THE POWER OF SOCIAL
The influence of social advertising on the internal psychological process and the decision- making of the consumer on Instagram.
Noor Jansen 13259024
MSc. in Business Administration - Digital Marketing (2021 – 2022) Amsterdam Business School, University of Amsterdam
Master Thesis Digital Marketing EC 20211118021129
Raoul Kübler
Final thesis submission: 28th January 2022
2 This document is written by Noor Jansen, who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document are 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 the completion of the work, not for the contents.
3
List of figures and tables ... 7
Figures ... 7
Tables ... 7
1. Introduction ... 11
2. Literature review ... 15
2.1 Important constructs ... 16
2.2 Models used to identify online consumer behavior in social commerce ... 17
2.3 Social learning theory – the purchase decision-making process ... 18
2.4 Summary of existing literature ... 19
3. Conceptual framework ... 22
3.1 Conceptual model ... 22
3.2 External interaction process ... 23
3.2.1 Social advertising ... 23
3.2.2 User-generated versus brand-generated social advertising ... 24
3.3 Internal psychological process – cognitive and affective appraisal ... 27
3.4 Decision-making – purchase intention ... 29
3.5 Advertised product – hedonic versus utilitarian ... 31
4. Research design ... 33
4
4.2 Operationalization of the variables ... 36
4.2.1 Independent variable – social advertising (user-generated vs. brand-generated) ... 36
4.2.2 Mediator – cognitive appraisal ... 36
4.2.3 Mediator – affective appraisal ... 37
4.2.4 Dependent variable – purchase intention ... 37
4.2.5 Moderator – advertised product (hedonic vs. utilitarian) ... 37
4.2.6 Control variables ... 39
4.3 Data collection ... 40
5. Results ... 40
5.1 Study 1 ... 41
5.1.1 Data cleaning ... 41
5.1.2 General data respondents ... 41
5.1.3 Skewness and Kurtosis ... 42
5.1.4 Outliers ... 42
5.1.5 Reliability ... 43
5.1.6 Correlation test ... 43
5.1.7 Manipulation check ... 45
5.1.8 Hypotheses testing ... 45
5.1.9 Summary of the results ... 49
5.2 Study 2 ... 49
5.2.1 Data cleaning ... 49
5.2.2 General data respondents ... 50
5.2.3 Skewness and Kurtosis ... 51
5
5.2.5 Reliability ... 51
5.2.6 Correlation test ... 52
5.2.7 Manipulation check ... 54
5.2.8 Hypotheses testing ... 54
5.2.9 Summary of the results ... 59
5.3 Results experiments ... 60
6. Discussion ... 60
6.1 Implications ... 67
6.1.1 Theoretical implications ... 67
6.1.2 Practical implications ... 69
6.2 Limitations & future research ... 70
7. Conclusion ... 73
Reference list ... 76
Appendices ... 86
Appendix A – Extensive literature overview ... 86
Appendix B – Operationalization of the variables ... 88
Appendix B1 – Mediator cognitive appraisal ... 88
Appendix B2 – Mediator affective appraisal ... 88
Appendix B3 – Dependent variable purchase intention ... 89
Appendix C - Survey questions study 1 ... 89
Appendix D – Survey questions study 2 ... 96
6
Appendix F – Data analysis study 1 ... 102
Appendix F1 – Robust normality check ... 102
Appendix F2 – Manipulation test ... 102
Appendix F3 – Linear Regression Analysis and assumptions ... 103
Appendix G – Data analysis study 2 ... 105
Appendix G1 – Robust normality check ... 105
Appendix G2 – Manipulation test ... 105
Appendix G3 – Linear Regression Analysis and assumptions ... 106
7 Figures
Figure 1 Conceptual model ... 22
Figure 2 Example of #MyMonsoon (Monsoon, 2021) ... 25
Figure 3 Interaction plot H6 ... 59
Figure 4 Advertisement brand-generated condition ... 91
Figure 5 Advertisement user-generated condition ... 92
Figure 6 Advertisements 1 and 2 user-generated and hedonic condition ... 97
Figure 7 Advertisements 1 and 2 user-generated and utilitarian condition ... 98
Figure 8 Advertisements 1 and 2 brand-generated and hedonic condition ... 99
Figure 9 Advertisements 1 and 2 brand-generated and utilitarian condition ... 100
Figure 10 P-P Plot study 1 ... 103
Figure 11 Scatterplot study 1 ... 103
Figure 12 P-P Plot study 2 ... 106
Figure 13 Scatterplot study 2 ... 106
Tables Table 1 Literature review ... 21
Table 2 Overview hypotheses ... 33
Table 3 2x2 factorial design online experiment ... 35
Table 4 Operationalization of the variables ... 38
Table 5 Stimuli for experiment 2 (Khan & Dhar, 2006, p. 266) ... 39
Table 6 Descriptive statistics study 1 ... 42
Table 7 Correlation Matrix study 1 ... 44
8
Table 9 Total effects PROCESS study 1 ... 49
Table 10 Descriptive statistics study 2 ... 50
Table 11 Correlation Matrix study 2 ... 53
Table 12 PROCESS model 7 study 2 ... 58
Table 13 Moderated mediation PROCESS study 2 ... 58
Table 14 Overview results study 1 and 2 ... 60
Table 15 Extensive literature overview ... 86
Table 16 Demographics study 1 & 2 ... 101
Table 17 Skewness & Kurtosis study 1 ... 102
Table 18 Manipulation check study 1 ... 102
Table 19 Multicollinearity study 1 ... 104
Table 20 Linear Regression Analysis study 1 ... 104
Table 21 Skewness & Kurtosis study 2 ... 105
Table 22 Manipulation check study 2 ... 105
Table 23 Multicollinearity study 2 ... 107
Table 24 Linear Regression Analysis study 2 ... 107
9 First and foremost, I want to thank my supervisor, Raoul Kübler, for supporting me throughout the thesis process. He has provided me with guidance and insightful feedback that has pushed me to a higher level. I also want to thank my second reader for taking the time to read my thesis.
Next, I want to thank my respondents, without whom I would not have been able to finish my thesis. Lastly, I want to thank my friends and family for supporting me throughout this process.
Writing my thesis during the COVID-19 pandemic has been challenging, requiring a lot of my perseverance and motivation. Fortunately, my friends and family were always there for me to listen to my troubles, but, more importantly, to provide me with new energy and motivation to finish my thesis.
