14 April 2014

Climate change communication: Key psychological research findings (and why you haven't heard about them yet) (1)

Research has identified a number of psychological barriers that can prevent people from believing in or acting on messages about climate change. Luckily, it has also suggested strategies for overcoming these barriers. First of a two-part report by Paul Connor.

Part 1 of 2 parts | Part 2

1. Climate change activists are pretty decent social psychologists. Social psychologists are terrible activists.

Most climate change activists I know are at least to some degree also social psychologists. They constantly consider questions like ‘how can we change the way people think’, ‘how can we make people care more’, and ‘what is the sound bite that is going to be most effective for this campaign?’. Generally, they hold reasonable theories about human psychology and societies. And for the most part, they’re willing to revise these theories as experience dictates.

Every email to supporters, every rally poster, and every social media post functions like a mini-experiment, the results of which are carefully analysed. Maybe, for example, more people showed up to the rally that had the funny poster. But maybe that was due to the nicer weather. More data is needed. And so on. This isn’t science, obviously. But it’s also not far away. Moreover, as we’ll see below, most climate change activists are already intuitively applying much of the advice the research has to offer for them. In my experience, then, as amateur social psychologists go, climate activists are actually pretty passable.

Social psychologists, however (and I include myself in this category), are terrible activists. The majority of them, admittedly, are driven by sincere desires to change the world for the better. They want to uncover important knowledge, and they want this knowledge to be applied in the world in salubrious ways. Yet typically, their lives consist of nothing but slaving away in laboratories, designing studies, poring over data, and investigating super specific questions. If they’re lucky, these answers will get published somewhere, presented to other academics at a conference, and they’ll keep their jobs a little longer. And normally, this is where the story ends. Rarely, if ever, do they take time out from academia to communicate their findings with the public. The result is knowledge that never makes it out of the ivory tower, and all of their good intentions amounting to little practical impact on the world.

One story I heard recently gave me a stark reminder of this. A journalist friend of mine met a young PhD candidate at a climate change adaptation conference. When he asked her about the importance of her work, she swiftly responded that there were a number of groups who could utilise and benefit from it. But when he asked if she had been in touch with any of these groups, she could only mumble something, about the work being available in journals.

I had two responses to this story. The first was no, lady. People are not going to just go out and find your work. People are busy. Moreover, people have never even heard of your ‘journals’. And even if people did somehow find your work, unless they were connected to an academic institution (which most people are not), they are not going to pay the $30-40 most journals charge for a measly PDF download. Seriously. It’s not going to happen.

My second response, however, was a painful realization that I am just as guilty. Last year, I spent a huge amount of time scouring databases for research related to climate change communication for my psychological honours thesis. As a result, I now have a fairly good grasp of the scientific literature in the area. I also happen to know many climate change campaigners who would love this knowledge shared with them. One of them, in fact, co-directs of one of Australia’s biggest climate campaigning groups, and even asked me specifically if I could produce readable research overview of the area.

Yet despite this, I have still managed to continually find excuses not to do it. And quite easily, too, I might add. I was busy. There were other people more qualified to write it. None of it would markedly change practice anyway. Generally, I want to move on to doing more research, rather than spend time regurgitating what I already know. And focusing on new research is a much better way of advancing one’s own academic career than reaching out to the public. All of these, I believe, are key reasons why academics often fail to take time to communicate their knowledge and research findings with the world outside academia.

Another reason is simply that no one pays them to do it. For the most part, academics are paid for their work. But engaging with the public will in most cases be pro bono, which makes it less attractive. Moreover, many academics already feel that they’re earning less than someone of their abilities would earn in the private sector, so in a sense already feel like what they are doing is partly voluntary. This renders it even less likely that they will take on additional tasks outside of their own research and teaching.

But to social psychologists, I say: get out there. If we really are motivated by the desire to have a positive practical impact on the world, then we can and should be doing far more to make sure our work find their way into the hands that will make use of them. And to climate change activists, I say keep reading. I have put together the following research overview for you, and as a sometime climate change activist myself, I feel confident in saying that I think there will be plenty there that will interest you.

