University of Amsterdam
Green Advertising in Germany’s Car Industry:
The Influence of Green Advertisements on the Consumer’s
Purchase Intention and Brand Attitude and the Roles of
Environmental Concerns and Ad Skepticism
Master Thesis M. Sc. Communication Science / Persuasive Communication
Introduction
Author: Robert Edbauer
Student number: 10602194
Address: Van Woustraat 53-3 1074 AC Amsterdam Email: robert.edbauer@student.uva.nl Mobile Contact: 0645864827
Submission date:30th January 2015 Discipline: M.Sc. Comm. Sc. / Persuasive Communication University of Amsterdam
Graduate School of Communication
Master thesis
Topic:
Green Advertising in Germany’s car industry: The Influence of Green Advertisements on the Consumer’s purchase intention and brand attitude and the role of environmental concerns and ad skepticism
Abstract
Green advertising is vastly expanding due to people’s rising environmental concerns and the tendency to consume environmentally friendly. However, statistics show a certain degree of refusal on the consumer’s side to purchase electronic or hybrid cars. Hence, this study deals with the question whether green advertisements are effective for this car industry. It
concentrates on one of the world’s leading countries in terms of car manufacturing, Germany, and investigates whether green car ads lead to more positive attitudes toward the car brand and to higher purchase intentions than non-green ads and whether environmental concerns or ad skepticism have a moderating influence on these relations. Results indicated that green car ads did not lead to more positive brand attitudes or purchase intentions as compared to non-green car ads, and that both environmental concerns and ad skepticism had no significant moderating impact. However, results also showed that more positive brand attitudes led to high purchase intentions among participants. Other factors such as ad likability, a person’s level of income as well as the already existing attitude toward the car brand were significantly related to the effectiveness of car ads instead and seemed to be more relevant than green ad claims. Based on these results, the study contributed meaningful implications for practitioners and the current scientific status quo: it clearly showed that green claims in ads were not successful in the car industry and that changes in attitudes toward a brand and purchase
intentions toward the advertised product after the ad exposure were more likely to be achieved by stimulating already existing positive brand attitudes and increasing the likability of an ad. While practitioners can take over these scientific implications and focus on ad likability, existing brand attitudes and socio-demographical factors when creating new advertising campaigns in the future, this study also contributes new scientific insights into the effectiveness of green ad claims in the car industry.
1. Introduction
In September 2013, the popular German automobile manufacturer BMW presented its latest e-car “BMW i3” at the Internationale Automobil-Ausstellung, the IAA (in English:
International Motor Show) in Frankfurt am Main, Germany. The IAA is the world’s largest motor show with nearly 1100 exhibitors and one million visitors every year setting new trends in the automobile industry (Stoller, 2013). When presenting the new e-car, BMW followed the most recent trend, a trend of ‘being green’. Being green was also the central theme of the IAA 2013 where new developments in the areas of electro mobility and hybrid cars were presented.
While the automobile-industry tries to successfully place so-called ‘green cars’ on the market, statistics indicate a certain degree of refusal on the consumers’ side towards this new development: the prestigious research organization Fraunhofer Society and its partner
PriceWaterhouseCoopers published a study in 2010 which found that – even though the
majority of people assign a very important role to electro-mobility within the car industry – 81 % of the respondents would not buy an e-car in the near future and 61 % hardly know anything about green cars in general (Fraunhofer Society & PriceWaterhouseCoopers study, 2010). One of the biggest problems in this matter could be the way automakers address consumers: by using advertisements that demonstrate the benefits of e-cars for the
environment, so-called ‘green ads’, the industry tries to persuade consumers and to increase profit and sales. However, reality shows that the lack of knowledge about green cars causes distrust and doubts that outweigh on the consumers’ side. It leads to stagnation in the field of electro mobility.
Examining research in the field of the automobile industry one can see that it mainly focused on forecasting green trends, the emergence of electric engines or the transformation of the automobile industry. Only little research has been conducted on how green car
advertising can be implemented effectively, leading to the consumers’ positive brand attitudes and resulting purchase behavior. Although green advertising and its effects as well as
sustainability have already been topic of scientific research, previous studies mainly concentrated on green advertising in other areas such as fashion, nutrition, and energy. However, these results can not be transferred to the automobile industry for two reasons: First, when buying a car, consumers need more time for the decision-making process due to the high prices of cars. Compared to fashion, nutrition or energy products, cars require a person’s high involvement and extensive thinking. The Elaboration Likelihood Model states that high and low involved consumers differ in the way they process information: while a low-involved person processes advertising in a heuristic way that focuses only on heuristic cues or emotional aspects, a highly-involved person processes ads through a central pathway focusing on the process of information, arguments and evaluation. Processing in the central path may take additional time, however, the persuasion effect is thus greater (Petty & Cacioppo, 1986). Second, when purchasing green cars, additional aspects such as the existence of charging stations in certain areas also need to be considered. The Model of
Responsible Behavior shows that – next to individual skills, knowledge or attitudes –
pro-environmental behavior is strongly influenced by situational factors (e.g. pro-environmental barriers such as charging stations). If situational factors stand against the pro-environmental behavior, the person will not perform the desired behavior (Hines, Hungerford & Tomera, 1987).
This study aims to fill the scientific gap of the effects of green car ads on consumers by examining to what extent green advertising in Germany’s car industry may influence a person’s purchase intention as well as brand attitudes as compared to non-green ads. It also studies whether the consumer’s ad skepticism and his or her basic environmental concern moderate this relationship because previous research showed that a person’s environmental
beliefs positively influence the brand attitude, which, according to the Theory of Reasoned
Action, in turn influences the purchase intention toward green products, and that skepticism
toward green ads has a negative impact on brand attitudes, which results into lower levels of purchase intentions (Albayrak, Caber, Moutinho & Herstein, 2011; Arnocky, Stroink & DeCicco, 2007; Obermiller, MacLachlan & Spangenberg, 2005; Tien & Phau, 2010). Therefore, the study presented in this paper deals with the following research question:
To what extent does green advertising in Germany’s car industry influence the consumer’s brand attitude and purchase intention as compared to non-green
advertising and what influence do the consumer’s ad skepticism and environmental beliefs have on this relationship?
By answering this question, we can help to gain both further scientific and practical insight into green advertising in the car industry and its effects on people’s brand attitude and behavioral performance. The results will indicate whether green car ads cause positive
attitudes toward the brand or increased purchase intentions compared to non-green car advertisements or whether green advertising in the car industry should be declared as an unsuccessful persuasion technique. This research shows professionals in the car industry whether green car advertisements indeed help to increase sales of e-cars and how to address consumers effectively.
2. Theoretical Background 2.1 Definition of concepts
Based on scientific studies, this chapter introduces definitions of the terminology used in the study.
