Master Thesis
Visual Persuasion in the Online Context
Does visual persuasion on image-centric social media have an effect on consumer attitude
towards the brand when mediated by source credibility and source likeability?
Diana H.P Le - 11138750 University of Amsterdam
Faculty of Economic and Business
MSc. In Business Administration – Marketing Supervisor – Dhr. A. C. J. Meulemans
Table of Contents
1. Introduction ... 6
2. Theoretical framework ... 9
2.1
Social media ... 9
2.2
Visual perception in advertisements ... 11
2.3
Visual Persuasion ... 14
2.4
Source of content ... 18
2.5
Brand personality ... 23
2.6
Consumer attitude towards the brand ... 27
2.7
Conceptual model ... 28
3. Methodology ... 30
3.1
Research design ... 30
3.2
Sample ... 31
3.3
Procedure ... 31
3.4 Measurements ... 34
4. Results ... 39
4.1
Preparatory analyses ... 39
4.2
Hypotheses testing... 45
5. Discussion ... 50
5.1
Findings ... 50
5.2
Limitations and future research ... 54
5.3
Implications ... 55
5.4
Conclusion ... 56
6. References ... 57
7. Appendix ... 63
Appendix 1. Stimuli ... 63
Appendix 2. Thesis Survey Welcome page ... 69
Appendix 3. Reliability ... 76
Appendix 4. Correlation ... 77
List of tables
Table 1: Aaker (1997) and Geuens et al’s (2009) models next to Golberg’s (1992) big five
human personality model.
Table 2. Experimental design conditions
Table 3. Celebrity destriptives
Table 4: covariates descriptives
Table 5. Shapiro-Wilk test for normality.
Table 6. Maughly’s test for sphericity
Table 8. Respondent’s favorite image
Table 9. Results of the hypotheses tests
Abstract
The purpose of this study is to explore the effect of visual persuasion on consumer attitude towards the
brand on the image-centric social media, Instagram. Additionally, this research investigates the
effectiveness of three different types of visual content, brand-generated content versus celebrity-endorser
generated content versus user-generated content. Visual content and its persuasion has been researched in
the context of traditional marketing however with the rise of image-centric social media platforms such as
Tumblr, Pinterest and Instagram , it is highly relevant to analyze this topic in the online environment. This
study uses a quantitative approach to examine the persuasiveness of visual content and how this persuasion
shapes consumer brand attitude by looking at source credibility and source likeability. Findings shows that
in the online context, visual persuasion is not as relevant as in the traditional media. Brand-generated
content is more likely to generate a positive consumer brand attitude and surprisingly user-generated
content does not or hardly influence the consumer attitude. Brand personality moderate the relationship
between visual content and the consumer brand attitude but depending on the brand personality dimension
the relationship is either positive or negative. In conclusion, on image-centric social media, brand
companies might want to post their own content on a profile specific to them as that is most effective.
Keywords: Visual persuasion, Instagram, Brand-generated content, Celebrity-endorsement, user-generated content, source credibility, source likeability, brand personality.
Statement of originality
This document is written by Student Diana H.P. Le 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.
Acknowledgement
First off, I would like to thank my supervisor A.C.J Meulemans for his help with developing an interesting
topic and model from an idea that was in my head. I want to express my gratitude for his guidance and
feedback throughout the difficult process of writing a thesis. Furthermore, I would like to thank Miss
Frederici, who answered my questions about SPSS fast and effectively.
I want to thank my parents and little sister for their support and encouragement throughout my
whole academic career. Their financial and mental support through all this years have been a great help and
has made these years so much easier.
Finally, I want to express my appreciation towards my friend, Sefanja Rubina, whom I only met
this year but has been my thesis partner and moral support through stressful times, even when she was in
1.
Introduction
In this day and age, everywhere you go you see advertisements of brands and products with celebrities,
attractive models and other elements that are visually appealing. There is no escape! With the proliferation
of social media, the availability of visual material has also increased. Since the web 2.0, the 21st century
has bought us many online platforms, which allows the transmission of mass-media internet-based
messages via social media (Mangold and Faulds, 2009; Kaplan and Haenlein, 2010; Berthon, Pitt, Plangger,
Shapirto, 2012). The web 2.0 has created new opportunities for businesses but also for users of the internet.
Applications of web 2.0 enables user to create user-generated content and to easily diffuse them under their
friends, family and even fans or followers (Berthon et al., 2012; Kaplan&nd Haenlein; 2010). More recent,
Berthon et al. (2012) proposed that creative consumers have become the new locus of value in web 2.0. In
other words, besides commercial content, user-generated content, under which are visual content, has
become a great part of the value-added content in social media (Berthon et al., 2012).
As Gurri, Denny and Harms (2010) stated that visual media can be found everywhere. The number
of images created and spread across the world with these new technologies is unbelievable. For example,
20 hours of new video material is uploaded on Youtube every minute (Youtube, 2015). Furthermore,
Youtube has over a billion users who watch hundreds of millions of hours of Youtube videos every day.
Photo sharing sites such as Photobucket and Flickr contain more than 10 billion photos with the latter
website sharing 1 million photos daily (DMR, 2015). However, compared to Facebook these websites still
have a lot to catch up. The total number of Facebook photos uploaded in 2015 is 250 billion with more than
300 million photos being uploaded daily (Zephoria, 2015; Ho, 2015). On Facebook, photos generate the
most engagement between users (Ross, 2014). Besides the previously mentioned websites, there are also
image-centric social media sites such as Tumblr, Pinterest and Instagram. Tumblr currently consists of
276.0 million blogs with 128.2 billion posts (Tumblr, 2015). However, more interesting is the social
platform Instagram that launched in October 2010. This young platform has grown into a community with
medium with over 30 billion photos published and an average rate of 70 million pictures uploaded daily. This content generate 2.5 billion likes (indication of liking a user’s photo) every day. Brands post content to promote their products or services are well-received on Instagram. 70 percent of the Instagram users have
ready looked for a brand on Instagram and 31 percent follow at least one brand account (Iconosquare, 2015).
These statistics indicate that Instagram can be an attractive platform for businesses, if used well. This is
proven with the fact that in 2015, 90 percent of the world top 100 businesses can be found on Instagram
(Simply measured, 2015). These statistics indicate that not only is there more visual material available, but
more people are paying a greater amount of attention to it. Due to the changes in the online atmosphere, we
are literally seeing the world differently from a few years ago. The continuous sharing and the increase of
social media use by brands demonstrate that it is important to understand the effects of these images
uploaded by brands and users on the consumer’s attitude towards the image and the brand, which will be the main topic of this paper is.