10 Social media channels are gradually changing into sales channels through social commerce, with Instagram taking the lead. As a result, the purchase decision-making process of the consumer simultaneously changes. This research confirms an additional social commerce construct, social advertising. Since managers are increasingly allocating a more significant share of their marketing budget towards social advertising, it is essential to understand its influence on the consumer’s purchase decision-making process. Thus far, few studies have investigated this, especially combined with the social learning theory. Consequently, this study researched the influence of social advertising (brand-generated versus user-generated) on the internal psychological process (cognitive and affective appraisal) and the decision-making of the consumer (purchase intention) on Instagram and whether this effect is moderated by the type of advertised product (hedonic versus utilitarian). Two experiments were performed; the first investigated the direct effect (N = 219) and the second the moderated effect (N = 448). The results show that the effect of social advertising on the decision-making of the consumer is mediated by the internal psychological process, such that brand-generated social advertising leads to a stronger cognitive appraisal and user-generated social advertising leads to a stronger affective appraisal. The moderation effect is only significant for user-generated social advertising on cognitive appraisal, with a stronger effect for the utilitarian advertised product.
In conclusion, the findings expand the available theoretical knowledge about online consumer behavior in the setting of social commerce and provide managers with practical guidelines on how to include social advertising within their marketing strategy.
Keywords: social learning theory, social commerce, external interaction process, internal psychological process, decision-making process, user-generated social advertisements, brand- generated social advertisements, hedonic and utilitarian advertised products
11 During Singles’ Day 2020 in China, companies sold 1.9 billion products within 24 hours (Harper, 2020). A total of 3.000 flights and cargo ships and three million people were needed to deliver these products (Harper, 2020). Singles’ Day, also known as double 11, is the world's most significant online shopping day (Liberto, 2021). According to Liberto (2021), the holiday originated in 1993 when a few students from China started the tradition. During this day, unmarried people treat themselves to gifts (Liberto, 2021). Singles’ Day accelerated when Alibaba started to offer deep discounts on its platform for 24 hours (He & Ma, n.d.). In 2020, Alibaba registered $74.1 billion worth of sales (Liberto, 2021). A big part of Alibaba’s success is due to its social commerce strategy (i.e., purchases made via social media) (Insider Intelligence, 2016).
According to Kaplan (2020), an example of Alibaba’s social commerce strategy is that one of China’s most popular influencers live-streamed to over 149 million viewers on Alibaba’s live platform during Singles’ Day. The viewers could purchase products without leaving the live stream (Kaplan, 2020). In 2017, 90% of Singles’ Day purchases were made on mobile phones (Quinlan, 2021). It is predicted that by 2022, there will be 1,069.53 million users of social media networks in China, making them the world’s largest social media market (Thomala, 2021). Consequently, Singles’ Day shows the potential that social commerce can offer companies and the need for researchers to understand the underlying processes that drive social commerce.
From January 2021 to January 2022, the number of active social media users worldwide increased by 10.1% (424 million users), which led to a total of 4.62 billion users worldwide (We Are Social & HootSuite, 2022). Subsequently, this increasement shows essential room for growth for companies on social media channels, such as Facebook, LinkedIn, Instagram, and
12 organizations planning to increase their social advertising investments in Instagram in 2022 (We Are Social & HootSuite, 2022). These impressive numbers indicate the need for more profound knowledge into investing in social media and the returns of these investments.
Over the last decade, social media have fundamentally influenced the consumer purchase decision-making process (i.e., need recognition, search, evaluation, purchase, and post- purchase) (Hudson & Thal, 2013). For instance, 76.1% of Internet users worldwide use social media to search for brand information, and 27.6% discover new brands and products via advertisements on social media (We Are Social & HootSuite, 2022). Over the past few years, Instagram has been slowly shifting from a social media channel to a sales channel (Baron &
Ciechomski, 2019). An interesting case study is L’Oréal, which hopped as one of the firsts on the social commerce trend and has started to invest in social commerce. Lubomira Rochet, the Chief Digital Officer, said: “Social commerce is an exciting new form of e-commerce that enables consumers, influencers, experts, beauty, or shop assistants to sell brands and products on social platforms through formats such as live shopping or live streaming. The rise of social commerce is a great opportunity for our brands to reinvent the consumer beauty experience worldwide.” (L’Oréal, 2020, p. 1). Unfortunately, L’Oréal is one of the exceptions.
Yadav et al. (2013, p. 312) define social commerce as “exchange-related activities that occur in, or are influenced by, an individual’s social network in computer-mediated social environments, whereby the activities correspond to the need recognition, pre-purchase, purchase, and post-purchase stages of a focal exchange.”. Social commerce falls under the e- commerce umbrella (Phaneuf, 2021) by adding e-commerce functionalities to social networks (Li & Ku, 2018). According to Zhang & Benyoucef (2016, p. 95), the difference between e- commerce and social commerce is that in e-commerce, consumers interact with online shopping sites separately, whereas, in social commerce, online communities support interactions. Social
13 online reviews (Zhang & Benyoucef, 2016). Social commerce empowers consumers, is user- centered, and offers an interactive online experience, whereas e-commerce offers its consumers little control, is product-centered and provides a one-way interaction (Li & Ku, 2018).
According to Lin et al. (2017), social commerce is an important part of social advertising because social commerce relies on social media to carry advertisements. Companies are trying to reach their target audience on social media through paid advertisements. Companies' investments in social advertising are growing due to its interactivity and better reach (Lin et al., 2017). According to Business Insider (2021), social media is going to be the fastest-growing advertising channel between 2021 and 2024. Social advertisement spending worldwide is forecasted to increase to $177 billion in 2022 and even up to $225 billion in 2024 (Business Insider, 2021). Research by Cooper (2020) shows that 27.6% of Internet users find new products and brands through paid social advertisements. These numbers show how important it is to deepen the knowledge about the influence of social advertising on the purchase decision- making process of the consumer. Similarly, it also raises a few questions for managers and researchers. What is the impact of social commerce, specifically social advertising (user- generated versus brand-generated), on the consumer’s internal psychological process and decision-making? Is this effect moderated by the type of advertised product (hedonic versus utilitarian)?
Research into social commerce started to receive more attention in 2004 with the launch of Facebook and has continued to grow ever since (Esmaeili & Hashemi G, 2019; Lin et al., 2017).
However, research has mainly focused on three social commerce constructs: ratings and reviews, social recommendations, and forums and communities (Hajli, 2015). On the other hand, social advertising, which is also an important social commerce construct, has received
14 process of the consumer has been researched to a lesser extent.
