Abstract
Research has found that many people view climate change as a psychologically distant, future threat, which leads them to be less motivated to engage in pro-environmental behavior. Engaging in episodic future thinking (EFT; projecting the self into the future to pre-experience future events) may facilitate the perception of future events as psychologically close, thereby increasing the perceived risk associated with those events. Therefore, engagement in EFT regarding climate change–related risks should induce higher risk perceptions and lead to acting pro-environmentally. In two experiments, we demonstrated that engaging in EFT to pre-experience climate change–related risk events was associated with a higher level of risk perception and a greater tendency toward pro-environmental behavior, including energy-saving use of air-conditioning (Experiment 1), willingness to participate in beach cleaning (Experiment 2), and choice of a meal with lower environmental impact (Experiment 2). The current research provides experimental evidence for an innovative approach to improving public engagement with climate change.
Keywords
Introduction
Climate change discourse among scientists, the media, and policy makers has mainly revolved around the future impacts of climate change over a long period of time (e.g., 50 to 150 years from now; van der Linden, Maibach, & Leiserowitz, 2015). Previous research has found that many people still perceive climate change as a psychologically distant, future threat (Jones, Hine, & Marks, 2017; Leiserowitz, 2005; Pidgeon, 2012; Spence, Poortinga, & Pidgeon, 2012; van der Linden et al., 2015). Prior research has shown that risk perception is a crucial factor in public engagement with climate change (e.g., Kahan et al., 2012; Lee, Markowitz, Howe, Ko, & Leiserowitz, 2015; Leiserowitz, 2005, 2006; Safi, Smith, & Liu, 2012; Semenza et al., 2008; Weber, 2006). Identifying a potential approach that can effectively enhance the awareness of climate change risk is an essential topic for environmental educators and policy makers, rendering the present research significant and appropriate. Episodic future thinking (EFT) refers to projection of the self into the future to pre-experience future events (Atance & O’Neill, 2001). Research on the connection between EFT and psychological distance has shown that EFT can lead people to perceive future events as psychologically close (Liberman & Trope, 2008; Trope & Liberman, 2003, 2010). Therefore, engaging in EFT to pre-experience risk events associated with climate change should induce higher risk perception and thereby increase the tendency toward pro-environmental behavior. In this article, we provide experimental evidence showing that engaging in EFT focused on the risk events related to climate change may enhance risk perception and promote pro-environmental behavior.
One possible explanation for the current low levels of engagement with climate change is that people may construe climate change risk as a remote threat (Dilling & Moser, 2007; Jones et al., 2017; Spence et al., 2012). The negative consequences of climate change tend to be perceived as an uncertain set of events that might occur far in the distant future (Leiserowitz, 2005; Spence et al., 2012; van der Linden et al., 2015). EFT, or projecting the self into the future to pre-experience future events, employs concrete mental representations, whereas semantic future thinking (SFT), which involves only semantic knowledge about future events, uses abstract mental representations. For example, the episodic/semantic distinction may be captured by the difference between envisioning specific events occurring during a forthcoming eco-mobility festival (episodic) versus having knowledge about eco-mobility (semantic). Furthermore, construal level theory (CLT) proposes that people may mentally represent the same event or object at high, abstract, or low concrete construal levels (Trope & Liberman, 2003). For instance, “attending a family reunion” can be mentally represented as “respecting tradition” (abstract, high-level construal) or “going to a restaurant” (concrete, low-level construal). From the perspective of CLT (Liberman & Trope, 2008), engaging in semantic thinking about future risky events related to climate change may refer more to semantic knowledge about the impacts of climate change (i.e., abstract representations), whereas engaging in episodic thinking about those events may include a more detailed association with perceptual properties (i.e., concrete representations). Thus, engaging in EFT (relative to SFT) should construe an event at more concrete levels.
