Healthy Influence – Persuasion Blog

When you break the Persuasion Rules, you will get what you deserve.

As a prime, on-going, case study in persuasion and scientific failure, I’ve documented the persistent sins, follies, and zealotries of the Climate Change Chorus.  Today, we get yet another brick in the oven of Global Warming.

First, recall past research.  We looked recently at the impact that temperature has on people’s beliefs about climate change.  Li et al. demonstrated that when it is hotter than usual, people express more acceptance of the theory.  When it isn’t, they don’t.  At first pass, this sounds okay, but if you are True Green Science, you know the failure here.  The theory says that both hotter and colder are the dangerous outcomes, not just hotter alone.  People are not getting that more complex message from the Persuasion Scientists in the Climate Change Chorus.  Worse still, it appears that people don’t think about their beliefs in that Long Conversation in the Head from the Central Route, but rather Cue off of their bodies (is it hot in here or is it just me?) to determine their allegiance.

Today’s example comes from a great series of experiments by Risen and Critcher that measure or manipulate ambient temperature, then relate that to beliefs about climate change.  Risen and Critcher also thoughtfully include a measure of political position (party) as a useful comparison since climate change clearly contains more political energy than scientific sense.  In the three main studies I’ll discuss here, people got a temperature manipulation then completed a similar survey of climate change beliefs.  Here’s the basic survey employed in the three experiments.

. . . participants completed a social/political questionnaire by indicating their degree of agreement with several political and social issues. The question about global warming was embedded within questions about four other current topics (firearm sales, charter schools, legalization of marijuana, and the firing of radio talk show host Don Imus). The critical question about global warming was adapted from a CNN poll (CNN/Opinion Research Corporation, 2007), “Which of the following statements comes closest to your view of global warming? Global warming is a proven fact. Global warming is a theory that has not yet been proven.” For all questions, participants responded on an 11-point scale, with lower numbers corresponding to a proven fact and higher numbers corresponding to an unproven theory . . . Participants reported their political ideology by checking the terms they believed applied to them (Republican, Democrat, Independent, Conservative, and Liberal).

Now, let’s detail those temperature manipulations and the results from each of the three experiments.

In Study 1 . . . Participants were taken outside under the pretense of judging the height of several campus landmarks . . . The experimenter recorded the outdoor temperature to the nearest degree Fahrenheit while the participant filled out the questionnaire.

What happened?

In fact, the effect of temperature, β = .24, t(63) = 2.04, p = .05, was as strong as the effect of ideology, β = .22, t(63) = 1.90, p = .06, and was not qualified by it, β = −.16, t(63) = 1.36, p > .17. Thus, outside temperature influenced liberals and conservatives similarly. Also, there was no relationship between temperature and attitudes toward the four other issues (ps > .21), indicating that temperature did not shift participants’ political positions more generally.

Those β values indicate the standardized regression coefficient, beta, and can be understood here the same way as correlations.  Thus the β of .24 for the temperature effect is about a Medium Windowpane, just like a Pearson r of .24.

Now for the second experiment.

In Study 2, participants were randomly assigned to complete the survey either in a small heated cubicle or in an identical nonheated cubicle.

Results?

Participants who responded in the heated cubicle believed global warming was more of a fact than those who responded in the control cubicle, β = .24, t(80) = 2.46, p = .02; the effect of political ideology was also significant, β = .39, t(80) = 3.98, p < .001. As in Study 1, the two variables did not interact, β = −.10, t(80) = 1.02, p > .31; liberals and conservatives were similarly influenced by the ambient temperature.

Finally, now, the third experiment.

In Study 3, participants were randomly assigned to complete the study in a small room that had been heated to approximately 27 °C (81 °F) or in a control room that was 23 °C (73 °F). The materials were equivalent to those used in Study 2 with a few small changes. First, all participants began by answering the question, “Before you express your views toward the following issues… to what extent would you say the room feels hot or cold?”

And, the results.

