Targeting Green Advertising: What Consequences
of Non-Green Consumption Should Be Displayed?
Master’s Thesis, Master’s Programme Communication Science Graduate School of Communication
Name: Lea Ellen Schwerdt Student Number: 11813318 Supervisor: Sandra Zwier Date: July 13, 2018
1
Abstract
Our modern way of life has a threatening impact on the planet and this is making it of crucial importance to further understand how consumers can be encouraged to willingly purchase more environment-friendly products. The present study investigated the role of green advertising messages, and particularly: a) how temporal distance and points of reference in green advertising messages impact their effectiveness, and b) if this is moderated by people’s level of environmental consciousness, long-term orientation and self-construals. An online experiment with 143 participants was conducted. The results showed that depicting immediate consequences was more successful in eliciting green purchase intentions than future
consequences, while a focus on consequences for the environment and the self were equally effective. The measured consumer traits were not found to moderate these effects. However, a high level of environmental consciousness, valuation of traditions and meta-personal self-construal were all found to be related to a higher purchase intention for green products. To strengthen generalizability, future research might repeat the current study with different product categories and a broader sample, in a laboratory setting and investigate effects on actual purchase behavior.
Key words: green consumption, green advertising, point of reference, temporal distance,
environmental consciousness, long-term orientation, independent self-construal, meta-personal self-construal, purchase intention
Introduction
Our modern way of life has a threatening impact on the planet since it is bringing about climate change and interrelatedly the disappearance of entire landscapes and animal species (Ripple et al., 2017). For that reason, the need to change our lifestyle as humans is increasingly urgent. Although this knowledge and the interest in and offer of environmentally sound products is rising, people often still prefer non-sustainable over sustainable products
2
(Chang, 2011; World Business Council for Sustainable Development, 2008). This is making it of crucial importance to further understand how consumers can be stimulated to willingly purchase more environment-friendly products, for instance by maximizing the effectiveness of advertising for these products. In the following, advertising for environment-friendly products will be called green advertising.
Generally, one way to position a product or brand in advertising is by distinguishing it from competitors. This entails pointing out the advantages of the product in comparison to competitors’ products so that consumers come to prefer it on the basis of those characteristics (Belch, & Belch, 2015; Schmuck, Matthes, Naderer, & Beaufort, 2018). For green products, the advantages in comparison to non-green products can either refer to the consumer
him/herself or the environment, which is called the point of reference (PR; Langer, 2013). For instance, self-related benefits may concern one’s health or financial gains, whereas
environmental benefits may be that the products do not contribute to global warming or threat to animal species. Furthermore, there can be differences in the time distance between the purchase of a green product and the occurrence of the beneficial outcome or likewise the purchase and the negative outcome of a non-green product, which is called temporal distance (TD; Chang, Zhan, & Xie, 2015). The consequences of (non-) green consumption may occur right now, thus at the time of the purchase and use of a product, or occur later on in the future, for instance in several years. In order to target consumers with green advertising messages that are appealing to them and thereby eventually stimulate purchase intentions (PI; Belch, & Belch, 2015), it is important to know which benefits of green products in comparison to non-green competitors appeal to them the most. To the knowledge of the researcher, it is not known yet how the above-mentioned factors, PR and TD, of green advertising messages affect consumers’ PI’s. This is an important insight since these two factors can be considered basic components of green advertising messages (Langer, 2013; Chang et al., 2015).
3
advertising message is equally effective in eliciting PI in all consumers. Marketers often target their marketing efforts, including advertising messages, to clusters of consumers with distinct characteristics in order to persuade them best that their products can fulfill their specific needs. For instance, target groups can be segmented on the basis of consumers’ personality, called “psychographic segmentation”, or specific needs and wants, called “benefit segmentation” (Belch, & Belch, 2015, pp. 47, 51, 53). Thus, it is generally beneficial to adjust advertising messages to clusters of consumers with distinct, relevant characteristics.
Environmental consciousness, the extent to which people are pro-environmentally
oriented, is one such trait that people possess to different degrees and that has been shown to influence how effective different green advertising messages are (Chang et al., 2015; Chan, & Lau, 2004). Furthermore, people’s level of long-term orientation and their self-construals are relatively steady personality traits that have partly been shown to have an impact on the effectiveness of different green advertising messages (Bearden, Money, & Nevins, 2006; Chen, Lee, & Huan, 2015; Arnocky, Stroink, & DeCicco, 2007). The degree of long-term orientation describes to what extent people value tradition and future instead of immediate outcomes (Bearden et al., 2006). People with a high long-term orientation may therefore value future benefits of green consumption more than short-term benefits. People’s independent
self-construal in turn has to do with the degree to which they distinguish themselves from
others as individuals and the meta-personal self-construal describes in how far people feel connected to all living beings (Chen et al., 2015). Those with a high independent
self-construal may therefore strive for the achievement of their individual goals irrespective of any benefit or harm to the environment, whereas those with a high meta-personal self-construal value harmony with other living beings and thus do not want to harm the environment or animals (Arnocky et al., 2007; Chen et al., 2015; Cross, Hardin, Gercek-Swing, 2011). It is still unclear in how far these individual characteristics influence the effect of green
4
advertising messages to consumers can increase their impact (Belch, & Belch, 2015), scrutinizing this issue further will provide important insights for making green advertising messages successful in eliciting PI across different types of consumers. Furthermore, it can contribute to scientific knowledge about the relation between different personality traits and motives for green consumption. Therefore, the current study sets out to answer the following research question:
RQ: To what extent do different levels of temporal distance and points of reference in
green advertising messages vary in their effectiveness to stimulate purchase intention and in how far is this moderated by consumers’ level of environmental consciousness, long-term orientation and type of self-construal?
Theoretical Background Purchase Intention
The current study will investigate how successful different green advertising messages are for consumers with different characteristics. The success of advertising messages will be measured by their purchase intention (PI), which will be explained here. PI is defined as an individual’s tendency to purchase a brand’s products (Spears, & Singh, 2004). More precisely, Spears and Singh (2004) define PI as follows: “Purchase intentions are an
individual’s conscious plan to make an effort to purchase a brand” (p. 56). Generally, it has been acknowledged that about 28% of actual behavior is explained by preceding behavioral intentions (Sheeran, 2002). Although PI thus does not perfectly predict green purchase behavior, it is one of several important determinants (Carrington, Neville, & Whitwell, 2010; Sheeran, 2002). In the present study, PI is therefore chosen as a crucial outcome of successful advertising messages for environmentally sound products.
5
When the benefits of green products are highlighted in advertising messages, the point of reference (PR) and temporal distance (TD) used are central features. In the following two sections, the existing literature about the effectiveness of these message characteristics and the subsequent hypotheses will be discussed.
