Graduate School of Communication
Master Thesis Persuasive Communication
Local versus global: manipulating the perceived
spatial distance of climate change
Name: Annabelle Gielen
Student ID: 10734740
Supervisor: Dr. M. Meijers
2
Abstract
Nowadays, the overall public awareness of climate change is high, but consumers' actual behavior is still not often sustainable. This may be due to perceiving climate change as a psychological distant issue. This research will examine how psychological distance (i.e., spatial distance: local versus global) of climate change affects individuals’ concern about the environment, which in turn influences people’s likelihood of forming pro-environmental attitudes and purchase intentions. In addition, the moderating role of consumer identity is included in the study. By doing an online experiment, 115 respondents were randomly assigned to either a locally or globally framed ad about a green product. The results showed that a fit between a local message type and a consumer with a local identity will lead to a higher purchase intention for a green product, mediated by environmental concern. No moderated mediation effects were found for ad and brand attitudes. Moreover, when
comparing the outcomes with the results when global identity was included as a moderator, it can be concluded that local identity is more important to take into account when creating a fit between message type and consumer identity. This study contributes to the scientific literature about the effects of a local versus global message type, mediated by environmental concern, on pro-environmental responses, and the moderating role of consumer identity.
3
Introduction
In its climate report, the Intergovernmental Panel on Climate Change (IPCC) indicates that climate change is largely the result of man-made greenhouse gas emissions and that the current climate trends will continue in the future (Van Dorland, 2014). Nowadays, the concerns for climate change increase, as well as the efforts made by organizations and consumers in order to encourage pro-environmental behavior (Grinstein & Riefler, 2015). Pro-environmental behavior expresses itself in activities that satisfy the needs of this
generation without harming future generations (Simpson & Radford, 2013). Previous research on the perceptions of climate change indicate that the overall public awareness of the issue is high, but consumers' actual purchase behavior is still not often sustainable (Wei et al., 2017). Consequently, the aim for many organizations and the society in general is to persuade consumers to behave themselves in a pro-environmental manner.
Spence et al. (2012) state that the disconnection between the overall public awareness of climate change and pro-environmental behavioral responses may be a result of perceiving the issue as psychological distant. In other words, climate change is perceived as an issue that takes place far away and concerns people in other parts of the planet. Therefore, in order to increase consumers’ pro-environmental responses, it is advisable to decrease the
psychological distance of climate change (Spence et al., 2012). For example, by making people realize that it can be a local problem which can have an effect on their own
environment. In contrast, other research suggests that perceiving climate change as a proximal issue may contribute to inaction. Namely, local impacts of climate change are generally perceived as less severe than the global impacts and climate change may therefore not be considered as a personally threatening issue (Spence & Pidgeon, 2010). There is limited and contradicted work within the area of psychological distance of climate change and the effect
4 on pro-environmental behaviors. More research is required, at which the present study will contribute. This research will examine the relationship between the psychological distance (i.e., spatial distance: local versus global) of climate change and pro-environmental responses (i.e., the attitudes and purchase intention for a sustainable product).
Furthermore, the influence of consumer identity on the effectiveness of a local versus global message is taken into account. Some studies state that consumer identity is a potential moderator for the relationship between a local versus global message on pro-environmental behavioral intentions (Grinstein & Riefler, 2015) or intentions to support a corporate social responsible activity (Russell & Russell, 2010). The present study takes into account the moderating role of consumer identity on environmental concern. It is expected that people who have a local (versus global) identity become more concerned about the environment, only when they are exposed to a local (versus global) message. In this way, a fit exists between information framing and consumer identity (White et al., 2011).
In addition, the mechanism of environmental concern potentially driving the effect on pro-environmental responses is investigated. Several studies did research about the influence of psychological distance on environmental concern and pro-environmental responses (Howansky, 2015; Spence et al., 2012; Singh et al., 2017; Whitmarsh, 2008). Although Whitmarsh (2008) did not find a significant effect of environmental concern driving the influence of psychological distance on actions towards climate change, other studies did (Spence et al., 2012; Singh et al., 2017; Howansky, 2015). As explained above, it is expected that a fit between a local versus global message and consumer identity will lead to more environmental concern, which in turn will lead to more pro-environmental responses.
To summarize, this study will explore how psychological distance (i.e., spatial
distance: local versus global) affects individuals’ concern about climate change, which in turn influences people’s likelihood of forming pro-environmental attitudes and purchase
5 intentions. In addition, the moderating role of consumer identity is included in the study. This leads to the following research question: To what extent does a fit between local versus global message type and local versus global identity leads to more positive ad and brand attitudes and purchase intention for a green product, and is this effect mediated by environmental concern? The answer to this question can provide relevant information for advertisers and campaigners who want to encourage pro-environmental behavior and are considering using local or global message types.
Theoretical framework
Local versus global message type
According to Smith and Petty (1996), using different ways of presenting a particular message, also known as framing, can influence people’s beliefs and behaviors. In the present study, climate change is framed as either a local or global issue. A local issue is
conceptualized as an issue related to a particular area close to where one lives, such as a city or home country. A global issue is conceptualized as an issue that has an effect on the whole world (Grinstein & Riefler, 2015). Alongside local versus global, the difference between a near versus distant framed message is also a common distinction within the literature. For example, Singh et al. (2017) measured distance by asking whether the relative geographic distance of climate change impacts was near where the respondent lives or far away. In this case, a distant impact of climate change could be a local issue, that takes place on the other side of the world. However, a global issue is a much broader concept that has an effect on the whole world in contrast to a local issue (Grinstein & Riefler, 2015). Despite this difference between local versus global and near versus distant, the concepts can also be divided into an abstract and concrete approach, whereby local and near are concrete terms and global and distant are abstract terms (Trope & Liberman, 2010). Both literature that uses the concepts
6 local versus global or near versus distant are included in the theoretical framework. However, the present study will use the typology local versus global to describe the different terms.
A local versus global message type is related to psychological distance, which refers to the extent to which an issue isnear or far away from the self (Trope & Liberman, 2010). Trope and Liberman (2010) state that there are four different ways of psychological distance: spatial, social, temporal, and hypothetical. This study will focus on the spatial distance of climate change, comparing a locally versus globally framed message, thus whether it is local issue or an issue that concerns the whole world.
According to the Construal Level Theory, people form mental construals which allow us to go beyond our direct experience (Trope & Liberman, 2010). On the one hand, issues that are perceived as nearby (i.e., local) are likely to be construed in a concrete way (low-level construals). On the other hand, the construals of issues that are perceived far away from the direct experience (i.e., global) are likely to become more abstract (high-level construals; Trope & Liberman, 2010). When it comes to climate change, McDonald et al. (2016) state that a low-level construal (i.e., local) can help people to show the personal relevance of the issue and eventually increase their intentions to act in a pro-environmental manner. Therefore, a local frame could help.
