Abstract
This research aims to analyze the chain-mediated effect of the different types of psychological distances (social, temporal, spatial, and probability) and the variables of the Social Identity Model of Collective Action (SIMCA) on the relationship between the use of social media and violent environmental collective action. The study sample consisted of 650 university students (M = 20.8, SD = 2.74) aged 18–35 years from Lima. Analyses were conducted by means of structural equation modeling (SEM) using the AMOS SPSS software, where a statistical model was performed for each type of psychological distance. The findings revealed two statistically significant paths that go from social media to violent environmental collective action, mediated, first, by each of the psychological distances and, second, by social identity and negative emotions (anger and fear). In addition, it was observed that only probability distance on its own acted as a mediator in the relationship between social media and violent environmental collective action. It was also observed that a path from social media to violent environmental collective action was mediated, first, by three types of distances (probability, spatial, and temporal) and, second, by participative efficacy.
Introduction
Climate change, a result of increased environmental temperature due to the accumulation of greenhouse gases, impacts different regions of the world, resulting in rising sea levels, forest fires, and excessive rainfall (Gates, 2021). However, evidence suggests that people perceive climate change as distant in time and/or as a problem specific to developing countries, which may affect their intention to participate in actions to mitigate this phenomenon (Jones et al., 2017). This inconsideration toward one of the greatest threats faced by humanity can be explained by the concept of psychological distance, which is based on the premise that people tend to pay more attention to events that are perceived as close in time or space, more likely to occur, and affecting people with whom we identify (Brügger et al., 2020).
Regarding its location and diverse ecosystems, Peru is one of the most vulnerable countries to climate change (Bergmann et al., 2021; Ministerio del Ambiente del PerúFederal Ministry for Economic Cooperation and Development, 2019). Potential environmental hazards include the increase of infectious diseases, flash flooding, landslides, the loss of the Amazon rainforest, and the melting of the glaciers in the Andes (Magalhães et al., 2019). In the last 20 years, Peru has experienced a series of environmental disasters such as oil spills, river contamination, landslides, and flooding of cities (Carey, 2014; León & Zuñiga, 2020; Pinto Herrera, 2010). In 2017, heavy rains, severe flooding, and landslides (known as “huaicos”) displaced nearly 300,000 people along the coast of Peru (Berdejo, 2021). These disasters were triggered by a high-level “El Niño” phenomenon associated with climate change, which causes the Pacific Ocean to warm and alter rainfall patterns. Over the last decade, deforestation in Peru’s Amazon region has increased, leading to significant environmental impacts (Luque Bustamante, 2020). Another pressing issue is the rapid retreat of glaciers in the Andes, which could lead to the migration of 28% of the highland population and threaten freshwater supplies. Peru also faces risks of drought, riverbank collapse, erosion, and extreme heat stress, particularly in the Amazon rainforest region (Berdejo, 2021).
Several studies have shown that a lower level of psychological distance is connected to actions to mitigate climate change and support for climate policies (Guillard et al., 2021; Sacchi et al., 2016; Singh, Zwinckle et al., 2017). However, these findings coexist with research that contradicts an alleged effect of psychological distance on supportive behaviors toward the fight against climate change (Chu & Yang, 2018; Rickard et al., 2016; S. Wang et al., 2019).
Given that reversing or mitigating climate change requires collective actions (Brechin, 2016), it is crucial to consider theories that explain coordinated group behavior. One of these theories is the Social Identity Model of Collective Action (SIMCA), whose variables have proven to be good predictors of massive environmental initiatives (Furlong & Vignoles, 2021; van Zomeren et al., 2019). In this regard, it is necessary to differentiate between normative and violent collective action. The former is framed within established social norms and involves actions through legal channels (e.g., signing petitions, peaceful protests, and voting), while the latter involves behaviors that disrupt social order (e.g., seizure of premises, damage to public/private property, and road blockades; Bamberg et al., 2015; Moskalenko & McCauley, 2009; Sabucedo & Arce, 1991; Thomas et al., 2012; Wright et al., 1990).
