Interpreting | 2009.06.22
Saved by the leptokurtosis | David McKie
The language of statistics is a puzzle. But it could help us cut back on the green shoot overgrowth
It came as a shock to stumble in the Guardian the other day across the word leptokurtic, a term I had never knowingly glimpsed before. True, it comes from the world of statistics, from where Larry Elliott gleefully disinterred it, and that is an area where non-subscribers must be constantly on the alert for the shock of the new.
It reminded me of the first time I heard the word heteroscedasticity, at a conference in Nuffield College, Oxford . I turned to the academic beside me and asked: "What does \'heteroscedasticity\' mean?" He unflinchingly replied: "I feared you might ask me that." In fact it meant, I think he explained, a group of indicators widely scattered around the centre, as opposed to homoscedasticity, where they are more closely clustered together. This was comforting, since half the concept was already so familiar. We talk hourly of heterosexuality and homosexuality, and hetero- and homogeneity. One had only to fathom scedasticity, and the worst would be over.
In fact, as I\'ve now discovered, leptokurtosis is a term quite widely employed. The world leptokurtic commands some 40,000 references on Google – against a mere 1,900 for its cousin platykurtic and a paltry 547 for mesokurtic. Collins conveniently describes it as indicating a distribution having kurtosis B 2 greater than 3, more heavily concentrated about the mean than a normal distribution.
Split the word into two and leptokurtosis falls obligingly into place. Kurtosis, which one might have assumed was some kind of complaint worth discussing with your chiropractor, comes from a Greek word indicating an arch or a bulge. Lepto- indicates that the arch or bulge is slender; platy-, that it is a flat, like a platypus or platyhelminth; and meso-, that it is middling. Thus a study of the state of the Australian economy by Lombardo, Flaherty and Fang , drawing on previous work by Ibbotson and Steiger and by Elton and Gruber, analyses 16 cases of statistically significant non-mesokurtosis, and finds that in only four cases was such leptokurtosis unaccompanied by significant skew.
Yet here we encounter a significant link between their recondite world and our own. The word skew, though deeply important to them, has never belonged solely to statisticians. We talk all the time of things being skewed. We may travel, while feeling a shade skewwhiff, across a skew bridge, perhaps on a skewbald pony. When statisticians say skew, they are almost talking our language; when they say kurtosis they probably lose us. Still, the underlying concept seems simple enough. To talk of a slender bulge would seem unexceptional in the context of fashion or architecture. To adopt such a formula, translated into simpler English by a joint committee of statisticians and etymologists, might even begin to deal with one of the most intractable problems of recent economic discussion: how to eliminate the merciless present-day harping (2,830,000 references on Google, and never more profuse than in the last week) on the concept known as green shoots .
The father of green shootism is usually taken to be Norman Lamont , who as chancellor after Black Wednesday claimed to have detected the first green shoots of recovery. I doubt that he truly invented it. I guess if you trawled long enough you would find it in a subordinate clause in the works of RH Tawney or JM Keynes . Lamont may have been subconsciously influenced by the opening lines of Chaucer\'s Canterbury Tales (as translated by Nevill Coghill), where he talks of April days when the sweet showers fall, piercing the drought of March to the root, and all the veins are bathed in liquor of such power as brings about the engendering of the flower; and when Zephyrus joins in by exhaling with his sweet breath in every grove and heath upon the tender shoots, and so on; though in this case these circumstances are said to engender a taste for going on pilgrimage rather than stampeding back into the high street.
And anyway green shoots are surely not always a happy sign. They may wither; they may be the beginnings of weeds. What we may be seeing now in the economy might be better described as a slender bulge, which, like some other kinds of slender bulge, may hold out some promise, though without any guarantee, of some kind of fresh birth. Could "green shoots" have reached their tipping point? Generations to come may, I suspect, routinely refer to the late spring of 2009 as a leptokurtic moment.
For more information, please visit
http://www.guardian.co.uk/commentisfree/[...]tatistics-economy-green-shoots
Instagram is adding automatic in-app text translation - The Verge views: 1530
Here’s a ‘Brand-New’ Massive Multilingual Dataset for Machine Translation views: 1529