The moderating effect of the fit between an individual’s personal views and the gender role portrayal on attitude towards the ad

(1)

MASTER THESIS 2016

The moderating effect of the fit between an

individual’s personal views and the gender role

portrayal on attitude towards the ad

University of Amsterdam

Faculty of Economics and Business

Master of Science in Business Administration

Track: Marketing

Under supervision of: Dr. Jonne Guyt

By:

Student: L.L. ULCELUSE

Student no: 10048707

(2)

1

STATEMENT OF ORIGINALITY

This document is written by Student Lucia Loredana Ulceluse, 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)

2

ACKNOWLEDGEMENT

First, I would like to acknowledge Professor Mr. Dr. Jonne Guyt for all the support and guidance

that he provided during this thesis development process.

Furthermore, I would like to express my gratitude towards my friends and family who always

gave me a helping hand when needed, especially my sister. This thesis would not have been

possible without her support. At last, my thesis is written.

I hope you enjoy reading this thesis.

(4)

3

1. Introduction

The social roles and purchasing power of both men and women are changing at a fast pace

(Catalyst, 2015), with women taking up more high level jobs and men reaching a 16% share of

all at-home parents in 2012, from a share of 10% in 1989, according to a research done by the

Pew Research Center in 2014 (Livingston and Parker, 2011). With internet playing a major role

in achieving equality of gender, race, class, social orientation etc. and empowering these

different groups that are being discriminated (Mehra et al., 2004), social changes are happening

at higher speed than we are used to, and social networking sites (henceforth SNS) are enabling

this fast change by providing people with a voice that either they did not previously had, or did

not reach as many people at once with just a simple like, tweet or share (Burgers, et al, 2000). The focus of this dissertation is one particular type of social change that has been heavily

influenced by social media interactions – gender-role portrayals in media, with a focus on

advertising. Although both genders are increasingly moving away from stereotypical gender

roles, not all marketers follow this trend or perceive that today’s consumer can find outdated

gender-roles offensive. For instance, a campaign from the company Dell advertising a new

learning platform called “Delta” severely backfired (Cresci, 2015). The company built the Delta platform to target women coders by using a pink layout, images featuring women coding on pink

notebooks offering tips on how to find cooking recipes online. After receiving backlash from

consumers on Twitter, the website was shut-down in less than 10 days. A similarly negative

reaction was received by the French company BIC when in 2012 they introduced the new line of

pens “BIC For Her” (Vinjamuri, 2012). The pens, targeting women specifically, with “an elegant

design”, thin enough to “fit a women’s hand” as the product description stated, received numerous negative and sarcastic comments on its Amazon.com page. The story was eventually

(6)

5

picked up by bigger media outlets like Jezebel, Forbes, the Washington Post, and even made into

a sketch on the popular US day-time television show “The Ellen Degeneres Show”. Seemingly

not learning from this experience, in 2015, the South African branch of the company, posted a

photo on its Facebook page, depicting a woman in a suit, with the tag line “Look like a girl, Act

like a lady, Think like a man, Act like a boss”. Posted on Women’s Day, the photo generated a

new wave of negative reviews on Amazon and comments expressing outrage on the company’s

Facebook page, prompting it to ultimately apologize to consumers and withdraw the photo

(Nudd, 2015).

On the other hand, campaigns and brands aware of these societal changes, and able to reflect

them in a timely manner in their product advertisements can also trigger a media frenzy, but of

the positive kind this time. A case in point is the “Real Beauty” campaign launched by the Unilever owned brand Dove. Challenging the idea that there is only one type of beauty and

encouraging women to see their own real, natural beauty and feel confident with themselves, the

main campaign ad became viral, accumulating over 114 million views on Youtube.com within

the first month of launch, and shared 3.74 million times over various social media channels

(Stampler, 2013). The effect of these campaigns shows that it is extremely important for

marketers and companies more broadly to understand why certain gender role portrayals can lead

to backlash while others to praise and emphasizes the need to provide marketers with a

framework that will help them better understand their target.

One potential explanation as to why marketers are taking more time to understand and adjust to

the changes in perceptions about gender roles and how they influence consumers behavior and

attitude, has been put forward by Bodenhausen and Galen (1988), who argue that in a world

(7)

6

our minds recur to stereotypes as a survival technique to judge and make quicker decisions.

Stereotypes, thus, are an evolutionary coping mechanism, which enable people to conserve their

resources as it allows them to use mental shortcuts instead of engaging in deep cognitive

processes (Neil et al., 1994).

However, since stereotypes are very simplified versions and descriptions of how people should

be, using them as consumer insights in ads can often backfire (Burgess and Borgida, 1999). As

Thurau et al, (2004) point out, consumers feel like they are not understood or feel excluded from

a specific community because they do not fulfill the stereotypical characteristics of their

gender-role. They have not only started to react to these restrictive and often offensive stereotypes, but,

due to social-media, they are able to share opinions and influence one another or offer instant

feedback on a campaign and influence its outcomes with the number of likes, shares, tweets or

comments they post. Social media enables people to point out inappropriate or offensive

portrayals and raise awareness among other consumers who might not have noticed these

nuances otherwise. Moreover, social media helps people form an opinion based not solely on

their own experience, but also on that of their friends from Facebook, Twitter or people whose

opinion they trust or value.

Responding to the changing roles of men and women is still a slow process for advertisers and

marketers, therefore the need to fully and truly understand the new consumer, his or her needs,

wants and how they see themselves is imperative if marketers want to create more authentic

products and more relevant and effective communication (Mathwick 2002).With the social and

economic changes that are currently happening in the world at a faster pace than ever, marketers

(8)

7

influencing their attitudes towards the ad and subsequently, towards the brand and the product

itself (Edell and Burke, 1987).