2. There has not been much experimental work done on the social psychology of climate change.

One of the first things I realized when I started looking for experimental psychological research into climate change is that there is actually surprisingly little of it. To be sure, there is a lot of research related to climate change, but the majority of it has to been correlational, rather than experimental. And this is an important difference.

Take the 2012 paper ‘The psychological distance of climate change’ by Nottingham University’s Alexa Spence and colleagues. This study asked a representative sample of 1822 Britons (1) a number of questions about how concerned about climate change they were, how prepared to act on climate change they were, and how distant they perceived climate change to be. There were a number of types of distance measured: geographical (whether they thought climate change would impact distant areas, close areas, or both), social (whether they thought climate change would impact people in developing countries, people in developed countries, or both), temporal (when they thought the impacts of climate change would be felt), and also general uncertainty about climate change (whether it is happening, what its causes are, whether there is a scientific consensus regarding it, and so on).

Spence and colleagues found that increased psychological distance of all kinds was significantly related (2) to lower concern about climate change. Interestingly, they also found that all forms of psychological distance were also related to lower preparedness to act except for one form of social distance – the more that Britons perceived that climate change would disproportionately affect developing countries, the more prepared they were to act on it (go Britons!).

Now, these findings are definitely interesting, but as mentioned above, they are correlational, rather than experimental. And this is especially important to remember for activists looking for practical advice to use in their work.

There is a simple saying that is repetitively drummed into the heads of all science students that correlation does not equal causation. Just because variables A and B are related doesn’t mean that A causes B, and it doesn’t mean that manipulating A will cause changes in B. In fact, B might cause A. Alternatively, A and B might have no direct causal connection at all, but both be controlled by another variable, C.

My first year statistics lecturer gave me an example of this that has always stuck with me. There is a significant negative correlation, he pointed out, between the amount of pirates in the world and global temperatures. As pirates have decreased, global temperatures have risen. But this does not mean that the drop in pirate prevalence has caused global warming. Increasing piracy (unfortunately!) is probably not a good method for combating climate change.

The relationship between the amount of pirates in the world and global warming. Keep in mind that the pirate figures are approximate. And that correlation is not causation!
 So it goes with much correlational research. Spence and colleagues found that A (the perception that climate change will disproportionately affect developing countries) was related to B (people’s willingness to act on climate change). But this does not necessarily mean that if one manipulates A (for example by promoting messages outlining climate change’s impact on the developing world) one can increase B, people’s willingness to act on climate change. Instead, the relationship between A and B may work in the other direction.

After all, the prediction that climate change will disproportionately impact developing countries, while common in sophisticated analyses of the issue, is not generally stressed in the mainstream media. So people’s understanding of this fact could simply be tied to how much reading they have done about the issue. And it may simply be the case that the same people who are already more willing to act on climate change are also more willing to read about it. If this is the case, then it is in fact B that is bringing about A, with the willingness to act causing people to learn more about the disproportionate impacts of climate change on the developing world. Therefore, focusing on changing perceptions about climate change’s impacts on the developing world will not necessarily increase willingness to act, despite there being a relationship between the two.

I am not saying that this is definitely the case. But we can’t rule out this possibility based on this study. And this, I think, is a good example of how research can easily be misinterpreted and misapplied. Spence and colleagues wrote that their results “suggest” that highlighting distant impacts of climate change will promote willingness to act on it. Manipulating A, in other words, will affect B. And sure, their results are at least compatible with this possibility. But generally (and I’m sure that this is why Spence and colleagues used the lukewarm term ‘suggest’), in order to prove that such causal relationships exist, we need to perform experimental research as well.

So what counts as experimental research? In a nutshell, experimental research is based on manipulating variables and observing the effects of that manipulation. If you manipulate variable A, hold all other variables constant, and observe an effect of your manipulation over variable B, then this gives you firm grounds for arguing that there is a causal connection between A and B. This allows much more confident conclusions to be drawn from research than if the findings are correlational alone.

With this article, then, I am going to focus on summarising what I think are the most informative findings of experimental research into climate change communication. Not only do I find it is a bit more useful, for the above stated reasons, but there is also much less of it out there, which (happily) will allow this to be an article, rather than a book.

Win win!

