Sponsored stories on Facebook:
The effects of celebrity endorsements and sponsorship disclosure
on persuasion knowledge and consumer responses
Master thesis Communication Science, track Persuasive Communication Graduate School of Communication
Name: Eva van der Aa Student number: 6088171 Supervisor: Dr. S.C. Boerman Date: 27-06-2014
Abstract
This study examined the effects of different sources (i.e., brand versus celebrity) and sponsorship
disclosure (yes versus no) in embedded Facebook advertisements on persuasion knowledge and
consumer responses. By conducting an online experiment (N = 401), we found that advertisements
posted by brands, lead to a less favorable attitude towards the advertisement and intention to share,
through the activation of consumers' conceptual persuasion knowledge (i.e., the recognition of
advertising). Additionally, exposure to a disclosure leads to the same mediation effects.
However, this effect is only apparent for consumers who recalled a disclosure. Furthermore, the
inclusion of a disclosure by advertisements posted by a celebrity leads to positive results for the
activation of conceptual persuasion knowledge and intention to share. This study contributes to our
understanding of embedded advertising on social networking sites. The results provide implications
Introduction
The purpose of advertising has always been to deliver a marketing message that connects with consumers (Wright, 2010). With the increasing popularity of social media platforms such as Facebook in 2008 (Facebook, 2014), the opportunities to create such a connection have increased. The strength of social media is namely the interaction between consumers and the brand and the possibilities to communicate in a direct and interactive fashion. In particular, the strength of Facebook is the reach that it generates. Worldwide, Facebook has 680 million users and accounts for approximately 90 percent of all time spent on social-networking sites (Lipsman, Mudd, Rich, & Bruich, 2012). Keeping these figures in mind, it is save to conclude that Facebook has grown to become an appealing platform for marketers and advertisers.
The increasing popularity of Facebook triggers marketers to find creative and innovative ways to advertise on this platform (O'Connor, 2012). Facebook offers the opportunity to reach the target group of companies in a less obtrusive way than traditional media, through sponsored
content. Here, the format of an advert is adapted to the shape of the Facebook messages that appear in the newsfeed of users. Due to this format, the commercial or advertising intent of the message is not immediately clear to consumers. To enhance this effect, marketers often make use of celebrities to promote their products. Prior research has shown that the use of celebrity endorsements in traditional media positively affect brand responses (Chu & Kamal, 2008; Dekker & Van
Reijmersdal, 2010; Keel & Nataraajan, 2012). This might explain why advertisers use the same technique in social media marketing.
Although these marketing techniques offer advantages for advertisers, consumer advocates question the appropriateness of using these techniques (Cain, 2011). Due to the format and the source of the advertisements, the persuasive intent of these embedded advertisements is not always clear to consumers. To avoid misleading consumers and to protect them from deception, the Federal Trade Commission (FTC) in the United States has set up guidelines, which marketers should
advertisements should include disclosures that warn consumers about the commercial intent of a message (FTC, 2013). Similar guidelines were proposed by the Dutch advertising foundation, stating that advertisements on social media should be clear and recognizable as such (Reclamecode Social Media, 2014).
Including a disclosure and thus reminding consumers that they are being exposed to
sponsored content could activate consumers' persuasion knowledge (Friestad & Wright, 1994). This implies that consumers retrieve previous experiences regarding persuasion and knowledge about how to cope with persuasion attempts. As consumers generally do not like to persuaded (Brehm, 1966), this could influence consumer responses in terms of attitude towards the advertisement, attitude towards the brand, and consumers' intention to share. However, as the inclusions of disclosures by the FTC are merely guidelines, they are not always applied to the content and therefore could not activate consumers' persuasion knowledge. Additionally, due to the fact that embedded advertising on Facebook can take different forms in terms of the source (i.e., celebrity endorsed advertisements posted by either a brand or a celebrity), it might be more complicated for consumers to activate their persuasion knowledge. This could affect consumer responses. To this date, it is unknown whether the use of celebrity endorsers via commercial or celebrity sources on Facebook instigates persuasion knowledge amongst consumers, and whether the addition of a disclosure for both advertising tactics affects these prior mentioned effects. Therefore, by conducting this experiment, the present study aims to extent literature within this domain.
In sum, the aim of this study is to examine to what extent the source of advertisements (i.e., brand or celebrity) and sponsorship disclosure influence consumer responses (i.e., attitude towards the advertisement, attitude towards the brand and intention to share) on Facebook. Furthermore, the activation of consumers' persuasion knowledge will be tested in order to see if it plays a mediating role in this process.
Theoretical Framework
Persuasion Knowledge
Persuasion knowledge refers to consumers' knowledge and beliefs about various advertising related issues, such as the goals and tactics marketers use to persuade them, the extent to which consumers find these techniques effective and appropriate, but also personal beliefs about how to cope with these persuasion tactics and goals (Hibbert, Smith, Davies, & Ireland, 2007). In the present study, persuasion knowledge refers to what consumers know about the tactics marketers use on Facebook advertising.
A theoretical framework that describes how knowledge influences responses to advertising is the Persuasion Knowledge Model (PKM), developed by Friestad and Wright (1994). According to this model, the development of persuasion knowledge is dependent on experience of persuasion and it continues to develop over time. However, the model stresses that the use of consumers' persuasion knowledge only can take place when they are aware of the persuasive attempt. For consumers to be aware of the attempt, they should be able to recognize advertising, the advertising source and to understand the selling intent of a message (Rozendaal, Lapierre, Van Reijmersdal, & Buijzen, 2011). According to John (1999), these are the first steps in the development process of persuasion knowledge. As these developments are mostly cognitive, Rozendaal et al., (2011) categorize them as conceptual advertising literacy. Because this type of advertising literacy involves cognitive advertising related knowledge, in the present study this will be referred to conceptual persuasion knowledge.
However, in order to study consumer responses to advertising, attitudinal persuasion knowledge must also be taken into account (Rozendaal et al., 2011; Boerman, Van Reijmersda l, & Neijens, 2012). This dimension of persuasion knowledge refers to the affective side of coping with persuasion attempts. Research on persuasion has shown that consumers do not like to persuaded and therefore show resistance (Knowles, & Linn, 2004). In addition, the reactance theory by Brehm (1966) explains that consumers can develop negative evaluations towards a persuasive message, as
influencing them would interfere with their choice of freedom.
Within advertising research, the number of studies investigating the effects of persuasion knowledge in embedded advertising is growing. Various formats, such as brand placement in television (Dekker & Van Reijmersdal, 2013), advergames (Van Reijmersdal, Rozendaal, & Buijzen, 2012) and advertorials (Kim, 2014) have been studied. To this date however, no investigation of embedded advertisements on Facebook has been conducted. As these types of advertisements can make it harder for consumers to activate their persuasion knowledge, it is of importance to study the effects regarding the source of Facebook advertisements. In contrast, sponsorship disclosures might activate consumers persuasion knowledge. Therefore, we propose that the activation of persuasion knowledge is dependent on both the source of Facebook
advertisements and sponsorship disclosure.