Point of reference. Although the main objective of green products is their
environmental soundness, their benefits or, equally, the disadvantages of analogous non-green products cannot only concern the environment but also the consumer him-/herself or other individuals or groups of people (Segev, Fernandes, & Wang, 2015; Langer, 2013). In the current study, whom or what the consequences in green advertising messages refer to is defined as point of reference (PR).
The results of existing research about different PR’s in green advertising messages are diverse, partly contradicting each other. For instance, in a study by Kareklas, Carlson and Muehling (2014), advertisement that included both egoistic, thus self-related, and altruistic, including societal and environmental benefits, predicted attitudes and PI’s in the context of organic food choices. Advertising messages featuring either only altruistic benefits or both egoistic and altruistic benefits of green food consumption appeared to be more effective than those displaying only egoistic advantages. Also, in a study by Tih, Thye, Ansary and Ahmed (2016), green advertising messages for an eco-friendly soap displaying environmental benefits increased PI more than those displaying benefits for the consumer. In line with that, Nordlund and Garvill (2002) found in a study that concern for the environment is a more important motive for environmentally friendly behavior than egoistic motives. However, opposite results were obtained in another study, showing that advertising messages for a green washing machine referring to benefits for the self were more effective in increasing brand attitudes and PI than those featuring environmental benefits (Segev et al., 2015). Likewise, there is existing research showing that egoistic motives and/or concerns for other people better predict the purchase of organic food (Magnusson, Arvola, Hursti, Aberg, & Sjödén, 2003), refraining
6
from driving a car and thereby reducing air pollution (Corbett, 2005), reducing carbon dioxide emissions (Howell, 2012) and the willingness to pay for wildlife (Ojea, & Loureira, 2007) than environmental concerns. Due to the contradictory results found in previous studies, it is thus not clear yet whether the self or the environment as PR in green advertising messages is more effective in eliciting PI. The current study is therefore supposed to support either one of the previous results and thereby answer the question:
Q1: To what extent do different points of reference of green advertising messages vary in their effectiveness to stimulate purchase intention?
Temporal distance. The featured advantages of green purchasing behavior or similarly the disadvantages of non-green purchasing behavior can also occur at different points of time. More precisely, some consequences of (non-) green consumption occur now and others more in the future (Miniero, Codini, Bonera, Corvi, & Betroli, 2014; Chang et al., 2015; Chen, & Chiu, 2016). In the current research, the temporal distance between the (non-) green purchasing behavior and the consequences mentioned in a green advertising message is called temporal distance (TD). Consequences occurring right now are referred to as having a low TD, whereas those occurring later on in the future are referred to as having a high TD. The Construal Level Theory by Trope and Liberman (2010) argues that people perceive events differently, depending on their TD’s. The more distant events are, the more they are represented in abstract features. Thus, temporally close events are generally thought of more concretely than future events. For instance, future consequences of non-green consumption may be thought of in little detail (e.g. there is less diversity in the ecosystem) while proximal consequences may be construed with many details concerning for example their course or exact effects (e.g. plants and animals are going extinct, the air quality is becoming worse…). To the knowledge of the researcher, there is no existing literature about the effects of different
7
levels of TD in green advertising on consumer reactions. Therefore, the current study aims to close this knowledge gap and answer the following question:
Q2: To what extent do different levels of temporal distance of green advertising messages vary in their effectiveness to stimulate purchase intentions?
Point of reference and temporal distance. As the consequences of (non-) green consumption depicted in advertising messages necessarily always have certain levels of PR and TD, it is also important to know how they influence each other’s potency, thus which combinations of both are most effective in eliciting PI in consumers. As stated above, temporally distant events are perceived as more abstract than temporally close events. The same is true for moral values, which are construed and perceived as abstract (Chen, & Chiu, 2016). They are therefore more likely to get activated when considering distant events and the other way around. This has been supported by a study showing that moral values are better predictors of behavioral intentions for situations with a high TD than with a low TD (Eyal, Sagristano, Trope, Liberman, & Chaiken, 2008). Also, in a study by Yang, Lu, Zhu and Su (2015), advertising messages for a natural drink displaying vague consequences of (non-) green consumption elicited more PI when they concerned the environment than when they concerned the consumers themselves. For green advertising messages displaying detailed consequences, there was no such difference. Since events with a high TD are perceived as more abstract and those with a low TD as more concrete, similar results may be obtained when investigating the PR and TD of green advertising messages. Based on the previous findings, it can be assumed that green advertising messages which target the consumers’ moral values, thus for instance with the environment as PR, will be more effective with a high TD than with a low TD. The opposite may be true for egoistic motives and temporally
8
will be more effective than a high TD.
H1: For green advertising messages with the environment as point of reference, a high temporal distance is more effective in eliciting purchase intention than a low temporal distance. For green advertising messages with the self as point of reference, a low temporal distance is more effective in eliciting purchase intention than a high temporal distance.
The Influence of Personal Predispositions on Message Characteristics Effects
Since targeting clusters of consumers on the basis of their individual characteristics is generally understood as a useful technique in advertising (Belch, & Belch, 2015), it can be assumed that the effectiveness of different PR’s and TD’s in green advertising messages differs for consumers with different traits. The present study therefore additionally aims to investigate to what extent consumers’ levels of environmental consciousness, long-term orientation and self-construals influence the effectiveness of different TD’s and PR’s of green advertising messages. These personality traits will be discussed in the following sections.
Environmental consciousness. The concept of environmental consciousness refers to the pro-environmental orientation people have. This entails their level of knowledge and concern for environmental issues, pro-environmental attitudes as well as values and behavior with the goal of reducing negative and likewise enhancing positive impacts on the
environment. The latter includes the purchase of environmentally friendly products (Chen, & Chiu, 2016; Kollmuss, & Agyeman, 2002). In the context of green advertising messages, it has been shown that environmental claims generate more positive responses from consumers with a high environmental consciousness than from those with a low environmental
consciousness (Chan, & Lau, 2004). In contrast, it can be assumed that people with a low environmental consciousness will react more positively to self-related benefits of green consumption, since they do not value the environment as much. Therefore, the following
9
hypothesis can be formulated:
H2: For consumers with low environmental consciousness, green advertising messages with the self as point of reference result in higher purchase intention than with the environment as point of reference. For consumers with high environmental consciousness, green advertising messages with the environment as point of reference result in higher purchase intention than the self as point of reference.