With green advertisements, professionals advertise green products that refer, for instance, to sustainability or environmental protection. They use a variety of appeals in order to persuade the customer to purchase the product. Advertisers do so by introducing appeals that emphasize a certain relationship between the environment and the product’s attributes. In other advertisements, more traditional appeals such as benefits of purchasing the product are being used (Schuhwerk & Lefkoff-Hagius, 1995). A brand attitude represents an attitude on the consumers’ side toward a brand that is created after being exposed to advertising (Mittal, 1990). It is an outcome of brand beliefs engendered by brand advertisement (Lutz, 1975; Olson & Mitchell, 1975). Prior research on brand attitudes shows that attitudes toward an ad can be understood as an important antecedent of brand attitude (Olson & Mitchell, 1975; Shimp, 1981). Ad skepticism explains the consumer’s skepticism toward advertising in general and can be defined as the tendency to disbelief advertising claims. While skepticism toward advertisements is a separate construct from skepticism toward other product
information sources such as personal selling, the tendency to disbelief any communicational form is a stable, learnt marketplace belief (Obermiller & Spangenberg, 1998). Environmental
beliefs refer to environmental concerns that indicate a person’s general orientation toward the
environment as well as his or her level of concerns toward environmental issues (Kim & Choi, 2005). Prior research indicates that environmental concerns and environmentally friendly behavior are positively correlated: when consumers show more favorable attitudes toward the environment, their intentions to stop buying products from polluting companies as well as their intentions to make personal sacrifices to slow down pollution are stronger, which leads to more environmentally friendly consumer behavior (Minton & Rose, 1997; Straughan & Roberts, 1999). Finally, a consumer’s purchase intention is the plan to purchase a
2.2 Prior research on sustainable cars in the automobile industry
Prior scientific research on sustainable cars in the automobile industry mainly focused on different hybrid business unit strategies as well as their benefits and disadvantages, the influential factors on consumer choice behavior regarding green products, the forecasting of green trends in the automobile industry by presenting a reliable forecasting model or on how the eco-innovations and technological diversity changed the car industry (Bakker, Sierzchula, Maat & Van Wee, 2012; Eggers & Eggers, 2010; Lin & Huang, 2012; Proff, 2000). There has hardly been any research conducted on the effectiveness of green advertising in the
automobile industry and how that affects a person’s brand attitude or purchase intention. Additionally, aspects such as ad skepticism and environmental concerns, which might have an influence on the effectiveness of green advertising concerning environmentally friendly vehicles on the consumer’s response, have not been researched yet. However, in order to study the aspects described above scientifically, theories and empirical studies from related scientific fields are taken into account that can be used to explain the relations between the concepts as researched in this paper and to state appropriate hypotheses. The following chapter therefore will describe the central concepts, their interrelations as well as underlying theories.
2.3 Green claims, non-green claims, purchase intent and brand attitude Within the last decades, there has been an influx of green claims in advertising. In order to meet increasing environmental concerns of consumers, companies placed green products on the market or assigned environmental attributes to their already existing products (Taghian, 2005). However, marketers often used meaningless, facile and unproven green claims in their ads to charge higher prices for green products even if they were identical to their non-green equivalents. This led to resistance on the consumer’s side, to an underperformance of green products and only limited success of green claims as compared to non-green ones (Crane &
Peattie, 2005; Simintiras & Martin, 1995; Wong, Turner & Stoneman, 1996). As consumers perceived that green products showed inferior quality and that they were unable to actually deliver the promised environmental benefits, they often did not purchase green products (Ginsberg & Bloom, 2004; Phau & Ong, 2007). However, even if consumers did not purchase sustainable products, they still held very positive attitudes toward and favorable perceptions of the brands that positioned themselves as green and showed high levels of purchase
intentions due to their strong environmental concerns as well as their appreciation of the
brand’s effort to support environmental protection (Davies, 1993; Hartmann, Ibáñez & Sainz, 2005, Phau & Ong, 2007). In fact, research showed that green advertising claims led to more positive attitudes toward the brand as compared to non-green advertising claims (Bech-Larsen, 1996; Eagly & Kulesa, 1997; Hartmann et al., 2005; Swenson & Wels, 1997; Benito Gómez, Noya & Paniagua, 1999). Therefore, this study constructs the following hypothesis:
H1: Green ads will lead to more positive brand attitudes as compared to non-green ads.
The Theory of Reasoned Action (TRA) indicates which behavioural determinants predict (intended) behaviour (e.g. purchase intention toward an e-car). It states that, next to a person’s subjective norm, the consumer’s attitude toward a brand or product is the best antecedent for predicting behavioural intention. Attitudes such as brand attitudes thus influence a person’s purchase behaviour via their behavioural intention. A person is more willing to perform certain behaviour when his or her attitude toward a brand is more positive. Scientific research on intended green product consumption is in line with the TRA, showing that individual values and attitudes (e.g. brand attitudes) influence a person’s purchase intention of green products (Mohamad, Lee & Ramayah, 2010). For instance, Kalafatis, Pollard, East and Tsogas (1999) showed that brand attitudes are one of the best antecedents for purchase intentions toward green products. Additionally, Vermeir and Verbeke (2006)
supported this by stating that positive brand attitudes strongly correlate with the purchasing intention toward green products. Based on this, the study as presented in this paper states the following hypothesis:
H2: The more positive the brand attitude, the higher the purchase intention.
2.4 The moderating influence of environmental concern
The influence of environmental concerns on behavioural performance and behavioral intentions has been researched since the 1970s. Schwartz’s (1973) moral norm-activation
theory of altruism, for instance, indicates that “[…] altruistic (including pro-environmental)
behaviour occurs in response to personal moral norms that are activated in individuals who believe that particular conditions pose threats to others (awareness of adverse consequences, or AC) and that actions they could initiate could avert those consequences (ascription of responsibility to self, or AR)” (Stern, 2000, p. 412). This shows that a person might enact pro-environmental behaviour when being confronted with perceived threats such as pro-environmental pollution: he or she develops pro-environmental behaviour due to their environmental
concerns as a response to the perceived threat of environmental pollution.
Kassarjian (1971) found that a general attitude of concern toward ecology is the best predictor of attitudes toward green brands and (intended) ecologically oriented consumer behaviour. According to his study, ecological concern made consumers more receptive to ecological appeals (Kassarjian, 1971). These results were corroborated by several other
studies, indicating that individuals who show higher levels of environmental concern are more likely to show positive attitudes toward the green brands, which in turn result into the
intention to enact in an ecologically conscious way, as compared to those who show lower levels of environmental concern when being confronted with environmental issues (Antil, 1984; Chan, 1996; Donaton & Fitzgerald, 1992; Hines et al., 1987; Kerr, 1990; Ottman, 1993;
Schlossberg, 1992; Shetzer, Stackman & Moore, 1991; Tucker, Rifon, Lee & Reece, 2012). Other research has also shown that environmental concerns have a moderating impact on brand attitudes: the higher the level of environmental concern is, the more positive are the brand attitudes (Black, Stern & Elworth, 1985; Dunlap & Van Liere, 1984; Fransson & Gärling, 1999). Therefore, this research hypothesizes that the influence of green ads on attitudes toward the brand is moderated by the level of environmental concerns:
H3: Participants who are exposed to green advertisements are more likely to have higher
levels of brand attitude when showing a high level of environmental concern than those participants with a low level of environmental concern.
