Between Two Giants: Advertising Effectiveness on Facebook and Twitter.
-Master’s thesis-
-Graduate School of Communication-
Author: Giacomo Riccobono
Student ID -10877754
Supervisor: Dr. Edith Smit
Master’s Programme: Communication Science
Track Specialization: Persuasive Communication
Abstract
Building on previous research on differences between communication media, the present study investigates how advertising either on Facebook or Twitter can have different effects on consumers’ attitudes. Moreover, the influence of three other variables are taken into
account; perceived interactivity is expected to be the underlying mechanism, while product involvement and social media usage are plausible moderators of the main relationship. Although the results of the experiment (N = 173) do not support the predicted effects, two interesting findings deserve attention for future studies in this field. Perceived interactivity do not mediate the main relationship, however, it significantly affects both attitude toward the ad and attitude toward the social media platform. Furthermore, the effect on attitude toward the social media platform is significantly greater on Facebook than on Twitter. No interaction effect have been found, yet product involvement significantly predicts attitude toward the ad whereas social media usage significantly predicts attitude toward the social media platform. The findings suggest that advertising effects are more determined by users’ differences than social media platforms.
Introduction
Nowadays there is no doubt that any company or organization, which aspires to be influential in the market, must convey its message/s in the most efficient way possible. The marketing revolution followed by the advent of the internet has definitely changed the
communication scenery. Traditional media, such as television and print, seem to be no longer the best and only way to advertise; instead, companies and organizations are increasingly reaching out to their audience through social media platforms (Fournier & Avery, 2011). According to the Social Media Marketing Industry Report (Stelzner, 2015), more than 90% of the companies use social media for marketing reasons and 92% are aware of the potential benefits deriving from these means. Such data seem to be proved by the constant growth in share of total online advertising spending in the last decade. Indeed, it has been registered a total spending of $23.68 billion in 2015 by advertisers worldwide, which corresponds to an increase of 33.5% compared to 2014 (Emarketer.com, 2015). In addition, although Facebook is still the first platform chosen by companies to advertise, the percentage of companies using Twitter raised from 46% in 2011 to 79% (see Figure 1) in 2015 and it is expected to increase more in the next years.
Figure 1. Percentages of social media platforms used by companies to advertise
0% 20% 40% 60% 80% 100%
Google+ Linkedin Twitter Facebook
Another interesting finding from Stelzner report (2015) regards the difference between ‘business to consumer’ (B2C) and ‘business to business’ (B2B) communication for marketing
activities. Indeed, it seems that the platform Linkedin is the most used for B2B
communication, however when it comes to communicate with the consumers, Facebook is the indisputable leader followed by Twitter. Based on the arguments provided so far, Facebook and Twitter are the social media platforms taken into account in this study which aims at investigating the effects that advertising on different platforms have on consumer outcomes. It is hypothesized that differences between Facebook and Twitter in terms of their objective characteristics are likely to lead to different attitudes toward both the ad and the social media platform. Furthermore, previous studies (Liu & Shrum, 2002; Song & Zinkhan, 2008) have extensively demonstrated the influential role played by perceived interactivity in conditioning advertising effectiveness. Hence, the present study assumes that different levels of
interactivity perceived by the users of the two platforms might be the underlying mechanism. However, it seems reasonable to assume that also other factors are likely to play an important role in attitude formation toward online advertising. Specifically, product involvement and social media usage are expected to moderate the main relationship. In line with a study conducted by Liu and Shrum (2009), it is predicted that depending on the degree of involvement with the product advertised and the experience (social media usage) with the social media platform, consumers will generate different attitudes.
In order to formulate the hypotheses of the study, an in-depth investigation of Facebook and Twitter characteristics is hereafter proposed. Furthermore, the following section will deal with the available literature concerning interactivity, product involvement and social media usage.
Theoretical background
The main aim of this study is to investigate the extent to which advertising either on Facebook or Twitter can influence consumers’ attitudes toward the ad and toward the
platform itself. Based on previous studies (Dijkstra, Buijtels, & Van Raaij, 2005), which were mainly focused on differences among the so-called traditional media (e.g. print, television, radio), it seems that each medium with its characteristics and peculiarities is likely to yield different outcomes. In particular, it is still current the theory of McLuhan according to which “the medium is the message”; indeed, the characteristics of the medium set the receiver in a
certain forma mentis which influence his reaction at a later stage (Levine & McLuhan, 1964). Furthermore, some objective media characteristics have been outlined and recognized as critical in shaping consumers’ responses to advertising. In line with these findings, it is
expected that a similar dynamic among the social media would exist, for instance between Facebook and Twitter.
Although the term social media is nowadays familiar to anyone, both marketers and academic researchers are still confused about what social media exactly are and what can be included in this category. According to Kaplan and Haenlein (2010, p. 61), social media can be defined as “a group of internet-based applications that build on the ideological and
technological foundations of Web 2.0, and that allow the creation and exchange of User Generated Content.” In addition, starting from the assumption that “there is no systematic way in which different social media applications can be categorized”, Kaplan and Haenlein (2010,
p. 62) have tried to come up with a first classification (see Figure 2). In their attempt, they have also tried to create a systematic framework able to comprehend both existing and forthcoming platforms. Thanks to their work it is possible to analyze both Facebook and Twitter under a theory-based perspective.
Social presence / Media richness
Low Medium High
Self- presentation/
Self- disclosure
High Blogs Social networking
sites Virtual social world Low Collaborative projects Content communities Virtual game worlds
Figure 2. Classification of social media by social presence/media richness and self-presentation/self-disclosure (Kaplan & Haenlein, 2010 p. 62).
A comparison between Facebook and Twitter
According to this social media classification it is possible to place Facebook and Twitter in two different categories. Indeed, although not mentioned in the figure, Twitter can be logically placed in the category of blogs since it is defined as a micro-blogging service. Therefore, a first distinction between Facebook and Twitter can be made on the basis of their nature, the former is a social networking site whereas the latter a micro-blogging service. Further, this classification is based on the combination of two theories from the media
research field, social presence theory and media richness theory, and two theories from social processes field, self presentation theory and self disclosure theory.