With the lightning speed in which this new digital age is developing, 4 years can be 4 decades. It
is important for the research literature on visual persuasion to be updated to include the new modalities and
communication channels used to spread the message. Marketing and advertisement in this day and age is
magnificently different from advertising practices ten years ago. To help managers make marketing
decisions, updated visual persuasion research that applies to the new media world is needed. This research
therefore explores whether visual persuasion has changed in the face of proliferation of image-centric social
media use. Whether the persuasion effects still holds true on these new platforms like Instagram. The
research question for this study is as follows:
Does visual persuasion on image-centric social media have an effect on consumer attitude
towards the brand when mediated by source credibility and source likeability?
The purpose of this study is to research whether visual content on image-centric social media sites positively
effect of brand-generated content, celebrity-endorsed content or user-generated conten.. Images have many
characteristics that can influence this effect, however with the scope and time of this study, it will not be
possible to test all the characteristics. Therefore, this research only tests the potential moderating effect of
brand-personality and mediation effect of source credibility and source likeability on the relationship
between visual persuasion and consumer attitude towards the brand. This is done through a survey that
measured each of the before mentioned constructs.
In practice, this information can be used by managers to create online marketing strategies that are
most effective and catered to specific social media platforms. Theoretically, this study contributes to the
extensive literature stream of visual persuasion by extending the topic to the online atmosphere. Moreover,
previous studies are outdated. This study shines new light on an old topic and examines visual persuasion
in a new but important context.
This paper is structured as follows. The next chapter reviews previous literature on visual
persuasion and visual perception in combination with concepts such as brand personality, and
gender-congruence. Thereafter, chapter two explains the theoretical framework. Additionally, this chapter also
depicts the conceptual framework and explains the variables and moderators. Chapter three will describe
the research design in detail and explain which items and questions were used in the survey. The results of
the survey can be found in chapter 4. Lastly, this paper will comments on the theoretical findings with a
conclusion, discussion and a separate paragraph discussing the managerial and academic implication and
2.
Theoretical framework
2.1
Social media
This paper focusses on the visual persuasion on image-centric social media, therefore this paragraph will
briefly discuss the general notion of social media in relation with marketing and consider the different
image-centric social media platforms.
In the past few years, social media has been an important topic of discussion for businesses. The
proliferation of social media platforms and users forced decision makers to find ways to use these
networking platforms, such as Facebook, Twitter and Instagram, to make profit (Kaplan & Haenlein, 2010).
Kaplan and Haenlein (2010) focus in their article on defining social media using concepts such as web 2.0
and user-generated content (UCG). The authors use these concepts interchangeably as they are
interdependent. Based on these notions they defined social media 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.” (Kaplan & Haenlein, 2010, p. 61). In other words, social media in general means different way in which users connect and interact with each other, that involves creating
word-of-mouth (WOM) and content but also sharing and commenting on this shared content (Kazım Kirtis
& Karahan, 2011). There are a wide range of social media, such as social networking sites (SNS), blogs,
microblogs (Facebook, YouTube, Twitter, Instagram, LinkedIn etcetera), discussion boards, chatrooms,
forums, which are focused on different types of content and interactions (Mangold & Faulds, 2009; Kaplan
& Haenlein, 2010).
Marketing research has studied the different ways social media is used to complement or replace traditional marketing activities (Mangold & Faulds, 2009; Kazım Kirtis & Karahan, 2011; Ashley & Tuten, 2015). Traditional marketing efforts are becoming increasingly ineffective, because consumers are turning
away from the traditional elements of the promotion mix, reducing their reliance on brand-generated
advertisement for information to make buying decisions (Mangold & Faulds, 2009). Unlike traditional
media, social media allows two-way conversations between the user and the brand. It is possible for users
Faulds, 2009). This new form of media focusses interaction and user generated content, original content
created by users without a commercial market context in mind (Kaplan & Haenlein, 2010). Social media
and UGC has become a major factor that influences consumer behavior like information acquisition,
purchase behavior, attitude and opinion, awareness and post-purchase evaluation and communication
(Mangold & Faulds, 2009). The rise of social media has given consumers more power, therefore it is
important for organizations to find ways to interact and engage with consumers to shape what is being
discussed about their brand online.
Photo-centric social media
Following the saying, a picture is worth more than a thousand words, visual social media platforms such as
Pinterest and Instagram are increasing in popularity. While Pinterest and Instagram usage growth has
doubled since 2012, the growth of other text-based platforms like Facebook and Twitter have slowed down.
These photo sharing social media apps capitalize on the ubiquity of mobile technology such as mobile
phones with cameras, microphones and GPS, that make it easy to capture, upload and share pictures
(PewResearchCenter, 2013; Fatani & Suyadnya, 2015). These user-generated pictures will spread through
social media as a form of silent WOM. As the images capture reality and the user experience, these UGC
have a positive tendency to influence and shape brands and their consumers’ attitude (Fatani & Suyadnya, 2015). Pictures are more attention grabbing than text as they are easier to understand and engage consumers
with ease. Moreover, images evoke more emotional feelings and reactions and portray information in a
more compact and efficient way than text (Pollard, 2016). Tumblr, Flickr, Pinterest and Instagram are a few
examples of more popular image-centric social media platforms, however compared to Facebook (a text
and image-based platform), they still have a lot of catching up to do.
In 2010, Instagram, a mobile phone social media application, started off serving as a medium for
online photography and visual storytelling. Currently, the platform has grown into a community with 300
million active users from celebrities, musicians and brands to teens and anyone with a creative mind
promotion, but also to provide information services to consumers. SNS are hubs for consumers to meet a
share ideas, opinions and experience, which influence purchasing decisions. Hence organizations are also utilizing these SNS’s to promote their products and communicate with customers (Fatani & Suyadnya, 2015). The facilities provided by Instagram allow brands to actively promote their brands on Instagram.
About 90 percent of the top 100 fortune brands already active on this platform. Instagram’s principal role
is photography and video sharing, but the platform also provide other features that can help spread the
brands content (hashtag) and generate fast feedback (like) (Fatani & Suyadnya, 2015). In general, Instagram
can enhance a brands online presence and showcases the brands identity through images, which leads to
interactions and engagement with consumers.
