“Is it for real or a sponsorship deal?”
The effects of sponsorship disclosures by influencers on persuasion knowledge and advertising outcomes in the context of Instagram for search versus experience goods
Esmée Oudman 12589330 Master Thesis
Graduate School of Communication Persuasive Communication
Dr. Saar Mollen June 25, 2020
Acknowledgements
Hereby I want to thank my supervisor Dr. Saar Mollen. Her guidance and feedback have made the Master’s thesis an inspiring challenge for me. I am very thankful for the new skills and insights that I have gained due to performing this research. Furthermore, I would like to thank everyone who participated in the pretests and experiment of my study, as well as the people who helped me recruiting participants. I would also like to express my gratitude and appreciation to my close friends Milou Nipshagen and Muriël de Groot who took the time to read through my work and provide valuable suggestions. Finally, special thanks go to my boyfriend Jeroen Grasmeijer who supported me emotionally throughout the process.
Abstract
Brands are trying to find their way to consumers by selling their products through influencers. However, consumers often do not recognize sponsored posts by influencers as advertising. Sponsorship disclosures may help consumers understand the commercial relationship between influencers and brands. Consequences of these sponsorship disclosures have been investigated in many studies, however, the results are mixed. It appears that the impact of disclosures may depend on the type of product that is promoted by the influencer. This study is the first to compare the effects of product types by distinguishing between search and experience goods. By conducting an online experiment among female Instagram users, this study investigated whether a disclosure on Instagram evokes critical attitudes (i.e., attitudinal persuasion knowledge), how this subsequently impacts advertising outcomes (i.e., brand attitude, purchase intention, and online behavioral intentions), and if these effects differ between an Instagram post including a search good (i.e., home decoration ) and an experience good (i.e., cosmetics). It was expected that a sponsorship disclosure will increase attitudinal persuasion knowledge which consequently leads to more negative persuasion outcomes for an experience good, whereas for a search good these effects were not expected. However, results showed that there was no interaction effect between disclosure and product type on persuasion outcomes, neither there was a moderated mediation effect. Unexpectedly, there was a significant main effect of both disclosure and product type on brand attitude and purchase intention. Hence, the study contributes to the research area of sponsorship disclosures as well as advertising practice, as it provides deeper insights into when and why disclosures are effective or not.
Keywords: sponsorship disclosures, influencer marketing, Instagram, persuasion knowledge, search goods, experience goods
Introduction
Nowadays the persuasive intent of marketeers within traditional advertising formats is easily recognized by potential buyers. Many consumers have learned to identify persuasive tactics and have become suspicious of the advertiser (Boerman, Willemsen & Van Der Aa, 2017). This often results in resistance and rejection of the offer that is communicated in the message (Friestad & Wright, 1994). Consumers’ skepticism challenges marketeers and requires new ways to sell products or services. Since consumers have greater trust in information and recommendations provided by peers (Batinic & Appel, 2013), and do not perceive their messages as persuasive attempts (Verlegh, Ryu, Tuk, & Feick, 2013), brands are trying to find their way to consumers by selling their products through influencers.
Influencers are opinion leaders with a large number of followers who are seen as credible and easy-to-relate sources (Boerman, 2020). Social media enable people to develop a relationship with influencers, which makes them interesting brand ambassadors (Tsai & Men, 2013). To shape their followers’ opinions, influencers receive products for free or are being paid to recommend the brand on their social media profiles (De Veirman & Hudders, 2019). For instance, by wearing an outfit and mentioning the brand in picture captions or tags. Since consumers choose whom they want to follow, they are likely to respond positively to
messages coming from influencers (Colliander & Dahlén, 2011). Currently, Instagram is the most popular platform for influencer marketing (Mediakix, 2019). On this platform, the sponsored content by influencers blends seamlessly with their own and other non-advertising content (Wojdynski, 2016).
However, consumers oftendo not recognize influencer marketing on Instagram as advertising (Evans, Phua, Lim, & Jun, 2017). Therefore, ethical concerns about influencer marketing are rising. The Federal Trade Commission (FTC) recommends influencers to disclose such sponsorships since this may help consumers recognize the influencer’s
commercial relationship with brands (FTC, 2019). This can be realized on social media by using hashtags, such as ‘#sponsored’ or ‘#ad’ (Evans et al., 2017). Nevertheless, it seems that many influencers are not using these disclosures yet (FCT, 2017). To draw attention to the importance of transparency around advertising on social media, in August 2019 the Dutch Stichting Reclame Code (SRC) launched a campaign named ‘#Ad Recognize advertising on the internet’. The SRC encourages influencers to be transparent about whether they have been paid to name a brand in, for example, a vlog or post (SRC, 2019).
Sponsorship disclosures are designed to help consumers recognize the persuasive intent of a message (i.e., conceptual persuasion knowledge), and to prompt people to
elaborate critically on the persuasive attempt (i.e., attitudinal persuasion knowledge; Friestad & Wright, 1994). It is important to distinguish between conceptual and attitudinal persuasion knowledge to understand disclosure effects, as recognizing the persuasive intent may not always induce critical and distrusting feelings (Ham, Nelson & Das, 2015). Consequences of sponsorship disclosures have been investigated in many studies. Some findings show that using disclosures on Instagram has a negative effect on advertising outcomes (De Veirman & Hudders, 2019; Evans et al., 2017), whereas other researchers found no effect (Johnson, Potocki, &Veldhuis, 2019), or even a positive effect (Boerman, 2020).
The mixed findings of previous research might be explained by the different products used in these studies. It appeared that the studies demonstrating negative effects selected products that are more experiential (e.g., donuts; Evans et al., 2017), while the studies that found no effect or a positive effect selected products considered as objective (e.g., a dress; Boerman, 2020). A relevant distinction to make in the context of this study is that between search and experience goods. Search goods are products of which the attributes are available for inspection (e.g., clothing, house furnishings, or cameras), which allows consumers to form a firsthand assessment of the product prior to purchase (Nelson, 1970). On the contrary,
experience goods need to be used to understand how well they fit your needs (e.g., food, cosmetics, or books), hence the product can be evaluated after purchase (Nelson, 1970).
Therefore, the impact of disclosures may depend on the type of product that is promoted by the influencer. The literature argues that purchasing experience goods is associated with more risk compared to search goods (Mitra, Reiss & Capella, 1999).
Subsequently, higher levels of risk predict how motivated people are to critically elaborate on the message (Petty & Cacioppo, 1984) and to verify its credibility (Wojdynski & Evans, 2020). Thus, it is expected that a disclosure is more likely to activate attitudinal persuasion knowledge and lead to negative responses for experience than for search goods.
Altogether, this study aims to gain more insight into when and why disclosures may affect persuasion outcomes. Furthermore, this study goes beyond studying the simple recognition of advertising as a result of disclosures, by also examining how consumers evaluate the persuasive attempt. Accordingly, it is one of the few studies that takes into account a multidimensional form of persuasion knowledge. Moreover, this study is, to the best of the author’s knowledge, the first to compare the effects of product types for sponsorship disclosures. Many studies on disclosures call for future research to provide insights into the differences between products (e.g., De Veirman & Hudders, 2019; Boerman, 2020). The current study will respond to this request by examining the fundamental
differences between search and experience goods that may help to reconcile earlier research showing disparate outcomes of disclosures. The study contributes to the research area of sponsorship disclosures as well as advertising practice, as it provides deeper insights into when and why disclosures are effective or not. Accordingly, the central questions addressed in this research are:
What are the effects of sponsorship disclosures by influencers on Instagram on consumers’ attitudinal persuasion knowledge and subsequent attitude toward the brand, purchase intention, and online behavioral intentions, compared to when no disclosure is present? And how do these effects differ between search and experience goods?