2.5 The moderating influence of ad skepticism
In daily life consumers are overwhelmed with advertisements. Coping with multiple advertising claims simultaneously requires effort. Therefore, consumers are often
unmotivated to evaluate the information and react with disbelief or skepticism toward the ad claim in order to simplify the processing process (MacInnis, Moorman & Jaworski, 1991). In their Persuasion Knowledge Model, Friestad and Wright (1994) describe how a person’s response to persuasive attempts of these advertising claims is influenced by his or her knowledge of persuasion strategies the advertiser is using. The consumer relies on three categories of knowledge: knowledge of the product, knowledge of the persuasion technique the advertiser is using and knowledge about the advertiser or agent himself. Based on this set of knowledge, the consumer counters with a persuasion-coping response (Friestad & Wright, 1994; Obermiller et al., 2005). A common persuasion-coping response toward advertising claims is disbelief and skepticism. As studies show, the consumer’s skepticism toward ads determines his or her advertising response: skepticism moderates responses such as brand
beliefs or attitudes toward the brand, which in turn influence a person’s purchase intention (Obermiller & Spangenberg, 2000; Obermiller et al., 2005). Prior research on skepticism toward advertising has indicated that the more skeptical consumers are toward ads, the less they like advertising, rely on it and attend to it (Obermiller et al., 2005; Kim & Lee, 2009). Based on these results, the research presented in this paper states following hypothesis:
H4: Participants who are exposed to green advertising have lower levels of brand attitude
when showing high levels of scepticism toward the ad as compared to those participants with a low level of scepticism toward the ad.
2.6 Conceptual model of the presented research
Figure 1 shows the conceptual model of this study. It clarifies the relations between
independent and dependent variables as well as the moderators and includes all hypotheses of this research as stated above.
3. Methods
3.1 Research design and procedure
Quantitative research is needed in order to test the hypotheses as stated above. Since this research contained one independent variable with two different levels (green vs. non-green
claims) and each experimental subject was randomly allocated to only one of the experimental conditions, a single factor in-between subjects design with a control group was chosen as a suitable research design (Bryman, 2012). Furthermore, this study contained a cross-sectional self-completion online questionnaire with an experimental character due to the two stimuli (green vs. non-green) and a control group. The questionnaire was sent to participants who were randomly assigned to one of the three conditions and were thus exposed to either full-color green car ads (green ad group), full-color non-green car ads (non-green ad group) or no car ad at all (control group). As of this, the benefits of two different research designs could be combined: while the cross-sectional self-completion online questionnaire enabled to collect a large body of data from people of different sociodemographic backgrounds and geographical areas and guaranteed strong external validity, the experimental character led to strong internal validity and showed the differences and variations between the groups (Bryman, 2012). Due to the randomization when assigning participants to the experimental groups, external factors such as individual differences were kept to a minimum; with this, it was possible to establish causal relationships between the variables (Bryman, 2012).
The self-completion online questionnaire including the advertising stimuli was solely in German since all the participants were German-speaking. This was necessary to ensure that participants fully understood the questionnaire and were able to answer in a correct manner. Additionally, all of the participants had access to the Internet in order to complete the survey. The questionnaire incorporated sociodemographic questions as well as questions concerning the ads presented in the survey, the participant’s brand attitude and purchase intention, as well as the participant’s skepticism toward ads and general environmental beliefs. The duration to complete the survey did not exceed 15 minutes, guaranteeing that potential respondents would
not decide against participating and were willing to complete the survey. Responses were sent back automatically and anonymously and were then statistically analyzed.
3.2 Pre-test
Among people of the intended target group, a pre-test was conducted (N = 15). The pre-test contained seven self-created car advertisements, either showing green or non-green ad claims. Based on the responses it could be decided which two advertisement stimuli were suitable for the self-completion online questionnaire. In order to rule out external influences such as brand or product liking as well as the familiarity with the brand or product, only two different car models from the same brand, namely BMW, were presented. Thus, the only difference between the ads could be found in the ad slogan and ad car description, which represented either green or non-green ads. The pre-test was implemented in German in order to guarantee that respondents had no difficulties in understanding the questions and the advertisings’ meanings.
The pre-test included questions concerning the likability, the persuasiveness, the perceived greenness and the perceived visual quality of the advertisements. Finally, the last question dealt with the perceived relevance of the product for the participant. Except for perceived greenness of the ad, all questions and their scales were based on existing and already tested scales in order to guarantee high scale reliability and – if tested – high validity.
Perceived greenness of an ad has not explicitly been asked in other studies. However, this
was of great interest for this pre-test in order to see which of the ads were perceived as particularly green or non-green. Therefore, a single-item bipolar measure was created, asking “How green (e.g. advertising environmentally friendly, sustainable products) do you think this ad is?” with answer options ranging from [1] “not green at all” to [7] “very green”. Ad
presenting seven answer categories ranging from “dislike very much” to “like very much” as used by Biel and Bridgwater (1990). The ad’s visual appeal was composed of six-, seven-point bi-polar adjectives that measured the degree to which a participant perceived the ad stimulus as aesthetically pleasant based on visual aspects of the ad (Cox & Cox, 1988). Scale items were “bad/good”, “unpleasant/pleasant”, “not likable/likable”, “unflattering/flattering”, “unattractive/attractive”, “not stylish/stylish”. Reliability was reported to be high with an alpha of .93. Validity of the scale was not explicitly tested; however, a factor analysis showed that all items loaded on the same dimension (Cox & Cox, 2002). Persuasiveness of the ads was measured by using a single-item bi-polar measure that asked: “How persuasive do you think this advertisement is?” with answer options ranging from [1] “not persuasive at all” to [7] “very persuasive”. As of this, the degree could be measured to which a participant thought a specific ad to be believable. An alpha of .88 was reported for this scale (Chang, 2003). Finally, for the participant’s perceived relevance of the ad a three-item-, seven-point Likert-type scale was used that measured the relevance of the ad’s message as well as the information in the ad (Laczniak & Muehling, 1993). The questions “Overall, how do you rate the product in the ad?” contained three items, namely “irrelevant/relevant to me”, “uninteresting/interesting to me” and “useless/useful to me”. Reliability was high with a reported Cronbach’s alpha of .96 (Muehling, Stoltman & Grossbart, 1990). In order to analyze the scores for the participant’s perceived relevance of the ad as well as the ads’ visual appeals, overall mean scores were conducted.
Three repeated-measures ANOVAs were conducted as to compare the means for ad likability, perceived greenness and perceived persuasiveness. Results showed that ad 7 (see appendix a) for all ads) insignificantly scored the highest on ad likability among ad 1, 3 and 5 (M = 5.40, SD = 1.18, p = 1,0) and significantly highest on ad likability as compared to ad 2 (M = 3.00, SD = 1.13, p = .012) and ad 6 (M = 2.93, SD = 1.49, p = .010). Additionally, ad 7
insignificantly scored highest on perceived greenness (M = 6.20, SD = 1.27) as compared to ad 1 (M = 4.80, SD = 2.01, p = .199), ad 3 (M = 5.67, SD = 1.23, p = 1.00) and ad 5 (M = 6.07, SD = 1.28. p = 1.00) and significantly highest on perceived greenness compared to ad 2 (M = 2.20, SD = 1.70, p < .001), ad 4 (M = 2.53, SD = 1.19, p < .001) and ad 6 (M = 1.53, SD = .74, p < .001). Lastly, ad 7 insignificantly scored highest on perceived persuasiveness (M = 5.47, SD = 1.30) among ad 1 (M = 4.27, SD = 1.62, p = .945), ad 3 (M = 4.67, SD = 1.50, p = 1.00), ad 4 (M = 5.00, SD = 1.77, p = 1.00) and ad 5 (M = 5.33, SD = 1.68, p = 1.00) and significantly highest on perceived persuasiveness among ad 2 (M = 3.40, SD = 1.64, p = .038) and ad 6 (M = 3.13, SD = 1.41, p = .008).