Consequently, social advertising is the center of attention of this research. This study uses the social learning theory (SLT) to research the effect of social advertising (user-generated versus brand-generated) on the purchase decision-making process of the consumer on Instagram and whether this effect is moderated by the type of advertised product (utilitarian versus hedonic). To fill the above-mentioned research gap, the following research question was formed:
What is the influence of social advertising (brand-generated versus user-generated) on the internal psychological process (cognitive and affective appraisal) and the decision-making (purchase intention) of the consumer on Instagram, and is this effect moderated by the type of advertised product (hedonic versus utilitarian)?
In light of the above, this study takes on a quantitative experimental approach to explore the effects of social advertising within the SLT. As a result, the objectives of this research are fourfold. The first objective is to explain the concept of social advertising using two different types of social advertising (user-generated versus brand-generated). Second, to study the effect of social advertising on the cognitive and affective appraisal of the consumer. Third, to investigate the influence of the cognitive and affective appraisal on the consumer's purchase intention. Fourth, to research the moderated effect of the type of advertised product (hedonic versus utilitarian) on the internal psychological process of the consumer.
This study extends the social commerce literature in multiple ways. The findings provide a comprehensive understanding of the consumer’s purchase decision-making process within the social advertising context. Moreover, this research extends the application scope of the SLT.
15 The study offers valuable insights due to its experimental research method, as surveys dominate the social commerce literature. The study extends the knowledge about social commerce by combining the type of advertisement (consumer-generated versus brand-generated) with the advertised product (hedonic versus utilitarian). Thus far, this combination has not yet been made, especially in combination with the SLT. Lastly, it extends the knowledge about social advertising, specifically for Instagram.
Moreover, this research has important implications for managers as it can help managers create a social advertising strategy. As mentioned earlier, organizations are increasingly investing more money in social advertising. Therefore, this research provides managers with valuable insights into applying social advertising within their business. Additionally, it gives them practical guidelines on what type of social advertisement (user-generated versus brand- generated) and what type of product (utilitarian versus hedonic) to advertise to achieve the wanted results on their investments.
The rest of the thesis is organized as follows: first, the literature on social advertising and the social learning theory is reviewed, followed by the conceptual framework and the development of the hypotheses. The following section describes the methodology. After that, the results are discussed. Subsequently, the discussion is described, after which the limitations, future research, and implications follow. Finally, the conclusion is discussed.
2. Literature review
This chapter explains the theoretical background of the study. It starts with describing the most important constructs and what is known and not (yet) known about social commerce and social advertising. Next, the models used to identify online consumer behavior in social commerce
16 available literature.
2.1 Important constructs
According to Yadav et al. (2013), social commerce comprises two important elements:
social media and commercial activities. Research into social commerce started receiving more attention in 2004 with the launch of Facebook and has continued to grow ever since (Esmaeili
& Hashemi G, 2019; Lin et al., 2017). The increasing attention that social commerce is receiving demonstrates the relevance of the topic. According to Esmaeili & Hashemi G (2019, p. 317), the most researched themes within social commerce are consumer behavior, social commerce website design, and social commerce adoption. Research into online consumer behavior in the context of social commerce can be broadly categorized as behavioral theories, social-related theories, culture-related issues, and consumers’ motives, benefits, and values (Zhang & Benyoucef, 2016). The current study investigates the online consumer behavior theme and, more specifically, behavioral studies combined with consumers’ motives, benefits, and values.
An important topic within online consumer behavior is the purchase decision-making process of the consumer. Huang & Benyoucef (2017, p. 42) refer to this as “a cognitive process resulting in the selection of a product, service, or course of purchase action from several alternatives.”. Research in the purchase decision-making process has used many different models. However, the social learning theory (SLT) has received less attention in identifying online consumer behavior in social commerce. The SLT states that cognitive and environmental factors influence human learning and behavior. The theory consists of three dimensions: the external interaction process, the internal psychological process, and the decision-making process (Chen et al., 2017). Hence, it is noteworthy to research the role of social advertising in
17 within social commerce studies, while experiments are underrepresented (Zhang & Benyoucef, 2016). Therefore, using an experiment as the empirical research method within this setting would be interesting.
Researchers have often used the three social commerce constructs described by Hajli (2015): ratings and reviews, social recommendations, and forums and communities. Multiple studies have used these social commerce constructs (Chen et al., 2017; Hajli, 2015; Lin et al., 2017; Zhang & Benyoucef, 2016). However, other social commerce constructs have received less attention while they have a lot of potential, such as social advertising.
Social advertising relies on social networking sites (SNSs) to generate, target, and deliver marketing communications. Several studies have investigated the effect of social advertising on consumer trust and purchase intention (Choi & Lee, 2017; Goh et al., 2013; Kim & Johnson, 2016; Krishnamurthy & Dou, 2008; Mayrhofer et al., 2020; Thompson & Malaviya, 2013).
Consumers perceive user-generated advertisements as more trustworthy than brand-generated advertisements (Thompson & Malaviya, 2013).
2.2 Models used to identify online consumer behavior in social commerce
The Stimulus-Organism-Response (SOR) theory and the five-stage consumer decision- making process are often used to identify the purchase decision-making process of the consumer within social commerce. The SOR theory explains that a stimulus triggers a response based on an organism’s internal feelings or behavior (Wu & Li, 2018). Wu & Li (2018) used the SOR theory to investigate the six marketing mix components of social commerce (i.e., social commerce needs, social commerce risk, social commerce convenience, social capital, social identification, and social influence) on consumer loyalty through consumer value. Their findings showed that all six marketing mix components significantly affected consumer value,
18 process explains five stages in the consumer’s decision-making process: need recognition, search, evaluation, purchase, and post-purchase (Hettiarachchi et al., 2018). Huang &
Benyoucef (2017) researched the effect of social commerce design on the consumer’s purchase decision-making. Their research showed that the design factors usability, functionality, and social features are essential at the five stages of the consumer purchase decision-making process.
The social learning theory has received less attention in identifying online consumer behavior in social commerce. While the SOR theory and the five-stage consumer decision- making process are often used in the social commerce context, they lack deepening concerning the internal interaction process of the consumer. The SLT splits this process into two, namely cognitive and affective appraisal.
2.3 Social learning theory – the purchase decision-making process
As mentioned above, compared to other theories, the SLT has received less attention from researchers in the social commerce context. However, it is an interesting theory to combine with social commerce. The SLT is special compared to other theories because it adds a social element; it shows that people can learn information and behavior by observing other people (Chen et al., 2017). The SLT has its roots in psychology and helps deal with behavioral concerns in a structured approach. All learning implies the integration of an external interaction process and an internal psychological process (McCullough Chavis, 2011).
