Very interesting article, Paula. I was particularly intrigued by the fact that the quota sampling approach can’t always be accurate. I think the issue of pre-election polls lies not necessarily in the existing ‚margin of error‘. The bigger issue is that it sold and accepted at face-value and obviously affects people’s voting behaviour as has been proven. It therefore is questionable whether we should be allowed to publish a polling prognosis before an election. It also makes me realise how much the characteristics, social and professional background, etc. of my sample will affect my data.
Very interesting article, Paula. I was particularly intrigued by the fact that the quota sampling approach can’t always be accurate. I think the issue of pre-election polls lies not necessarily in the existing ‚margin of error‘. The bigger issue is that it sold and accepted at face-value and obviously affects people’s voting behaviour as has been proven. It therefore is questionable whether we should be allowed to publish a polling prognosis before an election. It also makes me realise how much the characteristics, social and professional background, etc. of my sample will affect my data.
ReplyDelete