Most of the existing literature has focused on how the images and messages used in ads affect

cognitive responses towards the ad, with more recent research linking also emotional responses

to sex-role portrayals (Leigh et al., 1987, Ulrich & Holancova, 2004). However, to the author’s

knowledge, there is no research focusing on how the fit between the portrayal and people’s own

views influences their attitude towards the ad together with the fit between the portrayal and what the individual thinks society’s views are about that gender role portrayal. This is a gap that

the dissertation project intends to fill. It explores the extent to which this fit influence attitude

towards the ad. I argue that the higher the fit between the two, the higher the attitude towards the

advertisement, and implicitly, towards the brand (Shimp, 1981). The main reason we are looking

at attitude towards the ad is because there is extensive literature that shows that attitude towards

the ad is a predictor of attitude towards the brand and most importantly, of consumer behavior:

from the affect transfer hypothesis which shows the process through which a certain attitude

towards the ad causes a certain attitude towards the brand, to findings about the structure of the

relation between the two and how attitude towards the ad act as a mediator between the ad and

the consumer’s behavior (MacKenzie,1986; Shimp,1981). The project uses an online

survey/questionnaire with 32 questions, with data collected during the months of April and May

2016, from a total of 150 respondents.

The contribution of the present work is twofold. Firstly, it has theoretical implications, as it

contributes to the existing literature and our understanding of emotional consumer response in

advertising. By providing an insight into how gender role portrayals influence attitude towards

(9)

8

can backfire and those that challenge what are increasingly perceived as stereotypes become

popular with consumers and the media alike.

Secondly, it has practical implications, with spillover effects for businesses and managers alike.

The results of the research supply marketers with fresh customer insights, thus helping them

understand how to better segment and target consumers. The ramifications of these findings in

practice are rather substantial: a survey of 628 women taken by the lifestyle website SheKnows

in 2014, found that 52% of the respondents said that they have bought a product because they

like the way women were depicted in the ad (Stampler, 2014). These findings translate into

higher turnover for companies and brands that have managed to portray their target in a relevant

way: Dove and the “Real beauty” campaign have seen a sales increase from $2.5 billion to $4 billion, Nike’s revenues increased by 15% from one quarter to the other after they started to have campaigns more targeted at women (one ad from these campaigns has been used in the survey on

which the analysis of this research is based), while Getty images has seen a 52% increase in sales

five months after introducing more stock photos depicting women in powerful roles, with a

collaboration with Facebook COO, Sheryl Sandberg, entitled “Lean In”(Ciambrielo, 2014).

The rest of the thesis is structured as follows. Section 2 critically engages with the existing

literature on gender stereotyping, sex-role portrayals in media, attitude towards the ad, a priori

attitudes, purchase intent and social media word of mouth. Building on this review, section 3

presents the conceptual framework guiding this project and the hypotheses derived from it.

Section 4 develops on the research design and methodology employed, while sections 5 and 6

present the results and discuss the implications these results have on both theoretical and

practical grounds. The dissertation ends with a discussion on the limitations of this research and

(10)

9

2. Gender role portrayal in advertising – a literature review

Advertising is no longer about solely communicating information about products, which are then

used by consumers to make informed decisions. Rather, modern advertising industry does so

much more than that: it gives meaning and significance to our products, it defines social status

and contributes to and informs our understanding of social behaviors in a society by providing

cues about the current cultural values and beliefs (Domzal and Kernan, 1992). To achieve this in

an, often less than a minute advertisement, advertisers and marketers frequently resort to the use

of stereotypes, whether they are gender, race, age, religion or social class related. To show that a

product is genuine and high quality – whether it is a German car, Italian pasta or French wine,

the easiest way is to hyperbolize the properties of the product and/ or of the individuals using it

or producing it.

By itself, the stereotype does not carry either negative or positive values; it is only the context in

which it is used and the meaning that is given to it that can give it positive or negative

connotations (Frith and Mueller, 2003). According to Hamilton (1979), stereotypes contain not

only the perceiver's knowledge but also his/hers beliefs about different social groups.

Individuals, thus, interpret advertisements through their own belief system, against beliefs shared

by other people within the society (Frith and Mueller, 2003).

What happens, however, when these core beliefs change, while marketers still hang on to easy to

use stereotypes? In other words, what happens when the image of gender roles, or race or age or

class as depicted in advertisements no longer fits the individuals’ own perceptions? As the

examples used in the introductory sections show, the repercussions can be significant for a

(11)

10

and the company’s image as built through ads, that is, a fit between individual’s perceptions and the image presented in the ad, becomes paramount for ensuring a company’s success.

The following sections develops more on the concept of stereotypes and how they have been

used in advertisements in relation to gender, and identifies the literature gap that this projects

intends to fill.

2.1. Gender stereotypes

Stereotypes have been around for centuries and they have been used as a shortcut to form an

opinion on and classify people. Hence, the use of gender stereotypes in advertising has been

quite thoroughly researched (for an overview see Wolin, 2003).

Defined as holding certain expectations from someone because of their gender, gender-related

stereotypes are not only widely spread but also persistent over time (Bem, 1993). At a very broad

level, sex-role stereotypes consider women to be more communal (care for others, nurturing,

affectionate etc.) while men are perceived to be more agentic (aggressive, ambitious, leader

qualities) (Deaux and Kite, 1993). Although social roles, culture or social norms influence the

expectations we have from a specific gender, those expectations also in turn influence an

individual’s behavior. Deaux and Major (1987) looked at gender-related behavior and saw that the behavior of a certain gender is influenced by the expectations of the perceiver – building here

on the idea of expectancy conveying behavior – but also by the person’s self-concept, meaning

the identity they hold of themselves and the self-expectancies that are attached to it. On top of a

descriptive component about how men and women are usually perceived, gender roles also

contain a strong prescriptive component which indicates how men and women “should be” and

(12)

11

expectations and beliefs not only through the experiences they have with one another, but also

through the stories they see or hear around, especially the ones presented in the media. A

significant contributor to the way people tend to construct stereotypes about different genders is

the way men and women are portrayed in television and advertising. Previous studies have

shown that images presented on television influence what people perceive to be social reality and

that, because this medium is so ubiquitous and mainstream, it enables, and often encourages, the

development of very similar views across different typologies of people (Gerbner et al., 1986).

