The Effect of Celebrity Endorsement
on Purchase Intentions for
Fairtrade Products
The Social Dimension of Sustainability
Laura Bremer
The Effect of Celebrity Endorsement
on Purchase Intentions for
Fairtrade Products
The Social Dimension of Sustainability
Laura Bremer
Faculty of Economics and Business
Department of Marketing
Master Thesis
26.06.2017
Hauptstraße 36 – 31812 Bad Pyrmont – Germany +49 176 76 79 70 32
laurabremer@t-online.de S2911612
Dr. Wander Jager Dr. Jan Willem Bolderdijk
Abstract
Celebrity endorsement is a prevalent marketing strategy, in which companies often invest millions of Euros every year to create an associative link between their goods and the endorser in order to transfer the image and characteristics of the endorser onto the company’s product/service/brand. Many researchers have investigated the effectiveness of celebrity endorsement and it was found, that celebrities are more efficient in endorsing, than an unknown person. Being a celebrity is not enough to be an effective endorser. Some researchers state that there must be a match between the celebrity and the product/service/brand, while other researchers emphasize the importance of the attractiveness of a celebrity. As these studies have mainly been invested in the field of profit-oriented companies, there is a huge gap in the literature about how non-profit organizations have to deal with celebrity endorsement as a marketing strategy.
The current research contributes to the investigations of celebrity endorsement in the field of non-profit organizations, by testing the influence of celebrity endorsement on the purchase intention for pro-environmental products at hand of the non-profit-organization Fairtrade. The research also adds the dimension of credibility of the celebrity, as well as investigates if the purchase intention for Fairtrade products is increased due to a change in the attitude towards Fairtrade by the application of an endorsing celebrity. Promoting Fairtrade products is highly relevant, since such non-profit organizations contribute a lot to sustainably and ethically correct working conditions in developing countries. Non-profit organizations have to come up with persuading marketing strategies to stand out of all the other advertisements of profit-oriented organizations. These strategies should help to make people become aware of these products and make them more willing to purchase them. The results of this study indicate that although some celebrities are rated higher in expertise, trustworthiness and attractiveness regarding Fairtrade, they do not have a different impact on people’s purchase intention for Fairtrade products. Thus, a celebrity who is related to pro-environmental behaviour does not make people more likely to purchase Fairtrade products than an individual who is either a celebrity without any pro-environmental relation or a non-celebrity. But it has been shown, that when Fairtrade products are endorsed by a celebrity, this has a positive effect on the individual’s attitude towards Fairtrade, which in turn increases his/her likelihood to purchase Fairtrade products. Next to these findings, this research gives several advises to future researchers in the field of celebrity endorsement’s influence on the purchase intention for Fairtrade products.
Key words: Fairtrade, Fairtrade Products, Celebrity Endorsement, Credibility of Celebrity,
Table of Content
1 Introduction ... 1
1.1 Fairtrade ... 2
1.2 Marketing Strategy: Celebrity Endorsement ... 4
1.3 Research Question ... 5
2 Theoretical Framework ... 6
2.1 Celebrity Endorsement ... 6
2.2 Credibility of Celebrities ... 9
2.3 Attitude towards Fairtrade Products ... 13
3 Conceptual Model ... 15
4 Research Design ... 16
4.1 Data Collection Method ... 16
4.2 Pre-Survey ... 16
4.2.1 Participants and Procedure ... 18
4.3 Main Survey ... 18
4.3.1 Participants and Procedure ... 20
4.3.2 Plan of Analysis ... 20
5 Results ... 23
5.1 Frequencies ... 23
5.2 Reliability Analysis ... 23
5.3 Descriptive Statistics of the Sample ... 24
5.4 Manipulation Check ... 25
5.5 Testing for Hypothesis 1 ... 26
5.6 Testing for Hypothesis 2 ... 28
5.7 Testing for Hypothesis 3 ... 29
6 Conclusion & Recommendations ... 31
6.1 Limitations ... 32
6.2 Recommendations ... 34
6.3 Conclusion ... 35
7 References ... 37
7.1 Books and Journals ... 37
7.2 Pictures ... 41
7.3 Software ... 42
7.4 Websites ... 42
1
1 Introduction
As current concerns like the global warming, refugee influx, the huge difference between poor and rich, as well as many more sustainability issues question our behaviour every day, it becomes more and more important to conduct research which supports societal sustainability and pro-environmental behaviour. The marketing of sustainable products, such as solar panels, electric cars, organic food or ethically produced products, can be a powerful tool in making people more conscious about their behaviour and how they can contribute to the sustainability of our planet earth (Pickett-Baker & Ozaki, 2008).
But what exactly are sustainable products? The concept of the triple bottom line (TBL) explains this at hand of three dimensions: People, Planet and Profit (three Ps) (Slaper & Hall, 2011). It is a model which helps companies to measure their performance in terms of the three different dimensions. The People dimension measures how social and responsible a company acts. The Planet dimension helps to indicate how responsible a company operates in terms of the environment. And lastly, the Profit dimension, which accounts for the profits of the company (The Economist, 2009). When all three dimensions are well balanced, a company is sustainable in the long-run (Simmons, 2014). A sustainable product is therefore a product, which brings benefits from all dimensions.
2
the development of caring consumers (since the 1990s), the pressure by reference groups to support fair trade in the Third World, the interest of the media in fair trade, the increasing role of the concept of CSR (Corporate Social Responsibility) and the increasing power of suppliers (Strong, 1997).
1.1 Fairtrade
The internationally largest and best-known alternative approach to conventional trade is called Fairtrade. It is a fair-trade system, which consists out of Fairtrade International and 23 other organizations (fairtrade.net, About Fairtrade, 2017). Fairtrade’s geographical scope covers mainly Africa and the Middle East (here mainly parts of Western Africa), Asia and Pacific (here mainly parts in Pacific), and
Latin America and the Caribbean (here mainly parts from South America) (fairtrade.net, Our Standards, 2017). Now, there are approximately 1,6 Mio. certified Fairtrade farmers and workers around the whole globe. 46% of all farmers and workers of the Fairtrade certified plantations are women (fairtrade.net, Monitoring & Impact Reports, 2017). The non-profit, multi-stakeholder association works based on partnerships between producers, traders, businesses and consumers.
The main goal of the organization is “[…] to reform international trade rules and create a fairer economic system.” (fairtrade.net, Our Vision & Mission, 2017). Thereby the organization sets specific standards and better trading conditions so that small-scale farmers in poor countries get paid in a fair way. These are standards for small producer organizations, hired labour, contract production, as well as trader, textile and climate standards. They face poverty and empower farmers and producers (fairtrade.net, Standards, 2017). Additionally, the Fairtrade organization gives with the textile (prohibiting materials like chemicals etc.) and climate standards (offering companies to purchase Fairtrade Carbon Credits) a start to protect the environment and to become resilient against the climate change as well.
Hence, although Fairtrade’s main goals are directed towards better ethical standards in developing countries, thus the People dimension of the three Ps, they also partly cover pro-environmental activities from the Planet dimension.