The remaining part of this study proceeds as follows: First, theories and empirical research on sponsorship disclosures and product types are discussed. Second, the method will be outlined. Third, the results of the analyses are presented. Finally, implications and
limitations are discussed and directions for future research are provided.
Theoretical Framework
The effects of sponsorship disclosures on persuasion knowledgeThe effects of disclosures on advertising outcomes can be explained by the persuasion knowledge model (Friestad & Wright, 1994). This model implies that persuasion knowledge develops through life and enables consumers to recognize, evaluate, and cope with persuasion attempts. The use of influencers to promote products makes it harder for consumers to
recognize such content as advertising (Boerman et al., 2017; Evans et al., 2017). Disclosures are meant to inform consumers about the commercial intent of this new advertising format, thus facilitates in activating consumers’ persuasion knowledge (De Veirman & Hudders, 2019; Evans et al., 2017).
A distinction can be made between conceptual and attitudinal persuasion knowledge (Rozendaal, Lapierre, Van Reijmersdal & Buijzen, 2011). Conceptual persuasion knowledge concerns the recognition and understanding of advertising in general, and in new formats such as sponsored content (Rozendaal et al., 2011). This cognitive dimension represents how much consumers recognize the persuasion intent of the message’s sender (Ham et al., 2015).
Attitudinal persuasion knowledge refers to how consumers evaluate the persuasive attempt and whether it evokes critical attitudes, such as skepticism, or critical feelings about honesty, trustworthiness, and credibility (Rozendaal et al., 2011). Thus, this evaluative dimension goes beyond the mere recognition of advertising (Ham et al., 2015).
The persuasion knowledge model posits that once consumers recognize the persuasive intent of a message (i.e., conceptual persuasion knowledge), they develop critical and
distrusting feelings (i.e., attitudinal persuasion knowledge; Friestad & Wright, 1994). Thus, if consumers learn that the source’s action concerns a persuasion tactic, a ‘change-of-meaning’ will occur, which may impact persuasion outcomes (Friestad & Wright, 1994). For instance, if an influencer is expressing her happiness with a cosmetic product on Instagram, a
disclosure may trigger consumers to verify the credibility of this opinion, such as is this cosmetic product really as good as she states in the post. This is likely to affect their attitude toward the message and the product since experiencing such a ‘change-of-meaning’ is often considered to be unpleasant because people feel offended that someone is trying to trick them (Friestad & Wright, 1994).
However, the activation of persuasion knowledge will not always lead to negative outcomes (Nelson & Ham, 2012). When consumers do not perceive the tactic as
manipulative, the negative effect activated by persuasion knowledge is attenuated (Friestad & Wright, 1994). Accordingly, disclosures may trigger conceptual persuasion knowledge, but this does not always activate attitudinal persuasion knowledge (Ham et al., 2015). Thus, in order to understand the effects of disclosures it is important to take both dimensions into account (Ham et al., 2015). Therefore, the current study distinguishes between conceptual and attitudinal persuasion knowledge. Moreover, this study examines whether the effects of disclosures on persuasion knowledge and subsequent persuasion outcomes could be explained by product type.
The effects of sponsorship disclosures on advertising outcomes
Prior research has shown mixed findings of sponsorship disclosures on advertising outcomes. However, the majority of studies that investigated the impact of disclosures
indicated it is likely to result in negative effects. Boerman et al. (2017) found that people were less inclined to share, like, or comment on the Facebook post (i.e., online behavioral
intentions) when a disclosure was present versus absent. Additionally, research on disclosing sponsored content in blogs demonstrated that both brand attitude and purchase intention were lower in the disclosure compared to the no disclosure condition (Van Reijmersdal et al., 2016; Janssen, van Sprang & Fransen, 2017). Furthermore, using disclosures in sponsored online news stories negatively affected brand attitudes and users’ online behavioral intentions (Wojdynski & Evans, 2016).
In the context of Instagram, similar negative effects on brand attitude were found for disclosing sponsored content (De Veirman & Hudders, 2019). Moreover, Evans et al. (2017) indicated that disclosures negatively affected Instagram users’ attitude toward the sponsored brand and decreased their online behavioral intentions. In contrast to these studies, Johnson et al. (2019) found no effects by a disclosure on brand attitude and purchase intention.
Surprisingly, the results of another more recent study even found that online behavioral intentions were higher when the sponsorship was disclosed on Instagram, compared to when no disclosure was present (Boerman, 2020).
It should be noted, however, that most studies demonstrating negative effects selected more experiential products. For instance, coffee (Boerman et al., 2017), donuts (Evans et al., 2017), or energy bars (De Veirman & Hudders, 2019), are products that may lead to higher feelings of betrayal because they are considered to be difficult to evaluate prior to purchase (Nelson, 1970). On the contrary, the study that did find positive effects used a dress
(Nelson, 1970). Moreover, the study that found no effects selected three products for the experiment (i.e., sunscreen, fast food, and a vacuum; Johnson et al., 2019). Whereas sunscreen and food are generally more experiential, a vacuum could be considered as a product which features are objective (Hsieh, Chiu & Chiang, 2005). Accordingly, a disclosure is less likely to generate critical and distrusting feelings for these types of products.
This may explain why disclosures led to different effects in previous studies. However, most studies selected products that are considered as experiential, whereas few studies selected products that are objective. As the effects of sponsored content may differ between products (e.g., De Veirman & Hudders, 2019; Boerman, 2020), this study will examine further how product type impacts the effectiveness of disclosures.
Product type: Search and experience goods
As outlined above, the extent to which attitudinal persuasion knowledge becomes activated and the subsequent impact on persuasion outcomes likely depends on the type of product that is promoted by the influencer. A relevant distinction to make in this context is between search and experience goods. Search goods are products which features are stable, objective, and easy to evaluate (Hsieh et al., 2005). Accordingly, more pre-purchase
knowledge is available to the consumer (Nelson, 1970). Thus, with search goods you can assess its value before you purchase it (e.g., clothing, house furnishings, or cameras; Nelson, 1970). On the contrary, experience goods need to be used to understand how well they fit your needs. You can only ascertain the value by consuming the product, which happens well after the purchase decision (e.g., food, cosmetics, or books; Nelson, 1970).
As a result of these differences in pre-purchase knowledge, the recommendations of others such as friends or peers are more useful for purchases of experience versus search good (Nelson, 1970). This idea is confirmed in several empirical studies. Mitra et al., (1999) found
that the use of personal information sources (e.g., opinions of friends and relatives) was higher for experience good purchases, whereas impersonal information sources (e.g.,
promotion and advertising on traditional media channels) were more relevant for search good purchases. In the context of electronic word of mouth (eWOM), it was demonstrated that online recommendations had a greater effect on people’s purchase behavior for experience than for search goods (Senecal & Nantel, 2004; Huang, Lurie & Mitra, 2009; Park & Lee, 2009).