What is more, Pingre (1978) and Atkins and Miller (1975) have found evidence that advertising

has even influenced the gender role perceptions of children. In different studies they observed

that if women were portrayed doing jobs which have traditionally been associated with men, the

children were more likely to see these occupations as suitable for women too. These are

important findings that show the power that television and advertising have on people’s attitudes

towards gender stereotypes and role portrayals and how this medium can contribute to social

change by broadcasting more broad gender role portrayals.

However, Rudman and Glick (2001) have shown that when men and women defy the

expectations of their prescribed gender-role, such as displaying a more assertive behavior that is

traditionally associated with men (in the case of women), or displaying a more affectionate or

caring behavior that is usually associated with women (in the case of men), backlash effects may

arise They showed there is actually a paradox when women try to manage their image and the

way they come across to others: on one hand, women might not achieve a promotion because

they do not seem as competitive as men when they do not display an assertive or aggressive

behavior, but on the other hand, when they do display more authority they can receive backlash

(13)

12

expectations that comes with them. Rudman (1998) defined backlash effects as the price (both

social and economic) an individual pays for behaving counter-stereotypically. Other studies have

found evidence to the contrary, meaning that challenging these stereotypes and using more

progressive gender role portrayals has positive effects with certain audiences, especial female

audiences that hold a priori positive attitudes towards more modern gender role portrayals

(Whipple and Courtney, 1985). It is important to understand this phenomenon as even

well-intentioned portrayals might not have the effect they were intended to have. The

system-justification theory sheds some light into why these backlash effects appear. It argues that people

hold gender status beliefs as a way of protecting what they considered their just world, their just

system (Jost & Banaji, 1994) and any change can be perceived as a treat to their own status. The

degree to which this is a higher or lower threat may vary according to the extent to which that

particular individual is gender schematic. Deaux and Major (1987) classify individuals as gender

schematic – those that hold more traditional views and sex-role expectations, and gender

aschematic – those that hold fewer sex-role expectations (or none at all). Other factors that

influence how violations of sex-role expectations are perceived are situational cues, especially

cues that stress the biological sex, like interactions with the opposite sex, and therefore, gender

schematic individuals will tend to stereotype more than gender aschematic individuals, especially

when it comes to the opposite sex (Park & Hahn, 1988). However, the system-justification

theory does not provide an explanation as to why, in some cases, challenging stereotypes has

actually a positive effect or why conforming to stereotypes might backfire. Moreover, taking into

account all these social and economic changes, attitudes held a priori cannot fully explain the

attitudes and behaviors of consumers, particularly since these types of studies are looking at the

(14)

13

is precisely where a literature gap has been identified: exploring not only the consumer’s a priori

general attitudes or their justification system, but rather taking together the fit between their

views on the gender role portrayed in a specific ad and the fit between the gender role portrayal

and what they believe society’s views are about these gender role portrayals and how the two influence attitude towards the ad and purchase intent.

Next, a closer look at media portrayals of gender roles and its effects is going to be presented.

2.3. Media portrayals of gender-roles

Media has the capacity to create a “reality”, to establish relationship power dynamics and to shape beliefs. The “reality” that is presented in the media is often considered to be a reflection of how things actually are or how they should be approached (Street, 2011, Gerbner et al., 1980).

The advertising industry has a strong impact on the way people process information and form

opinions. A report by Svensson (2008) for the Committee on Women’s’ Rights and Gender

Equality within the European Parliament demonstrates how images and messages presented in

advertisements influence the way we tend to build our own identities and set standards for what

we should consider or expect to be “normal”. However, there is also empirical evidence showing that the relationship might be in the opposite direction. A meta-analysis based on 64 primary

studies on TV and radio commercials found that gender-related developments and new values in

society are the ones that influence role portrayals in advertising (Eisend 2010), and not the other

way around. In whichever direction the relation goes, there is not much evidence as to why a

(15)

14

2.4. Literature gap

2.4.1. Gender role portrayals in advertising and a priori attitudes

Although most core beliefs about gender are settled within our mind, society has evolved and

traditional gender roles are increasingly being challenged and, therefore, so are our attitudes and

expectations about gender roles. In recent years, we have seen how women take on roles

traditionally assigned to men and men are taking up more of a role in the family life and

caretaking of children, which used to be primarily assigned to women (Bardy, 2014). These

changes have translated also in the media with women and men being represented more diversely

in movies, TV shows and advertisements. Companies and brands that have observed and

embraced these cultural and societal changes, have capitalized on it with campaigns that

challenge gender stereotypes, gaining substantial public support, mainly expressed on social

media; at the other end of the spectrum, brands that still use outdated stereotypes on gender role

portrayals and fail to understand how consumers have evolved, have been less successful and

some have even experienced backlash from customers and the broader media (Lantagne, 2014).

The rapid growth of social media platforms has contributed to the ways consumers interact,

influence each other and drive change. Companies themselves are starting to value the type of

engagement they can build through social media with customers. Therefore, marketers need to

pay attention to these conversations and consider multiple segments when they use gender role

portrayals, not just the segments they are directly targeting. Another thing that they need to

consider is the extent to which a priori attitudes affect the attitude towards the advertisement

both when the information is being processed through a central route and when it is being

processed through a peripheral route. Studies have shown that when consumers evaluate ads not

(16)

15

the attitude towards the ad (Campbell, 1995). Moreover, the level of involvement has a great

impact on determining which persuasion route consumers are going to follow. For

low-involvement products and services, more emotional stimuli work therefore a priori attitudes

towards certain gender role portrayals will have a big influence on the attitude towards the ad,

whereas for a high involvement product, argument related to quality are more important, so the

effect of a priori attitudes towards gender role portrayals it is not going to be as high (Petty et al.,

1983; Sengupta and Dahl, 2007).

2.4.2 Gender role portrayals in advertising and the society’s views

Being a place where people can exchange and generate content, social media has played an

important role in helping consumers voice their opinions and show either their support or

disapproval of a campaign, brand or company.