The effects of climate change are relatively small for a lot of areas in the world, which results in too little personal experience to be concerned about climate change (Weber, 2006). However, when people experience the effects of climate change directly, by showing them the issue is also happening on a local level, including their own environment, this may lead to more concern (Milfont et al., 2014). For example, people could be exposed to a locally framed message about a forest fire close to where one lives due to the effects of climate change (Scannell & Gifford, 2013). Milfont et al. (2014) stated that reducing the spatial distance of climate change makes the outcome more salient, as the issue then becomes more concrete and
7 local. This in turn will lead to the motivation to engage in pro-environmental behavior
(McDonald et al., 2016).
Furthermore, as explained by the goal setting theory (Locke & Latham, 2002), setting proximal goals is relevant in order to induce motivations for action. Consistent with the study of Latham and Seijts (1999), proximal goals result in more behavioral intentions as these psychologically close outcomes give people the feeling that they can accomplish a particular task. For example, buying biological food in a local supermarket is a relatively easy task compared to combating global climate change. Therefore, presenting climate change as a local issue may lead to an increase in motivations to act in a pro-environmental manner.
In addition, assuming climate change as a global issue may lead to no
pro-environmental behavioral intentions, as people could construe the issue in more abstract terms and do not notice the tangibility of it (McDonald et al., 2016). The study of Pidgeon (2012) is consistent with this statement, indicating that climate change is mostly seen as a global issue, which give people the perception that it is not a personally threatening issue. Moreover, the media, personal or external developments may ensure that other issues (e.g., a financial crisis) are seen as more serious, which results in less attention for climate change (Pidgeon, 2012).
In contrast to the research that suggests a local frame, other researchersrecommend a global frame, indicating that the local impacts of climate change are seen as less severe than the global impacts (Spence & Pidgeon, 2010). With this in mind, it could be that a locally framed message will result in people only paying attention to the aspects of climate change that are seen as less important. In this way, the overall perceived seriousness of global climate change will be reduced,making them less likely to engage in pro-environmental action.
In line with this, it is possible that a more global view of climate change leads pro-environmental attitudes and behaviors, as someone then could see climate change as a
8 worldwide problem that requires action now (McDonald et al., 2016). Although many people believe that they cannot do anything about global climate change as individuals, Gifford (2011) stated that if one considers the bigger picture, one might be more willing to act on the issue. In addition, Liberman and Trope (1998) explain that by presenting a global (versus local) issue, the desirability of the outcome outweighs the difficulty of achieving it. In other words, people will aim attention at the benefits of the global outcome rather than the daily barriers that the behavior entails. With this in mind, a globally framed issue may lead to more desired behaviors, as the focus can be changed from the feasibility to the desirability of combating global climate change.
Moreover, as locally framed environmental issues become more personal, they may remind people more of their vulnerability than globally framed consequences of climate change. According to Vess and Arndt (2008), people are less concerned about the
environmentif they are reminded of their own vulnerability concerning the consequences of climate change. This would apply to those who do not derive self-esteem from
pro-environmental behavior. So if people see themselves as bad pro-environmentalists, locally framed issues will lead to a feeling of vulnerability, which in turn will lead to disengagement from climate change. Therefore, a global frame could help in order to motivate people to act in a pro-environmental manner.
In sum, research suggests that both concrete (i.e., local) and abstract (i.e., global) messages can be beneficial for encouraging pro-environmental attitudes and behaviors (White et al., 2011). Within the context of climate change, there are studies that recommend a global frame instead of a local frame (Spence & Pidgeon, 2010). Taking this view, it could be that increasing the spatial distance of climate change would promote action towards
9 evoke pro-environmental responses (Scannell & Gifford, 2013; Singh et al., 2017; Wiest et al., 2015; Shih & Lin, 2017).
In short, there is contradicted work within this research area. One of the goals of this study is to examine the effect of locally versus globally framed messages on
pro-environmental attitudes and intentions. The question remains if there is some explanation for the contradicted results about the effectiveness of a local versus global message on the attitudes and behavioral intentions that people have about climate change. Therefore, the difference between a local or global consumer identity is taking into account as a possible moderator. In the next section, the moderating role of consumer identity is further explained.
Consumer identity
Consumer identity can be divided into a local versus global identity (Tu, Khare & Zhang, 2012). People with a local identity feel connected to their local area and community and people with a global identity feel connected to the whole world and the global
community. It is expected that the effects of local versus global message framing on pro-environmental responses are moderated by whether someone has a local or global consumer identity. If a locally framed message is presented to a consumer with a local identity, then a fit exists between the consumer identity and information framing (White et al., 2011). The same applies to a globally framed message which is presented to a person with a global identity. Lee and Aaker (2004) state that better levels of fit lead to a feeling of fluency about the message, which in turn leads to favorable responses. The fit between the message frame and the consumer identity makes the message in a green advertisement more fluent and thus more effective in influencing someone’s pro-environmental responses. Therefore, it is expected that locally (versus globally) framed messages lead to pro-environmental attitudes and intentions only when the consumer has a local (versus global) identity.
10 The moderating role of consumer identity has been related to spatial distance and pro-environmental behavioral intentions within a variety of studies (Grinstein & Riefler, 2015; Russell & Russell, 2010; Salnikova, 2019). Grinstein and Riefler (2015) first conclude that for world citizens a global frame works better in order to have a greater purchase intention for a sustainable product compared to a local frame. However, when highlighting the local identity of the participants, the locally framed message became more effective in inducing
pro-environmental behavior than the globally framed message. In other words, world citizens are especially receptive to globally framed messages about climate change, because climate change is construed at a higher level. This high-level construal is consistent with their understanding of climate change as an abstract and global issue. Highlighting the local identity of the world citizens, however, can shift their global view of climate change (high-level construal) to a local view (low-(high-level construal). So locally framed messages can be effective for world citizenswhen the message triggers a person’s local identity (Grinstein & Riefler, 2015). The study of Russell and Russell (2010) also states that consumer identity moderates the intentions to support a corporate social responsible activity that is perceived in a local versus global way.
Salnikova (2019) did not find significant support for the effect of a globally (versus locally) framed message on the attitudes and purchase intention for a sustainable product for people with a stronger global identity. Although no significant results were found concerning the effect on pro-environmental attitudes and intentions, significant differences exist between the effect of a globally (versus locally) framed message on the engagement with
pro-environmental initiatives. The study demonstrated that people with a global identity are more engaged with information about pro-environmental initiatives when the message used a global (versus local) frame.
11 Furthermore, this research includes the potential mechanism of environmental concern driving the effect of spatial distance on pro-environmental attitudes and intentions. Spence et al. (2012) stated that if people are less concerned about climate change, they are also less likely to behave in a pro-environmental manner. Therefore, environmental concern is a possible mediator between a local versus global message type and pro-environmental
responses. In the next section, the underlying mechanism of environmental concern is further explained.
