THE EFFECTS OF ETHNIC SIMILARITY AND IDENTITY
ON ADVERTISEMENT
A study on the importance of ethnic similarity between the viewer of
an ad and the character in the ad, and the role of the viewer’s ethnic
identity.
Charlotte Verviers 11911751
Master Thesis, Master’s Program Persuasive Communication Graduate School of Communication
University of Amsterdam Supervisor: S. Schinkel
ABSTRACT
Ethnicity has an important influence on the effects of advertisement. Research has found that ethnic similarity between the viewer of and ad and the characters in an ad often increases brand attitude, ad attitude and purchase intention. However, this has not yet been examined in the Netherlands, where 12.73% of all citizens have a Non-Western background. Considering this relatively high percentage, marketers and advertisers should consider ethnic similarity as a strategy, in order to accommodate to all potential customers. This study investigates the effects of ethnic similarity on brand attitude, ad attitude and purchase intention in an ad for a typical Dutch food: stroopwafels. Furthermore, this study considers the role of the viewer’s ethnic identity strength, and – unlike previous research – recognizes that people can identify with more than one ethnicity. One hundred thirty-five people participated in the online experiment. The study contains four conditions (Ethnic Background Participant: White vs Non-White; and Ethnic Background Ad Character: White vs Non-White) with Ethnic Identity Strength as moderator. The experimental design is a combined strategy, as the moderator can not be manipulated. The findings show that ethnic similarity in advertisement does increase brand attitude, ad attitude and purchase intention for Non-White people as compared to White people. Moreover, ad attitude can be moderated by the strength parental ethnic identity of Non-White people. Based on these findings, implications and recommendations for future research are given.
TABLE OF CONTENT
Introduction p. 4
Theoretical Framework p. 8
Method p. 15
Results p. 21
Conclusion & Discussion p. 29
References p. 33
Appendix A: stimulus material p. 36
INTRODUCTION
An article from the Daily Mail (England) in November 2017 showed that Tesco’s Christmas commercial – showing a Muslim family celebrating Christmas in English-fashion – triggered a large amount of negative reactions. The article stated that Twitter users expressed their disapproval of the supermarket’s choice to put a Muslim family in their traditional Christmas commercial: “Tesco why are you showing Muslims celebrating Christmas in your advert.
That’s just wrong, we all know they don’t!!!” and “Tesco have just put an advert out showing families celebrating and enjoying Christmas dinner. Why was a Muslim family included in
this ad?” (Daily Mail, 2017). Could this be an example that people today are not open to the
idea that one can both be Muslim and yet celebrate the holidays of the country they live in? That people can identify with more than just one culture? The negative reactions to the characters portrayed in the Tesco Christmas ad suggest ethnicity is still a sensitive matter in advertisement.
The Netherlands, like England, is a country that is home to people from all kinds of different ethnic backgrounds. According to Statistics Netherlands there were 17,081,507 citizens in the Netherlands in 2017. 2,173,723 of these people are from a non-Western background, this constitutes 12.73% of the total Dutch population (CBS, 2017). Nearly twice as much as in 1996, when only 7.56% had a Non-Western background. These numbers include 1st, 2nd and 3rd generation immigrants. In the four largest cities in the Netherlands, up to 30% of all citizens have a non-Western background (CBS, 2018a). This relatively large percentage shows that the Netherlands circa 2017 is a multi-cultural society, meaning it could face the same issues as England did when showing this multiculturalism in the Tesco Christmas commercial. An example is the 2018 Soccer World Cup for which the Netherlands did not qualify, but Morocco did. The Moroccan team contained a notable amount of players who are Dutch citizens. To support these Dutch-Moroccans, a Dutch newspaper put the
Moroccan flag on their front page and converted the traditional Dutch soccer slogan (“Hup Holland Hup”) to “Hup Morocco Hup” (The Best Social, 2018), which provoked a high amount of negative reactions on the newspaper’s Facebook page. Claiming the paper is “betraying their country” and people asking if the paper “has gone mad” (Facebook, 2018).
And just like England, with their traditional English Christmas meals, the Netherlands is a country rich in cultural food traditions, especially stroopwafels (Marshall et al., 2016). Various websites on Dutch food history state that the average Dutch person eats twenty
stroopwafels per year (InfoNu, 2018). One can argue that this traditional food is well
integrated the Dutch culture. Considering the high level of citizens with a Non-Western background in the Netherlands, and the fact that the Netherlands has its own (food) traditions embedded in their culture, it is interesting to consider combining these factors in an advertisement. Would Dutch citizens react as negatively to an advertisement showing people from a Non-Western ethnicity enjoying a Dutch tradition, such as eating a stroopwafel?
In other countries, mainly the USA, a considerable amount of research has been done on the effects of ethnicity in advertisement. Sierra, Hyman and Heiser (2012) suggests that ads that have ethnically resonant cues, such as the ethnicity of the character in the ad, help create a more positive attitude towards the brand and the ad. Other studies have looked into the ethnicity of the character presented in advertising (Appiah, 2001a, 2001b), and found that when a viewer notices that a brand puts in effort to target their ethnic group, their attitude towards the ad and the brand was more positive (Sierra, Hyman & Heiser, 2012). It was found that when the ad portrayed characters that were similar in ethnicity to the viewers, this ethnic similarity would positively influence consumers' purchase intention (Donthu & Cherian, 1992; Hirschman, 1981; Webster, 1994) and ad attitude (Whittler, 1989; Green, 1999). A research conducted by Appiah (2001a) studied the influence of ethnic similarity between the viewer of and ad and the character in the ad, and found that when the viewer’s
ethnicity was similar to the characters in the ad, this would lead to a more positive ad attitude, brand attitude and purchase intention. Furthermore, they found that when someone related more strongly to their ethnicity they would especially appreciate ethnic similarity in advertisements. However, their research was limited, as it did not allow people to participate in the experiment if they would identify with more than one ethnicity. This shows that just like society, research has not been open to the idea that people can identify with more than one ethnicity.
Holland and Gentry (1997) also studied the effects of ethnic similarity in advertisement, however, they focused on cultural symbols. They matched ethnicity of the characters in the ad and cultural symbols portrayed in the ad with the viewer. They found that when there was a match of ethnicity and cultural symbols between the ad and the viewer, the viewer evaluated the ad more positively, resulting in higher brand attitude, ad attitude and purchase intention. Stroopwafels are seen as the most traditional Dutch food (Marshall et al., 2016), and could thus be seen as a cultural symbol. It is interesting to see if there will be the same effect Holland and Gentry (1997) found with the stroopwafel as a Dutch cultural symbol. When it comes to people who feel that they experience more than one ethnic identity, and they see a cultural symbol from one of their cultural identities and do (/not) see a character who is similar to their other cultural identity, will this influence their attitude towards the ad and the brand? And would it have an influence on their intention to buy the product?