Social media was defined by Kaplan and Haenlein (2010, p. 61) as “a group of Internet-based

applications that build on the ideological and technological foundations of Web 2.0, and that

allow the creation and exchange of User Generated Content”. Shares, likes, tweets or comments can now influence people’s attitudes towards a campaign and even help measure the success or

failure of an ad (Watts & Dodds, 2007). Although online engagement does not automatically

translate into offline behavior (Alhabash et al., 2015), the negative information people see on

social media about brands has been reported to have a stronger impact than the messages they

see in traditional media (Pace et al., 2014). Therefore, managing the brand content on social

media and paying attention to the electronic Word of Mouth (eWOM) happening around the

brand is of great importance for companies. Although one of the main reasons people engage

with brands on social media is for economic reasons, motives such as social interaction or

(17)

16

certain gender-roles and may have a strong influence on whether the attitude towards the ad that

consumers express online is positive or negative.

The “Like a girl” campaign from P&G’s brand Always was very successful and became viral on social media for aiming at redefining what the expression “like a girl” means and challenge the idea that feminine behavior has a lower status than masculine behavior (Deaux, 1984, Diaz,

2015). Similarly, a recent campaign from AXE – “Find your magic” – featuring different types

of men, each with his own quirks and personality - a departure from the brands’ stereotypical

portrayals of men in previous campaigns (Nudd, 2016) - received positive feedback on Twitter.

Taking into consideration that Positive Word of Mouth (PWOM) has been reported to have a

greater impact on purchase probability than Negative WOM (NWOM) (Vanhuele & Wright,

2008), marketers should not only try to avoid or manage backlash, but should rather aim at

implementing campaigns that will generate PWOM and reach more consumers organically.

Social media has not only given a voice to consumers, but has also allowed consumers to hear

the “voice of society” and therefore be influenced by it. 2.4.3. Gender differences in attitude towards the ad

Studies have shown that women and men react differently to advertisements (Stern & Holbrook,

1994; Meyers-Levy & Sternthal, 1991) and especially when it comes to gender issues (Prakash,

1992). However, there have been different findings regarding what gender role portrayals work

best and for whom: Whipple and Courtney (1985) found that modern portrayals work better for

women than more traditional and stereotypical ones, but they also showed that these portrayals

might have the opposite effect on people that are not targeted by them. Moreover, one would

expect that since men, on average, have more positive attitudes towards sex than women (Oliver

(18)

17

advertising, even if it is a gratuitous use of the sexual cue and might even be objectifying

women. However, studies have shown that using sexual images that have nothing or little to do

with the product leads to negative attitude towards the ad for both men and women (Simpson et

al., 1996). One potential explanation for this is that the viewers, both male and female, believe

such portrayals are unethical and manipulative (LaTour and Henthorne, 2003). Findings have

shown that one of the reasons why women react differently to advertisement is because they tend

to elaborate more than men, even when that message it is not directly targeted at them (Wood,

1966). These are consistent with our hypothesis than females are going to have a stronger

attitude towards the ad than men, both as a direct effect of gender, as well as due to the

interaction gender x gender role portrayal.

2.4.4. Attitude towards the ad and Purchase intent

It is becoming more and more important for marketers to understand what factors trigger certain

emotional responses from consumers and how these responses shape their attitudes towards the

ad and purchase intention. Orth and Holancova (2004) showed that a priori attitudes towards

gender role portrayals affect consumer emotional and attitudinal responses which in turn

translate into attitudes towards the brand and purchase intent. Moreover, they found that gender

has also an influence on consumer responses, with women responding more positively to modern

and progressive portrayals than men. There is extensive literature on how useful attitudes are to

predict behavior, how the attitude towards the ad is connected with purchase intent and how

these different construct vary from hedonic, products purchased for recreational or pleasure

related reasons, to utilitarian product, mostly purchased for practical and functional reasons

(Mitchell and Olson, 1981; Batra and Ahtola, 1991, Singh and Spears, 2004). Therefore, since

(19)

18

marketers and advertisers, this research is going to focus on the effect of gender, gender role

portrayals, the fit between the gender role portrayal and the respondent’s own views, and the fit

between the gender role portrayal and what the person thinks society’s views are about that gender role portrayal on attitude towards the ad. This latter effect further translates into purchase

intent. A question measuring purchase intent was also included in the questionnaire to confirm

these findings. This relation will be further discussed in the data analysis section.

In the next chapters, the main effect of the independent variable gender role portrayal on the

dependent variable attitude towards the ad is going to be tested first. Second, both the fit between

the portrayals from the ads and the respondent’s personal views on the gender role portrayal from the ad, and the fit between the portrayals and what the person thinks society’s view are about that specific gender role portrayal are going to be investigated both for their direct effect

on attitude towards the advertisement and for the interaction with the gender role portrayal. The

variable gender it is also going to be tested both for its direct effect and for the interaction

between gender x gender role portrayal. Female respondents are the reference group against

(20)

19

3. Conceptual framework

After looking at the existing literature, identifying a gap and formulating a research question, we

are now going to zoom in on the framework that will help us understand how the gender role

portrayal (henceforth GRP) in the ad influences attitude towards the ad and the associated

purchase intention.

We aim to find an answer by looking at three dimensions: the evaluation people give to the ad

when considering the gender role portrayal (in terms of traditional vs progressive), the fit

between the views expressed in the ad and the respondents own views and the fit between the

views expressed in the ad and what the respondent believes society’s views are.

Research question:

 How does the gender role portrayal in the ad influence the attitude towards the ad and purchase intent?

 How does the fit between the gender role portrayal in the ad and the person’s

view on that gender role portrayal influence the attitude towards the ad and

purchase intention?

 How does the fit between the gender role portrayal in the ad and the person’s

view on how society views the gender role portrayal influence the attitude

towards the ad and purchase intention?