Nonetheless, consumers make use of personal information sources for both product types (Mitra et al., 1999), since people have greater trust in information provided by peers than by brands themselves (Batinic & Appel, 2013). Therefore, also for search goods the recommendations of influencers are expected to be useful, but less necessary compared to experience goods. Since people rely more on the recommendations of others for purchasing experience versus search goods, it may be that consumers’ responses to sponsorship
disclosures on Instagram by influencers vary between the product types. The following section discusses two theories that may explain this: The elaboration likelihood model and psychological reactance theory.
The moderating effect of product type
First, the idea that consumers’ responses to disclosures differ between Instagram posts including search versus experience goods can be explained by the perceived risk associated with both product types (Mitra et al., 1999). Higher levels of perceived risk predict how involved consumers are in the decision-making process of a specific product (Dholakia, 1997). Involvement can be explained as motivation stemming from a degree of personal relevance of the message to the consumers (Petty, Cacioppo & Goldman, 1981). According to the elaboration likelihood model (ELM), the likelihood of elaboration on a message is
determined by the consumer’s motivation to process (Petty & Cacioppo, 1986). Heightened motivation, as a result of high involvement, leads people to scrutinize the message more carefully and critically (Petty & Cacioppo, 1984). Consumers who critically evaluate a message are more likely to consider the source’s motivation for creating or distributing the message (Wojdynski & Evans, 2020). Thus, consumers with greater levels of involvement are more inclined to verify the credibility of an influencer. This motivation may be triggered by a sponsorship disclosure.
Sah, Malaviya & Thompson (2018) empirically examined the differential effects due to automatic versus deliberative processing of disclosures. In line with the premises in the ELM, this study found that when participants carefully thought about the information
presented in the disclosure, this increased perceptions of bias, lowered their perceived trust in the blogger, and subsequently decreased the persuasive effects (Sah et al., 2018). However, when a disclosure was processed more automatically, consumers reported more trust in the blogger and the sponsor than consumers who read a post with no disclosure (Sah et al., 2018). These findings suppose that a disclosure may even lead to more positive effects when
processed automatically.
Based on the ELM, it is supposed that disclosures are processed differently between search and experience goods. Since information search is relatively easy for search goods (Nelson, 1970), purchasing a search good is associated with more certainty and less perceived risk (Mitra et al., 1999). On the contrary, experience goods are difficult to evaluate before purchase (Nelson, 1970), which leads to higher perceived risk (Murray & Schlater, 1990; Guseman, 1981). Prior research confirmed that people encounter a higher degree of risk for experience than for search goods (Mitra et al., 1999). The recommendations of others have the potential to diminish the risk that is involved with purchasing experience goods (Mitra et
al., 1999). However, when a disclosure reveals that an influencer has hidden motives, likely the perceived risk that is associated with experience goods is not reduced.
Therefore, it is expected that disclosures result in different effects between the two product types. According to the ELM (Petty & Cacioppo, 1984), the motivation to scrutinize the credibility of a message is likely to be greater for experience goods since these products are associated with a higher level of perceived risk (Mitra et al., 1999). Thus, consumers will elaborate more critically on disclosures when Instagram posts include experience goods. On the contrary, for search goods consumers are likely less motivated to critically evaluate a disclosure and to consider the credibility of the source since these products are associated with less perceived risk (Mitra et al., 1999). Thus, disclosures are expected to activate
attitudinal persuasion knowledge and subsequently negatively impact persuasion outcomes if influencers are promoting experience goods, whereas for search goods these effects are less likely to occur.
A second reason why consumers’ responses to disclosures vary for search versus experience goods can be explained by psychological reactance theory (Brehm & Brehm, 1981). The term ‘reactance’ refers to a negative emotional reaction that occurs when a person feels that someone else is taking away his choices, consequently this person will try to restore the threatened freedom by resisting the persuasive attempt (Brehm & Brehm, 1981). Hence, consumers’ freedom to avoid advertising messages is threatened when they encounter covert advertisements in an environment where they seek to consume informational or entertainment content (Edwards, Li & Hee, 2002). When a disclosure makes consumers recognize the Instagram post by an influencer as an advertisement, this violates their expectations and freedom to avoid advertisements, which consequently results in negative perceptions of both the message and the advertiser (Miron & Brehm, 2006). These negative evaluations of a
reactance, which means that higher levels of reactance are likely topredict higher levels of attitudinal persuasion knowledge (Wojdynski & Evans, 2020). Thus, attitudinal persuasion knowledge is likely to become activated when people feel that their freedom to avoid advertisements is being treated.
To what extent consumers experience reactance depends on the relevance of the advertising content to consumers’ situational motivational goals (Wojdynski & Evans, 2020). In other words, the negative effects of a disclosure are greater when the message does not match with the consumer’s goal in seeking content in the first place. For experience goods, the consumers’ goal is to find credible and honest user experiences, but a disclosure reveals that the influencer got paid to share an opinion. Therefore, the message contains hidden motives and does not meet consumers’ need for valid information. Since consumers who evaluate an experience good are strongly dependent on personal information sources, a disclosure implicates that the message’s content may be less relevant. In line with reactance theory (Brehm & Brehm, 1981), this may lead to higher feelings of betrayal which
consequently results in more negative effects.
For search goods, the recommendations of others are also relevant (Mitra et al., 1999). However, consumers are less dependent on this source since these products can be observed objectively and more easily (Hsieh et al., 2005). Yet, sponsored posts including search goods are able to deliver valid information (Bekcer-Olsen, 2003), thus the message’s content may still be relevant. When consumers do not perceive the recommendation as manipulative (Nelson & Ham, 2012), this can mitigate the harshness of negative responses to disclosures (Wojdynski & Evans, 2020). Therefore, a sponsorship disclosure is likely to result in greater negative effects for experience versus search goods.
The current study
This study will investigate whether product type is a moderating variable that may explain when and why sponsorship disclosures are effective, by distinguishing between search and experience goods. To test the interaction effect between disclosures and product type, the following hypothesis is proposed:
H1: A sponsorship disclosure in an Instagram post that includes an experience good will
lead to a more negative brand attitude (H1a), purchase intention (H1b), and online behavioral intentions (H1c), compared to when no disclosure is present. On the contrary, a sponsorship disclosure in an Instagram post that includes a search good will not lead to a more negative brand attitude, purchase intention, and online behavioral intentions, compared to when no disclosure is present.
Besides, this study will examine whether the effects of disclosures are mediated by attitudinal persuasion knowledge. A moderated mediated hypothesis is proposed:
H2: A sponsorship disclosure in an Instagram post that includes an experience good will
increase attitudinal persuasion knowledge which consequently leads to a more negative brand attitude (H2a), purchase intention (H2b), and online behavioral intentions (H2c), compared to when no disclosure is present. On the contrary, a sponsorship disclosure in an Instagram post that includes a search good will not increase attitudinal persuasion knowledge, and will therefore not consequently lead to a more negative brand attitude, purchase intention, and online behavioral intentions, compared to when no disclosure is present.
The conceptual model is shown in Figure 1.
Figure 1 Conceptual model
Method
Design and participantsThe hypotheses were tested with an online experiment with a 2 (disclosure: disclosure vs. no disclosure) x 2 (product type: search good vs. experience good) between-subjects design. In total 234 participants were recruited through invitations on social media. Only women aged 18 years and older who indicated to use Instagram regularly (on a daily, weekly, or monthly basis) could participate in the study. Participants who did not meet these criteria (n = 6) were excluded. Since this study is interested in the effects of a disclosure when consumers are made aware of this, participants who incorrectly remembered seeing a
disclosure in the control condition or not in the experimental condition were also excluded (n = 34). This left a final sample of 194 participants (disclosure and search good n = 46, disclosure and experience good n = 49, no disclosure and search good n = 52, no disclosure and experience good n = 47).
