By Joshua E. Keating
WASHINGTON As much as we might like to think we're capable of objectively evaluating the political candidates we vote for, factors like income, ethnicity and gender strongly affect political preferences. And now a wave of psychological research into voting behavior is bringing to light other, much less obvious predilections.
The gun show: A 2013 study in the journal Psychological Science found that men with large biceps are more economically selfish than their scrawnier counterparts. In other words, rich dudes with big guns opposed wealth redistribution policies more strongly than rich but less ripped bros.
Among poorer guys, the men with bigger muscles were more likely to support redistribution. The authors theorize that evolutionary factors are at work: Men with "greater fighting ability will more actively attempt to acquire or defend resources."
Because of Dixie: The Confederate flag provokes strong opinions in the United States, whether as a symbol of Southern pride or a white supremacist relic. But can just looking at the Stars and Bars subconsciously make someone more racist?
In 2010, researchers found that study participants who were shown a Confederate flag reported less willingness to vote for Barack Obama than those who were shown a neutral symbol. The same effect held true for a "hypothetical" black candidate.
Holy pollers: Some atheist groups and civil libertarians in the United States argue that putting polling stations in houses of worship violates the First Amendment's separation of church and state. But a 2012 study suggests that the location of voting booths might actually affect how you vote.
A survey in the Netherlands and Britain found that people walking near churches expressed more conservative views and had more negative feelings toward non-Christians than those polled strolling outside government buildings.
Rat race: During the 2000 U.S. election, Democrats accused George W. Bush's campaign of embedding a subliminal message in an attack ad the word "rats" flashed across the screen during a description of opponent Al Gore. (The Bush campaign denied it was intentional.)
Does that kind of thing even work? Actually, in a 2008 study, psychologists found that flashing the word "rats" did increase participants' negative feelings toward a hypothetical candidate. During Democratic California Gov. Gray Davis's 2003 recall referendum, the researchers also observed a drop in opinions of Davis among Republicans who were shown a subliminal photo of Bill Clinton.
Political cycle: The weirdest of all voting-behavior studies might be the 2013 Psychological Science paper that examined whether menstrual cycles affect female voting preferences. It found that ovulation made single women more liberal and women in committed relationships more conservative.
The study became a brief sensation online and was criticized for its methodology - it relied heavily on survey data and didn't control for factors like age and income - but the journal stood by it.
Joshua E. Keating is a staff writer at Slate.