> From: Erik Reuter [mailto:[EMAIL PROTECTED]
>> I was wondering if the public may have been clever enough to have been >> a bit mischievous during the polling, by making it appear closer, they >> may well have squeezed another billion or two out of our formerly stingy PM, > > Before I were to consider this rather unlikely possibility, I > would first consider the possibility that the election was rigged. > So, some random poll, of a few thousand, probably completely disinterested voters, in the middle of cooking dinner or watching the news, has more credibility than the national elections? Geez, you may run some hokey elections over there, but it's a serious and professionally run business here. What is the incentive for people to be truthful in polls? Even if they did know who they were going to vote for, why should they tell the truth? The election, they have to live with the outcome, the polls, well, who gives a damn, lets scare the government a little. The people know the politicians read the newspapers. Polls aren't like rainfall stats or surveys of kids heights, the data is not reliable and not related to what you are trying to predict. All it tells you is what people said in that poll. From it one may be able to draw inferences about their probable voting intentions, and perhaps note trends, but nothing absolute, nothing with any certainty. The only poll with sufficient sample size and with the required degree of motivation towards voter honesty is the one on election day. I would first consider the possibility that the electorate is a lot smarter than you give them credit for, and that data from election polls is not fit for statistical analysis. Andrew _______________________________________________ http://www.mccmedia.com/mailman/listinfo/brin-l