Furthermore, research has shown that when consumers with a high environmental consciousness are exposed to advertising claims with a high TD, they perceive the product to be more effective and have a more positive brand attitude than when they are exposed to claims with a low TD. For people with a lower environmental consciousness on the other hand, no difference has been found in the effectiveness of advertising claims with a high or low TD (Chen, & Chiu, 2016). However, Chen’s and Chiu’s study (2016) only featured vague consequences with a gain frame, thus benefits of green consumption behaviour (‘Let the earth be better tomorrow/ in ten years’). Negative consequences of non-green products, called loss frames, were not included. Since framing effects on the effect of different TD’s in green advertising messages have been shown before (Chang et al., 2015), it is important to know if a combination of a loss- and a gain-frame and also specific depicted consequences will lead to the same results. The following hypothesis is formulated:
H3: For consumers with low environmental consciousness, green advertising messages with a low temporal distance result in the same purchase intention as with a high temporal distance. For consumers with high environmental consciousness, green advertising messages with a high temporal distance result in higher purchase intention than with a low temporal distance.
10
Another study showed that in cases of temporally distant rather than close events, people are likely to prefer altruistic behaviours that target their individual moral values over selfish behaviours (Agerström, & Björklund, 2009). Therefore, for people with a high environmental consciousness, green advertising messages with a high TD and the
environment as PR that appeals to their pro-environmental orientation may be more effective in eliciting PI than green advertising messages with other combinations of PR and TD. Based on this and the first two hypotheses, it can be assumed that the effects of the two advertising characteristics will complement each other:
H4: For consumers low in environmental consciousness, green advertising messages with a low or high temporal distance and the self as point of reference result in the highest purchase intention. For consumers high in environmental consciousness, green advertising messages with a high temporal distance and the environment as point of reference result in the highest purchase intention.
Long-term orientation. People’s level of long-term orientation describes their tendency to consider rather the past and future or primarily the present time as important. People with a high long-term orientation value long-term benefits more than short-term benefits. Furthermore, they appreciate traditions and thereby also exhibit a long-term
orientation in direction of the past. In contrast, people with a low long-term orientation value short-term benefits over temporal distant values. The concept of long-term orientation thus consists of two distinct dimensions, namely planning and tradition (Bearden et al., 2006). In a study by Tangari and Smith (2012), savings on product packaging for energy efficient
products were evaluated better and elicited more PI by highly future-oriented consumers when they were framed with a high TD than when they were framed with a low TD. The opposite was true for people with a focus on the present (Tangari, & Smith, 2012). It can be
11
assumed that people with a high long-term orientation will also value future benefits of green consumption more than immediate benefits, which is assumed to be the other way around for people with a low long-term orientation.
It has also been implied that the level of long-term orientation affects a person’s ethical values. People with a high long-term orientation who value tradition and future outcomes put more emphasis on ethical values which are innately long-term. They might disapprove of unethical behavior firstly because it violates norms developed by traditions and secondly because relationships with other people can be negatively affected in the long run if the behavior is discovered by them (Bearden et al., 2006). Also, moral principles, just as future events or consequences, are generally perceived as more abstract than selfish motives (Chen, & Chiu, 2016). Therefore, future events and those that target moral values may match each other conceptually. Furthermore, research has shown that people with a high concern for future consequences of actions have more negative attitudes about non-green behaviors and are more willing to exhibit various environmentally friendly behaviors than those who mainly value proximal consequences of actions (Joireman, 2005; Joireman, Van Lange, & Van Vugt, 2004). In line with that, another study showed that people who are future-oriented regard environmental and those who have a focus on the present regard personal benefits for extending public transit options as most important (Joireman et al., 2001). Thus, consumers with a high long-term orientation are assumed to value environmental-friendly reasons for purchasing a green product and thereby the environment as PR in green advertising messages more than the self as PR. In contrast, individuals with a low long-term orientation may value self-related benefits of green consumption, thus the self as PR, more.
Furthermore, it can be assumed that the effects of PR and TD, taking people’s long-term orientation into account, will complement each other. On the grounds of the existing findings, the following hypotheses are formulated:
12
H5-H7: The interaction effects between long-term orientation, point of reference and temporal distance on purchase intention are expected to be similar to the interaction effects posed for environmental consciousness in H2 through H4.
Self-construals. People vary in the way they describe and view themselves in relation to their environment and other people, depending on their so-called self-construal. There are different dimensions of self-construals, which are all possessed by every individual to varying degrees (Arnocky et al., 2007; Chen et al., 2015; Cross et al., 2011; Kareklas, Carlson, & Muehling, 2012). In research and theory, a distinction is often made between independent and interdependent self-construals (Cross et al., 2011). A high interdependent self-construal describes individuals that construe themselves primarily in terms of relationships with other people or group memberships. In contrast, people with a highly independent self-construal perceive themselves especially in terms of individual, unique attributes and by distinguishing themselves from others. Their focus lies on individual goals (Arnocky et al., 2007; Chen et al., 2015; Cross et al., 2011). Importantly, recent literature suggests that people also possess a third type of self-construal to varying degrees which is especially relevant concerning
environmental issues and is named ‘meta-personal self-construal’. In comparison to people with a low meta-personal self-construal, those with a high meta-personal self-construal feel that they are interconnected with all life forms. They value harmony with other living beings and therefore do not want to harm the environment and animals (Arnocky et al., 2007; Chen et al., 2015).
In the context of green consumption behavior, there are several studies illustrating that individuals with varying self-construals behave on the ground of different motives. For instance, it has been shown that people with a high interdependent self-construal buy
ecological rice primarily for the benefits of their family while those with a high independent self-construal consume ecological rice more because of egoistic reasons such as personal taste
13
preferences. Individuals high on the meta-personal self-construal buy ecological rice especially due to a strong appreciation of nature and religious beliefs (Chen et al., 2015). Another study also showed that people with an independent self-construal have a more egoistic environmental concern whereas those with a meta-personal self-construal rather exhibit a more biospheric, thus truly environmental concern and more environmental conservation behavior in general (Arnocky et al., 2007).
In all, it seems that consumers have different motives for green consumption, depending on their self-construals. It is therefore probable that matching the benefits of a green product depicted in advertising messages with such motives and therefore personal levels of self-construals is beneficial. Specifically in the context of a green products’ benefits for either the environment or the self, mainly people’s independent and meta-personal self-construal are considered relevant. On the basis of the existing research, it is expected that consumers with a high independent self-construal will be more affected by ego-centric green advertising messages. In contrast, consumers with a high meta-personal self-construal are supposed to be more persuaded by green advertising messages with the environment as PR.
It can also be assumed that there is a correlation between people’s levels of
independent and meta-personal self-construal and the effectiveness of different TD’s of green advertising messages. One recent study combined measures of different PR’s and levels of abstraction of green advertising messages as well as a manipulation of self-identity (Yang et al., 2015). Participants were either primed with a collective level of self, analogous to a high interdependent or meta-personal self-construal as a consistent characteristic, or an
independent level of self, analogous to a high independent self-construal. In a nutshell, in the case of other people as PR, people with an induced collective self-identity reacted more strongly to abstract appeals than to concrete appeals. Conversely, those with an induced independent self-identity liked concrete more than abstract appeals. There were no such differences found when the messages referred to self-related benefits (Yang et al., 2015).