With respect to the effect of level of construal on psychological distance, the concept of temporal construal posits that a concrete (vs. abstract) representation of a future event may lead individuals to perceive the event as temporally closer (Liberman & Trope, 1998; Trope & Liberman, 2010). Moreover, vividness and concreteness can enhance the sense that the future is temporally closer (Liberman & Trope, 2008; Trope & Liberman, 2003). Thus, the psychological temporal distance of an event should be reduced when that event is represented in concrete (e.g., episodic) rather than abstract (e.g., semantic) terms. This implication is consistent with past research relating the perceived likelihood of events to the form of their representation. For example, Sherman, Zehner, Johnson, and Hirt (1983; see also Sherman, Cialdini, Schwartzman, & Reynolds, 1985) showed that the concreteness of an imagined event was associated with subjective evaluations of its likelihood. Jones et al. (2017) demonstrated that proximal (vs. distant) framing of climate change communications reduced the psychological distance of climate change impacts. Hershfield (2011) suggested that the degree to which an individual can imagine the future is correlated with the degree to which he or she will engage in future-oriented prosocial behavior, such as pro-environmental acts. Engagement in episodic thinking (relative to semantic thinking) involves more vivid and concrete mental representations (Atance & O’Neill, 2001; Chiou & Wu, 2017; Wu, Cheng, & Chiou, 2017). Hence, EFT focusing on the risks of climate change, which tends to be more vivid and concrete than SFT focusing on those events, should foster reduced temporal proximity to the risks of climate change.
The discounting of future risks is a pervasive feature of the processes by which immediate day-to-day concerns take precedence over future threats (Hirsh, Costello, & Fuqua, 2015; van Vugt, Griskevicius, & Schultz, 2014). Thus, understanding risk perception associated with the possible consequences of climate change is of paramount importance (Kahan et al., 2012; Leiserowitz, 2004, 2006). Risk perception may play important roles in shaping policy and in generating support for climate change mitigation (O’Connor, Bard, & Fisher, 1999; Safi et al., 2012). However, people cannot directly observe the future impact of climate change. The long-term impacts of climate change lie in the future, resulting in a remote distance between our present lives and the environmental impacts (Pahl, Sheppard, Boomsma, & Groves, 2014). The risk perception literature shows that climate change remains a temporally distant issue, and this construal is associated with lower risk perception and reduced concern (Pidgeon, 2012; Weber & Stern, 2011). Leiserowitz (2006) found that Americans’ moderate level of risk perception regarding climate change was associated with the perception that the threat of climate change is temporally distant. Spence et al. (2012) showed that when the psychological distance of climate change was reduced, the level of concern about climate change tended to be higher. Pahl and Bauer (2013) demonstrated that perspective taking in a future scenario of a person affected by negative environmental changes improved pro-environmental engagement (e.g., intentions, time spent engaging with information, and the number of brochures collected) because it reduced the psychological distance of environmental change. The above findings indicate that the psychological distance of climate change may serve as one determinant of the perception of climate change risk. Given that engagement in EFT involves vivid and concrete representations of future events, the closer psychological proximity of climate change induced by engaging in EFT should foster higher risk perception. Furthermore, a recent study demonstrated a connection between a proximal (vs. distal) frame for climate change communication and a greater intention to engage in pro-environmental behavior (Jones et al., 2017). Hence, the present study predicted that engagement in EFT pertaining to risk events associated with climate change would enhance climate change risk perception and thereby promote the tendency to act pro-environmentally.
Overview of the Current Research
Two laboratory experiments were conducted to test the hypothesis that engaging in EFT to pre-experience risk events related to climate change would promote a perception of climate change risk and therefore increase the tendency toward pro-environmental behavior. Experiment 1 examined whether engagement in EFT focusing on risks of climate change would be associated with a higher level of climate change risk perception and a higher likelihood of acting pro-environmentally (i.e., energy-saving use of air-conditioning, Whitmarsh & O’Neill, 2010). The mediating role of risk perception in the connection between engaging in EFT and pro-environmental behavior was tested. Experiment 2 was a replication study examining the effects of engaging in EFT regarding the risk events of climate change on other pro-environmental behaviors, namely, adopting a diet with lower environmental impact and participating in a beach-cleaning activity (Whitmarsh & O’Neill, 2010).
Experiment 1
Method
A total of 93 undergraduate students (45 males and 48 females; M age = 21.1 years, SD = 1.3) at a university in southern Taiwan were recruited to participate in this experiment. The sample size was determined by calculating the number of participants required to satisfy the omnibus F-test (number of groups = 3) under the following conditions: α = .05; ω2 = .10; and power (1 – β) = .80 (Kirk, 2012).