Even with participants’ attention directed to the temperature of the room, we found that participants in a hot room were more likely to indicate that global warming was a proven fact (M = 8.44, SE = 0.46) than were those in the control room (M = 6.85, SE = 0.43), β = .37, t(28) = 2.54, p = .02. As before, political affiliation also predicted belief, β = −.66, t(28) = 4.44, p < .001, but did not interact with the heat manipulation, β = .21, t(28) = 1.47, p > .15.

There’s a lot of data here, so use this Figure to assemble the main findings from the three experiments. Notice the main effects for both temperature and for political party and the absence of an interaction (the lines should cross, touch, or aim at each other, but don’t here).

Risen and Critcher thus replicate that prior work of Li et al. and find that people are more likely to believe the climate change hypothesis when they are hot. See that the results are consistent. Temperature always has an effect on beliefs. Political party affiliation always has an effect on beliefs. Crucially, the interaction of temperature and politics is never statistically significant and thus demonstrates the simple main effect for each variable. Finally realize that the effect sizes are typically in that Small to Medium Windowpane, about 40/60. You don’t need a huge sample to artificially inflate statistical significance; the effect is there.  Thus, liberals and conservatives think with their bodies and when they feel hot, Global Warming!, and when they don’t, Nevermind!

So.

Isn’t that good news for the Climate Change Chorus?

No.

If you think the hypothesis only holds for hotter, you’ve read the science with the Brooks Effect, drawing bad, simple, and incorrect conclusions because you don’t read the methods and results, just over heated discussions.   The hypothesis claims both hotter and colder outcomes, more extreme weather at both ends of the thermometer.   If the people are only hearing half the message, they are not hearing you.  You are failing at your persuasion.

Worse still, people are thinking about this hypothesis with their bodies and not their minds. The CCC has executed such a confused, noisy, and alarmist persuasion campaign that people clearly have no good idea of the point. They simply cue off of their body and thoughtlessly express a belief. Change the temperature, the body changes, and so does the belief, regardless of your political position.  That means people are not taking the Central Route on this with the Long Conversation in the Head which is what you’d expect if this was Falling Apples.

I would argue the reason for this failure in persuasion lies in the original weak science and the worse political motivation that drives the CCC.   If you take the time to actually read the methods and results from the climate science you realize how similar it is to all those bad health epidemiology studies I often dissect in this Blog.   Poor definition of the population of observations; convenience and biased sampling of observations; arbitrary mathematical transformations of datasets; the absence of mere statistical significance despite large sample sizes in many studies and when significance can be trumpeted, the effect sizes are so small they are barely distinguishable from random variation.  There is simply not much there, there, in the climate change literature to date.

Yet, the CCC aggressively mounted a persuasion campaign that the sky is falling and in a location near you, so you’d better look out. But, as the Risen and Critcher and the Li et al. studies demonstrate, people respond with strong Cue driven beliefs that show only surface and temporary strength and worse still miss the CCC warning for both hotter and colder.

Thus, if you truly wish to save the People from Themselves and along the way, save the Planet, what should you do? Nothing the Global Warming crowd does.  If there is science here, it is lost in persuasion.

Risen, J. L., & Critcher, C. R. (2011). Visceral fit: While in a visceral state, associated states of the world seem more likely. Journal of Personality and Social Psychology, 100(5), 777-793. doi:10.1037/a0022460

P.S. I’m giving short shrift to the great work from Risen and Critcher.  Please read it to see how you do good science to understand psychological phenomena.  They ran an additional series of experiments testing the mechanism that produces this effect “.  .  .  temperature as information (Studies 2 and 3), conceptual accessibility (Studies 4 and 5), and fluency of simulation (Studies 6a and 6b).” They conclude, ” . . . the results suggest that visceral states can influence one’s beliefs by making matching states of the world easier to simulate and therefore seem more likely.” Furthermore, Risen and Critcher make no arguments concerning the global warming debate and take no sides.  I’m reasoning from their research to my own conclusions.