Public opinion polls are big business in election seasons. Political campaigns hire professional pollsters to help craft the candidate’s language and image to appeal to a particular demographic group. Major media outlets conduct and report polls to attract audiences who want to know what people like them think. When results of a poll are repeated by anchors and pundits and echoed through social media channels, they begin to take on the air of authority and may have a band wagon effect on undecided voters.
A public opinion poll indicates how a percentage of individuals in a random sample of a target demographic group, e.g. likely voters, responded to a set of carefully worded questions at a particular moment in time. Polls taken this month may or may not be a reliable predictors of an election results.
To judge the reliability of a poll, the Pew Research Center advises asking these questions: who conducted and who sponsored the poll; who was surveyed; the size of the sample; how the poll was conducted: live telephone interviews or an opt-in on-line survey; what was asked; when was the poll conducted, and the margin of error. Learn more from Pew.
The margin of error or measure of uncertainty is often given at + or - 3 or 4% but, according to Pew, it is more likely 6% due to a number of unmeasurable errors such the wording and order of questions asked, respondents attitudes toward polls and what they were doing when the phone rang. Best advice—be wary. National polls that did not take the Electoral College into account were wrong in 2016.