Upon arrival at the laboratory, the participants were informed that they would be helping the experimenters to pilot several unrelated tasks that would be used in future studies. After providing written informed consent, members of each set of three participants were randomly assigned to the three study conditions (EFT, SFT, and control) via a permuted block randomization schedule. The sex proportions were identical in the three study groups. Before undergoing the thinking manipulation, all participants were instructed to read a report regarding climate change that focused on the negative impacts of global warming on Taiwan, such as seawater intrusion, floods, landsides, and water shortages. Under the EFT condition, participants were then instructed to write down three risky life events that would occur in the future under these climate change impacts. The participants were then asked to close their eyes and try to imagine the events they had listed, as specifically and vividly as possible (i.e., to imagine the setting and sequence of the events, the persons and objects present during the events, etc.; Chiou & Wu, 2017; D’Argembeau & Van der Linden, 2004; Wu et al., 2017). Participants were given 2 min to pre-experience each life event mentally. In contrast, participants under the SFT condition were instructed only to write about three risky life events that would occur in the future related to these climate change impacts. Thus, participants in the SFT group did not engage in prospective imagery to mentally pre-experience these risky events. We employed the yoking procedure to control the time spent in the two thinking conditions (Kirk, 2012). The mean time spent performing the manipulation task was approximately equal between the EFT and SFT groups. At the end of the manipulation, participants in both the EFT and SFT groups rated their representations of life events in terms of vividness and concreteness on 7-point scales (1 = not at all, 7 = very much). Participants under the control condition received no manipulation. The control group provided the baseline for risk perception and pro-environmental behavior against which the two thinking-manipulation groups were compared.
Subsequently, all participants were asked to complete an unrelated questionnaire, ostensibly to validate scales. They first completed a climate change risk perception scale (adapted from Leiserowitz, 2005) that included nine 7-point items (1 = very unlikely, 7 = very likely) assessing the seriousness of climate change risks, including the threat to nature, the current impact around the world, and specific impacts locally and worldwide (standard of living, water shortages, and rates of serious disease). The internal consistency of this scale was satisfactory (Cronbach’s α = .84). Participants’ responses to the risk-perception items were averaged to form a perceived risk scale. Higher scores on this scale reflected higher perceptions of risks related to climate change. The second part of the questionnaire consisted of demographic questions and an item regarding the tendency to act pro-environmentally. In the final part, participants were instructed to read and respond to the following text: “The government proposes that air-conditioning not be turned on until the room temperature is above 28° C, as this is an effective way to save energy. From now on, are you willing to turn on the air-conditioning only when the room temperature is higher than 28° C?” Participants rated their tendency to use air-conditioning in an energy-saving manner on a 7-point scale (0 = very unlikely, 6 = very likely).
Results and Discussion
Under the EFT condition, the participants’ ratings of vividness, M = 5.03, SD = 1.22; t(30) = 4.69, p < .001, and concreteness, M = 5.10, SD = 0.94; t(30) = 6.47, p < .001, were higher than the mid-point of the scale (test value = 4.0). Moreover, both vividness and concreteness were significantly greater under the EFT than those under the SFT condition—vividness: M = 2.81, SD = 1.14, t(60) = 7.42, p < .001, Cohen’s d = 1.88; concreteness: M = 3.32, SD = 0.98, t(60) = 7.26, p < .001, d = 1.83—indicating that the effects of the EFT and SFT thinking manipulations differed as intended.
As predicted, risk perception was associated with condition, F(2, 90) = 5.63, p = .005,
Descriptive Statistics by Condition in Experiment 1.
Note. Each condition involved 31 participants. Units of the measure are given in parentheses. Both the EFT and SFT conditions focused on the risk events of climate change. EFT = episodic future thinking; SFT = semantic future thinking.
More importantly, the tendency to use air-conditioning in an energy-saving manner differed among the three study conditions, F(2, 90) = 5.07, p = .008,
We further examined whether climate change risk perception mediated the link between EFT focused on risk events related to climate change and the tendency toward energy-saving use of air-conditioning. As the SFT and control groups did not differ on the dependent measures, these two groups were combined as the reference group for the dummy variable (1 = EFT, 0 = SFT and control). A bootstrap analysis (Preacher & Hayes, 2004) revealed that the bias-corrected 95% confidence interval (CI) = [0.28, 1.05] for the indirect effect of risk perception (B = 0.60, SE = 0.19; bootstrap resamples = 5,000) was significant (see Figure 1). Engagement in EFT-predicted participants’ perception of climate change risks (B = 0.46, SE = 0.14, t = 3.24, p = .002) and risk perception predicted the tendency toward energy-saving use of air-conditioning (B = 1.31, SE = 0.22, t = 5.99, p < .001). Furthermore, the connection between engaging in EFT and the tendency toward energy-saving use of air-conditioning (B = 1.11, SE = 0.35, t = 3.19, p = .002) was no longer significant (B = 0.51, SE = 0.31, t = 1.63, p = .108) when we controlled for risk perception. Thus, the mediation analysis suggested that a higher level of risk perception regarding climate change, induced by engaging in EFT to envisage risk events associated with climate change, increased the tendency to use air-conditioning in an energy-saving manner.