Environmental concern
Environmental concern in this study is seen as the concern people have about climate change in general, together with the concerns related to the personal and social impacts of climate change, focusing on the possible effects there might be on a person or on society in general (Spence et al., 2012). It is expected that the effects of a locally versus globally framed message on pro-environmental responses are mediated by environmental concern. As
explained before, Lee and Aaker (2004) state that better levels of fit lead to a feeling of fluency about the message, which in turn leads to favorable responses. Therefore, it is assumed that a fit between a local versus global message and a local versus global identity makes the message more fluent and thus more effective in increasing environmental concern. Accordingly, it is expected that a locally framed message in combination with a consumer who has a local identity will lead to greater environmental concern, which in turn will lead to more pro-environmental attitudes and intentions. The same is expected for a globally framed message in combination with an individual who has a global identity.
Spatial distance has been related to environmental concern and intentions within a variety of studies (Spence et al., 2012; Singh et al., 2017; Whitmarsh, 2008). Spence et al. (2012) claim that a local frame was associated with greater concern, whereby environmental concern acts as a partial mediator between psychological distance and the motivation to act in
12 a pro-environmental manner, reducing the direct relationship between these variables. The paper of Singh et al. (2017) also suggests a local perspective of climate change, as this makes people more concerned about the topic, which in turn has a positive effect on the support for climate adaptation policies. However, Whitmarsh (2008) found that there is a small difference in the concern and actions towards climate change between victims of climate change and other people.
In the above sections, it is shown that both local and global messages can be beneficial for encouraging pro-environmental attitudes and behaviors. Attitudes and intentions can be seen as relevant variables in order to predict actual behavior (Ajzen, 1991), and in this case pro-environmental behavior. Within this study, it is examined to what extent a local versus global type of message influences ad and brand attitudes and purchase intention for a green product. In addition, it is important to acknowledge that the effectiveness of a local versus global frame may be due to the consumers’ local versus global identity. It is also described that environmental concern is a possible mediator in the relationship between message type and pro-environmental responses. The conceptual model is shown in figure 1. So taking all the explained variables together, the indirect effect of a locally versus globally framed advertisement on the ad and brand attitudes and purchase intention for a green product, mediated by environmental concern, is expected to be moderated by the consumers’ identity. This leads to the following hypothesis:
A fit between message type (local versus global) and consumer identity (local versus global) leads to more positive ad and brand attitudes and purchase intention for a green product, and this effect is mediated by environmental concern.
13 Figure 1. Conceptual model
Methodology
Research strategy
Within this study, an online experiment was conducted, examining the effects of a local versus global message type on ad and brand attitudes and purchase intention for a sustainable product. It was decided to do this through an experiment, because then causal relationships between variables can be established (Hox et al., 2009). To explore the role of spatial distance (local versus global) in motivating positive attitudes and purchase intentions towards sustainable products, the perceived spatial distance of an environmental problem is manipulated. The mechanisms driving the effect of spatial distance on ad responses are also explored, by including environmental concern as possible mediator and consumer identity as a possible moderator. Furthermore, by conducting an experiment, possible confounding
variables can be kept under control, which benefits the internal validity (Hox et al., 2009).
Design
Within the main experiment, a one-factorial between-subjects design was used with as independent variable the different advertising forms, respectively a message with a local versus global context, and as dependent variables ad and brand attitudes and purchase intention for a green product. Consumer identity is taken into account as a continuous,
14 measured predictor of the effectiveness of a local versus global frame, and environmental concern is included as a mediator.
The participants were assigned to one of the two conditions to discover whether there is a difference between the local and global condition with regard to the effect on ad and brand attitudes and purchase intention for green products. A between-subjects design was chosen because this ensures that the participants cannot be influenced by the stimulus material of a different condition (Hox et al., 2009).Moreover, the participants were randomly
distributed over the experimental conditions. In this way, an attempt was made to distribute the participants with the same personality traits as evenly as possible over the conditions. Consequently, the results of the study should form a realistic reflection of reality and can be generalized to the natural population (Hox et al., 2009).
Sample
The respondents were approached through email and social media to participate in this study. They received a link which led them to an online questionnaire. Several participants were asked to further spread this link via e-mail and social media to family and friends. In this way, a snowball effect would occur in order to reach as much as participants as possible (‘t Hart & Snijkers, 2009).This should guarantee the external validity of the results.
The link was shared until 155 respondents filled out the questionnaire. The
respondents needed to be 18 years or older, with regard to the ethical accountability. After checking for the incomplete completion of the questionnaire (34 people) and participants who did not agreed to the informed consent (6 people), 115 participants remained. Subsequently, some Mahalanobis analyses were used in order to check for outliers. Outliers are data scores that are different compared to the rest of the data (Pallant, 2001). These analyses will notice any strange pattern of scores. Eleven outliers were found using the Mahalanobis analyses. However, removing the data form the outliers results in a smaller sample, and removing this
15 data did not lead to meaningful differences in the results of the studies. For this reason it was decided to keepthe outliers in the main analysis. The same accounts for the participants who gave an incorrect answer to the control question (14 people), which were also included in the analysis. Therefore, 115 participants remained who were included in the main analysis with an average age of 28 years (Mage = 28.02; SD = 10.95), of which 27 were male and 88 were
female. Most of the participants (N = 65) completed a Bachelor study.
Procedure
Within the online questionnaire of the main study, the participants were first shown a factsheet with a cover story that the research is being done with the aim of gaining more insight into the extent to which people attach importance to the environment. In addition, the participants read an informed consent form that stated that their participation is entirely voluntary and that their privacy will be guaranteed at all times. If the participants did not provide informed consent, they were automatically redirected to the end of the survey. If they provided informed consent, the participants were randomly exposed to the local message or the global message.
First, the participants had to answer some statements related to the moderator consumer identity. After those statements, they had to imagine that they were browsing through a magazine and saw the advertisement. The participants were then exposed to the local or global condition for at least ten seconds in order to prevent that they would rush through the questionnaire. Immediately after seeing the advertisement, the participants could provide answers to questions about how concerned they were about the environment,
measuring the mediator. In addition, they were asked to answer some statements related to their attitude towards the ad and brand and their purchase intention for the product. A manipulation check was then performed in order to check if the participants perceived the difference between the local and global advertisement. This was followed by a control
16 question about the condition the participants were into check whether they had viewed the advertisement properlyand did not rush through the questionnaire. Furthermore, the
participants were asked how connected they feel with the city Amsterdam. Their connection with Amsterdam was measured as a possible covariate to control for in order to check if this has any influence on the relationship between the independent variable and dependent variables (Hox et al., 2009). The demographic characteristics of age, gender, and education were also answered. Finally, the participants were asked if they had any comments regarding the survey.
Stimulus material
The participants received a questionnaire with either an ad about a product that
contributes to a sustainable Amsterdam (i.e., local condition) or an ad that contains a message about a contribution to a more sustainable world (i.e., global condition). The ads were print ads that could be encountered in a magazine. Between the two ads, an attempt was made to keep all elements as constant as possible to prevent confounding variables. Both ads advertised for a sustainable product, ‘the dopper’, a reusable water bottle. On the top of the page the main message was shown together with a picture of a blurred image of nature and the product hold by a hand. At the bottom right of the picture the website link of the dopper was visible where one can buy the water bottle. On the bottom of the page there was a text that explains that the bottle is produced sustainably with the use of renewable fuels and helps to reduce CO2 emissions (Dopper, 2019; Jansen, 2015, 18 March). The logo of the dopper could
be seen at the very bottom of the page.