It is evident that image-centric social media platforms, especially Instagram, has become
increasingly popular as a tool for businesses to use to promote their products and their brand, however little
scientific research can be found on this topic. The few researches found, discuss the consumer’s motivation to upload and share photos (Sheldon & Bryant, 2016), how the use of Instagram has a potential to ameliorate
loneliness (Pittman & Reich, 2016), and how Instagram can be used to shape tourism destination branding
(Fatani & Suyadnya, 2015).
2.2
Visual perception in advertisements
To be persuaded by an image in any sort of way, one must have some kind of perception and interpretation
of the image. Perception is defined by Krishna (2012) as awareness or understanding of information
acquired through the senses, which the author differentiate from sensation. Sensation is of neurological
nature and is when a stimulus touches the receptor cells of a sensory organ. This paper will not discuss the
psychological aspect of visual perception, instead visual perception in the advertising context is explored.
This paragraph will summarize and examine the existing body of visual perception literature.
Phillips (1997) discusses how consumers interpret complex visual ad images and whether the
interpretations match the intended message of the ad. In general, advertisement are understood by
how the consumer understand and interpret visual advertisement images, McQuerrie and Mick (1996, as cited in Phillips, 1997) came up with a framework by “defining advertising rhetoric as an expression that systematically deviates from expectation through a departure from convention” (Phillips, 1997, p 1.). This means that viewers come to an argument by perceiving what is different from the expected ‘normal’. They indicated that consumers learn how to respond to the deviations by interacting with the advertisement. In
other words, if an advertisement that does not make literal sense, it will be interpret figuratively (Phillips,
1997). For example, when consumers view an advertisement in which puppies are used to sell soft toilet
paper, consumers know not to take this literally. Advertisers do not want to make consumers think that
puppies will make their toilet paper softer and more pleasant to use. Besides the use of figurative
expressions, Phillips (1997) also shortly discusses pictorial metaphor in his paper. A pictorial metaphor is
stating that one picture figuratively has the same meaning as something else. This could be another picture,
object or verbal meaning. For instance, in a Hyundai ad, the advertisers put a car next to a cheetah. The
explanation of the metaphor can be understood as the car can be as fast as or faster than a cheetah. Thus,
consumers must use some cognitive processing as well as abstract thinking to interpret and transfer relevant
properties of one picture to another while dismissing the irrelevant similarities. Phillips (1997) in his article
found that consumers interpret an advertisement image by creating strong as well as weak implicatures. An
implicature is information that is implicitly communicated in an ad to the viewers, e.g. the meaning of a
message (Sperber and Wilson, 1986 as cited in Phillips, 1997). A strong implicature is the main meaning
of a message. Strong implicatures are obvious and cannot be missed. The weak implicatures are the secondary or less obvious meaning of a message. The latter can be intended by the advertisers or a viewer’s individual interpretation of an ad image. Consumers use problem solving methods and relevant stored
knowledge to analyses the message communicated through the ad and to create these implicatures (Sperber
and Wilson, 1986 in Phillips, 1997). The relevant stored knowledge can be cultural, product or advertising
information. Phillips (1997) also found that in general consumer’s interpretations match the intention of the
only with strong implicatures. The inferences or weak implicatures created by the consumers are not
intended (Phillips, 1997).
Imagery has also been explained as useful way to shape consumer attitude towards an ad or product.
Imagery is building a mental image of a concept or a relationship between two object. It was found that
using concrete wording in an ad can generate mental visual imagery which when mediated by vividness of
the mental image has positive linkage to consumer responses (Burns, Biswas, Babin, 1993). Moreover,
creating a visual image in your heads is linked to long-term memory. Thus, visualization whether actually
perceived through visual senses or mentally created can effect consumer’s attitude and intentions.
Visual perception has been approached from different directions within the advertising
environment. Imram (1999) studied the effect of visual perception on consumer’s evaluation of food
quality. The perception of food in affected by sensory factors, such as odour, taste and texture but also by
the price, nutritional content, the health beliefs and labelling. Food in mainly perceived and evaluated on
taste and texture. In his article the author depicted that hence the first encounter with food is usually visual,
vision also contribute to the perception of food products. Appearance tends to influence and modifies the
flavor perception and food acceptability (Imram, 1999). Similarly Hagtvedt and Patrick (2008) proposes
that the infusion of visual art in consumer products influences the perception and eventually the evaluation
of consumer products. They explain that consumer perceive the products to be more luxurious with the
inclusion of art in product packaging.
Other articles research visual biases on a more sensory level. For example, Krishna (2006) showed
that the elongation effect, objects that are taller are perceived to be more voluminous than shorter objects
of equal volume, is a visual bias and can be reversed through other senses, like touch. The expectation is
disconfirmed through haptic inputs which causes the reversal. This occurs because spatial dimensions such
as height and width is better perceived with different senses. Height can be gauged through vision while
width is better judge by touching and handling the object (Krishna, 2006). The author also shows that when
touch. In this case taller containers would appear bigger. In situations in which the visual sense is otherwise
occupied and the container is judged based on touch, wider containers seems bigger.
In this chapter, visual perception is highlighted from different aspect. From here on out we will be
looking at how visual perception persuade consumers to change their attitude.
2.3
Visual Persuasion
Several psychological models have been proposed to theorize the way visual content influences consumer
response and attitude towards the ad, for example, classical conditioning, affect-transfer theory and the
elaboration-likelihood model (Scott, 1994). ELM literature proposes that there are two routes to processing
information that eventually lead to attitude change (Morrison & Vogel, 1997). The ELM explains that the
central route and the peripheral route can be used to process information, which would lead to attitude
formation. In the central route one processes information in a cognitive coherent and reasoned way. In the
peripheral route the information is processed through peripheral cues, also called message cues. (Petty &
Cacioppo, 1983; Morrison & Vogel, 1997). Peripheral processing happens when the receiver evaluate a
message on cues other than the strength of the arguments. This can be based on the credibility of the source
of the message, or even the perceived amount of arguments in a message. These cues are short-cuts for the
receiver to process information when they are unable or not motivated. Peripheral processing is automatic
and unconscious, thus passive compared to cognitive processing (Petty & Cacioppo, 1983).