The mean age of the female participants was 24.74 years (ranging from 18 to 54, SD = 6.05). The current sample reflects the average Instagram user, as statistics show that women use the medium more than men and the core users are aged between 18 and 35 years (Statista, 2020). Most participants fell into this age range (94.3%), used Instagram daily (93.8%), had a Dutch nationality (95.4%), and completed higher education (93.8%).
Procedure
The study was presented as a study about people’s responses to Instagram posts. Participants first signed an informed consent and filled in their demographics. Participants who did not meet the inclusion criteria were immediately directed to the end of the study and thanked. After answering demographic and background questions, participants were requested to carefully look at an Instagram post and the description underneath. Participants were randomly assigned to one of the four conditions, and either saw an Instagram post about a search or experience good that did or did not include a disclosure. Following this exposure, the participants answered questions about brand attitude, purchase intention, and online behavioral intention, followed by attitudinal persuasion knowledge, reactance (i.e., perceived intrusiveness), and questions pertaining to the manipulation checks (i.e., conceptual
persuasion knowledge and perceived product type). This particular order of questions made sure that brand responses (e.g., attitude and intentions) were not primed by questions that revealed the commercial nature of the Instagram post (e.g., conceptual persuasion
knowledge). The questionnaire ended with one control question, namely whether they were familiar with the influencer before participating in this study. Finally, participants could leave comments and were debriefed and thanked.
Pretests
Two pretests were conducted in order to find appropriate stimulus materials for the experiment. The first pretest (n = 26) aimed to select an appropriate search and experience good. See Appendix 1 for a detailed description of this pretest. Moreover, the goal of the second pretest (n = 21) was to select one influencer and two pictures by this person that corresponded to the products selected in pretest one. See Appendix 2 for a detailed description of this pretest
Stimulus materials
Based on the two pretests stimulus materials were developed that consisted of two posts from the selected influencer (i.e., Rachel van Sas), showing either a search good (i.e., home decoration) or an experience good (i.e., cosmetics), which included a disclosure (‘#ad’) or not. See Figure 2 for two examples of the material. Apart from these manipulations of sponsorship disclosure and product type, the pictures were quite similar in appearance (comparable location, the influencer held a similar pose, and the picture was made from the same angle). Moreover, the designed posts included the same number of likes and both pictures did not include a brand name. The fictitious brand ‘Folily’ was created for the purpose of the experiment and was mentioned in the description of both the search and experience good conditions.
Measures
Brand attitude. Participants were asked how they felt about the brand mentioned in the Instagram post, by indicating on a seven-point semantic differential scale their agreement with five statements: “I think this brand is … unappealing/appealing, unpleasant/pleasant, boring/interesting, negative/positive, bad/good” (Bruner & Kumar, 1999).
Figure 2. Two examples of the stimulus materials: Including the search good (left) or the experience good (right),
and a disclosure (‘#ad’) or not.
Factor analysis revealed that the items load on one factor (Eigenvalue = 3.47, explained variance = 69.35%, = Cronbach's alpha = .89). The measure of brand attitude consisted of the mean score of the five items (M = 4.83 , SD = .91).
Purchase intention. Participants were asked to indicate on a scale ranging from 1 (strongly disagree) to 7 (strongly agree) to what extent they agreed with the statements: "I would like to try the brand Folily", "I want to buy the brand Folily", "I would buy other products of the brand Folily", and "I will look for the brand Folily in a store" (Spears & Singh, 2004). Factor analysis revealed that the items load on one factor (Eigenvalue = 3.12, explained variance = 78.01%, Cronbach's alpha = .90). The mean score of the four items was used as a measure of purchase intention (M = 3.57, SD = 1.23).
Online behavioral intentions. To measure participants’ intention to engage online with the post, they were asked to indicate on a scale ranging from 1 (strongly disagree) to 7
(strongly agree) to what extent they agreed with the statements: “I would comment on the post on Instagram", "I would like the post on Instagram", "I would share the post via a private message on Instagram", and "I would save the post on Instagram" (Boerman, 2020). These items include all the ways a person can engage with an Instagram post online. Factor analysis revealed that the items load on one factor (Eigenvalue = 2.01, explained variances = 51.91%, Cronbach's alpha = .66). This alpha value suggest that the internal consistency is reasonably reliable. However, the scale could not be improve by deleting an item. Thus, the measure of online behavioral intentions consisted of the mean score of the four items (M = 2.42, SD = 1.01).
Attitudinal persuasion knowledge. Participants were asked on a seven-point semantic differential scale how they felt about the Instagram post using five questions: “I think the Instagram post was … dishonest/honest, not credible/credible, untrustworthy/trustworthy, biased/unbiased, unconvincing/convincing” (Ohanian, 1990). Factor analysis revealed that the items load on one factor (Eigenvalue = 3.52 , explained variances = 70.44%, Cronbach's alpha = .89). These five items were reversed coded and the mean was used as a measurement of attitudinal persuasion knowledge (attitudinal PK; M = 4.16 , SD = 1.18).
Manipulation checks. Conceptual persuasion knowledge was measured by asking participants to indicate on a scale ranging from 1 (strongly disagree) to 7 (strongly agree) to what extent they agreed with the statement: "The Instagram post that I saw was an
advertisement" (Boerman, 2020; Evans et al., 2017), to check if this differed between the disclosure and no disclosure conditions. To check whether the manipulation of product type was successful, participants were asked to indicate on a scale ranging from 1 (strongly
disagree) to 7 (strongly agree) to what extent they agreed with the statement: “It is important for me to use this product before I can evaluate its quality".
Control variables. Disclosure recognition was measured by asking participants whether they remembered seeing a sponsorship disclosure (‘#ad’) in the Instagram post (yes/no). The frequency of Instagram usage was measured by asking participants how often they used Instagram (never/yearly/monthly/weekly/daily). Furthermore, participants were asked if they knew Rachel van Sas before seeing her Instagram post in this study (yes/no; 82.0% indicated they did not). In addition, people’s age in years, sex, nationality, and education level were measured.
Exploratory variable. Perceived intrusiveness was included as a measurement of reactance by asking participants to indicate on a scale ranging from 1 (strongly disagree) to 7 (strongly agree) to what extent they agreed with the following statements "If the post would show up on my Instagram timeline, I would find it … distracting, disturbing, forced,
interfering, intrusive, invasive, obtrusive” (Edwards et al., 2002). Factor analysis revealed that the items load on one factor (Eigenvalue = 4.34, explained variances = 62.02%, Cronbach's alpha = .89). The measure of perceived intrusiveness consisted of the mean score of the seven items (M = 3.48 , SD = 1.12).
Pre-analysis plan
To test whether the hypotheses can be confirmed or rejected, a pre-analysis plan was formulated. The interaction effect between disclosure and product type on persuasion outcomes will be tested using Two-way ANOVAs, with disclosure and product type as factors, and brand attitude, purchase intention, and online behavioral intentions as the dependent variables. Subsequently, the moderated mediation effects will be tested by using Model 8 of the PROCESS version 3.5 in SPSS (Hayes, 2018). All analyses will use 5.000
bootstrap samples to estimate bias-corrected bootstrap confidence intervals. For each dependent variable, the model will be run with the disclosure condition as the independent variable, product type as moderator, and attitudinal PK as mediator. The no disclosure conditions will be dummy coded as 0, the disclosure conditions as 1, the search good
conditions as 1, and the experience good conditions as 2. Finally, exploratory analyses will be conducted with perceived intrusiveness as a mediator, again by using Model 8 in PROCESS. See Appendix 3 for a detailed version of the pre-analysis plan.