For ad likability, Mauchly’s test was significant indicating that sphericity had been violated, χ(7) = .01, p < .001. Therefore, degrees of freedom were corrected using Greenhouse-Geisser estimates of sphericity (ε = .504). Results showed that there was a significant effect of likability among the ads, F(1,14) = 916.00, p < .001. This suggests that some of the ads were significantly scoring higher on ad likability than others. Mauchly’s test was also significant for perceived greenness of the ad, χ(7) = .04, p = .011. The degrees of freedom were corrected by using Greenhouse-Geisser estimates of sphericity (ε = .526). Again, results showed that there was a significant effect of perceived greenness existing among the advertisements, F(1,14) = 712.14, p < .001, which indicated that some ads were significantly scoring higher on perceived greenness than others. Finally, for perceived persuasiveness of the ad, Mauchly’s test was insignificant and sphericity had not been violated, χ(7) = .07, p = .066. Results showed that there was a significant effect of perceived persuasiveness of the ad among the advertisements, F(1,14) =647.76, p < .001, which meant that some ads were significantly scoring higher on perceived persuasiveness than others.
Additionally, two repeated-measures ANOVAs were conducted to compare the means for perceived visual quality and relevance of the ad. Results showed that ad 7 scored
insignificantly (and significantly compared to ad 6 (M = 3.68, SD = 1.00, p = .031)) second highest on visual quality (M = 5.36, SD = .99, p = 1.00) compared to all ads except ad 4 (M = 5.38, SD = .90, p = 1.00), which was also chosen for the study’s survey as described below. In regards to perceived relevance of the ad, ad 7 scored insignificantly second highest (M = 3.69,
SD = 1.91, p = 1.00). Ad 5 scored higher on relevance of the ad as compared to ad 7 and a
one-sample t-test showed that the difference was significant (M = 3.91, SD = 1.75, p < .001). However, as ad 7 scored highest on most of the other questions and was more clearly identified as green compared to ad 5 (M = 6.07, SD = 1.28), being an essential factor of this research, ad 7 was chosen as the green ad of this study. For visual quality of the ad, Mauchly’s test was significant showing that sphericity had been violated, χ(7) = .04, p = .013. The degrees of freedom were corrected by using Greenhouse-Geisser estimates of sphericity (ε = .530). It can be assumed that there was a significant effect of visual quality of the ad among the seven advertisements, F(1,14) = 818.17, p < .001. Thus, some ads were significantly scoring higher on visual quality than others. For product relevance, Mauchly’s test was also significant, χ(7) = .02, p = .001. Again, the degrees of freedom were corrected (ε = .447) and there was a significant effect of product relevance among all ads, F(1,14) = 93.83,
p < .001. Thus, some ads were significantly scoring higher on product relevance than others.
Ad 4 showed the same picture and car model but differed in slogan and description (non-green) and could therefore be understood as the counterpart of ad 7. It scored mostly insignificantly highest for visual quality among all ads (M = 5.38, SD = .90, p = 1.00) and also highest for ad likability (M = 4.93, SD = 1.79, p = 1.00) and perceived persuasiveness (M = 5.00, SD = 1.77, p = 1.00) among all non-green ads. Furthermore, respondents clearly understood that ad 4 was non-green (M = 2.53, SD = 1.19, p < .001). Ad 2 (M = 2.20, SD = 1.70, p < .001) and ad 6 (M = 1.53, SD = .74, p < .001) were more clearly identified as non-green as compared to ad 4 and multiple one-sample t-tests showed that these differences were
significant. However, since the non-green ad had to show the same car as its green counterpart (ad 7) in order to avoid biased results in product purchase intentions and only ad 4 showed the exact same car as ad 7, ad 4 was chosen as the non-green ad of this study. All results as described above and more detailed information on all mean scores are illustrated in multiple tables in the appendix (section c)).
3.3 The sample’s characteristics
The study’s target population consisted of German inhabitants with a minimum age of 30. Groups such as students or career entrants who are highly unlikely to be able to afford green cars were thus excluded. Due to the research design as described above, a minimum of 90 respondents were required in order to guarantee meaningful results. A non-probability
sampling, to be more precise a convenience sample, of respondents with Internet access was
chosen. As participants forwarded the questionnaire to other people of the target group, this research also contained snowball effects (Bryman, 2012). Via emails, social networks such as Facebook, LinkedIn or the German equivalent Xing, invitations to participate were sent out. A total of 102 participants completed the survey (N = 102). The respondents were randomly assigned to three different groups: one group consisting of participants who were exposed to the green car ad (N =35), another group who received the non-green car advertisement (N =33) and lastly one group who received no car ad at all (N = 34). Hence, participants were equally distributed between the different groups. Participants were aged between 30 and 73 (M = 55.6; Mo = 63). 46.1 % of the participants indicated to be male, 53.9 % to be female. 7.8 % stated that they completed primary, main or middle school (German: Haupt-/Real- oder Volksschule), 16.7 % completed high school (German: Abitur), 21.6 % a Bachelor’s and 42.2 % a Master’s degree at a University. 11.8 % of the participants completed the PHD/doctoral level (M = 3.33, Mdn = 4, Mo = 4). The majority of participants either earned less than 20.000
Euros (21.6 %) or between 30.000 to 39.999 Euros (19.6 %). 10.8 % indicated that they earned 90.000 Euros or more within one year (M = 4.15, Mdn = 4, Mo = 1).
3.4 Observed variables
3.4.1 Independent variable Green versus non-green advertisements
The study’s independent variable was car advertising showing two dimensions, either green or non-green claims. Additionally, one control group did not receive any advertisement at all. Both ad stimuli showed the same car, however, the ads differed in ad slogan and ad
description. Before being exposed to the ad, all participants were asked to answer questions regarding ad skepticism, environmental concerns, socio-demographics, brand attitudes and purchase intentions after the ad exposure.
3.4.2 Dependent variables Purchase Intention
The participants’ purchase intention was regarded as a dependent variable. In order to measure purchase intention after being exposed to one of the stimuli the question “After having seen the car advertisements, how likely are you to buy the advertised car?” was included in the self-completion questionnaire. As used by Dijkstra et al. (2005), respondents were asked to indicate their opinion on a seven-point-scale ranging from [=1] ‘very unlikely’ to [=7] ‘very likely’ in order to answer this question. For this scale, an alpha of .89 was reported. Thus, the reliability of the scale was high. The variable ‘purchase intention’ was treated as an interval level of measurement.
Brand Attitude
was measured by using a seven-point-, semantic differential scale of seven bipolar items such as “dislike very much/like very much”, “useful/useless“, “valuable/worthless”,
“unimportant/important”, “beneficial/not beneficial”, “not fond of/fond of” and
“unenjoyable/enjoyable”. This is consistent with the scale used by Machleit and Wilson (1988). Within the questionnaire, answer options were balanced out and negative and positive endpoints were mixed in order to minimize answer order effects as well as acquiescence. The seven bipolar items were combined into one variable, which indicated the overall attitude. The variable ‘brand attitudes’ was measured on an interval level and the scale showed a high reliability, with a Cronbach’s alpha between .92 and .97.
