The imprecise look
Effect of ethnically ambiguous models in fashion
advertising
Name: Migiza Victoriashoop
Student number: 10807977
Programme: Communication Science
Institution: Graduate School of Communication University of Amsterdam
Track: Persuasive Communication
Assignment: MSc Thesis Communication Science
Supervisor: Dr. Sophie Boerman
Table of content Abstract……….2 Introduction………...3 Theoretical background………...5 Method……….11 Results………..17 Conclusion………23 Discussion……….23 References………26 Appendix………..30
Abstract
This study examines the effectiveness of ethnic ambiguity. Ambiguity in
advertisements is a much-used strategy by marketers. The ambiguity of the message allows advertisers to appeal to multiple target audience with the use of a single message. Ethnicity has proven to be an effective message cue for appealing to different target audiences. Therefore, this study assesses the effectiveness ethnic ambiguity as a cue as opposed to the use of distinctive ethnicity as a cue. Furthermore, self-referencing has proven to explain the effect of ad model exposure on attitude toward the advertisement and purchase intention. Self-referencing was therefore tested as a mediator. The effect of objective ethnicity as a
moderator of self-referencing was also assessed. The extent to which consumers engage in self-referencing upon exposure depends on the ethnicity of their ethnicity.
The effectiveness of ethnic ambiguity is tested against the distinctive ethnicity via a 3 (ad model ethnicity: White vs. ambiguity vs. Black) x 2 (objective ethnicity: White vs. non-White) between-subject factorial design. Beforehand conducting the main study a pretest was conducted. This allowed the selection of stimulus material most suited for the experiment.
The results indicate that ethnic ambiguity does not effectively cause more positive attitude toward the advertisement and purchase intention, through the process of
self-referencing, taking objective ethnicity into account. Rather, findings indicate that exposure to the resulted in lower scores on self-referencing as opposed to exposure to distinctive models. Furthermore, exposure to distinctive models resulted in significantly higher mediation effects on attitude toward the advertisement and purchase intention. No, significant effect were found to support the moderated mediation.
Introduction
In 2008 the fashion industry received much criticism for its lack of ethnic diversity (Pilkington, 2008). The majority of the industry’s marketing outings portray White (78.70%) models (Dries, 2014). Though this criticism was mainly focused on the lack Black models featured, it did bring about an increase in the total percentage of the portrayal of ethnic
models from 13.00% in 2008 to 17.30% in 2013 (Sauers, 2013; Wilson, 2014). Furthermore, a new look as spotted during the fashion shows following the criticism; the imprecise look (Disgraisian, 2008). According to one of the industry’s top casting directors the models that portray this look are ethnically ambiguous (Betty, 2008). This means their face (i.e. the way they look) cannot easily be placed with a certain ethnicity, race or geographical area. This look allows fashion designer who prefer the use of ethnic models, to do so without fear of losing their appeal to the industry’s main target audience, namely White consumers (Freeman, 2014).
The use of ethnically ambiguous ad models is a form of strategic ambiguity (Puntoni, Vanhamme, & Visscher, 2011). Puntoni et al. (2011) have described strategic ambiguity as the use of cues that can be interpreted in multiple ways. These interpretations allow marketers to appeal to multiple target audiences with a single message. Research has shown that
ethnicity in advertisements plays an important role in affecting consumers’ attitude toward the advertisement and purchase intention (Green, 1999; Torres & Briggs, 2007). Therefore, the use of ethnically ambiguous ad models could serve as an effective marketing strategy to reach multiple target audiences with one message. The aim of this study is to assess whether the use ethnically ambiguous models is an effective strategy for appealing to multiple target
audiences as opposed to the use of ethnically distinctive models.
The effect of the ethnicity of the ad model (ad model ethnicity) on attitude toward the advertisement and purchase intention is mediated by self-referencing. This process allows
consumer to link information in the advertisement to their self-concept (Martin, Kwai-Choi Lee, & Yang, 2004). The self-concept is defined as the idea or a set of ideas of who a person is (Markus & Wurf, 1987; Oyserman, 2004). As a construct the self-concept helps organize and memorize experiences.
Ad model ethnicity forms a cue that evokes self-referencing (Torres & Briggs, 2007). By linking that information to the self-concept consumers are provided with a frame of reference through which to process the information of the advertisement (David, Morrison, Johnson, & Ross, 2002). Research has shown that consumers belonging to ethnic minorities are more susceptible to ethnic cues in advertisements compared to those belonging to ethnic majorities (Lee, Fernandez, & Martin, 2002). The Distinctiveness Theory explains this by stating that an individual’s ethnicity helps them distinguish themselves from others (McGuire & McGuire, 1982). Therefore the objective ethnicity of consumers moderates the effect of ad model ethnicity on self-referencing. Furthermore, the mediating process of self-referencing on attitude toward advertising and purchase intention is therefore also moderated by objective ethnicity.
The study’s research question is to which extent the use of ethnic ambiguity is more effective in developing positive attitude toward the advertisement and purchase intention, through self-referencing, when taking the objective ethnicity of consumers into account. This research builds forth on prior research on strategic ambiguity (). Therefore, more insight can be gained on the circumstances under which strategic ambiguity is effective and when not. Secondly, much research has been done on targeting ethnic minorities (Martin et al., 2004; Torres & Briggs, 2007), however, to the knowledge of the researcher, no prior research has combined strategic ambiguity and ethnic ambiguity. Therefore, this research can provide insight in both ethnic targeting and strategic ambiguity. Furthermore, findings from this
research offer marketers information on the extent to which they could apply ethnic ambiguity.
Theoretical background
On average the fashion industry spends over $500 billion (over €400 million) on advertising on a yearly basis (Sebastian, 2014; Neikova, 2014). The main purpose of
advertisements is to stimulate the sales of products amongst the target audiences (Lavidge & Steiner, 1961). In order to determine whether an advertisement is successful in achieving that goal, outcome variables, such as attitude toward the advertisement (Aad) and purchase
intention can be used. Research on fashion advertising has shown that these variables play an important role in assessing the effectiveness of advertisements (Lee, Fernandez, & Martin, 2002; Torres & Briggs, 2007).
Aad is defined as an individual’s tendency to respond in a favorable or unfavorable manner to a particular advertising stimulus during a particular exposure occasion (Lutz, 1985). Torres and Briggs (2007) have shown that a favorable response to the advertisement can be achieved by appealing to the consumer, which means letting the consumer feel targeted by the advertisement. This is be done by using models that resemble the consumer (Green, 1999). Purchase intention is defined as an individual’s readiness and willingness to purchase a certain product or service (Ajzen & Fishbein, 1980). According to the Theory of Planned Behavior, an individual’s intention forms an important predictor for the execution of actual behavior (Ajzen & Fishbein, 1977). Therefore this study uses Aad and purchase intention to determine the effectiveness of the use of ethnically ambiguous ad model as opposed ethnically distinctive ad models.
The ethnicity of the ad model (ad model ethnicity) portrayed in advertisements has proven to affect Aad and purchase intention (Green, 1999; Torres & Briggs, 2007). Research by Green (1999), for example, has shown that Black consumers developed a more favorable Aad and purchase intention when exposed to Black models, as opposed to exposure to non-Black models. The American Psychological Association (2003) has defined ethnicity as an individual’s acceptance of norms, mores, and practices of the culture of origin and the simultaneous feeling of belonging to the group. The ethnicity of the model portrayed in the advertisement, then, represents a culture of origin, a group. Consumers feel targeted by the advertisement when the portrayed model represents a group they are part of (Green, 1999; Puntoni, Vanhamme, & Visscher, 2011; Pantoja & Segura, 2003).