Climate change risk perception mediated the effect of episodic future thinking about climate change–related risk events on the tendency toward energy-saving use of air-conditioning.
In addition, ratings of both vividness (r = .37, p = .003) and concreteness (r = .32, p = .011) in the manipulation check were associated with the tendency toward energy-saving use of air-conditioning. We employed the multiple mediation model (Preacher & Hayes, 2008) to assess the indirect effects of vividness and concreteness. A bootstrap analysis (bootstrap resamples = 5,000) revealed that the bias-corrected 95% CI of both vividness [−0.43, 1.53] and concreteness [−0.65, 0.92] did not exclude zero, indicating that the indirect effects of vividness (B = 0.48, SE = 0.50) and concreteness (B = 0.18, SE = 0.40) were not significant (MacKinnon, Fairchild, & Fritz, 2007). Moreover, when climate change risk perception, vividness, and concreteness were simultaneously treated as mediators in the multiple mediation model, the indirect effects of vividness (B = 0.19, SE = 0.39, 95% CI = [−0.52, 0.98]) and concreteness (B = −0.03, SE = 0.38, 95% CI = [−0.80, 0.68]) were also nonsignificant. However, the indirect effect of climate change risk perception remained significant (B = 0.47, SE = 0.22, 95% CI = [0.13, 1.03]). These findings indicate that the experimental effect was mediated by climate change risk perception. Given that both ratings of vividness (r = .39, p = .002) and concreteness (r = .36, p = .004) in the manipulation check correlated positively with climate change risk perception, the lack of indirect effects of the manipulation check ratings might be due to correlations among the three measures (i.e., multicollinearity).
Results from our first experiment provided support for the hypothesis that engaging in EFT to pre-experience risk events related to climate change may foster higher risk perception and promote a tendency toward energy-saving use of air-conditioning. Moreover, the mediation analysis indicated that climate change risk perception mediated the association between engaging in EFT related to climate change risks and the tendency toward acting pro-environmentally. However, given that the tendency toward pro-environmental behavior in Experiment 1 was based on a self-report measure, the subsequent experiment employed behavioral measures.
Experiment 2
Method
A total of 102 undergraduates (56 females, 46 males; M age = 20.9 years, SD = 1.2) at a university in southern Taiwan participated in this experiment for extra course credit. This experiment involved only two thinking conditions (i.e., EFT and SFT) because the control condition had been examined in the first study. Pairs of participants of the same sex were assigned randomly to receive either the EFT or SFT manipulation; sex proportions in the two study conditions were identical. The sample size was determined by calculating the number of participants required to test a directional hypothesis regarding the mean difference between two independent groups under the following conditions: α = .05, d = 0.50 (medium effect size; Cohen, 1988), and power = .80.
Upon arrival, participants were informed that they would be helping the experimenters to test several tasks that would be used in future studies. After providing consent, all participants were instructed to read the same report regarding climate change as was employed in Experiment 1. The thinking manipulations were identical to those used in Experiment 1. Participants under the EFT condition were instructed to write down three risk events related to climate change impacts that would occur in the future, and then engage in EFT to pre-experience those events. Those under the SFT condition were only instructed to write down three risk events related to climate change impacts that would occur in the future. At the end of the thinking task, participants in both groups rated the vividness and concreteness of their representations of these risk events.
Following the thinking manipulation, participants were asked to complete the climate change risk perception scale employed in Experiment 1 for purposes of validation. After participants completed this questionnaire, the experimenter entered the test room and presented a participation reward sheet, which included two options for a reward meal box (vegetarian and non-vegetarian). The price and caloric content of the two options were identical, but the level of carbon dioxide emissions was lower for the vegetarian than for the non-vegetarian meal box. While the participants were selecting their participation rewards, the experimenter pointed out a poster hanging on the door that announced an upcoming beach-cleaning activity. Participants were invited to provide their contact information if they would be willing to participate in the beach-cleaning activity. The willingness to participate in beach cleaning and the choice of a meal box with lower environmental impact (i.e., a vegetarian meal box) served as indicators of pro-environmental behavior.