Self presentation is defined by Naegele and Goffman (1956, p. 61) as “the desire people have to control the impression that other people form of them when interacting”. For
instance, in the context of social media it can be said that people tend to create personal profiles in order to present themselves in the Internet environment. Such a presentation occurs through self disclosure, that is the act to share personal information such as thoughts and feelings (e.g. a tweet on Twitter or a post on Facebook) with others. By looking at Figure 1 and according to the definitions just provided, it would seem that under this perspective Facebook and Twitter do not differ too much from each other. The two platforms share both
the self presentation function, people spontaneously decide to register, and the self disclosure aspect since people decide to share their personal information.
As mentioned before, the other dimension of the Kaplan and Haenlein’s classification is more focused on the medium and composed of two theories, social presence and media richness theory. The former states that media differ in the “degree of contact – either this
being acoustic, visual or physical - they allow to emerge between the communication partners” (Parker, Short, Williams, & Christie, 1978, p. 78). The latter is strictly related and posits that “the ultimate goal of any communication is the resolution of ambiguity and the reduction of uncertainty” (Daft & Lengel, 1986, p.563). According to Figure 1, under this
perspective blogs (Twitter in this context) are less effective than social networking sites, because they are only text-based. On the other hand, Facebook is placed in the upper level since it enables users to share pictures, videos, and other forms of media. However, such a differentiation is not convincing because it does not take into account all the characteristics of Twitter which will be reported in the next section. In this specific case it can be argued that comparing Twitter with the blogs of Kaplan and Haenlein’s classification might be
misleading.
According to Kietzmann, Hermkens, McCarthy and Silvestre (2011), it is possible to further define social media based on seven functional building blocks, namely: identity, conversation, sharing, presence, relationship, reputation, and groups (see Figure 3).
Concept Definition
Identity It refers to the revelation of personal information (it can be compared with the already mentioned self disclosure concept).
Conversation The extent to which users communicate with other users in a social media setting.
Presence The extent to which users can know if other users are accessible (e.g. ‘Friends Around Me’ for both Facebook and Twitter).
Relationships The extent to which users can be related to other users.
Reputation The extent to which users can identify the standing of others, including themselves, in a social media setting. (e.g. ‘friends’ on Facebook, ‘followers’ on Twitter).
Groups The extent to which users can form communities and sub-communities. (e.g. Twitter has lists, Facebook groups).
Figure 3. Definitions for the seven functional building blocks of social media
Although similar under many of these aspects, we retain Facebook and Twitter to be different about ‘conversation’ and ‘relationship’. According to the definition provided in Figure 2, it
can be inferred that the type of conversation vary depending on the social media setting. Diverse reasons drive people to communicate with each other on social media platforms, for instance the desire to meet people or the need to make their voice heard (Beirut, 2009). For instance, from a report of Kaplan and Haenlein (2011) it seems that the communication happening on Twitter is more about conversation than identity. Indeed, Twitter is centered around ‘tweets’ of maximum 140 characters which are most of the time of an ephemeral
nature and do not necessarily need a response. On the other hand, we believe Facebook posts to be slightly different and more often based on identity rather than conversation. These posts do not have any length limit, so allowing users to express their thoughts in a less restrictive way. Kietzmann, Hermkens, McCarthy, and Silvestre (2011) conclude that this diversity of conversations translate into format and protocol implications for firms which, for marketing related reasons, want to host or track these conversations.
The second block taken into account is ‘relationships’, that is the various forms of connection among users available on social media platforms. Such a connection is generally crucial for determining the what-and-how of information exchange. For instance, in the case
of Twitter, the ‘relationship’ aspect does not matter too much since every user is free to ‘follow’ whoever he wants without any restriction. On the contrary, Facebook is based on a reciprocal friendship concept, which “means that one person has to send a friend request or an
invitation to join Facebook and the other person has to confirm this request in order for them to become friends” (Facebook.com, 2015). Therefore, it can be assumed that Facebook and
Twitter differ from each other in terms of the meaning and value given to the relationships in these platforms. Once again, according to Kietzmann et al. (2011) such a difference might create implications for firms seeking to engage with their users.
The research conducted by Dijkstra, Buijtels and Van Raaij (2005) may be useful as well in order to distinguish Facebook and Twitter in terms of their objective characteristics. Although their study focused on differences between traditional media (e.g. television and print) and the Internet, it proposes a good way to outline media characteristics and
consequently it suits well in this context. They specifically looked at modality of the medium and control over the medium. Modality refers to the mode of presentation of the medium (e.g. text, audio, video) which is assumed to “affect processing by directly evoking cognitive and
affective reactions or indirectly influencing the processing of and reactions to other sensory modes” (Dijkstra, Buijtels & van Raaij, 2005, p. 378). For instance, to read a newspaper it is
required only one sensory mode, the eyes, whereas when watching the television it is also important to listen. No previous studies have compared Facebook and Twitter under this perspective, however we retain them rather similar since both are mainly text media. The same argument may apply for control over the medium. Facebook and Twitter can be
considered media with internal pacing since the consumer freely decides whether to attend the ad, and can process the information at his own pace and sequence. In this circumstance, involvement and motivation to process the ad play an important role on consequent consumers’ attitudes, however this issue will be dealt with later in this section.
In this first section an analysis of the characteristics of Facebook and Twitter have been provided in order to find differences and similarities between these two platforms. To recap, with the support of the Kaplan and Haenlein classification and the theories from media research and social processes field, a comparison of Facebook and Twitter has been proposed. In conclusion, based on the fact that the two platforms present similarities but also
differences, the first hypothesis of this study is formulated as follow.
H1: Depending on the social media platform used to advertise (Facebook or Twitter),
consumers will score differently on attitudes toward the ad and toward the social media platform.