White (78.70%) and Black (9.80%) ethnicities are most frequently featured in fashion advertisements (Dries, 2014). Because the ethnicity of these models is distinctive, meaning that their ethnicity can easily be determined, consumers who do not belong to the ethnic group these models represent do not feel targeted (Puntoni, Vanhamme, & Visscher, 2011). The ethnicity of the ambiguous model, however, allows for more consumers to feel targeted. Daniel Peddle, one of the industry’s top casting directors, has described the look of these model as not easily ascribable to a certain ethnicity or geographical area (Betty, 2008; Disgraisian, 2008). The use of these models allow designers who want to feature ethnic models, to do so without fear of loosing appeal with the main (White) target audience (Betty, 2008; Disgraisian, 2008). This reasoning touches upon a much-proclaimed fear within the industry that Black (i.e. ethnic) does not sell (Freeman, 2014). Ethnic models would supposedly cause fewer sales amongst the White consumers. Because the ethnicity of ambiguous ad models is difficult to determine, more consumers can infer the model to represent their ethnic group as opposed to the use of distinctive models (Washburn, Till, & Priluck, 2004).
Ambiguity as a strategy
The use of ambiguous ad models in advertising is a form of strategic ambiguity. Puntoni et al. (2011) have described strategic ambiguity as the use of ambiguous cues in messages with the aim of bringing about multiple interpretations amongst diverse target audiences. Because these different interpretations occur, strategic ambiguity allows marketers to reach multiple target groups with a single message. As ethnically ambiguous ad models cannot easily be placed with a certain race or geographical area, more consumers are allowed to feel targeted by the advertisement. The ethnic ambiguity permits more consumers to infer traits with the ad model that are representative of the consumer’s own ethnicity (Washburn, Till, & Priluck, 2004). This causes the advertisement to appeal to more (diverse) consumers, which results in a more positive Aad and purchase intention.
H1: Exposure to ambiguous ad models causes a more positive a) Aad and b) purchase
intention amongst consumers as opposed to exposure to White or Black ad models.
Self-referencing
Whether consumers feel targeted by the model in the advertisement can be explained through a process called self-referencing (Martin, Kwai-Choi Lee, & Yang, 2004). Self-referencing is defined as a strategic process through which individuals relate incoming information to their self-concept. An individual’s self-concept is defined as the knowledge of the self, meaning the idea, or a set of ideas, of who an individual is (Markus & Wurf, 1987; Oyserman, 2004). The self-concept is an important structure through which individuals organize their experiences and memories.
When the self-concept is activated, through the incitement of self-referencing, the internal cues become accessible (Escales, 2007). These internal cues form an individual’s memories and experiences (Bellezza, 1984). The accessibility of these cues causes individuals
to engage in thoughts about the self. Research has shown that the process of self-referencing is spontaneously be evoked by ethnic cues (Torres & Briggs, 2007). According to the
Distinctiveness Theory an individual’s ethnicity forms a discriminating personal traits by which individuals can distinguish themselves from others (McGuire & McGuire, 1982). The self-concept forms a multidimensional structure of which ethnicity forms a key dimension (Martin, Kwai-Choi Lee, & Yang, 2004). Therefore, ethnic cues spontaneously activate the self-concept though self-referencing. As a result exposure to the ad model allows consumers to relate incoming information to the self-concept.
The extent to which self-referencing occurs, however, is determined by the way the information is interpreted (Puntoni, Vanhamme, & Visscher, 2011). In order for consumers to determine whether they are targeted by the advertisement a social comparison is made (David, Morrison, Johnson, & Ross, 2002). According to the Social Identity Theory (Tajfel, 1982) individuals have a tendency to categorize other individuals into groups to which they then compare themselves. This comparison helps consumers judge the similarity of the ad model’s ethnicity to their own ethnicity (Martin, Kwai-Choi Lee, & Yang, 2004). According to Green (1999) consumers can better relate to others who look like them, causing consumer to
interpret the advertisement as targeted at them. Therefore the similarity judgment provides consumers with a frame of reference by which relevance of the advertisement to the self-concept is determined (Forehand & Deshpandé, 2001; Ratneshwar, Barsalou, Pechmann, & Moore, 2001). Because the ethnicity of ambiguous models is difficult to determine,
consumers have to infer traits with the model that are similar to their ethnicity (Washburn, Till, & Prilluck, 2004; Puntoni, Vanhamme, & Visscher, 2011). This causes consumers to be more engaged in thoughts of the self, and therefore self-reference more compared to exposure to distinctive ad models (Debevec & Iyer, 1988).
H2: Exposure to ambiguous ad model causes higher self-referencing amongst consumers as opposed to exposure to White and Black ad models.
Objective ethnicity
Ethnicity has been defined as an individual’s acceptance of norms and practices, and the feeling of belonging to the group (American Psychological Association, 2003). This definition touches upon two types of ethnicity described in literature; felt ethnicity and objective ethnicity (Deshpandé & Stayman, 1994). Felt ethnicity refers to strength of
individual’s affiliation to the ethnic group they belong to, meaning how strongly they identify with their ethnic background (Torres & Briggs, 2007). Objective ethnicity refers to the
demographic classification of individuals (Kwai-Choi Lee, Fernandez, & Martin, 2004). This classification is usually based on individuals’ countries of origin (Bevolking; kerncijfers, 2014). The objective ethnicity of consumers plays an important part in the social comparison made by consumers (Pantoja & Segura, 2003).
According to the Distinctiveness Theory ethnicity especially forms a salient cue to minorities (McGuire & McGuire, 1982). In social contexts ethnicities that differ from the main ethnic group, are more easily noticed. This causes minorities to more aware and susceptible to ethnic cues. In the context of the fashion industry the largest target group consists out of White consumers (Dries, 2014; Freeman, 2014). As a result the majority (78.70%) of advertisements feature White models compared advertisements featuring ethnic (21.30%) models. Therefore, the use of ambiguous ad models as an ethnic cue forms a salient cue for non-White consumers.
H3: Exposure to ambiguous ad models causes higher self-referencing as opposed to exposure to White or Black ad models. This effect will be more prominent amongst non-White consumers as opposed to White consumers.
Self-referencing as an effective strategy
Cognitive Psychology has shown that self-referencing is an effective strategy for learning and comprehension, because it aids the generation of self-relevant thoughts (Escales, 2007). Similar results have been found for the mediating effect of self-referencing on the development of more positive Aad and purchase intention (Martin, Kwai-Choi Lee, & Yang, 2004; Torres & Briggs, 2007). Because self-referencing generates more positive thoughts information in the advertisement is linked to pleasant self-relevant thought (Debevec & Iyer, 1988; Burnkant & Unnava, 1995). This causes consumers to develop more positive Aad and purchase intention. Exposure to ambiguous ad models spontaneously evoke self-referencing through which more positive Aad and purchase intention develop as opposed to exposure to distinctive ad models
H4: Exposure to ambiguous ad model has a more positive effect on a) Aad and b) purchase intention, through the process of self-referencing, as opposed to exposure to White or Black ad models.
Self-referencing moderated by objective ethnicity
The mediating effect of self-referencing on Aad and purchase intention upon exposure to ad model ethnicity is moderated by consumers’ objective ethnicity. Self-referencing is evoked by ethnic cues in the advertisement (Torres & Briggs, 2007). However, the extent to which consumers engage in self-referencing is determined by the way in which consumers interpret the incoming information. When consumers are exposed to ethnically ambiguous models the similarity judgment takes more elaboration. Because consumer have to infer for traits that are similar to their own ethnicity, engagement in self-referencing is higher. Because minority consumers are more susceptible to ethnic cues the positive effect of the mediating process of self-referencing is more prominent amongst minority consumers.
