Advertisement research : effectiveness of ads that use female role portrayals

(1)

Advertisement research:

Effectiveness of ads that use female role

portrayals

Master thesis author: Angela Pruteanu (10856617)

Supervisor: Dr. Frauke Mattison Thompson

MSc. In Business Administration – Marketing Track

University of Amsterdam

(2)

2 Statement of Originality

This document is written by Student Angela Pruteanu who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

3

ACKNOWLEDGEMENTS

I would like to thank my supervisor, Dr. F.H. Mattison Thompson, for her assistance, genuine encouragement in this journey called “Master Thesis” and her strong belief in the end result.

Ms. Mattison Thompson experience in doing research helped me in certain stages of my thesis. Her confidence in my ability to deliver the master thesis in the first submission period, despite of my full-time internship, encouraged me to push my limits and accomplish this goal.

Moreover, I would like to thank my partner, Alper Ozyurt, for his simon-pure passion about gender parity topic and the related topic of ideologies towards women roles in societies. Also, for his endless support and contribution in proofreading this paper.

Finally, I would like to thank my parents for believing in me and for their patience in the busyness of the last few months.

(5)

5

ABSTRACT

When it comes to the attitudinal reactions towards stereotypic advertising, the findings are contradictory and scarce. Nevertheless, by using evolutionary theory, this paper attempts to explain female responses to career-oriented versus decorative women role portrayals in advertisements. Prior studies endeavored to examine women responses only to decorative female portrayals in ads. By contrast, this study focused on expanding the scope by juxtaposing women attitudes towards ads and brands that leverage decorative and career-oriented female role portrayals in their advertising.

Relevant hypotheses were tested by conducting an experiment. The experiment examined how romantic and fear-inducing content influences the effectiveness of ads that use female role portrayal. The results indicate that fear will lead to more favorable attitudes toward the ad and the brand when a career-oriented women portrayal than a decorative one is used. Additionally, romantic desire will lead to more favorable attitude toward the ad when the female model is depicted as decorative object rather than as career-oriented.

(6)

6

INTRODUCTION

Women portrayal in advertising is a topic that has extensively been touched upon in many research papers over the last five decades. The main motivation behind the interest is the persistence and extensive adoption of stereotypic portrayal in advertising, and the possible social impact of it (Gulas & McKeage, 2000) in fortification or modification of society’s beliefs and norms, according to Pollay work (as cited in Plakoyiannaki, Mathioudaki, Dimitratos, & Zotos, 2008). For instance, women depiction in advertising has triggered societal concerns regarding the pressure women might face due to idealized beauty images we see in media (Bissell & Rask, 2010). A significant stream of articles is also available on how/whether the way women are portrayed has changed over time due to the contemporary structure of the society and women’s movement (Grau & Zotos, 2016; Zotos & Tsichla, 2014).

According to Baxter, Kulczynski, & Ilicic (2016), most of the research in this field are content-analysis driven ones (e.g. Baxter et al., 2016; Eisend, Plagemann, & Sollwedel, 2014; Knoll, Eisend, & Steinhagen, 2011; Plakoyiannaki et al., 2008; Plakoyiannaki & Zotos, 2009), and in almost all the cases the focus is on the way women are portrayed in traditional media like print and TV (Knoll et al., 2011; Plakoyiannaki & Zotos, 2009). When it comes to the attitudinal reactions towards stereotypic advertising, the findings are contradictory and usually provide weak theoretical justification (Kyrousi, Panigyrakis, & Panopoulos, 2016, p. 773). Videlicet, some papers found that women have more favorable attitudes toward the ads which portray them in a traditional way, whereas others found evidence for positive attitudes towards ads that portray women in a non-traditional way, i.e. career oriented (as cited in Zawisza & Cinnirella, 2010, p. 1769).

(7)

7 However, little research is available on the response of women to ads that portray them in a stereotypic way, taking into consideration the emotionally arousing context that the ads are placed in. The reason why this is interesting to research is the fact that advertisements are always placed alongside a text in case of print advertising, after a TV programme in case of television or, lately, with the digital age, ads are placed after video content (e.g. YouTube) or as banner ads on different websites.

Therefore, this study will try to shed light on this issue by addressing the gap of whether the responses of female audience to stereotypic ads are influenced by the elicited emotions (i.e. fear

and romantic desire) of a preceding online medium,by using the evolutionary approach. More

concretely, it will cast light on whether the emotionally inducing context (i.e., love and fear triggered by what consumers previously read or saw) influences the attitudes of female consumers towards brands and advertisements that depict women weak (i.e. decorative) versus strong and independent (i.e. career-oriented). The practical implication of the findings is expected to help managers make better decisions in regards to advertisement execution strategies and message targeting.

The following literature review will first focus in brief on the existing literature on gender role in advertising, by touching upon some core definitions and categories of female portrayals. Subsequently, the extant literature on effectiveness of traditional and decorative versus non-traditional female advertising will be briefly discussed. Then, the evolutionary psychology will be introduced as a way to explain the responses to advertising. The literature review section will conclude with an identified research gap and question, a set of proposed hypothesis and the expected contributions to theory and practice this study will bring.

(8)

8

LITERATURE REVIEW

Female Role Stereotypes and Advertising

To start off, gender stereotypes in advertising definition is needed for a better understanding of the matter. According to Ashmore and Del Boca, gender stereotypes refer to beliefs that a set of attributes differentially define men and women (Zotos & Tsichla, 2014). Moreover, gender stereotypes in advertising relates to the propensity of advertisements to depict men and women according to some pre-established traits, role behaviors, occupational status and physical characteristics (Eisend, 2010). Stereotypical featuring of women in advertising normally involves depiction of women in traditional, decorative, and non-traditional fashion or as equal to men (Plakoyiannaki & Zotos, 2009). Description of aforementioned categories can be seen in Figure 1. A number of scholars provided a general perspective of existing studies, due to the large amount of

research in this area. For example,Zotos and Tsichla (2014) formed a holistic perspective on the

female stereotypes phenomenon in print advertising from the origins. After a review of articles in this field, the authors concluded that female stereotyping is still alive, however slight changes in types and patterns occurred in comparison with findings from few decades ago. For instance, traditional portrayal of women in dependency and housewife roles seem to have reduced to the benefit of decorative images (Zotos & Tsichla, 2014).

What is more, the coding instrument (Figure 1) which contains categories for female role stereotypes proposed and embraced by many studies in this field for over the past several decades was suggested to be rethought (Zotos & Tsichla, 2014). This was recommended by the authors as a consequence of nowadays internet context and the companies’ extensive use of online advertising. Nonetheless, to the author of this thesis’ knowledge, there is just one study that suggested that

(9)

9 the coding scheme is still relevant and its usage can be also extended to virtual contexts

(Plakoyiannaki et al., 2008). The findings ofPlakoyiannaki et al. (2008) indicate that “traditional” or

“decorative” ways of female portrayal are evident in all audience contexts; to wit female, male and general-audience web pages, nevertheless some “non-traditional” roles are also present.

Another holistic outlook on the studies of advertising and gender stereotypes since 2010, found a recent trend in the advertising area called fermentising which represents use of ‘empowered’ women’ in advertising (Grau & Zotos, 2016). Therefore, a depiction of women in media outside the stereotypic spectrum can be lately observed as well. However, as mentioned previously, decorative roles are still a frequent occurrence alongside with traditional ones.

Figure 1.

Coding instrument: Categories for female role stereotypes

Category Description

Women in Traditional Roles:

- Dependency - Housewife

Dependent on male’s protection; in need of reassurance; making unimportant decisions.

Woman’s place is at home concerned with tasks of housekeeping; primary role is to be a good wife.

Women in pursue of physical attractiveness and beauty.

Sex is related to product.

Engaged in activities outside the home (e.g. football, golf).

Professional occupations; non-professional; entertainer; blue-collar. Women the expert.

Women depicted as equal to men.

Women in Decorative Roles:

- Women concerned with physical attractiveness - Women as sex objects

Women in Non-traditional Roles: - Women in non-traditional activities - Career-oriented - Voice of authority

Women portrayed as equal to men:

- Neutral

(10)

10 Effectiveness of Traditional and Decorative vs. Non-traditional women portrayal in advertising

Concerning the literature with regards to consumers’ responses to stereotypic advertisements, most of the scholars agree that women have a tendency to believe that ads in general do not portray them in a realistic fashion (Kyrousi et al., 2016). Even so, the few studies which analyzed explicitly attitudinal reactions to certain stereotypic ads found diverging results in terms of female consumers’ attitude towards those ads.

Respectively, in terms of effectiveness of traditional versus non-traditional gender advertising, a main issue refers to the fact that findings on female portrayal are contradictory. On the one hand, women’s preference for non-traditional female featuring in ads (i.e., “superwomen” or egalitarian portrayals) rather than traditional ones was noted by the Jaffe and Berger, and the Orth and Holancova studies (as cited in Zawisza & Cinnirella, 2010). On the other hand, a recent study by Vantomme et al. concluded lower explicit preference, yet higher implicit preference, for ads depicting women in a traditional role in comparison to the ad featuring females in non-traditional way for a fictitious mobile phone (as cited in Zawisza & Cinnirella, 2010).

Indeed, when it comes to advertising effectiveness, Eisend et al. (2014) emphasizes the fact that

stereotypic depictions “can be helpful or detrimental”. The study looked at how the presence of humor is impacting the effectiveness of advertising which also includes gender-role portrayal. According to the results, humor is found to improve consumers’ attitude towards the ad and brand, especially if non-traditional stereotyping is used, and women audience tend to evaluate gender depictions more positively in humorous ads, in particular when non-traditional stereotyping is used instead of traditional one (Eisend et al., 2014). Nevertheless, the authors

(11)

11 concluded that there is a need to examine other factors than humor that could possibly influence consumers’ attitude to role portrayals in advertisements.

Another paper that analyzed attitudinal reactions towards the stereotypic ads which portray females in decorative way, found that women’s attitude can influenced by the medium context. Namely, when women are primed by the context (i.e. article they read, video they see) to engage in a self-promotion competitive strategy (i.e., acquiring traits that are favored by the opposite sex), their attitude towards the ad is more favorable (Kyrousi et al., 2016). According to the authors, the opposite is happening when primed to engage in competitor derogation, as in the latter case women perceive the female model in the ad as a potential rival and tend to derogate her physical attractiveness (Kyrousi et al., 2016). This article used the evolutionary theory to explain the reasoning behind women reactions only towards advertisements that depict the same-sex others in a decorative way. In this manner, it will be worthwhile to extend this approach over the non-traditional women role portrayals in advertising as well.

The evolutionary theory and emotions

Evolutionary psychology, according to Saad and Gill, represents “a paradigm that seeks to account for cultural phenomena via the biological underpinnings of humans” and it can provide explanations for many facets of consumer behavior, even attitudinal or behavioral reactions to advertising (as cited in Kyrousi et al., 2016, p. 772; Saad, 2004). That is to say, the principles at the

core of this philosophy are used to explain diverse aspects of human behaviorsthat derive from

fundamental adaptive problems like mate attraction, fear of predators or making friends (Griskevicius et al., 2009).

The evolutionary approach has been recently adopted by scholars to investigate responses to advertisements that use persuasion heuristics (Griskevicius et al., 2009) or, for instance, female

(12)

12 stereotypic representations (Kyrousi et al., 2016; Saad, 2004). How the evolutionary psychology helps to understand consumers’ responses to ads that depict women in a decorative way, in particular, was addressed by Kyrousi et al. (2016) in case of women respondents and by Saad (2004) in case of male respondents. The evolved mechanism of mate attraction was used to explain men’s attitudes towards ads that represent the opposite sex in a decorative way (Saad, 2004). Alternatively, Kyrousi et al. (2016) used female-to-female competition as a mechanism within the evolutionary psychology field to explain women’s responses to ads that use female decorative role depiction. Nevertheless, no study compared female consumers’ attitudes towards the ads and brands that portray women in a decorative versus non-traditional (i.e. career-oriented) fashion, by capitalizing on evolutionary theory principles.

From an evolutionary standpoint, Keltner et al. (cited in Griskevicius et al., 2009, p. 386) states that emotions activates certain motivational subsystems that manage the energy in ways that could deal with the adaptive problem. In this manner, emotions trigger certain cognitions and behaviors that could successfully solve the adaptive problem such as attracting mates or protecting from predators (Griskevicius et al., 2009).

Likewise, emotions can be activated by arousing stimuli (Griskevicius et al., 2009). As a consequence, according to Keltner et al. (cited in Griskevicius et al., 2009, p. 386), these specific emotion states should motivate individuals to act and think in ways that are in line with those emotions (Griskevicius, Shiota, & Neufeld, 2010). In this regard, an elicited fear emotion can trigger the motivation for self-protection. Conversely, love or romantic desire can trigger a mate attraction motivation (Griskevicius et al., 2010). Thus, the attitudes towards ads are context-dependent on the preceding elicited emotions.

(13)

13 Gap and Research Question

By relying on the evolutionary theory, it was previously explained how emotions can motivate people to act and think in ways that are in line with those states. As a result, some scholars agreed that attitudes towards ads are context-dependent; more specifically are influenced by the preceding elicited emotions (e.g. Griskevicius et al., 2009; Kyrousi et al., 2016). The paper by Kyrousi et al. (2016) was the first to capitalize on evolutionary psychology for explaining women consumer’s attitudes towards decorative role portrayal in advertising. Nevertheless, no study investigated the impact of emotions such as fear and romantic desire on women responses to ads that use non-traditional (i.e. career-oriented) versus decorative role portrayal.

That is to say, little research is available on the response to ads that use stereotypic women role portrayal, taking into consideration the medium context that the ads are placed on television, print and online medium (Kyrousi et al., 2016). According to (Griskevicius et al., 2009), the medium context is arousing specific emotions that influence the advertising effectiveness. In this way, the effectiveness can be improved by placing the advertisement in a more suitable media context (De Pelsmacker, Geuens, & Anckaert, 2002).

The existing literature on attitudinal reactions towards ads that depict women in various roles have reported findings that are contradictory (Kyrousi et al., 2016, p. 773). This paper seeks to address this disagreement by capitalizing on the fact that women responses to stereotypic role portrayal (i.e. decorative and non-traditional career-oriented) are context dependent, and the evolutionary theory can explain them.

Based on the aforementioned gap in the literature, this study aims to investigate how the effect of emotions aroused by a preceding context on advertising effectiveness (i.e. attitude towards the ad

(14)

14 and the brand) depends on the way women are featured in advertisements. Respectively, the research question that will be addressed is:

“How do specific emotions such as fear and romantic desire impact the attitude towards

the ad and brand that portray women in a decorative or career-oriented fashion?”

More specifically, this study aims to investigate how romantic desire and fear as emotions elicited by a preceding context affect the female consumers’ attitudinal responses towards the ad and the brand that portray women in a decorative versus non-traditional (i.e. career oriented) way. For this purpose, this study will control for variables such as age of the respondents, advertised product category and its “perceived gender”, gender role ideologies and relationship status of women participants.

(15)

15 Hypotheses

As it was mentioned in the literature review, according to Keltner et al. (cited in Griskevicius et al., 2009, p. 386), emotion states should motivate individuals to act and think in ways that are in line with those emotions (Griskevicius et al., 2010).

In this regard, by capitalizing on the evolutionary approach, an elicited fear emotion triggers the motivation for protection. Thus, when women are in a state of fear, they seek for safety, self-protection or something that would give them the feeling that they are in control of the situation. Respectively, an advertisement that portrays female weak (i.e. decorative way) is expected to lead less favorable attitudes than in the case of advertisements that depict women stronger and more independent. To put it differently, when women audience is in a state of fear, relying on the evolutionary theory, the advertisement that portrays the model ambitioned and independent (i.e. career-oriented) should give them a feeling of safety and lead to more favorable attitudes.

In summary,

Hypothesis 1(a, b): Fear will lead, in the case of women, to more favorable attitudes towards the advertisement (H1a) and the brand (H1b) when non-traditional (i.e. career oriented) women role portrayal than stereotypic (i.e. decorative) women role portrayal in advertising is used.

By using evolutionary psychology, romantic desire triggers the motivation for mate attraction (Griskevicius et al., 2009). Respectively, when women are in a state of romantic desire and are motivated to attract mates, an advertisement that portrays the female model attractive (i.e. decorative way) is expected to be more effective, considering that romantic desire should support characteristics that are perceived important in the dating market (Griskevicius et al., 2009). Thus,

(16)

16 the sex appeal generally used in ads that portray women in a decorative way should lead to higher advertising effectiveness when the romantic desire was primed. However, the advertisement that portrays women ambitioned and independent (i.e. career-oriented) signals characteristics that are not desired in the dating market (Bursztyn, Fujiwara, & Pallais, 2016). Therefore, this type of role portrayal is expected to be counter-effective for women respondents that are in a state of romantic desire.

Hypothesis 2(a, b): Romantic desire will lead, in the case of women, to more favorable attitudes towards the advertisement (H2a) and the brand (H2b) when stereotypic (i.e. decorative) women portrayal in advertising compared to non-traditional (i.e. career oriented) women role portrayal is used.

Overview of the main Hypotheses:

H1a Fear will lead to more favorable attitudes towards the advertisement when non-traditional (i.e. career oriented) than decorative women role portrayal is used in the advertisement.

H1b Fear will lead to more favorable attitudes towards the brand when non-traditional (i.e. career oriented) instead of decorative women role portrayal is used in the advertisement.

H2a Romantic desire will lead to more favorable attitudes towards the when stereotypic (i.e. decorative) women portrayal in advertising compared to non-traditional (i.e. career oriented) women role portrayal is used.

H2b Romantic desire will lead to more favorable attitudes towards the brand when stereotypic (i.e. decorative) than non-traditional (i.e. career oriented) women role portrayal is used in the advertisement.

(17)

17 Conceptual Model

The research question can be visualized in the conceptual model below.

H1a, H1b H2a, H2b

Explanation of the relationships in the model:

The conceptual model illustrates that the effect of emotions triggered by a preceding context (i.e. fear and romantic desire) on attitudes towards the ad and the brand depends on the way women are portrayed in the advertising (i.e. decorative or career-oriented). The women role portrayal used, namely decorative and career-oriented role portrayal, represents both an independent variable and a moderator, because this study will investigate how the effect of emotions on the ad effectiveness (i.e. attitude towards the ad and the brand) depends on the way women are featured in the advertisement. Moderator Women1Portrayal: - Decorative - Career-oriented Dependent variables:

• Attitude toward the ad • Attitude toward the

brand Independent variables Emotions: -Fear -Love Control variables: • Age • Relationship status

• Gender role ideologies (i.e. traditional, transitional, egalitarian)

(18)

18 Expected contributions

Results of the study are expected to have important theoretical implications and practical ones, especially in the area of advertisement execution strategies and message targeting.

Prior research has provided diverging evidence regarding women attitudes to ads that use stereotypic female portrayal (see Zawisza & Cinnirella, 2010, p. 1769). And only recently, a scant amount of literature started to use certain evolved mechanisms to explain the responses to ads (Kyrousi et al., 2016; Saad, 2004). However, no study yet undertook the direction of comparatively analyzing women’ responses to both decorative and non-traditional (i.e. career-oriented) portrayal in advertising, by capitalizing on the evolutionary psychology principles.

The response to this research gap will provide evidence on how an emotionally-arousing content impacts the advertising effectiveness. More exactly, it will cast light on how the emotionally inducing content (e.g., videos or articles that consumers previously read or saw) influences the reactions of female consumers towards the ad which depicts women weak (i.e. decorative) versus strong and independent (i.e. career-oriented). In this way, this study will contribute to the existing literature which is contradictory when it comes to females’ attitudes towards advertising that make use of various female role portrayals.

From a managerial standpoint, the findings of this study are expected to emphasize the need for advertisers to examine more carefully the medium context in which ads showing women in decorative and non-traditional (i.e. career-oriented) way are placed. In this respect, advertisers can get insights on which emotions (i.e. fear or romantic desire) should be stimulated so that the investment made will yield to higher advertising effectiveness.

(19)

19

METHODOLOGY

Experimental design

Decorative Career-oriented

Fear Fear – Decorative portrayal in ad Fear – career-oriented woman in ad

Love Love – Decorative portrayal in ad Love – career-oriented woman in ad

For the purpose of identifying the relationship between variables priorly introduced, I commenced with an experimental approach. By means of independent variable manipulation, I aimed to provide evidence why certain ads that use stereotypic women portrayal are being more favorably perceived by female audiences than others. The experiment can be regarded as a cross-sectional study, namely because it studied a particular phenomenon at a specific moment in time.

In order to answer the research question, experiments were run using a between-subjects 2 (emotion: fear, love) X 2 (women portrayal in advertising: decorative, career-oriented) experiment design.

Subjects

As a source, female university students aged between 20-26 years were used as respondents. As it will be further explained in the Variables section, a number of existing studies pointed out that

Women Portrayal: - Decorative - Career-oriented

• Attitude toward the ad • Attitude toward the

brand • Emotions:

-Fear -Love

(20)

20 gender of the respondents play an important role in forming attitudes towards the stereotypic advertisements (e.g. Theodoridis et al., 2013, p. 152). Respectively, as the current study assessed only women’s attitude towards the stereotypic ads and the brands, the gender variable is left out. What is more, a student sample was preferred in the endeavor to keep respondent’s age similar since the age of participants has been found to influence attitudes towards stereotypic advertisements (Theodoridis et al., 2013).

For each block of questions, a minimum of 40 respondents participated. To reach to female students who are typically pursuing their masters at University of Amsterdam, UvA email system was used. In addition, master students’ groups on social media were used in order to target a broader audience of students.

Procedure

For this study, conducting online experiments using Qualitrics was chosen. The reason behind this is the fact that with the digital age, ads are placed after video content (e.g. YouTube) or as banner ads on different websites. So, by conducting online experiment with Qualitrics, the online medium that female respondents often use for watching a video or reading an article was preserved. What is more, the design of laboratory experiments is often characterized as artificial. This has become a concern for the external validity of the results, according to Peler Harris work on designing and reporting experiments (Kyrousi et al., 2016, p. 773). Since priming emotions in an online setting where the researcher does not have a perfect control over the participant’s environment is a challenging task, this study included a pretest and manipulation checks that assess whether respondents are primed, as intended, with romantic desire and fear.

(21)

21 In the pretest, participants were shown two advertisements and two video clips. In other words, all participants were assigned the same online survey where they had to rate two movie clips on love

and romantic desire, and fear and motivation to protect themselves.Simultaneously, the female

students that took part in the pretest had to evaluate the two advertisements on the usage of decorative and career-oriented women portrayal. The pretest questionnaire was designed in a way that all respondents initially rated the chosen advertisements. By doing so, it was aimed that emotions triggered by the video clips would not influence the evaluation of the ads. The operationalization of the variables in the pretest is comprehensively explained in the

Operationalization of variables section of this paper.

The pretest provided enough evidence that the two videos managed to trigger romantic desire and fear; and, ads were perceived by the respondents as “showing the female model in a stereotypic fashion” as “decorative objects” and “depicting the women models in a non-traditional manner” as “career-oriented”. All these things considered, the decision of keeping the same emotion-arousing content in the further experiment was made. Subsequently, the main experiment was conducted. In the main experiment, however, subjects were randomly assigned to one of the following four conditions. A quarter of the students had to watch a short video clip designed to trigger romantic desire followed by an ad that shows a woman in a decorative fashion, whereas another quarter watched the same video clip but followed by an ad which depicts four career-oriented women. Alternately, one-fourth watched a video designed to trigger fear followed by an ad that uses decorative women portrayal, while the remaining part watched the same video clip followed by an ad that uses career-oriented women role portrayal.

As mentioned previously, data was collected online. All items used in the experiment are based upon previously designed and validated scientific tests to ensure internal validity. All items were

(22)

22 scored on a 7-point Likert scale, which was treated as an interval scale. Treating Likert scales as interval is a much discussed point, but it is common practice to use Likert scales as interval scale

and both non- and parametric tests can provide robust answers(Norman, 2010).

Operationalization of variables Dependent variable

Dependent variables used in the research are: attitudes toward the ad and attitudes toward the

brand (i.e. advertising effectiveness). A four-item, seven-point scale regarding (i.e. Good/Bad, Like/Dislike, Irritating/Not Irritating, and Interesting/Uninteresting) was used to test respondent’s attitudes toward the advertisement, as it was recommended by Mitchell and Olson 1981 (Kyrousi et al., 2016, p. 773). For the second variable, a five-item with seven-point scale regarding (i.e.

Unappealing/appealing, Bad/Good, Unpleasant/Pleasant, Unfavorable/Favorable, and

Unlikable/Likable) was used to test respondent’s attitudes toward the brand (Kyrousi et al., 2016, p. 773).

Based on pretest data, the reliability analysis showed (α = .890) for the fear triggering video and (α

= .967) for the romantic desire triggering video. Respectively, both dependent variables are

registering a high reliability score well above the threshold of 0.7 (Field, 2013). However, an additional manipulation check was included in the actual experiment which resulted in similarly high reliability scores.

Independent variable

Independent variables in the experiment are: fear and romantic desire as emotions.

Based on previous research (Griskevicius et al., 2009), the operationalization of the aforementioned variables was done by making the respondents view a clip with romantic scenes

(23)

23 from “The sunrise” movie which depicts how an attractive man and woman fall in love as they both travel through Europe, and “The shining” movie which showed a mad man chasing a woman with an axe and attempting to break the locked door in which woman was hiding. In this way, fear and romantic desire were primed. For assessing the effectiveness of manipulation, a seven-point scale was used as follows: Respondents were asked in a pretest (n = 31) to indicate the extent to which they felt (1a) fear, (1b) motivation to protect themselves, (2a) romantic desire, and (2b) motivation to attract a romantic partner.

I consider that the operationalization of the stimuli is representative of the conceptual constructs, as the movie clips explicitly are selected to trigger the motivation of respondents to protect themselves and to attract a partner – elements explained by the evolutionary theory as triggered by fear and romantic desire (Griskevicius et al., 2009).

In addition to the pretest, a short manipulation check was included in the main experiment (M=5.62, SD=.985) for the romantic desire triggering video and (M=5.79, SD=1.249) for the fear priming video. The reliability analysis that was run based on the manipulation check data showed (α = .819) for the fear triggering video and (α = .725) for the romantic desire triggering video (see Table 1). Respectively, both being above the threshold of 0.7 (Field, 2013). An overview of the variables, description, items and reliability score is displayed in Table 1.

Moderator variable – Ad pretesting

Moderators: Women Portrayal in decorative or career-oriented fashion.

To ascertain whether respondents have a clear understanding of the differences among the two versions of the advertisements, a group of 31 women students participated in a pretest where they had to rate the advertisements. It is important to note that the subjects for the pretest were

(24)

24 selected to be representative of the subjects that eventually took part in the final experiment. To wit, 31 female students, aged between 20 and 26, took part in the pretest.

The operationalization of the two moderating variables was done with a pretest by asking respondents to rate the ads on two statements. Firstly, female students were asked the extent to which they believed that the ad shows the female model in a stereotypic fashion and the degree to which they thought that the advertisement depicts the model as a decorative object (see Figure 2). Secondly, participants were asked to which extent they believed that the ad depicts women in a non-traditional manner and that the women models are shown as career-oriented (see Figure 3). For the first ad (Figure 2), as recommended in (Kyrousi et al., 2016, p. 773), a two-item with seven-point scale varying from “strongly agree” to “strongly disagree” was used to test whether respondents are perceiving the advertisement as stereotypical (α = .892). As mentioned previously, the respondents were asked the extent to which they think that the advertisement depicts woman in a stereotyped manner and the extent to which they believe that the ad shows woman as a decorative object.

(25)

25 The ad that uses decorative female portrayal (Figure 2) was also used in the (Kyrousi et al., 2016) research paper and comprises a female model lying on a couch with a fictitious brand name for a web radio (i.e. Tempo), and “The new web radio” tagline. This ad was selected by the aforementioned researchers on the basis of a pretest (n = 22). The pretest conducted by the authors also found that this specific ad was correctly perceived by the respondents (α = .070) (Kyrousi et al., 2016, p. 773).

For the second ad (Figure 3), similarly, a two-item with seven-point scale varying from “strongly agree” to “strongly disagree” was used to test whether respondents are perceiving the advertisements as non-traditional women portrayal (Kyrousi et al., 2016, p. 773). As previously noted, the respondents were asked the extent to which they think that the advertisement depicts women in a non-traditional fashion and the extent they believe that the ad shows woman as career-oriented (α = .892).

The ad using non-traditional women portrayal (Figure 3) comprises of four female model which are in their offices/work place with a fictitious brand name for a consulting firm (i.e. ABC Consulting), and “Be bold and make a change. Join ABC Consulting!” tagline. This ad was selected on the basis of the pretest (n = 31) that was conducted before the main experiment.

The criteria for this specific advertisement was make it easy for respondents to notice that the women are portrayed as career oriented. For this reason, pictures of women in their offices were chosen. Moreover, some additional cues as the professional posture were important in choosing the pictures. Another criteria was to use an industry that is considered by generally regarded by students as “gender neutral”.

(26)

26 In accordance with the literature review, it was important to control for the product/industry type (Kyrousi et al., 2016, p. 781). Therefore, as most of the respondents were female students at the Master in Business Administration programme, it was assumed that they perceive management consulting industry as “gender neutral”.

Nevertheless, it was crucial to assess the ‘neutrality’ with a pretest, where the results were found to be in accordance with these assumptions.

Figure 3.

Control variables

In order to capture the impact of emotions on attitudinal responses to ads that use stereotypic role portray, it is essential to control or account for factors that can influence this relationship.

Previous research has found that attitudes towards stereotypic ads differs by the gender role ideologies (i.e. traditional, transitional, egalitarian) that consumers have (Baxter et al., 2016), by consumers’ age and gender (Theodoridis et al., 2013) or by the product category being advertised and its ‘perceived gender’, according to Debevec and Iyer (cited in Kyrousi et al., 2016, p. 774). As mentioned, some of the existing studies pointed out that respondent’s gender play an important role in forming attitudes towards the stereotypic advertisements (e.g. Theodoridis et al.,

(27)

27 2013, p. 152). However, as the current study assessed only women’s attitude towards the stereotypic ads and the brands, the gender variable is left out.

Age:

What is more, it has been also found that respondent’s age account for differences in responses to stereotypic ads (Theodoridis et al., 2013, p. 152). For controlling this variable, the target group of respondents represented female students, therefore in most of the cases respondents are in the same age category of 18-24.

Product type and its perceived gender:

As it was aforementioned, the attitudes towards stereotypic ads was found to differ by the advertised product type and the ‘gender’ that people link to the product (Whipple & Courtney as cited in Kyrousi et al., 2016). It is important to control for this variable as, according to Whipple &

Courtney (as cited in Tsiataki, 2014),people tend to prefer the sex of the model in advertising to be

congruent with the advertised product’s gender-image. For controlling it, this study used a neutral product category as in the previous literature (Kyrousi et al., 2016, p. 781) and assessed the ‘neutrality’ with a pretest.

Relationship status:

A recent paper reported that single women avoid actions that signal career ambition because these are undesired in the dating or even marriage market (Bursztyn et al., 2016). In this regard, the relationship status becomes of interest in explaining the potential differences in women respondents’ attitudes towards ads that use female role portrayal (e.g. decorative, non-traditional).

(28)

28 Generally, ads that portray women in a stereotyped ‘non-traditional’ manner (i.e. career oriented) are emphasizing strong women characteristics such as ambition or assertiveness. At the same time, it is possible that respondents in a romantic desire state that are single, accounting on evolutionary psychology and mating process, would have more favorable attitudes towards the ad and brands that feature women characteristics as ‘desired’ in the dating market (i.e. ad that uses decorative portrayal). By contrast, the advertising that portrays women ambitioned and independent (i.e. career-oriented) could to be counter-effective for single female respondents that are in a romantic desire state, as it signals characteristics that are not desired in the dating market.

Respectively, because romantic desire can trigger mate attraction motivation (Griskevicius et al., 2010), single respondents are expected to respond differently to ad portraying a career oriented woman than respondents that are in a relationship. Hence, for controlling purposes, this variable is included in the data analysis.

Gender role ideologies (i.e. traditional, transitional, and egalitarian):

Also, Baxter et al. (2016) found out that respondents’ ideologies on gender role (i.e. traditional, transitional, and egalitarian) have a moderating role on their attitudes towards stereotypic ads. Even though, according to Bem (as cited in Tsiataki, 2014), societal norms can prescribe specific roles or characteristics as being more appropriate to men or women; people differ in the extent to which they do or not conform to these norms (Baxter et al., 2016). Some people, despite their cultural background, may have certain ideologies that might be in line or differ substantially from the societal norms. Therefore this study elicited these ideologies (i.e. traditional, transitional, and egalitarian) to be able to control them in the data analysis.

(29)

29 Op er at ion aliz at io n /M ea su re Ta b le 1. O p er at ionaliz at ion o f v aria b les in th e St u d y V aria b le/ con st ru ct A tt itu d e tow ar d s th e ad A tt itu d e tow ar d s th e b ran d Emo tio n s: lov e an d fea r D ec or at iv e/ C ar ee r-or ien ted w om en p or tra yal in th e ad ver tise men t Fou r-it ems w it h sev en -p oin t sema n tic d if fe ren tial sc ale (Mische l & Ols on , 1981) Fiv e-it ems w it h sev en -p o in t (Ky ro u si et al., 20 1 6, p . 773) . Tw o -it ems for each emo tio n , sev en -p oi n t sc ale (G risk evicius et al., 2 0 09 ) tes ted in th e p re tes t. Tw o -it ems for each p or tra yal, sev en -p o in t Lik er t sc ale tes ted in th e p re tes t • Bad /Go o d * • D isl ike /L ike * • Ir ritatin g/ N o t irr ita tin g* • Un in te rest in g/ In te re sting * • Un ap p ealin g/ ap p ealin g • Bad /Go o d • Un p leasan t/ P leas an t • Un fav o rab le/ Fa vo rab le • Un likabl e/ Likab le • The ext ent t o w h ich y o u fe lt ro m an tic desire • To wh ich ex tent d id yo u feel m o tiva ted to attr act a ro m an tic p artn er? • The ext ent t o w h ich y o u fe lt fear* • To which ext ent d id yo u fee l m o ti vate d to p ro te ct yo u rs elf? • The ad vertis em ent s h o ws wome n in a st ere o ty p ic fas h io n • The ad vertis em ent depi cts wome n a s dec o rati ve o b jec ts • Th e ad verti se m en t d ep ict s w o m en in a n o n -trad itiona l m an n er. • The ad vertis em ent s h o ws wome n a s c ar ee r-o rien ted 0.9 05 0.9 64 0.7 25 0.8 19 0.8 92 D escrip tio n It ems R elia b ilit y (C ron b ach ’s α ) *R ec od ed

(30)

30 DATA ANALYSIS AND RESULTS

The data analysis of the main experiment started with a look over the descriptive statistics and output of the reliability check. The results of the descriptive statistics and the reliability analysis can be viewed in Table 2.

Table 2

Descriptive statistics and correlations among study variables

*, indicates that correlation is significant at the .05 level **, indicates that correlation is significant at the .01 level Numbers in parenthesis represent Cronbach Alphas.

Table 2 exhibits the name of variables used in this study, the means, standard deviations, reliabilities (Cronbach alphas) and the correlation between variables.

The attitude towards the ad and the brand scale has high reliability, with a Cronbach’s Alpha = .905 and, respectively, .964. In Table 3, the corrected item-total correlations indicate that all the items of the aforementioned dependent variables have a good correlation with the total score of the scale (all above .30). In addition, none of the items would considerably affect reliability of the data if they were eliminated.

The correlation analysis was run in order to quantify the intensity of the relationship between the variables used in this study. No significant correlation was found, except the one between the two dependent variables, to wit: attitude towards the ad and attitude towards the brand that uses women portrayal in advertisement (α = .822, p = .01). The reason for this result is that the simple

Variable Mean S.D. 1 2 3 4 5 6

1.Emotions

2.Women Portrayal 3.Attitude towards the Ad 4.Attitude towards the Brand 5.Relationship Status

6.Gender Role ideology

.5086 .5013 .5143 .5012 .005 4.1486 1.4513 .024 -.019 (.905) 4.0969 1.4811 .044 -.066 .822** (.964) 1.47 .500 -.085 .156* -.110 -.141 (.944) 2.77 .500 -.105 -.112 -.168* -.096 .086 (.944)

(31)

31 effects were found to be in different direction which can be seen in the Moderation Analysis section of this paper. This results in a strong evidence of an interaction (F(1, 168) = 10. 6161, p= .001) and, therefore, no evidence for a main effect of the independent variable on the dependent one. In addition, we see a negative and significant correlation between gender role ideology and attitude towards the ad (p = 0.05).

Table 3 Reliability statistics Scale Mean if Item Deleted Scale Variance if Item Deleted Corrected Item-Total Correlation Cronbach's Alpha if Item Deleted How would you describe your overall

opinion of the advertisement? - Bad:Good*

12.2800 19.145 .873 .845

How would you describe your overall opinion of the advertisement? - Dislike:Like*

12.4571 20.077 .828 .863

How would you describe your overall opinion of the advertisement? - Irritating:Not Irritating*

12.0914 21.762 .642 .927

How would you describe your overall opinion of the advertisement? - Uninteresting:Interesting*

12.7314 19.588 .813 .868

How would you describe your overall attitude towards the brand that placed the online advartisement? - Unappealing:Appealing

16.49 35.918 .824 .967

How would you describe your overall attitude towards the brand that placed the online advartisement? - Bad:Good

16.25 36.566 .891 .956

How would you describe your overall attitude towards the brand that placed the online advartisement? - Unpleasant:Pleasant

16.39 35.470 .912 .953

How would you describe your overall attitude towards the brand that placed the online advartisement? - Unfavorable:Favorable

16.40 34.506 .941 .948

How would you describe your overall attitude towards the brand that placed the online advartisement? - Unlikable:Likable

16.44 34.133 .921 .951

(32)

32 Factor analysis

Furthermore, the factor analysis was run in order to investigate the similarities between the two dependent variables. A factor is defined as a dimension which represents the concise statement of the correlations between a set of variables (Kline, 2014). The factor analysis will help us to evaluate how good the two scales that we used are, regardless of the fact that they were previously validated by other researchers.

In particular, as recommended by Mitchell and Olson 1981, a four-item, seven-point scale (i.e. Good/Bad, Like/Dislike, Irritating/Not Irritating, and Interesting/Uninteresting) was used to test respondent’s attitudes toward the advertisement (Kyrousi et al., 2016, p. 773).

For the second dependent variable, a five-item with seven-point scale regarding (i.e. Unappealing/appealing, Bad/Good, Unpleasant/Pleasant, Unfavorable/Favorable, and Unlikable/ Likable) was used to test respondent’s attitudes toward the brand (Kyrousi et al., 2016, p. 773). A principal axis factoring analysis (PAF) was conducted on the scales. The Kaiser–Meyer–Olkin measure verified the sampling adequacy for the analysis, KMO = .921. Bartlett’s test of sphericity χ² (36) = 1844.989, p < .000, indicated that the correlations between items were acceptably large for PAF.

A primal analysis was run to in order to obtain eigenvalues for each item in the data. Two distinct constructs had eigenvalue over Kaiser’s criterion of 1, and together it explained 78.46 % of the variance. The first component explained 68.23 % of variance, followed by another construct that explained 10.24 % of the variance. The items clustered on the same factors imply that factor 1 represents Attitude towards the brand, and factor 2 represents Attitude towards the ad.

(33)

33 What is more, no items of one factor showed high cross-loadings on the other factor. In other words, all the principal loadings were at least twice as big as the secondary loadings.

Table 4

Item

Rotated Factor Loadings

𝐴_{𝑏𝑟𝑎𝑛𝑑} 𝐴_𝑎𝑑

How would you describe your overall opinion of the advertisement? - Bad:Good*

-.025 .963

How would you describe your overall opinion of the advertisement? - Dislike:Like*

-.073 .977

How would you describe your overall opinion of the advertisement? - Irritating:Not Irritating*

.071 .711

How would you describe your overall opinion of the advertisement? - Uninteresting:Interesting*

.055 .790

How would you describe your overall attitude towards the brand

that placed the online advartisement? - Unappealing:Appealing .786 .115

that placed the online advartisement? - Bad:Good .831 .122

that placed the online advartisement? - Unpleasant:Pleasant .969 -.028

that placed the online advartisement? - Unfavorable:Favorable 1.030 -.078

that placed the online advartisement? - Unlikable:Likable .958 -.006

Eigenvalues 6.823 1.024

% of Variance 68.823 10.240

Note: Factor loadings over .40 appear in bold

*Recoded

In agreement with Kaiser's criterion, examination of the scree plot revealed a levelling off after the second factor. Thus, the second factor was retained and rotated with an Oblimin with Kaiser Normalization rotation. Table 4 exhibits the factor loadings results after rotation.

In final consideration, the component correlation Matrix shows that there is a high positive relation between the factors (.794).

(34)

34 Moderation Analysis

For further testing the four hypotheses, moderation analysis was run with Process macro written by Andrew F. Hayes for SPSS. Moderation analysis is used as we want to test whether the effect of emotions on the outcome variables (Hayes, 2012), namely effectiveness of advertisements that use women portrayal (i.e. attitude towards the ad and towards the brand), depends on the way women are portrayed in ads.

To conduct the analysis, Process macro was used as in the current model both the independent and the moderating variables represent dummy variables (i.e. 0 and 1). Respectively, in order to run the Moderation Analysis and be able to calculate the interaction between the variables automatically, Process macro is the most appropriate method to do so.

In this case, the simple moderation analysis helped in our aim to determine whether the effect of the independent variable (i.e. emotions: romantic desire, fear) on attitude towards the ad/brand is influenced by the third variable (i.e. women portrayal).

The statistical and conceptual Process model in the case of simple moderation can be seen in Figure 4, where Women portrayal stands for the moderator, Emotions for the independent variable, Attitude towards Ad/brand - dependent variable, and Emotions x Women Portrayal the interaction between independent variable and moderator. The arrows stand for “predictor of”, which means that if an arrow points from one variable to another, then the variable from where the arrow originates is a predictor in the statistical model of the final variable (where the arrow ends) (Hayes, 2012).

(35)

35 Figure 4. PROCESS Model

Conceptual model

Statistical Model

Source: (Hayes, 2012)

The analysis conducted with Process revealed an existing interaction between emotions triggered by the preceding context and women portrayal in advertisements (F(1, 169)=14.3313, p = .000). In order to test the specific hypothesis that were formulated based on the literature review, main effects, interaction effects and simple main effects are analyzed.

In the descriptive statistics section, it can be seen that no significant correlation was found between the independent variable and dependent variables. As it was stated previously, the reason for the result obtained is that the simple effects were found to be in different direction resulting in a strong evidence of an interaction (F (1, 169) = 14.3313, p = .000) and, therefore, no evidence for a main effect. The result can be seen in Table 5 and visualized in Figure 4.

Emotions Attitude _towards

Ad/brand Emotions Emotions x Women Portrayal Women Portrayal c1 c2 c3 Attitude towards Ad/brand Women Portrayal

(36)

36 Table 5. PROCESS Procedure for SPSS Release 2.16.3

Model = 1 Y = AadTOT X = ForLove M = WPort Statistical Controls:

CONTROL= Con_RelS Con_RId Sample size 175 Outcome: AadTOT Model Summary R R-sq MSE F df1 df2 p .3160 .0998 1.9522 4.3410 5.0000 169.0000 .0010 Model

coeff se t p LLCI ULCI constant 5.2594 .5993 8.7757 .0000 4.0763 6.4426 WPort .7429 .2972 2.4992 .0134 .1561 1.3297 ForLove .8339 .2882 2.8933 .0043 .2649 1.4029 int_1 -1.6132 .4261 -3.7857 .0002 -2.4545 -.7720 Con_RelS -.2781 .2185 -1.2733 .2047 -.7094 .1531 Con_RId -.3913 .1827 -2.1182 .0356 -.7560 -.0266 Product terms key:

int_1 ForLove X WPort

R-square increase due to interaction(s):

R2-chng F df1 df2 p int_1 .0767 14.3313 1.0000 169.0000 .0002 Conditional effect of X on Y at values of the moderator(s):

WPort Effect se t p LLCI ULCI ,0000 .8339 .2882 2.8933 .0043 .2649 1.4029 1,0000 -.7793 .3130 -2.4898 .0137 -1.3972 -.1614 * Estimates are based on setting covariates to their sample means.

Level of confidence for all confidence intervals in output = 95, 00

There is a significant interaction effect found (F (1, 169) = 14.3313, p = .000). However, the model can only explain 9, 98 % of the variance in the dependent variable (Rsq = .0998).

According to the moderation analysis output, when women are featured in advertisement in a career-oriented fashion (WPort = 0), there is a positive and significant effect seen of emotions as independent variables on attitudes towards the ad (p = .004) and towards the brand (p = .014). By

(37)

37 comparison, when women are depicted in a decorative way in advertisements (WPort = 1), there is a negative and significant effect of emotions on attitudes towards the ad (p = .0137) and a negative, but not significant, effect of emotions on attitudes towards the brand (p = .0810).

Simple Effects Analysis

In order to further test the specific hypothesis of this study, a series of tests for simple main effects was performed. In Table 6 it can be seen an overview of the results obtained in the case of attitude

toward the ad and attitude toward the brand dependent variables.

Consistent with H1a, when fear was elicited, the attitude toward the ad that used career-oriented women portrayal was more favorable than the one that used decorative female portrayal (M career-oriented = 4.59, M decorative = 3.82). Likewise, consistent with H1b, the attitude towards the brand that uses career-oriented women portrayal in their ads was more favorable that towards the brand that uses decorative female portrayal (M career-oriented = 4.57, M decorative = 3.81).

Consistent with H2a, the attitude towards the ad were more favorable when decorative female

portrayal was used in the advertisement than career-oriented portrayal (M decorative = 4.72, M

career-oriented = 3.55) when romantic desire was primed. Consistent with H2b, the attitude towards the brands that were advertised in the ad was more favorable when ad portrayed the women in a decorative fashion than in a career-oriented one (M decorative = 4.48, M career-oriented = 3.59). However, the findings (see Appendix 3) reveal a negative, but not significant, effect of emotions on attitudes towards the brand (p = .0810). Therefore, hypothesis H2b, according to which romantic desire will lead to more favorable attitude toward the brand when decorative than non-traditional (i.e. career-oriented) women role portrayal is used in the advertisement is rejected.

(38)

38 Overall, as H1a and H1b, as well as H2a, were supported on the basis that a significant interaction was found by conducting moderation analysis which Process Model 1 reported automatically. What is more, the means shows a significant difference as explained previously, hence providing support for the proposed set of hypothesis. However, H2b was rejected on the basis of an insignificant interaction found by Process Model 1.