Results
Manipulation checksParticipants perceived the Instagram posts more as an advertisement in the disclosure (M = 6.38, SD = .95) than in the no disclosure conditions (M = 5.92, SD = 1.30), F(1, 192) = 2.89, p = .006. This did not differ significantly between the search good (M = 6.03, SD = 1.12) and the experience good conditions (M = 6.26, SD = 1.20). Thus, the manipulation of disclosure was successful. It must be noted, however, that the perceptions of conceptual persuasion knowledge were high in both conditions.
The manipulation of product type was also successful: Participants found it more important to use the product they saw on the picture before they could evaluate its quality in the experience good (M = 5.69, SD = 1.34) than in the search good conditions (M = 4.81, SD = 1.55), F(1, 192) = 17.82, p < .001.
Randomization checks
To check if the four experimental groups were comparable on several important characteristics, a randomization check was conducted. The groups did not differ significantly with respect to age, F(3, 190) = 0.18, p = .912, nationality, χ 2 (18) = 16.96, p = .526,
education, χ 2 (12) = 9.91, p = .624, Instagram use, χ 2 (6) = 5.74, p = .453, and influencer
familiarity, χ 2 (3) = 0.20, p = .978. Therefore, none of these variables will be controlled for in
the analyses.
Interaction effects
The interaction effects between disclosure and product type were tested using Two-way ANOVAs. H1 predicted that a disclosure in an Instagram post including an experience good would lead to more negative persuasion outcomes, compared to no disclosure. For search goods these negative effects were not expected. Table 1 shows the means and standard deviations of the dependent variables for the four conditions.
Brand attitude. In contrast to the expectations formulated in H1a, the analysis showed that the interaction between disclosure and product type was not significant for brand attitude, F(1, 190) = 0.02, p = .903. Therefore, H1a was not supported: The disclosure did not have a more negative effect on brand attitude for the post including an experience versus a search good. However, the analysis revealed that there was a main effect of disclosure: Brand attitude was significantly lower when participants were exposed to an Instagram post with a disclosure (M = 4.69, SD = .88) compared to no disclosure (M = 4.96, SD = .92), F(1, 192) = 4.10, p = .044, η2 = .021, but the effect size should be interpreted as small. Unexpectedly, it
also appeared that there was a main effect of product type: Brand attitude was significantly lower for participants who were exposed to a post including an experience good (M = 4.55, SD = .90) compared to a search good (M = 5.11, SD = .82), F(1, 192) = 15.41, p < .001, η2 =
.099. The effect size is considered as medium.
Purchase intention. Contrary to the expectations formulated in H1b, the analysis showed that the interaction between disclosure and product type was also not significant for purchase intention, F(1, 190) = 0.29, p = .591. Thus, H1b was not supported: The disclosure
did not have a more negative effect on purchase intention for the posts including an experience versus a search good. Again, the analysis did show a main effect of disclosure: Purchase intention was significantly lower in the disclosure (M = 3.36, SD = 1.20) compared to the no disclosure conditions (M = 3.77, SD = 1.24), F(1, 192) = 5.20, p = .024, η2 = .027,
but the effect size is considered as small. Surprisingly, there was also a main effect of product type: Purchase intention was significantly lower for participants who were exposed to a post including an experience good (M = 3.26, SD = 1.30) compared to a search good (M = 3.88, SD = 1.08), F(1, 192) = 12.89, p < .001, η2 = .064. The effect size should be interpreted as
medium.
Online behavioral intentions. In contrast to the expectations formulated in H1c, the analysis showed that the interaction between disclosure and product type was neither significant for online behavioral intentions, F(1, 190) = 0.54, p = .463. Thus, H1c was not supported: The disclosure did not have a more negative effect on online behavioral intentions for the posts including an experience versus a search good.
Table 1. Means and standard deviations of brand attitude, purchase intention, and online behavioral intentions for
the conditions of disclosure and product type.
Note: The results of the Two-way ANOVAs showed that none of the means differ significantly between the four
conditions at p < 0.05.
Search good Experience good
No disclosure Disclosure No disclosure Disclosure
Brand Attitude 5.23 (.77) 4.97 (.87) 4.66 (.98) 4.42 (.81)
Purchase intention 4.02 (1.05) 3.77 (1.11) 3.50 (1.38) 3.02 (1.18)
Online behavioral Intentions
Moderated mediation effects
The moderated mediation effects were tested using Model 8 in PROCESS. H2
predicted that a disclosure in an Instagram post including an experience good would increase attitudinal PK which consequently leads to more negative persuasion outcomes, compared to no disclosure. For search goods these effects were not expected. Table 2 presents the full regression results for these analyses.
Firstly, the regression model with attitudinal PK as dependent variable and disclosure and product type as independent variables was significant, F(3, 190) = 3.46, p = .018.
However, the strength of the prediction is moderate (R2 = .23). In contrast to the expectations,
the analyses showed that attitudinal PK is not significantly predicted by the interaction between disclosure and product type (b = 0.22, p = .512), nor by the disclosure (b = -0.11, p = .836). However, for completeness the full results of the moderated mediation analyses can be found below.
Brand attitude. The regression model with brand attitude as dependent variable and disclosure and product type as independent variables was significant, F(4, 189) = 21.66, p < .001. Therefore, the model can be used to predict brand attitude (R2 = .56). Unexpectedly,
there was no significant interaction effect between disclosure and product type (b = 0.11, p = .627), nor a direct effect by disclosure (b = -0.33, p = .335) on brand attitude. There was a relation between attitudinal PK and brand attitude (b = -0.35, p < .000), indicating that the more critical and distrusting feelings people had toward the Instagram post, the lower their brand attitude. Moreover, the index of moderated mediation was not significant (index of moderated mediation = -0.08, se = 0.12, CI -0.31; 0.17), since the confidence interval crosses zero. This implies that there is no moderated mediation effect, meaning that the presence of a disclosure does not negatively impact brand attitude via attitudinal PK, and that these indirect
Table 2. Moderated mediation effect of disclosure on brand attitude, purchase intention, and online behavioral
intentions (OBA), via attitudinal persuasion knowledge (APK), with product type as a moderator in the relationship between disclosure and APK, and between disclosure and the dependent variables.