Considering the expansion of social media, many studies can be mentioned regarding the relationship between the use of social media and normative collective actions, both virtual and in-person (Alberici & Milesi, 2018; Gerbaudo, 2016; Halpern et al., 2017). Concerning climate change, it has been concluded that these platforms raise awareness on environmental issues and enable the realization of collective actions, facilitating their sustainability and validity over time (A. A. Anderson, 2017; Ballew et al., 2015; Liu, 2016; Şen & Şen, 2016; Wamuyu, 2018). One of the groups most likely to use these platforms is university students (Basauri Delgado, 2023), who leverage them to organize and advocate for collective goals, such as reducing global warming (Meng et al., 2023)
Most of the research on environmental collective actions focuses on the normative type (Matthews et al., 2023). However, due to the imminence of climate change, violent initiatives are becoming more frequent, attracting greater public attention, and raising awareness of the seriousness of the issue (BBC, 2019; Bugden, 2020); yet, few studies analyze violent environmental collective action (Pauls et al., 2021). Landman and Naumann (2023) found that when peaceful normative collective action fails to produce results, people turn to the violent one, which is predicted by low collective efficacy and perception of injustice. Likewise, when there is a lack of government support for environmental care, individuals tend to turn to violent collective action and identification with activists (Duit, 2010).
This can be seen in the Peruvian context in what happened in the spill of nearly 12,000 oil barrels due to a failure during an unloading operation at La Pampilla refinery in the province of El Callao by the energy enterprise Repsol, contaminating several beaches in Lima (Paucar, 2023; Toledo-Leyva, 2022). This event caused the death of 1,852 coastal fauna and endangered the health of 198 species (El Peruano, 2023). As a result of this event, several normative and violent protests took place, and this was probably due to the exposure on social media of the event (Infobae, 2022). Among the violent protests, those affected vandalized the facade of the Spanish embassy because Repsol is a Spanish company (Valdivia Blume, 2023). Also, several roads were blocked by the protesters (Gestión, 2023).
However, other events far away from Lima do not receive the same media coverage nor do they promote similar protests. An emblematic case is that of the oil spills which occur weekly in the Peruvian Amazon. Between 2000 and 2019, there have been a total of 474 oil spills, of which 65% are due to pipeline corrosion and operational failures; however, this has low media coverage (Kámiche, 2023; León & Zuñiga, 2020). This brings harmful consequences to the environment due to the degradation and deforestation of the biodiversity of the rainforest. One example of this is the reduction of chlorophyll due to oil spills, which leads to dryness and ultimately the death of plants (Arellano et al., 2017; De Oliveira et al., 2020). Besides, oil contamination in marine ecosystems damages mangrove forest and marsh plants, the former stores carbon while the latter prevents shoreline erosion (Rapp, 2023; Wong, 2024).
As mentioned above, violent collective action is promoted through social media, and psychological distance has proven to be a good predictor of pro-environmental behaviors (Guillard et al., 2021; Sacchi et al., 2016). With the increase in violent environmental actions in recent years and the urgency of climate change, we find it necessary to create a model explaining how the use of social media can lead to environmental collective actions. This model will also clarify the relationship between psychological distance and participation in violent environmental initiatives.
Theoretical Framework
Social Media and Collective Action
There is research that associates the use of social media with collective actions through the variables of the SIMCA. It was found that if there are thought-provoking discussions, political identity can mediate between moral obligation and taking part in collective action (Alberici & Milesi, 2018). Furthermore, various studies demonstrate how different forms of efficacy (e.g., self-efficacy for using virtual platforms and political efficacy) mediate between digital activism and political collective action; Halpern et al., 2017; Ruiz-Dodobara et al., 2021). In this sense, anger resulting from the perception of injustice and identification with activists have also proven to be good mediators between virtual platforms and collective initiatives (Chan, 2017; Odağ et al., 2016).