3
In total, Fairtrade’s product portfolio includes 17 different product industries: bananas, cocoa, coffee, cotton, flowers, sugar, tea, composite products, carbon credits, fresh fruit, gold, honey, juices, rice, spices and herbs, sports balls and wine (fairtrade.net, Products, 2017). The highest sales volumes have been generated by bananas, cane sugar, cocoa and coffee (fairtrade.net, Monitoring & Impact Reports, 2017). Fairtrade products can be recognized by the typical Fairtrade logo (see Figure 1). This logo is used by marketers to emphasize that the product is produced based on ethical principles. It is a marketing tool, which helps to influence an individual’s purchase intention.
Particularly, consumers’ motives to buy ethical products are still debatable. Some authors found that people buy them due to self-interest as well as ethical motivations (Yamoah et al., 2014), while others discovered habits as well as self-satisfaction as a reason (Davies & Gutsche, 2016). David and Gutsche (2016) particularly found that people feel happy when buying fair trade products, since they fulfil responsibility amongst themselves as well as by exercising social justice.
Until now, many studies have been undertaken testing what drives peoples’ buying decisions (Stern, 1999). Using the reputation of well-known logos (called leveraging of secondary associations in a marketing manner), as the one of the Fairtrade organization, is only one tactic. Others can be product placement, price promotions, event marketing and sponsorship, give-aways or celebrity endorsement.
To encounter concerns as for example global poverty, ethical products need more attention and sales should be increased. Therefore, the role of marketing in this field becomes more and more crucial and it has to be found a way in which those products can be sold best.
4
1.2 Marketing Strategy: Celebrity Endorsement
In celebrity endorsement, a celebrity is used to communicate information (feelings, expertise as an argument etc.) about a product/service/brand via an advertisement to the audience. It should increase acceptance and reception of the communicated message (Cohen, 1972). Some researchers also view them as a reference group, which is used by individuals as evidence concerning values and what is right or wrong in public (Fireworker & Friedman, 1977). Celebrity endorsement is usually part of a whole marketing campaign, which also incorporates other marketing mix activities.
5
attend activities of non-profit organizations, donate their money to them or found their own charitable organizations, either for their good-will and/or their image. Emma Watson for instance, shows her special engagement in ethical production, because she created together with Alberta Ferretti an organic and fair-trade clothing line. Celebrities showing such an engagement create a specific image towards pro-environmental behaviour and are rather perceived as experts in this field. Therefore, they fulfil the most important criteria to become a celebrity endorser for pro-environmental and non-profit organizations and could help to promote sustainable/ethical products very much (Ecorazzi, 2011).
1.3 Research Question
With the background of ethical, hence Fairtrade, products and the possibility to use celebrities, having a connection to pro-environmental behaviour, as endorsers for such products, the following research question will be the focus in this research paper:
6
2 Theoretical Framework
In the following sections, the concepts used in this study will be discussed separately. Each explained concept bears a new hypothesis. Afterwards (in section 3), the Conceptual Model for this study is shown and shortly explained.
2.1 Celebrity Endorsement
Celebrity endorsement is a prevalent marketing strategy, which is used since the late nineteenth century (Erdogan, 1999). Generally, it can be seen as part of the marketing mix, which also incorporates branding, pricing, placing, and other promotional activities, e.g. the leveraging of secondary associations, direct marketing, digital marketing, social media, etc. Celebrity endorsement helps advertisers to create an associative link between a brand and a celebrity in order to project the feelings (favourably positive ones), associated with the celebrity, on the brand (Till & Shimp, 1998). There is evidence, that on average, celebrity endorsement has a positive impact on stock returns and it is generally seen as a valuable marketing investment (Agrawal & Kamakura, 1995). Further research also reveals that celebrities are more effective than non-celebrities (Mathur, Mathur & Rangan, 1997). Those mainly used for marketing reasons are athletes, actors/actresses, singers (Ohanian, 1990), product class experts or the company president (Byrne, Whitehead & Breen, 2003). Their roles in advertisement can vary a lot. They can be utilized as actors, spokesperson, endorsers and testimonials (Erdogan, 1999).
While celebrity endorsement can support marketing campaigns, it can also bear some risks. The authors Till and Shimp (1998) state that companies do not have control over a celebrity’s future behaviour, especially for brands, where the celebrity is the main attribute for the consumer to evaluate the brand on.
7
But how does celebrity endorsement exactly work? What is the ‘magic secret’ behind this marketing strategy?
Recent literature explains how authors tried to analyse the way in which consumers process advertisements, including those who utilize celebrity endorsers. Petty and Cacioppo (1986) built much of the ground floor to this literature by establishing the Elaboration Likelihood Model (ELM). It shows the impact of involvement on an individual’s response to an advertisement. Specifically, it states that the effectiveness of communication tools, thus of inducing persuasion, depends on the level (either low or high) of the elaboration likelihood of the communication situation.
A low elaboration likelihood, leaves to assume that the peripheral route is more appropriate (Petty, Cacioppo & Schumann, 1983). This route is mainly gone with the help of so called peripheral cues, i.e. the use of an emotional link between consumer and product/service/brand, celebrities, images, attractiveness or expertise of the communication source (Szmigin & Piacentini, 2015; Chaiken, 1980; Petty, Cacioppo & Goldman, 1981). The name of a product/service/brand and the visual impression of a celebrity can, in terms of low-involvement products like chocolate and bananas (latter one is considered in this study later), be the main elements quickly processed by a consumer. However, textual arguments, either strong or weak, will be ignored and have no influence on the individual’s purchase intention or final buying decision (Kahle & Homer, 1985). Thus, when celebrity endorsement is used to promote low-involvement products, it rather seems to be processed as a peripheral cue via the peripheral route.
8
By using a celebrity endorsing a product/service/brand, either as peripheral cue or as a strong argument, the advertisement has the chance to stand out of the other advertisement clutter and therefore to reach a bigger audience (Sherman, 1985). This is especially difficult, if a product should be launched in different countries. Celebrities who are worldwide known can help companies to overcome barriers in terms of language and culture (Erdogan, 1999). They can be very effective, due to their “[…] symbolic aspirational reference group associations.” (Byrne et al., 2003). Which means that they represent a specific group (aspiration group), of which others (the audience) desire to belong to (Escalas & Bettman, 2003). Especially people, who are prone to following what others do and attach a lot of importance to others’ opinions about them, are easily influenced by such reference groups.
Transferring the latter arguments onto the non-profit organization Fairtrade, since people need to become more aware of ethical products, the usage of a celebrity might be helpful in promoting these products to stand out of the other advertisements and to reach more attention. Since Fairtrade is an internationally operating organization, the usage of an internationally known celebrity could be even handier, since this could help to save money for the non-profit organization (only one celebrity is needed), and simultaneously overcome linguistic and cultural barriers when introducing the advertisement campaign worldwide. The celebrities used in the Fairtrade promotion are expected to serve as a role model, so that the audience learns a lesson from them, and parallelly aspires to be like him/her, which finally persuades and leads to a purchase of the Fairtrade product.
Based on the findings in the literature cited above, it is assumed that a celebrity is more persuading and therefore more likely to positively influence peoples’ purchase intention for Fairtrade products. Therefore, the following hypothesis has been stated:
9
This hypothesis represents the main effect of the conceptual model. In the sections discussed in the following, concepts will be introduced, which represent additions to this main effect, which might empower the influence of celebrity endorsement on the purchase intention for Fairtrade products.