The theoretical framework that will help us answer these questions is presented in the research

design model in Figure 1 bellow. Further on, the factors and variables included in the model will

be discussed as well as the relationship between them. Hypotheses will be then formulated and

(21)

20

Figure 1: Research Design Model

The independent variables included in the research model are: age (as control variable), gender,

the gender role portrayal in the ad, the fit between the gender role portrayal in the ad and the

person’s own views about the gender role portrayal in the ad and the fit between the gender role portrayal in the ad and what the person thinks society’s views are about that gender role portrayal. Gender and the last two variables mentioned above – the two types of fit – are also

going to be tested as the moderators of the relation between the independent variable gender role

portrayal in the ad and the dependent variable attitude towards the ad. The effect of attitude

towards the ad on purchase intent is going to be discussed at the end.

Further on we are going to explain the main independent variables and the three moderators. The

(22)

21

progressive the GRP in the ad is, having as a frame of reference the stereotypical gender role

portrayals society usually attributes based on gender (Wagner and Banos, 1973; Wiles and

Tjernlund, 1991). The fit between the GRP in the ad and the person’s own views represents how

much the respondent’s views on gender roles is in line with the views presented in the ad. The independent variable – fit between GRP in the ad and the respondent’s perception of how society

views the gender role portrayal – it is similar to the first type of fit except in this case the fit it is

connected to what the respondent thinks society in general sees the GRP in the ad, and not about

his own personal views. For the independent variable gender, females are going to be taken as

the reference group in the analysis part.

Further on, hypotheses will be formulated and arguments based on existing literature will be

provided to support our research design model.

In the literature review section, evidence was presented supporting the idea that gender role

portrayals affects consumer’s emotional responses, which in turn shape consumer attitudes.

Moreover, the evidence showed that modern portrayals of women can be successful with a

certain target, while the same “progressive” portrayal can backfire with audiences that do not

feel targeted by this portrayal and result in negative emotional responses and subsequently in not

wanting to buy the product (Whipple and Courtney, 1985). However, a direct effect of a specific

gender role portrayal on the general attitude towards the ad has not been tested yet. The

following general directional hypothesis it is formulated:

H1: The gender role portrayal in the ad positively influences attitude towards the ad.

Further on, the aim it is to find an explanation why this backlashes might happen by looking at

(23)

22

the ad. The higher the fit is between the two variables, the higher the attitude towards the ad will

be. This it is extremely important to understand, because, although there is evidence that ads to

which females react positively to cause negative reactions in men (Batra and Holbrook, 1987),

there it is not much literature exploring the underlying reasons why this it is happening.

H1 A: The fit between the GRP in the ad and the individual’s own views on that GRP positively influences attitude towards the ad.

H1 B: The fit between the GRP in the ad and the individual’s own views on that GRP positively moderates the relation between the GRP in the ad and attitude towards the ad.

Jost and Banaji (1994) showed that society’s views play a major influence on how people form their own views and that is very hard to just long established perceptions. Therefore, a high fit

between the GRP and what the person thinks society’s views are is also going to positively influence the attitude towards and purchase intent.

H1 C: The fit between the GRP in the ad and what the individual thinks society’s views are about that GRP positively influences attitude towards the ad.

H1 D: The fit between the GRP in the ad and what the individual thinks society’s views are about that GRP positively moderates the relationship between the GRP in the ad and attitude towards the ad.

Orth and Holancova (2004) showed that women and men react differently to gender role

portrayals in advertising, with women having a stronger attitude towards gender role portrayals

than men. This it is because women are more aware of gender stereotypes than men are (Lull et

(24)

23

advertisements than men are. We expect women to have a stronger attitude towards the ad than

men both when gender has a direct effect on the dependent variable and when gender interacts

with the gender role portrayals and moderates the relation between the gender role portrayal and

attitude towards the ad.

H2: There is a significant gender - gender role portrayal interaction effect on attitude towards the ad with downstream effects to purchase intent. Women and men will react significantly different to the gender role portrayal.

Age will not have a significant effect on attitude towards the ad. There have been several studies

showing both significant differences between different age groups and also no differences at all

(Eisend 2010; Kim and Lowry 2005). Moreover, since our sample is mostly students aged 18-35,

we believe there will not be much variation in the attitude towards the ad when it comes to age.

Since the sample is mainly European, students, aged 18-35 and considering that views on gender

roles are constantly and increasingly changing, the results of the research are valid for a short

timeframe of 2 to 3 years. The access to internet and increase use of social media will also

(25)

24

4. Research design

The relation between gender role portrayals and attitude towards the ad will be further analyzed.

First, the effect of the first type of fit will be analyzed – the fit between the GRP and the

respondent’s own view – and further on the second type of fit – the fit between the GRP and the respondent’s perception of what society’s view are. The effect of these two moderators will be tested for attitude towards the ad and the downstream effects on purchase intention. Gender and

age will be used as control variables.

The chapter is structured as follows: the research design, sample, procedure, manipulation and

presentation of the variables.

4.1.Research design, sample and stimuli

The data on which the analysis will be performed comes from an online questionnaire, collected

during the months of April and May 2016, for a total of 150 respondents. This approach provides

the appropriate way to collect the necessary data for the variables that are going to be studied as

well as for the relation between them and the moderators that have an effect on this relation.

Furthermore, the quantitative data can be analyzed and validated in SPSS to test the type of

hypotheses this research aims to test.

Respondents were randomly selected by distributing the survey on different social media

platforms. The reason for this is to provide the researcher with a diverse sample that will allow

for variance in terms of age, gender and evaluations and avoid a skewed distribution. The

respondents filled in an online survey with questions based on six different advertising

campaigns. The campaigns were chosen in such a way as to provide traditional, modern and

neutral gender role portrayals. We expect the fit to be more relevant in the two extremes: the

(26)

25

gender role portrayals. The campaigns were chosen based on the same frame of reference as the

Gender Role Portrayal (Wagner and Banos, 1973; Wiles and Tjernlund, 1991). Moreover, a

social media analysis was done by collecting and analyzing the brand sentiment from tweets

posted by users voicing their opinions about three of the campaigns used in the questionnaire to

confirm our evaluation of the commercials. All three commercial had mostly positive comments,

as expected since they were chosen for this research for their more modern portrayals of gender

roles. The three campaigns are from Ariel, Axe and Nike and aimed at portraying progressive,

modern and respectively neutral gender roles portrayals.