Table 6 displays the variables used in this study and the means and standard deviations for each of the 4 combinations derived from the 2 (emotion: fear, love) X 2 (women portrayal in advertising: decorative, career-oriented) experiment design.

As an illustration of the results obtained after running the moderation analysis with Process, a plot of the significant interaction effect was created by using the regression coefficients. In Figure 4 it can be seen a crossover interaction. In this case, there are no main effects because of the fact that effect of the independent variables (i.e. Romantic Desire and Fear) on the dependent variable is opposite for each of the two levels of the moderator (i.e. Decorative, Career-oriented women portrayal).

Table 6 Effectiveness of advertisement as a function of emotion and women portrayal

Career Oriented Decorative

Type of women portrayal used in advertisement

4.72 (1.21) 4.48 (1.42) 3.83 (1.45) 3.81 (1.42)

Notes: Means and standard deviations are in parentheses.

First raw presents the results for Attitude towards the Ad and second raw for Attitude towards the Brand. Romantic Desire 3.55 (1.09)

3.59 (1.21)

Fear 4.59 (1.47)

4.57 (1.44)

(39)

39 Figure 4 Effects of Stereotypic Women portrayal and Type of emotions on attitude towards the Ad

Figure 5 Effects of Stereotypic Women portrayal and Type of emotions on attitude towards the Brand 1 2 3 4 5 6 7 8

Romantic Desire Fear

At tutid es tow ar d the Ad Career-oriented Women Portrayal Decorative women potrayal 1 2 3 4 5 6 7 8

Romantic Desire Fear

A tt utid es t ow ar d the Br and Career-oriented Women Portrayal Decorative women potrayal

(40)

40 Overview of the main Hypotheses:

H1a – Supported: Fear, in the case of women, will lead to more favorable attitudes towards the advertisement when non-traditional (i.e. career oriented) than decorative women role portrayal is used in the advertisement.

H1b – Supported: Fear, in the case of women, will lead to more favorable attitudes towards the brand when non-traditional (i.e. career oriented) instead of decorative women role portrayal is used in the advertisement.

H2a – Supported: Romantic desire, in the case of women, will lead to more favorable attitudes towards the when stereotypic (i.e. decorative) women portrayal in advertising compared to non-traditional (i.e. career oriented) women role portrayal is used.

H2b – Rejected: Romantic desire, in the case of women, will lead to more favorable attitudes towards the brand when stereotypic (i.e. decorative) than non-traditional (i.e. career oriented) women role portrayal is used in the advertisement.

(41)

41 GENERAL DISCUSSION, LIMITATIONS AND IMPLICATIONS

Researchers for a long time have been interested in the topic of female role portrayals in advertising; however, only a limited number of studies investigated how women actually respond to women depictions. The ones that attempted to do so, as described in the literature review section, came up with diverging evidence.

As a result, this research paper commenced with the following question: How might distinct affect-arousing contexts impact women responses to ads that use female role portrayal? The approach to answer the question was to examine how a peculiar positive affective state (i.e. romantic desire) and a particular negative affective state (i.e. fear) impact the effectiveness of advertising that uses one of the two female role portrayals: decorative and career-oriented.

For this purpose, the evolutionary theory was used to make specific predictions about how the medium context influences women attitudes concerning advertisements with female role portrayals. In this case, the type of female role portrayal used in ads, namely as decorative objects or career-oriented ones, moderates this effect.

Overall, results of the experiment confirmed the proposed hypotheses (i.e. H1a, H1b, H2a) which were based on evolutionary theory principles. As the evolutionary model suggests, different emotions can motivate people to act and think in certain ways that are in line with those affective states (Griskevicius et al., 2009). Consistent with this perspective, romantic desire and fear effect on the effectiveness of advertising that uses female role portrayal must be different because these particular emotions would motivate people to think differently, thus react differently to the same advertisements. Respectively, this would play a role in explaining the diverging results that prior studies found regarding with women reactions to ads that use stereotypic female role portrayal,

(42)

42 such as Jaffe and Berger’s, and Orth and Holancova’s studies (as cited in Zawisza & Cinnirella, 2010).

Taking the case of fear, this negative emotion causes the decorative female portrayal to backfire when used in an advertisement. Per contra, the same stereotypic women portrayal was found more effective following a romantic content.

As previously discussed, results of the experiment can be explained by capitalizing on the evolutionary model. For instance, when fear and motivation for self-protection is triggered, people seek for safety or something that would give them the feeling that they are in control of the situation (e.g. Griskevicius et al., 2009; Kyrousi et al., 2016). Respectively, the findings demonstrate that in a state of fear, an advertisement that portrays women as decorative objects, thus weak, was counter effective. On the other hand, when fear was primed, the advertisement that portrays women ambitioned and independent (i.e. career-oriented) was perceived more favorably than in the case of decorative role portrayal.

Equally important, when romantic desire and the motivation for mate attraction are induced, people seek to have characteristics that are desired in the dating or marriage market (i.e. salient differentiation, physical attractiveness) (Bursztyn et al., 2016; Griskevicius et al., 2009). Consistent with the evolutionary theory, the results of this study demonstrate that when women are in a state of romantic desire, the advertisement that portray female model in a decorative fashion lead to more favorable attitudes toward the ad than in the case of career-oriented women portrayal. Thus, when a woman was depicted in the ad as a decorative object, the primed romantic desire lead to higher advertising effectiveness in terms of attitude toward the ad. By contrast, the same emotion lead to considerably less favorable attitudes toward the ad when the advertisement portrayed

(43)

43 women ambitioned and independent (i.e. career-oriented), thus leading to a counter-effective influence.

To the best of the author of this master thesis’ knowledge, this study represents the first attempt to account for female responses to career-oriented versus decorative women role portrayals in advertisements through the evolutionary theory. Prior studies (e.g. Griskevicius et al., 2009; Kyrousi et al., 2016) attempted to examine women responses only to decorative female portrayals in ads, and it was done by capitalizing on the evolved mechanism of intrasexual female competition. By contrast, this study focused on expanding the scope by juxtaposing women attitudes towards ads and brands that leverage decorative and non-decorative female role portrayals in their advertising.

All things considered, this study shed light into why some women have diverging attitudes towards the ads that portray them in a decorative fashion and the ones that portray them career-oriented. Interestingly, with all the societal concerns regarding the pressure women face due to their idealized portrayals in media (Bissell & Rask, 2010), one would assume that women indeed have a negative attitude towards advertisements that leverage a decorative female portrayal and a positive one towards non-traditional female depictions (i.e. career-oriented). However, the medium context that ads are placed in and the emotions triggered beforehand seem to have an impact on this equation. In accordance with the literature review, stereotypic women portrayal is still present in the internet context (Zotos & Tsichla, 2014), in fact decorative female portrayal is still vastly used in female-audience websites (Plakoyiannaki et al., 2008) alongside with some emerging female portrayals like the non-traditional ones.

Advertisement research : effectiveness of ads that use female role portrayals