In previous literature, images within ads with verbal arguments and product attributes are said to
be processed via the peripheral route. Images that show a product are realistic and no not need to be interpret
to be understood by consumers. Visual content are treated as sensory analogous and non-symbolic material
(MacInnis & Price. 1987 in Scott, 1994). However Scott (1994) argues that images are often laden with
symbolic elements. For example, in a message that is carried by an image without verbal text a trope is
present. A trope is a rhetoric argument in figurative form. To be able to interpret such a trope in an image,
consumers must realize that the image is not a realistic representation but a figurative communication.
processed cognitively and not processed peripherally. So in that case, an image needs to be deciphered and
interpret for consumers to understand the message. Under the theory of visual rhetoric, images are created
in such a way that the focus, graphics and layout is arranged so that it represents the intended message. In
other words, the design of a visual ad would be knowledgeable and inferential, which is not possible when
the image is automatically or unconsciously processed (Scott, 1994). “All pictures must be interpreted according to learned patterns – just likes reading words or recognizing numbers – and thus must be processed cognitively rather than merely absorbed” (Scott, 1994, p. 269).
Visual persuasion tactics
In their paper, Gurri, Denny and Harm (2010) suggest that visual images are powerful in their persuasion
because we are unaware of their persuasion effect. We are developed a learning culture that is based on
writing and reading messages in text form. From the old age, people were taught to read words but are
never taught to read images. However, they propose that with the continuously increasing use of
technology, internet and social media, the world is shifting towards a visual-centric landscape (Gurri et al.,
2010). Previous literature tends to discuss the contrasting differences between verbal or written material
and visual material (Johns, 1998). The most important distinction between these two types concerns the
emotional impact. Whereas visuals call upon emotional thoughts, verbal or written content tends to send
people down a rational and more logical track of thought (Joffe, 2008). For example, Boholm (1998) found
that the images used during a certain news reportage generated stronger emotional engagement and stronger
personal concern than the text did, because the images amplify the persuasiveness of the written message
(Seo, Dillard, Shen, 2013). In other words, the visuals evoked a higher level of emotive thoughts in the
audience than did the written content (Boholm, 1998). Boholm (1998) links this effect to images being able
to illustrate risks that usually do not happen in everyday life. Thus making these risks more subjectively
relevant. Moreover, visual images appears to be more memorable, which is associated with the vivid quality
of visual material. Information is considered vivid if it is imagery provoking, emotionally interesting, and
close to the audience in a temporal, spatial or sensory way (Joffe, 2008). The vividness of a visual influences
influencing the pathos, ethos and logos. Playing into the pathos - feelings and emotions - foregoing research
argues that bringing an audience in a certain state of emotion allows persuasion. The persuasive influence
of fear and humor has been researched thoroughly. Humor is found to be effective in advertisement to
influence the audience attitude towards the ad as well as towards the brand (Zhang & Zinkhan, 2006).
However, there is still a debate on the effect of fear in ads. While there is still an ongoing use of fear in ads,
fear-evoking campaigns deemed to be of no success (Joffe, 2008). Other emotions, such as disgust and
love, are found to have persuasion power as well (Joffe, 2008; Till, Stanley, Priluck, 2008).
In the case of love, celebrity endorsement is a good example. By manipulating the ethos - ethics
and credibility - advertisers use celebrities in advertisements. This is proven to be more effective than
unknown actors. As a celebrity has an active image in the mind of the audience and are easily recognizable,
it allows the celebrity endorser to convey a vivid message. Furthermore, as the consumer processes the
information in the ad, the presence of the celebrity is used to make positive inferences about the advertised
product. Celebrity or credible endorsers are used to convince the consumer that the product if trustworthy
and reliable. Combining an endorser’s credibility with emotions, the affect transfer theory explains that the affect felt towards the celebrity is imposed on the advertisement and the endorsed product (McCracken,
1989 as cited in Till et al., 2008). Similar to the affect transfer theory, visual images can be used in classical
conditioning. In this case, the image is used to produce an affect that then is associated and attached to the
brand in an unconscious way (Scott, 1994). Affect elicited from a certain visual stimulus are attached to the
brand during the conditioning. After the conditioning, the affects response are automatically evoked within
the consumer when viewing the brand, even without the image (Till et al., 2008; Scott, 1994).
Lastly, advertisers influence logic by adding rhetoric arguments into an ad or image. Scott (1994)
proposes that visual images can be symbolic artifacts that come forth from a particular culture (Johns,
1998). These visuals are capable of highly rhetorical task through variating the arrangement of the
viewpoint, context, style, but also through the interactions with and the references to other written texts and
in a rhetoric symbolic system with the purpose of persuasion. Rhetorical theory explains that a message can
be used by the sender to influence an audience. To be able to fulfill these rhetorical tasks these visuals must
be capable of representing concepts, actions, modifiers, metaphors, abstractions in such a way that they can be used by the audience to ‘invent’ a complex argument. This invention of an argument consists of positioning, strategy and concept, but also the benefits that are promised to the customer, arguments to
support the promise and the relationship of the brand or product to competitive alternatives (Scott, 1994).
So in other words, the visual image should be capable of clearly translating the organizational positioning,
concept as well as the products benefits and promises to the customer. The visual differs from numbers and
alphabets in basic ways, however images are also capable of transferring messages, making declarations
and comparisons (Scott, 1994). The reliance on analogy between computer and the human thought process
has limited the study of visual symbols.
Online persuasion
Persuasive messages can be delivered in many forms and through variation of modalities and
communication channels. Previous research has examined the persuasion effect of these different
modalities, for example, live speeches, various forms of text, and videos (Johnson, 1994, Kardash &
Scholes, 1995; Lord, Ross & Lepper, 1979 as cited in Murphy, Long, Holleran & Esterly, 2003). Other
research has examined the effectiveness of different communication channels to deliver the persuasive
message. Previous studies propose that live or prerecorded video can enhance the persuasive effect of the
message by positive receiver involvement and positive source credibility (Andreoli & Worchel, 1978 &
Eagly & Chaiken, 1993 as cited in Murphy et al., 2003).
Social networking sites (SNS) such as Facebook, Twitter and Linkedin are one of the most popular
websites in today’s age. They are also among the most persuasive, as for them to succeed, users should take on certain behaviors, like making a profile, connect to friend et cetera. Fogg and Iizawa (2008) use SNS to
study the effect of persuasion online. They used the behavior chain for online participation by Fogg and
involvement and true commitment. All SNS are designed in such a way that they persuade users to make a
true commitment (phase three). The SNS studied in the research (Facebook and Mixi from Japan) persuade consumer to return through email updates, updates of friends’ activities and through pokes and asiato, a form of footprints (Fogg & Iizawa, 2008, p. 3).