Topic of the section below, is credibility of celebrities, which means the perception of an individual about a celebrity’s expertise, trustworthiness and attractiveness. Also, the fit between a celebrity and the product/service/brand will be discussed. A high degree in credibility as well as a good celebrity-product congruence can help to strengthen the effectiveness of celebrity endorsement, as well as to overcome some of the named hurdles and cons companies face when using celebrity endorsement.
2.2 Credibility of Celebrities
Celebrity endorsement is most effective when the celebrity is similar or related to a product’s/service’s/brand’s attributes (Kahle & Homer, 1985). This match also has an impact on his/her credibility. If a celebrity has a specific expertise, which can be related to the product itself or it’s usage, the credibility of the endorser increases immediately and is therefore more persuasive. The characteristics of a communicator determine how he/she can persuade the audience, and it is difficult to find the appropriate person for a product/service/brand (Ohanian, 1990).
The credibility of a communication source, as in this study the celebrity, has been widely investigated by many studies, whereby every researcher had his/her own definition for the credibility of a source and measured it on differing sets of dimensions. The mostly appearing dimensions have been sociability, trustworthiness, expertness, competence, attractiveness and likability (Ohanian, 1990). These dimensions are comprised in the source-credibility model and the attractiveness-based model, which represent the fundamental theory behind the concept of celebrity endorsement. In the source-credibility model the persuasiveness of a message depends on its extent of expertise and trustworthiness (Fennis & Stroebe, 2016). The source-attractiveness model states that the effectiveness of celebrity endorsement relies on the celebrity’s familiarity, similarity, likability and attractiveness to the audience (Maddux & Rogers, 1980).
10
predicated on psychometrically accepted procedures, Ohanian (1990) developed a valid scale to measure the credibility of a celebrity endorser. This scale can additionally be used for finding the most appropriate average consumer or to evaluate in instance politicians. The three components of the scale are trustworthiness, expertise and attractiveness.
Trustworthiness is the extent to which an individual accepts and is confident of a message.
It can have a huge effect on someone’s degree of persuasion and attitude towards the product/service/brand (Ohanian, 1990). Expertise, also applied to authoritativeness, competence, expertness or qualification (Ohanian, 1990), is the degree to which the celebrity is perceived to have skills and knowledge concerning a specific topic, that he/she had learned through experience or practice (Fox, 2008). Attractiveness can be referred to as either facial or physical attractiveness and is the degree to which someone’s face or whole body is perceived as being chic, sexy or likable. Attractiveness is often used by individuals to judge other people, and is positively related to a positive change in attitude (Ohanian, 1990). The evaluations of an ad as well as of the appealed product/service/brand turns out more satisfactory, when the ad included attractive models (Benoy Joseph, 1982). In mass media, physical attractiveness may be a crucial attribute, since it is highly salient (Maddux & Rogers, 1980).
Creating the link between a celebrity and a product/service/brand has been conceptualized as ‘match up hypothesis’ (Lee & Koo, 2016), which states that the message which should be communicated about the product or service promoted, ought to be conform with the image conveyed by the celebrity (Kahle & Homer, 1985). Kanungo and Pang (1973) name the ‘match up hypothesis’ in their article ‘fittingness hypothesis’, which again emphasizes a necessary fit between the celebrity and the product/service/brand.
11
successful career and know-how on the Nike shoe. Thus, in this example, the audience of the MJ-basketball advertisement is appealed to directly think about the relation between the celebrity and the product itself. They process this relation via the central route and see the celebrity as strong argument for buying the product.
Having no relation between the celebrity and the product/service/brand can lead to negative product or service evaluations (Kanungo & Pang, 1973). In the digital age, these bad evaluations can be for example negative eWOM, negative recommendations or reviews, as well as public complaints on social media platforms. An example for a celebrity having no relation to the product she endorses is Helena Bonham Carter, a British actress. She was the ambassador of Yardley, an international English-based cosmetics producer, although the celebrity never uses make-up and cosmetics, as she said in interviews (Independent, 2005). The other way around, thus when a celebrity is related to too many products, is called ‘overexposure’. It means that a celebrity endorser can lose his/her credibility, due to the fact that he/she endorses many different products at the same time (Muda, Musa & Putit, 2012). For instance Tiger Woods, a professional American golfer, promoted not only golf equipment, which would have been related to his golf career, but also watches, gaming console games, cars, pharmaceutical products or motor-cycles (Kelley, 2017).
The source-credibility model and the attractiveness-based model can also be applied to the match-up hypothesis. In this case, they are renamed to: Expertise-based product
congruence and attractiveness-based product congruence (Lee & Koo, 2016).
In expertise-based product congruence, product involvement plays a role. It influences the degree of attention spent by an individual on an advertisement and the main message (cognitive route). The celebrity is either not recognized or ignored. Attractiveness-based
product congruence comprises the communication of the message (cognitive route), but
12
Analogue to the ‘match up hypothesis’, Kamins (1990) found that the attractiveness of a celebrity is more effective, when the product is related to attractiveness as well. Hence, the effectiveness of a celebrity depends on his/her relatedness and similarity to the product. In the perfume industry, almost every advertisement includes a celebrity endorsing the product. Hereby the celebrity’s attractiveness is very important, since it reflects the product’s characteristics and gives it an image. The product itself is an attractive one, because it is perceived as a luxury product, mostly bottled in a specifically beautiful and attention awakening glass flagon, which has the ability to make the customer attractive too. Therefore, there is an interplay of the celebrity’s beauty and the product’s beauty, while the celebrity’s expertise is in this case mostly his/her attractiveness and image.
It is recommended that marketers should evaluate potential celebrities on all three components, thus on Ohanian’s (1990) scale, to find the most appropriate endorser for their product/service/brand. The scale ensures, that the celebrity will match the product/service/brand on all dimensions (attractiveness, trustworthiness and expertise). In this study, the scale is used to figure out if a high credibility and a relation to pro-environmental behaviour positively influence the purchase intentions for Fairtrade products. Chen and Huddleston (2009), who investigated the effect of several marketing communication tools on the purchase intention for Fairtrade coffee, also found that the ‘matching’ of the endorsing celebrity and the product are crucial. In their case, sports stars were not very effective in endorsing Fairtrade coffee. The authors therefore advise to use celebrities, who are related to pro-environmental behaviour. Considering this, the following hypothesis will focus on the moderating effect of the credibility of a celebrity endorser on consumer’s purchase intentions for Fairtrade products:
13
Testing the credibility of the celebrity as a moderator can help to understand if a relation of a celebrity to pro-environmental behaviour, thus a high credibility, strengthens his/her persuasiveness for purchasing Fairtrade products. Since Fairtrade is an already well-established and world-wide known non-profit organization, most of the people are familiar with the concept and have already built an attitude towards it. But how can attitudes be influenced? Is it possible that celebrities could have an impact on someone’s attitude towards Fairtrade products? The following section is about attitudes, how they are built and influence purchase intentions.
2.3 Attitude towards Fairtrade Products
Attitudes are not born with someone, but established over years while growing up. Their creation is mainly influenced by someone’s social surrounding, also called social norms, which are most often family and friends (Terry, Hogg & White, 1999). But also culture and religion can play a role in how someone forms his/her attitudes towards different kinds of things (Rashid, Sambasivan & Rahman, 2004; Fam, Waller & Erdogan, 2004).