14
Those findings are interesting since, as elaborated above, a message’s TD is comparable to its level of abstraction. Generally, a high TD is perceived as more abstract than a low TD, which is thought of more concretely (Trope, & Liberman, 2010). Therefore, similar results may be obtained with TD rather than level of abstraction as a characteristic of green advertising messages as well as different self-construals as steady personality traits instead of a
manipulated temporary self-concept. Temporally distant benefits of (non-) green consumption in advertising messages will probably be more effective than immediate benefits for people with a high meta-personal self-construal. This may also be the case because they construe themselves holistically as part of all living beings, which may also include future living beings. In contrast, immediate benefits depicted in green advertising will assumably be more effective than future benefits in cases of a high independent self-construal.
Furthermore, on the grounds of the preceding hypotheses and therefore slightly divergent from the results obtained by Yang and colleagues (2015), the effects of the
messages’ TD and PR are hypothesized to complement each other. On the basis of all existing findings, the following hypotheses can be formulated:
H8,10,12: The interaction effects between independent self-construal, point of reference and temporal distance on purchase intention are expected to be opposite to the interaction effects posed for environmental consciousness in H2 through H4.
H9,11,13: The interaction effects between meta-personal self-construal, point of reference and temporal distance on purchase intention are expected to be similar to the interaction effects posed for environmental consciousness in H2 through H4.
For an overview of all hypotheses and the conceptual model, see Figure 1, Appendix A.
15
Selection of Participants
For this research, a convenience sample of adults was used. The participants were approached personally via social media and in social media groups that students and
researchers join. In order to avoid ethical concerns, only adult participants of 18 years or older were included. Children are generally also known to react differently to advertising (Nairn, & Fine, 2015). Furthermore, only individuals with a good level of English skills were able to take part in the study since the stimulus materials and questionnaires were provided in English only. The participants were informed about those conditions at the beginning of the study and therefore filtered by means of self-selection.
Characteristics of Participants
A hundred and sixty-five participants started to participate in the current study, of which 22 had to be excluded from the analyses because they did not finish the questionnaires. The final research sample thus consisted of 143 people, including 102 women (71.33%), 40 men (27.97%) and one person who identified as belonging to another gender identity (0.70%). On average, the participants were aged 25 (M = 24.63, SD = 4,18). Most participants had obtained a bachelor’s degree (n = 95, 66.43%), 23 participants had a master’s degree (16.08%), 21 a high-school diploma (14.69%) and four another diploma (2.80%) as their highest achieved diploma. Furthermore, most participants were currently students (n = 118, 82.52%), 23 were working (16.08%) and one had another employment status (0.70%). Eighty of the participants (55.94%) had lived in Germany and 24 (16.78%) had lived in the
Netherlands for the longest part of their life. The 39 remaining participants (27.27%) had lived in Algeria, Austria, Belgium, Chile, Djibouti, Dominican Republic, Estonia, Finland, Japan, Lithuania, Malaysia, Republic of Moldova, Sweden, Switzerland, Viet Nam, USA, Great Britain, Spain, Italy, Greece, France, Bulgaria and Australia for the longest period of their lives.
16
The current study investigated a possible causal relationship between different points of reference (PR) and levels of temporal distance (TD) in green advertising messages and consumers’ purchase intention (PI), while taking the participants’ levels of environmental consciousness, long-term orientation and self-construals into account. In order to be able to detect causality, an online experiment was conducted and the experiment was executed via the online survey platform Qualtrics Survey Software. The messages’ PR and TD were
systematically manipulated and the participants were randomly assigned to one of four conditions. A 2x2 design was applied, with the message’s PR (two levels: self vs.
environment) and TD (two levels: present vs. future) as between-subject factors. There were two nominal levels of PR (the self vs. the environment) and two nominal levels of TD (present vs. future; see Table 1, Appendix A). Measures of the participants’ levels of
environmental consciousness, long-term orientation and independent and meta-personal self-construal were included as quantitative moderators. PI as the dependent variable was
measured on a quantitative scale. Stimulus Material
As the independent variables, the PR and TD of the consequences of green
consumption depicted in advertising were manipulated. For this study, organic clothes were chosen as the advertised object. Clothes are consumer goods that are purchased a lot by people, regardless of for instance their gender, age or culture (Fashionunited.com, 2018; Gaille, 2016). Therefore, clothes seemed to suit well as the object for the current study which included adult participants with different backgrounds. In order to exclude the influence of possible pre-existing brand attitudes, a fictional brand of organic clothes, ‘YourLook’, and its logo were invented and used. The same basic advertisement was used in each condition (see Appendix B). The advertising pictures, layout and basic body copy were thus the same across conditions, so that possible confounding variables due to the design were avoided. Also, the text featuring the consequences of non-green purchasing behaviour included both a gain and a
17
loss frame in all conditions. Such an even framing strategy was supposed to eliminate any unwanted framing effects which have been found in previous studies (Chang et al., 2015; Langer, 2013; Segev et al., 2015, Kareklas et al., 2012). The negative consequences of purchasing non-organic clothes functioned as the loss frame, whereas the statement that purchasing the organic clothes could prevent such consequences provided a gain frame. Lastly, only the parts of the advertisement describing the impact of not purchasing the ecological clothes were adjusted to the four conditions in terms of TD and PR. The structure and wording of those sentences were kept as similar as possible, so that the conditions only differed regarding these two factors.
Experimental Factors
Two levels of the PR were integrated into the study. More precisely, the consequences of purchasing non-organic clothes referred to either the environment or the consumer him-/herself. For the environmental consequences, the perish of fish populations was chosen (see Appendix D). On the other hand, health problems like allergies and asthma were chosen as self-related consequences (see Appendix D).
The two levels of TD utilized in the study referred to immediate or future
consequences of purchasing non-organic clothes, respectively. In the final stimulus material, immediate consequences were described by the expressions ‘directly when you wear them’ or ‘right now’. Future consequences were illustrated by the expressions ‘several years after you wore them’ and ‘in several years’ (for the final stimulus material, see Appendix D).
Pilot Studies
In order to check whether the created stimulus material would be perceived as intended, two short pilot studies with a within-subject design were conducted.