Furthermore, Holland and Gentry (1997) also found that people who identify more strongly with an ethnic group will have a stronger affective response to advertisements that focus on these ethnic identities. Meaning that when a viewer puts more importance on their ethnic background, they respond more positively towards advertisements that portray their ethnicity.
Research has shown that ethnic similarity between the viewer of an ad and the characters in an ad has a positive influence on ad attitude, brand attitude and purchase intention (Donthu & Cherian, 1992; Hirschman, 1981; Webster, 1994; Whittler, 1989; Green, 1999; Appiah, 2001a; Sierra, Hyman & Heiser, 2012). If these results of ethnic similarity are the same in the Netherlands, Dutch advertisers could be missing out on potential customers if they ignore the effect ethnic similarity has on advertisement. Therefore, this study is of great societal relevance for advertisers and marketers. Furthermore, the outcome of this study will contribute to the existing literature of the effects of ethnic similarity on brand attitude, ad attitude and purchase intention through assessing a possible moderating effect of the strength of the viewer’s ethnic identity and taking into consideration that people can identify with more than one ethnicity.
In sum, ethnic similarity seems an important factor in advertising, influencing the viewers’ attitude towards the ad, the brand and their purchase intention. Furthermore, research showed that cultural symbols have a positive influence on ad attitude, brand attitude and purchase intention when matched with the viewer. Moreover, the Netherlands has proven to be ethnically diverse, strengthening the argument to consider ethnic similarity in advertising. Lastly, researchers found that when someone identified more strongly with their ethnical background, ethnic similarity in advertisement deemed of greater importance to them than people who did not identify as strongly with their ethnical background. Together, these factors lead to the following research question:
What is the effect of ethnic similarity between the viewer of and ad and the characters in an
ad for traditional Dutch products on brand attitude, ad attitude and purchase intention? And
THEORETICAL FRAMEWORK
Because of the growing diversity in society it has been made clear for marketers that they need to communicate with various groups of different cultural backgrounds (Holland & Gentry, 1997), because people from different cultural backgrounds perceive messages differently (Pitts, Whalen, O’Keefe, & Murray, 1989). Researchers argue that the effectiveness of an ad increases when the cultural environment of the intended audience is taken into account, meaning that the content – the characters, symbols and the values depicted in the ad – of a message fits with the cultural background of the viewer (McGuire, 1984). When the cultural background of a viewer is represented in an ad – showing similar looking characters, symbols related to their culture and portraying their values – it makes it easier for them to see similarities between themselves and the characters in the ad. This is crucial since people respond more favourably to people who they perceive are similar to themselves (Pitts et al., 1989). Studies have shown that when viewers feel they are similar to the characters presented in an ad they are more influenced by the message (Appiah, 2001a). Aaker, Brumbaugh and Grier (1996) found that when similarity between the viewer of an ad and the characters in the ad is high, the viewer feels that the ad is intended for them, resulting in a more positive attitude towards the ad and the product.
This effect of favourability towards people who are similar to you can be explained by Identification Theory, which states that during an interaction people automatically examine their similarity with the person in question and make similarity judgements, in a way that people adopt behaviours of people who they feel they can identify with because they see similarities between themselves and the person they are interacting with (Kelman, 1961). In advertising, race is seen as the most significant cue of similarity between the audience and the character in an ad (Appiah, 2001a). These findings on similarity and ethnicity by Kelman (1961) and Appiah (2001a) would mean that an ad where the ethnicity of the viewer is
similar to the ethnicity of the character in the ad would allow the viewer to more easily identify with the character in the ad, resulting in the viewer being more open to adopting the behaviour proposed in the ad (Sierra, Hymand & Heiser, 2012). The proposed behaviour in advertising would mean buying the advertised product. Thus, ethnic similarity between the viewer of an ad and the character in the ad would then result in a higher purchase intention for said viewer.
A meta-analysis conducted by Sierra, Hyman and Heiser (2012) focused on how ethnic similarity between a viewer of an ad and the character in an ad influences how the viewer is affected by the ad. As ethnic similarity allows people to more easily identify with the people presented in the ad. The meta-analysis showed that brand attitude, ad attitude and purchase intention are generally enhanced when there is ethnic identification with an ad. Meaning that the ethnicity of the viewer and the character in the ad are similar in such a way that it allows the viewer to more easily identify with the character presented in the ad. Ethnic identity is defined by Phinney (1992) as “a person’s knowledge of his or her membership in a social group and the value and emotional significance attached to that membership” (Appiah, 2001-B, p. 1). This ethnic identification thus refers to how connected one feels to their ethnicity. Sierra, Hyman and Heiser (2012) found that there was a positive correlation between a viewers’ ethnic identification with an ad and the attitude the viewer had towards the ad. Furthermore, they stated that a viewer responds more positively towards the advertised brand when the ethnicity of the characters in the ad is similar to the viewer. Moreover, they found that ethnic identification with an ad positively relates to a higher purchase intention.
These effects of ethnic identification were mainly tested in studies where participants were either White, Black, Hispanic or Asian in the USA. However, they were not tested in a setting where participants were allowed to identify with more than one ethnicity. While
mixed race couples, and thus people who might identify with more than one ethnicity, are more and more common. In the Netherlands 16% of all couples are mixed race (CBS, 2018b). Nonetheless, the predictions made by Sierra, Hyman and Heiser (2012) could be the same for White and Non-White people in the Netherlands due to the fact that – mixed race or not – people appreciate ethnic similarity in advertisement, leading to the following hypotheses:
H1a: There will be a more positive attitude towards the brand when the ethnicity of the
viewer and the character in the ad are similar, as compared to when the ethnicity of the
viewer and the character in the ad are not similar.
H1b: There will be a more positive attitude towards the ad when the ethnicity of the viewer
and the character in the ad are similar, as compared to when the ethnicity of the viewer and
the character in the ad are not similar.
H1c: There will be a higher purchase intention when the ethnicity of the viewer and the
character in the ad are similar, as compared to when the ethnicity of the viewer and the
character in the ad are not similar.
A study conducted by Appiah (2001a) claims that companies are reluctant to use Black characters in their advertising, because they were afraid it would not be appreciated by mainstream – in their case White – viewers. However, empirical research has shown that the ethnicity of the character in the ad is not the most important similarity cues for White viewers and thus does not have a large effect on White viewers’ attitude towards the ad or product. Appiah (2001a) states that this smaller effect of ethnic similarity on White viewer’s attitude towards the ad or product is due to the fact that White viewers (the majority group in Western
society) are less conscious of and concerned with their ethnicity, as compared to Non-white viewers (the minority group in Western society), and thus look for other cues of similarity, such as clothing, values, and the like.