Since Ajzen (1991), an attitude is the extent to which an individual evaluates a specific behaviour in a favourable or unfavourable way. In his model, Theory of Planned Behaviour (TPB), the author describes attitude as an independent determinant of intention. Its central point is the intention to act in a given behaviour, e.g. a purchase. This intention is built by motivations, which are perceived behavioural control, subjective norm and attitude
towards the behaviour (as already defined above). The outcome of the intention is then the behaviour.
Attitudes are claimed to be the best predictors for purchase decisions (Fennis & Stroebe, 2016). They are inferred by cognitive, evaluative or affective and behavioural information (Fennis & Stroebe, 2016). An attitude is formed, when something is new or unknown to a target audience. The formation, as well as a change of existing attitudes, is mainly influenced by cognitive processes that need rather less thought and little cognitive elaboration, e.g. peripheral cues. Therefore, celebrity endorsement can be in instance a good tactic to influence the formation or the change of existing attitudes towards anything that can be evaluated positively or negatively (Fennis & Stroebe, 2016).
14
marketing tactic for Fairtrade products, due to a change in the attitude towards Fairtrade by the endorser. Therefore, the following third hypothesis has been developed:
H3: A celebrity endorser leads to a positive attitude towards Fairtrade, which in turn increases the purchase intention for Fairtrade products.
15
3 Conceptual Model
Figure 2: Conceptual Model
As pictured in Figure 2, the conceptual model’s main relationship, which is simultaneously the first Hypothesis (H1), consists out of two parts: the influence of celebrity endorsement
(X) on the consumer’s purchase intention for Fairtrade products (Y). The moderation of
this relationship, the second hypothesis (H2), is the influence of the Credibility of a
Celebrity (W1). The third hypothesis is, that the main relationship (X Y) is assumed to
be mediated by the Attitude Towards Fairtrade Products (M).
16
4 Research Design
4.1 Data Collection Method
This study contained a pre-survey and a main survey. The pre-survey as well as the main survey both included three conditions: Condition 1: showing pictures of celebrities being related to pro-environmental behaviour, Condition 2: showing pictures of celebrities not related to pro-environmental behaviour, and Condition 3: the control group, where there were pictures of ‘normal’, non-famous people (in this study called “fake-celebrities”) shown (see Figure 3 below).
4.2 Pre-Survey
The pre-survey examined which celebrities and non-celebrities were appropriate to use in the main survey in order to increase the study’s external validity. For each condition, four pictures of either two men and two women (mostly one blonde and one brunette) have been collected. In terms of Condition 1, the selection of the celebrities has been based on the celebrities’ pro-environmental activity and attitude. For Condition 2, celebrities that are not linked to pro-environmental behaviour have been selected. Online magazines and blogs gave information about the celebrities (Mazzoni, 2015; Goode, 2015; Pustetto, 2015; Martin, 2003 & Oxfam International. (n.d.)). Based on the screening of the online magazines and blogs, Stella McCartney, Emma Watson, Chris Martin and Al Gore had been used as the celebrity sample for Condition 1. A sample consisting of Lady Gaga, Kim Kardashian, Louis Hamilton and Boris Becker represented Condition 2. For Condition 3, pictures of ‘normal’, un-famous people have been randomly found with the help of the Google images search (see also Figure 3).
17
a short description of what Fairtrade is and the Fairtrade logo, have been added. The questions asked in the pre-survey have been the same for all conditions. The survey was randomly assigned to the participants via the web-based survey program Qualtrics.
If a participant had been assigned to Condition 1 for example, he/she saw the posters of the celebrities related to pro-environmental behaviour. Then he/she was asked, if the persons shown on the posters are celebrities or not, based on a 5-Point-Likert scale. Afterwards, the participant had to rank (giving them a ranking from one to four, where one was “most” and four was “least”) the pictured persons in terms of expertise, trustworthiness and attractiveness. These three components summed up the Source-Credibility Scale developed by Ohanian (1990), as already discussed above. Afterwards, the participant had to indicate
18
if he/she knows, if the celebrities or fake-celebrities on the poster have done anything good for the environment (also on a 5-Point-Likert scale). If so and if he/she remembers, there was space given to type in what exact activities he/she knew the celebrities or fake-celebrities did. The pre-surveys can be found in the Appendix of this paper (see Table 1 in the appendices).
4.2.1 Participants and Procedure
The pre-survey was available via an anonymous link for 24 hours and was shared privately via Facebook, WhatsApp and e-mail. It was only available in English. In total, 56 people had participated, of which 45 responses could have been used after cleaning the data from not completely answered respondents. An overview of the descriptive information about the pre-survey’s respondents’ demographic characteristics can be found in the appendices in Table 2. In order to analyse the responses, the mean for each celebrity in each condition of the rankings has been calculated via SPSS (IBM Corp., 2016).
Figure 4: Final Selection of Celebrities for Main Survey
Also, the means for the two other questions have been calculated per condition and celebrity. The results have shown that from Condition 1 ‘Emma Watson’ is most appropriate, for Condition 2 ‘Kim Kardashian’ and for Condition 3, the control group/fake-celebrity, ‘Ivy Caleb, the brunette fake-celebrity who wears a hat’ should be used (see also Figure 4 above).
4.3 Main Survey
19
about the organization. All of them could have been answered on a 5-point Likert scale. These items made up the third control variable, called “Fairtrade Knowledge”.
The second section was about (2) the “Celebrity/Fake-Celebrity” and measured the independent variable (X). The specific poster, which had been assigned as appropriate for each condition in the pre-survey (see Figure 4), was shown to the participant. After looking at the poster, the participant was asked to rate the pictured person in terms of expertise, trustworthiness and attractiveness based on 15 criteria adapted from Ohanian (1990) on a 5-point Likert scale. The participant was also, as in the pre-survey, asked to specify if he/she perceives the pictured person as being pro-environmental or not pro-environmental (also on a 5-point Likert scale). If the participant remembered any pro-environmental activities of the pictured person, he/she had the chance to specify them within a text field, but this was optional (seven participants specified some activities they remembered). The following section was called (3) “Congruence”, in which the participant had to evaluate the fit between the pictured person and the ad. Therefore, three questions from Lee and Koo (2016) had been adopted, which were rated on a 5-point Likert scale, too. This section helped to measure the first control variable called “Congruence of Advertisement and Celebrity” (CV1).
In the fourth section, (4) the “Attitude towards Fairtrade” section, questions adopted from Yamoah et al. (2014) had been asked in order to measure the mediator effect. They were also scaled on a 5-point Likert scale.
Afterwards, the fifth section, (5) the “Purchase Intention”, measuring the dependent variable, came up. Here, a mixture of questions from different authors (all scaled on a 5-point Likert scale) had been asked (Bearden, Netemeyer & Haws, 2011; Baker & Churchill, 1977; van Doorn & Verhoef, 2011), as well as a confrontation took place, in which the participant had to choose for a product, namely either Fairtrade bananas or non-Fairtrade bananas. Fairtrade bananas had been chosen in this context, since it is a product which can be bought in almost every supermarket, which is well-known and which is one of those Fairtrade products generating the highest sales volume.
20
which questions adopted from Bearden, Netemeyer and Teel (1989) were asked, and could also been answered on a 5-point Likert scale. The questions specifically measured if people are likely to listen to what others say or attach importance to what products others use. Finally, in the last section (7), the participants were asked to give personal information (e.g. their gender, age, nationality, degree of education and current monthly net household income).