3.4.3 Moderators
Ad skepticism was regarded as a moderator in this study. In order to measure ad skepticism,
the SKEP scale as invented by Obermiller et al. (2005) was used. The SKEP scale is a nine-item scale rating advertising on different dimensions as follows: “We can depend on getting the truth in most ads”, “Advertising’s aim is to inform the consumer”, “I believe advertising is informative”, “Advertising is generally truthful”, “Advertising is a reliable source of
information about the quality and performance of products”, “Advertising is truth well told”, “In general, advertising presents a true picture of the product being advertised” and “I feel I’ve been accurately informed after viewing most advertisements”. The participants rated these statements by using a five-point response scale, ranging from [=1] strongly disagree to [=5] strongly agree. Later, the items of this scale were combined to get an overall score for ad skepticism and then divided into two groups with [=1] low ad skepticism and [=2] high ad skepticism. This was achieved by using a median split. The scale showed high reliability with Cronbach’s alpha being .85. Test replications of the scale confirmed this result and
additionally showed high validity (Obermiller & Spangenberg, 1998). Ad skepticism was measured on an interval level of measurement.
The participant’s environmental concern was also used as a moderator in this research. Tarrant and Cordell (1997) developed the environmental behavior index that can be taken into account when measuring the degree of environmental concern of a person. Hence,
respondents were asked to rate different items such as “I make a special effort to buy paper and plastic products that are made from recycled materials”, “I have switched products for ecological reasons”, “When I have a choice between two equal products, I purchase the one less harmful to other people and the environment” or “I have boycotted or avoided buying the products of a company because I felt that company was harming the environment” in order to see their concerns for the environment and (intended) environmental behavior; this was adapted from previous research as stated by Lin and Huan (2011). In total, ten questions concerning the environmental concerns were asked. All questions were answered by a five-point Likert scale from [=1] strongly disagree to [=5] strongly agree. The items of this scale were combined in order to get an overall score for environmental concerns and then divided into two groups with [=1] representing low environmental concerns and [=2] representing high environmental concerns, by means of a median split. The reliability of the scale was high, Cronbach’s alpha being .88. Environmental concern was measured on an interval level of measurement.
3.4.4 Control variables
In this research, ad likability and already existing brand attitudes were taken into account as control variables. As mentioned in the description of the study’s pre-test, ad likability was measured by using a single-item bipolar measure presenting seven answer categories ranging from “dislike extremely” to “like extremely” as introduced by Biel and Bridgwater (1990). Already existing brand attitudes were measured by using a seven-point, semantic differential scale of seven bipolar items such as “dislike very much/like very much”, “useful/useless“,
“valuable/worthless”, “unimportant/important”, “beneficial/not beneficial”, “not fond of/fond of” and “unenjoyable/enjoyable” as stated before. This is consistent with the scale used by Machleit and Wilson (1988). Furthermore, sociodemographic questions were assessed in order to see whether age, gender, education and income had any impact on the results.
3.4.5 Manipulation checks
The manipulation check of this study was whether the ad was perceived as being green or non-green. Therefore, the single-item scale “How ‘green’ (e.g. advertising environmentally friendly, sustainable product) do you think this ad is?” was stated with answer options ranging from [=1] not green at all to [=7] very green. As the stimuli clearly showed either non-green or green cars, failing to answer this question correctly led to the assumption that those respondents did not understand the advertisement properly, filled out the questionnaire
randomly or did not read the advertisement description. Hence, these observations would have been dropped in order to ensure that results were not biased. However, as the ads could easily be distinguished in green and non-green claims, all participants answered this question correctly and no observation had to be dropped. Within the self-completion survey, the ads perceived as most green and most non-green in the pre-test were chosen in order to ensure that the manipulation was successful.
4. Data analysis plan
In this study, several analyses and tests were conducted. In order to control for possible alternative explanations between the study’s variables, bivariate correlation analyses for the participant’s sex, level of income, general brand attitude, general ad likability and age on both dependent variables brand attitude and purchase intention were run. Additionally, chi-square tests for the participant’s educational background and brand attitude as well as purchase
intention were conducted. The scales used in this research were based on existing scales from scientific articles. However, their level of reliability was tested again so as to guarantee highly reliable scales in this study. Additionally, in order to test the study’s hypotheses, the
following statistical calculations were applied: while H2 was tested by means of a hierarchical
multiple regression with brand attitude after the ad exposure being the independent, and purchase intentions after the ad exposure being the dependent variable while controlling for the covariates income, general ad likability and already existing brand attitude toward BMW, H1, H3 and H4 were examined by applying ANCOVAS including moderators and covariates
since three group means were compared.
5. Results
5.1 Controlling for alternative explanations
Bivariate correlation analyses and a chi-square test were conducted in order to control for alternative explanations. Results showed that there were positive correlations between the dependent variable “brand attitude after ad exposure” and the general brand attitude (r = 0.595, N = 67, p < 0.001), the general likability of the ad (r = 0.629, N = 67, p < 0.001) and the participants’ income (r = 0.299, N = 67, p = 0.014). Furthermore, there were positive correlations between the dependent variable “purchase intention of the product” and the control variables general ad likability (r = 0.572, N = 68, p < 0.001), general brand attitude of a person (r = 0.549, N = 68, p < 0.001) and income (r = 0.330, N = 68, p = 0.006). The control variables sex, age and educational background were insignificant. As of this, general ad likability, income and general brand attitude were controlled for within the follow-up analyses.
5.2 Reliability of scales
A principal component analysis (PCA) showed that the 6 items of the scale “attitude toward the brand BMW after ad exposure” formed a single uni-dimensional scale with only one component having an eigenvalue above 1 (EV = 4.47, R2 = 74.54). All items correlated
positively with the first component, the variable “I think the brand BMW is
worthless/valuable” had the strongest association (factor loading was .90). Omitting items would not have improved the scale’s reliability, which was very good, Cronbach’s alpha = .93.
In order to test the scale “general ad skepticism”, a PCA indicated that the 8 items formed a single uni-dimensional scale with only one component having an eigenvalue above 1 (EV = 4.96, R2 = 62.00). The items correlated positively with the first component, the
variable “Advertising is generally truthful” had the strongest association (factor loading was .86). Omitting any of the items would not have improved the reliability of the scale. The scale’s reliability was very good, Cronbach’s alpha = .90.
The reliability of the scale “general attitude toward the brand BMW” was also
analyzed and showed that 7 items formed a single uni-dimensional scale. It showed that only one component had an eigenvalue above 1 (EV = 5.01, R2 = 71.58). All items correlated
positively with the first component. The variable “I think the brand BMW is very
unenjoyable/enjoyable” had the strongest association (factor loading was .89). The reliability of the scale was very good and would not have been improved by deleting any of the items, Cronbach’s alpha = .93.
Finally, the scale that presented the participants’ environmental concerns was tested. In contrast to prior analyses, it showed that the 10 items formed two components with eigenvalues above 1 instead of only one (factor one: EV = 4.23, R2 = 42.25, factor two: EV =
purchase the one less harmful to other people and the environment” loaded the highest for component one (factor loading was .86), for component two the item “I have given money to an environmental group” loaded the highest (factor loading was .71). The reliability analysis showed that the scale’s reliability would have been improved only marginally when the item “I have written a letter or called the member of congress or another government official to support strong environmental protection” was deleted (Cronbach’s Alpha = .83). Since the reliability of the scale was already very high (Cronbach’s alpha = .83) and the item deletion would have solely improved the scale by .002, the scale was not changed and remained as already used in existing research.
