There are all kinds of polls quoted in both the print and broadcast media these days. To my knowledge, very few of them are intentionally designed to arrive at a particular answer. The polls sponsored by political candidates or parties might be good examples of some that might be a bit self-serving. I don't trust any of them at all.

There would be little purpose, it seems to me, for well-known news organizations--in this case the New York Times and CBS--to purposely design poll questions and a sampling technique to arrive at answers they desire. Maybe I'm too trusting, but why would they do that? There are all kinds of professional pollsters, statisticians and academics that would be all over such poll sponsors or those they hire to design and conduct the polls with criticism of their apparent intent and the inaccuracies that would result from a poorly-designed survey or sample.

I haven't read any such criticisms of any of the myriad of polls presented and discussed on various media outlets. In fact, for the most part the results of the polls are all reasonable comparable--within the standard error they assign to their results, of course.

What's tough, of course, is when a poll taker is trying to determine who might win an election, as an example, and various polls show results around 50%. The standard error alone can make polls of close election races difficult to interpret or believe. But in this case--the NYT/CBS poll on healthcare of a few days ago--the poll answers to many of the questions weren't anywhere near 50%. Those polled seemed to be presenting consistent answers and ones which probably accurately reflect the feelings of the group being measured by the poll sample.

Of course, another variable of polling is the timing of the poll. An excellent example of the effect of time and new information might be the difference in poll results of the feelings of South Carolina residents regarding their Governor. I'm quite certain that a poll taken two weeks ago would have substantially different results than one taken today. That can also happen as the details of the proposed healthcare plan become known. I wouldn't be surprised at all if the "numbers" move around depending on the details of the proposed legislation.