Regarding the use of social media to combat climate change, some research associating these variables can be described. Wamuyu (2018) found that the use of a Facebook page to protect the environment in poor neighborhoods (e.g., reporting a pipe burst and illegal littering) engaged residents in a pro-environmental use of this platform (e.g., sharing the page with friends) and influenced the intention to continue using social media to support environmental initiatives. Liu (2016) described how several online resources (Sina Weibo, QQ, WeChat, Facebook, Twitter, etc.) enabled Chinese citizens to protest against petrochemical projects in the 2007–2014 period. It was found that virtual spaces enabled the exchange of information (about past demonstrations and mistakes) and promoted participation in collective actions, allowing for rapid public involvement. Another example of online environmental activism is the “Right to Water” website, which opposes water privatization in Turkey. This platform allows the organization of several local campaigns across Turkey under one single movement and promotes in-person collective actions, such as protests (Şen & Şen, 2016). It is worth mentioning that in addition to promoting environmental actions, social media allow these actions to remain in force and continue over time (A. A. Anderson, 2017; Ballew et al., 2015).
Psychological Distance and Support of Actions Against Climate Change
The construal level theory holds that the representation of future objects and events depends on their degree of abstraction. The level of abstraction is represented by the so-called psychological distance: objects and events that are distant in time and space are perceived as more abstract and require a higher level of interpretation in contrast to their psychologically close peers, which are perceived as more concrete and with detailed characteristics (Trope & Liberman, 2010). Psychological distance is measured based on four dimensions: spatial, referring to the physical distance between the perceiver and an event; temporal, referring to how much time separates the perceiver from an event; social, referring to the similarity between the perceiver and the individual(s) experiencing an event; and probability, referring to the likelihood of occurrence of an event (Spence et al., 2012). Psychological distance impacts our perception of risk concerning climate change and, therefore, our behaviors. For example, the risk of running out of gasoline in the distant future requires a higher level of abstraction than the concrete possibility of running out of gas in a low-traffic area (Zwickle & Wilson, 2013). Therefore, individuals may be more willing to promote gasoline consumption instead of supporting the search for alternative means of transportation.
Psychological distance is described as a predictor of environmental actions in different studies. Sacchi et al. (2016) found that commitment to the environment, pro-environmental attitudes, and the intention to develop climate-friendly behavior were negatively related to psychological distance. Jones et al. (2017) found that when comparing a group exposed to a video describing climate impacts at distant locations and a group exposed to a video describing climate impacts at nearby locations, the latter had lower scores on psychological distance, higher levels of concern about climate change, and greater intention to engage in climate change–mitigation actions. Singh et al. (2017) observed that the lower the psychological distance, the higher the concern and support for adaptation policies in the face of climate change. These authors also found that concern about climate change fully mediated the relationship between psychological distance and support of adaptation policies; in addition, they proposed that the more effective adaptation policies were perceived to be, the less concern there was about the impact of climate change and the weaker the effect of psychological distance was (Singh et al., 2017).
However, other studies have shown no effect of psychological distance on environmental behavior. Based on responses from university students in the United States and Singapore, Rickard et al. (2016) found that the manipulation of spatial and temporal distance did not affect the support of policies against climate change (e.g., regulating carbon dioxide). Yet, a moderating effect of political orientation was observed. For example, conservatives in the United States expressed more support for policies against climate change when they were informed that climate change impacts would occur close to where they lived. S. Wang et al. (2019) found that spatial proximity to climate change predicts intention to engage in environmentally friendly behavior. Despite this, manipulation of psychological distance did not increase pro-environmental behavior. Schuldt et al. (2018) found that manipulating the perceived geographic proximity of the Maldives (closer or farther from where participants were) influenced the description (concrete vs. abstract) of a video about the climate change impact on those islands, but that this manipulation of distance (close vs. distant) did not influence support for anti-climate change policies.
Social Media and Psychological Distance
Research on social media and psychological distance has focused on the commercial sector. Kim et al. (2016) found that U.S. participants exposed to Facebook pages characterized by high psychological distance (News Feed) responded better to advertisements requiring a high level of interpretation than those exposed to more concrete advertisements. On the contrary, individuals who watched advertisements on the Timeline, a Facebook page, perceived as more proximate (low psychological distance), had more positive attitudes toward advertisements with a high amount of information, that is, requiring a low level of interpretation. Sung et al. (2020) replicated the study with a sample of Korean and U.S. citizens to check for cultural differences. It was found that the results for U.S. participants were identical to those of the study carried out by Kim et al. (2016). However, Korean participants preferred messages requiring high levels of interpretation, regardless of the page (e.g., TimeLine or News Feed). Lee et al. (2018) analyzed whether psychological distance was a good predictor of the intention to support a Facebook initiative related to child welfare in Africa. These authors concluded that cognitive proximity (based on the prominence, relevance, and knowledge of an issue) and emotional proximity (including empathy and connection) were associated with supporting the virtual campaign (Lee et al., 2018).
