I spy with my little eye a disclosure on the blog : a study on how disclosures in blogs affect brand responses and the mediating role of persuasion knowledge

(1)

I spy with my little eye a disclosure on the blog

A study on how disclosures in blogs affect brand responses and the mediating role of persuasion knowledge

Name: Ann Mortensen Student ID: 10484818 Master’s Thesis Graduate School of Communication Persuasive Communication Supervisor: Guda van Noort Date of completion: June 27th

(2)

Abstract

Blog advertising is becoming increasingly popular amongst both consumers and advertisers. When a blog is being used for advertising it is required by law to include a disclosure (Federal Trade Commission, 2013). Despite the societal relevance disclosures in blogs have so far only been studied in a very limited extent and the findings have been inconclusive. Building upon Friestad and Wright’s persuasion knowledge model (1994) this study aims to answer how a disclosure (vs. no disclosure) mediated through persuasion knowledge affects affective responses negatively and cognitive responses positively. The study also examines how the timing of the disclosure in the blog post (before vs. after) mediated through persuasion knowledge affects affective and cognitive responses positively. A content analysis was

conducted in the pilot study in order to examine when and how disclosures were used in blogs in order to create the material for the main study. The material was tested in a pretest prior to the main study. An online between-subject experiment with three conditions (disclosure before the blog post vs. disclosure after the blog post vs. no disclosure) was conducted on a sample of 108 participants. The hypotheses were tested using Hayes’ process (2008) with a sequential mediation. None of the hypotheses were supported, but the outcome still holds several theoretical and practical implications as well as implications for future research, which are discussed.

(3)

Online marketing and advertising campaigns using social media have been a hot topic for scholars and advertisers alike in the last years, but there are still areas to explore. One of these is blogs and blog advertising. The 2013 Digital Influence Report (Higgins, 2013) show that even though blogs are perceived as a being highly trustworthy and popular among consumers and even though they are one of the most influential online media to affect consumer’s buying decisions (Chiang & Hsieh, 2011; Yang 2011) only 11% of most brand’s social media budget in 2013 was spend on blogs (Higgins, 2013). In order for scholars and marketers to realize the full potential of blogs and blog advertising more research is needed to know how blog

advertising affects consumer responses, especially when they include a disclosure, which they are required to by law when an advertisement is present (Federal Trade Commission, 2013).

A blog is a frequently updated website, where content is posted in reverse order on a regular basis (Schmidt, 2007). This thesis will focus on personal blogs and blog advertising, i.e. when the blogger is being paid to mention or review a product (Zhu & Tan, 2007). When a blog is being used for advertisement it needs to include a disclosure. In advertising a disclosure refers to the revelation or declaration of a persuasive intent of a message (Boerman, Van

Reijmersdal, & Neijens, 2014; Campbell, Mohr & Verlegh, 2013).

Prior research on disclosures in other media, such as TV, print and word-of-mouth marketing (Boerman, Van Reijmersdal & Neijens, 2014; Abendroth & Heyman, 2013; Van Reijmersdal, Tutaj & Boerman, 2013; Van Reijmersdal, Neijens & Smit, 2005) demonstrated that the affective and cognitive response to the disclosure differ depending on the medium it is placed in, therefore it is of highly theoretical importance to examine how the response to disclosures in blogs compare to those of other media. So far, only a few studies have examined the effects of disclosures in blogs (Campbell, Mohr & Verlegh, 2013; Nekmat & Gower, 2012; Zhu & Tan, 2007) and the findings of these studies have been inconclusive, therefore it is of highly theoretical relevance to further examine this topic in order to reach an understanding of what the effects of disclosures in blogs are. This study will thus contribute to the existing literature on disclosures by examining disclosures within the context of blogs.

(6)

The use of blog advertising is increasing (Higgins, 2013) and regulations from the Federal Trade Commission (Federal Trade Commission, 2013) state that online advertising, such as blog advertising, needs to be accompanied by a disclosure. It is therefore of highly practical relevance to study the timing of the disclosure: Marketers need to consider, where to place the disclosure strategically in order to gain the greatest persuasive effect. The aim of this study is to examine how the presence of a disclosure and the timing of it affect affective and cognitive responses in blog advertising.

When the advertisement is embedded in the content of the blog post it might be difficult for the consumers to recognize the persuasive intent. This is beneficial for certain advertising strategies, such as covert marketing (Campbell, Mohr & Verlegh, 2013), but it might have a negative effect on the consumers’ responses if they become aware that a message contains a persuasive intent (Pelsmacker & Neijens, 2012), which happens when the consumers’ persuasion knowledge is activated. When a disclosure is present the consumer’s persuasion knowledge is activated (Boerman, Van Reijmersdal & Neijens, 2014), which in turn affects the affective response negatively and cognitive responses positively (Friestad & Wright, 1994). When the disclosure is placed before the blog post, it will give the consumer more time to process the information, which has a positive effect on cognitive responses and it will also warn the consumer about the persuasive intent in advance, which has a positive effect on affective responses (Friestad & Wright, 1994). Based on Friestad and Wright’s persuasion knowledge model (1994) this study will examine how persuasion knowledge mediates the effect of disclosures and timing on affective and cognitive responses.

This study aims to answer the following research question:

To what extent does a disclosure compared to no disclosure affect affective responses positively and cognitive responses negatively and how is this mediated by persuasion knowledge and to what extent does a disclosure placed before the blog post positively affect affective and cognitive responses compared to a disclosure placed after the blog post and how is this mediated by persuasion knowledge?

(7)

Theoretical Framework

In this section the characteristics of blogs and blog advertising will briefly be defined

followed by an introduction to disclosures and the persuasion knowledge model. Finally, the effects of disclosures and the timing of disclosures on affective and cognitive responses will be described and the hypotheses will be presented.

Blogs

Blogs as a medium have their roots in the 1990’s, where they saw the light as web journals, which later changed to web logs, weblogs, wee blogs and finally blogs (Tremayne, 2012). The platform is still changing as well as increasing in popularity: Recent numbers show that there were more than 152 million blogs on the Internet by the end of 2013 and that a new blog was created somewhere in the world every half second (Gaille, 2013). On top of this

marketers are realizing the potential of using blogs for advertising (Chiang & Hsieh, 2011).

