Expectations matter most in how undecided voters decide. It's important to understand how late-deciders tend to vote. This knowledge (especially in Iowa and New Hampshire) could decide the outcome.
Primary voters tend to decide whom to for at the last minute.
In 2008, 16 percent of voters said they made their choice on election day. Which means we have a long way to go to know how Iowa will turn out.
I read a report (summarized here) that showed late-deciders tend to break for the candidate they expect to win. (Which may explain why most campaigns release a shock poll about their candidate's surge a day or two before the election.) There's also a study of how fish choose leaders that supports this theory via Science Daily:
"Their consensus arises through a simple rule," said David Sumpter of Uppsala University. "Some fish spot the best choice early on, although others may make a mistake and go the wrong way. The remaining fish assess how many have gone in particular directions. If the number going in one direction outweighs those going the other way, then the undecided fish follow in the direction of the majority."
But it's very difficult to know a particular voter's expectations, and those expectations vary widely by the voter's interest in politics.
Some late deciders are highly interested in the election while others are not. While less-interested people are more likely to stay home, many will vote. For interested voters, news and polls will provide a lot of information, so polls will make a difference. For the less interested, personal connections and conversations mean everything. Even conversations overheard at a store can make a difference.
Who Will Win Iowa?
Looking at Iowa, it's impossible to pick a winner. It's very possible that more Iowans have made up their minds this year because of the incredible amount of press coverage, but the polls show that's not the case. The latest Fox News poll finds:
A third of Republican caucus-goers say they may change their mind (33 percent). Even one in four Trump supporters says they may ultimately go with another candidate (25 percent).
That's consistent with past elections. For instance, looking at CNN's exit poll from the 2012 Iowa Caucus (which Santorum won) we find:
Santorum won the Iowa caucus in 2012 by 34 votes over Romney, and his support surged on the day of the caucus.
How did Santorum do it?
Santorum's supporters did a fantastic job of getting people to the caucus locations and of influencing fence-straddlers at the caucuses. My guess is that a bunch of that 35 percent who decided on the day of the caucus heard Santorum's name many times shortly after they arrived.
Does that mean the candidate with the best organization will win Iowa?
Remember, Ron Paul's campaign was famous for its grassroots organization, but that didn't help them in the caucuses. At the same time, Romney had a decent organization, too, but country clubbers might have shown less enthusiasm than Santorum's blue collar forces.
Cialdini's 6 Principles of Persuasion
So, besides organization and expectations, what else might influence a person to vote one way or the other this late in the game?
There are other factors to influence. According to researcher Robert Cialdini and his colleagues, there are six principles of persuasion: reciprocity, consistency, liking, authority, scarcity, and social proof.
Clearly, social proof is the big deal in primary elections and accounts for late deciders breaking for the expected winner. But don't discount the others.
Liking is important. If a candidate's supporters turn off a voter, their numbers might not matter. Undecided or weakly decided voters will also pay attention to how candidates' supporters treat supporters of rival candidates. So treating everyone at the cause with respect might win some converts.
Likewise, authority can play a deciding role in elections. Low-interest voters might be influenced by one political celebrity's personal request than by 100 peers.
Consistency is very powerful but also tricky. For example, people who describe themselves as "very conservative" are very likely to support Cruz. But how people have voted before influences them as well.
Reciprocity can help. Supporters of a particular candidate who offer caucus-goers rides or explanations of the process will influence people who feel they must repay the favor.
Scarcity is a little tricky here, but it's huge in getting voter to the polls or caucus. "This is your only chance to choose the next president."
We Just Don't Know
Which is all a long way to get to this: we won't know the winner until the votes are counted, and news reports of a candidate's surge can be counteracted by what voters see on the ground on caucus day.