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
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.
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.
3
Table of Contents
1. Introduction ... 4
2. Gender role portrayal in advertising – a literature review ... 9
2.1. Gender stereotypes ... 10
2.3. Media portrayals of gender-roles ... 13
2.4.2 Gender role portrayals in advertising and the society’s views ... 15
2.4.4. Attitude towards the ad and Purchase intent ... 17
3. Conceptual framework ... 19
4. Research design ... 24
4.1. Research design, sample and stimuli ... 24
4.2. Procedure and manipulation... 26
4.3. Strengths and limitations of the research design ... 27
5. Data analysis ... 29
5.1. Validity ... 29
5.2. Analysis... 29
6. Discussion ... 34
7. Limitations and further research ... 37
References ... 39
4
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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.