Perceived interactivity
Companies are more and more relying on social media tools to advertise their brand and products, implying an important economic investment for them and a need for researchers to investigate the underlying mechanisms that determine consumers’ responses to this type of
advertising (Ta k ran & Y lmaz, 2015). A common element of the numerous definitions of social media is the conversation aspect, implying that a core characteristic of social media is its level of interactivity (Romero & Fanjul, 2010). Based on this assumption and previous studies (Yaakop, 2013), this study assumes that interactivity might be the underlying mechanism, which shapes consumers’ attitudes toward the ad and toward the social media platform. According to Liu and Shrum (2002, p. 57), interactivity can be defined as “the degree to which two or more communication parties can act on each other, on the
communication medium, and on the messages and the degree to which such influences are synchronized”. Although the term interactivity has been widely used in the communication
field even before the advent of social media platforms, researchers agree that it is the factor distinguishing traditional and new media (Morris & Ogan, 1996; Chung & Zhao, 2004).
A fundamental distinction needs to be made between actual and perceived
interactivity. The former refers to objective measurements of interactivity and it is also called feature-based interactivity (Song & Zinkhan, 2008). The above-mentioned objective
measurements may be for instance the observation of the number and type of interactive features on a Web site. The latter is the interactivity “subjectively experienced by the users and therefore often referred to as experiential interactivity” (Voorveld, Neijens & Smit, 2011, p. 77). Measurements for this type of interactivity might be questions about feelings or experiences of the consumers during their visit to the Web site.
In line with a study conducted by Song and Zinkhan (2008), it is assumed in this study that it is the Website’s perceived interactivity that affect consumer responses. Indeed, in their study they concluded that what matters is not the number of interactive features in a Website, yet their effectiveness which eventually enhance the users’ interactivity perception.
Furthermore, it has been proved that websites containing the same actual interactive functions were then evaluated differently by users in terms of perceived interactivity (Lee, Park, & Jin,2006).
A few attempts (Johnson, Bruner, & Kumar, 2006; Liu, 2003; McMillan & Hwang, 2002; Wu, 2006) have been made in order to identify the dimensions defining the perceived interactivity construct. Yet, there would seem to be common ground on the three dimensions identified by Liu and Shrum (2002), which are active control, two-way communication, and synchronicity. Active control is characterized by voluntary and instrumental action that directly influences the controller’s experience (Liu & Shrum, 2002). Some navigational tools
such as hyperlinks, site maps, and search options, or the capability to customize products or information on the Web site have been identified as facilitators of this dimension (McMillan & Hwang, 2002; Song & Zinkhan 2008). Two-way communication refers to the ability for reciprocal communication between companies and users, and users and users (Liu & Shrum,
2002). This dimension of perceived interactivity can be facilitated by features such as feedback forms, Web logs, chat rooms, surfer postings, e-mail links, and online ordering facilities (Liu & Shrum, 2002; McMillan & Hwang, 2002; Song & Zinkhan, 2008). Synchronicity refers to the degree to which users’ input into a communication and the
response they receive from the communication are simultaneous (Liu & Shrum, 2002). This last dimension of perceived interactivity can be improved by features that reduce the space-time between the user and the Web site.
To recap, since different levels of perceived interactivity are likely to affect consumers’ responses toward advertising, it is now necessary to analyze Facebook and
Twitter under this perspective, taken into account the three dimension just defined. As mentioned earlier, to our knowledge no previous studies have investigated such a topic comparing these two media. Therefore, an attempt of comparison between Facebook and Twitter is made in this study by looking at the features of the three dimensions constituting perceived interactivity. Facebook Twitter Active control Hyperlinks Site maps Search options Customize info Two-way communication Feedback forms Web logs Chat rooms Surfer postings Email links
Online ordering facilities
Synchronicity Space-time user platform
Figure 4. Comparison between Facebook and Twitter in terms of their interactive features.
From an accurate analysis of the features of the three dimensions of perceived interactivity, it would seem that Facebook and Twitter share the same characteristics. Consequently, the attempt of comparison made in Figure 4 may be useless since it does not highlight any difference between the two platforms. Yet, it gives reason to expect Facebook and Twitter outcomes to be similar. However, as mentioned previously, recent studies have challenged the traditional theories about interactivity concluding that “simply adding features does not guarantee a high level of perceived interactivity” (Song & Zinkhan, 2008, p. 109).
Therefore, even though the two platforms seem to be similar in terms of interactive features, it is not straightforward to conclude that they will be perceived as interactive in the same way. In conclusion, although different views are available, we expect a different pattern of outcomes for Facebook and Twitter related to their perceived interactivity.
H2: The consumers’ perceived interactivity of the social media platform will significantly
differ between Facebook and Twitter.
In addition, this study predicts perceived interactivity to be the underlying mechanism influencing the consumers’ attitudes. A third hypothesis is formulated on the basis of the
theory proposed hereafter.
According to a study conducted by Mangold and Faulds (2009), advertising on social media platforms influences the consumers’ behavior as well as their opinions and attitude formation. The authors also found a correspondence between high levels of perceived interactivity and the elaboration of positive attitudes toward the advertisement. In another study, Coyle and Thorson (2001) manipulated some features such as video and audio and
observed that interactive web ads were more effective in terms of consumers’ attitudes.
Again, McMillan, Hwang and Lee (2003) investigated the effects of perceived interactivity on attitude toward the website concluding that the more interactive users perceive a website to be, the more positive attitudes they will have toward the website.
Furthermore, in line with a recent study conducted by Leung, Bai, and Stahura (2013), in this study attitude toward the ad and attitude toward the social media platform are both taken into account. Indeed, attitude toward the website has been widely recognized to be as important as the attitude toward the ad when evaluating the advertising effectiveness (Cheng & Wells, 1999). It is important to specify that the majority of the studies so far conducted have specifically looked at the variable attitude toward the website. However, since social media are types of websites we do retain reasonable based our assumptions oh these studies although using a different terminology.
