but reached the conclusion that to say anything that's in any way  useful and that won't be misconstrued by fools or knaves will be so time consuming that it will seriously interfere with doing stuff I really need to be getting on with.

It's an important issue, requires careful discussion and, perhaps most importantly, requires that some attention be paid to disciplinary context. It also requires an understanding that science doesn't take place in a knowledge vacuum. Usually no single experiment or piece of data analysis is decisive (at least in the fields I work in).

Already the usual cranks and buffoons are bellowing nonsense about it down their megaphones, but since a defining feature of crankiness is incorrigibility in the face of  reason it is a waste of time to include them in the conversation.

For what it is worth I  find  the ASA statement in itself  to be entirely reasonable. It doesn't condemn the use of p-values; it does point out that p-values are often used or interpreted inappropriately. Who could disagree with that? There are good  comments by Senn, Mayo and Benjamini among others. You can catch the latter on Mayo's blog

I even liked Gelman's 'garden of forking paths' routine though I fear where that would take us with regard to most of the quantitative empirical work in sociology.  I tend to think of most of the work I do in terms of estimation and data smoothing rather than hypothesis testing but whatever you are doing you do need to think about all of the data dependent decisions you have made and how those might have gone the other way if the data had been different.