Want to Sway Climate Change Skeptics? Ask About Their Personal Strengths (And Show Pictures!)

Read time: 6 mins

Readers of my posts over the last half year will be familiar with the phenomenon of motivated reasoning, in which people’s subconscious emotional impulses lead them to respond, in a biased way, to information that challenges their deeply held beliefs and worldviews. We’ve been focusing on this so much because I believe it explains a great deal of what we here call climate change denial, and the resistance to inconvenient science (and inconvenient facts) in general.

One important researcher on motivated reasoning is Dartmouth political scientist Brendan Nyhan. In Mother Jones, I described one of his previous studies, demonstrating how motivated reasoning can lead to a “backfire effect” when people are confronted with politically inconvenient information:

Take, for instance, the question of whether Saddam Hussein possessed hidden weapons of mass destruction just before the US invasion of Iraq in 2003. When political scientists Brendan Nyhan and Jason Reifler showed subjects fake newspaper articles (PDF) in which this was first suggested (in a 2004 quote from President Bush) and then refuted (with the findings of the Bush-commissioned Iraq Survey Group report, which found no evidence of active WMD programs in pre-invasion Iraq), they found that conservatives were more likely than before to believe the claim. (The researchers also tested how liberals responded when shown that Bush did not actually “ban” embryonic stem-cell research. Liberals weren't particularly amenable to persuasion, either, but no backfire effect was observed.)

So how do you persuade people, if not with factual corrections of the sort run by newspapers? That’s what a new paper by Nyhan and Reifler has undertaken to study.

This time, the contested issues under examination were whether the 2007 troop “Surge” decreased insurgent attacks in Iraq (it did), whether the U.S. economy added jobs during 2010 under President Obama (it did), and whether global average temperatures have risen since 1940 (they have). Those who opposed the Iraq war and supported troop withdrawals were disinclined to credit George W. Bush’s surge with having worked. Those who oppose President Obama are disinclined to credit him on the economy, or to generally believe in global warming—especially that it is human caused.

Nyhan and Reifler once again confronted partisans with information on these subjects that (presumably) contradicted their beliefs—but there was a twist. This time, the contradictory information was sometimes presented in the form of a convincing graph, showing a clear trend (in attacks, jobs, or temperatures). And second, sometimes the individuals went into the manipulation after having undergone a “self-affirmation” exercise, in which they were asked to describe a positive character attribute or value that they possessed, and a situation in which showing that attribute or trait made them feel good about themselves.

And in both cases, the manipulation worked—although by different means.

Presenting an unequivocal graph was powerful enough to change people’s views, even as presenting technical text (at least in the rising temperatures case) was not. Meanwhile, getting people to affirm their values and sense of self also decreased their resistance, presumably because they felt less threatened by challenging information after having had their egos reinforced and their identities bolstered.

This is a really important development, in several ways. First, it shows that scientists who communicate in wonk text, or cluttered graphs that are hard to follow, are shooting themselves in the foot. For instance, here is the wonk text in question, straight from a NASA press release*—the text that failed to work where a graph succeeded:

Groups of scientists from several major institutions — NASA's Goddard Institute for Space Studies, the National Oceanic and Atmospheric Administration's National Climatic Data Center, the Japanese Meteorological Agency and the Met Office Hadley Centre in the United Kingdom — tally data collected by temperature monitoring stations spread around the world. All four records show peaks and valleys that vary in virtual sync with each other. They each show an increase in average global surface temperatures of approximately 0.5 degrees Celsius over the last three decades. Data from each source also indicate that the last decade is the warmest since 1940.

By contrast, here is the graph that worked (it is also the image accompanying this post).

But I think the finding about self-affirmation is even more important. Because what this shows is that people are clearly resisting facts because these threaten their identities—which means that arguing back at them factually will only make them more defensive and engender a backfire effect. By contrast, approaching them in an emotionally sensitive and aware manner, and making them feel less threatened, will open them up. (Sometimes, at least.)