In the local ad, the main message on the top of the page stated “Contribute to an eco-friendly Amsterdam” (see Appendix I). The global ad showed a different text which was “Contribute to an eco-friendly world” (see Appendix II). The text at the bottom of the page was slightly adjusted to the local or global frame, indicating that by buying the dopper the
17 participant helps to move towards a sustainable Amsterdam versus towards a sustainable world.
Pre-test
To check whether the local message was perceived as more local than the global message, this was checked in advance by doing an online pre-test. Furthermore, it was checked if the advertisements were perceived as credible ads.
A within-subjects design was used with the advertising form as independent variable and perceived spatial distance was included as dependent variable. Both stimulus materials were shown to the participants (N = 16) with an average age of 25 years (Mage = 25.06; SD =
9.18), of which 3 were male and 13 were female. Counterbalancing was used in order to prevent that the participants will be influenced by the order in which they see the different stimulus materials. In this way, the order of exposure cannot influence the results of the study (Pollatsek & Well, 1995). While seeing the first ad for at least ten seconds, the participants were asked to answer some statements. After the first ad, the participants saw the second ad and also answered the same statements.
Perceived spatial distance. First, the participants were asked to answer six statements on a 7-point Likert scale that ran from 1 (Very strongly agree) to 7 (Very strongly disagree). Please note that a lower score indicates a higher agreement regarding the statements. The items are derived from the study of Spence at al. (2012), but within the present study some statements have been adjusted to generate a local and global scale. The local statements were: “The ad focusses on my local area”, “The ad pays attention to the place where I live”, and “The ad highlights an issue that is close to my local environment”. The global statements were: “The ad focusses on areas that are far away from here”, “The ad pays attention to the whole world”, and “The ad highlights an issue that affects places outside my local
18 Two reliability analyses were conducted to test whether reliable scales could be made of the three local items and three global items. For the local items, this analysis showed a Cronbach’s alpha of .84, from which can be concluded that the three local items form a reliable scale. For the global items, a Cronbach’s alpha of .53 emerged. However, when the item “The ad pays attention to the whole world” was deleted, the analysis showed a
Cronbach’s alpha of .60, incremented to a moderate reliable scale. Therefore, this item was deleted for the remainder of the pre-test.
A paired sample t-test was conducted to check whether the local condition was
perceived as more local than the global condition. A significant difference was found, t(15) = -6.86, p < .001, 95% CI = [-3.06, -1.61]. The local condition was clearly seen as more local (M = 2.10, SD = 0.69) than the global condition (M = 4.44, SD = 1.00). However, the local condition did not significantly differ from the global condition in terms of globality, t(15) = 2.06, p = .058, 95% CI = [-0.03, 1.91]. The local condition was perceived just as global (M = 4.16, SD = 1.27) as the global condition (M = 3.21, SD = 1.24). Despite that the conditions did not significantly differ from each other in terms of globality, the local condition was
perceived as more local than the global conditionand this was also a highly significant effect. Therefore, it was decided that the conditions differ enough in the degree of spatial distance to be included in the main study.
Credibility. Furthermore, in order to check if both ads were perceived credible, the participants were asked to answer three statements on a 7-point Likert scale that ran from 1 (Very strongly agree) to 7 (Very strongly disagree). The items measuring ad credibility are the same as used in the study of Cotte et al. (2005), respectively “The ad is believable”, “The ad is truthful”, and “The ad is realistic”.
From the reliability analysis a Cronbach’s alpha of .63 emerged, from which it can be concluded that a moderate reliable scale could be made of the items measuring ad credibility.
19 Finally, a paired sample t-test showed no significant difference, t(15) = -1.54, p = .144, 95% CI = [-0.45, 0.07], from which can be concluded that the local ad was perceived as credible (M = 2.83, SD = 0.70), as the global ad (M = 3.02, SD = 0.56).
Measurements
Consumer identity
First, the participants were asked to what extent they agreed with the following
statements regarding the moderator consumer identity: “My heart mostly belongs to the whole world”, “I identify that I am a global citizen”, “My heart mostly belongs to my local
community”, “I care about knowing global events”, “I identify that I am a local citizen”, “I respect my local traditions”, “I believe people should be made more aware of how connected we are to the rest of the world”, and “I care about knowing local events”. This scale was derived from Tu et al. (2012). A 7-point Likert scale that ran form 1 (Strongly disagree) to 7 (Strongly agree) was used to answer these statements.
A factor analysis showed that the eight statements resulted in multiple components. In addition, a reliability analysis indicated that the eight items could not be combined into one reliable consumer identity scale. The participants stated to have a local identity even as a global identity at the same time. People could feel connected to their local community, but the same people could also feel themselves as a world citizen. Therefore, it was decided to create a local identity scale and a global identity scale.
A factor analysis tested that the four statementsconcerning local identity had enough coherence to form a component with an eigenvalue higher than 1 (EV = 2.07, R2 = .52) and the new local identity variable is moderately reliable (Cronbach’s α = .69). The same was done for the global identity statements. The statements were sufficiently coherent to form a component with an eigenvalue higher than 1 (EV = 1.97, R2 = .49) and the created global
20 identity variable is moderately reliable (Cronbach’s α = .65). Both identity scales were then computed by the average of the four statements resulting in an index of local identity and global identity.
Environmental Concern
Subsequently, the mediator environmental concern was measured using a 7-point Likert scale that ran from 1 (Not at all concerned) to 7 (Extremely concerned). The
participants gave answers to the following three questions related to environmental concern, derived from Spence et al. (2012): “How concerned, if at all, are you about climate change?”, “Considering any potential effects of climate change which there might be on you personally, how concerned are you about climate change?”, and “Considering any potential effects of climate change there might be on society in general, how concerned are you about climate change?”. In addition, a factor analysis showed that the three questions had enough coherence to form a component with an eigenvalue higher than 1 (EV = 2.36, R2 = .79) and the new scale environmental concern is reliable (Cronbach’s α = .86).Finally, the environmental concern scale was computed by the average of the three questions.
Ad attitude
The dependent variable attitude towards the ad was measured in a similar way as Mackenzie and Lutz (1989) did. This attitude type was measured by asking the respondent about three statements on a 7-point Likert scale: “I think the ad is pleasant/ unpleasant, favourable/ unfavourable, and good/ bad”. The three characteristics of ad attitude had enough coherence to form one component with an eigenvalue higher than 1 (EV = 2.38, R2 = .79). Subsequently, the reliability analyses showed that the new scale ad attitude is reliable (Cronbach’s α = .86).
Brand attitude
21 statements used by Spears and Singh (2004), respectively “I think the brand is attractive/ unattractive, likable/ unlikable, and good/ bad”. Brand attitude was also measured by using a 7-point Likert scale. A factor analysis showed that the three characteristics of brand attitude had enough coherence to form one component with an eigenvalue higher than 1 (EV = 2.58, R2 = .86) and the new scale brand attitude is reliable (Cronbach’s α = .92). In addition, both attitude scales were then computed by the average of the three characteristics resulting in an index of ad attitude and brand attitude.