Based on these observations, the following hypotheses are proposed:
H1. Social media use, different types of psychological distances, moral convictions, social identity with activists, negative emotions (anger and fear), and group and participative efficacy predict violent environmental collective action.
H2. The relationship between social media use and the SIMCA variables (moral convictions, social identity with activists, negative emotions, and the two types of efficacy) is mediated by the different types of psychological distances.
H3. The relationship between the different types of psychological distances and violent environmental collective action is mediated by SIMCA variables (moral convictions, social identity with activists, negative emotions, and the two types of efficacy)
Method
Design
The study presents a multivariate cross-sectional design to create a model based on structural equation modeling (SEM; Preacher et al., 2007). The model is classified as multivariate because it presents latent and observed variables. Likewise, the research is cross-sectional, given that the variables were measured at a single point in time. In the proposed model, social media would predict violent collective action through chain mediation of the different types of psychological distances and the SIMCA variables.
Participants and Procedures
The services of the Analytical Unit of the Instituto de Investigación Científica of the Universidad de Lima (IDIC) were used to obtain the sample. The sample was of the non-probabilistic purposive type and comprised 650 (M = 20.8, SD = 2.74, 50% from private universities) undergraduate university students between 18 and 35 years of age from Lima, Peru, who use social media platforms. They were contacted at the exit of the universities where they were invited to participate in the survey. They were verified to ensure that they met the inclusion criteria: (1) they were undergraduate university students of public or private institutions, (2) they were in the 18–35 years age group, and (3) they used social media more than three times a week. In addition, the project that frames this article was reviewed and approved by the Ethics Committee of the Faculty of Psychology of the Universidad de Lima.
The Monte Carlo Power Analysis for Indirect Effects tool developed by Schoemann et al. (2017) was used to calculate the required sample size for the analysis. This approach was selected because it is considered the most appropriate to avoid underestimating the sample size needed to examine the indirect effect. The effect size was determined following Ferguson’s (2009) suggestion of a minimum association of .20. In addition, as recommended by Field (2009), a significance level of .05 was established, and statistical power of .80 was ensured, obtaining a minimum sample size of 600 participants, which was exceeded.
Measurements
The scales were developed in Spanish by the authors and provided content-based validity testing. A panel of reviewers was engaged to rate the representativeness of each item’s target dimension and clarity in wording (Sireci & Faulkner-Bond, 2014). The reviewers had a profile of academics with previous work on topics related to social psychology and/or scale construction. To this end, individual suggestions for changes in the wording of the items were considered for possible modifications.
Thereafter, a pilot test with 50 participants was carried out to ensure an adequate level of understanding. This demonstrated that the target population adequately understood the instruments. The test lasted 8–10 minutes, and these participants were not included in the final sample count.
McDonald’s omega coefficient was used to assess the reliability of the scales due to the Likert-type nature of the items of the instruments used, following the recommendation of Viladrich et al. (2017) and Hayes & Coutts (2020), who point out that such a coefficient would be less biased with the assumption of tau-equivalence.
The following variables were assessed in the research survey:
Social media use (M = 12.14; SD = 3.54; ω = .801) (Adapted from the study by Ruiz-Dodobara et al., 2023): To examine the social media use concerning climate change, three items were developed, one for each type of use: informative (“I use social media, such as Facebook, Instagram, Twitter, TikTok, and/or Instagram, to be informed about climate change”), expressive (“I use social media, such as Twitter, Facebook, WhatsApp, TikTok, and/or Instagram, to share news and opinions on environmental issues”), and participative (“I participate in online initiatives against climate change, for example, signing petitions, discussions of events, participation in forums, and energy saving actions”). The response format was a 5-point Likert-type items scale (from 1 = Strongly Disagree to 5 = Strongly Agree), with a minimum score of 3 and a maximum of 15. Even if this scale was used to measure different uses of social media, the confirmatory factor analysis showed that the scale behaves as unidimensional, having a good fit index with this structure (Comparative Fit Index [CFI] = .986, Root Mean Square Error of Approximation [RMSEA] = .091, Standardized Root Mean Square Residual [SRMR] = .018), and the factor loadings ranged from .622 to .770.