Originally, some of the adopted scales are measured on a 7-point Likert scale instead of a 5-point Likert scale. In this study, all scales had been set to a 5-point Likert scale (as described above) due to consistency and simplicity for the participant to answer.
The main-survey can be found in the Appendix of this study (see Table 3 in the appendices).
4.3.1 Participants and Procedure
The main survey was active for one week. It was available in English and German and spread via Facebook, Mail, Instagram and Whats App to fellow students, friends and family of the experimenter. To increase the attention for the survey and to incentivise potential participants, the chance to participate in a lottery at the end of the survey was promoted. The price were 30€ cash. In total, 248 people had participated, of which 202 responses could have been used after cleaning the data. Table 1 gives information about the descriptive information of the main survey’s respondents and the demographics will also be described in the results part (section 5.3) (see also Table 3 in the appendices).
4.3.2 Plan of Analysis
The first step to analyse the data was to check for the reliability of the different items which were supposed to be put together in one variable for the final analysis. Therefore, a reliability analysis had been undertaken for all item sets. Some of the items had to be removed from the different item sets to increase the Cronbach’s Alpha value of it.
21
1. Coding the fake-Celebrity/Control Group as number 1 was important, since the SPSS (IBM Corp., 2016) macro PROCESS by Andrew F. Hayes (2012) uses the category with the smallest value within a categorical independent variable as the reference group. This means, that the other two conditions were always compared to this group within the analyses by the computer program SPSS (IBM Corp., 2016).
The third step was to check for the frequencies of the data set, thus to control, if there were any missing values, outstanding deviations etc.
The fourth planned step was to run a “fake regression” to check the data for outliers and to get a first impression of the whole data set. The “fake regression’s” output was an overview of the entered variables, a model summary, an ANOVA, an overview of the coefficients, residuals statistics, a histogram, normal P-P plot of regression standardized residuals and finally a scatterplot. After deleting the outliers from the data set, the “fake regression” was repeated and the diagrams given in the output were checked again.
The next step was to mean centre all non-categorical variables to ensure a better interpretability of the results.
One manipulation check has been executed with the help of a univariate analysis of variance in SPSS (IBM Corp., 2016), namely to see, if the three celebrities (including the fake-celebrity) have been perceived as different in terms of their relation to pro-environmental behaviour, thus if they are good representatives of each of their assigned conditions.
To test the first hypothesis, a one-way-ANOVA has been executed in SPSS (IBM Corp., 2016) to see if the celebrities had a significantly different influence on the Purchase Intention for Fairtrade Products (Y).
The second hypothesis was tested via a moderation analysis. It was tested, if the Credibility of Celebrity (W1) has a moderating effect on the main effect relationship. This test for moderation was executed with the help of the macro PROCESS by Andrew F. Hayes (2012) in SPSS (IBM Corp., 2016) using model number one.
22
While conducting all these analyses it has always been controlled for the three variables “Congruence of Advertisement and Celebrity”, “Normative and Informational Influence” and “Fairtrade Knowledge” to consider specific influences. It was controlled for “Congruence of Advertisement and Celebrity” (CV), to see how people perceived the fit between the advertisement (Fairtrade poster) and the celebrity (Emma Watson, Kim Kardashian or Ivy Caleb (fake-celebrity)). The control variable “Normative and Influential Influence” (CV) was created, because people can be prone to what other people think about them or tell them, and thus can be easily influenced by others. The variable “Fairtrade Knowledge” (CV) helped to see if the participants knew much about the Fairtrade concept, or if it was new to them, or even if they had a wrong impression about it.
23
5 Results
5.1 Frequencies
The output of the frequencies showed that there were no missing values and the means (M) were mainly between three and four, which predicted a slight skewness to the right in the histogram, which was later also given by the ”fake regression” output. The standard deviations were around (SD =) 0,60, which is almost only half a point of the scales. One outstanding deviation was seen at the variable Congruence between Advertisement and Celebrity (CV) with a value of (SD =) 1,24271. The minimum and maximum was also always between the ranges of the 5-point Likert scale.
5.2 Reliability Analysis
To check for the reliability of the different items, which were supposed to be put together in one variable for the final analysis, a reliability analysis has been conducted in SPSS (IBM Corp., 2016). The following Cronbach Alpha values (α) can be reported after deleting items from some of the item sets: Purchase Intention for Fairtrade Products (Y) (9 items, α = ) 0,804 (one item deleted); Attitude towards Fairtrade Products (M) (3 items, α = ) 0,560 (one item deleted); Credibility of Celebrity (W1) (14 items, α = ) 0,958 (one item deleted); Fairtrade Knowledge (CV) (7 items, α = ) 0,851 (one item deleted); Congruence between Advertisement and Celebrity (CV) (3 items, α = ) 0,981 (no item deleted); and finally Normative and Informational Influence (CV) (11 items, α = ) 0,858 (one item deleted).
24
The “fake regression” showed that five respondents were outliers, which had then been excluded from the data set. The histogram of the second “fake regression” and the scatter plot showed an equal distribution around zero and the data values were closely approximated to the linear regression line in the normal P-P plot of standardized residuals showing a positive linear relationship with a Pearson’s correlation (r) of 0,201.
5.3 Descriptive Statistics of the Sample
The sample of analysis consisted of 202 (N = 202) respondents from 17 different countries (see Table 1). More than half of the sample were female (61,4%) (see Table 1). The respondents had mainly either graduated from high school (26,6%) or had a Bachelor’s Degree (40,6%) (see Table 1). Most of the participants were between 20 and 29 years old (20 – 24, 60,9%; 25-29, 28,2%; see Table 1). Around 60,0% of the respondents indicated that they had a current monthly net household income of less than €1000 (see Table 1). Comparing the sample of the main-survey with the one of the pre-survey, it can be concluded, that the pre-survey sample (See appendices Table 2) was representing the main-survey sample very well (see Table 1).
Demographic Variable Level Sample (n) Percentages
Nationality Afghanistan 1 0,50% Australia 1 0,50% Austria 1 0,50% Belgium 1 0,50% China 1 0,50% France 3 1,50% Germany 172 85,00% Hong Kong 1 0,50% Italy 3 1,50% Lithuania 1 0,50% Netherlands 7 3,50% Philippines 1 0,50% Poland 3 1,50% Rwanda 1 0,50% Sweden 1 0,50% Switzerland 3 1,50%
United States of America 1 0,50%
25
Current Monthly Net
Household Income < 1000€ 119 58,90% 1001€ - 2000€ 47 23,30% 2001€ - 3000€ 26 12,80% 3001€ - 4000€ 4 2,00% 4001€ - 5000€ 2 1,00% > 5001€ 4 2,00% 202 100,00%
Education Less than High School Degree 19 9,40%
High School Graduate (High School Diploma or
Equivalent including) 53 26,20%
Some College, but no Degree 14 7,00%
Associate Degree in College 11 5,40%
Bachelor’s Degree in College 82 40,60%
Master’s Degree 23 11,40% 202 100,00% Age 15-19 10 5,00% 20-24 123 61,00% 25-29 57 28,00% 30-34 4 2,00% 35-39 0 0,50% 40-44 1 0,50% 45-49 2 0,50% 50-54 2 1,00% 55-59 1 0,50% 60-64 2 1,00% 202 100,00% Gender Female 124 61,40% Male 78 38,60% 202 100,00%
Table 1: Descriptive Statistics of Main-Survey Sample Characteristics
5.4 Manipulation Check
26
Kardashian (Cond. 2) (M = -0,8883, SD = 0,68327). Emma Watson (Cond. 1) had the highest average rating in Credibility of Celebrity (W1) with (M =) 3,94 points, indicating that she is/would probably be a credible endorser for Fairtrade (see Figure 5). Ivy Caleb (Cond. 3/fake-Celebrity) followed with an average of (M =) 3,20 points, which indicates that the respondents were not sure if she was a celebrity and a credible endorser for Fairtrade (3 = neither credible nor not credible) (see also Figure 5). Kim Kardashian had an average rating of (M =) 2,23, which indicates that she is rather not perceived as a credible endorser for Fairtrade (see also Figure 5). Hence, it can be concluded, that the manipulation of the credibility of the endorsers functioned and that they were all good representatives of their assigned conditions (Cond 1: related to pro- environmental behaviour – Emma Watson, Cond 2: not related to pro-environmental behaviour – Kim Kardashian, Cond 3: fake-celebrity/non-celebrity – Ivy Caleb).