Ethnicity is a significant cue of similarity between a viewer and a character in ad (Appiah 2001a). This seems of greater importance for ethnic minorities, since for them ethnicity is more salient (Whittler, 1989); in a way that ethnic minorities are more aware of their ethnicity and put greater importance on their ethnicity (Phinney, 1992). White viewers, who are members of an ethnic majority in Western societies, are less conscious of the ethnicity of the character in an ad and look for other similarities than ethnicity between themselves and the ad character, such as lifestyle, appearance and values (Appiah, 2001a). In contrast, ethnic minority groups in Western societies (Non-White people) are more aware of their ethnic identity and place a higher importance on their ethnicity than majority group members do (Phinney, 1992). Due to the higher importance minority groups put on their ethnic identity, minority groups put a larger importance on ethnic similarity in advertisement and place a higher preference for characters who are ethnically similar to them, than majority groups do (Appiah, 2001a). Studies have shown that White viewers have similar reactions to advertisements with Black characters as to advertisements with White characters, whereas Black viewers have significantly more positive reaction towards ads with Black characters as compared to ads with White characters (Bush, Hair, & Solomon, 1979).
The more positive effect of ethnic similarity in advertising for ethnic minority groups could also be explained by Petty and Cacioppo’s (1986) Elaboration Likelihood Model (ELM). The model explains that attitudes are influenced by persuasive messages and that there are two different ways of message processing. The ELM states that the central route to persuasion happens when a viewer is more motivated to process the advertisement, due to the contents’ personal relevance and involvement for the viewer (Belch & Belch, 2015). The
viewer then pays closer attention towards the ad, and the attitudes formed via the central
route are believed to be more long term. The peripheral route, in contrast, happens when the
viewer has no motivation nor the ability to fully process the message. The viewer then pays less close attention to the ad, and the attitudes based on the ad are believed to be more short term. Hence, when an ethnic minority member looks at an ad that portrays a character ethnically similar to them, the personal relevance level of the message rises, thus making the viewer more motivated to process the message via the central route. Resulting in more long term attitudes towards the ad and the brand. These beliefs about the ELM, combined with the findings of Phinney (1992) and Appiah (2001a), would suggest that processing via the
central route would be especially beneficial for minority groups. Considering that ethnic
similarity in advertising is believed to be of greater importance for minority than majority groups, the following hypotheses are proposed:
H2a: The positive effect of ethnic similarity between a viewer and a character in an ad on
brand attitude will be stronger for non-white viewers, as compared to white viewers.
H2b: The positive effect of ethnic similarity between a viewer and a character in an ad on ad
attitude will be stronger for non-white viewers, as compared to white viewers.
H2c: The positive effect of ethnic similarity between a viewer and a character in an ad on
purchase intention will be stronger for non-white viewers, as compared to white viewers.
Studies have shown that ethnic similarity between a viewer of an ad and a character in an ad affects how a viewer evaluates the ad and that this effect would be higher for people who come from an ethnic minority. Phinney (1992) states that minority groups (Black people) are more mindful of their ethnicity due to the absence of Black people in mainstream (Western)
media. Holland and Gentry (1997) found that people who identify more strongly with their ethnic background have stronger emotional responses to seeing that ethnic background represented in advertising as compared to people who identity less strongly with their ethnic background. Sierra, Hyman and Heiser (2012) found that someone’s ethnic identity can influence the way they respond towards advertising. This ethnic identity entails the connection to a social group and one’s own identification as a group member, feeling like one belongs to a group and responds favourable towards their own group (Phinney, 1990). Sierra, Hyman and Heiser (2012) found that the strength of ethnic identity (SEI) has a moderating effect on ad attitude, brand attitude and purchase intention when the ad is ethnically resonant, in a way that individuals with a higher SEI appreciate ads more when the ads were resonant in ethnical cues representing their own ethnical background. Appiah (2001b) found similar effects in his research stating that the viewer’s SEI significantly influences the viewer’s evaluation of the advertisement. Moreover, they found that Black participants who had a stronger SEI with their Black background responded more positive when they were presented with ads including characters who were similar to them in their ethnicity, as compared to Black participants who had a weaker SEI.
The difference in importance of ethnicity between minority and majority group members could also be explained by McGuire’s (1984) Distinctiveness Theory, which states that a person’s more distinctive characteristics are more salient to them, as compared to more prevalent traits that are shared by more people in society. Moreover, the theory also states that a person’s ethnicity will be more prominent to those in an ethnic minority group as compared to people who belong to the ethnic majority group in their social environment (McGuire, 1978). Hence, in Western societies – where Non-White people are the minority – their most distinctive characteristic would be their ethnicity, resulting in a higher appreciation of an ad and a brand when this distinctive characteristic is portrayed.
The Black participants with a stronger SEI in Appiah’s (2001b) study were able to identify more with the characters in the ads that were similar in ethnicity, resulting in more positive attitudes towards the ad. According to McGuire’s Distinctiveness Theory this would be because their ethnicity is distinctive compared to the environment that they were in, where the majority is White. However, none of the studies said that majority groups can not have a strong SEI. The studies conducted by Appiah (2001b) and McGuire (1984) state that a strong SEI occurs with minority members, although majority group member – White people in Western society – could also be aware and proud of their ethnicity, resulting in a stronger SEI. This is worth exploring, and for this research that would mean that a persons’ attitude based on ethnic similarity in an ad, would be moderated by whether or not they have a stronger or weaker SEI. This lead to the following research question:
RQ1: Will the positive effect of ethnic similarity on brand attitude, ad attitude and purchase
intention be higher for people who have a stronger ethnic identity as compared to people who
have a weaker ethnic identity?
Figure 1
METHOD
Design & Sample
To answer the research question and test the hypotheses, an online experiment was conducted with a four (Ethnic Background Participant: White vs Non-White; and Ethnic Background Ad Character: White vs Non-White) by two factor (Strength of Ethnic Identity: Low vs High) between-subjects design (Table 1) with the dependent variables Brand Attitude, Ad Attitude, and Purchase Intention. This design is a combined strategy, since the moderator (SEI) can not be manipulated and is thus a quasi-independent variable in the model.