Scott (1994) says that images are cognitively processed, because in some way or another they
possess non-verbal arguments and/or an intended message that needs to be interpret by the consumer for
the consumer to understand the meaning behind the image. Moreover brands use different visual persuasion
tactics, such as celebrity endorsements and classical conditioning, in their visual content to positively
influence the consumer’s attitude towards ads and the brand. This suggest that images used by brands are
laden with rhetorical arguments and are intended to send a message to the consumer to persuade them
towards the brand. This study expects that brand-related images, which are more rhetorical are likely to be
more persuasive and therefore influence the consumer attitude towards the brand stronger.
H1. Image-centric social media visual persuasion is positively related to the consumer’s attitude towards the brand.
2.4
Source of content
Visual content can be posted online by all types of users, whether it be brands, celebrities or teens, their
family and friends. This paragraph discusses the different sources of visual content and how they might influence the consumer’s attitude towards brands.
There are different ways to utilize image-centric social media platforms for marketing
communications. On Facebook and Instagram, brands can choose to create their own profile and share their
own content or have their products be endorsed by celebrities and bloggers who use these platforms.
Nowadays, more than 90 percent of the top 100 businesses can be found on Instagram and their
communications are well received by their audience (Simply Measured, 2015). Besides content shared by
brands, the majority of the content that is shared on these channels is user-generated content (UGC). While
Facebook and Instagram to share their own experiences with friends and family. As previously mentioned
in the introduction, approximately 70 million images are shared on Instagram every day, Facebook tops this
with 300 million images being shared every day (Iconosquare, 2015; Ho, 2015). Brands want to leverage
the immense popularity of social media, yet these media have given consumers a podium to get their voice
heard, making them the new focus of content and value creation (Berthon et al., 2012). As Fournier and
Avery (2011) call them, brands are uninvited crashers of the social media party. Web 2.0 has given brands
the opportunity to listen to their consumers and to have a two-way conversations with their customers.
In-depth information about consumer preferences and lifestyles can be gathered for micro-targeting and
personalized messages. However, brands attempts to enter these online social spheres were rebuffed. Ads
posted before videos on YouTube are more than often skipped, promotions on Facebook and Instagram are
ignored. Despite social media providing brands a cheap and easy way to reach their consumers, social media
is created for people and their cultural conversations not for marketers to promote their activities. As brands
share their marketing communications on social media to persuade and convince people to buy their
product, they are perceived as disruptive and inauthentic (Fournier and Avery, 2011). Notwithstanding, it
is still important for brands to create online content to create brand awareness and brand resonance. For this
reason it is important for brands to choose the right social media strategy to come across as a more genuine
and honest brand.
On an image-centric platform, like Instagram, brands and users alike post images of their products,
experiences, special events and things that mean something to them. Brand organizations create and share
images and videos about their brand, these are called brand-generated content. While consumers share
images regarding their social and daily life, they also are share content related to brands. For example, they
upload a photo of their new Nike running shoes or a video of them driving their new Volkswagen car.
Motivations for consumers to contribute to brand-related content are personal identity- and self-related such
as self-presentation, self-assurance and self-expression. Social interaction, social identity and helping and
empowerment are also reasons for consumers to create and share their brand-related content (Muntinga, Moorman and Smit, 2011). Regardless of what the brands’ and consumers’ motivations are, every user of social media are exposed to brand-generated content and user-generated content.
In this paper, another source of content is considered, namely celebrity generated content (CGC).
Web 2.0 has made it easy for consumers to create and share their content, this gave rise to bloggers and
vloggers. Additionally, social media allows these bloggers to reach out and interact with their audience,
creating immense fan bases and communities. Their big following makes them pseudo-celebrities whose
opinions are very influential. Along with these bloggers, actual celebrities and other influential personalities
are using social media as well. More and more celebrities are on Instagram, sharing photos and short videos
of their daily activities to interact with their fans and to show fans a part of their life. These uploads can be
brand-related as they contain products of certain brands they use.
Source credibility
The revolution of web 2.0 facilitated the process of information sharing, leading to large quantities of
information being shared online. The easy accessibility of the information has created opportunities for
learning and personal development, but also made it difficult for consumers to judge the credibility of this
information as the source and the quality of information has become unclear (Metzger & Flanagin, 2013).
According to the source credibility model, the persuasiveness of a message depends on the trustworthiness
and the expertness of the source (McCracken, 1989). A source is trustworthy if it is willing to make valid
affirmations. Expertness is the perceived ability of the source to make these affirmations. Trusted sources
are credible and thus more likely to persuade consumers (McCracken, 1989).
Brand-related posts on Instagram can be divided into three types of content, brand-generated,
user-generated and celebrity endorser-user-generated content. Brand-user-generated content (BGC) is created by the
brand’s organization. User-generated content (UGC) is creative content created by the user for non-commercial purposes (Kaplan & Haenlein, 2010). Lastly, celebrity endorser-generated content can come
the company to recommend and promote a product, service or brand (McCracken, 1989; Lu, Chang, &
Chang, 2014; Sertoglu, Catli, Korkmaz, 2014).
Brand-related content shared in the online social media atmosphere can be considered electronic
word-of-mouth (eWOM) as the content transfers the creator’s opinions and experiences (Kim & Johnson,
2016). Consumers trust eWOM written by their peers more than they trust information provided by the
brands because peer written eWOM is independent of corporation’s commercial intent (Lee & Youn, 2009).
Metzger and Flanagin (2013) suggest that commercial information possess advertising cues, which in turn
can make consumers feel like they are being manipulated. This makes them mistrust the information as
they think it is bias. Moreover, eWOM written by other consumers are experience-based and thus more
useful and influential in providing product information (Lu et al., 2014). This suggest that UGC are
considered to be more trustworthy and credible. A study by Goh, Heng and Lin (2012) indicated that while
both UGC and BCG have a persuasive effect on the consumer, but the persuasive effect of UGC is 22 times
bigger. In the case of celebrity-generated content (CBC), it can be created either out of free will without
commercial intent or the celebrity can be sponsored by the corporation. Previous studies propose that
sponsored CBC must disclose the relation between the celebrity and the sponsoring organization, which
can influence the perceived source credibility. Supposing that these celebrities generate content because
they are payed, it makes the source bias and so less credible (Lu et al., 2014). Nevertheless, Lu et al. (2014)
showed that consumers do not have a negative attitude towards sponsored CBC, even when the content
reveals that the celebrity is sponsored. On Instagram, it is less obvious whether CBC is sponsored or not. It
is often not disclosed if the celebrity is paid, hence the content is considered an honest eWOM of
non-profitable value. The celebrity is perceived as another user genuinely supporting the product (Magnini,
2011).