Source of Facebook advertisements
According to the source credibility theory, people are more likely to be persuaded when the source demonstrates itself as credible (Petty & Cacioppo, 1986; Till, Stanley, & Priluck, 2008; Chu & Kamal, 2008). Research has shown that highly credible sources induce more positive consumer responses than less credible sources do (Chu & Kamal, 2008). According to Erdogan (1999), credible sources include components such as a high perceived level of expertise and trustworthiness by consumers. In addition, Amos, Holmes, & Strutton (2008) argue that attractiveness of the source is also of influence.
Keeping this information in mind, it is no surprise that marketers often choose to use a celebrity to endorse their products (Keel & Nataraajan, 2012). Research has shown that consumers assume celebrities endorse products because they genuinely believe they like and use these
products, without knowing they receive money for it (Silvera & Austad, 2004; Dekker & Van Reijmersdal, 2010). Therefore, with regard to advertisements on Facebook, it is plausible to assume that this effect might be enhanced when the advertisement, in addition to including a celebrity, is
also posted by a celebrity. This is in contrast to advertisements that are posted by a brand, by which consumers might associate exposure to a brand with commerce (Bhatnagar, Aksoy, & Malkoc, 2004). Therefore, consumers could more easily link a Facebook post by a brand to advertising, recognize the advertising source and understand the selling intent, in comparison to a Facebook post by a celebrity. Consequently, the recognition of the concept of advertising could elicit critical
feelings and therefore activate consumers' attitudinal persuasion knowledge. This assumption is in line with Van Reijmersdal, Neijens, & Smit (2005), who argue that consumers feel prudent towards advertisements as they are aware that marketing purposes are aimed at persuasion and increasing sales. Consequently, this could lead towards skepticism towards advertising (Boerman et al., 2012). Hence, the following hypothesis can be proposed: H1: Facebook advertisements posted by a brand
lead to a higher activation of consumers’ conceptual and attitudinal persuasion knowledge, in comparison to Facebook advertisements posted by a celebrity.
Sponsorship disclosures on Facebook
The activation of consumers' persuasion knowledge might also be dependent of the inclusion of sponsorship disclosures. The main purpose of a disclosure is to inform consumers about the
persuasive intent behind an embedded message (Campbell, Mohr, & Verlegh, 2007; Boerman et al., 2012). By including a disclosure, it is expected that consumers can more easily recognize the commercial nature of a message and thus activate their conceptual persuasion knowledge. Consequently, the activation of consumers’ conceptual persuasion knowledge can lead to an activation of their attitudinal persuasion knowledge (Boerman et al., 2012). In other words, this means that after consumers recognize the persuasive intent, they start to process the message more critically. Hence, the following hypothesis can be proposed: H2: Facebook advertisements that
Consumer responses
In addition to persuasion knowledge, we propose that consumer responses can also be dependent of the source of Facebook advertisements and sponsorship disclosure. In the following section,
possible effects of the source and disclosure by Facebook advertisements on attitude towards the advertisement, attitude towards the brand and intention to share will be discussed.
Source of Facebook advertisements
Research on celebrity endorsements in traditional media has shown that the inclusion of celebrities can positively influence consumer responses (Erdogan, 1999). For instance, Lafferty, Goldsmith and Newell (2002) found that the use of celebrities in traditional media elicits more favorable attitudes towards advertisements as well as towards the brand. This result was confirmed in the study conducted by Till et al. (2008). A possible explanation for these findings can be described by the meaning transfer model (McCracken, 1989), which states that through associations, certain qualities of celebrities can be transferred to the brand. In other words, the liking of a celebrity by consumers can lead to a greater liking of the brand and thus positively influencing consumer brand responses. Moreover, typical consumer behaviors on Facebook, such as the intention to share the advertisement, could also positively be influenced by celebrities. Research on online forwarding of e-mails has shown that consumers' motives can be dependent from the source of a message (Chiu, Hsieh, Kao, & Lee, 2007). Sources which provide emotional or substantive support (e.g., family and friends) are perceived as more credible and trustworthy by consumers, and therefore are more likely to be forwarded by consumers than messages coming from sources which are unfamiliar or commercial (Van Noort, Antheunis, & Van Reijmersdal, 2012). Even though it cannot be stated that celebrities have the same tie to consumers as family and friends do, previously discussed literature provides evidence that celebrities can be perceived as trustworthy and credible (e.g., the source credibility theory, Petty, & Cacioppo, 1986). Therefore, especially in comparison to Facebook advertisements posted by the brand in which the commercial intent is much clearer (Dekker & Van
Reijmersdal, 2013), the following hypotheses can be proposed: H3: Facebook advertisements
posted by a celebrity create a more favorable brand attitude and attitude towards the advertisement than Facebook advertisements posted by the brand.
In addition, we hypothesize:
H4: Facebook advertisements posted by a celebrity create a higher intention to share than
Facebook advertisements posted by the brand.
It is possible that the effects of the source of Facebook advertisements on consumer responses are mediated by persuasion knowledge. As previously discussed, exposure to brands might cause consumers to create associations towards commerce (Bhatnagar et al., 2004; Van Reijmersdal et al., 2005). This means that consumers are aware of a possible commercial intent of the message and thus activate their conceptual persuasion knowledge. As consumers are not fond of advertising and persuasive attempts (Brehm, 1966), once they acknowledge the source of the advertisement is not genuine or neutral, but rather persuasive, a change of meaning (Friestad & Wright, 1994) can occur. This can lead to more scrutinized and critical processing of the advertising content and thus
activating consumers' attitudinal persuasion knowledge. The critical processing of the content can lead to less favorable consumer responses, as consumers are already critical. Hence, the following hypothesis can be proposed: H5: Facebook advertisements posted by a brand lead to less favorable
consumer responses (i.e., attitude towards the advertisement, attitude towards the brand and intention to share) in comparison to Facebook advertisements posted by a celebrity, due to a higher activation of conceptual and attitudinal persuasion knowledge.
Sponsorship disclosure on Facebook
Next to the source of Facebook advertisements, it is also possible that the inclusion of a sponsorship disclosure influences consumer responses. Prior research has shown that consumers do not hold
favorable attitudes towards advertising in general (Brehm, 1966; Calfee & Ringold, 1994). In specific, a recent study by Kelly, Kerr and Drennan (2010) shows that this also accounts for advertisements on social media. In their study they argue that, due to negative experiences with advertising, irrelevance of the advertisement, skepticism towards the advertising message or skepticism towards the advertising medium, consumers try to avoid advertising. This finding is in line with previous research regarding persuasion and resistance (Knowles & Linn, 2004).