Table 1
The Four Similarity Conditions and Strength of Ethnic Identity Moderator of the Experimental
Design
Participants. All participants who were not within the approved age range (18 – 28),
who were not Dutch citizens and those who did not complete the survey, were excluded from the experiment. Resulting in one hundred thirty-five participants who agreed to participate in the experiment. Participants’ ages ranged from 18 to 28 (M = 23.70, SD = 2.45). Of the participants, 68.15% was female and 31.85% was male (see Table 2). Nearly 90% of all participants indicated that the Netherlands is their country of birth, with approximately 70% indicating to have a strong attachment to their country of birth.
Table 2
Sample of Participants
Notes. Mean with standard deviation between brackets for Age, Stroopwafel Likeability and Ethnic
Identity Strength. Quantities with percentages between brackets for Gender and Education. All mean differences between conditions were insignificant at the .05 level.
Stimulus Material
Participants were shown one of two advertisements. The only difference between the ads was the ethnicity of the characters in the ad. Both ads introduce a new brand of stroopwafels, a Condition Sample Similarity White Viewer (N = 36) No Similarity White Viewer (N = 37) No Similarity Non-White Viewer (N = 30) Similarity Non-White Viewer (N = 32) Total (N = 135) Age 24.14 (2.45) 23.03 (2.33) 24.17 (1.98) 23.53 (2.86) 23.70 (2.45) Gender % Man Woman 11 (30.56) 25 (69.44) 7 (18.92) 30 (81.08) 11 (36.67) 19 (63.33) 14 (43.75) 18 (56.25) 43 (31.85) 92 (68.15) Education % VMBO HAVO/VWO MBO HBO WO 0 (0.00) 2 (5.55) 3 (8.33) 11 (30.56) 20 (55.56) 1 (2.70) 2 (5.41) 2 (5.41) 12 (32.43) 20 (54.0)5 0 (0.00) 0 (0.00) 3 (10.00) 7 (23.33) 20 (66.67) 1 (3.13) 0 (0.00) 4 (12.50) 10 (31.25) 17 (53.13) 2 (1.48) 4 (2.96) 12 (8.89) 40 (29.62) 77 (57.04) Stroopwafel Likeability 3.70 (0.98) 3.69 (0.75) 3.88 (0.64) 3.42 (0.96) 3.67 (0.86) Own Ethnic Identity Strength 3.97 (0.78) 3.84 (0.67) 4.32 (.67) 4.23 (0.91) 4.07 (0.78) Parental Ethnic Identity Strength 3.92 (0.78) 3.79 (0.68) 4.30 (0.64) 4.21 (0.80) 4.04 (0.75)
traditional Dutch food product. This was chosen because it could be used as a cultural cue for the Dutch culture, that will be recognized by all participants. No other cultural cues have been implemented in the ad. The brand of the stroopwafels was called “Lot’s Stroopwafels” and a fictitious logo was designed, to ensure participants had no prior brand knowledge, and thus no existing brand attitude. To ensure participants knew that this was a new product the words “New” and “Now available in every supermarket” were implemented in the ad. The only manipulation that was made, were the characters in the ad. Both ads showed three characters (one boy, two girls) enjoying a plate full of stroopwafels, in one of the ads all three characters were Non-White, in the other ad all three characters were White. Both advertisements looked identical except for the ethnicity of the three characters. Furthermore, the characters in the ad were in the same age range (18 -28) to further stimulate similarity between the viewer of the ad and the characters in the ad. For the visual examples of both ads, see Appendix A.
Procedure
The online experiment (see Appendix B) was carried out between May 17th and June 4th 2018. An anonymous Qualtrics link was shared on Facebook to recruit participants. To reach the desired number of participants, ten participants were approached at the University of Amsterdam to participate in the experiment on a laptop. Participants were first told that the purpose of this experiment was to study stroopwafel advertisements and asked to agree with the informed consent. First some basic questions were asked regarding the age, sex and education level of the participants. Next, they were asked about their ethnicity. Unlike previous studies, participants had the option to fill in more than one ethnicity. Based on their answer regarding their ethnicity the participants were randomly assigned to one of the two ads, ensuring an equal amount of participants in each condition. Thereafter they were asked
to answer question regarding brand attitude, ad attitude and purchase intention. For the manipulation check, participants were asked to answer a question regarding the hair colour of the characters in the ad. Last, they were asked to what level they either agreed or disagreed with statements about their ethnic identity. Finally, participants were thanked for their participation and given an e-mail address if they had any further questions.
Pretest
To ensure the manipulation was perceived correctly, a pre-test was conducted, including the same questions as the actual experiment. First, all questions were approved by a chairman of Amsterdam United – a group for people of Non-Western backgrounds in Amsterdam – who has a Moroccan background herself, to ensure no one would be offended by the type of questions (specifically those involving ethnicity). The full experiment was carried out in a Word-document and sent to ten people, five with a White ethnic background and five with a Non-White ethnic background. All participants were asked to answer all the questions in the experiment. For the manipulation check a multiple-choice question was used, asking: “What was the hair colour of the people in the ad?”. Participants could then choose one of the following answers: “Red”, “Black” or “Blonde”. In the pre-test, all ten participants answered the manipulation check question correctly in correspondence with the ad that what shown to them (White Ad Character vs Non-White Ad Character). All pre-test participants were asked if they had further suggestion regarding the study. One participant suggested to first ask the questions regarding brand attitude, ad attitude and purchase intention, before asking questions about one’s ethnic identity. Since it could aggravate people, as the questions can be quite personal and this could affect their answers on the appreciation of the ad, brand and product. In accordance with this feedback, the order of the questions in the experiment were rearranged.
Measures
Dependent variables.
Ad Attitude. To measure ad attitude, a 5-item semantic differential scale was used,
based on the measurements by Sierra, Hyman and Torres (2009) and MacKenzie, Lutz & Belch (1986): Bad/Good, Unpleasant/Pleasant, Negative/Positive, Unattractive/Attractive and Unbelievable/Believable, on a 5-point Likert scale. A principal component factor analysis was conducted on the five items with varimax rotation. Results indicated the scale was indeed unidimensional, explaining 64.44% of the variance in the original five items. Reliability for the scale was good, Cronbach’s Alpha = .86 (M = 3.19, SD = .81), (1 = very negative, 5 = very positive).
Brand Attitude. To measure brand attitude, a 5-item semantic differential scale was
used, based on the measurements by Sierra, Hyman and Torres (2009) and MacKenzie, Lutz & Belch (1986): Bad/Good, Unpleasant/Pleasant, Negative/Positive, Unattractive/Attractive and Unbelievable/Believable, on a 5-point Likert scale. A principal component factor analysis was conducted on five items with varimax rotation. Results indicated the scale was indeed unidimensional, explaining 65.91% of the variance in the original five items. Reliability for the scale was good, Cronbach’s alpha = .86 (M = 3.31, SD = .75), (1 = very negative, 5 = very positive).
