The Forum > Article Comments > Deconstructing Newspoll > Comments
Deconstructing Newspoll : Comments
By Bob Ellis, published 27/3/2009Political polling: there are many ways a poll if manipulated can give you a false, convenient result.
- Pages:
-
- Page 1
-
- All
- Pages:
-
- Page 1
-
- All
"Every pollster, including us, was saying that Labor was in trouble in the Queensland election. The predicted swing was in the realm of 7 percent. From what you can tell, internal ALP research also seems to have supported this."
Are we to take it then that ALL pollsters are using, consciously or unconsciously, the skewing techniques imputed to Newspoll? That seems a little rich.
The nasty little sleeper in the article, in my view, is this:
"We live in a world where numbers are manipulated (by Enron, by Lehman Brothers, by Bernard Madoff, by Robert Mugabe) so commonly, so routinely, that economic meltdowns result from their manipulation."
I get the impression that the damage to most pollsters' credibility arising from the getting of most of the forecasts wrong is not viewed by Bob Ellis as enough 'punishment'. They have to be collectively smeared by a manufactured association with the notorieties mentioned. Do I run on too far in suggesting that the next item on the agenda may be the touting of the banning of opinion polls because electoral meltdowns are seen as arising from their manipulation?
Alternatively, one could rework the nasty little sleeper like this:
'We live in a little world where elections are manipulated so commonly, so routinely, that almost all the opinion pollsters are getting it wrong almost all of the time.'
A standing contradiction like that would have to be somehow dealt with, lest at some point credulity amongst the public at large became just too strained. Nothing like raising the acknowledged boogeyman of Newspoll and its apparent lack of transparency to get the ball rolling. If it was recognised as a standing contradiction, that is.
As a matter of interest, is the apparent voter turn-out for the Queensland State elections able to be determined from any publicly accessible official report of the real poll figures, the aggregate number of vote claims made? I haven't found this on the ECQ site.