Predictor B SE t p BC 50000 BOOT LL95 UL95 APK (mediator) Constant 3.52 0.36 9.67 .000 2.80 4.23 Disclosure -0.11 0.53 -0.21 .836 -1.15 0.93 Product type 0.36 0.23 1.55 .123 -0.10 0.82 Disclosure x product type 0.22 0.33 0.66 .512 -0.44 0.88 Brand attitude (DV 1) Constant 7.04 0.29 24.20 .000 6.46 7.61 APK -0.35 0.05 -7.32 .000 -0.44 -0.25 Disclosure -0.33 0.34 -0.97 .335 -1.01 0.35 Product type -0.45 0.15 -2.95 .004 -0.76 -0.15 Disclosure x product type 0.11 0.22 0.49 .627 -0.33 0.54 Purchase intention (DV 2) Constant 6.33 0.39 16.06 .000 5.55 7.10 APK -0.51 0.06 -7.90 .000 -0.64 -0.38 Disclosure -0.17 0.47 -0.36 .718 -1.09 0.75 Product type -0.34 0.21 -1.61 .109 -0.75 0.75 Disclosure x product type -0.07 0.30 -0.24 .809 -0.66 0.51 OBA (DV 3) Constant 4.17 0.34 12.41 .000 3.51 4.83 APK -0.42 0.05 -7.67 .000 -0.53 -0.31 Disclosure 0.06 0.40 0.16 .876 -0.72 0.85 Product type 0.04 0.18 0.25 .806 -0.31 0.39 Disclosure x product type -0.12 0.25 -0.47 .636 -0.61 0.38
Conditional effects on brand attitude for search and experience goods
Product type Bootstrap indirect effect Bootstrap SE BOOT LLCI BOOTULCI
Search good -0.04 0.08 -0.20 0.11
Note: N = 194. Unstandardized b-coefficients (B); boot SE; CI = 95% bias-corrected bootstrap confidence interval;
number of bootstrap samples = 5000
effects do not differ between search and experience goods. Thus, H2a was not supported. A conceptual diagram of the results for brand attitude can be found in Figure 3.
Figure 3. B-coefficients of the moderated mediation analysis for brand attitude
Note: * p < .05; ns = non-significant
Purchase intention. The regression model with purchase intention as dependent
variable and disclosure and product type as independent variables was significant, F(4, 189) = 21.89, p < .001. Thus, the model can be used to predict purchase intention (R2 = .56).
Unexpectedly, there was no significant interaction between disclosure and product type (b = -0.07, p = .809), nor a direct effect by disclosure (b = -0.17, p = .718) on purchase intention. There was a relation between attitudinal PK and purchase intention (b = -0.51, p < .000),
Conditional effects on purchase intention for search and experience goods
Product type Bootstrap indirect effect Bootstrap SE BOOT LLCI BOOTULCI
Search good -0.06 0.11 -0.29 0.15
Experience good -0.17 0.13 -0.43 0.09
Conditional effects on OBA for search and experience goods
Product type Bootstrap indirect effect Bootstrap SE BOOT LLCI BOOTULCI
Search good -0.05 0.09 -0.23 0.14
indicating that the more critical and distrusting feelings people had toward the Instagram post, the lower their purchase intention. Furthermore, the index of moderated mediation was not significant (index of moderated mediation = -0.11, se = 0.17, CI -0.43; 0.24). This implies that there is no moderated mediation effect, meaning that the presence of a disclosure does not negatively impact purchase intention via attitudinal PK, and that these indirect effects do not differ between search and experience goods. Thus, H2b was not supported. A conceptual diagram of the results for purchase intention can be found in Figure 4.
Figure 4. B-coefficients of the moderated mediation analysis for purchase intention
Note: * p < .05; ns = non-significant
Online behavioral intentions. The regression model with online behavioral intentions as dependent variable and disclosure and product type as independent variables was
significant, F(4, 189) = 16.33, p < .001. Therefore, the model can be used to predict online behavioral intentions (R2 = .51). Unexpectedly, there was no significant interaction between
disclosure and product type (b = -0.12, p = .636), nor a direct effect by disclosure (b = 0.06, p = .876) on online behavioral intentions. There was a relation between attitudinal PK and online behavioral intentions (b = -0.42, p < .001), indicating that the more critical and distrusting feelings people had toward the Instagram post, the lower their online behavioral intentions. In addition, the index of moderated mediation was not significant (index of
moderated mediation = -0.09, se = 0.14, CI -0.37; 0.18). This implies that there is no moderated mediation effect, meaning that the presence of a disclosure does not negatively impact online behavioral intentions via attitudinal PK, and that these indirect effects do not differ between search and experience goods. Therefore, H2c was not supported. A conceptual diagram of the results for online behavioral intentions can be found in Figure 5.
Figure 5. B-coefficients of the moderated mediation analysis for online behavioral intentions
Note: * p < .05; ns = non-significant
Exploratory analyses
Exploratory analyses were conducted to test whether the interaction effects between disclosure and product type were mediated by perceived intrusiveness. Furthermore, it was examined whether the unexpected main effects found for product type could be explained by attitudinal PK. The results and a discussion on the findings can be found in Appendix 4.
Discussion
This study examined if the effects of sponsorship disclosures by influencers on Instagram differ between search versus experience goods, and whether these effects can be explained by an activation of attitudinal persuasion knowledge. Disclosures were expected to
increase attitudinal persuasion knowledge and consequently result in more negative
persuasion outcomes for Instagram posts including experience goods, whereas these effects were not expected for search goods.
An online experiment was conducted in order to test the hypotheses. The results showed that there was no interaction effect between disclosure and product type on brand attitude, purchase intention, and online behavioral intentions, neither were these effects mediated by attitudinal persuasion knowledge. Therefore, both hypotheses are rejected. In contrast to the expectations, there appeared to be a main effect of both disclosure and product type. The findings indicated that the presence of a disclosure led to a more negative brand attitude and purchase intention compared to no disclosure, and similar effects were found for an experience versus a search good. These main effects did not occur for online behavioral intentions. The results lead to three main conclusions.
First, this study was the first to compare the effects of sponsorship disclosures between Instagram posts including a search versus experience good. This is a relevant distinction in this context because it was found that previous studies on disclosures selected different products that might have accounted for the conflicting findings. However, the current study shows that the type of product that is promoted by the influencer does not moderate the effects of the disclosure and neither the mediation effects of attitudinal persuasion knowledge. Thus, when distinguishing between search and experience goods, it seems that product type does not impact the effectiveness of disclosures. For future research it may be interesting to compare the effects of hedonic (i.e., purchased for luxury proposes) and utilitarian goods (i.e., purchased for practical uses) since luxurious products may lead to higher feelings of envy when an influencer discloses that the post is sponsored.
Second, the findings show that people who saw an Instagram post including a
to people who saw an Instagram post without a disclosure. Although this study did not
hypothesize a main effect of disclosure, these negative findings are in line with most previous studies on sponsorship disclosures (e.g., De Veirman & Hudders, 2019; Evans et al. 2017). This implicates that disclosures have a negative impact on brand attitude and purchase intention, irrespective of the type of product that is promoted by the influencer. In contrast to prior research (Boerman et al., 2017; Evans et al., 2017), people were not less inclined to share, like, or comment on the Instagram post when a disclosure was provided. It should be noted, however, that online behavioral intentions were generally very low in this study, which was also found in previous studies (Boerman, 2020; Boerman et al., 2017). This may indicate a floor effect: Most participants scored low on online behavioral intentions, regardless of whether a disclosure was present, thus there was very little variation possible between people who saw a disclosure and those who not.
Third, contrary to the persuasion knowledge model, this study shows that a disclosure did not activate attitudinal persuasion knowledge to a larger extent than when no disclosure was presented. This would mean that no actual ‘change-of-meaning’ occurred (Friestad and Wright, 1994): People did not develop more critical or distrusting feelings toward the
message including a disclosure, even though they recognized the persuasive intent. This is in line with the idea by Ham et al. (2015) that disclosures trigger conceptual persuasion
knowledge, but this may not always activate attitudinal persuasion knowledge. However, it was expected that this has to do with the type of product in the post, yet this was not the case either. Unexpectedly, disclosures did have a negative main effect on brand attitude and
purchase intention. For the reason that attitudinal persuasion knowledge did not mediate these effects, future research might consider to measure the evaluative dimension of persuasion knowledge in a different way than done in this study. For instance, the Persuasion Knowledge
Scale (Boerman, van Reijmersdal, Rozendaal & Dima, 2018) suggests to also take into account the appropriateness and liking of sponsored content.