Results and Discussion
With respect to the manipulation check, the ratings of vividness, M = 5.57, SD = 1.14; t(50) = 9.86, p < .001, and concreteness, M = 5.53, SD = 1.01; t(50) = 10.85, p < .001, under the EFT condition were higher than the mid-point of the scale. Moreover, participants under the EFT condition rated their mental representations during the thinking task as more vivid and concrete than did those under the SFT condition, vividness: M = 3.33, SD = 1.09, t(100) = 10.14, p < .001, d = 2.01; concreteness: M = 3.22, SD = 0.99, t(100) = 11.72, p < .001, d = 2.31, confirming that the thinking manipulation was satisfactory. In addition, participants’ adherence to a vegetarian diet was not related to the study condition, χ2 = 0.54, p = .461, indicating that the percentage of vegetarian participants did not differ between the EFT (three of 51, 5.9%) and SFT (five of 51, 9.8%) conditions and that the random assignment was satisfactory for creating equivalent groups.
As shown in Table 2, participants under the EFT condition exhibited higher risk perceptions related to climate change (M = 5.59) than did those under the SFT condition (M = 5.26), t(100) = 2.63, p = .01, d = 0.52. Furthermore, the two indicators of pro-environmental behavior were associated with the thinking condition. With respect to participants’ adoption of the meal with lower environmental impact, Table 2 shows that the vegetarian meal box was more often chosen as a participation reward by participants in the EFT group (24 of 51) than by those in the SFT group (14 of 51; χ2 = 4.19, p = .041; B = 0.85, SE = 0.42, p = .042, Wald = 4.12, odds ratio [OR] = 2.35, 95% CI = [1.03, 5.36]). A similar finding was found regarding the likelihood of participating in beach cleaning (Table 2). A logistic regression analysis showed that participants in the EFT group were more likely to participate in the beach-cleaning activity (34 of 51) than were those in the SFT group, 22 of 51; χ2 = 5.70, p = .017; B = 0.97, SE = 0.41, p = .018, Wald = 5.59, OR = 2.64, 95% CI = [1.18, 5.89].
Descriptive Statistics of the Measures in Experiment 2.
Note. Each condition involved 51 participants. Units of the dependent measure are presented in parentheses.
In addition, we employed linear regression and logistic regression analyses to test whether the perception of climate change risk mediated the link between the thinking manipulation and our two indicators of pro-environmental behavior. The SFT condition was treated as the reference group for the dummy variable (1 = EFT, 0 = SFT). In terms of the choice of a meal box with lower environmental impact, engaging in EFT predicted the perception of climate change risk (β = 0.25, t = 2.63, p = .01) and the perception of climate change predicted the choice of choosing a vegetarian meal box (Wald = 17.53, Z = 4.19, p < .001). When we controlled for climate change risk perception, the connection between engaging in EFT and choosing a vegetarian meal box (Wald = 4.12, Z = 2.03, p = .042) was not significant (Wald = 0.47, Z = 0.69, p = .492). A bootstrap analysis showed that the indirect effect (B = 0.94, SE = 0.47, 95% bias-corrected CI = [0.19, 2.03]; bootstrap resamples = 5,000) was significant, indicating that climate change risk perception served as a major mediator of the mediation model (see Panel A in Figure 2). In addition, the mediating role of climate change risk perception in the experimental effect was also observed with regard to the likelihood of participating in beach cleaning. Engaging in EFT regarding risk events of climate change was positively related to climate change risk perception (β = 0.25, t = 2.63, p = .01), and climate change risk perception predicted the likelihood of participating in beach cleaning (Wald = 20.64, Z = 4.54, p < .001). The connection between engaging in EFT and the likelihood of participating in beach cleaning (Wald = 5.59, Z = 2.36, p = .018) was no longer significant (Wald = 1.25, Z = 1.12, p = .263) when climate change risk perception was included in the model to test for mediation. Furthermore, a bootstrap analysis showed that the indirect effect was significant (B = 1.12, SE = 0.53, 95% bias-corrected CI = [0.24, 2.32]; bootstrap resamples = 5,000). Thus, these results suggested that a higher level of climate change risk perception, induced by engaging in EFT, increased the likelihood that participants would participate in beach cleaning (see Panel B in Figure 2).
Climate change risk perception mediated the effects of episodic future thinking about climate change–related risk events on the two indicators of pro-environmental behavior (Panel A: the choice of a vegetarian meal box; Panel B: the likelihood of participating in beach cleaning).
