New Poll Reveals Serious Liabilities for Hillary

[Jackpine Radical]

45. Thanks.

In re-examining what happened here, I see that I was the victim of my own confirmatory bias. That is, the poll presented information that supported my position, and that blinded me to some rather obvious methodological problems. I really try to fight that in myself, but sometimes I fail in some degree or another.

There is an underlying reality, as confirmed in actual electoral referendum results, that the public is much more in favor of individual progressive proposals (min. wage, pot, etc.) than their political spectrum self-labeling would indicate. I think that knowledge set me up to be less critical of the poll than I would like to think is my custom.


We do need to operate on the basis of good information, and I don't at all oppose conducting polls like this--sort of political marketing research. However, the results of these exercises do not reflect actual voting behavior, have no predictive power, and should not be presented to the public as reliable information in that respect. For any one of those questions, there is probably a way to frame it so as to get the opposite response from the respondent. The issue being tested is how to frame the issue to get a desired response, and is not even an appropriate methodology for eliciting quantitative opinion data.