What's Behind Divergent Marriage Amendment Polls?

Two recent polls, conducted within days of each other, offered widely divergent views on whether Minnesota voters will approve a hotly contested constitutional amendment this fall.

In November, voters will be asked whether the state constitution should ban same-sex marriage, to reflect current law.

One poll, conducted by SurveyUSA, showed the amendment winning with 50 percent of the vote to 43 percent opposing. Another, conducted by Public Policy Polling, puts the battle as a nearly even heat at 48 percent for and 47 percent against. 

But who's right? MinnPost's David Brauer lays out  what's known about each poll's inner workings in a recent column. 

His conclusion, while he doesn't say as much, appears to be "who knows?" Neither poll's internal workings give away an obvious hint of bias. For example, the poll suggesting more support for LGBT Minnesotans surveyed more older voters, who generally support same-sex marriage bans. The poll suggesting more support for the amendment surveyed more Democratic voters. 

Brauer does offer this consolation to the confused:

It's also important to note that normal statistical variation—the act of placing your statistical dipper into the vast ocean of voters—can produce differences.