Pilot study 1. In the first pilot study, 30 adult participants were provided with the first version of all four stimulus conditions in a randomized order (for the first stimulus material, see Appendix C). The participants had a mean age of 24 years (M = 24.20, SD = 1.79). Five
18
were males (16.67%) and 25 females (83.33%); 25 were students (83.33%) and four working (13.33%). Furthermore, nine had lived in the Netherlands for the longest part of their life (30%), 16 in Germany (53.33%) and five in other countries (16.67%). For each of the four conditions, two manipulation check items examined if the participants understood the TD and PR of the advertisements as intended. Furthermore, for each condition, one question with a five-point Likert-scale assessed the perceived believability of the depicted consequences of green consumption and two items with five-point Likert-scales assessed the perceived severity and the wish to prevent such consequences (called perceived severity from now on; for all scales, consult Appendix C). The results of the first pilot study showed that the initial
materials used to manipulate TD and PR were not successful because the participants did not perceive enough difference between the conditions with the present versus future as TD, and the self versus environment as PR. Up to nine participants did not understand the PR or TD as intended (see Table 4, Appendix C). Paired-sample t-tests were conducted comparing the means of the perceived believability and severity of the different conditions. Generally, the believability and severity of the conditions with the environment as PR were perceived as higher than those of the conditions with self as PR, whereas they should have been comparable across conditions (see Table 2 and 3 in Appendix C).
As a consequences of Pilot study 1, the stimulus materials were adjusted. In order to make sure that the participants payed enough attention to the information in the
manipulations, those parts of the sentences were made bold. Furthermore, clarifying words were added in order to emphasize the temporal occurrence of the consequences (‘directly when you wear them’ instead of ‘when you wear them’, ‘several years after you wore them’ instead of ‘years after you wore them’, ‘in several years’ instead of ‘in the years to come’). Additionally, for the conditions with the self as PR, information was added about how the consequences can occur (‘pesticides (…) can be absorbed by your skin’) and the
19
(‘allergies and asthma’ instead of ‘allergies and rashes’; for the revised stimulus material see Appendix D).
Pilot study 2. In order to ensure the suitability of the adapted stimulus material, a second pilot with the same within-subject design and the same scales was conducted with 16 other individuals. Their mean age was 25 (M = 24.81, SD = 1,56); five were males (31.25%) and eleven females (68.75%). Eleven participants were students (68.75%) and four working (25.00%). Three had lived in the Netherlands for the longest part of their life (18.75%), ten in Germany (62.50%) and three in other countries (18.75%). The results of this second pilot showed that the manipulations of TD and PR of the adjusted stimulus material were
comprehended as intended since maximally one participant did not perceive the manipulations as intended (Table 7, Appendix D). Furthermore, the perceived believability and severity of all conditions were non-significantly different (Tables 5 and 6, Appendix D). Therefore, this revised stimulus material was used for the main study.
Moderators
Environmental consciousness. Environmental consciousness as a moderator was measured with the New Environmental Paradigm (NEP), consisting of fourteen items answered on a seven-point Likert scale (consult Appendix E). This version of the NEP has often been used and found a valid and reliable scale for measuring environmental
consciousness (Chen, & Chiu, 2016; Harraway, Broughton-Ansin, Deaker, Jowett, &
Shephard, 2012). Factor analyses were conducted and led to the exclusion of all reverse coded items and thereby a final scale of eight items (Table 8, Appendix E). Its internal consistency was acceptable (Cronbach’s α = .743). The total mean score of the remaining eight items (Appendix E) was calculated for each participant, with one as the lowest possible mean score indicating a low level of environmental consciousness, and seven as the highest possible mean score indicating a high level of environmental consciousness (M = 5.54, SD = .72).
20
by a validated ratio scale consisting of two factors, one of which assessing the planning dimension and the other one the tradition dimension of long-term orientation. Each factor was assessed by four items, all of which answered on a seven-point Likert-scale (Bearden et al., 2006, see Appendix F). For the current sample, the results of a factor analysis confirmed that the scale was two-factored, distinguishing the tradition from the planning dimension of long-term orientation (Table 9, Appendix F). The internal consistency of the sub-scale measuring the tradition dimension was good (Cronbach’s α = .805), whereas it was acceptable for the sub-scale measuring the planning dimension (Cronbach’s α = .701). For the total scale and for both factors separately, the total mean score of all items was calculated in order to determine the participants’ score on the overall long-term orientation and on the planning and tradition dimension of long-term orientation. One as the lowest possible mean score indicated a very low level and seven as the highest possible mean score indicated a very high level of the respective long-term orientation (overall long-term orientation: M = 4.92, SD = .87; tradition dimension: M = 4.51, SD = 1.19; planning dimension: M = 5.24, SD = .90).
Self-construals. The independent self-construal and the meta-personal self-construal were examined by two separate scales since they are known to be independent from each other (Chen et al., 2015). The independent self-construal was assessed by the Refined
Independent Self-Construal Scale by Hackman, Ellis, Johnson and Staley (1999), based on the initial version of the Independent Self-Construal Scale by Gudykunst et al. (1996). Eleven items were answered on a seven-point Likert-scale (Appendix G). Factor analyses were conducted and led to the exclusion of four items (Table 10, Appendix G). The final scale with seven items (consult Appendix G) had an acceptable internal consistency (Cronbach’s α = .737).
The participant’s meta-personal self-construal was measured by the Metapersonal Self-Construal Scale developed by DeCicco and Stroink (2007). It consists of 10 items with a seven-point Likert-scale (2007, see Appendix H). A factor analysis constituted a one-factorial
21
solution (Table 11, Appendix H) of which the internal consistency was good (Cronbach’s α = .871).
For each scale, thus the independent self-construal and meta-personal self-construal, the participants’ total mean score was calculated with one as the lowest possible score indicating a low level and seven as the highest possible score indicating a high level of independent or meta-personal self-construal (independent self-construal: M = 5.71, SD = .65; meta-personal self-construal: M = 4.55, SD = .97).
Dependent Variable
To what extent the participants intend to purchase the brand’s products was measured by a scale which was designed and used previously by Van Reijmersdal (2016), as a modified version of a validated and often cited scale by Spears and Singh (2004). Answers on four items were given on a seven-point Likert-scale (see Appendix I). A factor analysis confirmed a one-factorial solution of the current scale in the sample (Table 12, Appendix I) and also the internal reliability was very high (Cronbach’s α = .927). For each participant, the total mean score was calculated to determine his/her purchase intention (PI), with a low value referring to a low PI and a high value describing a high PI (M = 3.98, SD = 1.36).
Manipulation Checks
In order to control if the manipulated elements in the stimulus material were perceived as intended, the same manipulation check items with dichotomous answer categories as in the pilot studies (see above) were integrated in the main study. As previously stated, one question checked if the participants perceived the PR as intended and another checked if they
perceived the TD as intended (Appendix C). Procedure
The main study on Qualtrics Survey Software took about five minutes to complete. In the beginning, the participants were provided with an information form, explaining the main purpose of the study, the procedure and contact details of the researcher and the ethical board
22
of the University of Amsterdam in case of any complaints. Then, after being informed about the anonymity of their data and the possibility to drop out any time they want, the participants were asked to give their consent to participate. The study started by a number of items tapping basic demographical information regarding their age, gender, highest achieved education, employment status and country of living. Subsequently, the participants were randomly assigned one of the four stimulus conditions and thus shown one of the advertisements. Thirty-five participants were randomly assigned the first (24.48%), 36 the second (25.17%), 37 the third (25.87%) and 35 (24.48%) the fourth condition. After viewing the advertisement, they were asked to fill in the scales measuring their PI, level of environmental consciousness, long-term orientation, meta-personal and independent self-construal. The two manipulation checks for the perceived PR and TD followed after that. Upon closure, the participants were given a short debriefing of the manipulation used in the study, the exact purpose of the study was revealed and participants were thanked.
