With less than a week before showtime, the polls are already jam packed. But the real stars of the eleventh hour are the undecided voters, who still (incredibly, miraculously) haven’t made up their minds. And apparently they’re more numerous than one might think, with one in seven voters saying they might go either way, according to a recent poll.

But are these vacillating voters really still deciding? Or could their minds have already formulated a choice, without their even realizing it? University of Virginia psychologist Brian Nosek suspects the latter is true, and has partnered with colleagues at Harvard and the University of Washington to test whether humans form mental associations that differ from what their conscious recognizes.

To do this, the researchers used the Implicit Association Test, which has been up and running for a decade now and has logged around 7 million responses. Since the 2008 presidential race began, the team has tested more than 25,000 voters on their implicit preferences concerning the two candidates. Of that group, around 4,000 (15 percent) declared themselves undecided. But the test results for a significant number of these so-called undecideds showed a clear implicit preference for Obama or McCain. Here’s how the team summarized their results:

• Implicitly, Democrats are strongly pro-Obama, and Republicans are strongly pro-McCain, similar to their explicit preferences.
• Independents are implicitly pro-Obama, on average, similar to current polling results.
• The most intriguing subsample is the large number of undecideds who appear to be leaning toward McCain implicitly and toward Obama explicitly.

The last finding, of course, being the elephant in the room (or the polling booth).

Jost started off by discussing the “Big Five Model of Personality,” which, according to his results, offers clues about the minds of hardcore liberals versus their conservative counterparts.

He found that liberals ranked themselves consistently higher on the “creativity and openness to new experiences” scale, while conservatives did the same for “conscientiousness and order & discipline/rule following.” These traits also manifested themselves in the subjects’ “behavioral residue,” with liberals having more colorful and cluttered bedrooms and offices while conservatives’ spaces were tidier and more scrubbed.

Jost then discussed the body of research on the relationship between political views and perceived threats. In particular, data has shown that reminders of death—from disturbing images to words briefly flashed on a screen—caused all people, regardless of political beliefs, to increase support for conservative ideas across the topical spectrum, from same-sex marriage to foreign policy. The latest of these studies, for example, found that the greater a person’s sensitivity to disturbing images—a large spider on a face, a maggot-filled wound—the higher a chance he or she had of supporting a conservative agenda.

All of which presents interesting questions about the influence of neurobiology on ballot casting, and how a predisposition for sensitivity to threats can push you towards a party. But what about the current election? Given that the current administration has overseen such massive, recent turmoil—not to mention that, for the first time in recent history, the GOP is running with a VP candidate that many consider unsafe—could that upset the balance for threat-sensitive voters? Maybe, said Jost, who told DISCOVER in an e-mail:

I think it’s possible that threat and uncertainty could work against the GOP this election, but *only* if (1) voters decide that Republicans are primarily to blame for creating financial, foreign policy, and environmental crises through bad policies, and (2) the Democrats are able to convince voters that they have sound solutions that will definitely work to fix the problems caused by conservative policies.

So there you have it. The first part shouldn’t be all that hard. And the second—well, guess we’ll have to wait and see.