However, the prior mentioned study does not take embedded advertisements into account. As previously discussed, due to the nature of these advertisement types, consumers might not recognize them as advertising and thus not show avoidance. The inclusion of a disclosure could therefore help consumers, as the main purpose of sponsorship disclosures is to raise awareness about the
commercial content of a message (Cain, 2011; Boerman et al., 2012). In the present study, the main goal of including a disclosure by Facebook advertisements would thus mean that consumers
understand that brands that are visible in a Facebook post appear because brands paid them to be there. Consequently, due to the change of meaning principle (Friestad & Wright, 1994), this could negatively affect consumer responses. Hence, the following hypothesis can be proposed: H6:
Facebook advertisements that include sponsorship disclosure lead to less favorable attitude towards the advertisement and attitude towards the brand.
In addition, we hypothesize:
H7: Facebook advertisements that include sponsorship disclosures lead to a lower intention to
share.
Similar to Facebook advertisements posted by brands, the effects of a sponsorship disclosure on consumer responses can be mediated by the activation of consumers' persuasion knowledge. As Boerman et al., (2012) argue, the witnessing of a disclosure could serve as a cue that reminds consumers about a commercial intention. As associations towards commerce can lead to the
activation of consumers' conceptual persuasion knowledge and consequently, a skeptical
perspective and processing of advertisements (Obermiller, Spangenberg, & MacLachlan, 2005), consumer responses could be affected. In line with this assumption, prior research has shown that the activation of persuasion knowledge negatively affects brand responses on television (Campbell, Mohr, & Verlegh, 2013; Boerman, Van Reijmersdal, & Neijens, 2014). As advertisements on social media are negatively perceived by consumers (Kelly et al., 2010), it is plausible to assume that including a disclosure, which could activate consumers’ persuasion knowledge, also negatively affects consumer responses on Facebook. Hence, the following hypothesis can be proposed: H8:
Sponsorship disclosure by a Facebook advertisement lead to less favorable consumer responses (i.e., attitude towards the advertisement, attitude towards the brand and intention to share), due to a higher activation of conceptual and attitudinal persuasion knowledge.
Interaction between the source of the Facebook advertisements and disclosure
A disclosure and the type of source of Facebook advertisements may interact with each other. As Facebook advertisements posted by a brand are more likely to be recognized as advertising, it could lead to an activation of consumers' persuasion knowledge. Therefore, the inclusion of a disclosure might be considered redundant, as the source already warns consumers about the commercial nature of the message. However, disclosures could provoke the activation of consumers' persuasion
knowledge when posted by a celebrity. In compliance with the source credibility theory and previous research regarding celebrity endorsements (Petty & Cacioppo, 1986; Silvera & Austad, 2004; Dekker & Van Reijmersdal, 2010), consumers might believe the post by the celebrity is honest and sincere, and therefore do not foresee any commercial or persuasive intention. Moreover, the possible effects on consumer responses might be stronger for Facebook advertisements posted by a celebrity that include a disclosure, due to the activation of persuasion knowledge. As consumers can associate celebrities with credibility and trustworthiness (Dekker & Van Reijmersdal, 2010), they might feel deceived once they activated their persuasion knowledge
(Boerman et al., 2014). This could negatively affect the attitude towards the advertisement, attitude towards the brand and the intention to share. The change of meaning principle by Friestad and Wright (1994) might explain this effect. Once consumers recognize the persuasive intention of the message posted by a celebrity, their meaning regarding the message changes, causing detachment and thus possible feelings of deception. Consequently, this could result in negative consumer evaluations. Hence, an interaction effect between Facebook advertisements posted by a celebrity and the inclusion of sponsorship disclosure on persuasion knowledge is expected. Therefore, the following hypothesis can be proposed: H9: The effects of disclosure on both conceptual and
attitudinal persuasion knowledge and consumer responses (i.e., attitude towards the advertisement, attitude towards the brand and intention to share) are stronger when the Facebook advertisement is posted by a celebrity, compared a Facebook advertisement posted by a brand.
Method
Study design, participants and procedure
This study employed a 2 (source: brand versus celebrity) x 2 (disclosure: yes versus no) between-subjects design. Through Facebook and personal communication, a total of 401 participants were recruited for this experiment. The average age of the participants was 29 (SD = 12.64) and 66.3% were female. Most of the participants had the Dutch nationality (67.1%) and the majority of the participants stated to have completed higher education (M = 3.58, SD = 1.56).
We conducted an online experiment using Qualtrics, where participants were first welcomed to take part of the study. In order not to influence the participants beforehand about the final
purpose of this study, we used a cover story. Participants were informed that the study was about Facebook usage and that it would examine how people react on different Facebook posts.
Furthermore, participants were explicitly told that there were no wrong answers and that all
information would be processed anonymously. To make sure that the participants agreed to take part in this study, an informed consent was signed.
The second page showed the stimuli. Participants were exposed to a Facebook
advertisement that was posted by either a brand or a celebrity, in which sponsorship disclosure did or did not appear. Participants were randomly assigned to one of four conditions. They were asked to view the advertisement, which was placed in the newsfeed of Facebook, as if they were visiting Facebook themselves. Once participants had viewed the advertisement, they could directly proceed to the questionnaire. The questionnaire started with a control variable regarding the Facebook advertisement, namely Facebook usage. Then, two consumer responses (i.e., attitude towards the advertisement and intention to share) were asked, followed by brand responses (i.e., brand recall, brand familiarity and brand attitude) and participants' persuasion knowledge (i.e., conceptual and attitudinal). Finally, the questionnaire ended with other control variables, namely recall and recognition of disclosure, and demographics. Participants were also given the opportunity to leave comments or remarks. In the questionnaire, consumer responses were intentionally asked before persuasion knowledge. By doing so, it could prevent consumer responses to be a result of the questions about advertising and trustworthiness, as these reveal the commercial nature of the advertisement.
Materials
The stimulus materials consisted of four edited iPad screenshots of a newsfeed on Facebook. The choice for the iPad screenshot was made due to the fact that the application contains less
information and less advertising, which leads to a more neutral and therefore less distracting newsfeed in comparison to one appearing on a pc. Apart from the manipulation of the source and the disclosure in the advertisement, the newsfeed in all conditions were identical to each other. The newsfeed belonged to a fictitious person.
The advertisement showed an international celebrity, David Beckham, drinking coffee from the brand 'Illy'. Alongside the advertisement posted by David Beckham, the following text
Beckham starts his day with a nice cup of coffee!”. The brand name was intentionally not mentioned in the advertisements, to stay in tune with the characteristics of an embedded advertisement.
Both David Beckham and the coffee brand 'Illy' were chosen due to their unisex character. By doing so, we expect to prevent inequalities based on attractiveness of the source and product, and therefore avoid an effect that is caused by gender. Furthermore, the brand 'Illy' was chosen because this brand is not too well-known. For this reason it is expected that some of the participants are familiar with this brand, but not all.