Figure 5: Average Values of "Credibility of Celebrity" (W1) per Condition (X)
5.5 Testing for Hypothesis 1
To test for hypothesis 1, thus to see, if the different conditions of the independent variable have varying effects on the dependent variable, a one-way-ANOVA has been conducted in SPSS (IBM Corp., 2016). The results have shown, that there was not a significant effect of Celebrity Endorsement (X) on the Purchase Intention for Fairtrade Products (Y) at the 0.05
0 0,5 1 1,5 2 2,5 3 3,5 4 4,5 5
Emma Watson Kim Kardashian Ivy Caleb / Fake-Celebrity
27
α-level for the three conditions [F(2, 199) = 0,493, p = 0,612]. This means that the different celebrities (including the fake-celebrity) have no different effects on the peoples’ purchase intentions for Fairtrade products.
Looking additionally at the means of the Purchase Intention for Fairtrade Products (Y) per condition, all participants, either they have seen the advertisement with Emma Watson (Cond. 1), Kim Kardashian (Cond. 2) or Ivy Caleb (Cond. 3/fake-celebrity), have given nearly the same answer (Cond. 1 M = 3,26; Cond. 2 M = 3,32; Cond. 3 M = 3,20) (see very balanced trendline in Figure 6). Additionally, the results of the moderation analysis conducted for Hypothesis 2 (see below), also indicate that there is no direct effect of Celebrity Endorsement (X) on the Purchase Intention for Fairtrade Products (Y) [Cond. 1:
t(8, 193) = -1,2471, p = 0,2139; Cond. 2: t(8, 193) = 0,3890, p = 0,6977].
Figure 6: Average Values of the "Purchase Intention for Fairtrade Products" (Y) per Condition (X)
This indicates that the celebrities have no differing effects on the participants’ purchase intentions compared to the fake-celebrity. Therefore, Hypothesis 1 needs to be rejected. It can be concluded, that people are not more likely to buy Fairtrade products when they are endorsed by a celebrity.
Following the paper of Baron and Kenny (1986), which describes the classic methods of moderation and mediation (Causal Steps Approach), the analyses test would have stopped here, because there was no significant effect of X on Y. But recent literature suggests to elide the insignificance of the main effect and to test further planned analyses to see, if the
0 0,5 1 1,5 2 2,5 3 3,5 4 4,5 5
Emma Watson Kim Kardashian Ivy Caleb / Fake-Celebrity
28
main effect might be influenced by third or fourth variables (Hayes, 2009). This may give some indications for specific patterns, which could be traced back to the literature described above, and might also help for future research. An overview of the found relationships can be seen in Figure 7 below, and a summary of the hypotheses and if they would be rejected or accepted is given in Table 3 below.
5.6 Testing for Hypothesis 2
To test for Hypothesis 2, a moderation analysis with the help of the macro PROCESS by Andrew F. Hayes (2012), and his model one has been executed in SPSS (IBM Corp., 2016). It was expected that the effect of Celebrity Endorsement (X) on the Purchase Intention for Fairtrade Products (Y) is impacted, thus moderated, by the Credibility of the specific
Table 2: Statistics for Moderation Analysis
Celebrity (W1). The overall model was significant (see Table 2). The model results have shown, that there was no significant moderating effect of the Credibility of Celebrity (W1) on the relation of Celebrity Endorsement (X) on the Purchase Intention for Fairtrade Products (Y) (see Table 2). The p-values of the interactions have been higher than the α-level of 0,05 and all of the 95% confidence intervals included a 0.
This means, that neither Emma Watson, nor Kim Kardashian compared to the fake-celebrity have an influence on the Purchase Intention for Fairtrade Products (Y), when the Credibility of the Celebrity (W1) moderates the relationship. Especially Emma Watson was
Model Summary R R2 MSE F DF1 DF2 p
0,326 0,1062 0,4838 2,8678 8 193 0,0049
Model Coefficient se t p LLCI* ULCI*
Constant -0,0832 0,0843 -0,9866 0,3251 -0,2495 0,0831
Cond. 1 Emma Watson -0,2018 0,1618 -1,2471 0,2139 -0,521 0,1174
29
expected to have a positive impact on the Purchase Intention for Fairtrade Products (Y) when moderated by the Credibility of the Celebrity (W1), because she is perceived to have a relation to pro-environmental behaviour, and is therefore assumed to be more persuasive in that context. But since the moderation analysis has shown that neither her, nor Kim Kardashian have a significant effect under the influence of the Credibility of the Celebrity (W1), Hypothesis 2 needs to be rejected. People are not more likely to purchase Fairtrade products, when the endorsing celebrity is related to pro-environmental behaviour, thus is a credible source (which means being trustworthy, having expertise and being attractive) in terms of a pro-environmental manner.
5.7 Testing for Hypothesis 3
To test for Hypothesis 3, a mediation analysis with the help of the macro PROCESS by Andrew F. Hayes (2012), and his model four has been executed in SPSS (IBM Corp., 2016). It was expected that people have a more positive attitude towards Fairtrade, when they see a Fairtrade promotion in which a celebrity endorses Fairtrade products, which then increases the likelihood to purchase Fairtrade products (X M Y). This means, that one of the celebrities, compared to the fake-celebrity, was expected to have a positive influence on the Attitude towards Fairtrade (M). Thus, the Attitude towards Fairtrade (M) was expected to positively mediate the relationship between Celebrity Endorsement (X) and the Purchase Intention for Fairtrade Products (Y).
The overall model was significant [R2 = 0,2518, F(6, 195) = 10,9358, p = 0,0000]. The results have shown, that there is a full mediation of the Attitude towards Fairtrade (M) on the main effect Celebrity Endorsement (X) on the Purchase Intention for Fairtrade Products (Y). This means that there was no significant main effect (X Y) to be found [Cond. 1:
Coeff. = 0,0350, SE = 0,1224, t(6, 195) = 0,2863, p = 0,7750, 95% CIs (-0,2063, 0,2764),
30
interaction effect results, because none of the confidence intervals of Cond. 2 (Kim Kardashian) included a zero [95% CIs (0,0953, 0,3442)]. Cond. 1 (Emma Watson) had, compared to the fake-celebrity (Ivy Caleb), no significant effect on the Purchase Intention for Fairtrade Products (Y) through the mediator [Cond. 1: Coeff. = - 0,0847, SE = 0,1123,
t(5, 196) = -0,7542, p = 0,4516, 95% CIs (- 0,3062, 0,1368)].