H5: Exposure to ambiguous ad model causes more positive a) Aad b) and purchase intention, through the process of self -referencing. This effect is more prominent amongst consumers with a non-White ethnicity as opposed to those with a White ethnicity.
Method
Design and sample description
For this study an online experiment was conducted with a 3 (ad model ethnicity: White vs. ambiguous vs. Black) x 2 (objective ethnicity: White vs. non-White) between-subject factorial design. Participants were recruited by posting the link to the online experiment on social media sites (Facebook, Twitter, and LinkedIn), spreading the link via e-mail and WhatsApp, and personally asking individuals to conduct the experiment. A convenience sample was used, because the recruitment relied on the personal network of the researcher (i.e. friends, family, and fellow students).
In total, 200 women (M = 26.20, SD = 8.47) participated in the experiment of which the majority indicated university (45.50%) as their highest level of education, followed by university of applied science (i.e. HBO; 26.00%), and vocational education (i.e. MBO; 18.00%). Women were chosen as participants because they form the largest group of consumers of jeans worldwide (Cotton Incorporated, 2015).
Procedure
After opening the link to the online experiment, an introductory text which stated the purpose of the study, the guarantees concerning the anonymity of participation, and the contact information of the researcher (see Appendix 8). This was followed by the informed consent, which had to be agreed upon before participants could continue the experiment. If
participants did not agree to the informed consent they were redirected to the end of the experiment, debriefed, and the experiment was ended.
Participants who did agree to the informed consent were subsequently asked to indicate their gender. As the purpose of this research was to assess the effect of ethnic
ambiguity the development of Aad and purchase intention amongst female consumers, all male participants had to be excluded from the experiment. Male participants were therefore
redirected to the end of the experiment, debriefed, and the experiment was ended. All female continued the experiment and were subsequently asked whether they had participated in the pretest of this study, ‘De onduidelijke look’ (‘the imprecise look’). As the stimulus material of the main study was constructed based on the results from this pretest, no participants with prior knowledge to the main study could participate in experiment. If participants had, they were redirected to the end of the experiment, debriefed, and the experiment was ended. All remaining participants were then randomly assigned to a fictitious advertisement (ad model ethnicity: White vs. ambiguous vs. Black; see Appendix 8). Before exposure participants were instructed to imagine they were browsing through a fashion magazine, upon which the
advertisement was shown.
After exposure participants were asked to indicate the origin and ethnicity that suited the ad model best (1= Dutch; 7 = Other, see Appendix 8). This was followed by several questions that assessed participants’ attitude toward the advertisement (Aad), purchase intention, and attitude toward the model (Amod). Then, participants were asked to answer questions concerning self-referencing, objective ethnicity and origin, and felt ethnicity. Subsequently, questions about their fashion involvement, ad model recognition, and product use followed. Finally, questions concerning participants’ personal information, such as age and educational level, were asked. After concluding these questions participants were
debriefed, with the possibility of commenting on the experiment, after which the online experiment was ended.
Stimulus material
As mentioned before, participants were exposed to an advertisement of a fictitious brand during the experiment with a relatively unknown model. This minimalized the chance of participants already having attitudes toward the model or advertisement. The advertisement was visible for seven seconds, and featured either a White, ambiguous or Black model. An exposure time of seven seconds was chosen, because research has shown that consumers spend about five seconds reading print advertisements (Young, 2015). As mentioned before, the fashion industry spends over $500 billion on advertising. Fashion magazines are one of the largest receivers of this amount of money (Sebastian, 2014; Neikova, 2014). This makes fashion magazines one of the major channels to reach their target group. As the experiment was conducted online two seconds were added to ensure proper loading of the advertisement onto the screen.
To determine which ad models to use for the main study a pretest was conducted amongst 25 female students (M = 23.52; SD = 4.67). Participants were exposed to nine fictitious advertisements featuring White, ambiguous, and Black models (see Appendix 7). The photographs of the models were retrieved from the website of Next Model Management (2015). The brand name and product information were added via a photo-editing tool. All advertisements portrayed a model wearing a blue denim jacket, the garment name, the product price, and the fictitious brand name. After exposure to each advertisement participants were asked to indicate the ethnicity that best suited the model. This was followed by questions about participant’s ethnicity, age, and educational level.
Jeans jackets were chosen as garment type, because it is one of the most commonly worn garment types worldwide (Cotton Incorporated, 2015). Furthermore, portraying a denim jacket allowed for the selection of pictures that showed models in a medium-wide picture frame. This frame made it easier for consumers to infer and observe the ethnicity of the model.
To determine which ethnicities suited the models best according to the participants all mentioned ethnicities were group based on their geographical area of origin (see Appendix 1). This division was based on the division made by CBS (Bevolking; kerncijfers, 2014). The results showed that for the White models all indicated ethnicities fall within regions that are mainly considered to inhabit White individuals. However, of the total ethnicities indicated, the majority of indications were appointed to ad model Roxanne (54.39%). Therefore she was chosen as the White model in the main study. The ethnicities indicated for the Black models fall within regions that are predominantly considered to inhabit Black individuals. However, of all the indicated ethnicities, those appointed to Monrose more clustered around Black-cultured regions (66.67%). Therefore Monrose was chosen as the Black model for the main study.
For the ambiguous ad models more diverse ethnic regions of origin were indicated. Of these models Neary and Ho were appointed the most diverse regions. However, when looking at the specific indicated regions indicated for Ho 25.00% clustered around Asia and 25.00% around Europe. For the model Neary the indicated ethnicities are less clustered (34.21% Europe, 23.67% Middle-America, and 18.42% North-Africa, see Appendix 1). These
percentages are much more dispersed compared to ad model Ho. Therefore Neary was chosen as the ambiguous model for the main study. Furthermore, the position of the product
information and the brand name were changed, in order to make sure that they were readable for participants (see Appendix 8).
Measures
Attitude toward the advertisement. Aad was measured with seven 7-point semantic differential scales (“Bad-Good”, “Not convincing-Convincing”; see Appendix 8). A similar scale was used by Martin Kwai-Choi Lee and Yang (2004). In this study participants had an average score of 3.88 (SD = 1.29, see Appendix 2). The Principle Component Analysis (PCA) indicated that all items correlated positively on a single dimension (eigenvalue 4.98; variance 71.17%), with item 4 having the highest factor loading (.89). The reliability of the scale was good with a Cronbach’s alpha of .94
Purchase intention. Purchase intention was measured with three 7-point semantic differential scales (“Not likely-Very likely”, “Not probable-Very probable”; see Appendix 8). A similar scale was used by Martin et al. (2004). In this study participants had an average score of 3.34 (SD = 1.66, see Appendix 2). The PCA indicated that all items correlated positively on a single dimension (eigenvalue 2.64; variance 87.95%), with item 2 having the highest factor loading (.96). The reliability of the scale was good, with a Cronbach’s alpha of .93.
Self-referencing. Self-referencing was measured with six 7-point scales (1 = Do not agree at all; 7 = Agree completely; see Appendix 8). A similar scale was used by Martin et al. (2004). In this study participants had an average score of 3.05 (SD = 1.37, see Appendix 2). The PCA indicated that all items correlated positively on a single dimension (eigenvalue 4.00; variance 66.63%), with item 2 having the highest factor loading (.87). The reliability of the scale was good, with a Cronbach’s alpha of .90.