In conclusion, the expectation of the current study is a positive linear relationship between perceived interactivity and attitudes formation.
H3: Higher levels of perceived interactivity will lead to more positive attitude toward the ad
and attitude toward the social media platform.
Involvement and social media usage
The main goal of this study is to investigate the advertising effectiveness of the two social media platforms Facebook and Twitter. As reported in the previous paragraph, it is expected that perceived interactivity will act as the underlying mechanism influencing attitudes toward the ad and toward the social media platform.
did not find homogenous results, because there would be other factors moderating the effect of interactivity on advertising effectiveness measures. Especially, involvement with the product advertised has been widely recognized to be a critical variable in influencing
consumers’ responses in the advertising environment (Chung & Zhao, 2004). In addition, it is
common practice for researchers dealing with attitude formation and change to use the
Elaboration Likelihood Model (Petty, Cacioppo, & Schumann, 1983). This is also the case for specific studies, for instance the one conducted by Lukka and James (2014), which looked at the formation of attitudes in the social media advertising environment employing the ELM. According to the ELM, involvement is defined as “the extent to which the attitudinal issue under consideration is of personal importance” (Petty, Cacioppo & Schumann, 1983, p. 847). For instance, in the case of this study where the attitudes toward an advertisement promoting a product are measured, it is presumable that the level of involvement users have with the product itself will affect their evaluations. In this regard, the model proposed by Petty and Cacioppo (1983) aims at explaining the formation of attitudes through a dual process theory. In particular, under high involvement conditions consumers use to form and change attitudes through the so-called central route. In such a situation, the individual evaluates as relevant the issue or the product in question and devote a great amount of cognitive effort in order to carefully evaluate it. On the other hand, when lowly involved with an issue or a product, consumers tend to process the information and elaborate attitudes through the peripheral route. In this case, it is likely that the arguments presented will not be intensely scrutinized and some information may also be overlooked.
However, not only the involvement with the product would be the reason for different outcomes. Logically, and according to the Liu and Shrum postulate (2009), also person factors such as the level of experience (social media usage) with the platform in which the product is advertised are likely to influence consumers’ attitudes. Specifically, when lowly
involved, consumers will be likely to do not pay attention to interactive features and positively evaluated both the ad and the social media platform regardless of their level of usage. On the contrary, under high involvement conditions (central route process) the level of usage will be determinant for the formation of attitudes. Being difficult for them ‘dealing’ with the platform, inexperienced users are expected to reduce the focus on central arguments and form negative attitudes. Experienced and highly involved users will be likely to form positive attitudes.
Based on their study, we propose a similar pattern in this context according to which both involvement and social media usage may moderate the relationship between advertising on social media platform and the consequent attitudes’ elaboration. In order to schematically sum up what has been said so far and to formulate the last hypothesis of this study, a revisited model from Liu and Shrum study is hereafter proposed (see Model 1).
H4: Individuals who are highly involved with the product advertised will have more positive
attitudes toward the ad and the social media platform when more experienced with the platform (high user experience).
high low
high low high low
Model 1. A model of interaction between involvement/social media usage and attitudes
Method section
Design and procedure
To test the four hypothesis a one factor between-subjects design experiment with two conditions was conducted. The participants were randomly assigned to either the Facebook or the Twitter condition. In both conditions, the same advertisement was displayed. While completing the survey, participants were asked to view the content as normally as possible to experience the context of the medium, and to avoid time constraints. In the beginning of the survey a set of instructions on how to proceed with the experiment were given to the
participants. Afterwards, they engaged with the stimuli followed by measurements of dependent variables, mediating variables, moderating variables and finally demographic variables.
INVOLVEMENT
Social media usage Social media usage Enhanced central processing Reduced focus on central argumentsNo effect on central processing
The experiment was carried out online, the participants decided spontaneously to take part in it by clicking the Qualtrics web-link spread out during the two last weeks of December 2015. The participants’ recruitment has been done through a convenience sampling method. The researcher and his colleagues spread the link of the survey among their social network. For legal reasons, only responses from participants aged 18+ were taken into account.
Sample
A total of 220 participants participated. Forty-seven participants were excluded from the analysis because either they did not complete the survey or they did not accept the terms and conditions of the survey. Therefore, the final sample consisted of 173 individuals, of which 51.4% were female. The mean age of the sample was 25.69 (SD = 4.87) years, ranging from 18 to 58. Males (M = 25.96, SD = 3.07) and females (M = 25.44, SD = 6.12) had a similar age on average. A similar mean age was observed also between conditions (see Table 1). The majority of the respondents were Italians (53.8%), yet there were other 31 countries represented in the sample. In regards to the participants level of education, which was
categorized into six levels, almost the totality of the participants (99.4%) had completed ‘high school’, 30.1% of the sample declared to have a ‘master’s degree’. Such data were expectable
since the respondents were gathered through snowballing method by the researcher.
Mean Standard Deviation N
Age
- Twitter 25.56 4.64 88
- Facebook 25.84 5.15 85
Table 1. Mean age for Facebook and Twitter conditions
Since the aim of the study was to investigate the effect of advertising in to different social media environments, the stimulus materials of the experiment consisted of two long images representing a Facebook as well as a Twitter page. The two images, elaborated with the software Photoshop, accurately reported all the functions of the two social media and were composed of regular content such as: friends’ posts, tweets, news, and sponsored content.
Both a fictitious profile name (John Red) and a profile picture were used to avoid any kind of bias. The manipulation consisted of a Nespresso advertisement which was added in the middle of both pages (see Appendix 1 and 2). According to a study conducted by Voorveld et al. (2013), the brand Nespresso was chosen for the purpose of the study. Indeed, it was expected to lead to the formation of two groups: highly and lowly involved consumers.