Purchase intention for a green product
The dependent variable purchase intention was measured by making use of the following statements related to the ad: “It is likely that I would buy this product”, “I am considering to buy this product”, “Presumably I would buy this product”, and “I am willing to buy this product”. These statements are related to the statements used by Dodds et al. (1991) and were answered with the use of a 7-point Likert scale. The four items had enough
coherence to form a component with an eigenvalue higher than 1 (EV = 3.36, R2 = .84) and the new scale purchase intention is reliable (Cronbach’s α = .94). Finally, the purchase intention scale was computed by the average of the four statements.
Manipulation check
The manipulation check for the present study was about perceiving the difference between the local and global advertisement. The manipulation check question was “Do you think the advertisement is focusing more on a local or global contribution against climate change?”. The respondents answered this question with a 7-point Likert scale that ran form 1 (Local) to 7 (Global).
Control question
The last question was about which advertisement the participants saw. This was done by using the question "Indicate what the title of the advertisement was", with the following
22 answer categories: "Contribute to an eco-friendly Amsterdam", "Contribute to an eco-friendly world", and "I don't know". As explained earlier, it was decided to keep the data from the participants who gave an incorrect answer to the control question in order to keep a larger sample and there were almost zero differences in the results of the studies.
Demographic characteristics
Gender was measured using a multiple choice question consisting the options male/ female/ different. Age was measured with an open question where the participants themselves could note their age. Furthermore, the level of education was asked with a multiple choice question “What is the highest degree or level of school you have completed?” Finally, the participants were asked if they feel connected to Amsterdam, answering this question with a 7-point Likert scale that ran form 1 (Strongly disagree) to 7 (Strongly agree).
Results
Randomization check
To check whether the respondents are randomly divided over the different conditions, respectively the local and global condition, three randomization checks were carried out. First of all, a chi-squared analysis showed that there is no difference between the conditions
regarding sex ratio, χ2 (1) = 0.074, p = .786. Subsequently, an independent samples t-test demonstrated that there was no difference in age of the participants between the local condition (M = 29.14, SD = 12.31) and global condition (M = 26.88, SD = 9.36), t(113) = 1.11, p = .270, 95% CI = [-1.78; 6.30]. Finally, an independent samples t-test showed that there was no difference in education of the participants between the local condition (M = 3.24, SD = 0.96) and global condition (M = 3.04, SD = 0.76), t(113) = 1.28, p = .204, 95% CI = [-0.11; 0.53]. Based on the randomization checks, it can be concluded that randomization was successful and that there are no disturbing variables in the study.
23 Manipulation check
Using an independent samples t-test, it was checked whether the local and global condition differed sufficiently from each other in terms of the degree of perceived spatial distance. First, the Levene's Test for equal variances was significant (Levene's F (1, 113) = 28.33, p < .001), indicating that there is a difference between the variances in the population, and therefore a more stringent p-value will be adhered too (p < .001). In line with the
expectations, a statistically significant difference emerged from the independent samples t-test, t(97.48) = -5.88, p < .001, 95% CI = [-2.66; -1.32], showing that the local condition was perceived as less global (M = 3.48, SD = 2.16) than the global condition (M = 5.47, SD = 1.39). Due to these significant differences with regard to spatial distance, it can be concluded that the manipulation was successful and that the results can be explained on the basis of the manipulation.
Moderated mediation
It was expected that once a fit arises between message type (local versus global) and consumer identity (local versus global), this would lead to more positive ad and brand attitudes and purchase intention for a green product, mediated by environmental concern. In order to analyze this hypothesis in SPSS, the PROCESS macro developed by Hayes (2013; Model 7; 5000 bootstraps, 95% CI) was used. Model 7 of the PROCESS macro for SPSS provides insights in the conditional indirect effect at different values of the suggested
moderator. A total of six analyses were conducted, as there are three dependent variables and both the local and the global identity scales were included as moderators.
The variable that indicates if the participants feel a connection with Amsterdam was included as a covariate in order to check if this has any influence on the results.However, the covariate did not show any significant results in all of the six models that there were run.
24 Therefore, it can be concluded that this variable had no significant influence on the
relationship between the independent and dependent variables (Hox et al., 2009).
Local identity as a moderator
First, the model with local identity as a moderator was run three times with 1.) ad attitude, 2.) brand attitude, and 3.) purchase intention as dependent variables. The results of the effects of the conceptual model with local identity as a moderator are shown in Table 1 (see Appendix III). In line with the expectations, the interaction between message type and local identity had a significant influence on environmental concern, b = -0.85, SE = 0.22, p < .001, 95% CI = [-1.29; -0.42]. Accordingly, a fit between a local message type and a
consumer who has a local identity led to greater environmental concern. There was also a significant main effect of a local versus global message type on environmental concern, b = 4.01, SE = 1.14, p = .001, 95% CI = [1.75; 6.27], showing that a local message led to more environmental concern compared to a global message. There was no main effect of local identity on environmental concern, b = 0.22, SE = 0.15, p = .150, 95% CI = [-0.08; 0.53]. In addition, a greater environmental concern 1.) led to more positive ad attitudes, b = 0.19, SE = 0.09, p = .039, 95% CI = [0.01; 0.37], 2.) had no effect on brand attitudes, b = 0.15, SE = 0.09, p = .112, 95% CI = [-0.04; 0.33], but 3.) did lead to a higher purchase intention, b = 0.41, SE = 0.12, p = .001, 95% CI = [0.17; 0.65].Lastly, there were no significant differences in the direct effects of a local versus global message type on 1.) ad attitude, b = 0.19, SE = 0.20, p = .330, 95% CI = [-0.20; 0.58], 2.) brand attitude, b = 0.06, SE = 0.20, p = .760, 95% CI = [-0.34; 0.46], or 3.) purchase intention for a green product, b = 0.10, SE = 0.26, p = .703, 95% CI = [-0.42; 0.62].
Finally, the results of the index of the moderated mediation with local identity included as a moderator are discussed, expecting that once a fit arises between a local
25 purchase intention for a green product, mediated by environmental concern. The outcomes show no evidence of conditional indirect effectconcerning the dependent variables 1.) ad attitude, Index = 0.16, SE = 0.12, 95% CI = [0.45; 0.02], nor on 2.) brand attitude, Index = -0.13, SE = 0.10, 95% CI = [-0.38; 0.03], as the bootstrapped confidence intervals include zero. However, the outcomes of the index of moderated mediation with 3.) purchase intention as the dependent variable did show evidence of a conditional indirect effect, Index = -0.35, SE = 0.19, 95% CI = [-0.77; -0.04]. Furthermore, the results of the indirect effects showed a significant indirect effect on 3.) purchase intention at moderate levels of local identity, b = -0.19, SE = 0.12, 95% CI = [-0.46; -0.01], and high levels of local identity, b = -0.45, SE = 0.24, 95% CI = [-0.98; -0.06]. Therefore, there is a significant indirect effect of message type on the purchase intention for a green product, via environmental concern, at moderate and high levels of local identity.