Based on the findings of the before mentioned authors, it is expected that UGC and CBC
are perceived to be more credible than BGC, while UGC is more credible than CGC. As follows
UGC will have more influence on consumer brand attitude than CBC, which in turn has a bigger
influence on consumer brand attitude than BGC.
H2a. User-generated images have more positive influence on consumer brand attitude than celebrity-endorsed images mediated by source credibility.
H2b. Celebrity-endorsed images have more positive effect on the consumer brand attitude than company-posted images mediated by source credibility.
Source likeability
The source attractiveness model explains that the persuasiveness of a message depends on the level of
familiarity, similarity and likeability of the source (McGuire, 1985 as cited in McCracken, 1989).
Familiarity is having knowledge of the source through exposure and similarity is the extent to which one
thinks the source is similar to oneself. Likeability or attractiveness is defined as attraction and affection for
the source based on the physical appearance, the personality and behavior (McCracken, 1989). In previous
studies, source attractiveness/likeability has been discussed as a component of source credibility (Sertoglu
et al., 2014). This study separates source credibility and source likeability because it is possible for
consumers to think that the source is credible, however finds the source to be unattractive. McGuire (1985
as cited in McCracken, 1989) suggest that messages are more effective the higher the level of likeability of
the source.
UGC are created by fellow Instagram users, who are just everyday people, hence it is easy to relate
to these users and sense a feeling of similarity. Moreover, these users posts about their experience and the
things they like, showing some of their personality, which could lead to likeability of the source. In contrast,
BGC is created by a formal and impersonal entity which makes it more difficult for the consumer to relate
to the brand even when the brand is liked. Previous studies suggest that attractive sources are more liked
and also preferred (Joseph, 1982 as cited in Sertoglu et al., 2014) and they are more likely to influence
consumers associated celebrities with certain meaning, derived from their public persona developed by the
way they are presented in the media. This meaning transfers to the brand they endorse (McCracken, 1989;
Sertoglu et al, 2014). Hence if the endorser is liked, this feeling will be transferred to the endorsed brand
creating positive consumer brand attitude.
Along these lines, this study expects that the source of UGC is perceived to be more similar to the
consumer than the sources of BGC and CGC and therefore more attractive and more likely to influence
consumer brand attitude. Additionally, it is expected that CGC is more likely to influence consumer brand
attitude because a celebrity is more familiar/similar to the consumer as a person than a brand entity.
H3a. User-generated images have more positive influence on consumer brand attitude than celebrity-endorsed images mediated by on source likeability.
H3b. Celebrity-endorsed images have more positive effect on the consumer brand attitude than company-posted images when mediated by on source likeability.
2.5
Brand personality
A brand is defined as “a name, term, sign, symbol, or design, or a combination of them which is intended
to identify the goods and services of one seller or group of sellers and to differentiate them from those of competitors” (Kotler, 1994, p. 442). A brand has a certain brand image, which is the brand associations consumers attach to the brand in their memory (Keller, 1993). Brand image is what consumers make of all
the sign emitted from the brand’s company or from external sources, which include brand personality (Louis and Lombart, 2010). A considerable body of literature has been dedicated to the research into brand
personality. Brand personality is defined by Aaker as “the set of human characteristics associated with a brand” (Aaker, 1997, p. 347). Using different brand personality characteristics managers can create a certain brand image and differentiate their brand from other brands. Aaker’s Big Five model represent one of the most widely used models in the research of brand personality traits typology. The model consist of
five brand personality dimensions: sincerity, excitement, competence, sophisticated and ruggedness (Aaker,
1). Even though Aaker’s big five model is most often used when analyzing brand personality, the proposed model has been criticized. Criticism focusses on the fact that Aaker’s definition and construct for brand personality is too broadly taken, which could lead to other facets of brand image and brand identity being
thrown into the measure. Moreover, the model is criticized for including personality traits – sophisticated
and ruggedness - that has no equivalent human traits (Azoulay & Kapferer, 2003; Geuens, Weijters and De
Wulf, 2009; Ambriose, 2004 as cited in Louis and Lombart, 2010). Additionally, Aaker’s research was
criticized for its non-replicability of the 5 personality dimensions culturally. In Spain, only three of the five
personalities were applicable. Similarly, in Japan four out of the five personalities were found (Aaker,
Benet-Martinez, & Garolera, 2001; Geuens, Weijters & De Wulf, 2009). Therefore Geuenset al. (2009)
created a new brand personality scale leaving out all non-personality items. They use a more rigorous
definition of brand personality created by Azoulay and Kapferer (2003, p 151): “brand personality is the set of human personality traits that are both applicable to and relevant for brands” to make sure that only personality traits are measured to not create construct validity problems. Furthermore, their scale is
appropriate for between brands between-category comparison as well as for between brand within-category comparisons, thus more generalizable. Building on Aaker’s model, Geuens et al. (2009) present a brand personality model based on five major dimensions with their own items (Table 1).
To make sure relevant personality dimensions are chosen to be used in this research, Geuens et al.’s (2009) scale can be interesting to look at in comparison with Aaker’s (1997) scale. Table 1. exhibits the brand personalities identified by the two author’s parallel to Goldberg’s (1992) big five human personality traits. The human personality traits extraversion and agreeableness are identified in both brand personality
models, however conscientiousness and openness were identified in only one of the models. Emotional
stability is missing in both the brand personality models. Both brand personality models established two
additional traits which could not be categorized under one of the big five human personality traits.
The effect of brand personality on consumer responses and consumer attitudes has been researched
For this relationship to be viable, consumers anthropomorphized the brands they interact with. In other
words, consumers designate human characteristics to the brands to make them more humanized, animated
and personalized (Fournier, 1998). Consumers have no difficulty in assigning personalities to inanimate
objects, likewise consumers effortlessly assign personalities to brands (Aaker, 1997). Anthropomorphizing
a brand can happen through transferring associations of spokespersons to the brand. When a spokesperson’s
personality fits with the brand, that personality can be associated and transferred to the brand. Secondly,
emotionality, thought and volition can be transferred to the brand to make it more humanized. This can be
done in the form of brand characters, in which these animated brand characters express their own emotions
and thoughts. Lastly, for the brand to be a human like relationship partner, the brand must be an active and Goldberg (1992) - Big Five
Human personality
Aaker (1997) - Big Five
Brand Personality Geuens et al. (2009)
Extraversion/ Surgency Excitement: daring, spirited, imaginative, up-to-date
Activity: active, dynamic and innovative.