Measures
Dependent variables
Attitude toward the ad. According to previous studies (Smit, 1999; Tutaj & Van Reijmersdal, 2012), attitude towards the ad is composed of three beliefs, which are information, amusement and irritation. In order to measure the variable, a battery of four items taken from a study conducted by Dijkstra, Buijtels, and Van Raaij (2005) was used. A general pre-question stated: “I find the advertisement that was displayed on the social media platform”: (1) “clear/confusing”, (2) “uninteresting/interesting”, (3) “appealing/unappealing” and (4) “dislikable/likable” (Tutaj & Van Reijmersdal, 2012; Dijkstra, Buijtels & Van Raaij
2005). All four items were measured on a scale ranging from 1 to 7. After having reverse recoded items 2 and 4, the attitude toward the ad scale was created by averaging the items. The scale resulted to be reliable with a Cronbach’s alpha of .79 (M = 3.87, SD = 1.28), and no need to withdraw any item from it to improve it.
Attitude toward the social media platform. In order to measure the attitude toward the
social media platform, a revisited scale by Chen and Wells (1999) was implemented. For the purpose of the study, the original term ‘website’ was replaced by ‘social media platform’. The
final five items were: (1) “I feel comfortable in surfing the social media platform”, (2) “I feel surfing the social media platform is a good way for me to spend my time”, (3) “I would like to visit the social media platform again in the future”, (4) “I am satisfied with the service
provided by the social media platform”, (5) “The social media platform makes it easy for me
to build a relationship with this company”. All five items were measured on a scale from 1 (strongly disagree) to 7 (strongly agree). The reliability analysis showed good levels of Cronbach’s alpha .77 (M = 4.49, SD = 1.05), therefore the items were averaged to compute
the attitude toward the social media platform scale.
Mediating variable
Perceived interactivity. Based on a study conducted by Voorveld, van Noort and Duijn
(2013), perceived interactivity was measured with eight items taken from the scales proposed by Liu (2003), and Song and Zinkhan (2008). Accordingly to the argumentation given in the theoretical background, all the three dimensions of perceived interactivity were taken into account. The first three items measured the two-way communication dimension: (1) “The social media platform gives me the opportunity to talk back”, (2) “The social media platform is effective in gathering visitors’ feedback”, (3) “The social media platform makes me feel it wants to listen to its visitors”. Similarly, three items were used to measure active control: (4) “The social media platform is manageable”, (5) “While I was in the social media platform, I could choose freely what I wanted to see”, (6) “I feel that I have a great deal of control over
my visiting experience at this social media platform”. The last dimension, synchronicity, was measured with two items: (7) “The social media platform processes my input very quickly”, (8) “I was able to obtain the information I wanted without delay”. For all the eight items,
response categories ranged from 1 (strongly disagree) to 7 (strongly agree). The reliability analysis revealed a Cronbach’s alpha of .82 (M = 4.60, SD = .90) which is good, hence the eight items were averaged to represent the perceived interactivity scale.
Moderating variables
Social media usage. Social media usage was measured as the amount of time an
individual spends on the social media platform each week. The respondents were asked “How many hours do you spend per week on these two social media platforms?”. They could
indicate the exact number of hours by moving an indicator on a slider bar ranging from 0 to 100. Actual scores observed ranged from 0 to 87. After performing a median split for both respondents in the Facebook (Mdn = 10; M = 14.97, SD = 14.85) and in the Twitter (Mdn = 1;
M = 4.54, SD = 7.23) condition, the high and low social media usage groups were created.
Involvement. The variable ‘Involvement with the product advertised’ was measured,
and not manipulated, taking into account two aspects: the experience as well as the importance of the product for the respondent. According to purpose of the study and the definition provided by Zaichkowsky (1994, p.61) – “respondents’ overall evaluation of how important the product is to their life” –, two questions were asked: (1) ‘How often do you use Nespresso products?’ and (2) ‘To me Nespresso is’. For both items the response categories ranged from 1 to 5. In the first question 1 meant ‘never’ whereas 5 ‘all of the time’, while in
the second question 1 corresponded to ‘unimportant’ and 5 to ‘important’. The involvement scale was computed by averaging the two items, resulting in a good level of Cronbach’s alpha of .83 (M = 4.98, SD = 2.25). After performing a median split (Mdn = 2.50; M = 2.48, SD = 1.13), respondents were respectively assigned to the high or the low involvement group.
Results
In order to ensure that the results of the study were not influenced by other variables, a preparatory analysis was implemented. Statistics showed that participants’ gender was similar across conditions (X2 (1, N = 173) = .28, p = .60) and the two conditions did not differ with respect to age (t (171) =-.37, p = .709). Furthermore, there were no significant differences between conditions regarding the participants’ educational level (X2 (3, N = 173) = 1.40, p = .71) (see Table 2). Consequently, we can assume that differences in the groups regarding the dependent variables cannot be caused by differences in these background variables.
Twitter Facebook Total
High school 23.9% 27.1% 25.4% Bachelor 44.3% 43.5% 43.9% Master 31.8% 29.4% 30.7% Total 100% 100% 100%
Table 2. Percentages of respondents based on their level of education
Hypothesis testing
As depicted in the conceptual model (Model 2), this study aims at testing both the mediation effect of perceived interactivity and the moderation effect of involvement with the product and the user’ experience with the platform. Since it is hypothesized that the
moderation will affect only the main effect, X Y, and neither X M nor M Y are moderated, this indirect effect is unconditional. That said, the hypothesis will be tested in two separate stages. First the mediation and then the moderation will be analyzed. In order to test the mediation, the three-step approach of Baron and Kenny (1986) is used, which allow us to determine whether each relationship is significant or not.
predictor of the dependent variables. A linear regression was conducted to predict attitude toward the ad, using Facebook and Twitter as predictors. Using Facebook or Twitter did not show to be a significant predictor of attitude toward the ad (β = -.01; t = -.127, p = .90). This was also confirmed by the mean values of the two conditions, respectively M = 3.85, SD = 1.25 for Facebook and M = 3.88, SD = 1.80 for Twitter. Similarly, a second linear regression showed that using Facebook (M = 4.64, SD = 1.02) or Twitter (M = 4.34, SD = 1.06) does not significantly predict attitude toward the social media platform (β = .14; t = 1.90, p = .06). Based on these results it can be assumed that hypothesis 1 is not supported.