Results Preparation of Analyses
Randomization checks. In order to check whether the participants were equally distributed across conditions in regard to their demographic data, randomization checks were conducted. A one-way ANOVA with age as the dependent variable and the experimental conditions as the independent variable was non-significant (F(3,139) = .57, p = .637) and thus showed that the participants’ age was comparable across the four conditions. Furthermore, Chi-Square Tests were conducted with the conditions and the remaining demographic
variables. All tests were non-significant, showing that there were no systematic differences in participants’ demographics across the four conditions. The participants gender (χ2 (6, N = 143) = 4.27, p = .640), highest achieved diploma (χ2 (9, N = 143) = 7.75, p = .560),
23
employment status (χ2 (6, N = 143) = 9.77, p = .135) and country1 (χ2 (6, N = 143) = 7.32, p = .293) were all equally distributed across conditions.
Manipulation checks. The manipulation checks assessed whether the participants perceived the point of reference (PR) and temporal distance (TD) of the stimulus material as intended. A Chi-Square Test with PR and the first manipulation check item, assessing the perceived PR, was significant. As the requirements of the Pearson Chi-Square Test were not met, the Fisher’s Exact Test was interpreted (χ2 (1, N = 143) = 70.45, p = .000). In the two conditions with the self as PR, 54 participants (76.06%) perceived the PR correctly and 17 (23.94%) incorrectly. In the conditions with the environment as PR, 67 participants (93.06%) perceived the PR correctly and five incorrectly (6.04%). It can be concluded from this that in the conditions with the environment as PR, almost all participants perceived the PR as intended. In the conditions with the self as PR on the other hand, almost one fourth of the sample believed that the environment was the PR.
Another Chi-Square Test with TD and the second manipulation check item, assessing the perceived TD, was significant as well. The requirements of the Pearson Chi-Square Test were not met and again, the Fisher’s Exact Test was therefore interpreted (χ2 (1, N = 143) = 13.25, p = .000). In the two conditions with the present as TD, 42 participants (58.33%) perceived the TD correctly and 30 (41.67%) incorrectly, whereas 51 participants (71.83%) in the conditions with the future as TD perceived it correctly and 20 (28.17%) incorrectly. Consequently, it can be concluded that the majority of participants perceived the TD as intended, but in both conditions, around one third did not.
All in all, the manipulations of the messages’ PR and TD tended to be understood correctly but were not completely successful for all participants. Implications of this for the validity of findings will be discussed in the conclusions section.
1 The country that the participants reported to have lived in the longest part of their life was recoded into
24
Hypotheses Testing
Point of reference and temporal distance. In order to assess whether different PR’s and levels of TD of the green advertising messages had an effect on the purchase intention (PI), a two-way ANOVA was conducted on SPSS. The messages’ PR and TD functioned as fixed factors and the PI as dependent variable in this ANOVA.
Q1 asked whether different PR’s vary in their effectiveness to elicit PI. The results of the analysis showed that the PI’s were not significantly different when the PR focused on the self versus on the environment, F(1,139) = .99, p = .322. Hence, the self and the environment as PR in green advertising messages did not differ in their effectiveness to elicit PI (see Table 1 below for means and standard deviations per condition).
Q2 asked whether different levels of TD vary in their effectiveness to stimulate PI. The result in this regard was marginally significant, F(1,139) = 3.31, p = .071. As can be seen in Table 1 below, green advertising messages featuring temporally proximal consequences of (non-) green consumption were more effective in eliciting PI than messages featuring
temporally distant consequences of (non-) green consumption.
Finally, it was hypothesized that green advertising messages with the environment as PR are more effective with a high TD than with a low TD, and that green advertising
messages with the self as PR are more effective with a low TD than with a high TD in eliciting PI (H1). This hypothesis for an interaction effect between PR and TD had to be rejected, F(1,139) = .73, p = .394 (see Table 1 for means and standard deviations). All results can be found in Table 13, Appendix K.
Table 1 – Purchase intention as a function of temporal distance and point of reference in the advertising message
PR/TD The present The future
The self M = 4.17
SD = 1.33
M = 3.57 SD = 1.45
25 The environment M = 4.20 SD = 1.39 M = 3.99 SD = 1.21 Note. N = 143.
Environmental consciousness. In order to assess whether the participants’ level of environmental consciousness influenced which PR and TD of the depicted consequences of green consumption are most effective in eliciting PI, a regression analysis with a moderated moderation model was conducted with SPSS and Hayes’ Process. PR functioned as the independent variable, TD as the first moderator variable and environmental consciousness as the second moderator. The assumptions of normality, linearity, homogeneity and
homoscedasticity were all met. The results of the regression analysis were significant,
F(7,135) = 3.03, p = .005, R² = .14.
H2 proposed that there is an interaction effect of the level of consumers’
environmental consciousness and the PR of green advertising messages on PI. H3 proposed that there is an interaction effect of the level of consumers’ environmental consciousness and the TD of green advertising messages on PI. H4 proposed that there is a three-way interaction effect of the consumers’ environmental consciousness, the TD and the PR of green
advertising messages on PI. All three hypotheses had to be rejected (see Table 2 for statistical values).
The analysis did show a significant effect. There namely was a positive relation between participants’ level of environmental consciousness and their PI, b = .72, t(135) = 2.28, p = .024. A high level of environmental consciousness was thus related to a high intention to buy the organic clothes. All results can be found in Appendix L (Table 14 and 15).
Table 2 – Moderating role of environmental consciousness
b t(135) p
26 Interaction with TD .08 .19 .850 Interaction with PR and TD -.53 -.85 .398 Note. N = 143.
Long-term orientation. In order to assess whether the participants’ level of long-term orientation has an influence on the effectiveness of different PR’s and TD’s of green
advertising messages in eliciting PI, a multiple regression analysis was conducted with SPSS and Hayes’ Process. PR functioned as the independent variable, TD as the first moderator and long-term orientation as the second moderator. The assumptions of normality, linearity, homogeneity and homoscedasticity were al met. The regression analysis was significant,
F(7,135) = 2.69, p = .0111, R² =.12.
