Research Article: Combining role-play with interactive simulation to motivate informed climate action: Evidence from the World Climate simulation

Date Published: August 30, 2018

Publisher: Public Library of Science

Author(s): J. N. Rooney-Varga, J. D. Sterman, E. Fracassi, T. Franck, F. Kapmeier, V. Kurker, E. Johnston, A. P. Jones, K. Rath, Felix Creutzig.


Climate change communication efforts grounded in the information deficit model have largely failed to close the gap between scientific and public understanding of the risks posed by climate change. In response, simulations have been proposed to enable people to learn for themselves about this complex and politically charged topic. Here we assess the impact of a widely-used simulation, World Climate, which combines a socially and emotionally engaging role-play with interactive exploration of climate change science through the C-ROADS climate simulation model. Participants take on the roles of delegates to the UN climate negotiations and are challenged to create an agreement that meets international climate goals. Their decisions are entered into C-ROADS, which provides immediate feedback about expected global climate impacts, enabling them to learn about climate change while experiencing the social dynamics of negotiations. We assess the impact of World Climate by analyzing pre- and post-survey results from >2,000 participants in 39 sessions in eight nations. We find statistically significant gains in three areas: (i) knowledge of climate change causes, dynamics and impacts; (ii) affective engagement including greater feelings of urgency and hope; and (iii) a desire to learn and do more about climate change. Contrary to the deficit model, gains in urgency were associated with gains in participants’ desire to learn more and intent to act, while gains in climate knowledge were not. Gains were just as strong among American participants who oppose government regulation of free markets–a political ideology that has been linked to climate change denial in the US–suggesting the simulation’s potential to reach across political divides. The results indicate that World Climate offers a climate change communication tool that enables people to learn and feel for themselves, which together have the potential to motivate action informed by science.

Partial Text

Scientific evidence supporting an urgent need to mitigate anthropogenic climate change is clear [1]. Yet, social science data are also clear: around the world, public understanding and concern over climate change are not commensurate with the risks we face [2]. Although public opinion favoring action to mitigate climate change is increasing, it is not strong enough to generate the individual and governmental actions necessary to meet international climate goals [3, 4]. The decision by the US to withdraw from the Paris accord [5] and shifts in US federal policy towards production of fossil fuels [6] further threaten global efforts to mitigate climate change [7]. Communication tools that are both scientifically rigorous and that motivate informed action on climate change are urgently needed [8].

Participants in World Climate take on the roles of parties to the UN climate negotiations and are challenged to create an international agreement that limits warming by 2100 to well below 2 °C above preindustrial levels. As in the UNFCCC process, participants specify Nationally Determined Contributions (NDCs) for the parties they represent while seeking to influence the other parties through face-to-face negotiations. Participants’ proposals are then entered into the C-ROADS climate policy model [26, 27], which provides immediate feedback about the expected climate outcomes of those decisions.

We seek to test whether World Climate helps people learn about climate change science while motivating them to learn more and increasing their intent to take action. Fig 4 summarizes prior theory showing how gains in knowledge, affect, desire to learn more, and intent to take action relate to one another. If the information deficit model of learning [9] were correct, then outcomes such as a desire to learn more and intent to take action (“Desire to Learn” and “Intent to Act,” respectively) would arise from gains in knowledge resulting from exposure to information about climate change. However, knowledge may neither function alone nor be sufficient to drive action. For example, knowledge about the causes and impacts of climate change is positively correlated with concern, an affective response [36]. Further, climate change knowledge and affect are thought to have a bidirectional, reinforcing relationship [37] (shown as links between “Knowledge” and “Affect,” Fig 4). Affect is also important in risk perception and support for climate action [38, 39], suggesting that changes in “Intent to Act” and “Desire to Learn” may be affected more strongly by affect than knowledge.