Objective ethnicity. The ethnicity of the participants was determined with a single-item scale (1 = Dutch; 7 = Other; see Appendix 8). A similar scale is used in research by the Dutch Central Agency for Statistics (Centraal Bureau voor de Statistiek; Bevolking;
kerncijfers, 2014). In this study, the majority of participants were Dutch (52.00%), followed by Surinamese (15.50%), and Moroccan (9.50%). These participants were then divided into two groups (White 53.00% vs. non-White 47.00%). For this division the same standard as used by the fashion industry (Dries, 2014).
Covariates
Additional measures were included in this study with the purpose of controlling for any confounding that might occur.
Attitude toward the ad model. Amod measured with five 7-point semantic differential scales (“Not credible-Very credible”, “Not appealing-Very appealing”; see Appendix 8). A similar scale was used by Lee, Fernandez, & Martin (2002). In this study the participants had an average score of 3.94 (SD = 1.33; see Appendix 2). The PCA indicated that all items correlated positively on a single dimension (eigenvalue = 3.92; variance 78.41%), with item 4 having the highest factor loading (.91). The reliability of the scale was good, with a
Cronbach’s alpha of .93.
Felt ethnicity. Felt ethnicity was measured with a single 7-point semantic differential scale (“Not strongly at all-Very strongly”; see Appendix 8). A similar scale was used by Lee et al. (2002). In this scale participants had an average score of 4.96 (SD = 1.46, see Appendix 2) on fashion involvement.
Fashion involvement. Fashion involvement was measures by eight 7-point semantic differential scales (“Not important-Very important”, “Boring-Exciting”; see Appendix 8). A similar scale was used by Josiam, Kinley, and Kim (2005). In this study the participants had an average score of 5.12 (SD = 1.19, see Appendix 2). The PCA indicated that all items correlated positively on a single dimension (eigenvalue = 6.43; variance 71.49%), with item 7
having the highest factor loading (.89). The reliability of the scale was good, with a Cronbach’s alpha of .95.
Product use. The general use of denim jackets was measured with three 7-point scales (1 = Not probable at all; 7 = Very probable; see Appendix 8). These scales were based on sales numbers from (Cotton Incorporated, 2015). In this study participants had an average score of 2.81 (SD = 1.54, see Appendix 2). The PCA indicated that all items correlated
positively on a single dimension (eigenvalue = 2.30; variance 76.62%), with item 2 having the highest factor loading (.89). The reliability of the scale was good, with a Cronbach’s alpha of .85.
Furthermore, model recognition was used as a single-item scale to determine whether participants recognized the ad model (1 = No; 2 = Yes; Appendix), of which 90.5% indicated not to recognize the model. Participants were also asked to indicate their age. Finally,
participants were asked to indicate their educational level on a single-item scale (1 = Lower education; 8 = Other; see Appendix 8).
Results
Randomization check
To determine whether the random assignment of participants to the different
conditions was executed successfully randomization checks were conducted. ANOVAs were conducted with ad model ethnicity on each covariate. These analyses indicated not significant differences between the three conditions concerning felt ethnicity (F (2,197) = 1.70, p = .19, η2 = .02, product use (F (2, 197) = .06, p = .95, η2 = .00), and fashion involvement (F (2, 197) = 1.83, p = .16, η2 = .02). Furthermore Chi-square tests for education (χ2 (12) = 3.83, p = .99) and ethnicity (χ2 (14) = 20.19, p = .12), showed no significant differences between the three conditions. However, concerning Amod (F (2, 197) = 9.35, p = .00, η2 = .09) and age (F (2,
168) = 4.27, p = .02, η2 = .05) significant differences were found between the three groups. In order to ensure that Amod and age did not cause confounding effects during the analyses, Amod and age were controlled for when testing the hypotheses.
Manipulation check
As the stimulus material of the main study was constructed based on result from the pretest, the perceived ad model ethnicity was assessed. This was done, to ensure that participants in the main study perceived the ad model in the intended ethnicity. The results shown that the White (Black) model was predominantly perceived as White (Black) compared to the other models (see Appendix 3). The places of origin and ethnicities indicated for this model are in line with places of origin and ethnicities that are considered White (Black). The places of origin and ethnicities indicated for the ambiguous ad model were more diverse compared to the distinctive models. Therefore, it was assumed that the stimulus material had the intended effect on the perception of participants.
The effect of ad model ethnicity on Aad and purchase intention
To test the main effect of ad model ethnicity on Aad and purchase intention a
MANCOVA was conducted, including Amod and age as covariates. This was done via the post hoc method using the Bonferroni statistic. When controlled for Amod and age, using Pillai’s trace no significant results were found for ad model ethnicity V = .00, F (4, 322) = .16, p = .96, η2 = .00. This means that exposure to ad model ethnicity did not bring about significant differences in scores on Aad and purchase intention between the different conditions (see Table 1). As exposure to the ambiguous ad model did not brig about significant higher scores on Aad and purchase intention compared to exposure to distinctive ad models, H1 was
rejected.
To test the effect of ad model ethnicity on the extent of self-referencing an ANCOVA was conducted, including Amod and age as covariates. This was done via the post hoc method using the Bonferroni statistic. When controlled for Amod and age a significant effect was found for ad model ethnicity on self-referencing F (2, 166) = 3.56, p = .03, η2 = .04. The results show that no significant differences in the scores on self-referencing were found between the ambiguous and White condition (p = .11) or between the White and Black condition (p = 1.00; see Table 1). However, between the ambiguous and Black condition a significant difference was found (p = .05). Participants in the ambiguous condition had a significantly lower score on Aad. Because exposure to the ambiguous ad model did not cause a more positive self-referencing H2 was rejected.
Table 1. Effect of ad model ethnicity on Aad, purchase intention, and self-referencing
Ambiguous White Black
Aad 3.73 (1.01)a 3.53 (1.26)a 4.24 (1.24)a
Purchase intention 3.27 (1.54)a 3.07 (1.66)a 3.74 (1.65)a Self-referencing 2.78 (1.17)a 2.85 (1.31)ab 3.46 (1.38)b
Note. Mean scores depicted with standard deviation between brackets. Scores are controlled for Amod and age.
a, b Means scores of different superscripts (in rows) differ significantly on p < .05.
The moderating effect of objective ethnicity on self-referencing
To test whether the effect of ad model ethnicity on Aad and purchase intention is moderated by consumers’ objective ethnicity an ANCOVA. For this test Amod and age were included as covariates. The ANCOVA was conducted via the post hoc method using the Bonferroni statistic. Significant main effects of ad model ethnicity (F (2, 163) = 5.58, p = .01, η2 = .06) and objective ethnicity (F (1, 163) = 8,76, p = .00, η2 = .05) on self-referencing (see
Table 1). In accordance with the results from H2 a significant difference was found in the score on self-referencing between the ambiguous and Black condition (p = .02) and no significant difference between White and Black condition (p = 1.00). However, contrary to the results from H2, a significant difference was found between the ambiguous and White condition (p = .04). Furthermore, a significant difference in the score on self-referencing was found between the White and non-White condition (p = .00, see Appendix 4). However, no significant interaction between objective ethnicity and ad model ethnicity (F (2, 163) = .34, p = .71, η2 = .00; see Figure 1). Because significant interaction between ad model ethnicity and objective ethnicity was found H3 was rejected.
The mediating effect of self-referencing on Aad and purchase intention
To test whether the main effect of ad model ethnicity on Aad and purchase intention was mediated by self-referencing PROCESS analysis with model 4 were conducted,
controlling for Amod and age. For each outcome variable three separate comparisons between the conditions were made (i.e. ambiguous vs. White; ambiguous vs. Black; White vs. Black).