Before the original questionnaire, a pre-test questionnaire was filled in by 15 respondents to test

if the stimuli provided were to be considered applicable. The second question about the fit

between the gender role portrayal and the respondent’s own views about the gender role portrayal was unclear to the respondents. There was a confusion whether the question was asking

about the respondents personal views on the gender role portrayal or whether the question was

asking about how the gender role portrayal it is seen by society in general. Therefore, the

question was rephrased and split into two questions: one regarding personal views and the other

one regarding society’s views. Moreover, respondents filled in the comment section with similar comments regarding how these ads will also influence them to buy or not buy the product,

therefore an additional question was added to measure purchase intention in addition to attitude

towards the ad. Both questionnaires can be found in the Appendix.

The data will was collected by having 150 respondents fill in the questionnaire online and

answering questions after watching 6 different videos with advertising commercials. The stimuli

(27)

26

Carl's Jr - All natural burger Axe - Find your magic

Nike - Better for it Coca- Cola - Parenting

Ariel - Share the load Hyundai - First date

Source: Youtube (link in titles)

4.2.Procedure and manipulation

The questionnaire starts by asking respondents to state their age and gender. Further on, the

commercials are in random order for every respondent to control for any influence on the

responses based on responses from previous questions. The respondents have to watch one

commercial and answer 5 questions about it and then move to the next commercial and answer

(28)

27

until the end of the survey which ended after respondents answered the last question from the last

commercial.

The first question asks respondents to evaluate the gender role portrayal in a commercial on a 5

point scale: “very traditional", “traditional”, “neutral”, “modern”, “progressive”. The next

question measures the first type of it fit – the fit between the gender role portrayal and the

respondent’s own view on a 5 point scale: from “none at all” to “a great deal”. The third question measure the second type of fit – fit between the gender role portrayal and what the respondents

thinks society’s views are – on the same 5 point scale: from “none at all” to “a great deal”. The fourth question measure attitude towards the ad on a 7 point scale from: “extremely negative” to “extremely positive”. The last question measures purchase intent on a 5 point scale: from “none at all” to “a great deal”. Each respondent has to answer the 5 questions for all the 6 commercials.

The control variables age and gender are collected in the beginning of the questionnaire by

asking respondents to “Please select your age” and “Please select your gender”.

All variables are measure with a single-item measure since it has been shown that for attitude

toward the ad (AAd) single-item measures can predict validity and should be used as well as for

other variables that consist of a concrete attribute/object (Bergkvist and Rossiter, 2007).

4.3.Strengths and limitations of the research design

By conducting an online survey any biases that could be caused during interviewing or any

personal interaction between the respondent and the researcher are being eliminated. Moreover, a

validity check on the data is going to be performed before any analysis is made and made sure

(29)

28

Although there is evidence of the validity of single-item measures (Bergkvist and Rossiter, 2007),

considering that our sample will be only 150 respondents the results might not be as significant

therefore for further validation the study should be replicate on a larger sample taking into

(30)

29

5. Data analysis

For this study a single-item measure has been used for each variable. A single-item measure can

be appropriate when the construct to be tested is clear to the respondents and is specific enough

(Wanous et al., 1997). Moreover, Scarpello and Campbell (1983) found that a single-item

measuring overall job satisfaction is better than a multiple-item measurement because the items

might not include things that are important to the employee and therefore the conclusions about

his overall job satisfaction might be wrong. Therefore, since the objective of this study is to look

at overall attitude towards the ad, purchase intent and the two types of fit, a single-item measure

is found to be appropriate.

5.1. Validity

The data file was first analyzed for any missing values – there are no missing values. Further on,

a frequency was performed check to see if there any errors in the data – there were no errors in

the data.

5.2. Analysis

The majority of respondents were between the age of 25-34 and 70 % of the respondents were

female. The graphs for this overview as well as the normality check can be found in the

Appendix.

The variables were operationalized as follows:

 Gender role portrayal in the (as seen by the respondent)  Ad Evaluation  Fit between the gender role portrayal and the respondent’s own views  Pfit

 GRP x Fit between the gender role portrayal and the respondent’s own views  Moderator1

(31)

30

 GRP x Fit between the GRP in the ad and what the respondent thinks society’s views are  Moderator 2

 Attitude toward the ad  AdAt  Gender: Females as reference group

 Gender x Gender role Portrayal  GenderModerator  Purchase Intent  PI

In the following pages the effect all three moderators (fit between gender role portrayal and

respondent’s own view and fit between the gender role portrayal and what the respondent think society’s views are and gender) it is going to be tested for attitude towards the ad together with each independent variable (gender, personal fit, society fit and gender role portrayal). Age it is

going to be used as a control variable.

5.2.1. Hypotheses testing

First, the effect of each independent variable it is going to be introduced in the model to test their

direct effect on attitude towards the ad and then the three moderates are going to be introduced to

test the interaction effects.

Model Summary

Model R R Square Adjusted R Square Std. Error of the Estimate

1 .714a .510 .503 1.349

a. Predictors: (Constant), Age, Gender, Gender Moderator, Pfit, Sfit, Moderator 1, Moderator2, gender role portrayal

ANOVAa

Model Sum of Squares df Mean Square F Sig.

1 Regression 1119.493 8 139.937 76.906 .000b

Residual 1077.196 592 1.820

(32)

31

All variables were introduced in the model and one linear regression was run. As seen in the

Figure above, the model was statistically significant p < .001. The R square is Rsquare = .510

Therefore the whole model explains 51% variance in attitude towards the ad, after controlling for

age. Coefficientsa Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta 1 (Constant) 10.126 2.085 4.856 .000 Women -.072 .146 -.017 -.494 .621 Age -.321 .095 -.098 -3.372 .001

Gender role portrayal .150 .040 .160 3.730 .000

Personal fit .811 .046 .550 17.484 .000

Society fit -.368 .065 -.189 -5.625 .000

Moderator1 -.334 .121 -.117 -2.765 .006

Moderator2 -.134 .106 -.046 -1.260 .208

GenderModerator .230 .098 .080 2.344 .019

a. Dependent Variable: Attitude towards the ad

As seen in the table above, seven out of four predictor variables were statistically significant.