Limitations and directions for future research
This study provides important insights into the effects of sponsored Instagram posts, but it does have some limitations. To begin with, this study used an influencer whom most participants did not know or follow on Instagram. This was done on purpose, to control for the possibility that people would already have an existing attitude toward the influencer that might have been hard to change. However, in real life, consumers choose which influencers they want to follow and with whom they develop a parasocial interaction. A parasocial interaction is the illusion of having an intimate and close relationship with a media
personality, but developing such a relationship takes time (Tsai & Men, 2013). Therefore, it is likely that people respond more positively to a message coming from someone whom they follow, than from a person whom they do not know (Colliander & Dahlén, 2011). Thus, even though a pretest was conducted to select an influencer toward whom people had the most positive attitude, it might still be that consumers were cautious with following this person’s recommendation because they did not perceive a parasocial interaction with this influencer yet.
Since purchasing an experience good generally involves more risk, it was expected that a disclosure would trigger people to verify the credibility and motivations of the influencer especially when she is recommending this product type. But instead of an
interaction effect, negative main effects were found for Instagram posts promoting experience versus search goods. These effects may be less pronounced when it concerns a
recommendation of a person who is self-chosen and therefore more likely to be trusted (Colliander & Dahlén, 2011). Thus, using an influencer whom most people did not know
might explain the negative main effects found for experience goods. Additional analyses indeed demonstrated that these negative effects were mediated by attitudinal persuasion knowledge. To establish the actual effects of Instagram posts including experience compared to search goods, future research should consider selecting an influencer with whom a selected sample of participants has already developed a parasocial interaction.
Moreover, it should be noted that the perceptions of the Instagram posts as advertising were high in both the disclosure and no disclosure conditions. Based on the findings of the second pretest, two Instagram pictures were selected that scored relatively low on perceived persuasive intent. However, the stimulus materials included an Instagram format and a description underneath the post, whereas in the pretest participants were only exposed to the pictures. For the reason that the pictures itself did not include a specific brand, a fictitious brand was mentioned in the description by tagging this brand’s name. Possibly, this tag in combination with praise for the brand might have inadvertently caused that the persuasive intent of both Instagram posts was too overt. Future research could prevent this by tagging the brand in the picture itself on the relevant product, like a recent study by Boerman (2020) also did, instead of mentioning the brand explicitly in the description. This would even make it unnecessary to add a description, which is likely to make the persuasive intent less obvious and effects of disclosures more pronounced.
However, the high levels of perceived advertising might also have to do with the fact that 93.8% of the participants completed higher education since persuasion knowledge is assumed to increase with experience and education (Friestad & Wright, 1994). Thus, to truly understand the effects of influencer marketing on Instagram, future research could consider recruiting a more diverse group of participants, especially in terms of education level.
Implications for theory and practice
The study’s findings have implications for both theory and practice. Firstly, the results contribute to the research area by confirming the idea that sponsorship disclosures on
Instagram lead to negative persuasion outcomes, as found in most previous studies (e.g., De Veirman & Hudders, 2019; Evans et al. 2017). In contrast to the expectations, and findings in prior research (Boerman et al., 2017; Boerman et al., 2012), these effects were not mediated by attitudinal persuasion knowledge. While the majority of studies draw on the persuasion knowledge model (Friestad & Wright, 1994) to explain the impact of disclosures, the findings of the current study implicate that other underlying mechanisms may explain the negative effects. Although perceived intrusiveness was included in an exploratory analysis, it was found to not mediate the effects as well. Future research should further investigate whether the influence of disclosures on persuasion outcomes may be explained by other measurements of reactance, such as cognitive (i.e., counterarguing) and affective resistance strategies (i.e., negative affect; Van Reijmersdal et al., 2016).
Moreover, this study responded to the request made by many researchers to investigate the effects of disclosures among different product types (e.g., De Veirman & Hudders, 2019; Boerman, 2020). However, the results did not differ between search versus experience goods. Thus, based on the insights of this study the disparate outcomes of disclosures in earlier research cannot be reconciled. Furthermore, the unexpected finding that the search good evoked more positive responses than the experience good conflicts with the common idea that consumers rely more on the recommendations of others for experience goods (Nelson, 1970). This is not in line with prior studies that demonstrated that eWOM had a greater effect on purchase behavior for experience than for search goods (Park & Lee, 2009; Huang et al., 2009). This underlines the importance of having a confidential relationship with the
influencer when he or she is recommending a product (Colliander & Dahlén, 2011), which was probably lacking in this study.
This leads to the implications for advertising practice. First, this study confirms the idea that disclosing a sponsorship by ‘#ad’, as suggested by the Dutch Stiching Reclame Code, may serve as a cue that raises conceptual persuasion knowledge: People perceived the Instagram post more as an advertisement when it included a disclosure compared to no disclosure. Thus, sponsorship disclosures can be considered as an effective tool that helps consumers recognize the commercial relationship between influencers and brands.
For brands and influencers, the insights of this study imply that disclosing a sponsorship does not lead to fundamental differences for experience versus search goods. Although it seemed that overall a search good evoked more positive responses than an experience good, this might have been because participants did not have a relationship with the influencer used in this study. For brands that sell experience goods, this highlights the importance of selecting an influencer whom consumers trust.
Conclusion
This study was the first to compare the effects of product types for disclosures, thus providing important insights for the research area of sponsorship disclosures. The findings implicate that the effects of disclosures do not differ between search versus experience goods, and that these effects are not mediated by attitudinal persuasion knowledge. Although two extensive pretests were conducted in order to find appropriate stimulus materials for the experiment, future research is needed to investigate if the findings are generalizable to other search and experience goods.
References
Batinic, B., & M. Appel. (2013). Mass communication, social influence, and consumer behavior: two field experiments. Journal of Applied Social Psychology, 43(7), 1358-1368.
Becker-Olsen, K.L. (2003). And now, a word from our sponsor–a look at the effects of sponsored content and banner advertising. Journal of Advertising, 32(2), 17–32. Boerman, S. C. (2020). The effects of the standardized instagram disclosure for micro- and
meso-influencers. Computers in Human Behavior, 103, 199-207.
Boerman, S. C., van Reijmersdal, E. A., Rozendaal, E., & Dima, A. L. (2018). Development of the Persuasion Knowledge Scales of Sponsored Content (PKS-SC). International Journal of Advertising, 37(5), 671-697.
Boerman, S. C., Willemsen, L. M., & Van der Aa, E. (2017). “This post is sponsored” Effects of sponsorship disclosure on persuasion knowledge and electronic word of mouth in the context of Facebook. Journal of Interactive Marketing, 38, 82-92.
Brehm, S.S., & Brehm, J.W. (1981). Psychological reactance: A theory of freedom and control. New York: Academic Press.
Bruner, G. & Kumar, A. (2000). “Web Commercials and Advertising Hierarchy-of-Effects,” Journal of Advertising Research, 40 (1/2), 35- 42.