Agreeableness Sincerity: honest, wholesome, cheerful, down-to-earth
Simplicity: ordinary and simple
Conscientiousness Responsibility:
down-to-earth, stable, responsible.
Emotional stability/ Neurotic
Openness/ Intellect Competence: reliable, intelligent, successful
Other dimensions Sophistication: upper class, charming;
Emotionally: romantic and sentimental
Ruggedness: tough, outdoorsy.
Aggressiveness: aggressive, bold
Table 1. Aaker (1997) and Geuens et al.’s (2009) model next to Goldberg’s (1992) big five human personality model.
contributing member of the relation dyad (Fournier, 1998). Consumers infer personalities from the brands
behavior such as the marketing activities and advertisements communicated to the consumers. Thus, the relationship consumers have with a brand depends partly on the brand’s personality and whether these personalities are liked by the consumer. Building on Fournier’s theory of consumer-brand relationship, Kim, Han and Park (2001) analyzed the direct and indirect effect of brand personality on brand loyalty. The
authors showed that consumers who develop brand identification through brand personality, are encouraged
to engage in a long-tern relationship with the brand. Consumers who identify themselves with a brand are
less likely to move away from the brand. Hence, it is important for companies to create distinctive and
attractive brand personalities to differentiate themselves from competitors. The attractiveness of a brand’s personality directly affects positive word-of-mouth (WOM), and it indirectly affects brand loyalty through
WOM (Kim et al., 2001). In a similar way, the paper by Matzler, Strobl, Stokburger-sauer, Bobovnicky and
Bauer (2016) discuss brand personality in the brand-self congruity context, in which self-congruity is the
mediator between brand personality and purchase intention. They found that brand personality by itself
does not affect the behavioral intentions but that it is mediated by whether consumers can associate themselves with the brand. Using Geuens et al’s (2009) brand personality model, they found full mediation for the trait activity, while partial mediation was found for emotionality, simplicity and responsibility.
However no mediation was found for aggressiveness (Matzler et al., 2016).
Drawn from the literature and the theoretical background on brand personality this research
concludes that sincerity and excitement would be the most appropriate personality traits to use to analyze
the moderating effect of brand personality. Literature (Aaker, 1997; Matzler et al., 2016) mentioned that
some traits are more effective than others, in this case literature showed that these two traits are the most
effective. Moreover, in table 1. the two traits are shown to have an equivalent human trait. Additionally, they are also represented in Geuens et al.’s (2009) model. This model is valid and reliable across product categories, brand and cultures, which is important as we are looking at international brands form different
product categories who can have different perceived brand images and brand personalities in different
nations and cultures.
H4a. The relationship between visual persuasion on and the consumer attitude towards the brand is positively moderated by brand personality.
H4b. The relationship between the type of visual persuasion and source credibility is positively moderated by brand personality.
H4c. The relationship the type of between visual persuasion and source likability is positively moderated by brand personality.
2.6
Consumer attitude towards the brand
A consumer’s attitude embodies his evaluation of a certain entity, for instance an advertisement or a brand (Ajzen and Fishbein, 1977). Some work has researched the effect of affective and cognitive antecedents of
attitude and preferences (Fabrigar and Petty, 1999; Zajone and Markus, 1982). Briefly, this attitudinal
evaluation of target objects consists of three components; the affective component, the cognitive component and the behavioral component. The first element is the consumer’s emotional response towards the entity. The second component is the beliefs and the thoughts about the evaluated entity. Lastly, the behavioral component is the consumer’s action towards the object of evaluation (Fabrigar and Petty, 1999). On the general notion, it is believed that a favorable attitude towards an object will lead to a favorable behavioral
action towards the object and thus not an unfavorable action. However, this could be deceptive as many
behavioral actions do not have any evaluative implications towards an object. In the theory of reasoned
action (TRA) by Fishbein and Ajzen (1975 in Armitage & Christian, 2003), the authors indicate that
consumers have a large number of believes about a certain action however only a few are important at a
point in time. TRA supplies that attitudes are determined by these salient underlying beliefs, consisting of
two components: outcome beliefs and outcome evaluations. Outcome beliefs are the beliefs about the likelihood that some particular outcome will occur. Outcome evaluations are the consumer’s evaluation of a particular outcome (Armitage & Christian, 2003). Gardner (1985) showed that ad and brand-related
beliefs are positively related with the consumer attitude towards the brand. Visual content posted on
Instagram can be considered a form of advertisement, it is thus expected that there is some influence of the
content on the consumer brand attitude.
2.7
Conceptual model
In this paragraph the conceptual model is presented. The model is divided into two models to make sure
the hypotheses are clearly depicted and to facilitate the understanding of the conceptual model. The first
figure shows the main relation between visual persuasion and consumer brand attitude moderated by brand
personality. The second figure shows the relationship between the different types of visual persuasion and
consumer attitude towards the brand mediated through source likeability and source credibility and
moderated by brand personality.
Figure 2. The conceptual framework representing the other relations of this research.
The following hypotheses will by tested:
H1. Image-centric social media visual persuasion is positively related to the consumer’s attitude towards
the brand.
H2a. User-generated images have more positive influence on consumer brand attitude than
celebrity-endorsed images mediated by source credibility.
H2b. Celebrity-endorsed images have more positive effect on the consumer brand attitude than
company-posted images mediated by source credibility.
H3a. User-generated images have more positive influence on consumer brand attitude than
celebrity-endorsed images mediated by on source likeability.
H3b. Celebrity-endorsed images have more positive effect on the consumer brand attitude than
company-posted images when mediated by on source likeability.
H4a. The relationship between visual persuasion on and the consumer attitude towards the brand is
H4b. The relationship between the type of visual persuasion and source credibility is positively moderated
by brand personality.
H4c. The relationship the type of between visual persuasion and source likability is positively moderated
by brand personality.
3.
Methodology
This chapter explains the research design and procedure of this study. First, a general research outline is
given by indicating the method used to gather the data. Secondly, this part also specify the sample pool and
the selected brands for this research. Lastly, the measurements of the independent variables and the
dependent variables are evaluated based on their reliability and validity.
