This time of year, new campaign polls seem to up every 15 minutes. Some basic guidelines can help readers navigate the resulting treacherous terrain.
Watch for trends: New national polls released Monday show Republican Mitt Romney making clear gains in the race for the White House following a strong debate performance last week. But the results are less uniform when it comes to Romney’s existing deficits on basic measures on likability and understanding the economic problems of average people. Look to see whether Romney can close these key gaps, and which candidate, if any, can convince voters that they are the one to lead the economic recovery. These are the critical underpinnings of the campaign, far more important to understand than whether voters right now prefer one candidate over the other by one or two percentage points.
Don’t obssess over the “margin”: Speaking of the horse race, avoid too much focus on the percentage point difference between candidates. More often than not the difference between being up five or eight points among a group of voters — independents for example — is basically meaningless. The margin of sampling error accompanying a poll applies to each candidate’s support, not the gap between them. If a survey has a three-point error margin (and note that error margins are widely underestimated), and a candidate goes from 52 percent in a poll one week to 50 in another week, that’s not a clear, statistically significant shift.
Be mindful of shortcuts: Popular polling averages can provide a useful quick glance at the status of a contest -- summarizing dozens or more polls in an apparently precise set of numbers. But important differences between polls are often obscured in the process, including methodology (e.g live telephone interviewers, automated calls, Internet polls), samples (e.g. likely vs. registered voters) and time-frames.
Track the same polls over time: Is there consistent, significant movement from one Washington Post poll to another? From one survey from the Pew Research Center to another? These are the most telling shifts, particularly when they run in a similar direction. Both Pew and Gallup shifted toward Romney in post-debate polls.
Blaming the polls is almost always a red herring -- struggling campaigns often take aim at the polls, revealing their candidates’ weaknesses. Even as partisans are publicly decrying properly conducted surveys, they are surely taking them seriously internally. To do otherwise would be to ignore the very voters who will be decisive.