In other words, Kim Kardashian has a positive influence on the Attitude towards Fairtrade (M), which again has a positive influence on the Purchase Intention for Fairtrade Products (Y). Due to this full mediation, Hypothesis 3 can be accepted, because celebrity endorsement increases the attitude towards Fairtrade, which finally increases the intention to purchase Fairtrade products.
Hypotheses Results Result
H1: People are more likely to buy Fairtrade products when they are endorsed by a celebrity. (X Y)
Rejected
H2: People are more likely to buy Fairtrade products, when the endorsing celebrity is related to pro-environmental behaviour, thus is a credible source (which means being trustworthy, having expertise and being attractive) in terms of pro-environmental concerns. (X Y moderated by W1)
Rejected*
H3: A celebrity endorser leads to a positive attitude towards Fairtrade, which in turn increases the purchase intention for Fairtrade products. (X M Y)
Accepted
31
6 Conclusion & Recommendations
The goal of this study was to find out if celebrity endorsement can increase the purchase intention towards pro-environmental products, hence Fairtrade products. The answer to this question will be discussed in the following, including limitations of this study, some recommendations and a conclusion incorporating managerial implications for the management of Fairtrade.
The test for hypothesis 1 has shown, that the celebrities had, compared to the fake-celebrity, no significantly different impact on the purchase intention for Fairtrade products. Therefore, the answer to the research question of this study is, that celebrity endorsement does not direct purchase intentions for pro-environmental products, hence Fairtrade products.
This outcome might be reasoned by the fact, that conscious consumers are found to be more likely to engage in a consumption behaviour, that is not popularly accepted, hence that is not followed by the mass (Webster, 1975). Celebrity endorsement is one of those marketing strategies mostly used, because its main advantage is to reach the mass due to the popularity of the celebrities. Celebrity endorsers are implemented in 9,6% of TV ads in the United States (Choi, Lee & Kim, 2005). Hence, conscious consumers might be reluctant to advertisement using such mass-appealing marketing strategies and therefore celebrity endorsement is not an effective marketing tactic to reach conscious consumers, who know the concept of Fairtrade and prefer to purchase sustainable goods.
Another reason might be, that conscious consumers, or also called “label-readers”, rather focus on the labels and information given on the product or in advertisement (Writer, 2017), and therefore focus on the slogan and informational text in the ad and do not spend attention towards the celebrity endorser, who therefore might lose his/her effect.
32
(1985), Ohanian, (1990) and Chen and Huddleston (2009), who all state, that a celebrity, who is related to the product is more persuasive. Looking at Lee and Koo’s (2016) findings, a celebrity who is related to the product/service/brand promoted, can also lead to information processing via the central route. But since the answers of the participants per group have been the same, it cannot be said, if Emma Watson made the people process the information in a deeper way than those, who saw Kim Kardashian or Ivy Caleb, the fake-celebrity.
The outcomes of this study might also be reasoned by the arguments given above, namely that conscious consumers are less likely to listen to celebrity endorsement as a mass marketing tool and therefore rather spend attention on facts and information given on the product or in ads.
Corresponding to the findings of Yamoah et al. (2014), it was found that the attitude towards Fairtrade has a positive influence on the purchase intention for Fairtrade products. The current study has shown that attitude towards Fairtrade is a full mediator for the influence of celebrity endorsement on the purchase intention for Fairtrade products. This means, that celebrity endorsement has a positive impact on the attitude towards Fairtrade, which then increases the purchase intention for Fairtrade products.
Surprisingly, Kim Kardashian instead of Emma Watson was shown to be the celebrity having a significantly positive effect on the Attitude towards Fairtrade, although she was not perceived to be related to pro-environmental behaviour. This can be explained by the fact, that the participants who saw Kim Kardashian in the survey, scored on average 0,1 points higher in attitude towards Fairtrade than those who saw Emma Watson promoting Fairtrade. Since this is such a small value, it cannot be assumed that Kim Kardashian has somehow a special influence on the participants. It must have happened due to coincidence, that people who had a higher attitude towards Fairtrade, had been randomly assigned for Condition 2 by the computer software.
6.1 Limitations
The surprising insignificant effect of the different celebrities on the purchase intention for Fairtrade products, indicates that this study has some limitations.
33
heterogenous, the answers might have been significantly different among the three conditions of the independent variable. Here, the sample size might play a role. The study investigated only responses of 202 participants. A sample of e.g. 5000 participants might have led to different outcomes. Most of the participants were between 20 and 29 years old. Reason for this might be that the survey was mainly spread across Facebook and Instagram, which are social media platforms essentially used by a younger generation. The participants had, for the most part, a higher educational degree, which allows for the assumption that they were familiar with the concept of Fairtrade and its impact, and therefore already had a positive attitude towards it. Knowing the concept of Fairtrade might have made them ignore further given information on the poster. 85% of the participants knew what the logo stands for and 80% indicated that they had already bought a Fairtrade product. This could explain why they all scored quite high and equal on the attitude towards Fairtrade independently from the celebrity they have seen in the ad presented in the survey. The average mean of the Fairtrade knowledge of the participants was 4,26. The higher education also reasons a need for higher cognition, which might also explain why the participants were rather processing the information given on every poster on a central route and therefore ignored the peripheral cue, namely the celebrity/fake-celebrity.
Another limitation why the different celebrities had no different influences on the participants could also be that they averagely were not prone to being influenced by others (Average of Normative and Informational Influence ((CV) = 2,3726). This gives another reason, why the celebrity/fake-celebrity did not have different influences on the participants purchase intentions.
34
& Homer, 1985), and might have lost his/her effectiveness.
Another limitation of the study might be, that the questions had to be answered on a 5-Point Likert scale. A scale giving more answer possibilities (e.g. 7 5-Point Likert scale) might have led to more varying results.
The item set, which was supposed to create the variable Attitude towards Fairtrade, indicated a very small Cronbach’s Alpha value, which indicates, that not all of the variables did not really measure the same construct. Therefore, the items of the scale should rather be aligned, expanded or exchanged to increase the validity of the total variable.
Lastly, the usage of only one product, namely bananas (either Fairtrade bananas or normal non-Fairtrade bananas), which is a low-involvement product might have biased the purchase intention as well. There could have been participants who do not like bananas and therefore never buy them, which is why it was hard for them to imagine the situation to purchase bananas in a supermarket, and therefore their responses in the survey were biased. If the product would have needed high-involvement of the consumer, thus much thought, this might have changed the way how he/she would have processed the information given on the poster in combination with the celebrity as peripheral cue.
6.2 Recommendations
35
the ad. The real supermarket environment would also lead to a bigger and more varying sample, since people of all ages do their groceries in a supermarket.
6.3 Conclusion
The current study contributes to the literature of celebrity endorsement in the field of non-profit organizations, by showing contradicting to recent research, that the credibility of a celebrity can also play no role and that there must be a third variable like attitudes, that makes the influence of celebrity endorsement on the purchase intention for Fairtrade products possible. In total, the current research gives a few important advises for future research in the field of celebrity endorsement and non-profit organizations like Fairtrade.
