At a briefing on Capitol Hill yesterday, Stanford University communication professor Jon Krosnick presented the best analysis to date estimating the impact of "ClimateGate" on public perceptions of climate change and of climate scientists. Stanford's Woods Institute for the Environment, where Krosnick is a faculty fellow, has put together a detailed news release on Krosnick's survey analysis. Also above is a YouTube clip of Krosnick explaining the research.

The full report should be read, but below I feature several key conclusions. Despite alarm over the presumed impact of ClimateGate, Krosnick's analysis reveals very little influence for this event. More research is likely to come on this issue and this is just the first systematic analysis to be released.

Yet there is an even more interesting question emerging here than the impact of ClimateGate on public opinion: If communication researchers have difficulty discerning a meaningful impact for ClimateGate, why do so many scientists and advocates continue to misread public opinion on climate change and to misunderstand the influence of the news media? As I argue below, an additional object of study in this case should be the factors shaping the perceptions of scientists and advocates.

--->Krosnick's analysis estimates that the percentage of Americans who believe in global warming has only dropped 5% since 2008 and that ClimateGate has had no meaningful impact on trust in climate scientists which stands at 70% (essentially the same as the 68% level in 2008).

---->According to Krosnick's analysis, the 5% shift has occurred among the 30% of the public already distrustful of scientists. Moreover, for this segment, ClimateGate is not the major factor shifting opinion about global warming but rather the most likely cause is the belief among this segment that recent temperatures are cooler and the weather overall is more stable. Here's how Krosnick explains the shift, discussing trends in several poll questions that track these beliefs:

"Katrina is a distant memory," Krosnick said. "2008 wasn't a year of giant-sized storms, but it was a year of lower temperatures. 2009 also saw the fewest storms since 1997. For some people - especially those who say that they have little or no trust in climate scientists - that's real information. They see that the weather appears to be more stable and that temperatures are cooler, and their reaction is, 'it stopped getting hotter, so maybe global warming isn't happening after all.'"

If the best analysis to date shows very limited impact for ClimateGate, why has there been such an outcry of alarm and such strong assertions from some scientists and advocates that ClimateGate has done serious damage to public opinion? In part it reflects an innocence among some scientists and advocates about the relevance and findings from social science research in the area. From the Stanford news release:

"The scientific community is overreacting to these events," Krosnick said. "In theory, it's possible that public regard for climate scientists has dropped sharply since our 2009 survey. But based on my 30 years of experience in this field, that's very unlikely, because American public opinion, even on a highly publicized and frequently debated issue, changes very, very slowly. So in a two-month period, it's unlikely that there would be a dramatic change. My guess is that relatively few Americans are aware of the media controversy or are paying attention to it, and even fewer are influenced by it."

Not only does innocence about public opinion research shape perception, but several other likely cognitive biases are likely at work. As I explained at the NYTimes' Dot Earth earlier this week, one bias relates to perceptions of media influence:

These biases are well understood and predicted by past research in communication. They include a tendency for individuals heavily involved on an issue to perceive almost all news coverage as hostile to their goals (even news coverage that favors their position); to presume much larger effects for a message on the public than the actual influence; and to apply a faulty quasi-statistical sense to where public opinion might actually stand on an issue, perceiving public opinion as hostile to their goals, no matter what the objective indicators might say.

Another strong bias among the science community also relates to political ideology. While peer-review and other norms in science help correct for the influence of ideology on basic scientific research, they don't correct for the conclusions and judgments that scientists might draw about political debates over policy and regulation. For example, one recent study finds that in light of uncertainty about risks related to nanotechnology, liberal-leaning nanotech scientists are likely to favor stronger regulation while conservative leaning scientists are likely to oppose regulation. In other words, above and beyond scientific expertise and training, ideology in this case plays a significant role in shaping the policy preferences of nanotech scientists.

Heuristic decision-making by elected officials, journalists, and the public is common. We are all cognitive misers, often relying on ideology and values to make sense of the complexities of the political world, and scientists are unlikely to be very different, especially on climate change where strongly framed assertions are made about who is to blame for societal inaction on the issue.

Here's how I described the relevance of these questions in a recent panel at Harvard's Kennedy School of Government:

Raising attention to these questions is important: Political strategists and commentators are calling upon scientists to become more directly politically involved on climate change and other science-related policy debates. Scientists are urged to "fight back," and encouraged to go so far as to organize political action committees and to openly support "pro-science" candidates.

This last trend also raises an important research question: more study of scientists as a social and professional group is needed, specifically examining the influence of scientists' own ideology and news media use on how they evaluate political leaders, define their roles in policy debates, form political opinions, come to support proposed policies, and participate politically. Consider that a Pew survey of AAAS members last year found that 55% of scientists self-identify as liberals compared to 20% of the public and that only 9% of scientists self-identify as conservatives, compared to 37% of the public. This ideological gap between scientists and the public--above and beyond professional expertise or technical knowledge--likely contributes significantly to how scientists differ from the public in their views on political leaders, proposed policy options, and who or what is to blame for policy failures.