With respect to supplementary analyses, participants’ ratings of vividness were positively correlated with the choice of a meal with lower environmental impact (i.e., a vegetarian meal box; rpb = .30, p = .002) and the likelihood of participating in beach cleaning (rpb = .26, p = .008). Higher ratings on concreteness were associated with the choice of a vegetarian meal box (rpb = .31, p = .002) and the likelihood of participating in beach cleaning (rpb = .31, p = .001). We further used the multiple mediation model (Preacher & Hayes, 2008) to examine whether ratings of vividness and concreteness in Experiment 2 could serve as mediators in the effects on the thinking manipulation of the two indicators of pro-environmental behavior. The indirect effects of vividness (B = 0.98, SE = 0.59, 95% CI = [−0.15, 2.06]) and concreteness (B = 1.07, SE = 0.64, 95% CI = [−0.19, 2.36]) were not significant when they were treated as mediators in the multiple mediation model with respect to the choice of a vegetarian meal box. Again, the indirect effect of climate change risk perception remained significant (B = 1.94, SE = 0.48, 95% CI = [0.15, 1.87]), whereas the indirect effects of vividness (B = 1.03, SE = 0.92, 95% CI = [−0.54, 2.95]) and concreteness (B = 0.15, SE = 0.72, 95% CI = [−1.29, 1.52]) were still nonsignificant when these three mediators were included in the model. A similar pattern was found regarding the likelihood of participating in beach cleaning. When climate change risk perception was not included in the multiple mediation model, the indirect effects of vividness (B = 0.42, SE = 0.47, 95% CI = [−0.62, 1.27]) and concreteness (B = 1.04, SE = 0.60, 95% CI = [−0.10, 2.27]) were not significant. When the three measures were treated as mediators in the model, only the indirect effect of climate change risk perception was significant (B = 1.14, SE = 0.57, 95% CI = [0.20, 2.28]); the indirect effects of vividness (B = −0.34, SE = 0.61, 95% CI = [−1.55, 0.89]) and concreteness (B = 0.45, SE = 0.73, 95% CI = [−1.00, 1.87]) did not reach statistical significance. Taken together, these findings indicate that climate change risk perception mediated the association between engaging in EFT and the tendency to act pro-environmentally in our second experiment.
In short, the connection between engaging in EFT regarding risk events related to climate change and the level of participants’ risk perception was replicated in Experiment 2. The effect of engaging in EFT to pre-experience climate change risks on the tendency to act pro-environmentally was also replicated in the results for the two behavioral indicators. The mediation hypothesis was thus replicated in this experiment by showing that engaging in EFT enhanced climate change risk perception and increased the likelihood of performing pro-environmental acts such as participating in beach cleaning and choosing a meal box with lower environmental impact.
General Discussion
Building on the notions related to the link between EFT and the psychological proximity of future events (Atance & O’Neill, 2001; Liberman & Trope, 2008; Trope & Liberman, 2003, 2010), and the relationship of psychological distance to risk perception (Leiserowitz, 2005; Spence et al., 2012) and pro-environmental behavior (Hershfield, 2011; Jones et al., 2017), we contend that engagement in EFT regarding risk events associated with climate change may heighten risk perception and help in promoting pro-environmental behavior due to EFT’s capacity to induce greater psychological proximity to climate change risks. We found that engaging in EFT to pre-experience risk events related to climate change promoted the inclination toward pro-environmental acts, as reflected in a greater expressed commitment to energy-saving use of air-conditioning (Experiment 1), greater willingness to participate in beach cleaning (Experiment 2), and a greater likelihood of choosing a diet with lower environmental impact, when compared with engaging in SFT, which entails only symbolic representation of the risks of climate change. Furthermore, we demonstrated that climate change risk perception mediated the connection between engaging in EFT and pro-environmental behavior. We provide the first demonstration that engaging in EFT to envisage the risk events of climate change can enhance risk perception, leading people to perform pro-environmental acts. These results have an important practical implication, namely, that projecting oneself into the future to pre-experience life events under the impact of climate change may be sufficient to promote mitigating behavior.