Moral convictions (M = 15.53; SD = 2.75; ω = .845) (Adapted from the study by Ruiz-Dodobara et al., 2023): Moral convictions, understood as absolute stances on moral issues, were examined with a four-item scale (i.e., “My opinion on climate change is an important part of my moral standards and values”). The response format was a 5-point Likert-type items scale (from 1 = Strongly Disagree to 5 = Strongly Agree), with a minimum score of 4 and a maximum of 20.
Psychological distance regarding climate change: Psychological distance refers to how closely climate change is perceived in terms of time, space, likelihood of occurrence, and impact on individuals. To measure this construct, four subscales were created for each type of distance with 5-point Likert-type items. The spatial distance scale (M = 6.88; SD = 2.23; ω = .634) consisted of three items (e.g., “The area in which I live will NOT suffer the effects of climate change.”); the social distance scale (M = 9.52; SD = 2.97; ω = .791) consisted of four items (e.g., “I identify with people who are affected by climate change.”); the probability distance scale (M = 5.95; SD = 2.18; ω = .836) consisted of three items (e.g., “The severity of climate change is being exaggerated.”); and the temporal distance scale (M = 3.79; SD = 1.50; ω = .684) consisted of two items (e.g., “The consequences of climate change will be mainly felt in the future.”). The spatial distance and probability distance subscales had score ranges from 5 to 15, the social distance subscale had a score range from 5 to 20, and the temporal distance subscale had a score range from 2 to 10. Higher values on these scales indicate a greater psychological distance.
Social identity with activists (M = 9.75; SD = 2.41; ω = .849) (Adapted from the study by Ruiz-Dodobara et al., 2023): This scale assessed identification with environmental activists and consisted of three items (e.g., “I have a lot in common with people involved in different types of climate change protests.”) with a 5-point Likert-type response format (from 1 = Strongly Disagree to 5 = Strongly Agree), with a minimum score of 3 and a maximum of 15.
Negative emotions (M = 29.74; SD = 5.62; ω = .881) (Adapted from the study by Ruiz-Dodobara et al., 2023): The scale of negative emotions comprised anger and fear, and eight items were used, with a 5-point Likert-type response format (from 1 = Strongly Disagree to 5 = Strongly Agree) (Anger: e.g., “I feel outraged by the lack of support to fight climate change.”; Fear: e.g., “It scares me to think about climate change.”), and with a minimum score of 8 and a maximum of 40.
Participative efficacy (M = 11.30; SD = 2.14; ω = .847) (Adapted from the study by Ruiz-Dodobara et al., 2023): For this variable, three items were used, with a 5-point Likert-type response format (from 1 = Strongly Disagree to 5 = Strongly Agree), to examine the perception of the importance of individual contributions to achieving environmental goals (e.g., “I believe I can make a significant contribution to controlling climate change through joint actions.”). The scale had a minimum score of 3 and a maximum of 15.
Group efficacy (M = 11.85; SD = 2.13; ω = .828) (Adapted from the study by Ruiz-Dodobara et al., 2023): Group efficacy was understood as the respondents’ perception of group efficacy toward actions to mitigate the effects of climate change (e.g., “I believe that we citizens, collectively, can fight climate change.”). For this variable, three items were used, with a 5-point Likert-type response format (from 1 = Strongly Disagree to 5 = Strongly Agree), with a minimum score of 3 and a maximum of 15.
Violent environmental collective action (M = 7.41; SD = 3.00; ω = .810): This scale was developed by authors based on the work of Barth et al. (2015) and Landman and Rothman (2020). The scale consisted of three 5-point Likert-type items (from 1 = Strongly Disagree to 5 = Strongly Agree) describing environmental actions that defy established norms. The items were (1) “I would participate in blocking roads to protest against climate change,” (2) “I would be willing to participate in damaging public or private property to protest climate change,” and (3) “I would take part in the seizure of premises to protest against climate change.”