In the second step, the aim of the analysis was to test whether using Facebook or Twitter significantly predicted consumers’ perceived interactivity. A linear regression was
conducted to predict perceived interactivity with Facebook and Twitter as predictors. The analysis showed that using Facebook (M = 4.61, SD = .93) or Twitter (M = 4.57, SD = .86) does not significantly predict consumers’ perceived interactivity (β = .03; t = .38, p = .70).
Therefore, the second hypothesis of the study is rejected as well.
In the third and last step, the role played by the mediating variable was tested as identified by hypothesis 3. Both the mediating variable (perceived interactivity) and the independent variable (social medium types) were regressed on the dependent variables (attitude toward the ad and attitude toward the social media platform). Perceived interactivity showed to be a significant predictor of both attitude toward the ad (β = .26; t = 3.54, p < .001) and attitude toward the social media platform (β = .56; t = 8.96, p < .001). Perceived
interactivity explained a significant part of the variance in scores on attitude toward the ad (R2 = .07, F(1, 173) = 12.56, p < .001) and attitude toward the social media platform (R2 = .32, F(1, 173) = 80.37, p < .001). As tested in the first step, the independent variable (social medium types) did not show to be a significant predictor of attitude toward the ad (β = -.02; t = -.23, p = .81). However, a different pattern of results was observed for attitude toward the
social media platform. Indeed, when perceived interactivity is included in the model, social medium types significantly predicts attitude toward the social media platform (β = .13; t = 2.04, p < .05). Specifically, the results revealed that more for Facebook (M = 4.64; SD = 1.06) than for Twitter (M = 4.34; SD = 1.02) higher levels of perceived interactivity lead to better attitude toward the social media platform.
The second part of the analysis consisted of testing whether the variables social media usage and involvement moderated the relationship between social media types and attitudes. Hence, a general linear model analysis was run which allow the researcher to check for interaction effects. It was first investigating the effect on one dependent variable, attitude toward the ad, and then the effect on the other one, attitude toward the social media platform. The analysis of variance did not show any significant interaction effects of involvement and social media usage (F(1,173) = .05, p = .81) on the main relationship between social media type and respondents’ attitude toward the ad. However, attitude toward the ad resulted to be
significantly affected by involvement (F(1,173) = 4.76, p < .05; M low involvement = 3.64, SD = 1.22; M high involvement = 4.09, SD = 1.30) in a way that the more involved with the product an individual is, the better his attitude toward the ad will be. A second analysis of variance did not show any significant interaction effects of involvement and social media usage (F(1,173) = .39, p = .53) on the main relationship between social media type and respondents’ attitude toward the social media platform. However, attitude toward the social media platform resulted to be significantly affected (F(1,173) = 7.02, p < .01; M low usage = 3.64, SD = 1.22; M high usage = 4.09, SD = 1.30) by social media usage in a way that the more experienced with the platform an individual is, the better his attitude toward the social media platform will be. Table 3 reports all the values of main and interaction effects of media, social media usage, and involvement toward the two dependent variables; values on the left refer to attitude
toward the ad, values on the right refer to attitude toward the social media platform.
A schematic representation of the overall findings, including different levels of significance, is proposed below in Model 2.
Source F Sig. F Sig.
Cond 0.75 .78 3.65 .58 Exp 25.56 .41 7.02 .009* Inv 25.84 .03* 2.81 .09 Cond*Exp 1.36 .24 1.31 .25 Cond*Inv 1.06 .30 1.18 .27 Exp*Inv .39 .52 1.81 .18 Cond*Exp*Inv .05 .81 .356 .53
Table 3. Interaction and main effects of media (Cond), involvement (Inv), and social media usage (Exp) on attitude toward the ad (values on the left) and attitude toward the social media platform (values on the right)
not significant p < .01 p < .01
p <.05
p <.01
Model 2. Conceptual model with levels of significance.
Conclusion
Decades of studies focused on the differences among media in terms of their
effectiveness as advertising tools have produced important results. Researchers agree on the assumption that media objective characteristics are likely to shape consumers’ responses. McLuhan (1964) emphasized even more the importance of an accurate choice of the medium, proposing the famous statement “the medium is the message”. The majority of the studies
looked at differences among traditional media, little has been said about new media such as social media platforms. This study attempted to investigate such an argument trying to answer three questions.
The first question was, are there differences in terms of attitude toward the ad and attitude toward the social media platform between Facebook and Twitter? According to the literature available on this field, it was not possible predict either Facebook or Twitter to be more effective in influencing consumers’ attitudes. However, after having highlighted some
differences between these two platforms, the first hypothesis of the study was formulated. It Facebook Twitter Social media usage Involvement Attitude toward platform Attitude toward ad Perceived interactivity not significant not significant
was expected that attitudes toward the ad and social media platform would have significantly differed between Facebook and Twitter. The results of the analysis did not show such an effect. Therefore, it can be argued that the social media platform itself does not really make the difference in the process of attitudes’ formation. Such a conclusion was expectable and
justifies the difficulties met in this study in comparing these two platforms.
The second question proposed in the study was, is the relationship between social media and attitudes mediated by perceived interactivity? Two hypotheses were formulated predicting an effect of social media on perceived interactivity and an effect of perceived interactivity on attitudes. However, the choice of the medium did not appear to influence the levels of consumers’ perceived interactivity. The results suggest that the two platforms are perceived as interactive nearly the same. Such findings are in line with part of theory
previously proposed in this paper and might prove that the number of interactive features (see Figure 4) actually affect the consequent outcomes. The other hypothesis specifically predicted that the higher the platform is perceived as interactive, the more positive will be the attitude toward the ad and toward the social media platform. The results supported the hypothesis and strengthen the findings of Coyle and Thorson (2001) and McMillan, Hwang and Lee (2003). Furthermore, the results suggested that when taking into account perceived interactivity, the relationship between social media type and attitude toward the social media platform became significant. In conclusion, it can be affirm that although no mediation effect has been found, the study confirms once more the criticality of perceived interactivity in shaping consumers’ attitudes.