The current research complements the existing literature by showing that engaging in EFT to pre-experience risk events associated with climate change can enhance risk perception with regard to climate change. This effect can be explained by the CLT of psychological distance (Liberman & Trope, 2008; Trope & Liberman, 2003, 2010), which proposes that future events that are mentally represented at concrete rather than abstract levels are perceived as temporally closer. The results of the two experiments indicate that engaging in EFT to pre-experience the risk events of climate change induced more vivid and concrete representations than engaging in SFT. Engaging in EFT (relative to SFT) may involve more concrete mental construal of risk events related to climate change and thereby induce individuals to perceive these events as temporally closer, leading to a greater perception of climate change risk. We provide the first experimental evidence showing a link between EFT and the larger conceptual framework of temporal construal. The observed link between using EFT to envisage future life events related to climate change and a higher level of risk perception is also congruent with prior research on ways to evoke visceral reactions to the risk of climate change (e.g., experiential factors, Leiserowitz, 2006; direct personal experience, Lujala, Lein, & Rød, 2015; reducing the psychological distance of climate change, Jones et al., 2017). Moreover, we found that, compared with control participants, those who engaged only in abstract representation of the risks of climate change did not show an increase in risk perception, indicating that engaging in SFT to construe the risks of climate change does not effectively enhance risk perception. Our findings regarding the ineffectiveness of SFT for enhancing risk perception also provide a viable explanation for the failure of the time-delayed, abstract, and statistical nature of many climate change communications and reports to arouse greater concern about the threat of climate change (Jones et al., 2017; Spence et al., 2012; Weber, 2006). Finally, in two experiments, we found that risk perception mediated the connection between engaging in EFT pertaining to risk events of climate change and the tendency toward pro-environmental behavior. These findings suggest that identifying ways to increase people’s concern about the risk of climate change may offer promising strategies for enhancing engagement with climate change.
Furthermore, environmental mental images refer to the images a person associates with environmental change (Boomsma, Pahl, & Andrade, 2016). The psychological barriers (e.g., lack of immediacy, remoteness of impacts, time lag, and uncertainty; Dilling & Moser, 2007; Gifford, 2011; Lorenzoni, Nicholson-Cole, & Whitmarsh, 2007) that prevent people from engaging with environmental change are related to a lack of vivid and concrete environmental mental imagery. Cognitive psychology provides insight into the facilitating effect of mental imagery on goals that maintain behavior (e.g., Kavanagh, Andrade, & May, 2005; Lindenberg & Steg, 2007) and the importance of mental imagery in guiding behavior (e.g., Connor et al., 2014; Holmes & Mathews, 2010). Preliminary research on environmental mental imagery found consistent and strong associations between mental images and the formation of pro-environmental goals and self-reported behavior changes (e.g., energy-related behaviors and general sustainability behaviors; see Boomsma et al., 2016, for a related review), suggesting that internalizing external environmental impacts as environmental mental images may play a crucial role in initiating pro-environmental behavior (e.g., Leiserowitz, 2006; Smith & Leiserowitz, 2012; Weber, 2006). Hence, it is reasonable to argue that tasks that enhance the generation of environmental mental images should promote pro-environmental intentions and behavior. In the current study, engaging in EFT to envisage personal life events associated with climate change involved concrete, vivid, and personally relevant mental representations. The effect of using EFT to pre-experience climate change on the tendency toward pro-environmental behavior seems to resonate with the notion that increased vividness, availability, and elaboration of environmental mental imagery may foster pro-environmental behavior (Boomsma et al., 2016).
With respect to limitations of this study and future directions, several issues are noteworthy. Our findings were obtained in a laboratory setting; thus, caution should be exercised when generalizing the results to real-world settings. Given that laboratory findings represent immediate effects, the temporal duration and persistence of the experimental effects require long-term investigation. The use of undergraduate samples limits the generalizability of our results to other populations. Although a nonintervention control group (included in Experiment 1) provided a baseline for climate change risk perception and the tendency to act pro-environmentally, a pre- and post-test design is necessary to detect the precise effects of EFT interventions on risk perception and pro-environmental behavior. Furthermore, we acknowledge that the lack of a control EFT condition is a limitation. The question of whether mere engagement in EFT per se, independently of its content, might increase the tendency to act pro-environmentally remains unanswered. In addition, the absence of a measure to assess psychological temporal distance did not allow us to elucidate the underlying mechanism in the link between experimental manipulation and climate change risk perception. CLT, developed by Liberman and Trope (2008), proposes four types of psychological distance: spatial or geographical distance, temporal distance, social distance, and certainty (e.g., how certain is it that an event will happen). In the current study, our thinking manipulation (i.e., engaging in EFT to pre-experience the risk events of climate change) may refer more to the representation of where it occurs and when an event occurs rather than to whom it occurs and whether it occurs. Thus, engaging in EFT about risk events related to climate change should be more closely related to the perceptions of spatial and temporal distance. However, CLT proposes that the four dimensions of psychological distance are interrelated (Trope & Liberman, 2010). Hence, it would have been a benefit to include a multidimensional measure of psychological distance to determine the impact of our thinking manipulation (i.e., EFT vs. SFT) on various types of psychological distance. Finally, the mean time spent performing the manipulation task was approximately equal between the EFT and SFT conditions under the yoking procedure. However, the actual amount of thinking time at the individual level between the two thinking conditions may not be identical. Therefore, the possible confounding of the amount of thinking time was a limitation.