0 0,5 1 1,5 2 2,5 3 3,5 4
Ambiguous White Black
Sc o re o n se lf -r e fer e n ci n g
Figure 1. Effect of A
adand objective ethnicity
on self-referencing
White Non-white
These separate analyses allowed the assessment of the effect of each condition of ad model ethnicity on Aad and purchase intention mediated by self-referencing.
The results show that participants in the ambiguous condition had a significantly lower score on self-referencing compared to those in the White (b = -.44; p = .04, see Appendix 5) and Black (b = -.54; p = .02) condition. Furthermore, no significant difference was found between participants in the White and Black condition (b = -.10; p = .66). These results are in accordance with findings from H2. Secondly, the results indicated a significantly positive effect of ad model ethnicity on self-referencing (b = .14; p = .02) for all conditions.
Concerning the indirect effect of ad model ethnicity on Aad, through self-referencing, no significant effect was found between White and Black condition (b = .01, 95% [-.11; .05]). However, between the ambiguous and White condition (b = -.06, 95% [-.16; -.01]) and the ambiguous and Black condition (b = -.08, 95% [-.20; -.02]), no significant differences were found. This means that participants exposed to the ambiguous ad model had a significantly lower score on Aad, mediated through self-referencing, compared to those exposed to the ethnically distinctive ad models. Based on these results H4a was rejected.
Concerning the effect of ad model ethnicity on self-referencing significant differences were found between the ambiguous and both the White (b = -.44, p = .04) and Black (b = -.54, p = .02) condition. This means that exposure to the ambiguous model lead to significantly lower referencing. Furthermore, significant effects of ad model ethnicity on self-referencing were found (b = .53, p = .00). For the indirect effect of ad model ethnicity on purchase intention no significant difference was found between the White and Black
condition (b = -.05, 95% [-.34; .20]). However, between the ambiguous and White condition (b = .23, 95% [.49; .02]) and the ambiguous and Black condition (b = .28, 95% [.64; -.06]) significant differences were found. This means that participants in the ambiguous
referencing, compared to those in the White or Black condition. Based on these finding H4b was rejected.
Self-referencing moderated by objective ethnicity
To test whether the mediating effect of self-referencing on Aad and purchase intention was moderated PROCESS analyses were conducted using model 8, controlling for Amod and age. For each outcome variable three separate comparisons were made between the
conditions (i.e. ambiguous vs. White; ambiguous vs. Black; White vs. Black). Through these analyses the effect of ad model ethnicity on Aad and purchase intention, mediated through self-referencing, taking the objective ethnicity into account.
The results show that the direct effect of objective ethnicity on self-referencing was significant (see Table). When comparing both the White and the Black group to the
ambiguous group, all non-White participants had a significantly lower score on
self-referencing (b = -.45, p = .03). Also, when comparing the Black and the White group all non-White participants had a significantly lower score on self-referencing (b = -.61, p = .00).
When assessing the moderated mediation effect on Aad as outcome variable, the results show that for each comparison a significant effect of self-referencing on Aad was found (b = .17, p = .01). However, the results indicate that the mediating effect of self-referencing on Aad was not significantly moderated by objective ethnicity when comparing both the White and the Black group to the ambiguous group (b = -.04, 95% BACAB [-.16; .05]). Similarly, no significant moderated mediation was found when comparing the Black group to the White group (b = .05, 95% BCAB [-.07; .23]).
Regarding the assessment of the moderated mediation effect on purchase intention as the outcome variable, the results show a significant effect of self-referencing on purchase intention (b = .57, p = .00). No significant results were found to support that the mediating
process of self-referencing on purchase intention was moderated by objective ethnicity when comparing the Black and the White group to the ambiguous group (b = -.12, 95% BCAB [-.51; .21]). When comparing the Black group to the white group no significant moderated mediation effect was found (b = -.16, 95% BCAB [-.31; .61])
Conclusion
The aim of this study was to assess whether the use of ethnically ambiguous ad models in fashion advertising is more effective in developing more positive attitude toward the
advertisement and purchase intention amongst consumers. In doing so, the effect on attitude toward the advertisement and purchase intention was mediated through self-referencing and by taking objective ethnicity into account. These effect were then compared to the effect generated by distinctive ad models. Findings indicate that exposure to the ambiguous model caused lower self-referencing compared to exposure to distinctive ad models. As a result were lower attitude toward the advertisement and purchase intention were developed. Therefore, no evidence was found to support the effectiveness of ethnic ambiguity over distinctive ethnicity in advertising. Ethnic ambiguity, as a form of strategic ambiguity, has therefore not been supported to be effective as a cue for fashion advertising.
Discussion
Prior research has shown that the ethnicity of the ad model affects the attitude toward advertisement and purchase intention of consumers (Green, 1999; Torres & Briggs, 2007). Ethnicity helps individuals to distinguish themselves from others in social contexts (McGuire & McGuire, 1982). The ethnicity of minority consumers is easily noticed, because it is different from the majority. This makes consumers belonging to ethnic minorities more susceptible to cues about ethnicities. This causes them to feel targeted and engage in more self-referencing (Torres & Briggs, 2007). However, evidence form this study showed no
evidence to support that non-White consumers engaged in higher self-referencing upon exposure to the ambiguous model compared to White consumers. Furthermore, use of the ethnic models has proven to be more effective in causing more positive attitude toward the advertisement and purchase intention compared to use of the ambiguous model. Therefore, the claim by Daniel Peddle, that ambiguous models could be used by designers to target ethnic audiences without losing appeal to their main (White) audience, is not supported.
Limitations and recommendations
The lack of evidence to support the hypotheses in this study could be explained by a number of reasons. First, in this study the ‘face’ of the ad model, meaning her appearance, was used as a single cue for ethnicity (Betty, 2008). This means that the ethnic appearance of the ad model formed the only informative cue by which consumer could determine their ethnic similarity to that of the model’s. Perhaps additional information concerning the ad model’s ethnicity should have been given, such as mentioning the country of origin or the use of ethnic clothes. When consumers are confronted with ambiguous information they infer additional information that might help them to process the message (Washburn, Till, &
Priluck, 2004; Lenoir, Puntoni, Reed, & Verlegh, 2015). Providing this additional information gives consumers a clearer frame of reference through which to process the message (David,
Morrison, Johnson & Ross, 2002). Because these were not provided in this study, it could
have prohibited or impeded consumers from relating to ambiguous ad model, causing them to develop less favorable attitudes and intentions compared to exposure to distinctive ad models.
Furthermore, results indicate no significant differences between the exposures to White or Black models for attitude toward the model and purchase intention. This indicates that the use of ethnic models does not damage the purpose of advertising, which is a
the use of other types of ethnicities in fashion advertising is needed to determine whether this effect applies to more ethnicities.
Additionally, the product type presented in the advertisements could have been a reason for the lack of supportive evidence for the effectiveness of ethnic ambiguity as a strategy. A jeans jacket was chosen, because jeans are one of the most commonly worn garments in the world and the jacket allowed the selection of models in a medium frame (Cotton Incorporated, 2015). However, the commonality of this product type could have affected the development of consumers’ attitude toward the ad and purchase intention. Because jeans form a very common garment, it could cause consumers to pay less attention advertisement. As a result consumers could have been less engaged in self-referencing, which resulted in less favorable attitude toward the advertisement and purchase intention. Therefore, further research on the type of products (i.e. high involvement vs. low involvement product types) portrayed when testing ethnic ambiguity is needed.