Purchase Intention. Based on the measurements by Diallo (2012) and Sierra, Hyman
and Torres (2009), participants were asked to what extent they agreed with the following statements regarding purchase intention on a 5-point Likert scale (1=Completely disagree, 5=Completely agree): “I would try Lot’s Stroopwafels”, “There is a high chance I would consider buying Lot’s Stroopwafels”, “The next time I buy stroopwafels, I would buy Lot’s Stroopwafels” and “There is a strong chance I will buy Lot’s Stroopwafels”. A principal component factor analysis was conducted on four items with varimax rotation. Results
indicated the scale was indeed unidimensional, explaining 73.65% of the variance in the original four items. Reliability for the scale was good, Cronbach’s alpha = .88 (M = 2.92, SD = 1.00), (1 = very negative, 5 = very positive).
Control variables.
Stroopwafel Likeability. To measure participants’ attitude towards stroopwafels
before the experiment, three statements regarding the product were created. Participants were asked to what extend they agreed with the following statements on a 5-point Likert scale (1 = Completely disagree, 5 = Completely agree): ‘I find stroopwafels tasty”, “I like to eat
stroopwafels”, and “I often buy stroopwafels”. A principal component factor analysis was
conducted for Stroopwafel likeability on the three items with varimax rotation. Results indicated the scale was indeed unidimensional, explaining 67.89% of the variance in the original three items. Reliability for the scale was good, Cronbach’s alpha = .73 (M = 3.67, SD = .86), (1 = very negative, 5 = very positive). See Table 2.
Own Ethnic Identity Strength. Based on the measurements for SEI by Sierra, Hyman
and Torres (2009) four statements were created. Participants were asked to what extent they agreed with the following statements on a 5-point Likert scale (1 = Completely disagree; 5 = Completely agree): “I am proud of my ethnicity”, “I am happy with my ethnicity”, “I feel strongly connected to my ethnicity” and “I have positive feelings towards my ethnicity”. To measure the strength of the viewer’s own ethnic identity a principal component factor analysis was conducted on four items with varimax rotation. Results indicated the scale was indeed unidimensional, explaining 75.2% of the variance in the original four items. Reliability for the scale was good, Cronbach’s alpha = .88 (M = 4.07, SD = .78), (1 = very negative, 5 = very positive). See Table 2.
Parental Ethnic Identity Strength. Based on the measurements for SEI by Sierra,
Hyman and Torres (2009) eight statements were created. Participants were asked to what extent they agreed with the following statements on a 5-point Likert scale (1 = Completely disagree; 5 = Completely agree): “I am proud of my fathers’/mothers’ ethnicity”, “I am happy with my fathers’/mothers’ ethnicity”, “I feel strongly connected to my fathers’/mothers’ ethnicity” and “I have positive feelings towards my fathers’/mothers’ ethnicity”. To measure parental ethnic identity strength, a principal component factor analysis was conducted on eight items with varimax rotation. Results indicated the scale was indeed unidimensional, explaining 68.74% of the variance in the original eight items. Reliability for the scale was good, Cronbach’s alpha = .93 (M = 4.04, SD = .75), (1 = very negative, 5 = very positive). See Table 2.
RESULTS
Randomization check
To check if participants’ age was comparable over the four conditions, a Oneway ANOVA was conducted with all four groups as independent variable and age as dependent variable. The ANOVA showed that participants’ mean age did not differ between the four conditions,
F (3, 131) = 1.76, p = .158. Second, to check the distribution between the conditions for
gender, a Chi-square test was conducted. There were no significant differences between the four conditions for gender, Χ2 (3) = 5.29, p = .152. Third, to check the distribution between conditions for education level, a Chi-square tests was conducted. There were no significant differences between the four conditions for education level, Χ2 (12) = 7.54, p = .820. Last, to check if participants’ stroopwafel likeability was comparable over the four conditions, a Oneway ANOVA was conducted with all four groups as independent variable and stroopwafel likeability as dependent variable. The ANOVA showed that participants’ mean
likeability for stroopwafels did not differ between the four conditions, F (3, 131) = 1.57, p = .200. Thus, randomization for age, gender, education level and stroopwafel likeability was successful. Since there were no differences between the conditions on these four variables, they were not included as covariates in the analyses.
Manipulation check
To check if participants had correctly perceived the hair colour of the characters in the ad in the stimulus material – thus verifying the manipulation – a Chi-square test was conducted, with all four groups as independent variable and the manipulation check as dependent variable. There was a significant difference between the four conditions for Manipulation
check, Χ2 (3) = 112.18, p < .001. The manipulation was thus successful, all participants perceived the manipulation correctly.
Hypotheses
Hypothesis 1.
To test H1a an Independent samples t-test was conducted with all four groups as independent variable and brand attitude as dependent variable. It was found that ethic similarity between viewer and ad character (M = 3.37, SD = .76) did not indicate a significantly more positive Brand attitude compared to no ethnic similarity between viewer and ad character (M = 3.24, SD = .74), t(133) = 1.019, p = .310.
To test H1b an Independent samples t-test was conducted with all four groups as independent variable and ad attitude as dependent variable. It was found that ethic similarity between viewer and ad character (M = 3.26, SD = .84) did not indicate a significantly more positive Ad attitude compared to no ethnic similarity between viewer and ad character (M = 3.12, SD = .78), t(133) = .999, p = .320.
To test H1c an Independent samples t-test was conducted with all four groups as independent variable and purchase intention as dependent variable. It was found that ethic similarity between viewer and ad character (M = 2.99, SD = 1.02) did not indicate a significantly more positive Purchase intention compared to no ethnic similarity between viewer and ad character (M = 2.85, SD = .98), t(133) = .804, p = .423.
Therefor hypotheses 1a, 1b and 1c are rejected. Although similarity seems to result in higher ad attitude, brand attitude and purchase intention the differences between similarity and no similarity groups are not significant. See Table 3.
Table 3
Hypothesis 1a, 1b, 1c: Results of Ethnic Similarity on Brand Attitude, Ad Attitude and Purchase Intention
Condition Dependent
Variable
Similarity Viewer and Ad Character (N = 68) No Similarity Viewer and Ad Character (N = 67) Brand Attitude 3.37 (0.76) 3.24 (0.74) Ad Attitude 3.26 (0.84) 3.12 (0.78) Purchase Intention 2.99 (1.02) 2.85 (0.98)
Notes. Mean with standard deviation between brackets. Significance levels are indicated with p <
.05*, p <.01**, and p <.001***.
