Global warming? Public attitudes often at mercy of the weather, study finds.

When much of North America was caught in the grip of unusually cold weather in early January, images of ice-encased lighthouses on Lake Michigan and hunched, bundled pedestrians crunching along snowy sidewalks elsewhere brought hoots from some circles asking "where is global warming?"

That question was less likely to be on the lips of people living along the US West Coast and up into Alaska, as well as northeastern Canada, eastern Asia, or much of northern Europe and Eurasia from Britain to locations deep in the Russian heartland. In these areas, temperatures were running from 5 to 30 degrees Fahrenheit above normal, according to the University of Maine's Climate Reanalyzer, a climate analysis and visualization project.

Social scientists have long recognized a tendency for many people to base their views on global warming at least in part on what they see outside the window, even at the time they are asked for their opinion. Now, a research team at Columbia University in New York suggests that this tendency may result in part from the way humans process information when confronted with complex issues.

In a new study, the team suggests that perceptions of today's temperature extremes and how unusual they seem trigger memories of past warm or cold spells in ways that lead people to overestimate the frequency of either kind of event in the past. Those estimates then feed into their stance on global warming and their level of concern over its effects, according to Lisa Zaval, a Columbia University graduate student and lead author of a formal summary of the study’s results, which were published online Jan. 12 by the journal Nature Climate Change.

Political leanings remain the single most significant predictor of where someone is likely to stand on climate change, the researchers acknowledge. But if one could cancel out political leanings, where a person stands may well rest on the mental equivalent of raising a moist fingertip to the wind and comparing the results with memories.

The study is part of an ongoing effort among social scientists to understand what they see as a gap between public perceptions and the scientific consensus that human-triggered climate change is real, notes Lawrence Hamilton, a sociologist at the University of New Hampshire who studies the issue but was not a member of the team reporting the results.

Over the past five years, a clear majority of Americans has agreed with the proposition that global warming is happening. The majority ranges from 71 percent to 63 percent depending on the year, according to polling data gathered by the Center for Climate Communications at George Mason University. But when asked about humans as the cause, the proportion of respondents who agree hovers around 50 percent.

The group whose opinions seemed to be the most affected by the immediate whims of weather are unaffiliated voters, according to a study Dr. Hamilton and colleague Mary Stampone published last year.

It's the climate analogue to undecided voters and elections, he suggests.

"A lot of people decide on who to vote for president the night before the election on the basis of what they saw on TV or the latest scandal," Hamilton says. And the phenomenon is most pronounced among those he dubs "low information voters."

Indeed, Ms. Zaval and her colleagues were testing three notions that – political leanings aside – could influence someone's opinion about global warming.

Some have posited that an individual's level of knowledge about the issue – in this case the important difference between weather and climate – affects their position. Others have looked to the wording on survey questions and the influence of terming the climate's warming trend as climate change or global warming.

Zavals's team wanted to add another potential influence: on complex issues, people often base their opinions on information that isn't the most relevant but is readily available and easily noticed. It's a simple approach that makes the most efficient uses of an individual's time, and it may well be hard-wired in to human cognition, some researchers have argued.

Others had noted it, but the team aimed to figure out why people use this.

The team conducted four experiments using surveys involving up to 686 people nationwide to test the influence of these factors while trying to minimize the influence of the 800-pound gorilla – political affiliation – on the results.

Varying the wording of questions by swapping global warming and climate change had no significant effect on people's responses.

The big surprise came when examining the level-of-knowledge issue, Zaval writes in an e-mail. "Teaching participants about the scientific distinction between transient local temperature and general global climate change did not eliminate the bias" in their viewpoints, she writes.

Finally, the team conducted two survey-based experiments to test the cognitive explanation. They found that the perception of temperature "today" played a more influential role in shaping a person's viewpoint on global warming than did perception's of the temperature "yesterday." And they found that people who thought "today" was unusually warm were far more likely than people who thought "today" was colder than usual to overestimate the number of unusually warm days in their area during the past year.

It may seem like a no-brainer to say more people believe in climate change with it's hot than when it's not, the University of New Hampshire's Hamilton acknowledges.

This latest study goes beyond to figure out how that happens, he adds.

"The answer is that people are using highly accessible, though not really relevant information rather than less accessible information that in principle should be relevant," such as scientific reports, he says.

"That's a cognitive point, and an interesting one" that is likely to get picked up as other researchers continue to examine the range of factors that influence public attitudes toward global warming.