In sum, the association between a local (versus global) message and purchase intention for a green product, mediated by environmental concern, significantly increases when a consumer has a local identity. Consequently, this hypothesis is supported. However, this only holds for the dependent variable purchase intention and not for the dependent variables ad and brand attitudes.
Global identity as a moderator
The same was done for the moderator global identity. The model with global identity as a moderator was also run three times with 1.) ad attitude, 2.) brand attitude, and 3.) purchase intention as dependent variables. The results of the effects of the conceptual model with global identity as a moderator are shown in Table 2 (see Appendix IV). The interaction between message type and global identity did not had a significant influence on
environmental concern, b = 0.24, SE = 0.21, p = .267, 95% CI = [-0.18; 0.65]. There was also no main effect of a local versus global message type on environmental concern, b = -1.54, SE
26 = 1.09, p = .161, 95% CI = [-3.71; 0.62]. Notably, there was a significant main effect of global identity on environmental concern, b = 0.35, SE = 0.15, p = .020, 95% CI = [0.06; 0.64], showing that a strong global identity led to more environmental concern compared to a weak global identity. Finally, as for the direct effects of environmental concern on the
dependent variables, the same results appeared as when local identity was included as a moderator. The same accounts for the direct effects of a local versus global message type on the dependent variables.
Lastly, the results of the index of the moderated mediation with global identity included as a moderator are discussed. The outcomes of the index of moderated mediation show no evidence of conditional indirect effect for the dependent variables 1.) ad attitude, Index = 0.04, SE = 0.07, 95% CI = [-0.04; 0.25], 2.) brand attitude, Index = 0.04, SE = 0.06, 95% CI = 0.03; 0.18], nor on 3.) purchase intention, Index = 0.10, SE = 0.13, 95% CI = [-0.08; 0.42].
In sum, the association between a global (versus local) message and 1.) ad attitude, 2.) brand attitude, and 3.) purchase intention for a green product, mediated by environmental concern, did not significantly increases when a consumer has a global identity. Consequently, this hypothesis is not supported.
Conclusion and discussion
In the present study it was investigated to what extent a fit between local versus global message type and local versus global identity will lead to more positive ad and brand attitudes and purchase intention for a green product. In addition, the mediating role of environmental concern was included in the study. This research was conducted among 115 respondents. Based on the resultswith local identity included as a moderator, it can be concluded that the association between a local (versus global) message and purchase intention for a green
27 product significantly increases when a consumer has a local identity, mediated by
environmental concern. Consequently, this hypothesis is supported. It is important to
acknowledge that this only holds for the dependent variable purchase intention and not for the dependent variables ad and brand attitudes.
With regard to the direct effects, it can be concluded that there is a significant main effect of a local versus global message type on environmental concern. A local message led to more environmental concern compared to a global message. These outcomes correspond to earlier research that has been done about the influence of psychological distance on
environmental concern (Spence et al., 2012; Singh et al., 2017; Howansky, 2015).Moreover, the interaction between a local versus global message type and local identity had a significant influence on environmental concern, indicating that a fit between a locally framed message and a consumer who has a local identity will lead to greater environmental concern. This result is in line with the studies of Grinstein and Riefler (2015), and Russell and Russell (2010), who also investigated the moderating role of consumer identity related to spatial distance and pro-environmental behavioral intentions. In addition, a greater environmental concern led to more positive ad attitudes and a higher purchase intention for a green product, but it had no effect on brand attitudes.
Furthermore, when comparing the outcomes with the results when global identity was included as a moderator, it can be concluded that local identity is more important to take into account when creating a fit between message type and consumer identity compared to global identity. If a global message type was presented, this did not lead to more positive ad and brand attitudes and purchase intention for a green product, mediated by environmental concern, when the consumer had a global identity.
28 Theoretical implications
This study contributes to the scientific literature about the effects of a locally versus globally framed message on pro-environmental responses. The study included the mediating role of environmental concern, and explored to what extent a fit between a locally versus globally framed message and consumer identity leads to higher environmental concerns, at which consumer identity was included as a moderator.
This research showed that local identity is a more important factor than global identity when creating a fit between message type and consumer identity. In the study of Grinstein and Riefler (2015) they explained that world citizens (i.e., people with a global identity) are more aware of global topics, and therefore may be more aware of the need for pro-environmental behavior. Similarly, other research indicated that people who value the well-being of the broader community (i.e., self-transcendence) are more willing to engage in pro-environmental behavior compared to people who value personal gain (i.e., self-enhancement; Corner et al., 2014). However, Grinstein and Riefler (2015) showed that world citizensare responsive to both locally and globally framed messages, indicating that a local message can trigger someone’s local identity. It is interesting to find out if consumers with a local identity could also be made receptive to globally relevant environmental issues by triggering their global identity. The study of Schoenefeld and McCauley (2015) explored the extent to which self-transcendence and self-enhancement values might interact with a local versus global message type to influence pro-environmental behavior, expecting that people with self-enhancing values would be more susceptible for a local message. However, they found that people with self-enhancing values who were exposed to a locally framed message were less motivated to act in a pro-environmental behavior manner compared to those who received a globally framed message, indicating a reactance effect.
29 Furthermore, it is important to acknowledge that the participants of this study
indicated to have a local identity even as a global identity at the same time. From this it can be concluded that the concepts do not exclude each other and could not form one scale. This suggests that people could have a strong local identity versus a strong global identity, but they could also have both a local and global identity or score low on both identity types. This finding corresponds to the research of Arnett (2002), who stated that people could develop a bicultural identity, indicating that people could feel themselves as a world citizen while still believing in the value of their local culture. However, for some people it is difficult to adapt to both local and global cultures which can lead to an experience of having neither a local or global identity (Arnett, 2002). More research is needed to examine the relationship between the concepts local versus global identity in order to explain the differences of the effect on pro-environmental behavioral intentions after presenting consumers with a locally versus globally framed message.
Limitations and future research
A limitation of the present study is that it only takes into account ad and brand attitudes and purchase intention for a green product as dependent variables. The study of Spence and Pidgeon (2010) indicated that a global frame works better to perceive climate change as more severe, whereas attitudes towards climate change mitigation were more positive when the participants were exposed to a local frame. Another paper focused on the engagement in climate change issues, where they stated that a local frame was more effective in order to increase engagement among the climate change topic (Scannell & Gifford, 2013). Other studies payed attention to behavioral responses after presenting their respondents with a local or global message. Examples are the support for climate adaptation policies (Singh et al., 2017), the intention to change non-sustainable behavior within a month (Wiest et al., 2015), and the willingness to engage in several pro-environmental behaviors (Shih & Lin, 2017;
30 Schoenefeld & McCauley, 2015). The studies of Singh et al. (2017) and Shih and Lin (2017) stated that a local frame was more effective in order to enhance pro-environmental behavioral responses, whereas other researchers also recommended a global frame in some cases (Wiest et al., 2015; Schoenefeld & McCauley, 2015). More research is needed, as there are
differences in recommending a local versus global frame in order to enhance various pro-environmental responses. Follow-up research could provide answers to what extent a fit between message type and consumer identity will lead to other pro-environmental responses compared to ad and brand attitudes and purchase intention for a green product, mediated by environmental concern, since little is known about this research area.