Colliander, J. & Dahlén, M. (2011). Following the Fashionable Friend: The Power of Social Media. Journal of Advertising Research 51(1), 313–20.
De Veirman, M., & Hudders, L. (2019). Disclosing sponsored Instagram posts: The role of material connection with the brand and message-sidedness when disclosing covert advertising. International Journal of Advertising, 1–37.
Dholakia, U. (1997). An investigation of the relationship between perceived risk and product involvement. Advances in Consumer Research, 24, 159–167.
Edwards, S.M., Li, H. & Lee, J.H. (2002). Forced exposure and psychological reactance: Antecedents and consequences of the perceived intrusiveness of pop-up ads. Journal of Advertising, 31(3), 83–95.
Evans, N. J., Phua, J., Lim, J., & Jun, H. (2017). Disclosing Instagram influencer advertising: The effects of disclosure language on advertising recognition, attitudes, and behavioral intent. Journal of Interactive Advertising, 17(2), 138–149.
Federal Trade Commission (2019). FTC Releases Advertising Disclosures Guidance for Online Influencers. Retrieved from
https://www.ftc.gov/news-events/press-releases/2019/11/ftc-releases-advertising-disclosures-guidance-online-influencers Federal Trade Commission (2017). CSGO lotto owners settle FTC's first-ever complaint
against individual social media influencers. Retrieved from https://www.ftc.gov/news- events/press-releases/2017/09/csgo-lotto-owners-settle-ftcs-first-ever-complaint-against
Friestad, M., & Wright, P. (1994). The persuasion knowledge model: How people cope with persuasion attempts. Journal of Consumer Research, 21(1), 1–31.
Guseman, D.S. (1981). Risk perception and risk reduction in consumer services. In Donnelly, J.H. and George, W.R. (Eds), Marketing of Services, American Marketing
Association, Chicago, IL.
Ham, C. D., Nelson, M. R., & Das, S. (2015). How to measure persuasion knowledge. International Journal of Advertising, 34(1), 17-53.
Hayes, A. F. (2018). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. Guilford Publications.
Hsieh, Y.C., Chiu, H.C. & Chiang, M.Y. (2005). Maintaining a committed online customer: A study across search-experience-credence products. Journal of Retailing, 81(1), 75-82.
Huang, P., Lurie, N. L. & Mitra, S. (2009). Searching for Experience on the Web: An Empirical Examination of Consumer Behavior for Search and Experience Goods. Journal of Marketing, 73(2), 55-69.
Janssen, L., van Sprang, B. & Fransen, M. (2017). Het effect van sponsorvermeldingen op merkevaluaties en geloofwaardigheid van bloggers: De rol van tweezijdige
boodschappen. Tijdschrift voor Communicatiewetenschap, 45(1), 19–36. Johnson, B. K., Potocki, B., & Veldhuis, J. (2019). Is that my friend or an advert? The
effectiveness of Instagram native advertisements posing as social posts. Journal of Computer-Mediated Communication, 24(3), 108–125.
Mediakix (2019). Influencer marketing 2019 industry benchmarks. Retrieved from https://mediakix.com/influencer-marketing-resources/influencer-marketing Mitra, K., Reiss, M. C. & Capella, L. M. (1999). An examination of perceived risk,
information search and behavioral intentions in search, experience and credence services. The Journal of Services Marketing, 13(3), 208–228.
Miron, A.M., & Brehm, J.W. (2006). Reactance theory - 40 years later. Zeitschrift Für Sozialpsychologie 37(1), 9–18.
Murray, K.B. & Schlater, J.L. (1990). The impact of services versus goods on consumers’ assessment of perceived risk. Journal of the Academy of Marketing Science, 18(1), 51-65.
Nelson, P. (1970). Information and consumer behavior. Journal of political economy, 78(2), 311-329.
Nelson, P. (1974), Advertising as information. Journal of Political Economy, 82(4), 729-54. Nelson, M. R. & Ham, C. D. (2012). The reflexive game: How target and agent persuasion
knowledge influence advertising persuasion. In E., Thorson, & S. Rodgers, S. (Eds.), Advertising theory, 174-187. New York: Routledge.
Ohanian, R. 1990. Construction and validation of a scale to measure celebrity endorsers perceived expertise, trustworthiness, and attractiveness. Journal of Advertising, 19(3), 3952.
Park, C., & Lee, T. (2009). Information direction, website reputation and eWOM effect: A moderating role of product type. Journal of Business Research, 62, 61–67.
Petty, R.E., Cacioppo, J.T. & Goldman, R. (1981). Personal involvement as a determinant of argument-based persuasion. Journal of Personality and Social Psychology, 41(5), 847–55.
Petty, R. E., & Cacioppo, J. T. (1984). The effects of involvement on responses to argument quantity and quality: Central and peripheral routes to persuasion. Journal of
Personality and Social Psychology, 46, 69–81
Petty, R. E., & Cacioppo, J. T. (1986). The elaboration likelihood model of persuasion. Springer.
Rozendaal, E., Lapierre, M. A., Van Reijmersdal, E.A. & Buijzen, M. (2011). Reconsidering Advertising Literacy as a Defense Against Advertising Effects. Media Psychology, 14(4), 333–54
Sah, S., Malaviya, P. & Thompson, D. (2018). Conflict of interest disclosure as an expertise cue: Differential effects due to automatic versus deliberative processing.
Organizational Behavior and Human Decision Processes, 147, 127-146. Senecal, S., & Nantel, J. (2004). The influence of online product recommendations on
consumers’ online choices. Journal of Retailing, 80, 159–169.
Spears, N., & Singh, S. N. (2004). Measuring attitude toward the brand and purchase intentions. Journal of Current Issues and Research in Advertising, 26(2), 53-66.
Statista (2020). Distribution of Instagram users worldwide as of April 2020, by age and gender. Retrieved from https://www.statista.com/statistics/248769/age-distribution-of-worldwide-instagram-users/
Stichting Reclame Code (2019). Campagne #Ad roept influencers op open te zijn over reclame. Retrieved from https://www.reclamecode.nl/news/campagne-ad-roept-influencers-op-open-te-zijn-over-reclame/
Tsai, W. S., & Men, L. R. (2013). Motivations and antecedents of consumer engagement with brand pages on social networking sites. Journal of Interactive Advertising, 13(2), 76– 87.
Van Reijmersdal, E. A., Fransen, M. L., Van Noort, G., Opree, S. J., Vandeberg, L., Reusch, S., Van Lieshout, F., & Boerman, S.C. (2016). Effects of disclosing sponsored content in blogs: How the use of resistance strategies mediates effects on persuasion.
American Behavioral Scientist, 60(12), 1458-1474.
Verlegh, P.W.J., Ryu, G., Tuk, M.A. & Feick, L. (2013). Receiver responses to rewarded referrals: the motive inferences framework. Journal of the Academy of Marketing Science, 41(6), 669-682.
Wojdynski, B. W. (2016). Native advertising: Engagement, deception, and implications for theory. The new advertising: Branding, Content and consumer Relationships in a data- driven. Social Media Era, 1, 203–236.
Wojdynski, B. W., & Evans, N. J. (2016). Going native: Effects of disclosure position and language on the recognition and evaluation of online native advertising. Journal of Advertising, 45(2), 157–168.
Wojdynski, B.W., Evans, N.J. & Hoy, M.G. (2020). The Covert Advertising Recognition and Effects (CARE) model: Processes of persuasion in native advertising and other