Blog characteristics and blog types.

Schmidt (2007) developed a framework to analyze and characterize blogs. He defines a blog as a frequently updated website, where content is posted on a regular basis in reverse order and where the readers have the opportunity to comment on the posts. It typically includes a blog roll in the sidebar with links to the bloggers’ own favorite blogs. Through the blog roll and the comment-section the bloggers and their readers form clusters and networks referred to as the blogosphere (Schmidt, 2007). There are no or minimum costs in producing a blog, it is visible to everyone and it is available indefinitely (Hennig-Thurau, Malthouse, Friege, Gensler, Lobschat, Rangaswamy & Skiera, 2010).

There are different types of blogs: personal blogs, corporate blogs and expert blogs in which journalists, academics or other specialists discuss a topic of their own choice (Schmidt, 2007). This study will examine a personal blog, where one or several blog posts contains blog

advertising. Compared to corporate blogs and expert blogs, the personal blog features user-generated content from a low-expertise source (Padmanabhan, 2008). What makes it

interesting to examine personal blogs is the use of personal blogs for advertising is increasing (Higgins, 2013) and the effects are unknown. So far a very limited amount of studies have

(8)

studied this phenomenon and the findings have been inconclusive, so in order to understand blogs and the use of them in advertising further research is needed.

Blogs as advertising.

Personal blogs can be used for blog advertising, that is “the paid sponsorship of bloggers to mention, review, promote, or sell products in blog writing” (as cited in Wang, Ghose, & Ipeirotis, 2012, p. 2; Zhu & Tan, 2007). The use of blog advertising and sponsored content in blogs is increasing, because blogs have a large reach, they are perceived as credible and the cost of the advertisement is low compared to other media (Higgins, 2013; Fu & Chen, 2012). Blog advertising takes the same form and qualities as the blogger’s original content and it usually serves as useful or entertaining information that describes the sponsored product in a favorably way (Sonderman & Tran, 2013). Marketers use blog advertising and sponsored content as a way to distinguish their advertisements from the advertising clutter (Boerman, Van Reijmersdal & Neijens, 2012) and to create product awareness and enhance brand images (Becker-Olsen, 2003), but new regulations in the US and Europe (Federal Trade Commission, 2013; International Chamber of Commerce, 2004) now state that online

advertising such as blog advertising need to be accompanied by a disclosure might jeopardize this opportunity and the disclosure might negatively affect the persuasion outcomes (Cain, 2011). Previous research on the effects of blog advertising have focused on word-of-mouth factors, reader comments and reader involvement (Cao, Knotts, Xu, & Chau, 2009; Fu & Chen, 2012; Zhu & Tan, 2007), but there is a lack of empirical studies that examine how disclosures affect blog advertising.

Disclosures

A disclosure literally means “the act of making something known” (Merriam-Webster’s online dictionary, n.d.). In advertising a disclosure refers to the revelation or declaration of a persuasive intent of a message (Boerman, Van Reijmersdal, & Neijens, 2014; Campbell, Mohr & Verlegh, 2013), but despite the regulations from the Federal Trade Commission and other agencies there are still no clear definition regarding the form, content and use of disclosures.

(9)

Form and content.

Disclosures in blogs take the form of a label or a small piece of text placed before or after the content in a blog post (Campbell, Mohr & Verlegh, 2013; Zhu & Tan, 2007). “This blog post is sponsored by product/brand/company” is often used as a template, e.g. “This blog post is sponsored by Land O’Lakes” (Drummond, 2014). The persuasion knowledge model, which will be introduced more thoroughly below, suggests that when a specific product is mentioned in the disclosure then the consumer’s topic knowledge is activated and when a disclosure mentions the brand then the consumer’s agent knowledge is activated (Friestad & Wright, 1994). The more information in the disclosure the easier it might become for the consumer to cope with the persuasive attempt.

The appearance and the content of disclosures have changed over time (Grubbs Hoy & Andrews, 2004) and to this day there is still no clear definition of what a disclosure should look like and what it should communicate. Cain (2011) argues that a problem regarding disclosures is that the text is often too small to read and disclosures in TV-shows specifically only appear in a very limited time span or and that. This problem could be solved by adding more information to the disclosure or by distinguishing it more clearly from the content, but solving the problem might not be in the interest of the people who benefit from the embedded advertising and product placement (Cain, 2011). Different groups lobby for and against clearer regulations regarding disclosures and recent as well as older studies (Edmonds, 2013; Brock, 2009) indicate that the debate is ongoing and that it is still not clear when a disclosure should be present or what it should include in online as well as more traditional platforms.

Cain (2011) suggests that a disclosure can serve an informing purpose, but it can also serve a warning purpose. The distinction between a disclosure and a warning seems to be that a warning is used when the usage of the advertised product can lead to health problems (Stewart & Martin, 2004; Cox III, Wogalter, Stokes & Murff, 1997), but according to Cain (2011) a disclosure have the ability to warn people about a persuasive attempt, but in order for it to do so the form and content of it might need to change. To sum up, there is no clear

definition of how or when a disclosure should appear and what it should communicate, therefore this study will examine the appearance and content of disclosures in blog

(10)

The persuasion knowledge model.

The persuasion knowledge model can be used to explain the underlying processes of the effectiveness of disclosures. Friestad and Wright’s persuasion knowledge model (1994) suggests that people’s persuasion knowledge develops continually over time and it influences their response to persuasive attempts: “Consumers' persuasion coping knowledge enables them to recognize, analyze, interpret, evaluate, and remember persuasion attempts and to select and execute coping tactics believed to be effective and appropriate” (Friestad & Wright, 1994, p. 3). Therefore, when a persuasive message is salient the chance of the consumer’s persuasion knowledge being activated is increased.