Analysis Plan
An analysis of the correlation matrix among the variables of interest was carried out, followed by the implementation of a multiple linear regression model, using as predictor variables social media use, the different types of psychological distances, moral convictions, negative emotions, social identity with activists, participative efficacy, and group efficacy. The outcome variable in this analysis was violent environmental collective action.
The inferential analyses were carried out with SEM analysis using the SPSS Amos IBM software, which allows plotting the model to be tested, establishing the relationships between the variables. First, to calculate multiple chain mediations, social media use was taken as a predictor variable, and the SIMCA variables were taken as outcome variables, with the different types of psychological distances as mediating variables. Subsequently, a chain mediation was proposed, where the predictor variables were the different types of distances, and the outcome variable was violent environmental collective action. This relationship was mediated by the SIMCA variables.
Given their adequate number of response options, the ordinal variables were treated as intervals using a Maximum-Likelihood approach (Suh, 2015). This approach was also combined with a 500-sample nonparametric bootstrapping approach given that it is not sensitive to sample size and normality distribution (Preacher & Hayes, 2004). The fit indexes CFI ⩾ .90, RMSEA ⩽ .08, and SRMR ⩽ .08 were established as adequate fit indicators (Keith, 2019).
Results
First, the correlation matrix among the variables of interest was analyzed (Table 1). A statistically significant relationship was found among all variables, except probability and temporal distance with violent environmental collective action.
Correlation Matrix.
Note. 1 = Social media use; 2 = Social distance; 3 = Temporal distance; 4 = Spatial distance; 5 = Probability distance; 6 = Moral convictions; 7 = Social identity; 8 = Negative emotions; 9 = Participative efficacy; 10 = Group efficacy; 11 = Violent environmental collective action.
p < .05. **p < .01. ***p < .001.
Furthermore, to verify the assumption of non-multicollinearity, the authors determined the variance inflation factor (VIF) index, which in no case reached a value close to or higher than 5, indicating that there is no evidence of endogeneity (Heiberger & Holland, 2015; Kline, 1998). The VIF statistic of the independent variables ranged from 1.03 to 2.38.
Likewise, a multiple regression model was carried out with violent environmental collective action as the outcome variable. It was noted that social identity, negative emotions (fear and anger), social media use, probability distance, and group efficacy were successful predictors of violent environmental collective action. Social identity and negative emotions were found to have the greatest effect.
An acceptable fit was obtained for the model including social distance (x2/df = 3.791, CFI = .870, RMSEA = .066, SRMR = .059). Likewise, it was noted that only social distance mediated the relationship between social media use and violent collective action in the presence of negative emotions, β = .080, b (SE) = .085 (.069), 95% CI [.066, .238], p = .026; and social identity, β = .144, b (SE) = .150 (.063), 95% CI [.065, .304], p = .004.
An acceptable fit was obtained for the model incorporating probability distance (x2/df = 3.807, CFI = .875, RMSEA = .066, SRMR = .061). It was observed that probability distance, β = −.266, b (SE) = −.281 (.157), 95% CI [−.685, −.089], p = .004, as the sole mediator, as well as accompanied by the negative emotions, β = .091, b (SE) = .095 (.044), 95% CI [.028, .207], p = .004, social identity, β = .090, b (SE) = .091 (.037), 95% CI [.033, .177], p = .011, or participative efficacy, β = .090, b (SE) = .092 (.062), 95% CI [.018, .246], p = .007, mediated the relationship between the use of social media and violent environmental collective action.
Concerning spatial distance, an acceptable fit was obtained (x2/df = 3.338, CFI = .892, RMSEA = .060, SRMR = .055). It was found that spatial distance mediated the relationship between social media use and violent environmental collective action through on negative emotions, β = .117, b (SE) = .125 (.062), 95% CI [.032,.269], p = .009; social identity, β = .127, b (SE) = .135 (.057), 95% CI [.060, .287], p = .004; or participative efficacy, β = .141, b (SE) = .150 (.093), 95% CI [.012, .432], p = .018.