3. Climate change messages can be threatening to some people. You may want to take this into account.

Possibly the best-known experimental study in the social psychology of climate change communication is the 2011 paper by Matthew Feinberg and Rob Willer of the University of California Berkeley entitled ‘Apocalypse soon? Dire messages reduce belief in global warming by contradicting just world beliefs’. This paper garnered a lot of attention when it was published, but unfortunately was somewhat misrepresented in the media, so it warrants some further examination.

Their paper actually contained two experiments. In the first, the researchers surveyed the climate change beliefs of 97 undergrad students, as well as the extent to which they reported holding ‘just world beliefs’. Put simply, just world beliefs are the belief that the world is just, and that everyone ultimately gets what they deserve. Widely studied in psychology, it has been shown to act for some people as a defence against anxiety, and to enhance the motivation to strive for things in the world. However, it has also been shown to enhance the tendency to view the victims of injustices as somehow having deserved their fate.

Anyway, after 3 to 4 weeks, the same undergrads were given one of two different climate change messages to read. The two messages were based on information pulled from the Intergovernmental Panel on Climate Change’s 2007 report, and were largely identical. An excerpt of the message is as follows: “As ocean temperatures continually rise, it is predicted that the frequency of category 4 and 5 hurricanes will also rise”. One of the messages, however, ended on an optimistic note regarding humanity’s ability to respond to climate change: “A drastic decrease in CO2 emissions would pretty quickly slow the rise in global temperatures, and in the long-run, would even allow the Earth to return to its normal temperature patterns”. The other ended pessimistically: “Global warming is going to change everything for the worse. It is just too big of a problem for science to grapple with. We don’t even know where to start”. After reading these messages, the undergrads completed another survey (social psychologists love their surveys) that again assessed their level of belief in climate change.

Feinberg and Willer found that when the undergrads received the message with the optimistic ending, their belief in climate change significantly increased. However, when they read the message with the pessimistic ending, their belief in climate change (nearly significantly) decreased. Interestingly, this effect was also found to depend on the level of people’s just world beliefs. The optimistic message increased belief in climate change among people with both low and high just world beliefs. However, the pessimistic message only decreased belief in climate change among people high in just world beliefs. Hence the title of the article: ‘Dire messages reduce belief in global warming by contradicting just world beliefs’.
Results from Feinberg and Willer’s first experiment. For people low in just world beliefs, there was little difference in the effect of the dire or positive messages over climate change belief. For people high in just world beliefs, however, there was a dramatic difference between the effect of the two messages.
Now, most climate change activists would likely find this study interesting, but perhaps wonder about its usefulness. After all, no real life climate change message would ever end with “We don’t even know where to start”. But this is where Feinberg and Willer’s second experiment was valuable. In it, they recruited 45 participants via Craigslist and ‘primed’ high just world beliefs in some and low just world beliefs in others. They then showed the participants two genuine videos produced by climate change campaigning groups (one of these videos was described as portraying a speeding train heading towards a small girl, the other as portraying children simulating the ticking of a clock. A Tck Tck Tck video perhaps?). After watching these videos, the participants were asked about their belief in climate change. The results showed that the participants primed to have high just world beliefs reported significantly lower belief in climate change than participants primed to have low just world beliefs, indicating that people’s just world beliefs do play a role in their reactions to real life climate change communications.
So what does this tell us? Should a part of your climate change communication strategy involve trying to convince people that the world is unjust? Probably not. But it is nonetheless certainly worth keeping in mind that some climate change messages will be threatening to certain people as a result of just world beliefs. Some people, for whatever reason, hold high just world beliefs (research has shown that these beliefs are more common in people with conservative, authoritarian political values, and in religious people). Consequently, there may be a tendency for these people to react to more dramatic or ‘dire’ messages negatively.

It is important, however, to keep this result in context. The difference observed between the high and low just world belief groups in Feinberg and Willer’s second experiment was relatively small. And it is also unclear whether this difference represented a loss of belief among the high just world group, or simply a lower gain in belief relative to the low just world belief group. Moreover, even the use of a message far more pessimistic than any climate change campaigning group would use in the first experiment did not quite decrease climate change belief to a significant level across all the research participants. Plus, even though Feinberg and Willer’s ‘optimistic’ message led to higher belief in climate change, they did not measure whether it also led to greater motivations to actually do something about climate change.