The variable gender does not have a direct significant influence on attitude towards the ad.

However, the gender x gender role portrayal interaction has a significance of p < .001 and a

positive Beta of β = .080. Since the reference group is women, this means that women react

stronger than men to different gender role portrayals. A closer look at the differences between

the two genders will follow in the next section.

The effect of the independent variable gender role portrayal on attitude towards the ad it is very

strong, with a p value of p < .001 and a positive Beta of Beta β = .080, which means that the

(33)

32

be. The fit between the gender role portrayal and the person’ own views about the gender role

portrayal it is also statistically significant with a positive Beta β = .550. Moreover, the interaction

between this fit and the gender role portrayal in the ad it is statistically significant with a

negative Beta of β = -.117 and a p value of p < .05, which means that the lower the fit, the

higher the attitude towards the ad is going to be. The fit between what the person thinks

society’s views are about the gender role portrayal and the gender role portrayal it is statistically significant with a negative Beta of β = -.189 and a p value of p < .05, which means that the

lower the fit the higher the attitude toward the ad it is going to be. However, the interaction

between this fit and the gender role portrayal in the ad it is not statistically significant ( p > .05).

5.2.3. Testing for gender differences in attitude towards the advertisement

The differences between men’s and women’s attitude towards the ad was tested with an analysis

of variance by looking at the interaction with the fit between the gender role portrayal in the ad

and the respondent’s own views on the gender role portrayal. As the results show in Table 5

below, women’s attitude towards the ad was higher than men’s attitude towards the ad when

Gender x Personal fit (Moderator1) interaction effect

Dependent Respondents Level of personal fit

None Little Moderate A lot A great deal

Attitude towards the ad Purchase intention females males females males 3.36 3.94 4.91 5.40 5.82 2.60 3.83 4.70 5.77 6.04 3.02 3.31 3.76 3.99 4.38 2.95 3.00 3.76 4.00 4.80

(34)

33

there was a higher fit, and it was lower than men’s when the fit was little or none at all. This

means women have a more positive attitude towards the ad when it comes to more modern

gender role portrayals and a more negative attitude towards than men when it comes to more

traditional portrayals. This is consistent with current findings that show that women react

stronger to gender role portrayals than men as they are more aware of stereotyped gender role

portrayals in advertisements and thus reacted less favorable to stereotypical portrayals (Lull et

(35)

34

6. Discussion

The objective of this research was to examine the effect on attitude towards the ad of gender, fit

between the gender role portrayal and the respondent’s own views about the portrayal and fit between the gender role portrayal in the ad and what the person thinks society’s views are about

the gender role portrayal. The tested model explained 51% of the variance in attitude towards the

ad. The results showed that the interaction (moderating effect) between the fit between the GRP

in the ad and the person’s own view about that GRP has an influence on people’s attitude

towards that ad and subsequently on their purchase intent in a negative direction. However, when

the direct effect of the fit on attitude towards the ad was tested, the direction was positive. These

differences are explained by the fact that when the interaction effect it is introduced in the model

the relation between the gender role portrayal and attitude towards the ad it is not linear

anymore. Therefore, there are situations when a high fit leads to a strong attitude and moments

when a low fit can also to a strong attitude. As seen in analysis of variance between the two

genders in Table 5, this phenomenon happens because people react more strongly in the

extremities, than in the middle. Therefore, usually when the fit is very low or very high the

attitude it is also strong. If there is a fit, but on a moderate level, it does not influence people’s attitude to a high degree. This is the reason why the six campaigns were evenly distributed: two

that portrayed more traditional gender role, two that portrayed more modern or progressive

gender roles and two that portrayed more neutral gender roles. As seen in the data analysis

section, the highest means were towards the left (for the very traditional) and the right (for very

progressive) which shows that people had a stronger attitude for more extreme portrayals,

whether they liked them or not. This it is quite common sense when taking into account that the

(36)

35

people either really like or really dislike, and not campaigns that stimulate a more “neutral”

feeling (Libert and Tynski, 2013). These findings build upon and bring a new perspective to the

existing literature about how a person’s own views about certain gender roles affect their attitude towards the ad when a specific gender role it is being portrayed.

When it comes to the fit between the gender role portrayal and what the person thinks society’s

views are about that specific role portrayal, the direct effect on attitude towards the ad it is

significant and negatively influences the dependent variable. This means that higher the fit, the

lower the attitude towards the ad. This it is consistent with current findings that show that people

are influenced by the opinion of people around them, especially by people similar to them

(Williams, 1995). The questionnaire posed the question “To what extent do you believe society’s

views align with the gender role portrayal”, therefore, since people are influenced by the opinion

they believe other people around them have, if that opinion it is a very commonly held one, it

will not stimulate a strong attitude as there it is no element of surprise that will elicit a strong

emotion (people will have a moderately negative or positive attitude) (Orth and Holancova,

2004). These findings are very important as current research focused mostly on the person’s own

attitudes and not so much on the effect of society’s opinion on the attitude towards the ad regarding gender role portrayals.