Finally, for the temporal distance model, an acceptable fit was obtained (x2/df = 3.341, CFI = .898, RMSEA = .060, SRMR = .055). It was found that temporal distance mediated the relationship between social media use and violent environmental collective action, only in the presence of negative emotions, β = .110, b (SE) = .116 (.057), 95% CI [.031, .263], p = .010; social identity, β = .111, b (SE) = .116 (.044), 95% CI [.054, .233], p = .004; or participative efficacy, β = .127, b (SE) = .135 (.119), 95% CI [.016, .463], p = .010.
Discussion
Despite the existence of research affirming that digital activism on platforms such as Twitter can generate changes in public attitudes and promote collective actions (Nucci & Hibberd, 2021; Suitner et al., 2023), it was found that even though social media use is a statistically significant predictor of violent collective action against climate change, its effect is weak (see Table 2). This finding is consistent with the concept of “slacktivism” proposed by Morozov (2009), who affirms that social media use would be limited to the virtual sphere, making people feel good about themselves, but these actions alone would not transpire into initiatives with a real impact on climate change. This is evidenced by the lack of statistical significance of the direct path of social media on violent environmental collective action.
Multiple Regression Model.
Note. N = 650. Control variables: Sex and age.
p < .05. **p < .01. ***p < .001.
Climate change probability distance alone was found to be a statistically significant mediator of the relationship between social media use and violent environmental collective action with a medium negative effect size (see Figure 2). However, the multiple linear regression results (see Table 2) showed that the probability distance positively predicted the outcome variable; hence, when people see climate change as imminent, their intention of violent collective action decreased. This may be because as climate change is seen as an event that is bound to happen, people do not see the value of taking risks that could endanger them (Corcoran et al., 2015).
It was also found that both spatial distance and temporal distance had a medium negative indirect effect near the statistical significance level of p < .05 in the relationship between social media use and violent environmental collective action, which may reflect a lack of statistical power for such effects. A possible explanation for these findings is that social media use related to environmental issues (e.g., news sharing and diffusion of events) would make the conception of climate threat less abstract and influence our risk perception (Maiella et al., 2020; Zwickle & Wilson, 2013), thereby reducing skepticism about the effects of climate change (Diehl et al., 2021). Therefore, if climate change is perceived as proximate, our need for extreme behavior outside the established order (violent collective action) would increase.
Concerning the chain mediation analyses, all models (Figures 1–4) showed that the SIMCA variables “Social Identity with Activists” and “Negative Emotions” enabled a path from social media use to violent environmental collective action, going through the four different types of psychological distances.
SEM model—Social distance and violent collective action.
SEM model—Probability distance and violent collective action.
SEM model—Spatial distance and violent collective action.
SEM model—Temporal distance and violent collective action.
Concerning the proposed paths, risk perception would increase through exposure to social media, leading to a greater need to act against climate change. An example of the influence of social media became clear in the context of COVID-19: Zhu and Liu (2021) found a bidirectional effect between media publications and risk perception, as well as the presence of a unidirectional effect from other users’ publications toward threat perception. Regarding emotions, several studies suggest that higher risk perception is related to fear and anger, which motivate normative and violent collective action against climate change (Gregersen et al., 2023; van Valkengoed et al., 2023). This perception of imminent risk associated with climate change would generate outrage toward individuals involved in environmental damage, such as political leaders favoring climate-damaging projects (Gregersen et al., 2023; Stanley et al., 2021). This way, emotions such as fear and anger may be related to aggression toward those who are considered responsible for environmental damage, which would lead to violent collective action (Bohannon, 2013; Böhm et al., 2023; Harth et al., 2013; Musher-Eizenman et al., 2004).
Furthermore, Alberici and Milesi (2018) found that the use of online platforms for sharing information, discussing, and other activities can influence the political identity of users, which in turn affects their willingness to engage in collective action. Based on the results and the connection of climate change with the political sphere, using social media would strengthen users’ identification with individuals who share similar political orientations, such as environmental activists, potentially leading to a radicalization of opinions (Chan, 2017). This way, the polarization found in virtual spaces such as echo chambers (Törnberg, 2018; Yarchi et al., 2021), along with a decrease in psychological distance, would allow for greater identification with activists and consideration of climate change as a proximal situation that affects the in-group (Tajfel, 1974). Furthermore, according to the literature, the path toward participation in environmental initiatives could be explained by the fact that identification with activists leads to the promotion of collective ecological actions (Brügger et al., 2020; Ruiz-Dodobara et al., 2023; van Zomeren et al., 2019; Wallis & Loy, 2021). In the violent sphere, it was found that identification with pro-environmental activism can predict violent environmental collective action, such as seizure of premises and property damage (Furlong & Vignoles, 2021; Schmitt et al., 2019).