Persuasion knowledge has two dimensions: The first one is the cognitive dimension in which advertising and its source is recognized as well as understood by the receiver (Boerman, Van Reijmersdal & Neijens, 2012; Rozendaal, Lapierre, Van Reijmersdal & Buijzen, 2011). This dimension is referred to as conceptual persuasion knowledge. The conceptual persuasion knowledge activates the attitudinal persuasion knowledge, the second dimension. In the affective dimension, known as attitudinal persuasion knowledge, critical attitudes, such as skepticism and disliking, are activated and used to evaluate the advertisement. Attitudinal persuasion knowledge involves critical feelings about honesty, trustworthiness and credibility about the advertisement (Van Reijmersdal, Tutaj & Boerman, 2013; Boerman, Van

Reijmersdal & Neijens, 2012; Rozendaal, Lapierre, Van Reijmersdal & Buijzen, 2011). In the case of blog advertising, conceptual persuasion knowledge is recognizing the fact that a blog post has a persuasive intent and thereby differs from other blog posts about personal topics. Attitudinal persuasion knowledge is the critical attitudes about the persuasive intent in the blog post, which consequently influences persuasion outcomes.

Persuasion knowledge has been operationalized differently in prior research, but most studies have measured both the recognition of the advertisement (i.e. conceptual persuasion

knowledge) and critical attitudes or skepticism towards it (i.e. attitudinal persuasion

knowledge) (Boerman, Van Reijmersdal & Neijens, 2014; Tutaj & Van Reijmersdal, 2012; Boerman, Van Reijmersdal & Neijens, 2012; Rozendaal, Lapierre, Van Reijmersdal & Buijzen, 2011). Based on previous empirical tests (Boerman, Van Reijmersdal & Neijens, 2014; Boerman, Van Reijmersdal & Neijens, 2012; Rozendaal, Lapierre, Van Reijmersdal & Buijzen, 2011) it is assumed in this study that conceptual persuasion knowledge automatically

(11)

activates attitudinal persuasion knowledge, therefore this will not be further tested. This study will measure and operationalize both conceptual and attitudinal persuasion knowledge, but the two dimensions will be discussed as a single construct, which is persuasion knowledge.

When a disclosure is present then the consumer’s persuasion knowledge is activated, which should lead to the advertisement being recognized as well as processed more critically (Boerman, Van Reijmersdal & Neijens, 2014). Critical processing, or skepticism, relates to the credibility of the sender (Cain, 2011) and being that blogs and bloggers are generally perceived to be credible (Johnson & Kaye, 2009; Mack, Blose, & Pan, 2008) it is interesting to examine, how the disclosures affect persuasion knowledge and in turn brand responses. Blog advertising and sponsored content are being used by marketers as a way to distinguish their product from the advertising clutter (Boerman, Van Reijmersdal & Neijens, 2012), but if a consumer’s persuasion knowledge is activated they recognize the advertising, which might nullify the tactics used in covert marketing. It is unknown how persuasion knowledge affects the effects of the disclosure on brand responses, but this study aim to examine this process.

Affective and cognitive responses

It is important to examine the affective and cognitive responses to disclosures in blogs in order to compare it to the responses in other media, such as TV-shows (Boerman, Van Reijmersdal & Neijens, 2014; Campbell, Mohr & Verlegh, 2013; Boerman, Reijermsdal & Neijens, 2012). It is also important to examine the affective and cognitive responses to brands from a marketer perspective, because the use of blog advertising is increasing (Higgins, 2013) and advertisers need to know what it implies for their brand or product to have it promoted on a blog, especially when a disclosure is present.

The effects of disclosures on affective responses.

Previous studies on disclosures in other media than blogs show contradicting results in the affective responses, such as brand attitude. First, disclosures in TV-programs have a negative effect on brand attitude (Boerman, Van Reijmersdal & Neijens, 2014; Campbell, Mohr & Verlegh, 2013; Van Reijmersdal, Tutaj, & Boerman, 2013), which is enhanced by the length of the disclosure (Boerman, Van Reijmersdal & Neijens, 2012). Contrary to this, disclosures in word-of-mouth marketing lead to positive reactions such as higher source credibility, fewer

(12)

negative feelings about the sponsorship and sharing of the content (Carl, 2008) as well as higher purchase intent and a greater attitude towards the product (Abendroth & Heyman, 2013). The affective consumer reactions to disclosures in advertorials placed in a print magazine are positive (Van Reijmersdal, Neijens & Smit, 2005), but it is argued that this might be because the consumers fail to notice the label, i.e. the disclosure, in the advertorials. When the consumers do not realize that there is an advertisement their attitude is more positive than when they recognize the disclosure and the persuasive intent (Kim, Pasadeos & Barban, 2001).

The above-mentioned findings indicate that there are different reactions to disclosures across media, therefore the responses to disclosures in blogs should be looked into. Zhu and Tan (2007) found in a previous study on blog advertising that disclosures in blog post are more effective in increasing behavioral intent, e.g. purchase intention and intention to recommend a product, when the sponsored product is considered low in involvement and the blogger is considered as low-expertise, whereas the effects are negative, when the product is considered as being high in involvement (Zhu & Tan, 2007). Two other studies have examined the effects of disclosures in blogs on affective responses and purchase intention: Nekmat and Gower (2012) found that the perceived credibility of the blogger, the attitude towards the product and purchase intention was higher when there was no disclosure present compared to when a disclosure was present. Contrary to this then Campbell, Mohr and Verlegh (2013) found no significant effects on brand attitude when comparing a blog post with a disclosure to one without, which is not in line with previous disclosure-findings. Being that the findings are inconclusive it is argued that here is a call for further research on the effects on disclosures in blogs.

Although prior studies are inconclusive this study argues that based on the persuasion knowledge model disclosures negatively influences attitude, because the higher level of processing also leads to a more critical processing of the information in agreement with the persuasion knowledge model (Boerman, Van Reijmersdal & Neijens, 2012; Friestad & Wright, 1994), which lead us to the first hypothesis:

Hypothesis 1: The effect of disclosures on affective responses is mediated by persuasion knowledge, such that a disclosure (vs. no disclosure) in blogs positively

(13)

influences persuasion knowledge and persuasion knowledge in turn negatively influences affective responses.

The effects of timing on affective responses.