5.3 Randomization Check
A randomization check was conducted in order to analyze whether variables such as gender, age or income and education levels were equally distributed among all conditions (condition 1 = green ads; condition 2 = non-green ads; condition 3 = no ad at all/ control group) as used in this study. Results of a one-way ANOVA showed that the variables “sex” (F (2,102) = .51, p = .603), “age” (F (2,102) = .51, p = .603),
“level of income” (F (2,102) = .09, p = .913), “brand attitude in general” (F (2,102) = 1.38,
p = .257) and “educational level” (F (2,102)
= .04, p = .961) were insignificant.
Furthermore, the analysis showed that all of the variables were equally distributed among all conditions (see Table 1). Therefore, it could be assumed that the randomization was successful.
5.4 Main analyses
The first hypothesis of this study examined whether green car advertisements led to more positive brand attitudes as compared to non-green car ads. An analysis of covariance (ANCOVA) was carried out in order to assess the influence of green car ads on brand
attitudes after the exposure while taking into account the covariates “general brand attitude”, “general ad likability” and “level of income”. The results showed that there was no significant main effect of the condition on a person’s brand attitude after the ad exposure (F (1,65) = .30,
p = .589). On average, green car ads scored insignificantly higher (M = 3.61, SD = 1.12) as
compared to non-green car ads (M = 3.26, SD = 1.06, p = .589). Hence, the first hypothesis was rejected. The covariates ad likability (F (1,65) = 17.63, p < .001) and general brand attitude toward BMW (F (1,65) = 11.10, p < .001) had a significant positive effect on the participants’ brand attitudes after the exposure of the ad. The covariate income was
marginally significant (F (1,65) = 2.94, p = .09). Table 2 shows the results as described above (as well as the results of those ANCOVAs being run for H3 and H4 which are explained later
Table 2: Results of the study’s ANCOVAS testing on the dependent variable brand attitudes including moderators and covariates (N = 67)
Variables Type III Sum of Squares df Mean Square F p ANCOVA H1 Ad Manipulation (Green vs. non-green ads) 0.18 1 0.18 17.14 0.589
General Brand Attitude 6.78 1 6.78 11.10 0.001***
General Ad Likability 10.77 1 10.77 17.63 0.000*** Income 1.79 1 1.79 2.94 0.092 Error 37.87 62 0.61 Total 872.75 67 ANCOVA H3 Ad Manipulation (Green vs. non-green ads) 0.88 1 0.88 1.21 0.282
General Brand Attitude 0.01 1 0.01 0.01 0.926
General Ad Likability 9.50 1 9.50 13.17 0.001*** Income 0.47 1 0.47 0.65 0.430 Environmental Concerns 10.50 28 0.38 0.52 0.950 Ad Manip.*Environmental Concerns 10.60 11 0.96 1.34 0.268 Error 16.59 23 0.72 Total 872.75 67 ANCOVA H4 Ad Manipulation (Green vs. non-green ads)
0.07 1 0.07 0.17 0.681
General Brand Attitude 1.68 1 1.68 3.91 0.059
General Ad Likability 9.88 1 9.88 23.07 0.000*** Income 0.25 1 0.25 0.59 0.451 Ad Scepticism 6.38 21 0.30 0.71 0.787 Ad Manip.*Ad Skepticism 6.45 13 0.50 1.16 0.361 Error 10.71 25 0.43 Total 805.72 64 Note: * p ≤ .05; ** p ≤ .01; *** p ≤ .001
In order to examine whether participants who were exposed to green ads were more likely to have higher levels of brand attitude when showing a high level of environmental concerns than those participants with a low level of environmental concern (H3), an ANCOVA was
run. Again, the covariates “general brand attitude”, “general ad likability” and “level of income” were taken into account. Results showed that there was no significant main effect of condition on a person’s brand attitude after the ad exposure (F (1,23) = 1.21, p = .282). The covariate ad likability (F (1,23) = 13.17, p = .001) had a significant positive effect on the participants’ brand attitudes after the exposure of the ad while the general brand attitude toward BMW (F (1,23) = .01, p = .926) and income (F (1,23) = .65, p = .430) were not significant. The variable ‘environmental concerns’ had no insignificant interaction effect on
brand attitudes (F (1,23) = 1.34, p = .268) (see Table 2). Hence, the third hypothesis was rejected.
A final ANCOVA was carried out in order to examine H4. It was tested whether
participants who were exposed to green ads had lower levels of brand attitudes when showing high levels of skepticism toward advertising as compared to those participants with low levels of ad skepticism. The results indicated that there was no significant main effect of condition on a person’s brand attitude after the ad exposure when controlling for levels of ad skepticism and the covariates income, ad likability and general brand attitude (F (1,25) = .17, p = .681). The covariate ad likability (F (1,25) =23.07, p < .001) had a significant positive effect on the participants’ brand attitudes after the exposure of the ad, the general brand attitude toward BMW a marginally one (F (1,25) = 3.91, p = .059). Income (F (1,25) = .59, p = .451) and ad skepticism were insignificant (F (1,25) = .71, p = .787). Ad skepticism had no significant interaction effect on brand attitudes (F (1,25) = 1.16, p = .361) (see Table 2). Hence, the fourth hypothesis was rejected.
H2 dealt with the question whether more positive brand attitudes lead to higher
purchase intentions after controlling for the participant’s income, his or her general likability of the ad and the attitude toward the brand BMW. This was examined by running a
hierarchical multiple regression analysis. Firstly, preliminary analyses were conducted in order to guarantee that no violation of the assumptions of normality, linearity or
homoscedasticity occurred. The assumptions were all satisfied as residual and scatter plots showed. Secondly, correlations between all of the variables were examined. Results showed that all predictor variables were significantly correlated with the dependent variable purchase intention toward the advertised product. The correlations were moderate to strong, ranging between r = .33, p = .003 and r = .64, p < .001. According to Tabachnick and Fidell (2001), only very strong or perfect strengths of associations can indicate multicollinearity. Hence, due
to the moderate and strong correlations between the variables, multicollinearity was unlikely to be a problem. As of this, it can be said that an examination of H2 through a hierarchical
multiple regression was suitable.
In the first step of the hierarchical multiple regression, three predictors were added: general ad likability, general attitude toward the brand and level of income. Results indicated that this model was statistically significant (F (3,66) = 17.92, p < .001), and explained 44 % of the variance in purchase intention toward the product after the ad exposure. In the second step of the model, the variable “brand attitudes after the ad exposure” was added. Now, the total variance explained by the model as a whole was 48 % (F (4,66) = 15.98; p < .001). By introducing brand attitude after the ad exposure, an additional variance of 4 % in purchase intention toward the advertised product could be achieved, after controlling for general ad likability, general brand attitude and level of income (R2 Change = .05; F (1,62) = 5.94; p =
.018). In the final model only one predictor, namely brand attitude after the ad exposure, remained statistically significant with a Beta value of β = .33, p = .013. Table 3 shows the results of the hierarchical regression model of purchase intention after the ad exposure as described above.