Relatedly, a limitation of this research is that it focuses on the intention to buy a green product instead of actual buying behavior. A consequence of this might be that people will give a socially responsible answer rather than give an honest answer to the question (Tourangeau & Yan, 2007). It is interesting to do a follow-up study about actual purchase behavior, for example, by comparing the selling figures of a green product in different stores where either the local ad or the global ad is shown. Perhaps it is also possible to look at other (i.e., less popular) brands. The dopper water bottle is a well-known product under Dutch students. It could be that the results show different outcomes if an unknown
pro-environmental brand is chosen. One might be more willing to buy a well-known brand rather than an unknown brand, as brand recognition can have a positive impact on attitudes and purchase intentions towards the brand (Laroche et al., 1996).
Practical implications
This research corresponds to the increasing concern for the environment. Every day it becomes more important that one acts in an environmentally conscious way (Karl &
Trenberth, 2003). The use of locally versus globally framed message types could play a significant role in encouraging pro-environmental behavior. This study showed that a fit
31 between a local message and a consumer with a local identity will lead to a higher purchase intention for a green product, mediated by environmental concern. No effects were found for people who have a global identity, indicating that people should have a local identity in order to be receptive for a message with the aim of enhancing pro-environmental responses.
Therefore, in order to implement this study, people need to have a local identity. Advertisers and campaigners can use this research to gain knowledge about how they can reach
consumers most effectively, and therefore encourage pro-environmental behavior.
References
‘t Hart, H. & Snijkers, G. (2009). De enquête. In H. Boeije, H. ‘t Hart & J. Hox (red.), Onderzoeksmethoden (pp. 208-245). Den Haag: Boom Lemma uitgevers.
Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50, pp. 179–211.
Arnett, J.J. (2002). The psychology of globalization. American Psychologist, 57(10), pp. 774- 783.
Corner, A., Markowitz, E. & Pidgeon, N. (2014). Public engagement with climate change: the role of human values. WIREs Climate Change, 5(3), pp. 411-422.
Cotte, J., Coulter, R.A. & Moore, M. (2005). Enhancing or disrupting guilt: The role of ad credibility and perceived manipulative intent. Journal of Business Research, 58, pp. 361-368.
Dodds, W.B., Monroe, K.B. & Grewal, D. (1991). Effects of price, brand, and store
information on buyers' product evaluations. Journal of Marketing Research, 28(3), pp. 307-319.
32 Dopper. (2019). Behind the bottle. At Dopper, sustainability is in our drinking water. Dopper. Retrieved from https://dopper.com/about/
Grinstein, A. & Riefler, P. (2015). Citizens of the (green) world? Cosmopolitan orientation and sustainability. Journal of International Business Studies, 46(6), pp. 694–714.
Gifford, R. (2011). The dragons of inaction: psychological barriers that limit climate change mitigation and adaptation. The American Psychologist, 66(4), pp. 290–302.
Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. Guilford Press.
Howansky, K. (2015). Getting warmer: the optimal distance for motivating climate change mitigation. The State University of New Jersey.
Hox, J., De Goede, M. & Boeije, H. (2009). Het experiment. In H. Boeije, H. ‘t Hart & J. Hox (red.), Onderzoeksmethoden (pp. 164-207). Den Haag: Boom Lemma uitgevers.
Jansen, H. (2015, 18 March). De flesjesboodschap van Dopper. One World. Retrieved from https://www.oneworld.nl/duurzaamheid/de-flesjesboodschap-van-dopper/
Laroche, M., Kim, C. & Zhou, L. (1996). Brand familiarity and confidence as determinants of purchase intention: An empirical test in a multiple brand context. Journal of Business Research, 37(2), pp. 115-120.
Latham, G.P. & Seijts, G.H. (1999). The effects of proximal and distal goals on performance on a moderately complex task. Journal of Organizational Behavior, 20(4), pp. 421–
429.
Lee, A.Y. & Aaker, J.L. (2004). Bringing the frame into focus: The influence of regulatory fit on processing fluency and persuasion. Journal of Personality and Social Psychology, 86, pp. 205–18.
33 Liberman, N., & Trope, Y. (1998). The role of feasibility and desirability considerations in near and distant future decisions: A test of temporal construal theory. Journal of Personality and Social Psychology, 75(1), pp. 5–18.
Locke, E.A. & Latham, G.P. (2002). Building a practically useful theory of goal setting and task motivation. A 35-year odyssey. The American Psychologist, 57(9), pp. 705– 717.
Mackenzie, S.B., and R.J. Lutz. (1989). An empirical examination of the structural
antecedents of attitude toward the ad in an advertising pretesting context. Journal of Marketing, 53(2), pp. 48 - 65.
McDonald, R.I., Chai, H.Y. & Newell, B.R. (2016). Personal experience and the ‘psychological distance’ of climate change: an integrative review. Journal of Environmental Psychology, 44, pp. 109–118.
Milfont, T.L., Evans, L., Sibley, C.G., Ries, J. & Cunningham, A. (2014). Proximity to coast is linked to climate change belief. PLoS ONE, 9(7), pp. 1-8.
Pallant, J. (2001). SPSS survival manual. Buckingham, UK: Open University Press.
Pidgeon, N. (2012). Public understanding of, and attitudes to, climate change: UK and international perspectives and policy. Climate Policy, 12(1), pp. 85–106.
Pollatsek, A. & Well, A.D. (1995). On the use of counterbalanced designs in cognitive research: A suggestion for a better and more powerful analysis. Journal of
Experimental Psychology: Learning, Memory, and Cognition, 21(3), pp. 785-794.
Russell, D.W. & Russell, C.A. (2010). Here or There? Consumer Reactions to Corporate Social Responsibility Initiatives: Egocentric Tendencies and Their Moderators. Marketing Letters, 21(1), pp. 65–81.
34 Salnikova, E. (2019). Think Global, Act Local: The Role of Global Consumer Identity in Food- and Sustainability-Related Consumer Behavior. Aarhus University Department of Management.
Scannell, L. & Gifford, R. (2013). Personally relevant climate change: The role of place attachment and local versus global message framing in engagement. Environment & Behavior, 45, pp. 60–85.
Schoenefeld, J. & McCauley, M. (2015). Local is not always better: the impact of climate information on values, behavior and policy support, Journal of Environmental Studies and Sciences, pp. 1–9.
Shih, T.J. & Lin, C.Y. (2017). Developing communication strategies for mitigating actions against global warming: linking framing and a dual processing model. Environmental Communication, 11(6), pp 840–858.