Summarized, it can be said that, based on this study’s findings, Fairtrade should use a celebrity to promote their products (especially Fairtrade bananas), since they can indirectly increase the purchase intention for Fairtrade products. Looking only at the effect a celebrity might have on the purchase intention for Fairtrade products, and ignoring their impact on the attitude towards Fairtrade, would have led to the outcome, that celebrities, either related or not related to pro-environmental behaviour, do not have an influence on the purchase intention for Fairtrade products. This was due to the fact, that the participants in this study showed to be well informed about the Fairtrade concept (average score on Fairtrade Knowledge (M) = 4,26), which motivates their supernumerary decisions for the Fairtrade bananas instead of the normal bananas in the end of the survey, without being influenced by any of the celebrities (see Figure 8). This is good news for Fairtrade, because it shows, that people between 20 and 29 years old with higher education are familiar with their concept, nonetheless, the credibility of the celebrity, although not enhancing people’s purchase intentions for Fairtrade products in this study, should again be checked during further research, since it can, based on recent research, have an impact. Thus, before using any available celebrity for the coming Fairtrade marketing campaign, the Fairtrade
74% 26%
Fairtrade Bananas Non-Fairtrade Bananas
36
37
7 References
7.1 Books and Journals
Agrawal, J., & Kamakura, W. A. (1995). The Economic Worth of Celebrity Endorsers: An Event Study Analysis. Journal of Marketing, 59(3), 56.
doi:10.2307/1252119
Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and
Human Decision Processes, 50(2), 179-211. doi:10.1016/0749-5978(91)90020-t
Baker, M. J., & Churchill, G. A. (1977). The Impact of Physically Attractive Models on Advertising Evaluations. Journal of Marketing Research,14(4), 538. doi:10.2307/3151194
Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173-1182.
Bearden, W. O., Netemeyer, R. G., & Haws, K. L. (2011). Handbook of
marketing scales: multi-item measures for marketing and consumer behavior research.
Los Angeles: Sage.
Bearden, W. O., Netemeyer, R. G., & Teel, J. E. (1989). Measurement of Consumer Susceptibility to Interpersonal Influence. Journal of Consumer
Research,15(4), 473. doi:10.1086/209186
Benoy Joseph, W. (1982). The Credibility of Physically Attractive
Communicators: A Review. Journal of Advertising, 11(3), 15-24. Retrieved from http://www.jstor.org/stable/4188405
Bezençon, V., & Blili, S. (2010). Ethical products and consumer involvement: what's new? European Journal of Marketing, 44(9/10), 1305-1321.
doi:10.1108/03090561011062853
Byrne, A., Whitehead, M., & Breen, S. (2003). The naked truth of celebrity endorsement. British Food Journal, 105(4/5), 288-296. doi:10.1108/00070700310477086
Chaiken, S. (1980). Heuristic versus systematic information processing and the use of source versus message cues in persuasion. Journal of Personality and Social
38
Chen, S., & Huddleston, P. (2009). A comparison of four strategies to promote fair trade products. International Journal of Retail & Distribution Management,37(4), 336-345. doi:10.1108/09590550910948565
Choi, S. M., Lee, W., & Kim, H. (2005). LESSONS FROM THE RICH AND FAMOUS: A Cross-Cultural Comparison of Celebrity Endorsement in
Advertising. Journal of Advertising, 34(2), 85-98. doi:10.1080/00913367.2005.10639190 Cohen, D. (1972). Surrogate Indicators and Deception in Advertising. Journal of
Marketing, 36(3), 10. doi:10.2307/1251034
Davies I., A., & Gutsche, S. (2016). Consumer Motivations for Mainstream ‘Ethical Consumption. European Journal of Marketing, 50(7-8), 1326–1347.
Doorn, J. V., & Verhoef, P. C. (2011). Willingness to pay for organic products: Differences between virtue and vice foods. International Journal of Research in
Marketing,28(3), 167-180. doi: 10.1016/j.ijresmar.2011.02.005
Doran, C. (2009). The Role of Personal Values in Fair Trade Consumption.
Journal of Business Ethics, 84(4), 549-563. Retrieved from
http://www.jstor.org/stable/40294760
Erdogan, B. Z. (1999). Celebrity Endorsement: A Literature Review. Journal of
Marketing Management, 15(4), 291-314. doi:10.1362/026725799784870379
Escalas, J. E., & Bettman, J. R. (n.d.). You Are What They Eat: The Infulence of Reference Groups on Consumers Connections to Brands. SSRN Electronic Journal. doi:10.2139/ssrn.412081
Fam, K. S., Waller, D. S., & Erdogan, B. Z. (2004). The influence of religion on attitudes towards the advertising of controversial products. European Journal of
Marketing, 38(5/6), 537-555. doi:10.1108/03090560410529204
Fennis, B. M., & Stroebe, W. (2016). The Psychology of Advertisement (2nd ed.). Abingdon: Routledge.
Fireworker, R. B., & Friedman, H. H. (1977). The Effects Of Endorsements On Product Evaluation. Decision Sciences, 8(3), 576-583. doi:10.1111/j.1540-5915.1977. tb01104.x
39 Wesley Longman Ltd.
Hayes, A. F. (2009). Beyond Baron and Kenny: Statistical Mediation Analysis in the New Millennium. Communication Monographs, 76(4), 408-420. doi:10.1080/03637750903310360
Kahle, L. R., & Homer, P. M. (1985). Physical Attractiveness of the Celebrity Endorser: A Social Adaptation Perspective. Journal of Consumer Research, 11(4), 954. doi:10.1086/209029
Kamins, M. (1990). An Investigation into the "Match-up" Hypothesis in Celebrity Advertising: When Beauty May Be Only Skin Deep. Journal of Advertising, 19(1), 4-13. Retrieved from http://www.jstor.org/stable/4188750
Kanungo, R. N., & Pang, S. (1973). Effects of human models on perceived product quality. Journal of Applied Psychology, 57(2), 172-178. doi:10.1037/h0037042
Lee, Y., & Koo, J. (2016). Can a Celebrity Serve as an Issue-Relevant Argument in the Elaboration Likelihood Model? Psychology & Marketing, 33(3), 195-208.
doi:10.1002/mar.20865
Maddux, J. E., & Rogers, R. W. (1980). Effects of source expertness, physical attractiveness, and supporting arguments on persuasion: A case of brains over beauty.
Journal of Personality and Social Psychology, 39(2), 235-244.
doi:10.1037//0022-3514.39.2.235
Mathur, L., Mathur, I., & Rangan, N. (1997). The Wealth Effects Associated ' with a Celebrity Endorser The Michael Jordan Phenomenon,". Journal of Advertising
Research, 37(3), 67-76.
Muda, M., Musa, R., & Putit, L. (2012). Breaking through the Clutter in Media Environment: How Do Celebrities Help? Procedia - Social and Behavioral Sciences, 42, 374-382. doi: 10.1016/j.sbspro.2012.04.201
Ohanian, R. (1990). Construction and Validation of a Scale to Measure Celebrity Endorsers' Perceived Expertise, Trustworthiness, and Attractiveness. Journal of
Advertising, 19(3), 39-52. Retrieved from http://www.jstor.org/stable/4188769