Pre-test
The Facebook advertisements were manipulated in terms of the source (brand versus celebrity) and disclosure (yes versus no). In order to see if the manipulation of the disclosure was successful, a pre-test amongst 22 participants was conducted. The pre-test consisted of two conditions. The first condition showed the disclosure in the original size as it appears on Facebook, whereas the second condition showed a bigger and therefore more prominently placed disclosure. Participants were randomly assigned to one of the two conditions.
Results showed that the majority of the participants in both conditions did not recall (54.5%) nor recognized (86.4%) both types of disclosure. No significant differences between the two groups were found for recall (χ² (1) = .714, p = .398) and recognition of the disclosure (χ² (1) = .268, p =
.605). In other words, participants did not notice both types of disclosure. In order to show an effect of disclosure in this experiment, the disclosure must be manipulated in such way that people do recognize it. Therefore, the disclosure was manipulated again so that it had a bigger size and appeared in bold (see Appendix A for the stimuli).
Measures
Persuasion Knowledge
We measured participants' persuasion knowledge using the conceptual and attitudinal dimensions of persuasion knowledge (Rozendaal et al., 2011; Boerman et al., 2012). Conceptual persuasion
knowledge was measured by asking participants to indicate on a 7-point scale that varied from “strongly agree” to “strongly disagree” to what extent they thought the Facebook post was “advertising” (recognition of advertising), encouraging to buy Illy Coffee (understanding selling intent), and to what extent participants believed Illy Coffee paid David Beckham to post the Facebook message (recognition of advertising source). The principal component analysis (PCA) showed that the items load on one factor (Eigenvalue = 1.559, explained variance = 51.98%). However, reliability analysis showed that the three items together did not make a reliable scale (Cronbach's alpha = .473). After correcting for the item “understanding selling intent”, alpha was increased to .680. The final scale for conceptual persuasion knowledge consisted of the mean scores of the items “recognition of advertising” and “recognition of advertising source” (M = 5.59, SD = 1.24).
Participants' attitudinal persuasion knowledge was measured by asking participants to indicate whether they agreed (1 = strongly disagree, 7 = strongly agree) with the statement, “ I think the Facebook post was..” followed by five attributes based on a scale measuring source
trustworthiness (Ohanian, 1990; Boerman et al., 2012): “honest”, “trustworthy”, “convincing”, “biased” and “not credible”. Initial principal component analyses revealed that the items load on two factors (Facor 1: Eigenvalue = 2.57, explained variances = 51.42%; Factor 2: Eigenvalue = 1.318, explained variances = 26.37%) and the items did not appear reliable (Cronbach's alpha = .472). However, after deleting the items “biased” and “not credible”, new factor analysis showed that the items represent a one-dimensional scale (Eigenvalue = 2.381, explained variances = 79.37%) and proved reliable (Cronbach's alpha = .864). The final scale of attitudinal persuasion knowledge consisted of the mean scores of the items “honest”, “trustworthy” and “convincing” (M
= 3.31, SD = 1.25).
Attitude towards the advertisement
Attitude towards the advertisement was measured using three 7-point semantic differential scales (Bruner, 2009; Campbell, 1995). Participants were asked to what extent they believed the
advertisement was bad/good, negative/positive and dislikeable/likeable. Factor analysis revealed that all items loaded on one factor (Eigenvalue = 2.278, explained variances = 75.92%, Cronbach's alpha = .841). The mean score of the three items were used as a measurement for the attitude towards the advertisement (M =4.40, SD =1.25).
Attitude towards the brand
Brand attitude was measured using six 7-point semantic differential scales (Bruner, 2009; Campbell, 1995). Participants could indicate whether they thought the brand was bad/good
unpleasant/pleasant, unfavorable/favorable, negative/positive, dislike/like, and poor quality/high quality. Factor analyses showed the six items represent a one-dimensional scale (Eigenvalue = 5.120, explained variances = 85.33%) and a reliability analysis showed the items also form a reliable scale (Cronbach's alpha = .965). The mean score of the six items were used as a measurement for brand attitude (M = 4.93, SD = 1.09).
Intention to share
To measure intention to share the Facebook advertisement, we used the Forwarding Model developed by Eckler, Petya and Bolls (2011). This model consisted of two items ('This Facebook post is worth sharing with others” and “I will recommend this Facebook post to others”) that measure intention to share. Participants were asked to indicate to what extent they agreed with the statements on a 7-point Likert scale ( 1 = strongly disagree, 7 = strongly agree). Factor analysis revealed that the items load on one factor (Eigenvalue = 1.718, explained variances = 85.93 %,
Cronbach's alpha = .832). The final measurement of intention to share consisted of the mean score of the two items (M = 2.20, SD = 1.26). Appendix B shows the complete questionnaire.
Results
Control variables
To ensure that the effects of the disclosure and the advertisement source were not caused by other differences between the experimental groups, a number of control variables were measured. In order to measure Facebook usage, participants were asked how often they checked their Facebook accounts. Most participants (30,9%) indicated that they check their Facebook account two to five times a day, whereas 22,2% stated to check their accounts six to ten times a day. In addition, the majority of the participants stated to be familiar with the brand Illy (74,6%). Moreover, brand recall was measured by asking participants if they could recall a brand appearing in the post and if so, which one. The majority of the people (65,1%) said they could recall the correct brand. In addition, age, education, on a scale from 1 (high school) to 6 (master of arts/science), sex and nationality were measured.
Randomization and manipulation check
The experimental groups did not differ in terms of sex, χ2 (1) = 2.66, p = .103, education, F(1. 395) = 0.24, p = .622, Facebook usage, F(1, 395) = 2.99, p = .085 and brand familiarity χ2 (1) = 0.771, p = .380. Keeping in mind however that the majority of all participants stated to be familiar with the brand (74.6%), brand familiarity will be included as a control variable in all analyses, so that it will not confound the effects.
Furthermore, in order to see if the manipulation of the disclosure condition was successful, a manipulation check was conducted. Participants were asked to indicate whether they could recall a label that informed them that the Facebook post was sponsored. Next, a sponsorship disclosure label was shown to the participants and they could indicate whether they recognized the label.
Participants in the no disclosure condition who indicated to recall or recognize a disclosure, were excluded from all analyses regarding disclosure effects. This also accounted for participants who were exposed to a disclosure, but did not recall or recognized it. After completing the manipulation check, the final sample of participants for analysis regarding disclosure consisted of 281 people.