Note: * p ≤ .05; ** p ≤ .01; *** p ≤ .001
6. Conclusion and discussion
The purpose of this study was to find out whether green advertisements for e-cars had a more positive impact on brand attitudes toward the advertised brand as compared to non-green ads, whether these brand attitudes had any impact on purchase intentions toward the advertised product and whether environmental concerns and general skepticism toward advertising had a moderating influence on these relations.
Overall, results indicated that green ads had no significant positive effect on brand attitudes in comparison to non-green advertisements. H1 was rejected. This stands in contrast
to prior studies, which indicated that green ad claims lead to more positive attitudes and favorable perceptions of the brand as compared to non-green claims (Bech-Larsen, 1996; Benito Gómez et al., 1999; Davies, 1993; Eagly & Kulesa, 1997; Hartmann et al., 2005; Phau & Ong, 2007; Swenson & Wels, 1997). One reason for the opposing results could be that the consumers’ resistance toward green claims has become too high due to untrue green claims: as already described before, marketers mostly use meaningless and unproven green claims in these ads in order to have a reason to charge higher prices for their supposedly “green”
R R2 R2 Change B SE β T Model 1 .46 .44 General brand attitude .63 .27 .31** 2.79 General ad likability .43 .12 .38*** 3.49 Income .16 .75 .21* 2.13 Model 2 .51 .48 .05 General brand attitude .40 .24 .19 1.68 General ad likability .27 .13 .24 1.99 Income .12 .07 .15 1.62 Brand attitude after ad exposure .59 .23 .33* 2.55
products. Due to the exposure to untrue ads in the past, the resistance on the consumers’ side could have grown very high, which in turn could have led to the resistance among the study’s participants toward the shown green ads. This could have led to an underperformance of green claims as compared to non-green claims (Crane & Peattie, 2005; Simintiras & Martin, 1995; Wong et al., 1996). As research showed, people cope with untrue advertising claims by applying various resistance strategies unconsciously and in some situations also consciously: they react to perceived false ad claims by counterarguing against the statements in an
advertisement or by derogating the ad source (e.g. the brand BMW) in order to protect their own attitude (Ahluwalia, 2000; Zuwerink Jacks & Cameron, 2003). Possibilities for
practitioners to overcome this resistance could be to push the choice into the future (e.g. the choice to purchase a green car) as customers would feel that they have more time to think about their buying behavior, more freedom to decide on purchasing such a high priced product and no pressure on purchasing the car. For instance, this can be achieved by
incorporating messages that make it clear to the customer that the e-car can be tried first for multiple days before one has to purchase it or that only parts of the high price for the e-car needs to be paid when purchasing it while the rest of the price can be paid in the following months (Knowles & Linn, 2004). Alternatively, instead of showing advertisements, brand placement (e.g. the subtle placement of an e-car in entertainment programs or media content) could lead to more general positive brand attitudes, because the learning effect can take place entirely outside of someone’s conscious awareness (Matthes, Schemer & Wirth, 2007). Another reason for the insignificant results could be that green claims are not effective in the car industry due to the very unique green product “e-car”: prices for electronic vehicles are generally high and require high cognitive thinking before attitudes or purchase intentions can be changed. In other areas such as clothing or nutrition with lower priced green products, the decision-making processes might be easier. People might feel that purchasing green clothes or
green nutrition products requires less investment risk and attitudes and purchasing intentions could be positively influenced more easily by green advertisements.
Furthermore, environmental concerns and ad skepticism had no moderating influence on brand attitudes. Hence, H3 and H4 were rejected. Previous research indicated that
individuals with high levels of environmental concerns were more likely to show positive attitudes toward green brands as compared to those with low levels of environmental concerns. They indicated that environmental concerns as well as ad skepticism had
moderating influences on brand attitudes: the higher the environmental concerns of a person were, the more positive was their brand attitude (Antil, 1984; Black et al., 1985; Chan, 1996; Donaton & Fitzgerald, 1992; Dunlap & Van Liere, 1984; Fransson & Gärling, 1999; Hines et al., 1987; Kerr, 1990; Kim & Lee, 2009; Obermiller et al., 2005; Ottman, 1993; Schlossberg, 1992; Shetzer et al., 1991; Tucker et al., 2012). Again, this study showed opposing results and stood in contrast to Schwartz’s (1973) moral norm-activation theory of altruism that stated that a person might enact pro-environmental behavior when being confronted with perceived threats such as environmental pollution as described before. However, due to the complex structure of environmental concerns, it remains unclear why some people enact and some people do not enact in pro-environmental behavior: as Kollmuss and Agyeman (2002) stated, hundreds of studies in the past years have tried to understand the complexity of this
phenomenon but failed to give a clear answer. Nevertheless, most of the studies made demographic, external (e.g. social and cultural) and internal factors (e.g. motivation,
awareness, values, emotions, locus of control, responsibilities etc.) accountable for it as pro-environmental behavior or behavioral intentions are strongly dependent on these factors (Kollmuss & Agyeman, 2002).
Research indicated that ad skepticism occurs as consumers are overwhelmed with ads and coping with multiple advertising claims simultaneously requires effort. Instead of
evaluating the ads’ information, consumers react with disbelief or skepticism toward the ad claim in order to simplify the processing (MacInnis et al., 1991). As studies have shown, skepticism moderated responses such as brand beliefs or attitudes toward the brand, which in turn influenced a person’s purchase intention (Obermiller & Spangenberg, 2000; Obermiller et al., 2005). In contrast, the present study found opposing results showing that ad skepticism did not moderate brand attitudes. Two reasons can account for this: first, the ad being shown in the present study incorporated emotional green claims (e.g. the slogan “Clean car. Clean planet. Clean conscience.”). Matthes and Wonneberger (2014) indicated that emotional appeals of green ads had no impact on green ad skepticism and green consuming intention. Second, there are more recent studies nowadays that refute prior assumptions of earlier studies as described above: Matthes and Wonneberger (2014) for instance stated that there was no positive relationship between green consumerism/ green consuming intention and general ad skepticism: A survey of Austrian consumers showed that green consumers perceive green ads to be more informationally useful as compared to non-green consumers. As of this, their green advertising skepticism decreased (Matthes & Wonneberger, 2014).
Finally, this study showed that the more positive brand attitudes toward the advertised brand BMW were, the higher was the purchase intention toward the advertised product. Therefore, H2 was approved and results were in line with the TRA stating that the consumer’s
attitude toward a brand or product is the best antecedent for predicting behavioral intentions. Furthermore, this was also approved by other studies, which stated that brand attitudes are one of the best antecedents for purchase intentions toward green products (Kalafatis et al., 1999; Vermeir & Verbeke, 2006). Results also showed that other factors such as the
participants’ levels of income, their already existing brand attitudes toward BMW and general likability of the ad had a significant positive impact on the brand attitude. This is in line with other existing studies, which showed that general ad likability, for instance, significantly
predicted brand attitudes (Bergkvist & Rossiter, 2008). Additionally, socio-demographics and prior existing attitudes can have an effect on follow-up attitudes as prior research has
indicated (Diamantopoulos, Schlegelmilch, Sinkovics & Bohlen, 2003).