Our sample consisted of 39 World Climate sessions conducted between September 2015 and October 2017 in locations in North and South America, Europe and Africa, with a total of 2,042 participants (Table 1). These sessions are broadly representative of the wide spectrum of educational and cultural settings in which World Climate is used, ranging from early secondary school to graduate school to sessions open to the public; participant ages from 11 to more than 75; and heterogeneous participant backgrounds from no prior education or interest in climate change to professionals whose career focus is climate change and sustainability.

As described below, we test multiple hypotheses to examine whether participants experienced gains in knowledge, affect, and desire to learn and do more; how those gains were related to each other, if at all; the potential influence of political views; and potential threats to validity. We apply a Bonferroni correction to reduce the likelihood of erroneously finding statistically significant results when multiple hypotheses are tested. For any conventional threshold for statistical significance, α, the Bonferroni correction is αadjusted = α/N, where N is the number of tests carried out. We conduct a total of 104 tests for statistical significance across the full set of t-tests and regression analyses. The Bonferroni-adjusted significant levels for α = 0.05, 0.01, and 0.001 are therefore p < 4.8 x 10−4, p < 9.6 x 10−5, and p < 9.6 x 10−6, respectively. That is, we reject the null hypothesis that World Climate had no impact for each individual test we conduct only if the probability of erroneously doing so is p < 4.8 x 10−4. Doing so yields a familywise error rate—the probability of erroneously rejecting any true null hypothesis across the full set of tests conducted—of 0.05. All statistical analyses were conducted using IBM SPSS Statistics, version 24. We used exploratory factor analysis (EFA) of survey results to test the learning model (Fig 4), specifically, to test for the presence of constructs capturing knowledge of climate change, affective responses to the issue, and participant intent to take action to address it. EFA reduces the dimensionality of the dataset by identifying latent variables in the surveys, if any, [45] and enables us to assess whether they correspond to the constructs in the learning model. The EFA identifies a number of constructs capturing participants’ knowledge of climate change, their affective engagement with the issue, and their intent to take action to address the problem. Here we assess whether World Climate led to changes in these constructs. The information deficit model of science communication [9] suggests that gains in knowledge lead to behavior change, as represented in Fig 4 by the hypothesized links from Knowledge to Intent to Act and from Knowledge to Desire to Learn. Yet, research demonstrates the importance of affect in risk perception and action [38, 39]. Worry, interest, and hope were strongly associated with support for climate change policy in a nationally representative survey in the U.S. [39]. Similarly, Leiserowitz et al. [38] found affect to be a strong predictor of climate change risk perception. Under the information deficit model, gains in knowledge about climate change should be positively associated with gains in people’s desire to learn more about climate change and intent to take action. In contrast, under an affect-mediated model of learning, gains in the emotions people experience would be associated with gain in their desire to learn more and intent to act. Here we ask how the gains in each construct identified in the EFA are associated with gains in the others, and with a wide range of session- and participant-level attributes such as where the session was held and participant socio-demographic characteristics. The effects of political ideology on climate change beliefs are well established and, in the US, free market ideology has been linked to climate change denial [12, 56, 57]. Is World Climate effective among those who oppose regulation and other collective action solutions for climate change? We now consider two potential sources of bias. First, survey completion was optional, raising the potential of bias from voluntary response sampling if participants with more extreme prior views about climate change or who had the strongest reactions to World Climate, positive or negative, were more likely to complete the surveys than those indifferent to the experience. Second, approximately half the participants were required to participate in World Climate as part of a course unrelated to climate change. However, the other half elected to participate, raising the possibility that these individuals were not representative of the populations in their nations. The World Climate role-play simulation offers an approach to climate change communication that enables people to learn for themselves through a scientifically grounded and socially engaging experience. Across a diverse set of participants, World Climate was associated with statistically significant gains in three areas: (i) knowledge of climate change causes, dynamics and impacts; (ii) affective engagement including greater feelings of urgency and hope; and (iii) a desire to learn more and intent to take action in the real world. The results are robust across diverse geographic, cultural, educational, and sociodemographic conditions, suggesting that World Climate is a versatile and effective tool for motivating action informed by science.   Source: