Master Thesis – MSc BA: Digital Business Track
How do you tell the best Instagram Story?
An experimental study on the effects of picture type in Influencer versus Brand
ephemeral Instagram Stories
Name: Ruby Dinkla Student number: 10555080 Submission date: 20/06/2018
Supervisor: F. Adigüzel
Version: Final Master Thesis
Study: Master Business Administration -
Digital Business Track
Faculty: Faculty of Economics and Business Word count: 14.617
Acknowledgement
I want to express my gratitude to my thesis supervisor, Feray Adigüzel. I want to thank you for guiding me through the process of writing my master thesis with a lot of
enthusiasm, clear feedback, advice and supporting ideas. Your detailed feedback helped me to keep a critical eye on my own work and challenge myself to perform better.
Next, I want to thank the University of Amsterdam for providing additional
methodological workshops to help me with designing my research and the statistical part of my thesis. I appreciate these courses to stay up to date and critical on my research design.
Besides, I want to thank all my respondents who participated in this study’s experiment.
Finally, I would like to thank my family and friends for their support during my thesis. I thank you for keeping me motivated throughout the whole research process.
Statement of originality
This document is written by
Ruby Dinkla
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.
Table of Content
Abstract 6
1. Introduction 7
1.1 Emergence of ephemeral Instagram Stories 7
1.2 Research question & Academic relevance 8
1.3 Managerial relevance 9
1.4 Thesis outline 9
2. Literature review & Conceptual model 10
2.1 Social media marketing & Instagram marketing 10
2.2 Instagram & Ephemeral content 12
2.3 Instagram advertising for fashion brands 12
2.4 Influencer marketing & Advertising source type (firm versus sponsored influencer) 13
2.5 Picture type 14
2.6 Brand attitude 15
2.7 Attitude towards the ad 17
2.8 Source credibility 18
2.9 Hypotheses & Conceptual Model 22
3. Method 23
3.1 Research design 23
3.2 Sampling 23
3.3 Manipulation material & Procedure 24
3.4 Variables 25
3.5 Data cleaning & Hypothesis testing 29
4. Results 30
4.1 Sample definition 30
4.2 Normality check 30
4.3 Reliability test & Validity test 31
4.4 Descriptive analysis 33
4.5 Manipulation checks 34
4.5.1 Source type 34
4.5.2 Picture type 35
4.6 Hypothesis testing 36
4.6.1 Main and interaction effects of content type and source type 36 4.6.2 Moderated mediation effect of self-related mental imagery 39
5. Conclusion 43
5.1 The main effect of picture type and source type 43
5.2 The mediating effect of self-related mental imagery 44
5.3 Scientific implications 45
5.4 Managerial implications 45
6. Limitations and future research 46
6.1 Limitations and solutions for future research 46
6.2 Additional future research possibilities 47
References 49
Appendix 56
Appendix 1: Pre-test procedure and materials 56
Appendix 2: Factor Analysis 64
Appendix 3: Correlation matrix 66
Appendix 4: Two-way ANOVA analysis for the main effect of source type on brand
attitude, ad attitude and source credibility 67
Appendix 5: Two-way ANOVA analysis for the main effect of picture type on brand
attitude, ad attitude 68
Figures & Tables
FiguresFigure 1: Conceptual model
Figure 2: Hayes (2017) moderated mediation model 7
Figure 3: Moderation effect of source type on self-related mental imagery
Tables
Table 1: Overview literature review
Table 2: Experimental conditions and number of respondents per cell in the collected data Table 3: Chi-square analysis for manipulation check of ‘source type’, option A (appendix 1) Table 4: Chi-square analysis for manipulation check of ‘picture type’, option A (appendix 1) Table 5: Chi-square analysis for manipulation check of ‘source type’, option B (appendix 1) Table 6: Chi-square analysis for manipulation check of ‘picture type’, option B (appendix 1) Table 7: Overview variables and measurement scales
Table 8: Normal distribution analysis Table 9: Factor analysis (appendix 2) Table 10: Correlation matrix (appendix 3)
Table 11: One-way ANOVA analysis for the control variables among experimental conditions Table 12: One-way ANOVA analysis & independent t-test for the descriptive statistics (M; SD) per variable among experimental conditions
Table 13: Chi-square analysis for manipulation check of ‘source type’ Table 14: Chi-square analysis for manipulation check of ‘picture type’
Table 15: Two-way ANOVA analysis for the main- and interaction effect of picture type and source type on brand attitude, ad attitude and source credibility
Table 16: Two-way ANOVA analysis for the main effect of source type on brand attitude, ad attitude and source credibility (appendix 4)
Table 17: Two-way ANOVA analysis for the main effect of picture type on brand attitude, ad attitude (appendix 5)
Table 18: Mediation analysis using Hayes PROCESS model 7 Table 19: Moderation analysis using Hayes PROCESS model 7 Table 20: Summary of results per hypothesis
Abstract
The use of ephemeral content as a marketing channel is growing, same as the use of Instagram Stories in marketing strategies. Despite the daily 300 million users of Instagram Stories and 14,5 million sponsored influencer Stories last year alone, little is known about which effects this rising content type has on consumers. As of today, the knowledge about the marketing effectiveness of Instagram as a channel is limited. As a consequence, comparison studies between brands and influencers as source and different content types of the Instagram Story (featuring a person wearing a product versus showing only a product) are lacking. To answer this gap, an online experiment is conducted to investigate whether source type (influencer versus brand) and picture type (product versus person wearing a product)
influences consumer’s brand attitude, ad attitude and source credibility (N=188). In this study, Instagram user’s perceived imagination of using a product, self-related mental imagery, is included as mediating variable. The experiment shows a significant effect of ephemeral picture type on brand attitude mediated by self-related mental imagery and moderated by the source type of the Instagram Story. Thus, Instagram Stories with an influencer as source and content including the picture type which features a person wearing a product leads to higher levels of brand attitude via higher levels of perceived self-related mental imagery than
Instagram Stories which are originated from a brand showing just a product. This is explained by the fact that consumers show increased levels of imagining themselves using a product triggered by influencers as source in combination with portraying a person wearing a product in the content of the Instagram Story. Based on this study, the importance of influencer marketing for today’s Instagram marketing strategy is confirmed. Also the importance of anticipating Instagram marketing content on the imagination of your target audience to activate consumer’s self-related mental imagery to leverage higher brand attitudes is proven.
Keywords: Instagram Stories, ephemeral content, influencer marketing, self-related
1. Introduction
1.1 Emergence of ephemeral Instagram Stories
Over the last years, the trend of ephemeral content
has conquered the social media
landscape. Ephemeral content is a fast-paced message type with a predetermined existence and short term visibility to others (Kerkhofs, 2016). Snapchat introduced ephemeral content to the world in 2011 (Snap.com, 2018). A similar feature is incorporated by other social
mediums, like Instagram.
Instagram is an online, image-based social media platform focused on visual content shared by individual- or corporate users. Instagram is one of today’s fastest growing social media networks reaching over 800 million monthly users in 2017 (Sheldon & Bryant, 2016; Statista, 2017). In 2016, Instagram introduced “Instagram Stories
”. This ephemeral content
feature enables the display of fast-paced content which exists for a predetermined 24 hours and is visible to viewers for a maximum of 15 seconds (Kerkhofs, 2016). Currently, more than 150 million Instagram users use the “Story
” feature on a daily basis and one third of the
viewed Instagram Stories are brand originated (Instagram Business, 2017).
Many brands incorporated social media platforms in their advertising strategy to promote their products and brand. These platforms provide both companies and consumers the possibility to share content about brands and their products (Mangold & Faulds, 2009; Kaplan & Haenlein, 2010). This creates additional promotion opportunities for companies as they can collaborate with third party endorsers to promote for them, known as influencer marketing (De Veirman, Cauberghe & Hudders, 2017). So, companies can use both owned content, firm-generated content (FGC), and (sponsored) consumer-generated content (CGC) of influencers as advertising channel. Last year, around 14,5 million Instagram Stories were posted by influencers (van de Ketterij, 2018).
The visual character of Instagram makes the platform suitable for advertising, branding and engagement purposes (Kietzmann, Hermkens, McCarthy & Silvestre, 2011; Killian & McManus, 2015). According to Smith (n.d.) Instagram is the king of engagement marketing tools because of its focus on visual content. Brands can engage with consumers as Instagram provides endless possibilities of picture types to create branded imagery. For example, brands can solely highlight the advertised products or create content that speaks to
consumer’s imagination by portraying persons using the advertised products. In this case, brands build on basic human motives of self-verification related to mental imagery and self reflection theories (Aydınoğlu & Cian, 2014). The questions is: Which content combination is
the most effective for Instagram Stories?
1.2 Research question & Academic relevance
Despite the existence of Instagram for 8 years, research regarding the marketing effectiveness of the platform is limited (Instagram Business, 2017). Previous research has focused on Instagram posts in relation to concepts of influencer marketing, sponsorship disclosure, brand attitude, uses and gratifications theory, ad elements and purchase intentions (Pieters & Wedel, 2004; Kim & Ko, 2012; Lee & Sin, 2016; De Veirman et al., 2017;
Djafarova & Rushworth, 2017). However, the focus has only been on Instagram posts for either FCG or (sponsored) CGC. Little research is performed to compare the effects of branded Instagram content versus sponsored influencer content. Campbell, Mohr & Verlegh (2013) show that brand attitude becomes less favorable for sponsored posts and research by Lawrence, Fournier and Brunel (2013) shows a higher brand attitude for sponsored content compared to traditional ads. But, these studies compared content from the same source. It is unclear if the outcomes are similar when sponsored posts are compared to brand posts.
Brands are increasingly using branded imagery similar to CGC fading the line between commercial and non-commercial content. However, little research is done to investigate the effectiveness of various content types. Studies on the effects of picture types (product versus person
) in a traditional media context show that consumer’s brand attitude
increases for advertisements including a person according to the occurrence of the underlying mental process “self-related mental imagery
” (Aydınoğlu and Cian, 2014). However, if the
same outcomes are valid in the context of Instagram Stories has not been researched.
Since social media research examining Instagram Stories as the medium has not been initiated yet, this study focuses on multiple literature gaps to expand the theoretical field of Instagram marketing. Taking its character into account, this research focuses on brand engagement aspects of Instagram Stories. The effectiveness of source type (brand versus
influencer
) and picture type (product versus person) is studied on the consumer’s brand
attitude, ad attitude and source credibility via an online experiment. Taking the influence of consumer’s visualization of the self with a product into account, the concept self-related
mental imagery has a mediating role in this study. In conclusion, the central research question and sub questions are:
RQ: What is the most effective type of Instagram Stories generated by influencer
versus brand on brand attitude and why?
SQ1: What is Instagram social media marketing and influencer marketing?
SQ2: How do different types of Instagram Stories (brand versus influencer) and picture
types (product versus person wearing a product) affect consumer brand attitude, ad attitude and source credibility?
SQ3: How is this effect mediated by self-related mental imagery?
1.3 Managerial relevance
Research on the effectiveness of Instagram Stories is valuable for both theory and practice. Besides the above stated theoretical additions, this research also provides new managerial insights into Instagram marketing. First of all, influencer marketing and Instagram advertising is a timely and relevant topic for marketers and brand managers. Nowadays, most companies use Instagram to be present at the marketplace. But, they have no specific strategy for Instagram. This study provides (fashion) brands with useful insights about how consumers perceive their (influencer) Instagram Stories marketing efforts. The study enables brands to gain additional knowledge on consumer insights and marketing effectiveness which can be used to optimize their marketing efforts and strategy.
1.4 Thesis outline
This study is focused on (1) what is Instagram social media marketing and influencer marketing, (2) how do different picture types in Instagram Stories (product versus person wearing a product) affect consumer brand attitude, (3) how does the source type moderates the effect of Instagram Stories types on brand attitude, ad attitude and source credibility and (4) how is this effect mediated self-related mental imagery? To answer these questions, this study has the following structure: a literature review on previous research, explanation of the research design and method, the research results, main conclusions, limitations and
2. Literature review & Conceptual model
The following section clarifies the constructs and variables presented in this study. The conceptual background is based on a review of existing academic literature about the
researched concepts. The section ends with the conceptual model of this study which illustrates the hypothesis and tested relationships between the variables.
2.1 Social media marketing & Instagram marketing
Social media marketing can be defined as the use of social media technologies and channels to create, communicate, deliver and exchange value offerings for an organization (Tuten & Solomon, 2017). As consumers are increasingly present on social media, brand simultaneously shifted their brand presence and communication towards the online
environment activating a revolution in brand communication and marketing. The popularity of social media continues to grow and the adoption of these new channels into corporate marketing strategies grows simultaneously (Killian & McManus, 2015). Over the past years, social media became the most popular channel through which brands attempt to strengthen the relationship between customers and their brand, products and company (Jaakonmäki, Müller & Brocke, 2017). Social media rapidly replaced traditional media and enable companies with unlimited new opportunities to advertise, build personal relationships and increase
engagement with consumers (Bruhn, Schoenmueller & Schäfer, 2012; Godey, Manthiou, Pederzoli, Rokka, Aiello, Donvito & Singh, 2016).
In addition, the era of the active consumer who is able to generate, create and share content of its own, has changed the marketing ecosystem into an interactive digital media landscape (Hanna, Rohm & Crittenden, 2011). Malthouse, Haenlein, Skiera, Wege and Zhang (2013) support this vision and state that social media has highly empowered consumers in such a way that they rule on social media.
Today’s social media landscape exists of a rich and diverse ecology of social media sites varying in scope and functionality (Kietzmann, Hermkens, McCarthy and Silvestre, 2011). This study is focused on Instagram, a social media platform concentrated on sharing visual content. According to Sheldon & Bryant (2016) Instagram is one of the fastest growing social network sites on a global scale. Instagram is focused on an “image first, text second”
formula creating a strong visual-oriented culture. The creative possibilities Instagram provides, enables brands to engage with their consumers on a higher level than other social media. Instagram demonstrates that pictures can speak louder than words as visual
communication through images can capture attention and enables more effective ways to communicate than purely textual communication (Lee, Lee, Moon & Sung, 2015; Lee & Sin, 2016). The authors claim that visual elements in communication are powerful stimuli that can affect consumers on an emotional and cognitive level.