Hypothesis 2.
To test H2a an Independent samples t-test was conducted with the two similarity groups (White vs Non-White) as independent variable and brand attitude as dependent variable. It was found that ethic similarity for the Non-White group (M = 3.68, SD = .60) indicates a significantly more positive Brand attitude compared to ethnic similarity for the White group (M = 3.10, SD = .79), t(66) = -3.38, p = .001.
To test H2b an Independent samples t-test was conducted with the two similarity groups as independent variable and ad attitude as dependent variable. It was found that ethic similarity for the Non-White group (M = 3.66, SD = .65) indicates a significantly more positive Ad attitude compared to ethnic similarity for the White group (M = 2.90, SD = .84),
t(66) = -4.11, p < .001.
To test H2c an Independent samples t-test was conducted with the two similarity groups as independent variable and brand attitude as dependent variable. It was found that ethic similarity for the Non-White group (M = 3.38, SD = .89) indicates a significantly more positive Purchase intention compared to ethnic similarity for the White group (M = 2.64, SD = .1.01), t(66) = -3.21, p = .002.
Therefor hypotheses 2a, 2b and 2c are accepted. The results show that when the Non-White participants saw an ad with characters who were ethnically similar to them, their attitude of the brand and the ad was more positive and they have a higher purchase intention as compared to when the White participants saw an ad with characters who were ethnically similar to them. See Table 4.
Table 4
Hypothesis 2a, 2b, 2c: Results of Non-White vs. White Ethnic Similarity on Brand Attitude, Ad Attitude and Purchase Intention
Condition Dependent Variable Similarity Non-White Viewer and Ad Character (N = 32)
Similarity White Viewer and Ad Character
(N = 36)
Brand Attitude 3.68 (0.60)*** 3.10 (0.79)***
Ad Attitude 3.66 (0.65)*** 2.90 (0.84)***
Purchase Intention 3.38 (0.89)** 2.64 (1.01)**
Notes. Mean with standard deviation between brackets. Significance levels are indicated with p <
Research Question 1.
To answer the research question a MANOVA was conducted with all four groups as independent variable and their own strength of ethnic identity (SEIown) as moderator on all three dependent variables. The MANOVA did not show a significant interaction effect on
Brand attitude, F (3, 125) = 2.05, p = .11, on Ad attitude, F (3, 125) = 1.41, p = .24, and on
Purchase intention, F(3, 125) = .43, p = .73. See Table 6.
Table 5 shows the mean scores on brand attitude, ad attitude and purchase intention before taking the viewer’s SEIown into account. The results show that the mean scores on brand attitude, ad attitude and purchase intention differed between the four conditions (Table 5) and between the strength of the viewer’s SEIown (Table 6). However, these differences were not significant. The results in Table 6 and Figure 2 show that in the Similarity White Viewer condition and the No Similarity Non-White Viewer condition, people who had a lower SEIown scored lower on brand attitude, ad attitude and purchase intention compared to people who had a stronger SEIown. In the No Similarity White Viewer condition, people who had a lower SEIown had a higher score on brand attitude, ad attitude and purchase intention compared to people who had a stronger SEIown. In the Similarity Non-White Viewer condition, there were mixed results. The people in this group who had a lower SEIown had a higher score on brand and ad attitude, but a higher score on purchase intention as compared to people who had a stronger SEIown.
Table 5
Mean Scores Ad Attitude, Brand Attitude and Purchase Intention All Participants and Per Condition Condition Variable Similarity White Viewer (N = 36) No Similarity White Viewer (N = 37) No Similarity Non-White Viewer (N = 30) Similarity Non-White Viewer (N = 32) Total (N = 135) Ad Attitude 2.90 (0.84) 3.06 (0.84) 3.18 (0.70) 3.66 (0.65) 3.19 (0.81) Brand Attitude 3.10 (0.79) 3.17 (0.84) 3.33 (0.60) 3.68 (0.60) 3.31 (0.75) Purchase Intention 2.64 (1.01) 2.76 (1.07) 2.96 (0.86) 3.38 (0.89) 2.92 (1.00)
Notes. Mean with standard deviation between brackets. Significance levels are indicated with p <
.05*, p <.01**, and p <.001***.
Table 6
Results Research Question 1: Interaction Effects Strength of Ethnic Identity and Ethnic Similarity on Brand Attitude, Ad Attitude and Purchase Intention
Condition Similarity White Viewer (N = 36) No Similarity White Viewer (N = 37) No Similarity Non-White Viewer (N = 30) Similarity Non-White Viewer (N = 32) Own Ethnic Identity Strength Low (N = 15) High (N = 21) Low (N =18) High (N = 19) Low (N = 8) High (N = 22) Low (N = 8) High (N = 24) Brand Attitude 2.87 (0.72) 3.27 (0.82) 3.39 (0.56) 2.97 (1.01) 3.25 (0.37) 3.35 (0.67) 3.78 (0.67) 3.65 (0.58) Ad Attitude 2.75 (0.71) 3.01 (0.93) 3.26 (0.62) 2.88 (0.98) 3.10 (0.47) 3.21 (0.78) 3.93 (0.68) 3.57 (0.63) Purchase Intention 2.42 (1.12) 2.80 (0.91) 2.83 (0.95) 2.70 (1.19) 2.81 (0.62) 3.01 (0.94) 3.31 (1.23) 3.40 (0.78)
Notes. Mean with standard deviation between brackets. Significance levels are indicated with p <
Figure 2
Results Research Question 1 in Plots: Interaction Effects Strength of Ethnic Identity and Ethnic Similarity on Brand Attitude, Ad Attitude and Purchase Intention
Note. X-Axis = Ethnic Identity Strength: Low vs. High. Y-Axis = Mean score. Conditions: Blue = Similarity White Viewer; Red = No Similarity White Viewer; Green = No Similarity Non-White Viewer; Orange = Similarity Non-White Viewer.
It is noteworthy that especially the scores of the Non-White participants were the opposite of what was expected from the results of Research Question 1. Based on these unexpected results, an additional analysis was conducted. To see if the Non-White participants would score differently based on the strength of ethnic identity of their parents (SEIparental), as compared to the strength of their own ethnic identity, a MANOVA was conducted using these parameters. The reason for looking into the attachment towards the ethnicity of the participants’ parents, is due to the aim of the study to be open to the fact that people can identify with more than one ethnicity. The results showed there was a significant interaction effect between the Non-White groups (similarity vs. no similarity) and the SEIparental strength (low vs. high) on ad attitude, F(1, 61) = 10.46, p = .002. See Table 7 and Figure 3. However, it still showed that when the Non-White participants in the similarity condition had a lower SEIparental, they scored higher on ad attitude, brand attitude and purchase intention, as compared to when the SEIparental was higher.