Effects of the source of the advertisement
In order to see which effects the source of the Facebook advertisement had, a MANCOVA was conducted with the source of the Facebook advertisement as independent variable, both dimensions of persuasion knowledge (i.e., conceptual and attitudinal), brand attitude, attitude towards the advertisement and intention to share as dependent variables and Facebook usage, brand familiarity and demographics (i.e., gender, sex and education) as covariates. These covariates were included in all other analyses. With respect to H1, the results showed that there is a significant difference of the advertisement source on the activation of persuasion knowledge, Pillai’s Trace = .03, F = 2.39(5, 392), p = .038, η2
= .030. When looking into each dependent variable individually, a significant
difference could be found between the sources for conceptual persuasion knowledge, F(1, 396) = 10.76, p = .001, η2
= .026, but not attitudinal persuasion knowledge, F(1, 396) = .60, p = .441, η2
= .002. Thus, the support for hypothesis 1 can only be found for conceptual persuasion knowledge. Table 1 shows that the majority of participants perceived the Facebook advertisement posted by the brand as advertising (M = 5.79, SD = 1.19), and therefore activated their conceptual persuasion knowledge. This means that exposure to a Facebook advertisement that is posted by a brand has a significant effect on conceptual persuasion knowledge, but not on attitudinal knowledge. Thus, people more often perceive content posted by brands as advertising, and therefore activate their conceptual persuasion knowledge, in comparison to Facebook advertisements posted by celebrities. Hence, H1 is partially accepted.
With regards to the effects of the advertisement source on the dependent variables (H3 and H4), the analysis showed that there is no significant difference between the two groups on brand
attitude, F(1, 396) = .70, p = .405, η2 = .002, attitude towards the advertisement, F(1, 396) = .28, p = .595, η2
= .001 and intention to share, F(1,396) = 1.29, p = .257, η2
= .003. This means that participants exposed to a Facebook advertisement that was posted by a celebrity do not create a more favorable attitude towards the advertisement (M =4.44, SD = 1.24) nor attitude towards the brand (M = 4.90, SD = 1.05) in comparison with Facebook advertisements posted by the brand (M = 4.36, SD = 1.26, M = 4.97, SD = 1.13, respectively). Moreover, being exposed to a Facebook
advertisement posted by a celebrity does not lead to a higher intention to share (M = 2.27, SD = 1.23) in comparison to a brand (M = 2.14, SD = 1.29). In other words, it does not matter whether participants are exposed to Facebook advertisements posted by the brand or the celebrity, no
significant effect for brand attitude, attitude towards the advertisement and the intention to share the advertisement can be found. Thus, H3 and H4 are not supported.
Table 1. Effects of source of the advertisement on conceptual and attitudinal persuasion knowledge,
attitude towards the advertisement, attitude towards the brand and intention to share.
Brand Celebrity
Conceptual persuasion knowledge 5.79 (1.19)a 5.40 (1.26)b
Attitudinal persuasion knowledge 3.25 (1.32)a 3.36 (1.19)a
Attitude towards the advertisement 4.36 (1.26)a 4.44 (1.24)a
Attitude towards the brand 4.97 (1.13)a 4.90 (1.05)a
Intention to share 2.14 (1.29)a 2.27 (1.23)a
Note. Mean scores with standard deviation between parentheses are portrayed
a
Means with a different superscript in the same row differ significantly at p < .05.
Effects of sponsorship disclosure
To test the hypotheses that predicted effects of disclosure, another MANCOVA was conducted with the experimental group (disclosure: yes versus no) as independent variable, and both dimensions of
persuasion knowledge, brand attitude, attitude towards the advertisement and intention to share as dependent variables. The sample used for analyses regarding disclosure consists of participants who have completed the manipulation check successfully (N = 281). With respect to H2, the results showed that there is a significant difference within the experimental group, Pillai’s Trace = .047, F
= (5,272) = 2.706, p = .021, η2
= .047. When looking at the variables independently, a significant difference between the groups was found for conceptual persuasion knowledge, F(1, 276) = 10.88, p = .001, η2 = .038. The effects on attitudinal persuasion knowledge were margina lly significant, F (1, 276) = 2.79, p = .097, η2
= .010. In other words, consumers who saw a disclosure had a higher recognition of advertising (M = 6.01, SD = 1.14) than consumers who were not exposed to a disclosure (M = 5.55, SD = 1.21). This however did not lead to an activation of their attitudinal persuasion knowledge in neither the disclosure condition (M = 3.12, SD = 1.29) nor the no disclosure condition (M = 3.34, SD = 1.25). Hence, H2 was accepted for conceptual persuasion knowledge.
With regards to the effects of disclosure on the consumer responses (H6 and H7), the analyses showed that no significant effect could be found for any of the consumer responses; attitude towards the advertisement: F(1, 276) = 1.724, p = .190, η2
= .006 brand attitude: F(1, 276) = .896, p = .345, η2 = .003, intention to share, F(1, 276) = .014, p = .908, η2 = . 000. This means that exposure to disclosure does not directly affect participant´s attitude towards the advertisement (M = 4.26, SD = 1.35), attitude towards the brand (M = 5.01, SD = 1.12) or their intention to share the advertisement (M = 2.17, SD = 1.16), in comparison attitudes towards the advertisement (M = 4.43, SD = 1.19), attitudes towards the brand (M = 4.93, SD = 1.03) or intention to share (M = 2.17, SD = 1.31) without a disclosure. Thus, hypothesis H6 and H7 are not supported. The means of the effects of disclosure are visible in Table 2.
Table 2. Effect of disclosure on conceptual and attitudinal persuasion knowledge, attitude towards
the advertisement, attitude towards the brand and intention to share.
No disclosure Disclosure
Conceptual persuasion knowledge 5.55 (1.21)a 6.01 (1.14)b
Attitudinal persuasion knowledge 3.34 (1.25) c 3.12 (1.29)d
Attitude towards the advertisement 4.43 (1.19)a 4.26 (1.35)a
Brand attitude 4.93 (1.03)a 5.01 (1.12)a
Intention to share 2.17 (1.31)a 2.17 (1.16)a
Note. Mean scores with standard deviation between parentheses are portrayed. a b
Means with a different superscript in the same row differ significantly at p < .05.
c d
Means with a different superscript in the same row differ significantly at p < .10.
Mediation effects
The source of Facebook advertisements
To test the mediation effect proposed in H5, we used the PROCESS macro in SPSS (Hayes & Preacher, 2013). This program estimates the path coefficients in a mediator model and generates 95% bootstrap confidence intervals for total and specific indirect effects of the advertising's source on consumer responses (i.e., brand attitude, attitude towards the ad and intention to share), through conceptual persuasion knowledge. Given the fact that the previous analyses only showed significant effects for conceptual, but not attitudinal persuasion knowledge, the mediation analysis will focus solely on conceptual persuasion knowledge. The PROCESS macro method uses 1000 bootstrapped samples to estimate the bias corrected and accelerated confidence intervals (BCACI).
Figure 3 represents the tested mediation model, with the advertising source as independent variable, the attitude towards the advertisement and intention to share as dependent variables, and conceptual persuasion knowledge as a mediator. The paths in these figures correspond with the results in Table 4. In these models, the c-path consists of the direct effect of the source on the dependent variables, independent of the effect of the mediator (c’), and the total effect of either the
source or disclosure on one of the dependent variables (c), which can be described as the sum of the direct effect and the indirect effect via he mediator (Hayes & Preachers, 2013).