However, looking at the individual effects, none were significant. H5 proposed that there is an interaction effect of the level of consumers’ long-term orientation and the TD of green advertising messages on PI. H6 proposed that there is an interaction effect of the level of consumers’ long-term orientation and the PR of green advertising messages on PI. H7 proposed a three-way interaction of the participants’ level of long-term orientation and the PR and TD of advertising messages on PI. All those three hypotheses had to be rejected (see Table 3 for statistical values). All results can be found in Appendix M, Table 16 and 17.
Due to the two-factorial nature of long-term orientation, additional multiple regression analyses were conducted with the sub-scales of the tradition and planning dimension of long-term orientation respectively.
Table 3 – Moderating role of long-term orientation
b t(135) p Interaction with PR -.47 -1.31 .192 Interaction with TD .30 .87 .384 Interaction with PR and TD -.03 .41 .681
27
Note. N = 143.
Long-term orientation, tradition. A regression analysis with PR as independent, TD as first moderator and the tradition dimension of long-term orientation as second moderator variable was conducted. All assumptions of linearity, normality, homogeneity and
homoscedasticity were met. The regression analysis was significant, F(7,135) = 2.80, p = .01,
R² = .13. When looking at the individual effects, the relation between tradition long-term
orientation and PI was marginally significant, b = .33, t(135) = 1.81, p = .073. A higher tradition long-term orientation was thus related to a higher PI, though marginally so.
However, the hypotheses could not be confirmed. Neither the interaction between TD and tradition long-term orientation, nor between PR and tradition long-term orientation, nor the three-way interaction between TD, PR and tradition long-term orientation was significant (consult Table 4 for statistical values). H5, H6 and H7 were thus also not supported by only taking the tradition sub-scale of long-term orientation into account (for all results, consult Appendix M, Table 18 and 19).
Table 4 – Moderating role of long-term orientation, tradition
b t(135) p Interaction with PR -.31 -1.21 .227 Interaction with TD .16 .64 .553 Interaction with PR and TD .17 .45 .654 Note. N = 143.
Long-term orientation, planning. Another regression analysis with PR as independent variable, TD as first moderator and the planning dimension of long-term orientation as second moderator variable was conducted. All assumptions of linearity, normality, homogeneity and homoscedasticity were met. The regression analysis was not significant, F(7,135) = 1.48, p = .179, R² = .07. Thus, hypotheses 5,6, and 7 could also not be supported by focusing only on the planning sub-scale of long-term orientation (Appendix M,
28
Tables 20 and 21).
Independent self-construal. In order to assess whether the participants’ level of independent self-construal influenced the effectiveness of different PR’s and TD’s of green advertising messages on PI, another multiple regression analysis with SPSS and Hayes’ Process was conducted. All assumptions of normality, linearity, homogeneity and homoscedasticity were met.
The regression analysis was not significant, F(7,135) = 1.13, p = .351, R² = .06. H8 proposed that there is an interaction effect of consumers’ level of independent self-construal and the PR of green advertising messages on PI. H9 proposed that there is an interaction effect of the consumers’ level of independent self-construal and the TD of green advertising messages on PI. H10 proposed that there is a three-way interaction effect of the consumers’ level of independent self-construal and the messages’ PR and TD on PI. Therefore, all three hypotheses had to be rejected. For the results, consult Tables 22 and 23 in Appendix N.
Meta-personal self-construal. In order to assess whether the participants’ level of meta-personal self-construal influenced the effect of different PR’s and TD’s of green advertising messages on PI, another multiple regression analysis with SPSS and Hayes’ Process was conducted. All assumptions of normality, linearity, homogeneity and
homoscedasticity were met. The regression analysis was significant, F(7,135) = 2.48, p = .020, R² = .11.
H11 proposed that there is an interaction effect of consumers’ level of meta-personal self-construal and the PR of green advertising messages on PI. H12 proposed that there is an interaction effect of consumers’ level of meta-personal self-construal and the TD of green advertising messages on PI. H13 proposed that there is a three-way interaction effect of consumers’ level of meta-personal self-construal and the PR and TD of green advertising messages on PI. Neither of the three hypotheses was supported (consult Table 5 for statistical values).
29
However, a significant relationship was found between the participants’ level of meta-personal self-construal and PI, b = .54, t(135) = 2.30, p = .023. The higher the meta-meta-personal self-construal, the higher was the PI (see Tables 24 and 25 in Appendix O for the complete results).
Table 5 -Moderating role of meta-personal self-construal
b t(135) p Interaction with PR -.21 -.61 .545 Interaction with TD .13 .41 .686 Interaction with PR and TD -.05 -.10 .919 Note. N = 143.
Discussion and Conclusion
The present study investigated whether different points of reference (PR) and levels of temporal distance (TD) of the consequences of non-green purchase behavior depicted in green advertising messages vary in their effectiveness to stimulate purchase intention (PI). Also, it was investigated if this is influenced by certain consumer traits, namely the level of
environmental consciousness, long-term orientation and/or self-construals. An online experiment was conducted with 143 adult participants which yielded mixed results. The self and the environment as PR in green advertising messages were comparably successful in eliciting PI. However, participants’ PI was higher when they were shown green advertising messages with consequences of non-green consumption occurring in the present than when future consequences were displayed. The messages’ effectiveness did not differ for people with different levels of environmental consciousness, long-term orientation or self-construals. Point of Reference
The finding that there was no effect of different PR’s on PI indicates that the self and the environment as PR in green advertising messages are equally successful in stimulating PI.
30
This suggests that consumers value egoistic and environmental motives for purchasing green products to the same extent. This conclusion neither supports nor stands out from previous research which yielded ambiguous results, either supporting the superiority of egoistic (Segev et al., 2015; Magnusson et al., 2003; Corbett, 2005; Howell, 2012; Ojea, & Loureira, 2007) or of environmental motives (Kareklas et al., 2014; Nordlund, & Garvill, 2002) for
environmental-friendly behavior. A possible explanation for those mixed findings is that the effectiveness of the PR in green advertising messages is dependent on the product or product category. Earlier studies were conducted with organic food, ecological washing detergents, a carbon dioxide reducing lifestyle and wildlife donations (Kareklas et al., 2014; Segev et al., 2015; Magnusson et al., 2003; Corbett, 2005; Howell, 2012; Ojea, & Loureira, 2007). It is possible that consumers either have more egoistic or more environmental motives for those green (consumer) behaviors, but not for purchasing organic clothes. In this context,
advertising messages with different PR’s may be equally effective because egoistic and environmental motives for purchasing organic clothes may be equally strong.