Gender and age have different degrees of influence. Age it is significant when it comes to

attitude towards the ad, with a negative Beta, which means that the younger the person is, the

lower the attitude towards the ad will be. This contradicts our hypothesis that age does not have a

significant effect on attitude towards the ad. However, there are findings that show that a

younger audience is more malleable in terms of learning new things and therefore more open to

(37)

36

As expected, women and men have different attitudes and evaluations of the gender role

portrayals. Both the direct effect of gender and the interaction of gender x gender role portrayal

are significant and positively influence attitude towards the ad. Women have a stronger attitude

than men when it comes to gender role portrayals which it is in line with the existing literature on

gender differences in attitude towards the ad. Studies have shown women are more aware then

men when it comes to gender stereotypes in advertising and (Lull et al., 1977) and tend to form

more negative attitudes then men towards these stereotypes. This it is consistent with our

findings in Table 5 that women tend to have a more negative attitude than men when it comes to

roles that are regarded as very traditional. Moreover, when it comes to more progressive

portrayals our findings are also consistent with the existing literature which states that more

modern portrayals of women are more effective for the female audience than are traditional ones

(Ford and LaTour, 1993). The same findings can be seen in Table 5, were women had a higher

attitude towards the ad then men when they considered the portrayal to be “modern” or

“progressive”. This it is extremely important for marketers to understand when they are targeted consumers, even when those consumers are male. As current findings show, women tend to

analyze deeper than men even advertising messages that are not addressed to them, and, since

social media enables people to speak their mind to a larger audience than just their close circle of

friends, backlashes on a campaign or a brand can arise even when the target was not initially

offended by the portrayal (Wood, 1966; Pace et. all, 2014).

Furthermore, the effect of attitude towards the ad translates further on to purchase intent. The

questionnaire also measured how purchase intent it is influenced by the gender role portrayal

through the effect on attitude towards the ad as the moderator. The result showed that attitude

(38)

37

consistent with the existing literature and studies that show that gender role portrayals have an

effect on attitude towards the ad which in turn extends to purchase intent (Whipple and

Courtney, 1985, Orth and Holancova, 2004). These findings are of importance for marketers

especially when they wish to target a broader and heterogeneous audience for which the same

gender role portrayal might have opposite effects. As seen in the results section, the most

favorable and least favorable attitudes were towards the extremes – either extremely traditional

portrayals or extremely modern portrayals. Thus, if their target is broader, marketers might want

to opt for more neutral portrayals of both gender that could reach and be relevant for a larger

target group.

7. Limitations and further research

The current study has several limitations that could serve as a starting point for future research.

First, it looks at only one type of portrayal for 6 different products, therefore the positive or

negative emotions of each portrayal cannot be generalized for the whole category/type of

product/type of service. Moreover, the research included low, medium and high involvement

products (from Coca-Cola to the Hyundai car), therefore several types of gender role portrayals

need to be further tested for each of these types of products to see whether the fit becomes less or

more significant depending on the combination of type of product and gender role portrayal.

Second, considering that gender has a significant influence both directly and as a mediator,

further research could explore further reasons why these differences appear in men and women

besides their a priori attitudes. Studies have shown that situational cues such as the marital status

or even biological reasons (menstrual cycle) can have an influence in the way men and women

perceive gender role portrayals (Buss, 2000). Moreover, a 2014 survey from the Harvard

(39)

38

roles overall, when asked specifically about their own lives, partners of families, men were less

likely to see their careers or household responsibilities on an equal level as women (Ely et al.,

2014). This means that although men do agree that certain role portrayals are offensive or sexist,

these attitudes might not reflect in totality what they actually think. Therefore, further research

should be made to test for implicit attitudes in an experiment setting where all the variables can

be controlled. Third, since the sample it is mostly European, further research can focus on the

differences between cultures as it is shown that culture has an effect on the way people perceive

and assign gender roles (Siu and Au, 1997). The second moderator – fit between the gender role

portrayal and what the individual thinks society’s view about that gender role portrayal – might be very interesting to explore for differences between more individualistic and more communal

cultures as the influence of society on consumer’s attitudes might differ from one culture to

(40)

39

References

Alhabash S., McAlister A.R., Lou C. & Hagerstrom A., (2015), From Clicks to Behaviors: The

Mediating Effect of Intentions to Like, Share, and Comment on the Relationship Between Message Evaluations and Offline Behavioral Intentions, Journal of Interactive Advertising,

Volume 15, Issue 2, p. 82-96

Anderson, J. C, & Gerbing, D. W, 1988, Structural Equation Modeling in Practice: A Review

and Recommended Two-Step Approach, Psychological Bulletin, Vol. 103, No. 3,411-423

Asur S., Huberman B.A., 2010, Predicting the Future with Social Media, http://www.arxiv.org arXiv:1003.5699v1.

C Atkins, M Miller, The effects of television advertising in children: Experimental evidence Paper presented to the Mass Communication Division, International Communication Association (1975)

Bardy Laveme H., (2014, Dec 21), Men and Women: Their Changing Roles, retrieved from http://www.huffingtonpost.com/laverne-h-bardy/men-and-women-their chang_b_6018414.html

Batra R, Ahtola O. T., (1991), "Measuring the Hedonic and Utilitarian Sources of Consumer

Attitudes," Marketing Letfers, 2 (April), 159-70.

Bem S. L., (1993), The Lenses of Gender: Transforming the Debate on Sexual Inequality. Yale University Press. Retrieved from http://www.jstor.org/stable/j.ctt1nq86n

Bergkvist L. and Rossiter J. R. (2007) The Predictive Validity of Multiple-Item Versus

Single-Item Measures of the Same Constructs. Journal of Marketing Research: May 2007, Vol. 44, No.

2, pp. 175-184.

Bodenhausen, Galen V., Stereotypic biases in social decision making and memory: Testing

process models of stereotype use. Journal of Personality and Social Psychology, Vol 55(5), Nov

1988, 726-737.

Burgers A., Ruyter K., Keen C., Streukens S., (2000) "Customer expectation dimensions of

voice-to-voice service encounters: a scale-development study", International Journal of Service

Industry Management, Vol. 11 Iss: 2, pp.142 – 161

Burgess D.; Borgida E., (1999), Who women are, who women should be: Descriptive and

prescriptive gender stereotyping in sex discrimination., Psychology, Public Policy, and Law, Vol

5(3), pp. 665-692.

Buss, D. M. (2000). The dangerous passion: Why jealousy is as necessary as love and sex. New York, NY: Free Press.

The moderating effect of the fit between an individual’s personal views and the gender role portrayal on attitude towards the ad