In line with prior MANCOVA findings, there were no significant total (c) or direct effect (c’) of the advertising source on attitude towards the advertisement, attitude towards the brand and intention to share. This means that the source of the advertisements did not directly influence any of the dependent variables.
As portrayed in Figure 3, the results show that exposure to Facebook advertisements posted by the brandhas a positive effect on conceptual persuasion knowledge (b = .40, p < .001).
Furthermore, conceptual persuasion knowledge has a significant negative effect on the attitude towards the Facebook advertisement (b = -.17, p < 0.001). In addition, a significant indirect effect of source on attitude towards the advertisement via conceptual persuasion knowledge could be found (indirect effect: -.07, SE = .03, 95% BCACI [–.14, -.02]). This means that exposure to the brand leads to viewers recognizing the persuasive intent of the message and therefore activating their conceptual persuasion knowledge, which in turn leads to a less favorable attitude towards the advertisement.
There was no evidence of an effect of the advertising source on brand attitude. The results demonstrate that conceptual persuasion knowledge does not directly influence the attitude towards the brand (b = .06, p = .176). Hence, no significant indirect effect on brand attitude was visible (indirect effect: .02, SE = .02, 95% BCACI [–.01, .07]. This means that attitude towards the brand was not significantly affected by the advertisement source. Therefore, H5 was not supported for brand attitude.
With regard to the intention to share (see Figure 3), a significant effect of conceptual persuasion knowledge on intention to share (b = -.22, p < .001) could be found, which means that there is a significant indirect effect of the advertising source on intention to share via conceptual persuasion knowledge (indirect effect: -.09, SE = 0.03, 95% BCACI [–.16, -.03].
persuasion knowledge can be viewed as a predictor for a less favorable attitude towards the
advertisement, as well as a lower intention to share the advertisement. However, brand attitude does not seem to be affected by an activation of conceptual persuasion knowledge. This means that participants do not think less favorable of the brand even though they are aware of the persuasive intent behind the Facebook message. Hence, H5 was confirmed for attitude towards the
advertisement and intention to share.
Figure 3. Mediation model: Effect of advertising source on attitude towards the advertisement and intention to share, via conceptual persuasion knowledge.
a b c, c’ Consumer responses Advertisement source Conceptual persuasion knowledge
Table 4. Results of mediation model: Effect of advertising source on attitude towards the
advertisement, attitude towards the brand and intention to share, via conceptual persuasion knowledge.
a b c (total) c’ (direct) Indirect effect
(95% BCACI)
Attitude towards the advertisement .40 (.12)* -.17 (.05)** -.07 (.12) .00 (.12) -.07 (.03) [-.14; -.02] Brand attitude .40 (.12)* .06 (.04) .08 (.10) .06 (.10) .02 (.02) [-.01; .07] Intention to share .40 (.12)* -.22 (.05)*** -.14 (.13) -.06 (.12) -.09 (.03) [-.16; -.03]
Note a, b, c and c’ are unstandardized b-coefficients (with boot SE between parentheses); BCACI = Bias corrected
and accelerated confidence interval using 1,000 bootstrap samples; N = 401. * p < .05, ** p < .01, *** p < .001
Sponsorship disclosure
In order to test the mediation effect proposed in H8, the same mediation analyses were conducted for the independent variable sponsorship disclosure. In line with prior MANCOVA findings, the results demonstrated that both the total (c) and the direct (c’) effects of disclosure on attitude towards the advertisement, attitude towards the brand and intention to share were not significant. This means that there is no direct effect of disclosure on any of the dependent variables. However, the results showed that disclosure has a significant positive effect on conceptual persuasion
knowledge (b = .51, p = 0.001, see Figure 4). Furthermore, conceptual persuasion knowledge
showed a significant negative effect on the attitude towards the Facebook advertisement (b = -.21, p < 0.001). Additionally, a significant indirect effect of disclosure on attitude towards the
advertisement via conceptual persuasion knowledge could be found (indirect effect: -.11, SE = .05, 95% BCACI [–.22, -.04]). This means that due to a disclosure, viewers recognized the sponsored
message as advertising and therefore activated their conceptual persuasion knowledge. In turn, this led to a less favorable attitude towards the advertisement.
Furthermore, conceptual persuasion knowledge did not directly influence the attitude towards the brand (p = .434). Therefore, no significant indirect effect of the disclosure on brand attitude via conceptual persuasion knowledge could be found (indirect effect: 02, SE = .03, 95% BCACI [–.03, .09].
Moreover, a significant effect of conceptual persuasion knowledge on intention to share (b = -.27, p < .001) was found (see Table 5 and Figure 4). For this reason, a significant negative indirect effect of disclosure on intention to share via conceptual persuasion knowledge (indirect effect: -.14,
SE = 0.05, 95% BCACI [–.16, -.06].
Thus, support for H8 was found for attitude towards the advertisement and intention to share the advertisement. However, exposure to a disclosure did not affect the brand attitude, in spite of participants recognizing the advertising intent.
Figure 4. Mediation model: Effect of sponsorship disclosure on consumer responses.
a b
c, c’
Table 5. Results of mediation model: Effects of disclosure on attitude towards the advertisement,
Consumer responses Sponsorship
disclosure
Conceptual persuasion knowledge
attitude towards the brand and intention to share, via conceptual persuasion knowledge.
a b c (total) c’ (direct) Indirect effect
(95% BCACI)
Attitude towards the advertisement .51 (.15)* -.21 (.06)** -.21 (.16) -.10 (16) -.11 (.05) [-.22; -.04] Brand attitude .51 (.15)* .04 (.05) .12 (.12) .10 (.13) .02 (.03) [-.03; .09] Intention to share .51 (.15)* -.27 (.06)*** .02 (.16) .16 (.16) -.14 (.05) [-.24; -.06] Note
a, b, c and c’ are unstandardized b-coefficients (with boot SE between parentheses); BCACI = Bias corrected
and accelerated confidence interval using 1,000 bootstrap samples; N = 281.
* p < .05, ** p < .01, *** p < .001
Interaction effects
To test the possible interaction effect proposed in H9, a MANCOVA was conducted with the two experimental groups as independent variables and both dimensions of persuasion knowledge, attitude towards the advertisement, attitude towards the brand and intention to share as dependent variables. The analyses showed a significant overall interaction effect of advertisement source and disclosure on the dependent variables, Pillai’s Trace = .06, F = 3.13 (5, 270), p = .009, η2 = .055.
Separate ANCOVA's however did not show significant effects for attitudinal persuasion knowledge, brand attitude and attitude towards the advertisement, but a significant interaction effect was found for conceptual persuasion knowledge, F = 5.13 (1, 274), p = .024, η2
= .02 and intention to share, F
= 5.70 (1, 274), p = .018, η2 = .02.