In fact, there is some evidence to suggest that too much optimism could actually de-motivate people. In their 2013 study ‘Positive fantasies dampen charitable giving when many resources are demanded’, Heather Barry-Kappes of New York University and colleagues found that positive ideation about the resolution of an issue can reduce people’s willingness to donate time or money to it. In their study, participants considered the issue of lack of medication in Sierra Leone. Some of them imagined an idealised resolution of the issue, while others either imagined a more realistic outcome, or imagined nothing. The study found that when the perceived demands of donating money or time were low (e.g. the request was for a $1 donation), there was no difference between groups. However, when the demands were perceived as high (e.g. the request was for a $25 donation), the group that imagined the idealized resolution was significantly less likely to be willing to donate both money and time.

Results from Barry-Kappes and colleagues’ first study. People who fantasized about an ideal resolution to a problem were significantly less likely to donate a larger sum ($25) than people who thought about a more realistic resolution.
So definitely, be positive. People need hope. And as Feinberg and Willer showed, if you don’t give it to them, then some people (often the more churchy, conservative types, but not always!) may start believing in climate change a little less as a result. But while you’re being positive, definitely also remember Barry-Kappes’ and colleagues’ results and don’t go overboard with it. Hooray for nuance!

4. Show windmills. Lots and lots of windmills.

The concept of people being turned off by pessimistic messages leads nicely to the next paper I want to discuss, which is the 2009 paper by Saffron O’Neill and Sophie Nicholson-Cole of the University of East Anglia entitled '’Fear Won't Do It'': Promoting Positive Engagement With Climate Change Through Visual and Iconic Representations’. Similar to Feinberg and Willer, O’Neill and Nicholson-Cole sought to investigate whether the use of ‘fear-appeals’ (defined as any attempt to motivate people to take actions to avoid an undesirable threat) in climate change communication could have unintended negative consequences.

They noted that research into the use of fear appeals in health-related persuasive communication has consistently shown that people’s ‘coping appraisals’ (their perceptions of health-related behaviours and their ability to perform them) are more important in determining their intentions and behaviour than their perceptions of the health threats themselves. Consequently, they theorised that using fear-inducing messages and imagery in climate change communications might act to reduce people’s engagement with the issue by negatively affecting their coping appraisals.

In their study, 10 young mothers, 10 young professionals, and 10 high school students were all provided an array of images related to climate change, which depicted a wide range of scenes and objects associated with the issue, such as ice caps melting, wind turbines, and so on. The participants were then asked to rank the images in two different ways. First, they ranked the images according to how personally important or unimportant they made climate change seem. Second, they ranked the images according to how able or unable the images made them feel to do anything about climate change.

Examining people’s responses, the authors noted that the images that were ranked as making climate change seem more personally important (images of starving children, floods in Bangladesh, dried up lakes, and rising temperature graphs) were also often the images that made people feel less able to do something about the issue. Converseley, the images that made people feel more able to do something about climate change (e.g., a thermostat, a cyclist, someone fitting low energy lightbulbs, a house with solar panels) were not ranked either high or low in terms of how important they made climate change seem.

List of the images ranked in O’Neill and Nicholson-Cole’s study both as making climate change seem relatively more important and making people feel less able to do anything about it.
Yet while these results are interesting, there are some major problems that need to be pointed out with using ranking in this way. Because ranking was used, the results don’t really tell us how much the different images made people feel that that climate change is important or that they could do something about it. Nor do they tell us how much difference was perceived between the high or low ranked images. It is possible, for example, that all the images made climate change seem highly important or unimportant, or were all highly empowering or disempowering. It is also possible that there were big differences between the high and low ranked images in important they made the issue seem, but small differences between the high and low ranked images in how empowering they were. Or vice versa.

So there is a lot that remains unknown regarding this study. But there was at least one particularly interesting aspect of the findings. This was the fact that out of all the images, there was one in particular that participants ranked as making climate change seem both personally important and as being empowering  the image of wind turbines. I laughed when I read this, because if there is one image that is absolutely ubiquitous in climate change campaigning communications, it is the wind turbine. In fact, the friend who asked me to write this research summary? Her group not only has a big picture of wind turbines on its home page, but it even has a windmill in its logo! This, then, seems to be an area where activists are already intuitively applying the advice the research has to offer. So kudos, activists.