Chen et al. (2017) researched the consumer’s purchase decision-making process in social commerce using the SLT. In their study, the purchase decision-making process consisted of three stages: the external interaction process, the internal psychological process, and the decision-making process (i.e., purchase intention). The external interaction process is a process
19 participation, communication, and cooperation (Chen et al., 2017). In the social commerce context, the external interaction process refers to the interaction with social commerce constructs. The internal psychological process refers to the attitude of the consumer after the interaction (Marsden, 2010) and consists of the cognitive and affective appraisal of the consumer (Chen et al., 2017). Finally, the decision-making process exists in the study of Chen et al. (2017) out of the purchase intention of the consumer.
The research of Chen et al. (2017) only considered the earlier mentioned three social commerce constructs researched by Hajli (2015): forums and communities, ratings and reviews, and social recommendations. Chen et al. (2017) showed that cognitive and affective appraisal are positively associated with a consumer’s purchase intention on a social commerce website.
Furthermore, they found that learning from forums and communities and ratings and reviews significantly influenced cognitive and affective appraisal attitudes. However, learning from social recommendations had no significant impact. As a suggestion for future research, they suggested to investigate other social commerce constructs, such as social advertisements (Chen et al., 2017). Other research that used the SLT looked at social media influencers combined with the social capital theory and social exchange theory (Chia et al., 2021). Nevertheless, no research has combined social commerce, specifically social advertising, with the SLT.
2.4 Summary of existing literature
Table 1 below portrays an extensive overview of the literature from which several conclusions can be drawn about the existing literature. First of all, an increasing amount of research has started to investigate social commerce and its effects. Topics that are often researched are the value of earned social media, especially within online communities, electronic word-of-mouth (eWOM), attitudes towards the advertisements, and user-
20 Research into social commerce has used different models and theories to explain consumer behavior. Still, a few have received more attention than others, such as the SOR theory, the Elaboration Likelihood Model (ELM), and the five-stage consumer decision-making process.
And lastly, Facebook dominates the research as chosen SNS.
Simultaneously, the table clearly demonstrates research gaps in the literature. The SLT is not often used to investigate consumer behavior within social commerce. More specifically, thus far, no study has combined the SLT with social advertising and the type of advertised product. Moreover, this research takes on an experimental research method, which is underrepresented in the literature. Additionally, this research uses Instagram as SNS, while most other studies have used Facebook. The extended version of the literature review below can be found in Appendix A in Table 15.
In table 1, other studies that were not included in the table are the five-stage consumer behavior decision-making process, the social identity theory, the self-categorization theory, the skepticism-identification model, the Unified Theory of Acceptance and Use of Technology (UTAUT2), Kaplan’s landscape preference model, the product trial theory, the attitudes and hierarchy response model, and the social exchange theory. Additionally, platforms that were not included in the table are Taobao, Naver, hedonic information systems, Twitter, and social media networks (SNSs) in general.
.
21
Study Survey Experiment
Content
analysis SOR TAM SLT ELM PKM
Other
theories Facebook
Social commerce
website No platform
Others platforms
Wu & Li (2018) X X X
Huang & Benyoucef
(2017) X X X
Chen, Lu & Wang (2017) X X X
Hajli (2015) X X X
Choi & Lee (2017) X X
Goh, Heng & Lin (2013) X X X
Kim & Johnson (2016) X X X
Mayrhofer, Matthes, Einwiller & Naderer
(2020) X X X
Thompson & Malaviya
(2013) X X X
Chia, Hsu, Lin & Tseng
(2021) X X X
Alalwan (2018) X X X
Steyn, Ewing, van Heerden, Pitt & Windisch
(2011) X X X
Lee & Kozar (2009) X X X
Chae, Stephen, Bart &
Yao (2017) X X X
Schulze, Schöler & Skiera
(2014) X X X
Kempf (1999) X X X
Colicev, Malshe, Pauwels
& O'Connor (2018) X X X X
van der Heijden (2004) X X X X
Duffett (2015) X X
Research method Used theory/model Used platform
22 This chapter explains the conceptual model of this study, which is based on the three stages of the social learning theory: the external interaction process, internal psychological process, and decision-making. After that, the hypotheses development is discussed, and lastly, the chapter ends with an overview of the proposed hypotheses and relations.
3.1 Conceptual model Figure 1
Conceptual model
The theoretical framework is the basis for the development of the conceptual framework in Figure 1. This model shows the relationships between the different variables. This study investigates the effect of social advertising (user-generated versus brand-generated) on the
23 internal psychological process and the decision-making process of the consumer. Furthermore, this study researches the influence of the moderator on the type of advertised product (hedonic versus utilitarian) on the effect of social advertising on the internal psychological process.
3.2 External interaction process 3.2.1 Social advertising
As mentioned before, the external interaction process is a process between the learner and their environment (Chen et al., 2017). A form of such a process is social advertising. According to Lin et al. (2017), social advertising is part of social commerce. Bakshy et al. (2012, p. 146) describe social advertising as “using information about consumers’ peers, including peer affiliations with a brand, product, and organization to target ads and contextualize their display”. Social advertising has two important advantages over offline advertisements (Lin et al., 2017). Firstly, through social media, the advertisement can be delivered to consumers with a certain user profile. Secondly, social advertising creates the possibility of generating online groups to build brand communities of loyal consumers. However, according to Lin et al. (2017), creating an effective social advertisement can still be quite challenging. If the advertisement is not informative, entertaining, creditable, or is even irritable, consumers can become skeptical of the brand and ignore future advertisements.
According to Lin et al. (2017), research concerning social media advertisements has mainly investigated the themes of organization, advertisement, and word-of-mouth. The acceptance and avoidance of advertisements and online social networking communities are less explored.
This study partly investigates the acceptance of social advertising by looking at the consumer’s purchase decision-making process.
Alalwan (2018) tested the influence of social media advertisement factors (i.e., interactivity, perceived relevance, performance expectancy, and informativeness) on purchase intention. The
24 results showed that interactivity was the most significant factor predicting purchase intention.
Additionally, it plays an important role in hedonic motivations (i.e., affective appraisal) and performance expectancy. On the other hand, informativeness was the second strongest factor to predict purchase intention and significantly predicted performance expectancy. It is noteworthy to research the role of social advertising on cognitive appraisal, compared to solely affective appraisal. Moreover, limited research has been performed into the influence of social advertising on the consumer’s purchase decision-making process.