The engaging character of Instagram is confirmed by an uses and gratification analysis. The uses and gratification theory (UGT) is a theoretical framework used to explain why individuals seek and choose different media to fulfill their needs and wants (Katz, Blumer & Gurevitch, 1974). The gratifications users receive via the use of their selected media, in turns, leads to the satisfaction of an array of needs (Katz et al, 1974; Whiting & Williams, 2013). According to Katz et al (1974) consumers are goal-oriented in their (social) media behavior and actively interpret media messages, including advertisements, to achieve the optimal levels of gratifications for their needs. Previous UGT research on Instagram, and comparable platforms like Pinterest and Snapchat, indicate that these image-sharing social platforms evolve around users motivated by: engaging in brand communities, connecting with people who share similar interests and needs, and following fashion (Lee et al., 2015; Phua, Kim & Jin, 2016). Instagram is used for the purpose of passing time, entertainment, and serves the purpose of escaping from everyday life routines (Phua, Kim and Jin, 2016).
Commercial brand use Instagram as part of their marketing mix to advertise, raise brand awareness, drive engagement and increase the conversation about brand and products (Ashley & Tuten, 2015; Lee et al., 2015; Phua et al., 2016). The gratifications associated with social media can be divided into two categories: utilitarian gratifications focused on
functional and cognitive aspects and hedonic gratifications centered around the emotional and affective aspects of social media usage (Xu, Ryan, Prybutok & Wen, 2012). The hedonic motivations demonstrated by the UGT of Instagram are linked to experiences, fashion, social entertainment and achieving entertainment (Xu et al., 2012; Pöyry, Parvinen, & Malmivaara, 2013). These UGT motivations and the hedonic character of Instagram are important to keep in mind as it influences the perceivement and effectiveness of advertisements.
2.2 Instagram & Ephemeral Content
This study focuses on a specific type of visual communication: ephemeral visual content. Ephemeral content is introduced by Snapchat, quickly followed by Facebook and Instagram. Instagram introduced this feature with “Instagram Stories” where users can share temporary content on their (corporate) accounts and ephemeral content is in the center of the user experience (Bayer, Ellison, Schoenebeck & Falk, 2016). Ephemeral content has a predetermined lifespan of availability and is perceived as highly engaging by consumers (Wakefield & Wakefield, 2018). This message type leverages a more personal connection and experience between brand and consumer, and because of the temporary character, consumers perceive a “fear of missing out” (Shechter, n.d.; Trounce, 2018). This is linked to the sense of urgency, scarcity and exclusivity as the content disappears in a short time giving the message a sense of privilege (Beckett, 2017). In addition, the ability to quickly release content
leverages brand authenticity and provides the opportunity to continuously engage with consumers. According to Anderson (2015) brands recognize the marketing possibilities of ephemeral content, especially as a go-to place for millennial marketing. However, while brands engage in social media marketing strategies and efforts, little academic marketing research is performed to investigate the effects on consumer attitudes, behavior and the adjunctive underlying processes (Hudson, Huang, Roth & Madden, 2016).
2.3 Instagram advertising for fashion brands
Social media can accelerate the connection between brand content and consumers, especially for the retail industry (Kim & Johnson, 2016). The visual character, possibility to connect with consumers on a personal level, and opportunities to create a personal experience for consumers makes Instagram a suitable platform, specifically for the retail industry. Both the fashion and beauty industry are the two most engaging categories on Instagram putting influencer marketing at the center of their marketing strategy (McCulloch, 2018). As the majority of the most followed brands on Instagram fall into the fashion industry, this sector has the focus of this study (Statista, 2017). Previous studies on UGT show that “fashion following” is one of the top usage motivations for Instagram and other image-based platforms (Phua et al., 2016; Sheldon & Bryant, 2016). This is linked to the hedonic motivations users
have for using Instagram, as they seek to meet their need of entertainment (Xu et al., 2012; Phua et al., 2016). In the scarce field of academic research on Instagram, the marketing effectiveness of Instagram Stories in a fashion context has not been researched yet. This study is focused on the fast fashion industry because its leading industry brands Zara and H&M score high in the overview of most followed Instagram brands by Statista (2017). Besides, fast fashion brands are consumer centered, active on social media like Instagram and their
constant renewing collections makes attracting attention and engagement on social media crucial (Barnes & Greenwood, 2006).
2.4 Influencer marketing & Advertising source type (firm versus sponsored influencer) The interactive character of social media changes the marketing landscape and the nature of information sources (Hanna, Rohm & Crittenden, 2011). According to Bruhn et al. (2012) are consumers empowered to communicate to other consumers making companies no longer the sole source of brand communication. The emergence of this phenomenon is referred to as consumer-generated content (CGC). CGC is described as: “a variety of new sources of online information that is created, initiated, circulated and used by consumers’ intent on educating each other about products, brands, services, personalities, and issues’’ (Mangold & Faulds, 2009, p. 357).
This development leads to the categorization of two distinct social media content types: firm-generated content (FGC) and consumer-generated content (CGC). Companies create FGC on their branded Instagram pages such as Instagram posts and Instagram Stories which feature their brand and products. Over the past years, brands discovered the potential of collaborating with social media influencers to promote their products or brands. Social media influencers are defined as “social media users with a sizable social network of others
following them for being a trusted tastemaker” (De Veirman, Cauberghe & Hudders, 2017, p. 798). Social influencers produce sponsored CGC on social media. On Instagram, this is expressed by Instagram posts and Instagram stories promoting the products and brand of the companies they collaborate with. According to De Veirman et al. (2017) around 75% of the marketers today use influencer marketing. Consumer influencers have the ability to bring people together who share common interests, for example in fashion and particular brands, and create a high level of brand equity (Booth & Matic, 2010). Bruhn et al. (2012) state that
social media are in general regarded as trustworthy information source and consumers trust the evaluation of products and brands by other consumers.
The two content types, FGC and CGC, are ways to communicate to consumers from two distinct source types. As FGC is categorized as traditional marketing content from brand to the consumer, CGC used in a marketing strategy can be categorized as “electronic word of mouth (eWOM)” (Phua et al., 2016). Research shows that content spread from consumer to consumer via “word of mouth” has more impact to leverage brand preference and brand attitudes than the content distributed by the brand itself (Booth & Matic, 2010). The authors claim that this effect is even more powerful in a social media setting. According to Booth & Matic (2010) content posted on corporate brand social media accounts have a far lower possibility of going viral than the same content being spread via the account of an influencer. With the popularity of Instagram, brands increasingly aim to incorporate influencer marketing with influencers who are perceived as trustworthy to advertise their products and brands on social media via sponsored advertisements (De Veirman et al., 2017). However, the academic evidence related to Instagram and the marketing effectiveness of FGC compared to sponsored CGC is still limited (Djafarova & Rushworth, 2017).
2.5 Picture type
Images are a prominent means of today’s online communication and in fact,
photography is turning into a social medium in its own right, being the raison d’etre for social media platforms, like Instagram (Kelsey & Stimson, 2008; Colliander & Marder, 2010; Bakhshi, Shamma & Gilbert, 2014). The most common type of content shared on social media are images of people as human faces in Instagram content leads to higher engagement levels (Bakhshi et al., 2014). The authors mention that previous studies towards offline behavior also result in human faces being powerful channels of nonverbal communication. The creative character of Instagram enables brands and sponsored influencers to portray various picture types on their accounts to interact with consumers (Colliander & Marder, 2010). For example, brands and influencers can post pictures of products, behind the scenes pictures or people using the advertised products.
Platforms like Instagram have the premise of being social, rather than commercial (Colliander & Marder, 2010). Therefore, companies and brands increasingly replace their professional studio photos for visual content that matches the aesthetic of CGC. The authors
state that many fashion brands, like Alexander Wang, ASOS and Topshop, currently use a more user-generated aesthetic in their imagery. This content is defined as “snapshot aesthetic” (Colliander & Marder, 2010). As the snapshot aesthetic is more congruent and meaningful in a social media setting, users should be able to process it more fluently, leading to more favorable responses to the advertised image.
Besides the aesthetics of pictorial content, the context of advertisements also influences consumer responses assessed from the perspective of the underlying mental processes. According to Aydınoğlu & Cian (2014) advertisements can depict two context types: either a visual representation of the product being advertised or a person who is a presumed user of the product to illustrate the potential use of the product. According to the authors, advertising effectiveness is reached if the ad enables consumers to relate to some aspects of the product or advertisement. Self referencing, the process of relating external stimuli to yourself, introduces additional persuasion advantages of an advertisement. Due to the vivid nature of pictorial content, they influence consumers by the imagery value
(Aydınoğlu & Cian, 2014). Previous studies have shown that imagery is an important aspect to influence attitude towards an advertisement or product (Burns, Biswas & Babin, 1993).
Building on the concept of self-referencing, this study focuses on the activation of self-related mental imagery through the use of visual advertising content in the Instagram Story context. Self-related mental imagery can be defined as: “consumer’s visualization of the self in the product purchase, trial or usage situations” (Aydınoğlu & Cian, 2014, p. 507-508). The results demonstrate that activating consumer’s visualization of the self in the product increases positive consumer responses towards the brand and advertisement message itself.
2.6 Brand attitude
Social media communication is a major factor in influencing various aspects of consumer behavior, including their attitudes towards the brand (Mangold & Faulds, 2009; Bruhn et al., 2012). Naylor, Lamberton and West (2012) disclose that social media are the right platform to find information about brands to develop preferences. Those preferences are related to the formation of attitudes. Brand attitude is defined as “an individual’s internal evaluation of the brand” (Spears & Singh, 2012, p. 55). According to Mitchell and Olson (1981) attitudes are directed at an object, in this case the brand, and are evaluative in nature demonstrating the imputation of a degree of (un)likeability of a brand as well as the extent to
which a consumer has an (un)favorable view of it. Brand attitude is an important objective of marketing and advertising activities and both scholars and practitioners use brand attitude to assess marketing effectiveness (Mitchell & Olson, 1981; Zarantonello & Schmitt, 2013). The authors state that in order to be effective, advertising messages must be able to influence consumers’ attitudes towards the brand presented in the ad.
According to the two-step flow theory specific individuals are recognized as opinion leaders who pass along content to others increasing the influence of the information
transmission (Katz & Lazarsfeld, 1955; Uzunoğlu & Kip, 2014). The theory proposes that mass communication by brands may not always directly affect the public. However, opinion leaders mediating in the communication flow leads to more effectiveness on the public by their intermediary factor in establishing social interactions. Influencers on Instagram can be regarded as digital influencers similar to the offline opinion leaders of the original two-step flow theory (Uzunoğlu & Kip, 2014). The influencers communicate messages and positively affect consumers. According to Uzunoğlu & Kip (2014) it is essential for brands to engage with online influencers in order to obtain a trustworthy, authentic online presence leading towards positive advertising outcomes.
This study focuses on addressing the literature gap on the effect of both FGC and CGC on brand attitude. Research by Campbell, Mohr and Verlegh (2013) found that sponsored content has a negative effect on consumer’s attitude towards the brand. Yet, Lawrence,
Fournier and Brunel (2013) found that advertisements originated from consumers resulted in a lower level of negative effect on attitude towards the brand in comparison to traditional
advertisements by brands. Thompson and Malaviya (2013) found that consumers perceive sponsored content by fellow consumers less persuading than branded content. Previous studies on the effectiveness of (celebrity) endorsers on consumer responses show that advertisements with (celebrity) endorsers as source lead to positive consumer responses towards the advertisement and product (Phua, Lin & Lim, 2018). Based on these previous research outcomes, it is hypothesized that Instagram Story ads including the influencer as source will lead to a higher brand attitude compared to the brand being the source.
According to Colliander & Marder (2010) consumers have the presumption that social media are non-commercial media platforms. As professional quality studio photos by brands are perceived as advertisement, brands started to transform their FGC using the “snapshot aesthetic” to yield positive responses. Research by Colliander & Marder (2010) shows that
images with snapshot aesthetic result in a more positive brand attitude. This outcome is supported by Aydınoğlu and Cian (2014) as their research illustrates that advertisements containing a product picture type leads to lower levels of self-related mental imagery and negatively influences brand attitude. This mental process is confirmed by Phua and colleagues (2018) in a celebrity endorser setting. Based on these previous findings, it is expected that Instagram Story ads with the picture type “person wearing a product” yields a higher effect on brand attitude. The same expectation on brand attitude is made for the influencer being the source type compared to the brand as source in combination with picture type “person wearing a product”.
2.7 Attitude towards the ad
According to MacKenzie, Lutz and Belch (1986, p.130) consumer attitude towards the ad can be defined as: “predisposition to respond in a favorable or unfavorable manner to a particular advertising stimulus during a particular exposure situation”. The attitude consumers hold towards the advertising message is seen as important factor to determine the ad’s
effectiveness. Biehal, Stephens and Curio (1992) state that a consumer’s attitude towards the ad significant positively influences on the overall attitude towards the brand. Hutter, Hautz, Dennhardt and Fuller (2013) researched the fit between the content of social media
advertising messages and the evaluation of those advertisements. The authors conclude that a positive ad attitude results in a more positive brand attitude. In this research setting, the Instagram Story portrays as the advertisement whereas attitude towards the ad is defined as the consumer responses towards the Instagram Story.
Attitude towards the ad is determined by various factors: content, creativity of the medium, source, ad images or pictures which communicate much about the brand being advertised (Biehal et al., 1992). This research context consists of two different picture types (product versus person wearing a product) and two various source types (brand versus influencer). Those picture- and source types could have deviating consumer responses towards the Instagram Story ad.
According to Lee & Koo (2012), whenever consumers feel the message is biased or persuasive, their beliefs on the source change, which influences their attitude towards the advertisement. Research of Lu, Chang & Chang (2014) concludes that exposure to sponsored
CGC content does not result in changed consumer attitude towards the advertisement. However, such effects for attitudes towards FGC advertisements have not been found.
Colliander & Marder (2010) demonstrate that advertising content which resembles the aesthetic of CGC results in positive consumer evaluations. The positive effects are a result of the content being perceived as honest and authentic rather than purely commercial. Besides, the role of self-related mental imagery on advertising effectiveness has been researched in a traditional media context (Gavilan, Avello & Abril, 2014). Here, positive effects of mental imagery on attitude towards the brand and ad attitude are concluded in the context of print advertisement (Burns, Biswas & Babin, 1993; Babin & Burns, 1997; Gavilan et al., 2014).
As the picture type “person wearing a product” resembles the snapshot aesthetic and shows results of higher ad attitude, it is expected that Instagram Stories including this picture type leads to a higher ad attitude than the picture type “product”. According to the current literature, it is expected that a positive ad attitude towards the Instagram Story is related to attitude towards the brand. Based on these previous findings, ad attitude is incorporated in the study. Since previous findings prove that ad attitude mediates the effect between
advertisement and brand attitude, ad attitude is not investigated as mediating effect in this study (MacKenzie et al., 1986; Biehal et al., 1992; Hutter et al., 2013).
2.8 Source credibility
An UGT analysis of Instagram concludes that the hedonic use of the platform leverages a peripheral mindset towards persuasion attempts. The Elaboration Likelihood Model (ELM) theory proposes that individuals experience two routes to persuasion: the central route versus the peripheral route. The peripheral route consists of an emotional, affective and heuristic route to persuasion. In this state, consumers tend to have a low level of elaboration where consumers place content under an ephemeral judgment grounded in
simplistic, visual and intuitive cues (Meyers-Levy & Malaviya, 1999; Petty & Cacioppo, 1986). In this state, messages are evaluated on cues rather than the strength of arguments, for example the credibility of the source (Petty & Cacioppo, 1983). This can lead to a lower level of critical advertisement assessment making users open to the use of the two-step flow
persuasion strategy.
The effectiveness of different advertising sources has been extensively researched for source credibility. Source credibility refers to the consumer’s perception of the
communicating source in terms of attractiveness, trustworthiness and knowledge in the area of the endorsed product (Ohanian, 1990). A positive evaluation of the source increases the acceptance of the message among recipients and enhances the value of the communicated message. Credible sources generally lead to enhanced persuasiveness of an advertising message and higher immediate attitude changes (Li & Zhan, 2011).
The rise of social media enables companies to advertise through different sources. In this study both FGC and the corresponding brand source and CGC by sponsored influencers propose the central source types which influence the way advertising messages are perceived by consumers. In the past, studies examined whether brands can benefit from sponsored CGC. This is the case whenever consumers evaluate content by non-brand sources to be more credible and authentic (Gensler et al., 2013). Consumers develop a positive attitude towards the brand being endorsed whenever the information is originated from a credible source (Luo, Luo, Schatzberg & Sia, 2013; Djafarova & Rushworth, 2017). As sponsored CGC by
influencers is categorized as eWOM, consumers trust the content provided by peers more than the information shared by brands because peers are evaluated as non-commercial and
independent of brands (Lee & Youn, 2009; Hautz, Füller, Hutter, & Thürridl, 2014). Yet, Kumar, Bezawada, Rishika, Janakiraman & Kannan (2016) found that FGC spread via company social media sources positively affect the attitudes of consumers.
Colliander and Marder (2010) emphasize that brands increasingly use a “snapshot aesthetic” to make their FGC similar to (sponsored) CGC as it leads to a more positive evaluation of the message. Using snapshot aesthetic results in higher perceived credibility of the brand’s and influencer’s Instagram account as it shows their expertise on the uses of the platform. Aydınoğlu and Cian (2014) support this outcome as the picture type “person wearing a product” leads to higher self-related mental imagery which humanizes the brand and source. Therefore, it is expected that sources using this picture type yields positive outcomes, in this study, related to a higher source credibility.
Concluding, the existing literature on this topic shows contradicting results. But, taking these results into account, it is anticipated that Instagram Stories originated from sponsored influencers have a more positive effect on brand attitude and source credibility than Instagram Stories originated from the brand itself.
Table 1: Overview literature review
Study Context Findings Method Contribution
Djafarova & Rushworth
(2017)
Instagram Celebrities influence the purchase behavior of young female users. Bloggers and Instafamous endorsers are perceived as more credible than celebrity endorsers.
Qualitative research
Investigating the credibility of the Instagram profiles of various online celebrity types on purchase
decisions of young female Instagram users.
Colliander & Marder (2010)
Instagram Ads with snapshot aesthetic content result in higher perceived source credibility than ads with traditional studio aesthetic.
Instagram ads with snapshot aesthetic results in a more positive brand attitude compared to ads with traditional studio aesthetic.
Experiment Examining the effects of snapshot aesthetic pictures versus traditional studio aesthetic social media advertisements on brand attitude and source credibility.
Aydınoğlu & Cian (2014)
Magazine ads
Creating ads which take consumer self-evaluations into account can result in more favorable attitudes towards the ad through a higher level self-related mental imagery. Ads portraying persons generate higher levels of self-related mental imagery than ads only showing a product mediating the effect on attitude towards the ad.
Experiment Providing additional evidence to existing marketing theories
demonstrating that self-relevance in commercial imagery can be achieved through picture use (portraying persons versus products) in advertising. Burn et al. (1993); Babin & Burns (1997)
Print ads Visual imagery mediates the effects of print ads on attitude towards the ad, attitude towards the brand and purchase intentions.
Consumer’s attitude towards the ad and brand attitude is mediated by the elaboration of imagery processing.
Experiment Investigating the effects of different pictorial advertisement content on consumer responses such as attitude towards the ad, brand attitude and purchase intentions.
Campbell et al. (2013)
Television product placement
Covert marketing increases brand recall and brand attitude.
Sponsorship disclosure can evoke the use of persuasion knowledge to correct evaluations.
Experiment Gain additional insights on the effect of covert marketing on consumer brand recall and brand attitude.
Lawrence et al. (2013)
YouTube campaigns
CGA are perceived as higher quality than traditional ads.
CGA lead to higher engagement levels than traditional ads. Consumers do not seem to mind if CGA is for personal gain, only mind whether the source is a consumer or brand.
Multimethod research
Investigating the differentiation effects of consumer-generated ads (CGA) on consumer engagement and trustworthiness.
Thompson & Malaviya
(2013)
Video ads Attributing the consumer as ad source increases persuasion if the audience has limited cognitive resources to scrutinize the message; is given background information on the ad creator; and has high loyalty towards the brand.
Experiment Studying the effect of awareness of advertising co-creation of
consumers on persuasion effects.
Phua et al. (2018)
Instagram ads
The level of congruence between endorser and product results in higher levels of positive attitude towards the ad, brand, product and usage intentions.
Experiment Investigate the existing knowledge on consumer engagement and brand responses towards social media based advertisements using celebrity endorsers.
MacKenzie et al. (1986)
Television ads
Attitude towards the ad exerts strong positive influence on brand attitude. Ad attitude positively influences ad-related cognitive responses. Peripheral route ad processing positively influences the effect of attitude towards the ad on brand attitude.
Experiment Investigate the relationship between attitude towards the ad in consumer responses and
confirming the relationship between ad attitude and attitude towards the brand.
Biehal et al. (1992)
Print ads Attitude towards the ad has a direct, positive effect on brand choice and brand attitude.
Attitude towards the ad has an indirect effect via brand attitude.
Experiment Examining how attitude towards the ad influences brand choice in a print advertisement setting highlighting running shoe ads. Hutter et al. (2013) Facebook; case car brand MINI
Engagement with Facebook fan page has positive effect on consumer brand awareness, WOM activity and purchase intention.
Online survey Analyzing the influence of social media activities on purchase intention, brand awareness and WOM activities. Lu et al. (2014) Sponsored blogs posts
Consumer’s attitude towards the advertisements remain the same whether the sponsored consumer receives (in)direct monetary benefits.
Experiment Examining the effect of sponsored recommending blogs on consumers attitudes towards the advertisement, blogger and advertised product. Luo et al.
(2013)
Online forum
Source credibility negatively moderates the effect of
recommendation credibility and recommendation adoption.
Secondary data analysis
Investigating which moderators have an effect on source credibility of online recommendation on recommendation credibility and adoption.
Hautz et al. 2014)
Video ads UCG results in a higher level of source credibility than FGC.
Experiment Examining consumer responses on source credibility of video ads and the mediation effect of the video content quality.
2.9 Hypotheses & Conceptual Model
Based on the literature review above, four hypotheses are formulated for this study. Since previous studies have proved the mediating effect of attitude towards the ad and source credibility between ad content and brand attitude, a separate hypothesis on the mediating effect of these variables is not included in this study, only for the mediating variable “self-related mental imagery”.
H1: Instagram Stories advertisements with source type “brand” will result in (a) lower
source credibility, (b) lower ad attitude, and (c) lower brand attitude compared to source type “influencer”.
H2: Instagram Story advertisements with content type “person wearing a product” will
result in (a) higher brand attitude, and (b) higher ad attitude compared to Instagram Story ad with picture type “product”.
H3: Instagram Stories advertisement with picture type “person wearing a product” and
source type “influencer” will result in (a) higher brand attitude, (b) higher ad attitude, and (c) higher source credibility compared to Instagram Story ad with picture type “person wearing a product” and source type “brand”.
H4: Instagram Story advertisement with picture type “person wearing a product” and
source type “influencer” will result in (a) higher self-related imagery resulting in (b) higher brand attitude compared to Instagram Story ad with picture type “person wearing a product” and source type “brand”.
3. Method
3.1 Research design
To answer the central research question a survey based online experiment is
performed. The experiment consist of a 2 (influencer versus brand) x 2 (product versus person wearing product) between-subjects full factorial design to test the hypothesized relationships. The experimental design enables the comparison between four different experimental groups to which participants are randomly assigned. The distinct conditions each focus on one of the centralized Instagram Story source types. According to Allen (2017) the use of random sampling and assignment of participants towards one of the experimental conditions, requires between 30 and 40 participants per conditions to leverage statistical results. Therefore, the sample size in this study exists of a minimum of 160 participants.
Table 2: Experimental conditions and number of respondents per cell in the collected data
Brand Influencer
Person C1 = Brand x person (N=40) C2 = Influencer x person (N=48))
Product C2 = Brand x product (N=51) C1 = Influencer x product (N=49)
3.2 Sampling
Based on the focus of this study, the sample group of this research can be defined as female Instagram users between the age of 18-35, considered as “millennials” (Phua, Lin & Lim, 2018). User statistics point out that the global user age for Instagram in 2018 is highest for users in the age categories 18-24 and 25-34 years. The age groups combined are
categorizes as millennials and represent 61% of all Instagram users (Statista, 2018). The statistics show that Instagram is slightly more used by females (Statista, 2018). Taken these numbers, along with the fast fashion brand perspective into account, an all female target audience is deemed most appropriate for this study. Besides, the overall target group of Zara, the used fashion brand in the stimuli material, is considered 0-45 years (Lopez & Fan, 2009). So, targeting the millennial audience in this research covers the largest population active on Instagram.
The study is performed using Qualtrics (qualtrics.com), an online survey program suitable to perform experiments. The online experiment is most convenient to reach participants during a short time scope like the timeframe of this study (Wright, 2005). Because access to any sampling frame for probability sampling is not in place, both
convenience- and snowball sampling method are applied. These sampling procedures provide an easy and quick way to gather data matching the time and resources available for this study. However, it might influence the representativeness of the research sample. Participants will be asked to participate in an online experiment via the main communication channels: social media, word of mouth and email. The research can be performed in participant’s natural environment, one of the main benefits of this sampling method.
3.3 Manipulation material & Procedure
In order to measure the effects of firm generated versus sponsored influencer
generated content on brand attitude, distinct types of manipulated content are created for the aim of the study. With the use of photoshop different branded Instagram Stories are
developed. Thecontent exists of the elements: brand logo, short text, an image and Instagram
Story source identifiers (profile picture and account name). Product and person images are chosen which can either be posted in an Instagram story of corporate account and by a
sponsored influencer. The images are similar across all conditions to ensure the consistency of the stimulus materials and exclude any influence of the content on the outcomes of the study.
In total, two versions of the four scenarios are prepared which have the design of real Instagram Story posts. The design is based on a research conducted towards the current Instagram Story posts of the central brand Zara and sponsored influencers. Scenario 1 consisted of the Instagram Story by the brand Zara with only indicating the products (jeans and blouse). Scenario 2 consisted of the Instagram Story by the brand Zara including a model wearing the products (jeans and blouse). The third and fourth scenario represented the
Instagram Story posts by a fictitious influencer indicating only the products (jeans and blouse) and the influencer wearing the products (jeans and blouse). The product (jeans and blouse) combination and the message (“check out our new @Zara collection for the brand and “check out the new @Zara collection” for the influencer) are kept equal for all source type scenarios to prevent any extraneous variable effect. The same model is used in the “person wearing the
product” conditions. An unknown model is used in the Instagram Stories to avoid any familiarity- and celebrity effects.
Two alternative manipulation scenario candidates (option A and option B) with different influencers and products (outfit) were created to select the most suitable scenarios for the main experiment. The data of the pretest was analyzed using a Chi-square analysis to test if the manipulation was successful. The results show that option A performed best and was therefore selected to use in the main study (see tables 3-6; appendix 1; Pre-test).
The main experiment consisted of the following procedure: first, the participants were shown an informed consent. Secondly, participants were asked several eligibility questions to ensure they fit the predetermined sample criteria, followed by a short explanation of the study. Then, participants were exposed to stimuli material and after that asked a set of questions to measure the (in)-dependent variables and control variables. Lastly, the manipulation check is performed together with and attention check and debriefing on the purpose of the study. The manipulation controls will determine whether the manipulation of the experiment succeeded. 3.4 Variables
Dependent variables
Brand attitude is chosen as dependent variable measured based on an existing and
previously proven reliable measurement scale. The four-item scale developed by Spears & Singh (2004) asks participants several statements about their attitude towards the brand measured via a 7-point Likert scale (1=unpleasant/7=pleasant; 1=unlikable/7=likeable; 1=bad /7=good; 1=unappealing/7=appealing).
Attitude towards the ad
, the second dependent variable, is measured according to a
four-item scale proven to be reliable in the previous study of Can and Kaya (2016). The statements asks participants their attitude towards the Instagram Story via a 7-point Likert scale (1=strongly disagree; 7= strongly agree).
Source credibility
is measured based on a five-item measurement scale proven to be
reliable in several previous studies (Ohanian, 1990). Participants are asked to describe their overall feelings about the source of the Instagram Story (1=undependable/7=dependable; 1=dishonest/7=honest; 1=unreliable /7=reliable; 1=insincere/7=sincere;
Mediating variable
Self-related mental imagery
is measured based on a measurement scale proven to be
reliable in the previous research Aydınoğlu & Cian (2014). The four-item measurement scale asks participants to what extent they imagine themselves using a certain product via a 7-point Likert scale (1=strongly disagree; 7= strongly agree).
Control variables
Besides the core variables presented above, several control variables are included in this study as they could affect the hypothesized expectations. Because of the marketing perspective and the use of a well-known existing brand, the control following control variables are incorporated in the study:
Persuasion knowledge
, tested as control variable, is based on the five-item
measurement scale of Boerman, Reijmersdal and Neijens (2012). Participants are asked about their opinion on the Instagram Story (1=dishonest/7=honest; 1=untrustworthy/7=trustworthy; 1=not convincing /7=convincing; 1=unbiased/7=biased; 1=not credible/7=credible).
Brand familiarity
is measured using a variable proven to be reliable in previous
research by Machleit, Allen & Madden (1993). The three-item scale asks participants to what extent they are familiar with the presented brand (1=unfamiliar/7=familiar;
1=inexperienced/7=experienced; 1=not knowledgeable/7=knowledgeable). Product attitude
is measured according to the three-item measurement instrument by
Becker et al (2011) asking participants statements describing their attitude towards the product shown in the Instagram Story (1=strongly disagree; 7= strongly agree).
Attention Checks
Attention source name
is measured by the multiple choice question “Who posted the
Instagram Story you just saw?
”. Respondents who answered the question correctly are
considered the most valid participants for testing the hypothesis. Attention brand name
is measured by the multiple choice question “Which brand did
you see in the Instagram Story you just saw?
”. Participants who answered this question
properly are considered the most valid participants for testing the hypothesis. Sponsored content check
is measured for participants in condition 4 (influencer x
wearing the brand’s clothes
?”. Participants who correctly answered the question properly are
categorized as the most valid participants for testing the hypothesis.
Table 7: Overview variables and measurement scales
Variables Source Scale Scale type
Independent variables
Source type Ephemeral Content type
- -
0=Brand, 1=Consumer
0=Picture showing a product, 1=Picture showing a person
Nominal Nominal
Dependent variable
Attitude towards the brand
Attitude towards the ad
Source credibility (Spears & Singh, 2004) (Can & Kaya, 2016) (Ohanian, 1990)
Describe the overall feeling about the brand you saw in the Instagram Story:
Q1:1 = Unpleasant; 7= Pleasant Q2:1 = Unlikable; 7= Likeable Q3:1 = Bad; 7= Good
Q4:1 = Unappealing; 7= Appealing
Please respond (strongly agree-strongly disagree) on the statements about the Instagram Story you just saw:
Q1:1 = I like this Instagram Story
Q2:1 = I am interested in this Instagram Story Q3:1 = This Instagram Story appeals to me Q4:1 = This Instagram post is nice
Describe the overall feeling about the source of the Instagram Story:
Q1:1 = Undependable; 7= Dependable Q2:1 = Dishonest; 7= Honest Q3:1 = Unreliable; 7= Reliable Q4:1 = Insincere; 7= Sincere Q5:1 = Untrustworthy; 7= Trustworthy 7-point Likert scale (ordinal) 7-point Likert scale (ordinal) 7-point Likert scale (ordinal) Mediating variables
Self-related mental imagery (Aydınoğlu & Cian,
2014)
Please respond (strongly agree-strongly disagree) on the following statements about the Instagram Story:
Q1:1 = I visually imagined myself using or trying the product after seeing this Instagram Story Q2:1 = I visually imagined myself shopping for this product after seeing this Instagram Story Q3:1 = I visually imagined how I would look after using this product after seeing this Instagram Story
Q4:1 = I imagined what it would be like to use this product after seeing this Instagram Story
7-point Likert scale
(ordinal)
Manipulation check questions Picture type Account type Source name Brand name Sponsored check - - - - -
Which type of image did you see in the Instagram Story?
(0= product, 1= person)
Which type of account posted the Instagram story?
(0=brand, 1=influencer)
Who posted the Instagram Story?
Q1:1 = Andy Q2:1 = Nicky Q3:1 = Zara Q4:1 = Mango
Q5:1 = I don’t remember
Which brand did you see in the Instagram Story?
Q1:1 = Andy Q2:1 = Nicky Q3:1 = Zara Q4:1 = Mango
Q5:1 = I don’t remember
In the Instagram Story, was the influencer wearing the brand’s clothes?
0= yes, 1= no Nominal Nominal Nominal Nominal Nominal Control variables Age Gender Nationality Instagram account ownership Instagram use intensity
Years of Instagram usage
- - - - - - Open question 0=Male, 1=Female Open question 0= yes, 1= no
Q1:1 = Less than once a week Q2:1 = 1-2 times a week Q3:1 = 2-3 times a week Q4:1 = 3-5 times a week Q5:1 = 5-7 times a week Q1:1 = 0-2 years Q2:1 = 3-5 years
Q3:1 = More than 5 years
Metric Nominal Nominal Nominal Ordinal Ordinal
Product attitude (Becker et al., 2011)
Q1:1 = This is a superior product Q2:1 = This is an eye-catching product Q3:1 = This is a high quality product
7-point Likert scale
(ordinal) Persuasion knowledge (Boerman,
Reijmersdal & Neijens,
2012)
I think the Instagram Story I just saw is:
Q1:1 = Dishonest; 7= Honest
Q2:1 = Untrustworthy; 7= Trustworthy Q3:1 = Not convincing; 7= Convincing Q4:1 = Unbiased; 7= Biased
Q5:1 = Not credible; 7= Credible
7-point Likert scale
Brand familiarity (Machleit, Allen & Madden, 1993)
Regarding the brand you just saw in the Instagram Story, two what extent are you:
Q1:1 = Unfamiliar, 7= familiar Q2:1 = Inexperienced, 7= experienced
Q3:1 = Not knowledgeable, 7= knowledgeable
7-point Likert scale
(ordinal)
3.5 Data cleaning & Hypothesis testing
The data gathered through the online experiment will be analyzed using SPSS. Before starting any analyses, the data is controlled for non-finished responses, missing values, and deviating values. Both non-finished responses and early-finished responses from participants who did not meet the research criteria after answering the eligibility questions are excluded from the data set.