As it is mentioned previously then blog posts with sponsored content are required by law to enclose a disclosure (Federal Trade Commission, 2013), therefore it is important to examine how advertisers should deal with this given fact in a strategic way. The law does not prescribe where a disclosure should be placed, thus bloggers and advertisers are free to place it up front, somewhere in the middle or below the blog. As previous studies in other media suggest that timing influences persuasion outcomes it is of importance to examine how the timing of the disclosure affects brand responses.

The persuasion knowledge model can be used to explain effects of timing of disclosures. Friestad and Wright’s persuasion knowledge model (1994) states that “when a person begins conceiving of an agent’s action, heretofore not identified as having any particular meaning, as a persuasion tactic a ‘change of meaning’ will occur” (Friestad & Wright, 1994, p. 13). People change or reinterpret their behavior and attitude, when they realize in hindsight that a persuasive attempt has been made or when they realize the true nature of the agent or the source (Van Reijmersdal, Tutaj & Boerman, 2013; Taillard, 2000; Friestad & Wright, 1994). This indicates that the timing of the disclosure affects the response to the disclosure and that it is being mediated by persuasion knowledge.

Previous studies have found that disclosures placed before the blog post in a blog (Campbell, Mohr & Verlegh, 2013; Nekmat & Gower, 2012) or during a conversation in word-of-mouth marketing (Abendroth & Heyman, 2013) have a positive effect on brand attitude compared to disclosures placed afterwards. Boerman, Van Reijmersdal and Neijens (2014) found opposite results in the study on disclosures in TV-shows, but this could be explained by the attributes attached to the different media: Perceived credibility and trustworthiness are important aspect of blogs and word-of-mouth marketing, perhaps more important than for TV-shows. When a disclosure is placed before the sponsored content it indicates that the sender is being honest and has no ulterior motives (Abendroth & Heyman, 2013), whereas if the disclosure was placed after the sponsored content this could result in annoyance (Campbell, Mohr &

(14)

Verlegh, 2013) and a ‘change of meaning’, because the agent’s true colors are being revealed (Friestad & Wright, 1994), which in turn would have a negative influence attitude. Based on this the following hypothesis is formulated:

Hypothesis 2: The effect of timing on affective responses is mediated by persuasion knowledge, such that a disclosure placed before (vs. after) the blog post positively influences persuasion knowledge and persuasion knowledge in turn positively influences affective responses.

The effects of disclosures on cognitive responses.

The persuasion knowledge model can be used to explain, how disclosures positively affect cognitive responses, such as brand memory. When a disclosure is present then the processing of information is increased (Friestad & Wright, 1994), which has a positive effect on memory. The disclosure functions as a cue or a prime that stimulates the recognition or recall of the brand or product (Van Reijmersdal, Tutaj & Boerman, 2013). Previous studies found that disclosures in TV-programs have a positive effect on brand memory (Boerman, Van

Reijmersdal & Neijens, 2014; Campbell, Mohr & Verlegh, 2013; Van Reijmersdal, Tutaj, & Boerman, 2013). Campbell, Mohr and Verlegh’s (2013) examined the effects of disclosures in blogs on brand memory. They found that no disclosure had a positive effect on brand memory compared to a disclosure, which again is not in agreement with the persuasion knowledge model or previous findings, therefore it is argued that further studies are needed to confirm the effects of disclosures on cognitive responses, thus this study will examine the effects of a disclosure versus no disclosure on brand memory in blogs.

Based on the above-mentioned theory and empirical findings it is argued that disclosures have a positive effect on cognitive responses, because disclosures lead to a higher level of

processing which in turn positively influences cognitive responses (Van Reijmersdal, Tutaj & Boerman, 2013; Friestad & Wright, 1994):

Hypothesis 3: The effect of disclosures on cognitive responses is mediated by persuasion knowledge, such that a disclosure (vs. no disclosure) in blogs positively influences persuasion knowledge and persuasion knowledge in turn positively influences cognitive responses.

(15)

The effects of timing on cognitive responses.

When a disclosure is placed before or during the sponsored content it positively influences cognitive responses (Boerman, Reijersdal & Neijens, 2014; Van Reijmersdal, Tutaj & Boerman, 2013): The longer time between the disclosure and the exposure of the product or brand, the longer the consumer will process the information, which in turn has a positive effect on brand memory (Boerman, Van Reijmersdal & Neijens, 2014). The disclosure serve as a information prime for the product or the brand, thus when it is placed before the

sponsored content then the consumer have time to activate their persuasion knowledge related to the agent and the topic (Friestad & Wright, 1994), which has a positive effect on cognitive responses (Van Reijmersdal, Tutaj & Boerman, 2013). Based on this the last hypothesis is formulated:

Hypothesis 4: The effect of timing on cognitive responses is mediated by persuasion knowledge, such that a disclosure placed before (vs. after) the blog post positively influences persuasion knowledge and persuasion knowledge in turn positively influences cognitive responses.

All of the above-mentioned hypotheses are illustrated below in the conceptual model in figure 1. Disclosure: - Before - After No disclosure Conceptual persuasion knowledge Attitudinal persuasion knowledge Affective responses Cognitive responses H2 + H1 - H3 + H4 +

(16)

Pilot Study

The form and the content of disclosures in blogs vary depending on the blogger. In order to test the hypotheses on the most realistic material possible a pilot study was conducted, which examined how disclosures look and what they communicate on a sample of blogs.

The aim of the pilot study is to examine blogs and disclosures in blog posts through a content analysis in order to create the experimental stimuli for the main study. The experimental material in the main study was based on the findings of the content analysis in the pilot study, thus increasing the external validity of the study. This study will use a food blog in the

material. The food blog was chosen, because food and drinks are considered low-involvement products and it has been found that when a disclosure is present then low-involvement

products have a more positive effect on responses to the brand compared to high-involvement products (Zhu & Tan, 2007). Also, food blogs are read by both men and women and different age groups (Norén, 2011). The product was a chocolate bar, because it is assumed that

chocolate is a low-involved product and that people in general like chocolate, thus product interest will most likely not influence the results of the study.