Taken together, the results of this study show that green advertisements in the car industry are not effective. Consumers are not interested in how ‘green’ a car is or which environmental benefits it has as compared to non-green cars. Instead, the clients’ intentions and attitudes are marginally dependent on already existing attitudes toward the brand as well as on income levels. Additionally to this, it is more relevant to clients whether they enjoy and like the ad in general as compared to whether it contains green claims. Ad likability should therefore play a bigger role for practitioners: they need to create ads, which are enjoyable and lead to high likability among viewers instead of demonstrating environmental values of a car. A possibility of how to increase likability could be the incorporation of well-known
celebrities with a positive image in the ad that is transferred to the car brand or the insertion of humor (Erdogan, 1999; Weinberger & Gulas, 1992).
The results of this study are substantial and contribute to existing literature. However, this study also has a few limitations. Firstly, it only included German participants. Therefore, results are hard to generalize to other populations and are only specific for one country. Secondly, age groups were not equally distributed. The majority of the respondents were either between 30 and 40 or above 60 years. Therefore, the group of 40 to 60 year olds was underrepresented. Thus, the chance exists that equally distributed age groups could have led to different results: participants aged between 40 and 60 years are often more likely to have a higher salary compared to 30 year olds or participants above 60 (who might already be retired) and are therefore more likely to be able and willed to afford a high priced e-car. The intentions to buy, and in turn the attitudes toward the brand, which affect the level of purchase
intentions directly as shown in this study, could have been stronger if this group would have not been underrepresented.
This study has shown that green advertisements are not successful in the car industry. Brand attitudes were not improved by green ads as compared to non-green ads and ad
skepticism as well as environmental concerns had no moderating influence on this relation. Instead, socio-demographic factors, prior brand attitudes and the likability of an ad seemed to influence the brand attitude significantly, which in turn significantly influenced a person’s purchase intention. Therefore, the likability of an ad and general high brand attitudes among customers should be generated in order to make green products more successful in the car industry.
Bibliography
Albayrak, T., Caber, M., Moutinho, L. & Herstein, R. (2011). The Influence of Skepticism on Green Purchase Behavior. International Journal of Business and Social Science,
2(13), 189 – 197.
Ahluwalia, R. (2000). Examination of Psychological Processes Underlying Resistance to Persuasion. Journal of Consumer Research, 27(2), pp.217-232. DOI: 10.1086/314321. Antil, J. H. (1984). Socially Responsible Consumers: Profile and Implications for Public
Policy. Journal of Macromarketing, 4(2), 18-39. DOI: 10.1177/027614678400500203. Arnocky, S., Stroink, M. & DeCicco, T. (2007). Self-construal predicts environmental
concern, cooperation, and conversation. Journal of Environmental Psychology, 27, 255-264. DOI: 10.1016/j.jenvp.2007.06.005.
Bakker, S., Sierzchula, W., Maat, K. & Van Wee, B. (2012). Technological diversity of emerging eco-innovations: a case study of the automobile industry. Journal of Cleaner
Production, 37, 211-220. DOI: 10.1016/j.jclepro.2012.07.011.
Bech-Larsen, T. (1996). Danish consumers’ attitudes to the functional and environmental characteristics of food packaging. Journal of Consumer Policy, 19, 339-63. DOI: 10.1007/bf00411413.
Bergkvist, L. and Rossiter, J. (2008). The Role of Ad Likability in Predicting an Ad's Campaign Performance. Journal of Advertising, 37(2), pp.85-98. DOI:
10.2753/joa0091-3367370207.
Benito Goméz, C., Noya, F. J. & Paniagua, A. (1999). Actitudes y comportamientos hacia el medioambiente en Espan ̃a. Opiniones y Actitudes, 25, 244-250.
Biel, A. L. & Bridgwater, C. A. (1990). Attributes of Likable Television Commercials.
Journal of Advertising Research, 30(6), 38-44. Retrieved online at 7th January 2015 from:
http://search.ebscohost.com/login.aspx?direct=true&db=ufh&AN=9605212068&site= ehost-live
Black, J. S., Stern, P. C. & Elworth, J. T. (1985). Personal and contextual influences on household energy adaptations. Journal of Applied Psychology, 70, 3-21. DOI: 10.1037//0021-9010.70.1.3.
Bryman, A. (2012). Social research methods. 4th edition. Oxford: Oxford university press.
Chan, T. S. (1996). Concerns for Environmental Issues and Consumer Purchase Preferences: A Two-Country study. Journal of International Consumer Marketing, 9, 43-55. DOI: 10.1300/j046v09n01_04.
Chang, Ch. (2003). Party Bias In Political-Advertising Processing. Journal of Advertising,
32(2), 55-67. DOI: 10.1080/00913367.2003.10639129.
Cox, D. S. & Cox, A. D. (1988). What Does Familiarity Breed? Complexity as a Moderator of Repetition Effects in Advertisement Evaluation. Journal of Communication
Research, 15(6), 111-16. DOI: 10.1086/209149.
Cox, D. S. & Cox, A. D. (2002). Beyond First Impressions: The Effects of Repeated Exposure on Consumer Liking of Visually Complex and Simple Product Designs.
Journal of the Academy of Marketing Research, 30 (32), 119-130. DOI:
10.1177/03079459994371.
Crane, A., Peattie, K. (2005). Green marketing: legend, myth, farce or prophesy? Qualitative
marketing research: An international Journal, 8(4), 357-370. DOI:
10.1108/13522750510619733.
Davies, J. J. (1993). Strategies for environmental advertising. Journal of Consumer
Marketing, 10(2), 19-36. DOI: 10.1108/07363769310039102.
Diamantopoulos, A., Schlegelmilch, B., Sinkovics, R. and Bohlen, G. (2003). Can socio-demographics still play a role in profiling green consumers? A review of the evidence
and an empirical investigation. Journal of Business Research, 56(6), pp.465-480. DOI: 10.1016/s0148-2963(01)00241-7.
Dijkstra, M., Buitjtels, H. E., & Van Raaij, W. F. (2005). Separate and joint effects of medium type on consumer responses: a comparison of television, print, and the internet. Journal of Business Research, 58(3), 377-386. DOI: 10.1016/s0148-2963(03)00105-x.
Donaton, K., Scott, L. & Fitzgerald, K. (1992). Polls Show Ecological Concerns is Strong.
Advertising Age, 63(15), 49.
Dunlap, R. E. & Van Liere, K. D. (1984). Commitment to the dominant social paradigm and concern for environmental quality. Social Science Quarterly, 64, 1013-1028.
Eagly, A. H. and Kulesa, P. (1997). Attitudes, attitude structure, and resistance to change.
Environment, Ethics, and Behavior, 5, 122-53.
Eggers, F. & Eggers, F. (2010). Where have all the flowers gone? Forecasting green
trends in the automobile industry with a choice-based conjoint adoption model.
Technological Forecasting & Social Change, 78, 51-62. DOI: 10.1016/j.techfore.2010.06.014.
Erdogan, B. Z. (1999). Celebrity endorsement: A literature review. Journal of Marketing
Management, 15, 291-314. DOI: 10.1362/026725799784870379.
Fransson, N., & Gärling, T. (1999). Environmental concern: Conceptual definitions,
measurement methods, and research findings. Journal of Environmental Psychology,
19, 369–382. DOI: 10.1006/jevp.1999.0141.
Fraunhofer Society & PriceWaterhouseCoopers (2010). Elektromobilität –
Herausforderungen für Industrie und öffentliche Hand. Study by the Fraunhofer