The last research question of the study was, are involvement and social media usage acting as moderators of the main relationship? Indeed, the fourth hypothesis of the study proposed an interaction effect of involvement with the product. It was expected that when highly involved, consumers with high levels of social media usage would have scored better
in terms of attitude toward the and toward the social media platform. The findings did not support the hypothesis, however two interesting main effects were found. Involvement with the product significantly predicted consumers’ attitudes toward the ad; the more involved a
consumer is, the better his attitudes toward the ad will be. Furthermore, the same positive linear relationship was found between social media usage and attitude toward the social media platform.
Overall, the findings of this study offer good starting points for future research in this field. Both the mediating variable and the two moderators showed to be significant predictors of the dependent variables. In the next section some limitations of the study are proposed with the aim of addressing future research in the right track.
Limitations and future research (draft)
A first limitation of this study is represented by the experimental methodology employed. Indeed, whilst convenient and cheap, online experiments entail some drawbacks. The most important is that the researcher cannot control and support the respondents while they are completing the survey. More than forty responses had to be excluded from the analyses since not completed. The second drawback of online experiments is that respondents might take part of the survey only to get a reward back. However, this is not the case of this study where no incentives were offered to the respondents.
The stimulus materials of the experiment may be improved in future studies. Both the Facebook and the Twitter pages, although well done, lacked of some important characteristics which are likely to have influenced the final outcomes. First of all, the two pages were
slightly different each other. On the Facebook page two images are shown, whereas on Twitter the images are three. Furthermore, by using two images it is not possible recreate the same conditions respondents have when browsing their Facebook or Twitter page. This might have particularly influenced the perceived interactivity results.
The fact that both the moderators were not pre tested might represent the third
limitation of this study. The results showed very different mean values in terms of hours spent a week between the two conditions. Many Twitter respondents actually declared to do not even have a Twitter account. It is presumable that such a difference may have had an impact on the results. The same reasoning can be made for the variable involvement with the product. Based on a previous study (Voorveld, Van Noort, & Duijn, 2013), the brand Nespresso was chosen because believed to lead to the formation of highly and lowly involved consumers. However, differences in terms of the sample were not taken into account. Therefore, pre testing the two moderators would have probably been better, yielding to more reliable results. Future research should focus more on differences between Facebook and Twitter. This study represents a real first attempt of comparison between these two social media platforms. However, it needs to be acknowledged that the strategy used here does not take into account of many other variables at stake. The comparison made with the help of other studies and theories can only represents a first step of this process.
The last suggestion for future research regards the choice of the dependent variables. In this study only affective responses were taken into account. It may be interesting to see how the same model can influence other types of responses such as recall of the ad or purchase intention.
References
Baron, R.M., & Kenny, D.A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic and statistical considerations. Journal of
Personality and Social Psychology, 51, 1173-1182.
Beirut,. (2009). Why Do People REALLY Tweet? The Psychology Behind Tweeting!.Blog.thoughtpick.com. Retrieved 6 December 2015, from
http://blog.thoughtpick.com/2009/08/why-do-people-really-tweet-the-psychology-behind-tweeting.html
Borgatti, S., & Foster, P. (2003). The Network Paradigm in Organizational Research: A Review and Typology. Journal Of Management, 29(6), 991-1013.
http://dx.doi.org/10.1016/s0149-2063_03_00087-4
Cheng, Q., & Wells, W.D. (1999). Attitude toward the site. Journal Of Advertising
Research, 39(5), 27-37.
Chung, H., & Zhao, X. (2004). Effects of Perceived Interactivity on Web Site Preference and Memory: Role of Personal Motivation. Journal Of Computer-Mediated
Communication, 10(1), 00-00. doi:10.1111/j.1083-6101.2004.tb00232.x
Coyle, J., & Thorson, E. (2001). The Effects of Progressive Levels of Interactivity and Vividness in Web Marketing Sites. Journal Of Advertising, 30(3), 65-77. http://dx.doi.org/10.1080/00913367.2001.10673646
Daft, R., & Lengel, R. (1986). Organizational Information Requirements, Media Richness and Structural Design. Management Science, 32(5), 554-571.
http://dx.doi.org/10.1287/mnsc.32.5.554
Dijkstra, M., Buijtels, H., & van Raaij, W. (2005). Separate and joint effects of medium type on consumer responses: a comparison of television, print, and the Internet. Journal Of
Business Research, 58(3), 377-386. http://dx.doi.org/10.1016/s0148-2963(03)00105-x
Emarketer.com,. (2015). Social Network Ad Spending to Hit $23.68 Billion Worldwide in
http://www.emarketer.com/Article/Social-Network-Ad-Spending-Hit-2368-Billion-Worldwide-2015/1012357
Facebook.com,. (2015). People are appearing on my friend list without me sending them a
friend request. | Facebook Help Center | Facebook. Retrieved 6 December 2015, from
https://www.facebook.com/help/215747858448846
Fournier, S., & Avery, J. (2011). The uninvited brand. Business Horizons, 54(3), 193-207. doi:10.1016/j.bushor.2011.01.001
Granovetter, M. (1983). The Strength of Weak Ties: A Network Theory
Revisited. Sociological Theory, 1, 201. http://dx.doi.org/10.2307/202051
Halkias, G., & Kokkinaki, F. (2013). Increasing advertising effectiveness through
incongruity-based tactics: The moderating role of consumer involvement. Journal Of
Marketing Communications, 19(3), 182-197.
http://dx.doi.org/10.1080/13527266.2011.592346
Johnson, G., Bruner II, G., & Kumar, A. (2006). Interactivity and its Facets Revisited: Theory and Empirical Test. Journal Of Advertising, 35(4), 35-52.