Method

A qualitative content analysis was conducted on a sample of 50 food blogs (see appendix 1). The food blogs were randomly chosen on www.bloglovin.com, a network of blogs. The only criteria for the blogs were that they were written in English and placed under the food and drinks category. Following the procedure of Zwier (2009) the blogs were first coded for basic features such as name, link to blog, size (measured by numbers of followers on bloglovin) and the date of the most recent blog post in order to make sure that the blog was up-to-date. The blogs were coded in order to see, if any of the blog posts contained sponsored content and disclosures (0 = no, 1 = yes) and if that was the case, the timing of the disclosure (1 = before, 2 = in text, 3 = after, 4 = multiple) and the mentioning of an agent (0 = no, 1 = yes). An example of a disclosure in a blog, which mentions an agent, is: “This blog post is sponsored by Land O’Lakes” (Drummond, 2014). The sub-aim of the content analysis was to examine the frequency of disclosures and sponsored content in blogs as well as the content of the disclosures. The procedure of the content analysis was to first skim the most recent blog posts in order to see, if there were any posts that contained sponsored content. Then, the search engine on the blog was used to find the following search words: sponsored (by), sponsor,

(17)

disclosure, association, partner, giveaway and opinion. Blog categories such as “sponsor post” were also included in the search. Due to the regulations regarding disclosures on blogs

(Federal Trade Commission, 2013) it is assumed that if blog advertising is present on a blog that it is either accompanied by a disclosure or there is a mentioning of sponsorship.

Results

The variables’ descriptives were recorded and the findings showed that 72% (n = 36) of the blogs had one or several blog posts with a disclosure. The disclosure was placed before the blog post in 11.1% of the blogs (n = 4), in the text in 8.3% of the blogs (n = 3), after the blog post in 69.4% of the blogs (n = 25) and multiple times in 11.1% of the blogs (n = 4). An agent was mentioned in 99.7% of the disclosures (n = 33). Based on the findings of the content analysis the disclosure in the experimental stimuli will include the mentioning of the agent and there will only be one disclosure per blog post. The findings of the pilot study indicate that the disclosures vary in regard to timing, content and form, which emphasize the importance of examining the content and form of the disclosure as well as the difference in effects of timing.

Main Study

The aim of the main study is to test the four hypotheses and thus examine the effects of disclosures and timing on brand responses as well as the mediating effect of persuasion knowledge.

Experimental stimulus

An already existing blog posts was changed to fit the experiment and used as the experimental stimulus. The blog post was from the blog www.jeroxie.com, an Australian blog with 70 followers on bloglovin. It is expected that the blog is unknown to the participants. In the before-condition the disclosure, “This blog post is sponsored by Cadbury”, was placed before the blog post. In the after-condition the disclosure was placed after the blog post. There was no disclosure present in the no disclosure-condition, which also functioned as a control condition. The material is attached in appendix 2.

(18)

Pretest of effectiveness of experimental stimuli

Prior to the experiment a pretest was conducted to test the effectiveness of the manipulation and the authenticity of the material, which was created based on the findings of the pilot study. 34 participants (58.8% female; Mage = 26.91, SD = 2.40) received a link to an online experiment per mail. The sample of the pretest resembled the main study, but it was taken from a different group. The participants were randomly placed in one of three conditions (disclosure before blog post vs. disclosure after blog post vs. no disclosure). After reading the blog post the participants were asked to fill in a questionnaire, which measured authenticity of the material, brand familiarity, product familiarity, previous buying behavior and the

effectiveness of the manipulation. Authenticity of the material was measured on a 7-point scale using two items: unrealistic/realistic and not authentic/authentic. There was a strong, positive correlation between the two items (r = .73, p < .01). The mean score of the two items is used as a measurement of authenticity of the material (M = 5.66, SD = 1.09). Brand

familiarity was measured by asking the participants if they were familiar with Cadbury before participating in the study (0 = no, 1 = yes). If they were then product familiarity (“Were you familiar with Cadbury Mousse Double Chocolate before participating in this study) and previous buying behavior (“Have you ever bought Cadbury Mousse Double Chocolate?”) were measured in a similar manner. In order to measure if the manipulation was successful the participants were asked if they noticed a disclosure (i.e. the mentioning of sponsored content or an advertisement) in the blog post (0 = no, 1 = yes).

The results showed that the participants perceived the material as authentic, based on the high mean score of authenticity of the material (M = 5.66, SD = 1.09), thus it is argued that the material can be used in the main study. 44.1% of the participants (n = 15) were familiar with the brand Cadbury, but none of those were familiar with the product Cadbury Mousse Double Chocolate, thus it is argued that the product can also be used in the main study. A chi-square test revealed that there was a significant association between the presence of a disclosure and whether the participants noticed it or not, x2

(2, 34) = 16.61, p < .01. Cramer’s V was used to measure the strength of the association, which was strong, V = .70. Based on the findings of the pretest it is argued that the manipulation was successful and that material and the brand can be used in the main study.

(19)

Research design, participants and procedure

An online between-subjects experiment with three conditions (no disclosure vs. disclosure before vs. disclosure after) was conducted to test the hypotheses. 113 participants (65.5% female, Mage = 27.62, SD = 7.24) took part in the experiment. After excluding the participants, who were already familiar with the blog and/or the product as well as those who had bought the product prior to this experiment, the final sample consisted of 108 participants (66.7% female, Mage = 27.47, SD = 6.92). A convenience sample was used in which the participants received an online invitation to an event through Facebook.

The procedure was as follows: On the event page they were given brief instructions and asked to click a link and precede to the online experiment. The participants were first asked a few demographic questions and then they were asked to read the instructions carefully, before they were randomly assigned to one of the three conditions. After they read the blog post they had to do a small filling-task in order to enhance the time between the exposure of the

material and the measurement of effects, which enhances the external validity of the study. The filler task was followed by a questionnaire in which the participants were asked, if they noticed the disclosure (manipulation check) as well as their brand responses (brand recall and recognition and brand attitude) and their persuasion knowledge. The questionnaire also measured covariates, such as brand familiarity and blog reading frequency. Finally, the respondents were debriefed and thanked for their participation. No incentives were given. It took seven days to collect the data.