As Figure 1 shows, for conceptual persuasion knowledge, a significant difference was found between an advertisement posted by a celebrity and disclosure (p < .001) and between the brand and
no disclosure (p < .001). This means that including a disclosure by a Facebook advertisement has a positive effect on conceptual persuasion knowledge, but only when the advertisement is posted by the celebrity. Disclosure did not cause any effects on conceptual persuasion knowledge when the Facebook advertisement was posted by the brand itself. Facebook advertisements without disclosure did not, as expected, lead to an increase in conceptual persuasion knowledge when posted by a celebrity (for means, see Table 3).
With respect to the intention to share the Facebook advertisement, pairwise comparisons only showed significant differences between the source of the advertisement (brand versus
celebrity) in the disclosure condition (p <.008). The means in Table 3 show that participants in the disclosure condition had a significant higher intention to share the Facebook advertisement when posted by the celebrity in comparison to the brand. This is also illustrated in Figure 2.
In sum, celebrity postings that include disclosures lead to an activation of conceptual persuasion knowledge and to a higher intention to share the Facebook advertisement. In other words, even though disclosure leads to a higher awareness and more critical processing of the Facebook posting, this does not influence the intention to share the advertisement when posted by a celebrity. Hence, hypothesis 7 was partially confirmed.
Table 3. Interaction effects between the source of the advertisements and sponsorship disclosure
No disclosure Disclosure
Brand Celebrity Brand Celebrity
Conceptual PK 5.93 (1.09)a 5.17 (1.23)b 6.09 (1.20)abc 5.96 (1.11)ac Attitudinal PK 3.36 (1.26)a 3.32 (1.24)a 2.94 (1.41)a 3.24 (1.20)a Ad attitude 4.39 (1.09)a 4.47 (1.28)a 4.06 (1.53)a 4.39 (1.21)a Brand attitude 4.97 (1.03)a 4.90 (1.03)a 5.09 (1.22)a 4.95 (1.06)a Intention to share 2.20 (1.33)a 2.13 (1.30)a 1.74 (.88)ab 2.46 (1.25)ac Note
All variables are scaled 1-7. N = 281: no disclosure brand n = 95, no disclosure celebrity n = 93, disclosure brand n = 38, disclosure celebrity n = 55.
abc
M eans with a different superscript in the same row differ significantly at p < .05.
Figure 1. Interaction effects between advertising source and disclosure on conceptual persuasion knowledge.
Figure 2. Interaction effects between advertising source and disclosure on intention to share.
Conclusion, discussion and implications
This study examined the effects of the source of Facebook advertisements (i.e., brand or celebrity) and sponsorship disclosure on persuasion knowledge and consumer responses (i.e., attitude towards the advertisement, attitude towards the brand and intention to share). The outcomes of this study emphasize that both the source of Facebook advertisements as well as sponsorship disclosures affect consumers’ conceptual persuasion knowledge, providing that consumers recalled the disclosure. In addition, the effects of Facebook advertisements posted by the brand and
advertisements with disclosure on attitude towards the advertisement and intention to share were mediated by conceptual persuasion knowledge. Finally, the source of Facebook advertisements and sponsorship disclosure interact with each other on consumers’ conceptual persuasion knowledge and intention to share.
The results of this study show that the source of Facebook advertisements is of important matter, as it was found that consumers could more easily recognize Facebook posts by brands as advertisements, in comparison to posts posted by a celebrity. This finding is in accordance with research regarding advertising, stating that brands are more likely to be associated with commerce
than celebrities (Bhatnagar et al., 2004; Van Reijmersdal et al., 2005). In addition, this assumption is in line with the source credibility theory (Petty & Caciaoppo, 1986), as consumers did not foresee a commercial intent by celebrities and therefore did not activate their conceptual persuasion
knowledge. However, consumers' attitudinal persuasion knowledge was not activated. This means that despite the recognition of the persuasive intent of a message when posted by a brand,
consumers did not critically process the sponsored content. This finding conflicts with the reactance theory (Brehm, 1966), which implies that consumers generally do not like to be persuaded and therefore show skepticism and critical feelings towards advertising. A possible explanation for this effect can be found considering the fact that consumers, despite the source, were exposed to a celebrity endorsed advertisement. The liking of a celebrity could cause consumers to not mind the persuasive intention of the message, and therefore not activating their attitudinal persuasion knowledge. Future research therefore could include celebrity liking as a measure in their study.
In addition to the source of advertisements, the results of this study also stress the importance of sponsorship disclosures by Facebook posts. As advertisements in the newsfeed of Facebook have the same format as posts generated by consumers, it was expected that consumers might have trouble differentiating these two types. We demonstrate that disclosures help consumers to make the distinction between commercial and consumer-generated content on Facebook, as we found that witnessing a disclosure activates consumers' conceptual persuasion knowledge. These findings show that disclosures fulfill their main purpose, which is to inform consumers about the presence of sponsored content and warn them about its persuasive intent (Cain, 2011). However, our findings regarding consumers' attitudinal persuasion knowledge conflict with prior research in this domain (Boerman et al., 2012). The recognition of the Facebook post as advertising due to sponsorship disclosure did not lead to distrust or critical processing of the message. In line with prior argumentation regarding the absence of an effect between the source and attitudinal persuasion knowledge, it is possible that celebrity liking also applies for sponsorship disclosure.
not found, the source or a disclosure only affects consumer responses when consumers recognize the post as advertising. Our findings demonstrated that an advertisement posted by a brand
negatively affected consumers' attitude towards the advertisement and intention to share, due to the activation of conceptual persuasion knowledge. In a similar vein, a disclosure provoked the same negative effects, through conceptual persuasion knowledge. Considering that consumers generally do not like advertising (Brehm, 1966), it is not surprising that the consumer responses regarding the advertisement are negative.
Furthermore, two interaction effects between the source of the advertisements and sponsorship disclosure were found. First, our findings demonstrate that advertisements by a celebrity that include a disclosure, activate consumers’ conceptual persuasion knowledge.
Additionally, advertisements by brands without disclosure also induced consumers’ activation of conceptual persuasion knowledge. In other words, consumers need to be warned about sponsored content when posted by a celebrity, whereas no warning is needed when posted by a brand, as consumers here directly could recognize a persuasive attempt and thus activate their conceptual persuasion knowledge. These findings stress the importance of a sponsorship disclosure. In line with the source credibility theory (Petty & Cacioppo, 1986), consumers do not immediately recognize a persuasive intent when posted by a celebrity and thus can be persuaded without being aware when not informed about this (Cain, 2011). Moreover, our findings are in concordance with previous literature about the efficacy of a disclosure on conceptual persuasion knowledge (Boerman et al., 2012).