Despite this interesting finding, however, O’Neill and Nicholson-Cole’s study largely missed the mark in terms of investigating the effect of fear-appeals in climate change communications. Another paper that examined the effect of fear in climate change communications was the 2001 study by Anneloes Meijnders from the Eindhoven University of Technology and colleagues entitled ‘Role of negative emotion in communication about CO2 risks’.

In their study, 162 community members of Eindhoven in the Netherlands were asked about their levels of concern about climate change, and then exposed to one of three videos. The first of these videos was a ‘low fear’ climate change video that explained climate change and some of its effects. The second was a ‘high fear’ video that presented the same explanation, but included dramatic images and music to induce a more fearful response. The third was a video related to computing that functioned as a control group. After the videos were watched, the participants read texts containing either strong or weak arguments in favour of switching to energy efficient light bulbs (perhaps a more novel idea in 2001!), and reported their attitudes and intentions towards their use. They then reported how much fear they had experienced as a result of the earlier videos. Three weeks later, the participants were contacted again and once more reported their attitudes and intentions regarding the energy efficient bulbs.

Perhaps the main thing to note from the experiment’s results is that while the high fear video produced significantly more fear, it did not produce any overall differences in people’s attitudes or intentions towards switching to energy efficient light bulbs. In fact, if anything, the higher fear video increased people’s focus on and engagement with the issue. Among people who had watched the high fear video, the relative strength of the arguments they read in favour of the bulbs made more difference than it did for the groups that watched the other videos, which indicated that the high fear group was more attentive to the arguments. It was also found that this effect was stronger for people who were originally lower in concern for climate change.

Yet even this study has its problems. While the high fear video produced more fear, it is also likely that it was simply more interesting than the other videos, with its dramatic images and music. Thus, heightened interest is what may in fact have increased the less concerned participants’ focus on the issue, rather than fear. This is a good example of what scientists call a confounding variable – something (in this case ‘interest’) besides the variable being manipulated (in this case ‘fear’) that differs between the experimental groups. Confounding variables make it hard to be sure what causes any observed differences between experimental groups, and so weaken our ability to make conclusions from results. So while these results could indicate that using a little bit of fear might make less concerned individuals more focused on climate change, they may also just mean that making your messaging more interesting will have that effect.

Unfortunately, then, psychological research to date really doesn’t have all that much advice for activists regarding the use of fear appeals in climate change communication. Which is a shame, because I know from experience that whether or not to use fear appeals is a question that many climate activists struggle with a lot. I hope that this situation gets rectified soon!


(1) Representative samples are those that keep key demographics in line with population distributions. For example, if 60% of Britons vote Tory, then you would need 60% of your sample to vote Tory, and so on.
(2) The term ‘significantly’ has special meaning in science. In this context, it indicates that a statistical analysis showed it to be unlikely that the relationships between these measured variables, e.g. the tendency for people scoring higher on the ‘geographical distance’ questions to score lower on the ‘concern about climate change’ questions, were unlikely to have occurred by chance.


Barry-Kappes, H.,  Sharma, E., & Oettingen, G. (2013). Positive fantasies dampen charitable giving when many resources are demanded. Journal of Consumer Psychology, 23(1), 128-135.
Feinberg, M., & Willer, R. (2010). Apocalypse soon? Dire messages reduce belief in global warming by contradicting just-world beliefs. Psychological Science, 22(1), 34-38.
Mejinders, A.L., Midden, C. J., & Wilke, H. A., (2001). Role of negative emotion in
communication about CO2 Risks. Risk Analysis, 21, 955–966.
O’Neill, S. J., & Nicholson-Cole, S. (2009). ''Fear Won't Do It'' : Promoting Positive Engagement With Climate Change Through Visual and Iconic Representations. Science Communication, 30, 355-379.
Spence, A., Poortinga, W., & Pidgeon, N. (2012). The psychological distance of climate change. Risk Analysis, 32(6), 957-972.

Paul Connor is a climate change campaigner and social psychological researcher at the University of Melbourne. He blogs at www.paulconnor.org.