Table 7
Results Research Question Additional Analysis: Interaction Effect of Parental Ethnic Identity and Ethnic Similarity for Non-White Participants on Brand Attitude, Ad Attitude and
Purchase Intention
Condition No Similarity Non-White Viewer (N = 30)
Similarity Non-White Viewer (N = 32) Parental Ethnic Identity Strength Low (N = 8) High (N = 22) Low (N = 8) High (N = 24) Brand Attitude 3.22 (0.37) 3.37 (0.68) 3.83 (0.57) 3.60 (0.61) Ad Attitude 2.87 (0.69)** 3.31 (0.68)** 4.10 (0.55)** 3.43 (0.58)** Purchase Intention 2.47 (0.96) 3.17 (0.75) 3.50 (1.17) 3,32 (0.73)
Notes. Mean with standard deviation between brackets. Significance levels are indicated with p <
.05*, p <.01**, and p <.001***.
Figure 3
Results Research Question Additional Analysis in Plots: Interaction Effect of Parental Ethnic Identity and Ethnic Similarity for Non-White Participants on Brand Attitude, Ad Attitude and Purchase Intention
Note. X-Axis = Ethnic Identity Strength: Low vs. High. Y-Axis = Mean score. Conditions: Red = Similarity Non-White Viewer; Blue = No Similarity Non-White Viewer.
CONCLUSION & DISCUSSION
The purpose of this study was to examine what the effect of ethnic similarity in advertisement was on brand attitude, ad attitude and purchase intention and what role ethnic similarity had. Considering the results, it can be concluded that ethnic similarity does improve brand attitude, ad attitude and purchase intention for Non-White people. And that ad attitude can be moderated by the strength of the parental ethnic identity of Non-White people.
The results showed that people who saw an ad with characters that were ethnically similar to them – regardless of whether they were White or Non-White – scored higher on brand attitude, ad attitude and purchase intention as compared to people who saw an ad with characters who were not ethnically similar to them. However, these results were not significant (H1a, H1b, H1c). Interestingly, the results did show significant differences between the scores on brand attitude, ad attitude and purchase intention between Non-White people in the similarity condition and White people in the similarity condition. Indicating that ethnic similarity has a more positive effect for Non-White people (H2a, H2b, H2c). Last, it was expected that a higher SEI would lead to a higher appreciation of ethnic similarity in advertising, resulting in a higher brand attitude, ad attitude and purchase intention (RQ1). The white participants scored as expected. They scored higher when their SEIown was strong in the similarity condition, and accordingly when their SEIown was low in the no similarity condition. Unexpectedly, the results showed that Non-White participants in the no similarity condition scored higher when their SEIown was high as compared to when it was low. Moreover, the Non-White participants in the similarity condition scored higher on both brand attitude and ad attitude, but did score lower on purchase intention when their SEIown was low as compared to when it was high. However, none of these results were significant. Especially the results of the Non-White participants in the similarity conditions were unexpected (Table 6/Figure 2). Considering the findings by Appiah (2001b) it was expected that when
Non-White people had a stronger SEIown, they would have a more positive brand attitude, ad attitude and purchase intention when they saw an ad with characters ethnically similar to them. Due to these unexpected results, another test was conducted to see how the Non-White participants would score on the three dependent variables when the strength they felt towards their parents’ ethnicity was taken into account. These result showed that the Non-White participants in the similarity condition scored significantly higher on ad attitude when their SEIparental strength was low as compared to when their SEIparental was high (Table 7/Figure 3).
One possible explanation for the unexpected outcomes of the effect of a high SEI on the Non-White participants in the similarity condition, could be the fact that they were not necessarily shown a character in an ad that was similar to their specific ethnicity. In the questionnaire of the experiment, participants were able to fill in their specific ethnicity (Appendix B, Figure 10) because this studied aimed to include all participants’ specific ethnicities. For example, almost 15% of the participants who identified with an ethnicity other than Dutch, stated that they were Armenian. If these Armenian viewers did not have a strong SEI and were exposed to the Non-White stimulus it could be that they had a higher appreciation for this similarity (resulting in higher brand attitude, ad attitude and purchase intention) because they saw people who were – just like them – Non-White and yet advertising a typical Dutch product. In contrast, the Armenian viewers who did have a strong SEI (ergo: strongly identify as being Armenian) that were exposed to the Non-White stimulus scored lower on brand attitude, ad attitude and purchase intention. This could be explained by
Identification Theory (Kelman, 1961), because potentially they could not identify with the
characters in the ad, as even though they were Non-White, they did not portray the specific ethnicity of these Armenian viewers. Thus, the higher someone’s SEI is, the more specific the ethnic similarity has to be in order to have a positive effect on ad attitude. As a recommendation for further studies it could be advised to manipulate the ad based on the
specific ethnicities of the viewers in a follow up experiment. By conducting a pre-test to see which ethnicities are most common and then manipulate the advertisements based on those specific ethnicities. And include questions to measure how strongly the viewer feels similar to the characters in the ad.
Moreover, it is noteworthy that the Non-White participants in the similarity condition showed the lowest likeability towards stroopwafels compared to all other conditions (see Table 2), yet they had the highest scores on brand attitude, ad attitude and purchase intention as compared to all other conditions (see Table 5). Even though these results were not significant, it can thus be concluded that the ad was most effective for Non-White people who saw an ad portraying other Non-White people. Supporting the hypothesis that ethnic similarity in ads does a positive effect on purchase intention for minority groups.
It is important to note that this study has some limitations. Due to the restricted time for this study, only 135 participants could partake in the experiment. More participants per condition would increase the likelihood of reflecting a true effect (Button et al., 2013). Another limitation of the study was that some of the participants might recognize the characters portrayed in the ad, making their opinions about the ad biased. Specifically, the recruited ad characters attend schools and universities in Amsterdam. It is likely that a predominant amount of participants attends either of these institutions.
Practical and Theoretical Implications
The knowledge gained from this study is useful for marketers and advertisers, as the results indicate that by implementing ethnic similarity in an ad, the appreciation for the ad, the brand and the likelihood of purchasing the advertised product could be enhanced. However, it is noteworthy that the overall scores for purchase intention were lower than the scores for brand attitude and ad attitude, in all conditions. Practitioners might consider choosing ethnic
similarity in advertisements as a strategy especially when they want to improve brand and ad attitude, rather than purchase intention.