Instagram has, after Facebook, the highest share of potential advertising audience reach versus the total population aged 13+ (We Are Social & HootSuite, 2022). However, Instagram is well on its way to surpass Facebook. According to the prediction of We Are Social &
HootSuite (2022), Instagram’s advertising reach versus the total population aged 13+ will increase from 20% in 2021 to 23.9% (1.478 million users) in 2022. In comparison, Facebook’s advertising reach will decrease from 36% in 2021 to 34.1% (2.109 million users) in 2022. Since Instagram will probably catch up with Facebook’s advertising reach, it is not only interesting but also essential to research the effects of social advertising on Instagram.
3.2.2 User-generated versus brand-generated social advertising
Companies can choose between two types of social advertisements: user-generated or brand-generated advertisements. The consumption of these two types of advertisements influences consumer mindsets and, thus, behavior (Schivinski & Dabrowski, 2016). User- generated content is any type of content created voluntarily by end-users and not by media professionals (Knoll, 2016). Examples of user-generated content are product reviews and blog posts. This study focuses on another form of user-generated content, namely user-generated advertising, where a brand advertises content made by consumers. Consumer-generated advertising is a form of user-generated content and refers to specific instances where consumers create brand-focused messages to inform, persuade, or remind others (Campbell et al., 2011, p.
25 87). Brand-generated advertising is content that the company oversees and is governed by the company’s marketing strategy (Schivinski & Dabrowski, 2016). Social media enables consumers to become gatekeepers and producers of brand-related content (Tsai & Men, 2017, p. 2). The advantage of consumer-generated advertisements is that it gains easier consumer acceptance in comparison to brand-generated advertisements because the consumer expects the consumer who creates the content to produce trustworthy content (Krishnamurthy & Dou, 2008). According to Bazilian (2017), 47% of millennials and 36% of baby boomers trust user- generated content, whereas only 25% of both trust branded content. An example of a company that uses user-generated content is Monsoon.
Figure 2
Example of #MyMonsoon (Monsoon, 2021)
26 Monsoon asked their consumers to share photos with the hashtag #MyMonsoon, see Figure 2. When Monsoon used these user-generated photos in their email campaign, the campaign's click-through rate increased by 14% (Greenbaum, 2020). Figure 2 is an example of how the hashtag is included on the website of Monsoon. Unfortunately, many companies struggle to combine user- and brand-generated content in their social advertising strategy.
Research into consumer-generated advertising has received more attention increasingly.
Thompson & Malaviya (2013) show in their research that consumer-generated advertising can create positive as well as negative effects. It can create positive effects because consumer- generated advertising can enhance the advertisement and brand evaluations. After all, the consumer can identify more with the consumer, who created the advertisement, than with the brand. According to Bazilian (2017), 76% of consumers believe that content shared by other users is more honest than advertisements of brands. However, it can also create negative effects because users are perceived as less competent to create advertisements, creating skepticism towards the brand’s advertisements.
The opposite of consumer-generated advertising is brand-generated advertising, in which the brand is presented as the source. Steyn et al. (2011) researched the source effects of consumer-generated advertising. They discovered no overwhelming evidence that consumers preferred consumer-generated advertisements over brand-generated advertisements. However, they did detect a significant difference in that consumers are more critical of the advertisement when labeled as either brand-generated or consumer-generated compared to the unlabeled advertisement group (Steyn et al., 2011).
Mayrhofer et al. (2020) investigated the purchase intention of disclosed advertisements, brand posts, and user-generated posts. The results showed that user-generated content led to a higher purchase intention compared to brand posts and disclosed advertisements. However, this study only looked at user-generated content with the user and not the brand as the source
27 (Mayrhofer et al., 2020). Consequently, this creates a gap in the literature to discover the purchase intention of consumer-generated versus brand-generated advertisements on social media. Additionally, less is known about the impact of those advertisements on the internal psychological process (cognitive and affective appraisal) of the consumer, which in turn could lead to purchase intention.
3.3 Internal psychological process – cognitive and affective appraisal
The second stage of the social learning theory is the internal psychological process which refers to the attitude of the consumer after interaction (Chen et al., 2017). This stage comprises two types of appraisals: cognitive and affective appraisal. These two types of appraisals are two distinct dimensions of the attitude theory. According to Kihlstrom & Park (2016), this theory provides a link between perception and memory. The amount of attention given to a particular event predicts the likelihood that the event will be remembered. Cognitive appraisals are based on the utilitarian aspects of attitude, while affective appraisals are based on emotions, feelings, and gut reactions (Lee & Kozar, 2009). According to Tavares et al. (2021), the cognitive and affective appraisals are derived from the product and the consumer. The consumer side involves perception, whereas the product side involves cognitive attributes such as functionality, usability and affective attributes such as pleasure and hedonism (Tavares et al., 2021).
According to Tavares et al. (2021), cognition is used to interpret, comprehend, and understand.
Affection refers to sentimental responses and promotes learning and experience in interaction with the product. Both cognition and affection are needed to make sense of the world and complement each other (Tavares et al., 2021). According to Lee et al. (2012), both appraisals affect the attitude of the consumer when they evaluate a product. The emotions and mood of the consumer may impact the beliefs regarding an object.
28 Choi & Lee (2017) studied the influence of user- and marketer-generated content in social network services on consumer trust. They discovered that user-generated content has a stronger effect on consumers’ cognitive trust because it can be interpreted as a general evaluation of a product. According to Choi & Lee (2017), marketer-generated content is often perceived as persuasive, of which consumers are generally skeptical. In addition, consumers believe that companies mainly advertise the positive aspects of their products, while other consumers have little motivation for misrepresentation in their reviews. However, their study focused on perceived trust, while this study focused on perceived appraisal.
Research on the effects of brand-generated advertising is divided (Colicev et al., 2018). On the one hand, the ‘truth effect’ leads to credibility in consumers' minds due to repetition.
However, on the other hand, consumers might perceive brand-generated advertising as disguised advertising and are, as a result, skeptical towards it (Colicev et al., 2018).
Based on the fact that brand-generated advertisements usually contain more practical information, this study posits that brand-generated social advertisement has a stronger positive effect on cognitive appraisal compared to user-generated social advertising. Therefore, the following hypothesis was formed:
H1: Brand-generated (vs. user-generated) social advertising has a stronger positive effect on cognitive appraisal.