Nevertheless, there are also several potential problems that I see with the study, and its global warming portion in particular.

First, none of the study’s manipulations were done in a really partisan context that would have gotten people’s political emotions firing, priming them to be really, really defensive. For instance, people were asked if jobs increased, but they weren’t asked whether “President Obama’s unfairly maligned stimulus worked to help save the economy from disaster, in contradiction to the bogus claims of many Tea Partiers.”

Similarly, the most hotly contested issue in the climate debate is not whether the world has warmed, but whether humans are responsible for that warming. Many deniers will agree that warming has occurred, but then claim that it’s natural. So they might not have found the information presented in the study very threatening. And once again, it wasn’t presented in the most partisan and emotionally arousing way—e.g., they weren’t shown evidence to prove that “Al Gore is right about global warming and those who have been irresponsibly attacking him, like Rush Limbaugh, don’t know what the hell they’re talking about.”

Second, while I am not surprised that John Q Climate Skeptic cannot refute a definitive temperature graph, I think that those who occupy climate denial blogs—a very small proportion of the total public, but individuals who are very engaged on this issue and very intense in their beliefs—would be more than happy to give it a try. In fact, we see them picking apart and undermining graphs, like the Hockey Stick graph, all the time.

In other words, if high bias is combined with high sophistication (as in the case of the most engaged climate deniers), I don’t think the graphical treatment is going to work. Nyhan and Reifler write that “graphs may be effective in reducing misperceptions because they are more difficult to counter-argue”—but some will still be able to. Al Gore showed lots of graphs in An Inconvenient Truth, and that hardly stopped him from being attacked (by people who don’t know what the hell they’re talking about ;>).

Will the self-affirmation work on such folks? I would imagine at least to an extent. It would make them less defensive. But some people are so set in their beliefs that they are virtually unchangeable.

Luckily, the new study suggests they’re a relatively small proportion of the overall population.

*CORRECTIONThe graph and text discussed above were based on, but did not exactly duplicate, the NASA press release. I regret the error.

Get DeSmog News and Alerts


Sorry, Chris, but Nyhan says nothing about the “anthropogenic” behind global warming, and explicitly states that in his paper. You’re overinterpreting him: http://socraticgadfly.blogspot.com/2011/09/chrismooney-get-real-about-agw.html

Check out page 32 of the report.


“As in the previous two studies, Graph is effective. It reduces misperceptions about global temperature change for both groups (p < .01; columns 1 and 2) and makes strong Republicans more likely to acknowledge that global warming is real and man-made (p < .01; column 4), though it has no effect on beliefs about global warming among respondents who do not identify as strong Republicans (column 3).


As I understand that quote, their results do show that graphs are able to make strong Republicans more likely to acknowledge global warming is real and man-made.

Maybe I am missing something, but it seems Nyhan does say something about the anthropogenic portion of global warming.  Where does Nyhan explicitly state that he has excluded it in his paper?

On the not fully addressing the “anthropogenic,” there’s this, from page 34, body text of the PDF:

Since our stimuli only concern temperature change (and not the role of humans in
causing it), we restrict our attention to the predicted probabilities that respondents will agree that “Global warming is just a theory” in Figure 4.

(The “stimuli” is items such as the graph.) So, Nyhan is saying his research isn’t directly, at least, measuring the feelings on the cause of global warming, just the fact that it is or not.

Nyhan also self-limited the study, per page 31, body text of the PDF:

Second, we excluded respondents who failed to pass a pre-treatment attention filter designed to make sure that subjects were reading survey questions rather than clicking through mindlessly (the question is available in the online appendix).

How much does that correspond to reality? Given the level of motivated reasoning of the typical denialist, they’re trolling for how to reframe information to fit back into their own worldview.