Singh, A.S., Zwickle, A., Bruskotter, J.T. & Wilson, R. (2017). The perceived psychological distance of climate change impacts and its influence on support for adaptation policy. Environmental Science & Policy, 73, pp. 93-99.
Simpson, B. & Radford, S. (2013). Situational variables and sustainability in multi-attribute decision-making. European Journal of Marketing, 48(5/6), pp. 1046–1069.
Smith, S. M. & Petty, R. E. (1996). Message framing and persuasion: A message processing analysis. Personality and Social Psychology Bulletin, 22(3), pp. 257–268.
Spears, N. & Singh, S. (2004). Measuring Attitude toward the Brand and Purchase Intentions. Journal of Current Issues and Research in Advertising, 26(2), pp. 53-66.
35 Spence, A. & Pidgeon, N. (2010). Framing and communicating climate change: the effects of distance and outcome frame manipulations. Global Environmental Change, 20, pp. 656-667.
Spence, A., Poortinga, W. & Pidgeon, N. (2012). The psychological distance of climate change. Risk Analysis, 32, pp. 957-972.
Tourangeau, R. & Yan, T. (2007). Sensitive Questions in Surveys. Psychological Bulletin, 133(5), 859-83.
Trope, Y. & Liberman, N. (2010). Construal-level theory of psychological distance, Psychological Review, 117, pp. 440-463.
Tu, L., Khare, A. & Zhang, Y. (2012). A short 8-item scale for measuring consumers’ local global identity. International Journal of Research in Marketing, 29(1), pp. 35-42.
Van Dorland, R. (2014). Klimaatverandering zet door. Nederlands Tijdschrift voor Natuurkunde, 80(2), pp. 48-51.
Vess, M. & Arndt, J. (2008). The nature of death and the death of nature: The impact of mortality salience on environmental concern. Journal of Research in Personality, 42(5), pp. 1376–1380.
Weber, E.U. (2006). Experience-based and description-based perceptions of long-term risk: Why global warming does not scare us (yet). Climatic Change, 77(1-2), pp. 103– 120.
Wei, C.F., Chiang, C.T., Kou, T.C. & Lee, B.C. (2017). Toward sustainable livelihoods: investigating the drivers of purchase behavior for green products. Business Strategy and the Environment, 26, pp. 626–639.
36 White, K., MacDonnell, R. & Dahl, D. (2011). It’s the mind-set that matters: The role of construal level and message framing in influencing consumer efficacy and conservation behaviors. Journal of Marketing Research, 48(3), pp. 472–485.
Whitmarsh, L. (2008).Are flood victims more concerned about climate change than other people? The role of direct experience in risk perception and behavioural response. Journal of Risk Research, 11(3), pp. 351−374.
Wiest, S.L., Raymond, L. & Clawson, R.A. (2015). Framing, partisan predispositions, and public opinion on climate change. Global Environmental Change, 31, pp. 187–198.
37 Appendix I
38 Appendix II
39 Appendix III
Table 1. Effects of conceptual model with local identity as a moderator Bootstrap 95% CI
Direct effects Coefficient SE p Lower CI Upper CI
Message type Environmental concern 4.01 1.14 .001 1.75 6.27 Local identity Environmental concern 0.22 0.15 .150 -0.08 0.53 Message type * Local identity
Environmental concern
-0.85 0.22 < .001 -1.29 -0.42
Environmental concern Ad attitude 0.19 0.09 .039 0.01 0.37 Environmental concern Brand attitude 0.15 0.09 .112 -0.04 0.33 Environmental concern Purchase intention 0.41 0.12 .001 0.17 0.65
Message type Ad attitude 0.19 0.20 .330 -0.20 0.58
Message type Brand attitude 0.06 0.20 .760 -0.34 0.46 Message type Purchase intention 0.10 0.26 .703 -0.42 0.62
Indirect effects Effect size BootSE Lower CI Upper CI
Message type Environmental concern Ad attitude (low level Local identity)
0.07 0.08 -0.03 0.28
Message type Environmental concern Ad attitude (moderate level Local identity)
-0.09 0.07 -0.25 0.01
Message type Environmental concern Ad attitude (high level Local identity)
-0.21 0.15 -0.56 0.02
Message type Environmental concern Brand attitude (low level Local identity)
0.06 0.07 -0.03 0.23
Message type Environmental concern Brand attitude (moderate level Local identity)
-0.07 0.06 -0.21 0.02
40 Brand attitude (high level Local identity)
Message type Environmental concern Purchase intention (low level Local identity)
0.15 0.14 -0.06 0.50
Message type Environmental concern Purchase intention (moderate level Local identity)
-0.19 0.12 -0.46 -0.01
Message type Environmental concern Purchase intention (high level Local identity)
-0.45 0.24 -0.98 -0.06
Moderated mediation Index BootSE Lower CI Upper CI
Ad attitude -0.16 0.12 -0.45 0.02
Brand attitude -0.13 0.10 -0.38 0.03
Purchase intention -0.35 0.19 -0.77 -0.04
Note. Bootstrap resampling = 5000. CI = confidence interval.
Appendix IV
Table 2. Effects of conceptual model with global identity as a moderator Bootstrap 95% CI
Direct effects Coefficient SE p Lower CI Upper CI
Message type Environmental concern -1.54 1.09 .161 -3.71 0.62 Global identity Environmental concern 0.35 0.15 .020 0.06 0.64 Message type * Global identity
Environmental concern
0.24 0.21 .267 -0.18 0.65
Environmental concern Ad attitude 0.19 0.09 .039 0.01 0.37 Environmental concern Brand attitude 0.15 0.09 .112 -0.04 0.33 Environmental concern Purchase intention 0.41 0.12 .001 0.17 0.65
Message type Ad attitude 0.19 0.20 .330 -0.20 0.58
Message type Brand attitude 0.06 0.20 .760 -0.34 0.46 Message type Purchase intention 0.10 0.26 .703 -0.42 0.62
41 Message type Environmental concern
Ad attitude (low level Global identity)
-0.11 0.11 -0.39 0.02
Message type Environmental concern Ad attitude (moderate level Global identity)
-0.06 0.05 -0.18 0.02
Message type Environmental concern Ad attitude (high level Global identity)
-0.02 0.06 -0.15 0.13
Message type Environmental concern Brand attitude (low level Global identity)
-0.08 0.00 -0.32 0.02
Message type Environmental concern Brand attitude (moderate level Global identity)
-0.04 0.05 -0.16 0.02
Message type Environmental concern Brand attitude (high level Global identity)
-0.02 0.05 -0.13 0.09
Message type Environmental concern Purchase intention (low level Global identity)
-0.23 0.18 -0.66 0.01
Message type Environmental concern Purchase intention (moderate level Global identity)
-0.12 0.09 -0.34 0.03
Message type Environmental concern Purchase intention (high level Global identity)
-0.05 0.13 -0.29 0.24
Moderated mediation Index BootSE Lower Upper
Ad attitude 0.04 0.07 -0.04 0.25
Brand attitude 0.04 0.06 -0.03 0.18
Purchase intention 0.10 0.13 -0.08 0.42
