Does built in spin makes MS trial polls useless?
Sponsored or not, the questions too often dictate the answer
We have a tendency to be harsh on pollsters, and usually with good reason. There have been several polls asking the American public what they think about the DoJ's breakup proposal for Microsoft, but the point isn't so much whether or not particular polls have been Microsoft-funded: it's their general uselessness. Frequently the questions are loaded, and the opinions produced could therefore validly follow from the questions. Take the recent Harris poll questions, for example. Had Microsoft "treated competitors fairly"? Randomly selected respondents could not possibly give a meaningful response to that if they had no experience of Apple or Linux - all they can do is defend what they imagine to be their "choice" of operating system. As for whether Microsoft was "monopolist", that's a legal question that requires a knowledge of what "monopolist" means in a legal sense. After all, being a monopolist is not a crime - it's leveraging the monopoly that is illegal. The real purpose of polls is very often to seek support for a political stance by the commissioner of the poll, and it's for this reason that the questions are too often really bent. An outfit called the Portrait of America commissioned Rasmussen Research to find out about what the court of public opinion thought about the US versus Microsoft case, after Judge Jackson's determination that Microsoft had acted illegally. But was it really objective to ask whether respondents agreed with Bill Gates' statement that "This ruling turns on its head the reality that consumers know - that [Microsoft] software has helped make PCs affordable to millions"? Register readers will not be surprised that 59 per cent agreed, and just 11 per cent disagreed. Now how many of these respondents knew that the only increased costs in real terms on PCs has been the cost of Windows? Microsoft's breakup was seen as bad by 59 per cent and good by 19 per cent. So far as the outcome is concerned, 36 per cent say it would be good for America if Microsoft won, with 27 per cent thinking it bad. It is cases like this that destroy what little merit there is in public polls. ®