Measures

Cognitive responses.

Cognitive responses were operationalized as brand memory and measured with unaided recall and recognition in line with previous disclosure-studies (Van Reijmersdal, Tutaj & Boerman, 2013; Russell, 2002). The participants were first asked if they saw any brands in the blog post. If they did they could fill out which brands. If Cadbury was mentioned, then brand recall was coded as 1, otherwise it was coded as 0. Afterwards the participants were shown five different chocolate bars and asked, which of the brands they recognized if any. If Cadbury Mousse Double Chocolate was recognized, then brand recognition was coded as 1, otherwise it was coded as 0.

(20)

Affective responses.

Affective responses were operationalized as brand attitude and measured using a 6-item semantic differential scale based on Bruner (1998) and as used in e.g. Van Noort, Voorveld and Van Reijmersdal (2012) and Li, Daughterty and Biocca (2002). The items were measured on a 7-point scale: bad/good, unappealing/appealing, pleasant/unpleasant (reversed),

unattractive/attractive, interesting/boring (reversed), dislike/like. A factor analysis revealed that the items load on one factor (EV = 3.59; R2_{= 59.89%; Cronbach’s alpha = .86). The mean} score of the six items is used as a measurement of brand attitude (M = 4.85, SD = .95).

Conceptual and attitudinal persuasion knowledge.

The two dimensions of persuasion knowledge were measured by two measures developed by Boerman, Van Reijmersdal & Neijens (2012). Conceptual persuasion knowledge was

measured on a 7-point scale (strongly disagree/strongly agree) by asking the participants to what they extent they agree with the statement: “The item about Cadbury Mousse Double Chocolate was advertising” (M = 5.04, SD = 1.62). A single-item indicator has been used in other similar studies (Boerman, Van Reijmersdal & Neijens, 2014; Van Reijmersdal, Neijens & Smit, 2005) and it is, according to Rossiter (2011), an adequate measure for a concrete, single object. Attitudinal persuasion knowledge was measured on a 7-point scale by asking the participants if they agreed (strongly disagree/strongly agree) with the statement, “I think the item about Cadbury Mousse Double Chocolate was…” followed by five attributes based Ohanian’s (1990) scale of source trustworthiness: honest (reversed), trustworthy (reversed), convincing (reversed), unbiased and credible. A factor analysis revealed that the items load on one factor (EV = 2.93; R2

= 58.69%; Cronbach’s alpha = .82). The mean score of the six items is used as a measurement of attitudinal persuasion knowledge in which a higher score

corresponds to more distrust and a lower score corresponds to less critical feelings (M = 4.12, SD = 1.00).

Control variables and manipulation check.

A number of variables were measured to make sure that the effects of the study were caused by the manipulation and not any other differences between the experimental groups, thus increasing the internal validity of the study. Brand familiarity, product familiarity and

(21)

previous buying behavior was measured similar to the pretest. Blog familiarity was measured by asking the participants if they were familiar with the blog www.jeroxie.com (0 = no, 1 = yes) and blog reading frequency was measured by asking the participants to indicate how often they read blogs (1 = daily, 2 = weekly, 3 = monthly, 4 = less than monthly, and 5 = never; M = 3.07, SD = 1.33). The measures of brand familiarity, product familiarity, previous buying behavior, blog familiarity and blog reading frequency are in line with previous

disclosure-studies, such as Boerman, Van Reijmersdal and Neijens (2014) and Tutaj and Van Reijmersdal (2012). The manipulation check was measured similar to the pretest. Finally, the participants were asked to fill in gender, age and their English proficiency (1 = poor, 7 = excellent; M = 5.76, SD = 1.21). Brand familiarity and blog reading frequency will be tested in the analysis as covariates, whereas product familiarity, previous buying behavior and blog familiarity was used to exclude participants.

Results

Prior to testing the hypotheses it was tested whether the control variables correlated with the dependent variables. First, blog reading frequency was tested as a covariate. The correlation showed that blog reading frequency did not correlate with brand attitude (r = -.08, p = .43) or recognition (r = -.02, p = .82), but there was a negative, weak correlation between blog reading frequency and recall (r = -.21, p = .03). This indicates that the more frequent the consumers read blogs the less likely they are to recall the product. Second, Brand familiarity was tested as a covariate. The correlation showed that brand familiarity did not correlate with brand attitude (r = -.15, p = .13) or recognition (r = -.06, p = .53), but there was a positive, moderate correlation with recall (r = .44, p < .01). This indicates that the more familiar the participants are with Cadbury the more likely it is that they would recall the brand. Based on this blog reading frequency and brand familiarity will be used as covariates in the hypotheses with recall as the dependent variable.

Manipulation check

A chi-square test was used to check if the manipulation was successful. The test revealed that there is a significant association between the presence of a disclosure and whether or not the participants noticed it, x2

(22)

size, which was small V = .28. A cross tabulation analysis was also conducted, which revealed that 56.8% of the participants in the before-disclosure condition (n = 21) and 63.2% in the after-disclosure condition (n = 24) indicated that they saw the disclosure. 69.7% of the participants (n = 23) in the no-disclosure condition indicated that they did not see the disclosure. This means that the manipulation was effective, but the findings suggest that it could have been clearer. In order to examine whether the participants who did not see the manipulation in one of the disclosure-conditions (and vice versa) affected the results, the analysis below was also run on a smaller sample (n = 68) in which the participants who failed to recognize the manipulation were removed. The findings of this analysis did not differ from the results presented below thus the results of the larger sample (N = 108) was used in the remaining analyses. It is argued that the respondents who claimed to have seen the disclosure were perhaps more familiar with the brand or disclosures in general. Because of this and that the chi-square test did reveal a significant correlation and the fact that the smaller sample did not change the results no participants were excluded based on the manipulation check.