Practical implications
Based on the findings from this study no evidence was found that supported the claim by the fashion industry that ethnically ambiguous models could help them to appeal to minority target groups without losing appeal to the (main) White target audience (Betty, 2008). Rather than aiding the development of more positive attitude toward the advertisement and purchase intention, the use of ambiguous models resulted in less favorable attitude and intention, through the process of self-referencing. However, the use of distinctive models did proof to be successful. As no significant difference could be found on the use of White or Black models on attitude toward the advertisement and purchase intention, the use of ethnic models has proven to be effective more effective in appealing to ethnically diverse
(Freeman, 2014). Evidence indicates that marketers can use Black models to appeal to their main (White) target audience.
References
Ajzen, I. and Fishbein, M. (1977). Attitude-behavioral relations: A theoretical analysis and
review of empirical research. Psychological Bulletin, 84(5), 888-918.
Ajzen, I. and Fishbein, M. (1980). Understanding Attitudes and Predicting Social Behavior.
Englewood Cliffs, NJ: Prentice Hall.
American Psychological Associations, (2003). Guidelines on multicultural education,
training, research, practice, and organizational change for psychologists. American Psychologists, 58(5), 377-402.
Bellezza, F. S. (1984). The self as a mnemonic device: The role of internal cues. Journal of
Personality and Social Psychology, 47(3), 506-516.
Betty. (2008, August 27). New generation: Daniel Peddle. Retrieved from [Blog]
http://models.com/oftheminute/?p=4111
Bevolking; kerncijfers (2014, April 8). Bevolking; kerncijfers retrieved from CBS [web page]
http://statline.cbs.nl/Statweb/selection/?VW=T&DM=SLNL&PA=37296NED&D1=0-2,19-28,52-58,68&D2=0,10,20,30,40,50,(l-1)-l&HDR=G1&STB=T
Burnkrant, R. E., & Unnava, H. R. (1995). Effects of self-referencing on persuasion. Journal
of Consumer Research, 22(1), 17-26.
Cotton Incorporated Lifestyle Monitor: Easing apparel sourcing cost pressures (2015, April
Deshpandé, R., & Stayman, D. M. (1994). A tale of two cities: Distinctiveness theory and
advertising effectiveness. Journal of Marketing Research, 16(4), 57-64.
David, P., Morrison, G., Johnson, M. A., & Ross, F. (2002). Body image, race, and fashion
models social distance and social identification in third-person effects. Communication
Research, 29(3), 270-294.
Debevec, K., & Iyer, E. (1988). Self‐referencing as a mediator of the effectiveness of sex‐role portrayals in advertising. Psychology & Marketing, 5(1), 71-84.
Disgrasian. (2008, October 10). The ‘ethnically ambiguous’ are so last season. Retrieved
From [online newspaper] http://www.huffingtonpost.com/disgrasian/the-ethnically
unambiguou_b_129365.html
Dries, K. (2014, February 14). New York fashion week: diversity talks but white faces walk.
Retrieved from [blog] http://jezebel.com/new-york-fashion-week-diversity-talks-but-white-faces-1522416724
Escalas, J. E. (2007). Self‐referencing and persuasion: narrative transportation versus analytical elaboration. Journal of Consumer Research, 33(4), 421-429.
Freeman, H. (2014, February 18). Why black models are rarely in fashion. Retrieved from
[online newspaper]
http://www.theguardian.com/commentisfree/2014/feb/18/black-models-fashion-magazines-catwalks
identification, media placement, and degree of racial composition. Journal of
Advertising, 28(1), 49-64.
Kwai-Choi Lee, C., Fernandez, N., & Martin, B. A. (2002). Using self-referencing to explain
the effectiveness of ethnic minority models in advertising. International Journal of
Advertising, 21(3), 367-380.
Lavidge, R. J. and Steiner, G. A. (1961). A model for predictive measurements of advertising
effectiveness. Journal of Marketing 25(6), 59-62.
Lutz, R. J. (1985). Affective and cognitive antecedents of attitude toward the ad: A conceptual
framework. In L.F. Alwitt and A. A. Mitchell (Eds.), Psychological processes and
advertising effects; Theory, research and application (45-63). Hillsdale, NJ: Lawrence
Erlbaum Associate.
Markus, H., & Wurf, E. (1987). The dynamic self-concept: A social psychological
perspective. Annual review of psychology, 38(1), 299-337.
Martin, B. A., Kwai-Choi Lee, C., & Yang, F. (2004). The influence of ad model ethnicity
and self-referencing on attitudes: Evidence from New Zealand. Journal of
Advertising, 33(4), 27-37.
McGuire, W. J., & McGuire, C. V. (1982). Significant others in self-space: Sex differences
and developmental trends in the social self. In J. Suls (Ed.), Psychological
perspectives on the self, 71-96. Hillsdale, NJ: Erlbaum.
identity. Retrieved from [blog]
https://www.notjustalabel.com/editorial/fashion-advertising-campaign-big-business-brand-identity
Next (New York). Retrieved from [web page] http://nextmanagement.com/
Oyserman, D. (2001). Self-concept and identity. In A. Tesser & N. Schwarz (Eds.), The
blackwell handbook of social psychology, (499-517). Malden, MA: Blackwell.
Pantoja, A. D., & Segura, G. M. (2003). Does Ethnicity Matter? Descriptive representation in
legislatures and political alienation amongst Latinos. Social Science Quarterly, 84(2), 441-460.
Pilkington, E. (2008, June 20). Italian Vogue’s black issue breaks fashion barrier. Retrieved
from [online newspaper]
http://www.theguardian.com/lifeandstyle/2008/jun/20/fashion.usa
Puntoni, S., Vanhamme, J., & Visscher, R. (2011). Two birds and one stone. Journal of
Advertising, 40(1), 25-42.
Sauers, J. (2013, February 18). Fashion week’s models are getting whiter. Retrieved from [blog]
http://jezebel.com/5985110/new-york-fashion-weeks-models-are-getting-whiter
Sebastian, M. (2014). Native is the new Black: fashion editors are producing brand content.
Retrieved from [blog] http://adage.com/article/media/fashion-editors-producing-branded-content/292560/
Tajfel, H. (1982). Social psychology of intergroup relations. Annual review of
psychology, 33(1), 1-39.
moderating role of involvement. Journal of Advertising, 36(3), 97-108.
Washburn, J. H., Till, B. D., & Priluck, R. (2004). Brand alliance and customer‐based brand‐equity effects. Psychology & Marketing, 21(7), 487-508.