Next, the data will be analyzed for descriptive statistics, skewness, and normality tests. The descriptive and frequency analysis are used to define the research sample and the
skewness and normality test will determine whether the variables are normally distributed. The multi-item scales (self-related mental imagery, brand attitude, attitude towards the ad, source credibility, persuasion knowledge, brand familiarity, product attitude) are tested for their reliability and validity. The reliability is tested based on Cronbach’s alpha and an
exploratory factor analysis is used to test construct validity. These analysis will show whether the measurement scales needed for analyzing the hypothesis are considered valid and reliable.
To analyze the hypothesis, both ANOVA analysis and linear regression analysis are used to test the mediation and moderation effects via SPSS PROCESS macro (Hayes, 2017). A two-way ANOVA analysis is used to compare the means between the two experimental conditions (brand versus influencer and product versus person) related to the dependent variables (H1, H2, H3). The ANOVA analysis is used to test whether the main effect of the variables (brand attitude, source credibility, and attitude towards the ad) are significantly different for the source types (brand and influencer) and picture types (product versus person wearing a product).
The PROCESS macro for SPSS is used to analyze whether the variable self related mental imagery mediates the effect of Instagram Story source- and picture type on brand attitude. For the mediation hypothesis, the magnitude of the indirect effect is analyzed. If the upper and lower confidence levels of the indirect effects contain 0, there is no mediation effect. The moderated mediation effect of source type through self-related mental imagery
will be tested based on the conditional indirect effect. Here, mediation is tested using Hayes’ PROCESS (2017) model 7 (H4).
4. Results
4.1 Sample definition
The data for this study is collected over a time period of two weeks in April and May 2018. In total 220 participants started the online experiment of which 188 completed the research, creating a 85,5% response rate for this study. Corresponding, this research has a research sample of 188 participants (N=188). In line with the predefined sampling, the
respondents exist of 100% females in possession of an Instagram account with an average age of 23,5 years (SD = 2.5). The majority of the participants is Dutch (92,6%), 78,2% of the respondents spend more than 8 times a week on Instagram and 52,7% owns an Instagram account for 3-5 years. Of the participants, 72,3% has previously bought something of the focal brand Zara.
4.2 Normality check
Both skewness and normality tests are performed to test the normality of the variables. The skewness test shows that the acceptable empirical criteria of values between -1 and +1 is not met for the variables ‘brand familiarity
’, which is substantial negatively skewed and
‘brand attitude
’, which shows a moderate negative skewness. This indicates that the most
frequent scores are grouped towards the higher end of the distribution. All variables show acceptable criteria for normal distribution except the brand related variables. This could be explained by the fact that this study uses the well-known brand “Zara” in the experimental materials. Therefore, the deviations from a normal distribution are disregarded and the original measurement scales are used to test the hypothesis.
Table 8: Normal distribution analysis
Skewness Skewness Variable N Statistic Std. Error Lower CI Upper CI Brand attitude 188 -0,789 0,177 -1,13592 -0,44208
Ad attitude 188 -0,259 0,177 -0,60592 0,08792
Source credibility 188 -0,363 0,177 -0,70992 -0,01608
Persuasion knowledge 188 -0,238 0,177 -0,58492 0,10892
Brand familiarity 188 -1,115 0,177 -1,46192 -0,76808
Product attitude 188 -0,113 0,177 -0,45992 0,23392
4.3 Reliability test & Validity test
Reliability tests
This research consists of seven latent measurement scales used as mediating-, dependent-, and control variable. In order to test the reliability of the multiple latent
measurement scales used in this study, a Cronbach’s alpha reliability analysis is performed. All latent variables brand attitude
(Cronbach’s Alpha =.951); ad attitude (Cronbach’s Alpha
=.948); self-related mental imagery
(Cronbach’s Alpha =.935); source credibility (Cronbach’s
Alpha =.906); persuasion knowledge
(Cronbach’s Alpha =.890); product attitude (Cronbach’s
Alpha =.774); brand familiarity
(Cronbach’s Alpha =.854) show high reliability (see table 9,
appendix 2; table 10, appendix 3). According to the corrected item-total correlations, all items have a good correlation value all with a total score of the scale (all above .30). None of the items would have a substantial increasing effect on the reliability score if they were deleted.
Factor analysis
Next, a principal axis factor analysis (PAF) is conducted for all seven latent
measurement scales. The Kaiser-Meyer-Olkin measure verified the adequacy of the sampling for the analysis, KMO = .915. The Bartlett’s Test of Sphericity χ² (378) = 4633.657, p < .001, indicates that correlations between the items are sufficiently for the analysis. In the initial analysis an Eigenvalue is obtained for each component in the data. Six components showed an Eigenvalue conforming the criteria of Kaiser (Eigenvalue > 1) cumulatively explaining
71,49% of the variance. In line with the criteria of Kaiser, an examination of the scree plot showed a levelling off after the sixth factor. The expected seventh component scored an Eigenvalue of .925, to which the variable can be considered trivial. However, based on the value of the component loadings (>0.40), the reliability analysis (Cronbach’s Alpha >.70), no significant increase in reliability if the lowest scoring scale-item would be deleted, and the use of all scales in various previous studies, the seventh component is kept in the factor analysis. In this case, the factors cumulatively explain 73,81% of the variance. Hence, seven factors are
incorporated and rotated using an Oblimin with Kaiser normalization rotation. The factor loadings after rotation are displayed in table 9. The items that load on the same factors suggest that the factors from one to seven represent the variables: 1-ad attitude; 2-source credibility; 3-brand familiarity; 4-self-related mental imagery; 5- brand attitude; 6-persuasion knowledge; 7-product attitude (see table 9; appendix 2).
Correlation matrix and control variable test
Next, a correlation matrix (see table 10; appendix 3) is performed to analyze how the variables relate to each other. Based on previous literature, it is expected that age, Instagram usage intensity, the years of Instagram usage, persuasion knowledge, brand familiarity and product attitude possible relate to brand attitude, ad attitude, source credibility, self related mental imagery and source- and content type. The correlation matrix illustrates that the control variables “age, brand familiarity, persuasion knowledge and product attitude”
correlate to one or multiple core variables of the research. In order to analyze whether the four experimental conditions significantly differ on these control variables, an one-way ANOVA analysis is performed. The results (table 11) show a statistically significant effect of product
attitude
among the four experimental conditions, F(3, 184) = 2.69, p < .05. The Tukey
post-hoc tests illustrates that the perceived level of product attitude was significantly higher in the brand x product
experimental condition compared to the influencer x person conditions (p
= .05). Therefore, the control variables product attitude
is incorporated in the analysis of the
hypothesis.
Table 11: One-way ANOVA analysis for the control variables among experimental conditions
SS DF MS F Sig. Age 5.40 3 1.80 0.28 0.84 Error 1185.17 184 Total 1190.57 187 Usage in years 0.56 3 0.19 0.39 0.76 Error 88.44 184 Total 88.99 187 Usage intensity 0.34 3 0.12 0.09 0.96 Error 224.57 184 Total 224.92 187