Testing hypothesis 1

Hypothesis 1 stated that the effect of disclosures on affective responses is mediated by

persuasion knowledge, such that a disclosure (vs. no disclosure) in blogs positively influences persuasion knowledge and persuasion knowledge in turn negatively influences affective responses. In order to test this hypothesis Hayes’ process for SPSS (2008) was performed with a bias-corrected bootstrapped confidence interval of 1000 samples on model 6 with a sequential mediation (Hayes, 2013). The model explained 30.25% of variance in brand attitude. The presence of a disclosure did not predict conceptual persuasion knowledge (β = -.14, p = .68), but conceptual persuasion knowledge significantly predicted attitudinal

persuasion knowledge as assumed (β = -.18, p < .01). Attitudinal persuasion knowledge significantly predicted brand attitude (β = .53, p < .01), but unlike what was assumed the effect was positive. The total indirect effects were also tested using process analysis. The results show that the indirect effect between disclosure (vs. no disclosure), persuasion

knowledge and affective responses was not significant (95% CI = [-.06, .09]), thus hypothesis 1 is not supported.

(23)

Hypothesis 2 stated that the effect of timing on affective responses is mediated by persuasion knowledge, such that a disclosure before (vs. after) the blog post positively influences

persuasion knowledge and persuasion knowledge in turn positively influences affective responses. In order to test this hypothesis Hayes’ process for SPSS (2008) was performed with a bias-corrected bootstrapped confidence interval of 1000 samples on model 6 with a sequential mediation (Hayes, 2013). The model explained 39.37% of variance in brand attitude. Timing did not predict conceptual persuasion knowledge (β = -.26, p = .46), but conceptual persuasion knowledge significantly predicted attitudinal persuasion knowledge as assumed (β = -.23, p < .01) and attitudinal persuasion knowledge significantly predicted brand attitude (β = .56, p < .01), but positively, not negatively as assumed. The total indirect effects were also tested using process analysis. The results show that the indirect effect between timing, persuasion knowledge and affective responses was not significant (95% CI = [-.14, .05]), thus hypothesis 2 is not supported.

Hypothesis 3 stated that the effect of disclosures on cognitive responses is mediated by persuasion knowledge, such that a disclosure (vs. no disclosure) in blogs positively influences persuasion knowledge and persuasion knowledge in turn positively influences cognitive responses. In order to test this hypothesis Hayes’ process for SPSS (2008) was performed with a bias-corrected bootstrapped confidence interval of 1000 samples on model 6 with a sequential mediation (Hayes, 2013). The analysis was performed with both recognition and recall as dependent variables.1

The presence of a disclosure did not predict conceptual persuasion knowledge (β = -.14, p = .68). Conceptual persuasion knowledge significantly predicted attitudinal persuasion knowledge as assumed (β = -.18, p < .01), but attitudinal persuasion knowledge did not predict recognition (β = -.35, p = .29). The total indirect effects were also tested using process analysis. The results show that the indirect effect between disclosure (vs. no disclosure), persuasion knowledge and recognition was not significant (95% CI = [-.10, .03]). In the analysis of recall the covariates blog reading frequency and brand familiarity were included. The presence of a disclosure did not predict conceptual persuasion knowledge (β = -.24, p = .47), but brand familiarity did significantly affect it (β =

(24)

-.68, p = .04). Conceptual persuasion knowledge significantly predicted attitudinal persuasion knowledge as assumed (β = -.17, p < .01), but attitudinal persuasion knowledge did not predict recall (β = .49, p = .07). The analysis further revealed a direct effect of disclosure vs. no disclosure on recall (β = -1.74, p < .01) as well as a significant effect of the two covariates, blog reading frequency (β = -.49, p = .01) and brand familiarity (β = 2.59, p < .01). The total indirect effects were also tested using process analysis. The results show that the indirect effect between disclosure (vs. no disclosure), persuasion knowledge and recall was not significant (95% CI = [-.03, .13]), thus hypothesis 3 is not supported.

Hypothesis 4 stated that the effect of timing on cognitive responses is mediated by persuasion knowledge, such that a disclosure before (vs. after) the blog post positively influences

persuasion knowledge and persuasion knowledge in turn positively influences cognitive responses. In order to test this hypothesis Hayes’ process for SPSS (2008) was performed with a bias-corrected bootstrapped confidence interval of 1000 samples on model 6 with a sequential mediation (Hayes, 2013). The analysis was performed with both recognition and recall as dependent variables. Timing did not predict conceptual persuasion knowledge (β = 26, p = .48), but conceptual persuasion knowledge significantly predicted attitudinal

persuasion knowledge as assumed (β = -.23, p < .01). Attitudinal persuasion knowledge did not predict recognition (β = -.35, p = .38). The total indirect effects were also tested using process analysis. The results show that the indirect effect timing, persuasion knowledge and recognition was not significant (95% CI = [-.02, .25]). In the analysis of recall the covariates blog reading frequency and brand familiarity were included. Timing did not predict

conceptual persuasion knowledge (β = 33, p = .38), but conceptual persuasion knowledge significantly predicted attitudinal persuasion knowledge as assumed (β = -.22, p = .01). Attitudinal persuasion knowledge did not predict recall (β = .64, p = .07), but brand familiarity had a significant effect on the depend variable (β = 3.52, p < .01). The total indirect effects were also tested using process analysis. The results show that the indirect effect between timing, persuasion knowledge and recall was not significant (95% CI = [-.45, .03]), thus hypothesis 4 is not supported.

(25)

Discussion

Disclosures in blog advertising have so far only been studied to a limited extent and so far results are inconclusive. Therefore this study has examined whether the consumers responses to disclosures in blogs. The aim of this study was to examine the effects of disclosure and the timing of disclosures on affective and cognitive responses and furthermore how this process was mediated by persuasion knowledge, thus there were three main aims: First to examine the effects of a disclosure vs. no disclosure, second to examine the timing of the disclosure

(before vs. after) and third to examine the underlying effects of the disclosure, that is

persuasion knowledge. To reach these aims a pilot study was first conducted to examine the content and form of the disclosure. The results showed that the content and the tming of the disclosure varied. Then an online experiment was conducted using a food blog and Cadbury Mousse Double Chocolate in the material was conducted to test the hypotheses. The study tested three conditions: the first one with a disclosure placed before the blog post, the second one with the disclosure placed after and no disclosure in the last condition. The first two hypotheses examined the effects of disclosure vs. no disclosure and the timing of the disclosure (before vs. after the blog post) on affective responses. The last two hypotheses examined the effects on cognitive responses. It was assumed for all hypotheses that

persuasion knowledge, that is recognition of the persuasive attempt and the critical attitudes toward it, mediated the effect.