As mentioned above, user-generated advertising gains more consumer acceptance than brand-generated advertising because the consumer can identify more with the content and expects it to be trustworthy. However, brand-generated advertisements usually contain more practical information and facts as marketers are more knowledgeable about the products and services (Goh et al., 2013).
29 Kim & Johnson (2016) examined the influence of positive brand-related user-generated content using the SOR theory. They found that brand-related user-generated content activated both the emotional (activated pleasure) and cognitive (perceived information quality) responses of consumers. These findings are in line with research by Lee et al. (2012), which states that heightened enjoyment positively influences the affective appraisal of the consumer. Vazquez et al. (2020) researched the influence of user-generated content on the consumer’s online experience. They found that user-generated content positively influences the emotions of consumers. The effect was most powerful for pleasure, which can be compared to affective appraisal.
Based on the findings that user-generated advertisements activate both emotional and cognitive responses of consumers, this study states that user-generated social advertising has a stronger positive effect on affective appraisal compared to brand-generated social advertising.
Therefore, the following hypothesis was formed:
H2: User-generated (vs. brand-generated) social advertising has a stronger positive effect on affective appraisal.
3.4 Decision-making – purchase intention
The last stage of the SLT is the decision-making of the consumer (Chen et al., 2017). This study looks at the purchase intention of the consumer as part of their decision-making. The consumer's purchase intention can describe the consumer’s behavior in social commerce (Chen et al., 2017). Previous research has proposed a direct relationship between consumers’ cognitive and affective appraisal and behavior intention, such as purchase intention (Chen et al., 2017;
Lee & Kozar, 2009; van der Heijden, 2004). Chen et al. (2017) show that cognitive and affective appraisals are positively associated with a consumer’s purchase intention on a social commerce
30 website. The attitude that consumers form from their cognitive and affective appraisals plays an important role in their purchase decision-making process (Chen et al., 2017). Positive cognitive and affective appraisals can change the attitude of the consumer and, as a result, trigger purchase intention. However, Chen et al. (2017) researched the purchase intention within a social commerce website, while this study examines purchase intention within Instagram. Lee & Kozar (2009) researched the usability of online stores. They found that if both the affective and cognitive appraisals are high, the consumer’s purchase intention is higher.
Van der Heijden (2004) investigated the user acceptance of hedonic information systems. The results showed that perceived enjoyment (comparable to affective appraisal) and perceived ease of use (comparable to cognitive appraisal) are strong determinants of intention to use. Lee et al.
(2012) explored the effect of presentation modes and filler interfaces on perceived waiting time online. They found that affective and cognitive appraisal positively influence the intention to use the website.
The studies described above show the relationship between consumers’ cognitive and affective appraisal and purchase intention in different settings. However, it is expected that these findings are applicable within this research. Based on these findings, it is proposed that the higher the perceived cognitive and affective appraisal, the higher the purchase intention.
Therefore, the following hypotheses were formed:
H3: The higher the perceived cognitive appraisal, the higher the purchase intention.
H4: The higher the perceived affective appraisal, the higher the purchase intention.
31 3.5 Advertised product – hedonic versus utilitarian
Companies can advertise their products via their social media channels. The products they display in their social advertisements can be categorized as hedonic or utilitarian. According to Okada (2005), hedonism stands for wants or vices and offers experiential enjoyment, such as clothes or jewelry. Utilitarianism stands for shoulds or virtues and offers practical functionality, such as a vacuum cleaner. People are motivated to have fun, but, simultaneously, this raises issues such as guilt and the need for justification (Okada, 2005). Chae et al. (2017) found that utilitarian products are often explained through actions and choices, whereas hedonic products are often explained through reactions and feelings. Schulze et al. (2014) researched differences between hedonic and utilitarian Facebook campaigns’ sharing mechanisms. They found that the hedonic sharing mechanisms are the opposite of the sharing mechanisms for utilitarian campaigns. The main reason for this is that consumers do not visit Facebook to learn about utilitarian products; instead, they rely on simple cues and heuristics to process the messages (Schulze et al., 2014). Unsolicited direct messages and incentives for the receiver are less effective for utilitarian products (i.e., hedonic sharing mechanism). In contrast, broadcast messages from strangers and direct messages from friends are more effective for utilitarian products (i.e., utilitarian sharing mechanism). The study of Schulze et al. (2014) focused on Facebook; it is interesting to test these results on other social media platforms, such as Instagram. Accordingly, it is noteworthy to research the influence of hedonic versus utilitarian products on the internal psychological process of the consumer.
Kempf (1999) researched the attitude formation in the product trial and the distinct roles of cognition and affect for hedonic and functional products. Product trial refers to a consumer’s first usage experience with a brand (Sun et al., 2017). Kempf (1999) found that affective responses have a greater effect on a hedonic product (i.e., a computer game) than cognitive responses. Especially arousal was shown to be an important determinant. The opposite was
32 shown for a utilitarian product (i.e., grammar-checking software) in that the trial evaluation process appeared to be more cognitive in nature. Pleasure was a significant explanatory variable for both products (Kempf, 1999).
Colicev et al. (2018) researched the effects of owned and earned social media on consumer mindset metrics and shareholder value. They found that owned social media is more likely to increase purchase intention for firms with a high reputation and high-involvement utilitarian brands (Colicev et al., 2018).
Based on the findings that explain that the type of product influences the consumer’s reaction, this study posits that the type of advertised product moderates the effect of social advertising on cognitive appraisal such that a utilitarian advertised product has a stronger positive effect on cognitive appraisal. Therefore, the following hypothesis was formed:
H5: The effect of social advertising on cognitive appraisal is moderated by the type of advertised product, such that a utilitarian (vs. hedonic) advertised product has a stronger positive effect on cognitive appraisal.
On the contrary, Goh et al. (2013) found that user-generated content on social media is more impactful on online consumer behavior than marketer-generated content, especially for hedonic goods. Thus, based on the findings described above, it is proposed that a hedonic advertised product moderates the effect of social advertising on affective appraisal. Therefore, the following hypothesis was formed:
H6: The effect of social advertising on affective appraisal is moderated by the type of advertised product, such that a hedonic (vs. utilitarian) advertised product has a stronger positive effect on affective appraisal.
33 Table 2 shows an overview of the hypotheses tested in this study, the relationship between the variables, and the prediction.
Table 2
Overview hypotheses
4. Research design
This chapter explains the chosen research approach used to test the above-described hypotheses.