http://dx.doi.org/10.2753/joa0091-3367350403
Kaplan, A., & Haenlein, M. (2010). Users of the world, unite! The challenges and opportunities of Social Media. Business Horizons, 53(1), 59-68.
http://dx.doi.org/10.1016/j.bushor.2009.09.003
Kaplan, A., & Haenlein, M. (2011). The early bird catches the news: Nine things you should know about micro-blogging. Business Horizons, 54(2), 105-113.
http://dx.doi.org/10.1016/j.bushor.2010.09.004
Kietzmann, J., Hermkens, K., McCarthy, I., & Silvestre, B. (2011). Social media? Get serious! Understanding the functional building blocks of social media. Business
Horizons, 54(3), 241-251. http://dx.doi.org/10.1016/j.bushor.2011.01.005
Lee, K. M., Park, N., & Jin, S.-A. (2006). Narrative and interactivity in computer games. In P. Vorderer & J. Bryant (Eds.) Playing video games: Motives, responses, and
Leung, X., Bai, B., & Stahura, K. (2013). The Marketing Effectiveness of Social Media in the Hotel Industry: A Comparison of Facebook and Twitter. Journal Of Hospitality &
Tourism Research, 39(2), 147-169. http://dx.doi.org/10.1177/1096348012471381
Levine, S., & McLuhan, M. (1964). Understanding Media: The Extensions of Man. American
Quarterly, 16(4), 646. http://dx.doi.org/10.2307/2711172
Liu, Y. (2003). Developing a scale to measure the interactivity of websites. (2003). Journal
Of Advertising Research, 43(2), 207 - 216.
Liu, Y., & Shrum, L. (2002). What is Interactivity and is it Always Such a Good Thing? Implications of Definition, Person, and Situation for the Influence of Interactivity on Advertising Effectiveness. Journal Of Advertising, 31(4), 53-64.
doi:10.1080/00913367.2002.10673685
Liu, Y., & Shrum, L. (2009). A Dual-Process Model of Interactivity Effects. Journal Of
Advertising,38(2), 53-68. http://dx.doi.org/10.2753/joa0091-3367380204
Lukka, V., & James, P. (2014). Attitudes toward Facebook advertising. Journal Of
Management And Marketing Research, 14, 1-26.
McCarthy, I., Lawrence, T., Wixted, B., & Gordon, B. (2010). A multidimensional
conceptualization of environmental velocity. Academy Of Management Review, 35(4), 604-626. http://dx.doi.org/10.5465/amr.2010.53503029
McMillan, S., & Hwang, J. (2002). Measures of Perceived Interactivity: An Exploration of the Role of Direction of Communication, User Control, and Time in Shaping Perceptions of Interactivity. Journal Of Advertising, 31(3), 29-42.
http://dx.doi.org/10.1080/00913367.2002.10673674
McMillan, S., Hwang, J., & Lee, G. (2003). Effects of structural and perceptual factors on attitudes toward the website. Journal Of Advertising Research, 43, 400-409.
Morris, M., & Ogan, C. (1996). The Internet as Mass Medium. Journal Of
Computer-Mediated Communication, 1(4), 0-0.
http://dx.doi.org/10.1111/j.1083-6101.1996.tb00174.x
Naegele, K., & Goffman, E. (1956). The Presentation of Self in Everyday Life. American
Parker, E., Short, J., Williams, E., & Christie, B. (1978). The Social Psychology of Telecommunications. Contemporary Sociology, 7(1), 32.
http://dx.doi.org/10.2307/2065899
Petty, R., Cacioppo, J., & Schumann, D. (1983). Central and Peripheral Routes to Advertising Effectiveness: The Moderating Role of Involvement. J CONSUM RES, 10(2), 135. http://dx.doi.org/10.1086/208954
Romero Calmache, M., & Fanjul Peyró, C. (2010). La publicidad en la era digital: el microsite como factor estratégico de las campañas publicitarias on-line. Comunicar, 17(34), 125-134. http://dx.doi.org/10.3916/c34-2010-03-12
Stelzner, M. A. (2015). 2015 Social Media Marketing Industry Eeport: How marketers are
using social media to grow their businesses.
Smit, E. (1999). Mass media advertising. Amsterdam: Het Spinhuis.
Song, J., & Zinkhan, G. (2008). Determinants of Perceived Web Site Interactivity. Journal Of
Marketing, 72(2), 99-113. doi:10.1509/jmkg.72.2.99
Ta k ran, N., & Y lmaz, R. (2015). Handbook of Research on Effective Advertising Strategies
in the Social Media Age. Hershey: IGI Global.
Tutaj, K., & van Reijmersdal, E. (2012). Effects of online advertising format and persuasion knowledge on audience reactions. Journal Of Marketing Communications, 18(1), 5-18. http://dx.doi.org/10.1080/13527266.2011.62076
Voorveld, H., Neijens, P., & Smit, E. (2011). The Relation Between Actual and Perceived Interactivity. Journal Of Advertising, 40(2), 77-92. http://dx.doi.org/10.2753/joa0091-3367400206
Voorveld, H., van Noort, G., & Duijn, M. (2013). Building brands with interactivity: The role of prior brand usage in the relation between perceived website interactivity and brand responses. J Brand Manag, 20(7), 608-622. http://dx.doi.org/10.1057/bm.2013.3
Wu, G. (2006). Conceptualizing and Measuring the Perceived Interactivity of
Websites.Journal Of Current Issues & Research In Advertising, 28(1), 87-104. http://dx.doi.org/10.1080/10641734.2006.10505193
Yaakop, A. (2013). Like It or Not: Issue of Credibility in Facebook Advertising. Asian Social
Science, 9(3). http://dx.doi.org/10.5539/ass.v9n3p154
Zaichkowsky, J. (1994). The personal involvement inventory: Reduction, revision, and application to advertising. Journal Of Advertising - Taylor & Francis.