Regarding the indirect path that includes participative efficacy, this research showed that the use of social media could decrease probability, spatial, and temporal distance. This reduction positively predicts the perception of participative efficacy, potentially promoting violent environmental collective action. These findings are consistent with recent literature, which has shown that the association between identification with activists and the intention of collective action was partially mediated by the perception of individual benefits and efficacy (Besta & Zawadzka, 2019; Blackwood & Louis, 2012). The empowering effect of social media could be a tentative explanation for the path found: Conversations occurring through social media would increase the notion of urgency of an event (Zwickle & Wilson, 2013) that is perceived as probable and close in time (Ge et al., 2021; Guo & Hou, 2023; Zhu & Liu, 2021). Besides, the persuasive support (Bandura, 2003) that a person could receive for their opinions and intentions through these virtual spaces would reinforce their perception of participative efficacy, leading them to engage in higher-risk environmental actions (Chu & Yang, 2020; Saab et al., 2016). Therefore, increased risk perception, due to an event’s perceived immediacy and certainty of occurrence, would strengthen the individuals’ sense of self-efficacy and their willingness to act (Choi et al., 2017), in this case, toward environmental protection.
It was found that at a lower psychological distance, regardless of its type, there is a greater moral conviction about the need to address climate change (Table 1). However, in this study, higher moral conviction does not transpire into violent actions challenging the status quo of individuals. This finding is in line with previous works, which suggests that the mere violation of individual moral convictions alone is not sufficient to motivate actions that challenge social norms, thus explaining the lack of a mediating effect in this sense (Pauls et al., 2021; Tausch et al., 2011). This is consistent with the literature given that engagement in violent actions may be against moral conventions (Galvin & Harris, 2019). Therefore, if the outcome variable consisted of normative environmental collective actions, perhaps moral convictions would have played a predictive role in the tendency to participate in these actions (Becker & Tausch, 2015).
No statistically significant indirect chain effect, including psychological distances and group efficacy, was found. A possible explanation for these findings is that violent actions are more related to the individual than to the group, as they depend more on individual decisions (Neyra, 2023).
Regarding the practical meaning of our findings, since the study identifies the variables that promote violent environmental collective action, social media communicators should propose messages that increase climate change consciousness with less abstract language, thereby decreasing psychological distance (Kim et al., 2016). At the same time, to promote normative environmental collective action, communicators should seek to increase group efficacy. This could be achieved by showing messages that demonstrate large-scale initiatives already being taken by activists or other groups to mitigate climate damage. These actions would promote an alternative way that could avoid the activation of negative emotions and participative efficacy, both connected to violent behaviors.
With respect to future research, it is important to note that in the context of social media, psychological distance (except for probability distance) by itself does not predict violent environmental collective action, so in further work on the impact of this variable, we recommend integrating emotions, such as anger and fear and identification with activists.
As for limitations, this study is based on a non-experimental cross-sectional design, which does not allow for the establishment of cause-and-effect relationships between variables and limits the conclusions to the specific population selected. Likewise, given that the instruments collected were self-administered, social desirability may have been a factor involved, even if measures were taken to avoid this. Besides, the study was focused on university students, which restricts the extrapolation of the results in different populations. It is important to note that university students have a strong tendency to use social media. Therefore, the effect of these platforms on psychological distance could be enhanced in this individual in contrast to other age groups. We, therefore, suggest that future studies should explore the relationships between these variables in different samples.
In conclusion, the study gives insights about the relationship between social media and violent environmental collective action, through psychological distance and the SIMCA variables. The research highlights the importance of emotions, identity with activists, and participative efficacy in understanding the relationship between the predictive and outcome variables.
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 article was developed as part of the research project “Social media, psychological distance and collective action against climate change” of the Instituto de Investigación Científica (IDIC) of the Universidad de Lima. Project code: PI.50.001.2023