Next, the variables are operationalized accordingly. Finally, the results are discussed in the next chapter.
4.1 Research approach
This study aims to investigate the effect of user-generated versus brand-generated advertisements on the internal psychological process (cognitive versus affective appraisal) and
34 the decision-making (purchase intention) of the consumer on Instagram. This effect is possibly moderated by the moderator type of advertised product (utilitarian versus hedonic). In this study, social advertising is the independent variable, cognitive and affective appraisal are the mediators, purchase intention is the dependent variable, and the type of advertised product is the moderator.
Thus, this research takes a deductive approach, confirming the social learning theory (SLT).
This study uses a quantitative research method to test the relationships between the variables and analyze the moderating effect of additional variables. To be more precise, the research method will be an online experiment using the online survey design tool ‘Qualtrics’. Two experiments will be conducted; the first experiment will measure the direct effect of social advertising, and the second experiment will measure the moderated effect of the advertised product. Another reason this research uses two instead of one survey is that the respondents might become biased after seeing multiple advertisements and start to understand the goal of the survey. Although the questions in both surveys are the same, the used social advertisements are different. Both experiments are a between-subjects design, which means that respondents are randomly assigned to only one of the treatment conditions (Keren, 2014). According to Keren (2014), a between-subjects design results in a higher degree of sensitivity compared to a within-subjects design. This higher degree of sensitivity is due to the exclusion of individual differences. A manipulation check will be conducted within both experiments to test whether the manipulations have the intended effects. Once the results of the online experiments are collected, they will be analyzed using SPSS.
The first experiment will research the direct effect of social advertising (brand-generated versus user-generated) on the cognitive and affective appraisal and purchase intention of the consumer on Instagram. Participants will be randomly assigned to one of the two conditions:
35 user-generated or brand-generated. Appendix C contains the questions and advertisements that were used for this survey. Study 1 investigates H1 – H4.
The second experiment will investigate the effect of the moderator type of advertised product. This experiment will have a 2 (social advertising: user-generated versus brand- generated) x 2 (advertised product: hedonic or utilitarian) factorial between-subjects design, as shown in Table 3. The participants will be asked questions about their internal psychological process (cognitive and affective appraisal) and their purchase intention within each condition.
Appendix D contains the advertisements that were used for this survey. Study 2 investigates all six hypotheses.
As explained in the literature review, Chen et al. (2017) researched the effects of social commerce constructs in combination with the SLT. This study investigates the effect of an additional social commerce construct, social advertising. The dependent variables of the current study (affective and cognitive appraisal and purchase intention) are the same as the study conducted by Chen et al. (2017).
Table 3
2x2 factorial design online experiment
36 4.2 Operationalization of the variables
The experiments consist of an independent variable, two mediators, one dependent variable, one moderator (only experiment 2), and control variables.
4.2.1 Independent variable – social advertising (user-generated vs. brand-generated)
The independent variable in this study is social advertising. The respondent will be shown a user-generated or brand-generated social advertisement to test the direct effect of the type of social advertisement on the internal psychological process and purchase intention of the consumer. The user-generated social advertisements describe that the fictional brand advertised pictures made by consumers, and the users were tagged in the advertisement.
Two fictional brands were used in the experiment to prevent bias if the participants already knew the brand. The brand used in study 1 sells sunglasses (Opticz), and the brand used in study 2 sells backpacks (Bagpack). Those two products were specifically chosen because they attract both females and males. In the second experiment, which looks at the moderation effect, all respondents were shown two advertisements - one with a male and one with a female. Those two pictures were used to ensure that the respondents could identify with the social advertisements. The manipulation was checked with a manipulation check question. The created social advertisements for Bagpack can be seen in the survey for study 1 in Appendix C, and the advertisements for Opticz can be seen in the survey for study 2 in Appendix D.
4.2.2 Mediator – cognitive appraisal
One of the two mediators in this study is cognitive appraisal. As explained earlier, cognitive appraisal is based on the utilitarian aspects of attitude (Lee & Kozar, 2009). In both experiments, the cognitive appraisal of the respondents (after being exposed to the social advertisements) is measured using four statements by Martín-Consuegra et al. (2019) on a 7- point Likert scale. Martín-Consuegra et al. (2019) researched the relationship between brand involvement in the context of luxury brand-related activities on social media and the moderating
37 effect of hedonic and utilitarian motivations. First of all, their research uses social media as a platform; this offers a good fit for this research. Moreover, the hedonic and utilitarian motivations they use can be compared to cognitive and affective appraisal in this study. The statements are shown in Table 4. After answering questions about the cognitive appraisal, the participants were asked about their affective appraisal of the social advertisement.
4.2.3 Mediator – affective appraisal
The other mediator in this study is affective appraisal. Affective appraisal is based on a consumer’s emotions, feelings, and gut reactions (Lee & Kozar, 2009). The affective appraisal of the respondents is measured by four statements which are, again, based on research of Martín-Consuegra et al. (2019) on a 7-point Likert scale. Lastly, the participants were asked questions to measure their purchase intention.
4.2.4 Dependent variable – purchase intention
The dependent variable in this study is purchase intention. The purchase intention of the participants is measured by answering four statements by Duffett (2015), also on a 7-point Likert scale. Duffett (2015) researched the influence of advertising on Facebook on Millennials’
purchase intention and whether this is effect is influenced by usage and demographic variables and characteristics. Since Duffett (2015) researched the purchase intention of social advertising, the four statements fit this current research well.
4.2.5 Moderator – advertised product (hedonic vs. utilitarian)
The moderator in this study is the type of advertised product, which can be divided into utilitarian and hedonic. The moderation will be manipulated by showing one-half of the respondents the utilitarian sunglasses and the other half the hedonic sunglasses (Khan & Dhar, 2006), as shown in Table 5. Khan & Dhar (2006) researched whether a prior choice, which activates and boosts a positive self-concept, subsequently licenses the choice of a more self- indulgent option. One of their studies operationalized hedonic versus utilitarian by giving their
38 participants a choice between two sunglasses. ‘Sunglasses A’ were expensive and hedonic, and
‘sunglasses B’ were cheaper and utilitarian. A pretest confirmed this. Because this operationalization was already pretested by Khan & Dhar (2006), there is no need to pretest this operationalization in this current study.
Table 4
Operationalization of the variables
39 Table 5
Stimuli for experiment 2 (Khan & Dhar, 2006, p. 266)
4.2.6 Control variables
Next to the variables needed for the experiment, multiple control variables were included in the survey. First of all, the variable gender. In the study of Chen et al. (2017), on which this study is partly based, only one of their control variables appeared significant, which was gender.