You did say that, but you buried it near the bottom of your post. Since that basically … somewhat disconfirms the stance of this blog post, shouldn’t you have noted that more prominently? I mean, the whole angle of the post is, hey, how much can we learn from Nyhan … and the answer might just be, in light of this “not a lot,” or might be … “converting denialists isn’t as simple as some people would like to believe,” or, per Nyhan’s comment about “self affirmation,” which you didn’t pick up … that “denialists have ‘self-healing’ mechanisms.”

The issue isn’t about “poor self worth” which purportedly leads people into denial (according to your strawman argument) but of holding two incongruent attitudes, which contradicts their self-image as a rational being, and thus their self worth.

Skeptics of climate science can accept that global warming is occurring but are unable to accept the anthropogenic aspect of the problem because it threatens their core identity as market fundamentalists, thus the need to compartmentalize the anthropogenic component.

From my own observations, use of the term “natural” tends to trigger this rationalizing compartmentalization. It’s a term that has no real explanatory value. Scientists would use the term forcing to describe temperature inputs but “natural” changes would be ones that are supposed to occur, given the connotation of the word. Frank Luntz is known for achieving framing effects through the use of certain individual terms that carry emotional baggage. I suspect a Luntz-like disinformation strategy is in play since it’s such a rudimentary mistake that no scientist would make but it has utility as a propagandizing tool for persuading low information voters, as well as some of the better informed ones, too.

Anyway, this is an interesting article with some practical advice.

I can attest to the power of self-affirmation.  Presenting science from peer-review literature does not mean all that much to deniers.  What works better for me is diplomatically trying to understand what their real issue happens to be.  I don’t even discuss the science with a denier all that much anymore, just go straight to the discussing the solutions and risk management. 

Nyhan’s study is a LOT more nuanced and hedged than Chris’ take on it.

on body text page 4 of the PDF is this:

However, unlike previous studies, we find little evidence that affirmation increases the persuasive power of corrective information.

Followed by this a few pages later:

Individuals who encounter dissonant information that is threatening are thus motivated to restore their feelings of self-worth; resolving the dissonance directly is just one of many ways that this goal can be accomplished. Steele supports this claim with a series of experiments showing that individuals who completed an exercise in which they affirmed personally important values and thereby felt secure in their self-worth did not engage in dissonance reduction, suggesting that their need to do so had been eliminated.

(That’s my emphasis within the pull quote.) In other words, rather than getting affirmation from a “scientific elitist,” a John/Jane Doe can “pull a Stuart Smalley” and say, “I’m smart enough to have resisted such elitism.” Nyhan goes on to add that the self-affirmation process can be contingent on personal or situational relevance.

There’s also questions of how tightly self-worth and worldview offer and much more. Nyhan’s saying this process isn’t easy and it’s not simple. If we accept that, we won’t promote simple-to-simplistic strategies for “conversion.” For more on that, and more on why Chris’ reading doesn’t do Nyhan enough justice, hit my blog post: http://socraticgadfly.blogspot.com/2011/09/chrismooney-get-real-about-agw.html

Before you changed your blog post, you had admitted that you did not “plow through the 56-page pdf” either.

In any case, I don’t have any skin in the game either way.  If the study covers just temp rise denial and excludes the anthropogenic cause, whatever.  It still is a useful contribution to psychology.  If it ends up getting strengthened/weakened by further research, I’ll update my views.  I myself just read the portions that dealt with climate and the conclusions section.  I would have to digest it further to fully understand the results, maybe wait to see an abstract.

Let’s focus on the real reason why you are here, though.  You seem to be anxious to show that Chris Mooney is wrong.  Why is that?  Has he been wrong before?  Do you perceive your political views to be opposed to his?

Ah, here is the quote from the cached page.


>> “I didn’t plow through Brendan Nyhan’s 56-page PDF upon which Mooney bases his post; I just scanned the start. It’s got other problems, though. First, he assumes denialists are denialists due to poor self-worth, I think, if he’s listing boosting their self-worth as another “conversion strategy.” I wonder how “lasting” the effects are, as well as how strong.”