3.1
Research design
The design of this study was an experiment in the form of an online survey with experimental stimuli, i.e.
the visual content. Existing visual stimuli were taken from Instagram and visual stimuli simulating posts
from Instagram were developed to be used in the online survey. The online questionnaire consists of three
conditions (Table 2). The first condition measured a base-line for control and comparison. This condition
measured the respondent’s brand attitude without visual persuasion. In this condition the stimulus (Appendix 1) was not presented to the respondent. Only the brand logo was presented. The second condition
presented the stimuli of two brands that were perceived to be exciting. The last condition presented the
stimuli of two brands considered to be sincere. In condition 2 and 3, stimuli representing brand-generated
content, user-generated content and celebrity-generated content were shown for each brand. An online
experiment is chosen to explore the relation between visual persuasion and the consumer attitude towards the brand with the influence of source credibility, source likeability and the brand’s personality. Additionally, this study assess the differential effect of the various types of content on the consumer attitude
towards the brand. A pre-test was conducted to choose brands with the appropriate brand personalities.
As this is a study measuring consumer behavior, it is important to have a relatively large sample size for
reach many respondents in a short amount of time. The negative aspect of an online experiment is that there
is a possibility of error and bias, such as social desirable answered and common method bias that should be
expected in self-reported questionnaires.
Condition 1 Condition 2 Condition 3
Consumer brand attitude towards the four brands
Instagram posts of exciting brands posted by the organization, a fellow Instagram
user and a celebrity
Instagram posts of sincere brands posted by the organization, a fellow Instagram
user and a celebrity Table 2. Experimental design conditions
3.2
Sample
A convenience sampling technique was used to recruit the respondents. Most of the participants are
therefore friends and family of the researcher. This technique allows generation of large amount of data in
a short time period, however the selection of the participants is not random which could result in a less
representative sample pool. In total, 308 respondents participated in the survey. The survey has a dropout
rate of 25.6% resulting in 229 respondents actually completing it and 79 participants dropping out. The
respondents were randomly assigned to one of the conditions. 84 respondents were assigned to the control
condition, 75 people were assigned to the excitement condition and lastly, 70 respondents were assigned to
the sincere condition. 30.1% of the participants were male and 69.9% were female. The participants were
aged 55 year-old and younger. With 60.7% participants falling into the category, 18-24 year-old. The level
of education was relatively high with 67.7% of the respondents obtained a university degree. In condition
2 and condition 3, respondents were asked if they used Instagram. 73.1% indicated that they used Instagram
and 26.9% indicated that they did not.
3.3
Procedure
Pre-test
The pre-test had two objectives; 1. To select two brands that were perceived as exciting and two brands that
were perceived as sincere. 2. To measure the perceived level of excitement and sincerity for the selected
the first part focused on selecting and measuring exciting brand while the second part focused on selecting
and measuring sincere brands. Firstly, to select the brands the respondents were asked to rank a set of 8
pre-provided brands. Subsequently, respondents were asked to rate their top 3 brands on Aaker’s (1997) and Geuens et al.’s (2009) brand personality items on a 5-point Likert scale. The survey was online for a week in which 87 respondents started the survey, however only 43 completed it.
The ranking results showed that Red bull was the most exciting brand. Additionally, Disney was
the second most exciting brand followed by Nike and Coca Cola. For sincere brands, Ben & Jerry’s was the considered the most sincere, followed by Dove and Toms respectively. These brands were then rated
on the items by Aaker (1997) and Geuens et al. (2009). The results shows that Red bull (M= 4.01, SD =
0.79, N= 28), Disney (M= 4.08, SD =0.57, N=26), Nike (M= 4.29, SD =0.51, N=19), Coca Cola (M= 3.66,
SD = 0.71, N= 21) were considered relatively more exciting than other brands that did not make the top 3.
While for sincere brands the level of sincerity of the 8 brands were measured around the same level, Ben
& Jerry’s (M= 3.68, SD =0.57, N= 29), Dove (M= 3.95, SD = 0.54, N=25), Toms (M= 3.96, SD = 0.65, N=17). Therefore the two sincere brands were selected based on their ranking.
Brand selection
Two brands were chosen to represent the Aaker’s (1997) excitement personality dimension and two brands will be chosen to represent Aaker’s (1997) sincerity dimension. The brand selection process started off with an orientation search through google search engine. On first look several examples of brands are mentioned;
for excitement, Virgin and MTV and for sincerity, Dove. After a deeper and more throughout search
through both Google and Instagram, a set of brands were chosen to be tested in the pre-test. After testing a
total of 16 brands for their perceived excitement and sincerity, 4 excitement brands and 3 sincere brands
were taken into consideration. As the research is conducted using images retrieved from Instagram the
brands are selected based on two requirements. First off, the brand has to be active on Instagram. Moreover,
different source types. The following brands were chosen; for excitement, Disney and Red bull and for sincerity, Dove and Ben & Jerry’s.
Stimuli
The visual stimuli consist of a total of 12 images, for each brand 3 images. Instagram is a social media app
used on mobile phones, however the images are in the context of Instagram on a personal computer. When
viewed respondents were asked to imagine the visual stimuli on their personal Instagram time line. For each
brand the respondents were shown an Instagram posts created by the brand, Instagram user and celebrity. If available on Instagram, the stimuli was directly taken from the brand, user or celebrity’s account, however in the case that a stimulus was not available, stimuli was developed to imitate an Instagram posts with the
username and avatar clearly visible. Celebrity-generated content for the four brands was not obtainable on
Instagram, therefore for the purpose of this study, the celebrity-generated stimuli was developed to simulate
Instagram posts. In the photos the brand or the product is visible to make sure that the proper brand is being
rated. The exciting condition consists of images representing Disney posted by the organization, a fellow
user and Demi Lovato, a singer, and images representing Red bull posted by Red bull, another Instagram
user and Neymar da Silva Santos Júnior, a soccer player. The sincere condition presented images
representing Dove posted by the company, a user and Taylor swift and images representing Ben & Jerry
posted by the company, a fellow Instagram user and Elton John. In each image, the username and avatar
clearly indicated each of the sources. The stimuli can be found in the appendix 1.
Pilot test
A pilot test was conducted among 10 peers to validate the survey and check for mistakes. The respondents
pointed out a few mistakes in the survey. Moreover, after going through the data generated by the pre-test,
it became clear that some of the respondents were confused by the counter-indicative items. The mistakes
were fixed and the scales were re-arranged for the online experiment.
Online experiment
Aforementioned, the experiment was conducted through an online survey spread through Facebook with a
