Reading polls correctly is crucial, as it will misinterpret the trends in major elections and change the course of the decisions being made. It is about more than checking the headline numbers; it means digging into the data and context.
“One of the major lessons from big elections, like the 1994 Pataki-Cuomo race, the 2016 U.S. presidential election is how to look at the internals of a poll. Most polls report topline results, meaning overall support for a candidate,” John Zogby, founder of the nationwide online survey service and opinion polling analysis company John Zogby Strategies, said. “This might mask an important trend within subgroups that are counted in the topline results. Often, demographic breakdowns of the voters tell more about the dynamics of an election than do the overall percentages.”
Another lesson is to know that voter behavior is fluid. For example, the late-deciding voters in 2012 proved to be critical in re-electing Barack Obama; early polling missed that trend. Remember, polls represent a snapshot in time, not forecasts of what will happen in the future. It is important to understand opinions may change closer to Election Day.
It also is important to think about margin of error and statistical significance. Small differences in the results of polls the media often seize on may be less meaningful than they seem because they may fall within the margin of error.
Finally, remember bias can enter either in the wording of questions or in the sample selection itself. The good pollster asks unambiguous, unbiased questions and samples a diverse representative population. By focusing on these factors, readers are better placed to understand the polls and be insulated from surface-level results.
You may also want to read the following article: What Does Robert F. Kennedy, Jr., Actually Want?