Second, our results demonstrate an interaction effect between the source of the advertisements and disclosure on the intention to share the advertisement. The inclusion of a disclosure increases consumers’ intention to share the advertisement, when posted by a celebrity. In other words, consumers do not mind the persuasive intention of a message coming from a celebrity, as their intention to share is not negatively affected by this recognition. This finding differs from our expectations based on the source credibility theory (Petty & Caciaoppo, 1986) and the change
of meaning principle (Friestad & Wright, 1994), as the recognition of advertising did not cause detachment or feelings of deception, but rather an increase of intention to share. These findings suggest that marketers can benefit from including celebrities in their social media strategies. This study was the first to show how the source of Facebook advertisements and sponsorship disclosure influences conceptual persuasion knowledge and subsequently, affects consumer responses. Theoretically, this study implies that conceptual persuasion knowledge is of essential value for understanding the effects of both the source of Facebook advertisements and disclosures. When the source or a disclosure induces an activation of conceptual persuasion knowledge, consumer responses regarding the advertisement (i.e., attitude towards the brand and intention to share) become qualitatively different. However, even though previous research has put emphasis on taking both conceptual and attitudinal persuasion knowledge into account (Rozendaal et al., 2011; Boerman et al., 2012), the present study shows that the effects of both the source of advertisements as well as disclosure only account for conceptual persuasion knowledge. This might be due to the fact that previous research solely focused on the effects of television, in which there was no celebrity endorsement. The extent to which consumers like a celebrity on a personal level could possibly explain the non-appearance of the effect. This assumption can be supported by Dekker and Van Reijmersdal (2010), who found that celebrity credibility influenced consumers’ brand attitude. Therefore, our findings suggest that future research regarding embedded advertising on social media should include both types of persuasion knowledge and celebrity liking into their study.
This study also holds important practical implications. Our findings suggest that marketers should invest in celebrity endorsements on Facebook which include the celebrity posting the advertisement, but also contain a disclosure. In doing so, they do not only comply with the guidelines from the FTC, they also generate a high intention to share. Moreover, celebrities also benefit most from the above stated combination. Advertisements posted by the brand, which in combination of a disclosure or through conceptual persuasion knowledge, decreased consumers'
intention to share or led to negative advertisement attitudes, respectively. Because these advertisements also show the celebrity, it is possible that the effects could spill over.
Furthermore, this study underscores the importance of consumers' recall of disclosures. As the results of our pre-test demonstrated, the original disclosures on Facebook were not significantly noticed. Without witnessing a disclosure, consumers are still susceptible for persuasion without being aware. As our results showed, this affected persuasion knowledge and consumer responses. The present study demonstrates that a bolder type of disclosure is capable of informing consumers about the commercial intent of the message. Therefore, we recommend representatives regarding the regulations and guidelines of sponsorship disclosures on social media to incorporate
requirements concerning the format of disclosures into their policies.
More research is needed regarding celebrity endorsements and disclosure on Facebook. Because the present study solely focused on Facebook advertisements that appear in the Facebook newsfeed, the effects of other sponsored content on the social network site, such as advertisements placed on the right side of the newsfeed, are yet to be studied. Furthermore, this study examined the effects of one type of disclosure on Facebook. Future research could also focus on other types of disclosure, which can include warnings that come from the source of the advertisement. This may take the form of putting a hashtag with the words ‘ad’ or ‘sponsored’ behind their message. Finally, considering that Facebook is a social network site in which consumer engagement is key, future research could also examine which effects social normative influences have on persuasion knowledge and consumer responses. As Facebook offers the possibilities to “like” and place comments beneath posts, this could lead to different effects.
References
Amos, C., Holmes, G., & Strutton, D. (2008). Exploring the relationship between celebrity endorser effects and advertising effectiveness. A quantitative synthesis of effect size. International Journal of Advertising, 27(2), 209-234.
Bhatnagar, N., Aksoy, L., & Malkoc, S. A. (2004). Embedding brands within media content: The impact of message, media, and consumer characteristics on placement efficacy. The psychology of entertainment media, 99.
Boerman, S. C., Van Reijmersdal, E. A., & Neijens, P. C. (2012). Zijn sponsorvermeldingen in televisieprogramma’s effectief? Tijdschrift voor Communicatiewetenschap, 40(4), 46-59.
Boerman, S. C., Van Reijmersdal, E. A., & Neijens, P. C. (2012). Sponsorship disclosure: Effects of duration on persuasion knowledge and brand responses. Journal of Communication, 62(6), 1047-1064.
Boerman, S. C., Van Reijmersdal, E. A., & Neijens, P. C. (2014). Effects of sponsorship disclosure timing on the processing of sponsored content: a study on the effectiveness of European disclosure regulations. Psychology & Marketing, 31(3), 214-224.
Brehm, J. W. (1966). A theory of psychological reactance. New York: Academic Press.
Bruner, G. C., Hensel, P. J., & James, K. E. (2009). Marketing scales handbook. Chicago: American Marketing Association.
Cain, R. M. (2011). Embedded advertising on television: disclosure, deception, and free speech rights. Journal of Public Policy & Marketing, 30(2), 226-238.
Calfee, J. E., & Ringold, D. J. (1994). The 70% majority: Enduring consumer beliefs about advertising. Journal of public policy & marketing, 228-238.
Campbell, M. C. (1995). When attention-getting advertising tactics elicit consumer inferences of manipulative intent: the importance of balancing benefits and investments. Journal of Consumer Psychology, 4(3), 225-254.
Campbell, M. C., Mohr, G. S., & Verlegh, P. (2007). Examining effects of product placement and sponsorship disclosure: a flexible correction approach. In: Proceedings of the 34th International La
Management.
Campbell, M. C., Mohr, G. S., & Verlegh, P. W. (2013). Can disclosures lead consumers to resist covert persuasion? The important roles of disclosure timing and type of response. Journal of Consumer Psychology, 23(4), 483-495.
Chiu, H. C., Hsieh, Y. C., Kao, Y. H., & Lee, M. (2007). The determinants of email receivers’ disseminating behaviors on the Internet. Journal of Advertising Research, 47(4), 524-534.
Chu, S. C., & Kamal, S. (2008). The effect of perceived blogger credibility and argument quality on message elaboration and brand attitudes: an exploratory study. Journal of Interactive Advertising, 8(2), 26-37.
Dekker, K., & Van Reijmersdal, E. (2010). Waarschuwingen, beroemheden en brand placement: de effecten van type waarschuwing en geloofwaardigheid op kijkerreacties. Tijdschrift voor
Communicatiewetenschap, 38(4), 320-337.
Dekker, K., & Van Reijmersdal, E. A. (2013). Disclosing celebrity endorsement in a television program to mitigate persuasion: how disclosure type and celebrity credibility interact. Journal of Promotion Management, 19(2), 224-240.
Eckler, P., & Bolls, P. (2011). Spreading the virus: emotional tone of viral advertising and its effect on forwarding intentions and attitudes. Journal of Interactive Advertising, 11(2), 1-11.
Erdogan, B. Z. (1999). Celebrity endorsement: a literature review. Journal of marketing management, 15(4), 291-314.