There’s this, page 30, body text of the PDF:

Second, we excluded respondents who failed to pass a pre-treatment
attention filter designed to make sure that subjects were reading survey questions rather than clicking through mindlessly (the question is available in the online appendix).

Well, how closely does that approximate reality?

On the not fully addressing the “anthropogenic,” there’s this, from page 34, body text of hte PDF:

Since our stimuli only concern temperature change (and not the role of humans in
causing it), we restrict our attention to the predicted probabilities that respondents will agree that “Global warming is just a theory” in Figure 4.

So, Nyhan is saying his research isn’t directly, at least, measuring the feelings on the cause of global warming, just the fact that it is or not.

That’s why I said, in response to Chris, that even if you take an overly simple reading of Nyhan, all you do is convert a “denier” into a “minimalist.”

Hey, it wasn’t that hard for me to find all this. C’mon, now.

This is a man who, on a Facebook thread, salutes Democrats for working better with big business, evidently buying the Obama line that “we’re already over-regulating businesses.” And other things. I didn’t claim, and haven’t claimed in the past, that Chris is wrong about the science; I have claimed in the past, and still claim, that he’s politically naive in some ways. (And, I’m not alone in that claim.)

As for not “plowing through,” I did plow through, and found nothing to greatly change my original impression of the post. So Chris did note the non-anthropogenic; he buried it 3/4 of the way down, perhaps because it undercuts his original idea. And, that said, because he missed, or ignored, Nyhan’s other caveats, he draws the wrong sociolgical lessons, too.

Were DeSmog Blog insightful enough to have hired me, I would have written a far different post than his, titled: “New Research Says Motivated Reasoning May Not Always Be Easy to Overcome” and subhead “Self-Affirmation Can Inoculate Against Cognitive Dissonance.” That’s what Itook away from Brendan Nyhan’s paper.

>> “As for not “plowing through,” I did plow through, and found nothing to greatly change my original impression of the post. “

That’s not the point.  You criticized Mooney’s article on the Nyhan/Reifler paper before even analyzing it yourself.  Now that you apparently have analyzed the paper, you come to the same conclusion as before, maybe through motivated reasoning?  I don’t think much credibility can be assigned to your comments on the matter, really.

I’ll stick with what I have learned from this article by Mooney and from reading portions of the Nyhan/Reifler paper.  If new info comes up or the authors further clarify in a journal publication, I’ll be open to modifying my understanding of the matter.  I would imagine most rational people would do the same.

>> “So Chris did note the non-anthropogenic; he buried it 3/4 of the way down, perhaps because it undercuts his original idea.”

Uh, he hid it in plain sight?


With no intended criticism of you, Otter, it’s pointless to argue the anthropogenic issue although nonetheless tempting. SG’s initial comment was irrelevant. It might have been interesting to hear what he/she thought the point of contention was but that argument was never made.

The subject of this topic is the Backfire Effect – a psychological/cognitive effect. Chris’ comments are on point since it’s difficult to say that the experimental design countered the Backfire Effect. Subjects weren’t challenged in their beliefs so the reality-denying effect of ideology never came in to play.

I find SG’s comments illuminating in that he/she apparently considers the anthropogenic issue relevant to the discussion of the Backfire Effect. It’s apparently an attempt to validate  his/her skepticism of a portion of climate science, as if arguing against the causes of global warming is somehow less contentious if one admits to the fact that it exists.

I’ve always considered this change in argumentative strategy as a significant event in itself, in that it showed a possible disingenuous intent. I haven’t ruled out other explanations, such as denial, etc., but the willingness to accept the fact of GW and deny the anthropogenic causes smells of intentional deception and bad faith. Primarily, in those cases in which this does apply, it’s a shifting the burden of proof tactic. We’re supposed to knock down an endless series of pseudo-theories that no one has attempted to actually prove. They’re like Potemkin theories, without any legitimate purpose and instead contrived to guile uninformed lay people.