Wilson, J. (2014, February 18). Diversity on the runway is slightly better, but it’s still super white
(INFOGRAPHIC). Retrieved from [online newspaper]
http://www.huffingtonpost.com/2014/02/18/runway-diversity-report-jezebel-new-york-fashion-week_n_4808263.html
Young, S. (2015, April 16). Insights from PRS eye-tracking of ad readership. Perception
Research Services. Retrieved from [web page]
http://www.prsresearch.com/prs-insights/article/a-campaign-for-improvement-getting-more-from-print-advertising/
32 Appendix 2. Mean scores of and standard deviation ad model ethnicity on measures
M SD
Attitude toward the advertisement (Aad) 3,88 1,29
Purchase intention 3,34 1,66 Self-referencing 3,05 1,37 A ppe n dix 1. Pre test fre que nc y of pe rc eiv ed ad mode l ethni cit ies Cond ition M od el n am e G eog rap h ical re gion S ub -S aha ra n Af ric a Nor th -Af ric a Ea st -Asia W est -Asia Nor th -Ame ric a Midd le -Ame ric a S outh -Ame ric a Eur ope No spe cific countr y mentioned T otal W hi te J C re w 0 0 0 0 2 0 0 24 0 26 R ox anne 0 0 0 0 0 0 0 31 0 31 Ambig uous Ne ar y 1 7 0 5 0 9 1 13 2 38 Ho 0 6 10 4 0 2 3 10 5 40 Ar ca njo 0 0 1 6 1 1 1 19 1 30 B lac k Monrose 4 0 0 0 0 12 12 4 4 36 J Cre w 4 1 0 2 1 9 12 6 6 41
34 M ain st udy f re que nc y of p er ce iv ed ad m ode l e thnic iti es M od el n am e G eog rap h ical r egion S ub -S aha ra n Af ric a Nor th -Af ric a Ea st -Asia W est -Asia Nor th -A mer ica Midd le -Ame ric a S outh -Ame ric a Eur ope No spe cific countr y mentioned T otal R ox anne 0 1 1 1 2 3 3 71 2 84 Ne ar y 0 9 4 5 0 25 19 31 7 100 Monrose 10 4 0 1 3 28 37 5 10 98 A ppe n dix 4. M ain effec t of ad mode l ethnicity and o bjec ti ve e thni cit y on s elf -re fere nc ing A d m od el e thn icity Ambig uous W hit e B lac k O b je ctive e thn icity
A pp en di x 5 . Effec ts of ad mode l ethnicity on A ad a nd purc hase int enti on me diat ed by se lf -re fere n cin g A ad (D V) G rou p (r ef er en ce ) X→M ed Me d →y x→y ( tot al) x→y (dire ct) In d ire ct e ff ec t [95% B CaC I] A m b ig u ou s (Wh ite ) -.44 (. 21)* .14 (.06)* -.10 (. 16) -.04 (. 16) -.06 (. 04) [ -.16; -.01] * A m b ig u ou s (B la ck ) -.54 (. 22)* .14 (.06)* -.12 (. 17) -.04 (. 17) -.08 (.04) [ -.20; -.02] * Wh ite (Bl ac k) -.10 (. 22) .14 (.06)* -.01 (. 17) -.00 (. 17) -.01 (. 04) [ -.11; .05] P u rc h as e in te n tion (D V) A m b ig u ou s (Wh ite ) -.44 (. 21)* .53 (.09)* -.04 (. 27) .20 (.2 5) -.23 (. 12) [ -.49; -.02] * A m b ig u ou s (B la ck ) -.54 (. 22)* .53 (.09)* -.09 (. 28) .19 (. 26) -.28 (.14) [ -.64; -.06] * Wh ite (Bl ack ) -.10 (. 22) .53 (.09)* -.06 (. 28) -.00 (. 26) -.05 (. 14) [ -.34; .20] *Mea ns a si g nific ant di ff er enc e wa s found on p < .05
Appendix 7. Questionnaire Pretest Geïnformeerde Toestemming A ppe n dix 6. M ode rated m ediat ion of se lf -re fere n cing and objec ti ve e thni cit y A a d (D V) G rou p (r ef er en ce ) Ob je ctive e thn icity (z ) gr ou p (r ef er en ce ) X → M Z → M M → Y X → Y M od er ate d m ed iation [95% B CaC I] A mbi g uous (W hit e) W hit e (n on -W hit e) -.19 (. 55) -.45 (. 21) * .17 (. 06) * .69 (. 42) -.04 (. 05) [ -.16; . 05] A mbi g uous ( B lac k ) W hit e (n o n -W hit e) -.30 (. 55) -.45 (. 21) * .17 (. 06) * .70 (. 42) -.04 (. 05) [ -.16; . 05] B lac k (W hit e) W hit e (n on -W hit e) -.32 (. 59) -.61 (. 21) * .17 (. 06) * -.46 (. 45) .05 (. 07) [ -.07; . 23] P u rc h ase in te n tion (D V) A mbi g uous (W hit e) W hit e (n on -W hit e) -.19 (. 55) -.45 (. 21) * .57 (. 09) * .40 (. 67) -.12 (. 17) [ -.51; . 21] A mbi g uous ( B lac k ) W hit e (n on -W hit e) -.30 (. 55) -.45 (. 21) * .57 (. 09) * .41 (. 67) -.12 (. 17) [ -.51; . 21] B lac k (W hit e) W hit e (n on -W hit e) -.32 (. 59) -.61 (. 21) * .57 (. 09) * -.05 (. 71) -.16 (. 23) [ -.31; .61] *Me ans a si g nific ant di ff ere nc e wa s found on p < .05
Hierbij verklaar ik dat ik op een duidelijke manier geïnformeerd ben over de aard en de methode van dit onderzoek, zoals die beschreven is in de online uitnodiging voor dit onderzoek.
Ik ga vrijwillig akkoord met deelname aan dit onderzoek. Hiermee behoud ik het recht mijn toestemming voor dit onderzoek terug te trekken, zonder daar enige verklaring voor te hoeven geven. Ik ben mij ervan bewust dat ik mijn deelname op ieder moment in het onderzoek stop kan zetten
Als mijn onderzoeksresultaten gebruikt worden in wetenschappelijke publicaties of op enige wijze publiek gemaakt worden, zal dit op een dusdanige wijze geschieden dat mijn
anonimiteit volledig gewaarborgd blijft. Mijn persoonlijke gegevens zullen niet doorgespeeld worden aan derden zonde mijn expliciete toestemming
Als ik meer informatie wil ontvangen over dit onderzoek, nu of in de toekomst, kan ik contact opnemen met M. Victoriashoop (modereclame@hotmail.com). Mocht ik klachten hebben over dit onderzoek, dan kan ik contact opnemen met een aangewezen lid van de Ethische Commissie, vertegenwoordigd door ASCoR op het volgende adres: ASCoR secretariaat, Ethische Commissie, Universiteit van Amsterdam, Postbus 15793; 1001 NG Amsterdam; 020-5253680; ascor-secr-fmg@uva.nl.
Ik begrijp de hierboven gepresenteerde tekst en ga akkoord met deelname aan dit onderzoek. Akkoord
Je krijgt nu een aantal advertenties te zien. Wij zijn geïnteresseerd in hoe de modellen uit de advertenties overkomen en vragen je daarom om jouw mening te geven.
Welke etnische achtergrond(en) past/passen volgens jou het best bij dit model?(Meerdere antwoordmogelijkheden zijn toegestaan)
White ad models J. Crew Roxanne Ambiguous ad models Neary Ho Arcanjo Black ad models Monrose J, Crew
o Nederlands o Surinaams o Antilliaans o Marokkaans o Turks
o Afrikaans, namelijk (noem land) ____________________ o Aziatisch, namelijk (noem land) ____________________ o Anders, namelijk (noem land) ____________________
Wat is jouw leeftijd? __________________
Wat is jouw hoogst genoten opleiding? Lager onderwijs VMBO (LBO/MAVO) HAVO VWO MBO Universiteit Anders, namelijk ____________________
Wat is jouw geslacht? Man
Wat is jouw etnische achtergrond? Nederlands Surinaams Antilliaans Marokkaans Turks
Afrikaans, namelijk (noem land)____________________ Aziatisch, namelijk (noem land) ____________________ Anders, namelijk (noem land) ____________________
Debriefing
Bedankt voor je deelname aan dit onderzoek. Hierbij wordt nogmaals benadrukt dat deelname volledig anoniem is en dat er op vertrouwelijke wijze omgegaan zal worden met de verkregen informatie uit dit onderzoek.
Mocht je op de hoogte willen blijven van de resultaten uit dit onderzoek, dan kun je contact opnemen met M. Victoriashoop via modereclame@hotmail.com.
Klik op de onderstaand knop om de vragenlijst af te sluiten.