Temporal Distance
To the knowledge of the researcher, there was no previous research about the effectiveness of different TD’s in green advertising messages on PI. In the current study, although marginally so, green advertising messages displaying immediate consequences of non-organic clothes were more successful in eliciting PI than those featuring future
consequences. This indicates that a low TD tends to be more effective in green advertising messages than a high TD, suggesting that stressing immediate consequences of green consumption encourages people more than future consequences. This matches the fact that humans generally value immediate over future rewards because waiting for a benefit is perceived as a cost (Wittmann, & Paulus, 2007).
Point of Reference and Temporal Distance
31
combinations of PR and TD, suggesting that all combinations of the two characteristics of green advertising messages are equally effective in eliciting PI in consumers. This contradicts the findings of an earlier study which showed that moral values are better predictors of
behavioral intentions for situations with a high TD than with a low TD (Eyal et al., 2008). Therefore, although moral motives may be better predictors of future than immediate behavioral intentions, this may not be the case for the temporal occurrence of the consequences of behavior.
Personality Traits
No effects were found of the participants’ level of environmental consciousness, long-term orientation or self-construals on the effectiveness of different PR’s or TD’s in green advertising messages on PI. Also, there was no effect found of those personality traits on the effectiveness of different combinations of PR’s and TD’s in green advertising messages. Firstly, this indicates that consumers’ pro-environmental orientation, long-term orientation and self-construals do not influence what PR appeals most to them. Both self-related and environmental consequences displayed in green advertising messages may be equally
effective in stimulating PI regardless of these personality traits. This in turn indicates that for people with a high and a low environmental consciousness, long-term orientation and
independent and meta-personal self-construal, both egoistic and environmental motives for green consumption are equally relevant. Also, the findings suggest that these personality traits do not influence the effect of green advertising messages displaying either immediate or future consequences of non-green consumption. Thus, consumers who score high on
environmental consciousness, long-term orientation and independent and meta-personal self-construal seemingly value immediate and future consequences of (non-) green consumption to the same extent as those who score low on these personality traits. Also, all combinations of PR and TD in green advertising messages seem to be equally effective for all levels of these consumer traits, indicating that self-related and environmental consequences of green
32
consumption are equally effective for people with a high and low environmental
consciousness, long-term orientation and independent and meta-personal self-construal, regardless of their temporal occurrence.
Those findings contradict earlier research which led to the conclusion that
environmental consequences of (non-) green consumption displayed in advertising messages would be more effective than self-related consequences for people with a high rather than a low environmental consciousness (Chan, & Lau, 2004), long-term orientation (Joireman et al. 2001) and meta-personal and a low rather than a high independent self-construal (Chen et al., 2015, Arnocky et al., 2007) and that a high TD in green advertising messages would be more effective than a low TD for people with a high rather than a low environmental consciousness (Chen, & Chiu, 2016), low long-term orientation (Tangari, & Smith, 2012), and
meta-personal and a low rather than a high independent self-construal (Yang et al., 2015).
However, those studies had noticeable differences from the current one. First of all, some of them investigated completely different environmental-friendly behaviors than green consumer behavior, such as general green lifestyle choices or societal environment-friendly opinions (e.g. Joireman et al. 2004). Other studies found effects in the context of consumer behavior. However, they used purely practical products (refrigerators, paper towels, detergents, food) as advertising objects (e.g. Chan, & Lau, 2004). It is thus possible that in the case of such
products, where consumers elaborate on their choices mainly based on practical reasons, a PR and TD in green advertising messages matching their personality traits are relevant additional arguments for choosing a certain product over another. However, in the case of clothes, this effect may be overshadowed by their self-expressive nature. Clothes are, besides their practical value, often used for social, symbolic and self-expressive reasons (Chernev, Hamilton, & Gal, 2011; Kleine, Schultz Kleine, Kernan, 1993). Thus, it is possible that, depending on consumers’ levels of environmental consciousness, long-term orientation and self-construals, different PR’s and TD’s in persuasive messages vary in their effectiveness to
33
stimulate general environment-friendly behavior and attitudes as well as consumer behavior only when it concerns purely functional products.
Other Findings
The present study yielded three significant effects. First of all, the participants’ level of environmental consciousness was positively related to their PI. It thus seems that consumers high in environmental consciousness tend to value organic clothes more than those with a low environmental consciousness, regardless of the sort of consequences portrayed in advertising messages. Possibly this means that consumers with a high environmental consciousness regard additional (immediate and future health-related and environmental) benefits of clothes as more relevant than those with a low environmental consciousness. Also, there was a positive relation between the consumers’ meta-personal self-construal and PI, just as a marginal positive relation between the tradition dimension of long-term orientation and PI. Apparently, consumers who perceive themselves to be part of all living being on earth and those that highly value traditions also have a greater interest in organic clothes than those who do not feel a strong connection to other living beings or who value traditions less. Again, this may be because they find health-related and environmental benefits of clothes that
complement their functional and self-expressive features more relevant. Limitations
Just as in every experiment, there are a number of limitations that have to be taken into account when interpreting the findings. To begin with, the manipulation checks showed that not all manipulations were successful. Although two pilot studies resulted in stimulus materials that were perceived as intended, a high percentage of the sample in the main study did not report the correct PR and/or TD of their condition. Therefore, the results of the current study have to be interpreted with caution. Some participants who had been shown a green advertising message with the self as PR reported that the consequences of non-green consumption concerned mainly the environment. This did almost not happen the other way
34
around, indicating that when it comes to organic clothes, consumers have the tendency to believe that they are especially beneficial to the environment, but not themselves. Concerning the message’s TD, many participants reported that the consequences of non-organic clothes would be immediate although they were described to occur in the future, or the other way around. On the one hand, future research should therefore assure that the TD of the advertising messages is clear, for example by using more explicit expressions such as ‘immediately’ or ‘in 20 years’. On the other hand, this indicates that participants could not recall the advertising messages, probably because they did not read them attentively.
This may have been due to the fact that the experiment was conducted online. As a result of this, the participants could take part on their own devices anywhere and at any time, for instance in a distracting environment and while doing something else. Therefore, the possibility cannot be ruled out that the results were influenced by a reduction of attention and would be different in a controlled laboratory setting. On the one hand, the fact that the
participants were in a natural environment with possible distractions can be regarded as a benefit for the external validity of the findings. Usually, consumers happen to encounter advertising in their everyday life, for instance when they are strolling, watching TV or, in the context of an online advertisement, surfing the Internet. Therefore, the fact that the
participants took part in a natural, possibly distracting environment brought the study closer to actual (online) advertising situations. On the other hand, this may have restrained the study’s internal validity. If the participants were distracted, it is possible that they did not report their personality traits and PI in a valid way.
The assumption that the participants did not devote their full attention to the
participation in the study is supported by two additional observations. First, the mean duration of participation was about half an hour, although taking part actually only should have taken five to ten minutes. Secondly, this assumption is supported by the pattern of items of the Revised New Ecological Paradigm Scale measuring the participants’ level environmental