Our findings also suggest several interesting avenues for future research. Besides the indicators of pro-environmental behavior employed in the present research, future research might examine whether EFT focusing on the risk events of climate change is also effective in promoting other mitigating actions (e.g., buying environmentally friendly products; saving water by taking shorter showers; turning off lights when they are not in use; and walking, cycling, or taking public transport for short journeys). A recent study showed that interacting with one’s weight-reduced self in a virtual environment could lead to better regulation of dietary practices (Kuo, Lee, & Chiou, 2016). Participants who were exposed to vivid messages about the amount of energy used for heat when experiencing a virtual shower used cooler water than those exposed to vague messages (Bailey et al., 2015). Given that vivid and concrete mental representations of events may induce closer psychological proximity (Liberman & Trope, 2008; Trope & Liberman, 2010), the question of whether the application of virtual reality technology to simulations of concrete future consequences of climate change would enhance risk perception is worthy of investigation. Moreover, we suspect that the enhancement of risk perception by EFT related to climate change is not invariant. Several personal factors may moderate this association. For instance, future-oriented individuals are less likely to discount temporally remote consequences (Nagin & Pogarsky, 2003; van Gelder, Hershfield, & Nordgren, 2013). It would be interesting to explore whether the effect of engaging in EFT on perceptions of climate change would be more prominent for people with higher future orientation or for those with lower tendencies toward discounting (i.e., the tendency to focus on distant, larger benefits of pro-environmental acts in the context of climate change engagement). In addition, a connection to the future self has been shown to play a crucial role in individuals’ construals of future events (Hershfield, 2011). Future research could examine whether individuals with high future self-continuity are more susceptible to this EFT intervention. Furthermore, future research could examine the role of implementation intention (Gollwitzer, 1999), that is, forming a personal action plan, specifically planning when, where, and how one will act (e.g., “If situation Y is encountered, I will perform goal-directed behavior X”). Numerous empirical studies have shown that goal-pursuit interventions can robustly promote goal attainment (Sheeran, 2002; Sheeran & Webb, 2016) and increase the performance of intended actions (Gollwitzer & Sheeran, 2006; Prestwich & Kellar, 2014). Thus, a combined application involving engagement in EFT, with a focus on implementation intention, in the domain of climate change may greatly increase successful behavior change and, thus, pro-environmental behavior. Finally, CLT suggests that engaging in low-level construal (e.g., answering “how” questions; Freitas, Trope, & Gollwitzer, 2004) is more likely to yield concrete representations, whereas engaging in high-level construal (e.g., answering “why” questions) is more likely to yield abstract representations (Liberman & Trope, 1998, 2008). In line with the present findings that EFT (relative to SFT) induced more vivid representations and greater risk perception, we predict that construing climate change risk at low (relative to high) levels should be associated with higher perceived risk and construing climate change engagement at concrete (relative to abstract) levels should be related to pro-environmental acts.
In conclusion, the current research demonstrated that engaging in EFT to pre-experience risk events related to climate change could promote the tendency to act pro-environmentally more effectively than SFT. Given that visceral reactions to the risks of climate change encourage pro-environmental behavior (Weber, 2006), an EFT intervention with a focus on climate change impacts may promote mitigating behavior by enhancing risk perception. Increased concern about climate change may not be achievable through the abstract and statistical nature of many climate change communications (Jones et al., 2017; Lujala et al., 2015; Weber, 2006). To increase personal involvement with climate change, it may be helpful to engage frequently in prospective imagery that vividly and concretely envisages some of the future risks of climate change.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was partially supported by the Ministry of Science and Technology, Taiwan (MOST 107-2410-H-110-037-MY2).