Appendix 8. Questionnaire main study Introduction
Beste participant,
Hierbij wil ik je uitnodigen om deel te nemen aan een onderzoek dat gedaan wordt onder toezicht van de Graduate School of Communication, onderdeel van de Universiteit van Amsterdam.
Het onderzoek waarvoor ik jouw deelname vraag heet ‘De onduidelijke look’. Tijdens het online experiment krijg je een aantal reclamebeelden te zien van het merk MA’VI. Bij ieder beeld wordt je gevraagd een vraag te beantwoorden, gevolgd door een aantal vragen over jezelf. Alleen vrouwen mogen deelnemen aan dit experiment. Het doel van dit onderzoek is om inzicht te krijgen in de manier waarop consumenten reclameboodschappen verwerken. Het experiment duurt ongeveer 7 minuten.
Omdat het onderzoek uitgevoerd wordt onder de verantwoordelijkheid van ASCoR, Universiteit van Amsterdam, wordt gegarandeerd dat:
1) ouw anonimiteit gewaarborgd wordt en dat jouw persoonlijke informatie onder geen beding niet doorgespeeld wordt aan derde partijen, tenzij je daar expliciet toestemming voor hebt gegeven.
2) Je participatie aan dit onderzoek mag weigeren of stop mag zetten zonder daar enige reden voor af te hoeven leggen. Tot 24 uur na deelname mag je jouw toestemming om jouw
3) Deelname aan dit onderzoek stelt je niet bloot aan enig merkbaar risico of ongemak, de onderzoeker zal je niet opzettelijk misleiden en je zult geen expliciet beledigend materiaal te zien krijgen.
4)Niet later dan vijf maanden na afronding van het onderzoek, zal de onderzoeker in staat zijn het onderzoeksrapport te verstrekken en de algemene bevindingen uit te leggen.
Voor meer informatie over het onderzoek en de uitnodiging om deel te nemen, kun je op elk moment contact opnemen met de onderzoeksleider M. Victoriashoop
(modereclame@hotmail.com).
Mocht je klachten hebben over het verloop van het onderzoek en de procedures die het vergt als gevolg van jouw deelname aan het onderzoek, kun je contact opnemen met het
aangewezen lid van de Ethische Commissie dat ASCoR vertegenwoordigd op het volgende adres: ASCoR Secretariaat, Ethische Commissie, Universiteit van Amsterdam, Postbus 15793, 1001 NG Amsterdam; 020-5253680; ascor-secr-fmg@uva.nl.
Alle klachten of opmerkingen zullen vertrouwelijk behandeld worden.
Hiermee hoop ik je voorzien te hebben van voldoende informatie. Ik wil van de gelegenheid gebruik maken om je bij voorbaat te bedanken voor de deelname aan dit onderzoek, hetgeen ik op prijs stel.
Met vriendelijke groet,
Geïnformeerde Toestemming
Hierbij verklaar ik dat ik op een duidelijke manier geïnformeerd ben over de aard en de methode van dit onderzoek, zoals die beschreven is in de online uitnodiging voor dit onderzoek.
Ik ga vrijwillig akkoord met deelname aan dit onderzoek. Hiermee behoud ik het recht mijn toestemming voor dit onderzoek terug te trekken, zonder daar enige verklaring voor te hoeven geven. Ik ben mij ervan bewust dat ik mijn deelname op ieder moment in het onderzoek stop kan zetten.
Als mijn onderzoeksresultaten gebruikt worden in wetenschappelijke publicaties of op enige wijze publiek gemaakt worden, zal dit op een dusdanige wijze geschieden dat mijn
anonimiteit volledig gewaarborgd blijft. Mijn persoonlijke gegevens zullen niet doorgespeeld worden aan derden zonde mijn expliciete toestemming.
Als ik meer informatie wil ontvangen over dit onderzoek, nu of in de toekomst, kan ik contact opnemen met M. Victoriashoop (modereclame@hotmail.com). Mocht ik klachten hebben over dit onderzoek, dan kan ik contact opnemen met een aangewezen lid van de Ethische Commissie, vertegenwoordigd door ASCoR op het volgende adres: ASCoR secretariaat, Ethische Commissie, Universiteit van Amsterdam, Postbus 15793; 1001 NG Amsterdam; 020-5253680; ascor-secr-fmg@uva.nl.
Ik begrijp de hierboven gepresenteerde tekst en ga akkoord met deelname aan dit onderzoek. Akkoord
Wat is jouw geslacht? Man
Vrouw
Debriefing geslacht
Bedankt voor je interesse. Helaas zijn wij voor dit onderzoek specifiek op zoek naar vrouwelijke participanten. Klik op de onderstaande knop om het onderzoek af te ronden.
Heb je al eerder deelgenomen aan het onderzoek 'De onduidelijke look'? Nee
Ja
Debriefing deelname pretest
Bedankt voor je interesse in dit onderzoek. Helaas zijn wij specifiek op zoek naar
participanten die niet eerder hebben deelgenomen aan het onderzoek 'De onduidelijke look' Klik op de onderstaande knop om dit onderzoek af te ronden.
Instructie
Stel je voor dat je een modetijdschrift aan het doorbladeren bent. Je krijgt nu een advertentie uit het tijdschrift te zien. Wij zijn geïnteresseerd in hoe het model uit de advertentie op jou overkomt. Daarom vragen wij je om jouw mening te geven.
Welke afkomst past volgens jou het best bij dit model?(meerdere antwoorden zijn toegestaan) o Nederlands o Surinaams o Antilliaans o Marokkaans o Turks
o Afrikaans (noem land) ____________________ o Aziatisch (noem land) ____________________
o Anders, namelijk (noem land) ____________________
Welke etnische achtergrond past volgens jou het best bij dit model?(Meerdere antwoordmogelijkheden niet toegestaan)
Blank Zwart Aziatisch Latina
Attitude toward the advertisement
Ik vind de advertentie die ik net gezien heb
Slecht ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ Goed
Niet overtuigend ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ Wel overtuigend
Niet informatief ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ Wel informatief
Niet interessant ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ Wel interessant
Niet plezierig ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ Wel plezierig
Niet aantrekkelijk ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ Wel aantrekkelijk
Niet aangenaam ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ Wel aangenaam
Purchase intention
Ik zou het product uit de advertentie die ik net gezien heb kopen
Waarschijnlijk niet ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ Waarschijnlijk wel
Niet aannemelijk ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ Wel aannemelijk
Uitgesloten ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ Niet uitgesloten
Niet interessant ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ Wel interessant
Niet plezierig ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ Wel plezierig
Niet aantrekkelijk ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ Wel aantrekkelijk
Attitude toward the mode
Ik vond het model uit de advertentie die ik net gezien heb
Niet geloofwaardig ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ Erg geloofwaardig
Niet aantrekkelijk ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ Erg aantrekkelijk
Niet overtuigend ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ Erg overtuigend
Niet aangenaam ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ Erg aangenaam
Self-referencing
Geef aan in hoeverre je het eens bent met de volgende stellingen
Helemaal mee oneens Mee oneens Een beetje mee oneens Noch mee oneens, noch mee eens Een beetje mee eens Mee eens Helemaal mee eens De advertentie heeft betrekking op mij ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ Ik kan mezelf gemakkelijk vergelijken met het model uit de
advertentie ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ
De advertentie lijkt te zijn gemaakt met mij
in gedachten ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ
Ik kan gemakkelijk een vergelijking maken tussen mezelf en het model uit de
advertentie ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ
Ik kan gemakkelijk voorstellen dat ik het product uit de
advertentie gebruik ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ ᴏ
Het model uit de advertentie spreekt voor een doelgroep waar ik onderdeel van