This study provides partial support for the existing literature on ethnic similarity as a way of increasing brand attitude, ad attitude and purchase intention, as it showed that there were only significant differences on these three dependent variables between ethnic similarity for Non-White people and White people. This study adds to the existing literature on the effects of ethnic identity strength on ethnic similarity in advertisement. As there is a high probability that the specific ethnicity of the viewer and the attachment the viewer has towards their parents’ ethnicity, play a substantial role on how ethnic similarity works between the viewer of and ad and a character in an ad.
REFERENCES
Aaker, J., Brumbaugh, A., & Grier, S. (1996). The non-target and viewer distinctiveness: The impact of target marketing on processing and attitudes (Research Paper No. 1391). Palo Alto.
Appiah, O. (2001a). Black, White, Hispanic, and Asian American adolescents' responses to culturally embedded ads. Howard Journal of Communication, 12(1), 29-48.
Appiah, O. (2001b). Ethnic identification on adolescents' evaluations of advertisements. Journal of Advertising Research, 41(5), 7-22.
Belch, G. E., & Belch, M. A. (2015). Advertising and Promotion: An Integrated Marketing Communications Perspective. (10th Edition), New York, NY: McGraw-Hill
Education.
Bush, R. F., Hair, J. F., & Solomon, P. J. (1979). Consumers’ level of prejudice response to black models in advertisements. Journal of Marketing Research. 16(3), 3417345. Button, K. S., Ioannidis, J. P., Mokrysz, C., Nosek, B. A., Flint, J., Robinson, E. S., &
Munafò, M. R. (2013). Power failure: why small sample size undermines the reliability of neuroscience. Nature Reviews Neuroscience, 14(5), 365. Centraal Bureau Statistiek. (2017). Retrieved from
http://statline.cbs.nl/StatWeb/publication/?DM=SLNL&PA=37325&D1=a&D2=0& 3=0&D4=0&D5=0-4,137,152,215,232&D6=0,4,9,(l-1)
l&HDR=G2,G1,G3,T&STB=G4,G5&VW=T
Centraal Bureau Statistiek. (2018a). Retrieved from https://www.cbs.nl/nl nl/achtergrond/2016/47/bevolking-naar-migratieachtergrond Centraal Bureau Statistiek. (2018b) Retrieved from https://www.cbs.nl/nl
Daily Mail. (2017). Retrieved from http://www.dailymail.co.uk/news/article 5071521/Viewers-complain-Tesco-s-advert-features-Muslim-family.html Diallo, M. F. (2012). Effects of store image and store brand price-image on store brand
purchase intention: Application to an emerging market. Journal of Retailing and
Consumer Services, 19(3), 360-367.
Donthu, N., and J. Cherian. (1992). Hispanic Coupon Usage: The Impact of Strong and Weak Ethnic Identification. Psychology and Marketing, 9 (6), 501–510.
Facebook. (2018). Retrieved from
https://www.facebook.com/Metro/photos/a.664124570302122.1073741851.1205683 1324412/1746088678772367/?type=3&theater
Green, C. L. (1999). Ethnic Evaluations of Advertising: Interaction Effects of Strength of Ethnic Identification, Media Placement, and Degree of Racial Composition. Journal
of Advertising 28 (1), 49–64.
Hirschman, F. C. (1981). American Jewish Ethnicity: Its Relationship to Some Selected Aspects of Consumer Behavior. Journal of Marketing, 45 (3), 102–110.
Holland, J. L., & Gentry, J. W. (1997). The impact of cultural symbols on advertising effectiveness: A theory of intercultural accommodation. ACR North American
Advances.
InfoNu. (2018). Retrieved from https://eten-en-drinken.infonu.nl/producten/109743-een nederlandse-lekkernij-de-stroopwafel.html
Kelman, H. C. (1961). Process of Opinion Change. Public Opinion Quarterly, 25(1), 57-78. MacKenzie, S. B., Lutz, R. J., & Belch, G. E. (1986). The role of attitude toward the ad as a
mediator of advertising effectiveness: A test of competing explanations. Journal of
Marshall, S. J., Livingstone, K. M., Celis-Morales, C., Forster, H., Fallaize, R., O’Donovan, C. B., & San-Cristobal, R. (2016). Reproducibility of the online Food4Me food frequency questionnaire for estimating dietary intakes across Europe. The Journal of
nutrition, 146(5), 1068-1075.
McGuire, W. (1984). Search for the Self: Going beyond self-esteem and the reactive self. In R. A. Zucker, J. Aronoff, & A. I. Rabin (Eds.), Personality and the prediction of behavior (pp.737120). NewYork: Academic Press.
Petty, R. E., & Cacioppo, J. T. (1986). The elaboration likelihood model of persuasion.
Advances in experimental social psychology, 19, 123-205.
Phinney, J. S. (1992). The multigroup ethnic identity measure: A new scale for use with diverse groups. Journal of adolescent research, 7(2), 156-176.
Pitts, R. E., Whalen, D. J., O’Keefe, R., & Murray,V. (1989). Black and white response to culturally targeted television commercials: A values-based approach. Psychology &
Marketing, 6(4).
Sierra, J. J., Hyman, M. R., & Heiser, R. S. (2012). Ethnic identity in advertising: A review and meta-analysis. Journal of Promotion Management, 18(4), 489-513.
Sierra, J. J., Hyman, M. R., & Torres, I. M. (2009). Using a model's apparent ethnicity to influence viewer responses to print ads: A social identity theory perspective. Journal
of Current Issues & Research in Advertising, 31(2), 41-66.
The Best Social. (2018) Retrieved from https://www.thebestsocial.media/nl/metro-is-op dreef-tegen-aangebrande-nederlanders-onder-marokko-post/
Webster, C. (1994). Effects of Hispanic Ethnic Identification on Marital Roles in the Purchase Decision Process. Journal of Consumer Research, 21 (2), 319–331.
Whittler, T. E. (1989). Viewers’ processing of actor’s race and message claims in advertising stimuli. Psychology of Marketing, 6.
APPENDIX A Figure 4
Stimulus Material White Ad Characters
Figure 5
APPENDIX B
Figures 6 – 22 show the questions the participants were asked in the online experiment. Figure 6
Figure 7
Figure 8
Control Variable Stroopwafel Likeability
Figure 9
Country of Birth, Scroll Down Menu
Figure 10
Figure 11
Ethnic Background (2/3)
Figure 12
Figure 13
Brand Attitude
Figure 14
Figure 15
Purchase Intention
Figure 16
Figure 17
Strength of Ethnic Identity 1
Figure 18
Strength of Ethnic Identity 2
Figure 19
Figure 20
Strength of Ethnic Identity 4
Figure 21
Strength of Ethnic Identity 5
Figure 22