The other variables, age, education, and occupation, were insignificant. However, this study will include the control variable age because, in this current setting, it might provide significant insights. Instead of using education and occupation, this study will investigate the variable income per household. This variable was chosen because it is important to know whether income influences purchase intention. Additionally, the respondents will be asked where their home is located. This question was specifically chosen because the respondents for the experiment will be collected using Prolific, which is not constrained to a specific country.
Similarly, the respondents will be asked questions about their Instagram usage. It is a prerequisite to have an active Instagram account in order to participate in the study. The respondents are asked how many times a day they use Instagram and whether they have purchased a product via Instagram. Lastly, this study also includes a seriousness check and an
40 attention check. According to Aust et al. (2013), a seriousness check improves the validity of the data. An attention check allows differentiation between high-quality and low or unreliable responses and does not compromise the scale validity (Kung et al., 2018).
4.3 Data collection
The survey was created with Qualtrics. This platform allowed the creation of surveys with different experimental conditions and, more importantly, random assignment of the respondents in the different conditions. Moreover, Qualtrics allows for a direct transfer to SPSS with which the data will be evaluated. Both surveys were written in English. The surveys had the following order: the intro of the experiment (which included the requirement that the participant should have an active Instagram account), informed consent, the social advertisement(s), affective appraisal, cognitive appraisal, purchase intention, attention question, manipulation check question, demographic questions, Instagram usage, and, lastly, a seriousness check. Both surveys took approximately 5 minutes to complete. The survey questions for study 1 can be found in Appendix C, and for study 2 in Appendix D. The respondents for both studies were collected with Prolific. Prolific makes sure that the study's internal validity is high because the surveys reach a high diversity of respondents. As a result, these responses are of high quality.
5. Results
In this chapter, the results of the two conducted experiments are discussed. First, the data is cleaned, after which the general data about the respondents will be discussed. After that, the variables will be checked on normality distribution, outliers, and reliability. Next, the correlation, manipulation check, and hypotheses are being tested. Finally, the chapter ends with a summary of both experiments.
41 5.1 Study 1
5.1.1 Data cleaning
A total of 233 respondents completed the survey through Prolific. Of those respondents, five did not have an active Instagram account and could thus not complete the survey. Six respondents were removed from the dataset because of what they filled in at the seriousness check (one respondent ‘not at all’, and five respondents ‘only a little’). In addition, three outliers of affective appraisal statement 3 were below -3 and were removed. After removing the missing data, the seriousness check, and the outliers, the final set comprised 219 respondents. The randomizer was performed well since almost all variables are normally distributed. The results do not include the attention check since it did not influence the outcomes.
5.1.2 General data respondents
Appendix E, Table 16, gives an overview of the demographic data of both studies. Below is given a summary of the data. There is almost an equal distribution between males (48.9%) and females (48.4%). More than half of the respondents (59.8%) are aged 18 to 24 years. This result could be due to the platform that was used to collect the respondents, which students often use as a source of income. More than half of the respondents (61.2%) are located in Europe. A little less than half of the respondents (42%) earn less than €25.000, after which the largest group is €25.000 - €49.999 (31.1%). The distribution across Instagram usage is fairly equally distributed; 2 – 5 times a day (33.8%), 5 – 10 times a day (25.6%), and 10+ times (24.7%). About half of the respondents have never purchased a product via Instagram (55.7%), and about one-third of the respondents one to two times (30.1%). The descriptive statistics of the variables can be seen in Table 6. Affective and Cognitive Appraisal are both distributed almost the same. However, purchase Intention has a lower mean in comparison to Cognitive and Affective Appraisal.
42 Table 6
Descriptive statistics study 1
Note. N = 219.
5.1.3 Skewness and Kurtosis
According to Hopkins & Weeks (1990), skewness represents the asymmetry of the data distribution. The kurtosis values represent deviations from a normal distribution. If the skewness and kurtosis are close to 0, it can be concluded that the data is normally distributed (Hopkins & Weeks, 1990). The skewness and kurtosis of Affective Appraisal, Cognitive Appraisal, and Purchase Intention were all between -1 and +1 (Table 17, Appendix F1). These results show that all three constructs are normally distributed, and no further actions need to be taken.
5.1.4 Outliers
Osborne & Overbay (2004) showed in their research that removing extreme scores in the data can improve the significance of the results. The outliers were calculated using the z-score, which shows how far the data point is from the mean. The only item that contained outliers was the third item of Affective Appraisal (i.e., ‘I perceive this social advertisement as pleasant’).
This item contained three outliers that were outside of z > ï3 (z = -3.378). Those three outliers were excluded from the dataset.
43 5.1.5 Reliability
Before calculating the reliability, the items were transformed into constructs by calculating the mean. The Cronbach’s Alpha measures the internal consistency of multiple items in a construct (Cronbach, 1951). A good scale has an a-value of > .70. Affective Appraisal has high reliability, with Cronbach’s Alpha a = .844. The corrected item-total correlations indicate that all the items have a good correlation with the total score of the scale (all above .30). Moreover, three of the four items would reduce the reliability if the items were removed. One item would improve the reliability, but not more than > .10, which is why the item is not removed.
Cognitive Appraisal has high reliability, with Cronbach’s Alpha a = .920. The corrected item-total correlations indicate that all the items have a good correlation with the total score of the scale (all above .30). Moreover, all four items would reduce the reliability if the items were removed.
Purchase Intention has high reliability, with Cronbach’s Alpha a = .930. The corrected item-total correlations indicate that all the items have a good correlation with the total score of the scale (all above .30). Moreover, all four items would reduce the reliability if the items were removed.
5.1.6 Correlation test
A correlation analysis demonstrates the correlation between two or more quantitative variables (Gogtay & Thatte, 2017). The correlation matrix offers information about the strength and direction of the correlation between variables. In order to perform the correlation matrix, the variables gender, social advertising, and annual household income had to be dummy coded.
Gender is coded as 0 = male and 1 = female. Social advertising is coded as 0 = male and 1 = female. The annual household income had option 6 (i.e., ‘Prefer not to say’); this was transformed into a missing case since it is an ordinal variable.
44 Note. N = 219. Statistical significance: ***p < .001, **p < .01, *p < .05. Social Advertising is coded as 0 brand-generated and 1 user-generated.
Gender is coded as 0 male and 1 female.
