Saturday, September 24, 2011

[special to FGAQ]

A stern rebuke to the inanely blathering Gallup Organization 

George Gallup, Jr

I have written before here on the pages of FGAQ [and yes, I am well aware of the fact that these are not actual 'pages', but you must forgive an old pollster his habits] about the pioneering work of the statistical analyst who bequeathed to me my name, and how his career was nearly ruined by his faulty prediction of a Dewey victory in the 1948 presidential election, an unforced error committed through the simple omission of the proper follow up question.
It is a lesson that I never forgot once I took over the helm of the mighty ship Gallup. Indeed, I was nearly legendary for my follow-up questions, and there is a simple reason why - at the time when I would be making my initial query, my mind would be whirling with the formulation of a follow-up. [As an example, the 1996 presidential elections were preceded by a number of polls showing Bob Dole trouncing Bill Clinton, a result that I found to be highly unlikely. Prior to our own poll, I instructed our interviewers to ask the following to anyone who selected Dole - "Oh, really?" Needless to say, our poll was on the money, and our glory continued to grow.]
Sadly, current Gallup CEO Jim Clifton - a man who is, alas, neither blood nor spiritual kin - has not always been as keen on this sort of follow-up and follow through. I have upbraided him on numerous occasions for his slovenly work ethic to little avail. [More accurately, I have seriously reproached his secretary, as Mister Clifton finds a way to avoid my calls]. The current Gallup Organization is always polling, polling, incessantly polling, with little regard to the importance or significance of the matter at hand. Furthermore, after completing one of these ludicrous samples, Gallup has now taken to formatting them into ludicrous attention-seeking opinion pieces, as though the facts were unable to speak for themselves.
No real newsman reads a Gallup poll for it's entertainment value or 'political slant', they read the topic, and they note the numbers. This is the reason that I nearly suffered apoplexy today when I read the latest piece of excretory brouhaha to bear the Gallup logo - 'Majority Rates Obama Same or Worse Compared With Bush'.
With trembling hands I began to delve into the raw data, scant as it was. My keen pollsters eye had immediately fixed on the word 'same', a loaded word if there ever was one. I shuddered to think of the many ways in which that short adjective could be abused, and I was right to shudder. To wit: 34% of respondents did indeed say that Obama was worse than Bush, while 22% said that he was 'the same', and I cannot deny that this equals 56%, thus making the headline technically true. What may not be immediately clear to the unskilled analyst, however, is the fact that 43% responded that Obama was better than Bush, while the coalition of the wishy-washey malcontents still provided 22% who voted 'same'. For those of you keeping score at home, that equals 65%, meaning that the only way to honestly report the results would have been 'Majority Rates Obama Same or Better Compared With Bush'. This is the way that facts become fictions, and I vigorously object.
But the shame does not end there. Indeed not. Imagine, if you will, if Clifton's clowns had reported 'Obama beats Bush 65% to 56%' and you will immediately come to the understanding that even in a mathematically challenged nation such as ours, such an outcome would be impossible, and that the integrity of the poll had been completely undermined by the reckless use of the word 'same'. [Memo to Jim Clifton: Back in my day we had a somewhat more complex word - 'undecided'. You would be well advised to look it up and learn it.]
Hoping for a silver lining, I proceeded to look at the party breakdown and was astounded by what I found - the vast majority of those who though Bush was worse were Democrats and the vast majority of those who thought Obama was lesser were Republicans. Stop the presses!
Although polling methodology is summarily included, in all honesty I do not think that any such survey was ever taken, and have decided to take my own meticulous survey to find out how many people believe that this was a product of Clifton's fevered imagination as opposed to those who feel it was a prank perpetrated by the college interns. But be advised, Mister Clifton, you have failed, and in doing so I shall look upon the current management of Gallup with ever greater disdain.

1 comment:

  1. This is incredible! I like the idea!!! So pleased I merely stumbled across your blog That is lovely! I stumbled upon your site only a few moments back as well as dropped crazy about this kind of! I have so much cloth and not considerably to do with this, this helps! Thanks!