The results showed that none of the hypotheses were supported. Based on this we can conclude that a disclosure compared to no disclosure does not have a positive effect on memory or a negative effect on affective responses nor are the effects of disclosures on persuasion outcomes mediated by persuasion knowledge. Furthermore, a disclosure placed before the blog post compared to after the blog post does not have a positive effect on affective responses or cognitive responses and neither is the effect mediated by persuasion knowledge. Even though the hypotheses were not supported the analysis did reveal some significant results, which should be elaborated on as well as a discussion of why no significant results were found in relation to the hypotheses. Before elaborating on the theoretical and practical implications, the limitations are being discussed.

(26)

Limitations

This study has made several contributions, both theoretical and practical, to the study of disclosures in blogs, but as with any research there are some limitations. First of all, the covariate, blog reading frequency, was related to recall, which indicate that whether or not the participants were familiar with readings blogs or not influenced their response. It would have been interesting to conduct the experiment on a more representative sample such as a group of blog readers. It is also suggested that a larger sample size could possible have influenced the results and perhaps lead to more significant results. It is recommended that future studies use a larger sample of blog readers, which would make the results more generalizable. Secondly, it would also be interesting to test the hypotheses using a different design. The online

experiment was conducted with an unknown blog, thus the participants had no prior attitude toward the blog or the blogger, which most blog readers do. By using a different design for example a longitudinal field study then other factors such as involvement with the blog and the blogger could also be included. Advertisers may use blogs to reach an audience that is already interested in their brand or product. Therefore, it would be interesting to examine the pre-attitude towards the brand to see if this intensifies after being exposed to blog advertising as well as how a disclosure affects this. The third and last limitation relates to the material (a food blog) and the product (chocolate). A different type of blog and product might result in different findings. Future studies should test other product categories as well as other blog genres and take into account other covariates, such as product interest or prior brand attitude.

Theoretical implications

Although the current study did not support the assumptions, the outcome has several

important theoretical implications as well as implications for future research. The three main theoretical implications relate to persuasion knowledge, brand responses and the timing of the disclosure. First of all, one of the main assumptions of this study was that the presence of a disclosure on a blog would automatically activate the consumer’s persuasion knowledge of the online content. The persuasion knowledge model states that when an advertising attempt is being made then the consumer’s persuasion knowledge is being activated, which enables them to recognize and analyze the advertisement (Friestad & Wright, 1994). Thus, a

disclosure in online content should help the consumer recognize the persuasive intent of the advertisement. The findings of the current study confirmed previous studies (Boerman, Van Reijmersdal & Neijens, 2014; Boerman, Van Reijmersdal & Neijens, 2012) in regard to that

(27)

conceptual persuasion knowledge (i.e. the recognition of the advertisement) automatically activates attitudinal persuasion knowledge (i.e. critical attitudes). The results furthermore showed that attitudinal persuasion knowledge significantly predicted brand attitude, so that the higher the consumer’s attitudinal persuasion knowledge the more positive their brand attitude. This is not in line with previous findings (Boerman, Van Reijmersdal & Neijens, 2012). However, this study argues that an increase in attitudinal persuasion knowledge might not necessarily lead to a negative attitude. Although perceptions of the ad and the brand are often related, then even if the consumer distrusts and dislikes the ad he/she might still like the brand or the product. Previous studies such as Mackenzie, Lutz and Belch (1986) have found that the attitude toward the ad affects brand responses, but further studies are needed to examine this in the context of blogs where a disclosure is present.

To sum up, the findings of this study did not support Friestad and Wright’s persuasion knowledge model (1994) in that persuasion knowledge is activated when a disclosure is present and in turn affects the persuasive outcome. Thus in this study persuasion knowledge had no mediating effect when a persuasive attempt was being made. The findings of this study showed that the presence of a disclosure or the timing of it did not automatically activate persuasion knowledge. These findings are similar to those of Van Reijmersdal, Tutaj and Boerman (2013), who found that the presence of a brand placement disclosure did not increase skepticism as assumed. They explained this non-effect by the fact that the mere information about sponsored content (a disclosure) does not necessarily result in increased skepticism – this will be discussed further in regards to the practical implications of this study.

The findings of this thesis regarding persuasion knowledge raise the question as to what extent persuasion knowledge is automatically activated when a persuasive attempt is being made such as in the case that a disclosure is present. One of the explanations that no significant findings or mediating effects are found could lay in the different reactions consumers have to online versus offline media. Previous studies (Ha & McCann, 2008; Cho & Cheon, 2004) have suggested that consumers react differently to online advertising as compared to offline advertising in traditional media, such as TV and print, and that ad avoidance (e.g. “banner blindness”) might be occurring more frequently online than offline (McCoy, Everard, Polak & Galletta, 2007; Sundar, Narayan, Obregon & Uppal, 1998). Tutaj

(28)

and Van Reijmersdal (2012) compared two types of online advertising: banner ads and sponsored content. They found that banner ads were more likely to activate the consumer’s persuasion knowledge than sponsored content, which indicates that consumers might have more difficulties recognizing and understanding the advertising attempt, even when it is presented in sponsored content such as a blog post with a disclosure. This might explain why in the current studies that persuasion knowledge was not activated even though a disclosure was present. Further studies should be conducted to examine the role of persuasion

knowledge and whether or not it is automatically activated when it is clear to the consumer that a persuasive attempt is being made, for instance when a disclosure is present.

The second theoretical implication relates to the cognitive responses, operationalized as brand memory and brand attitude. If we disregard the mediator then this study partly confirmed the findings of previous studies (Boerman, Van Reijmersdal & Neijens, 2014; Van Reijmersdal, Tutaj & Boerman, 2013; Boerman, Van Reijmersdal & Neijens, 2012) in regards to that a disclosure as compared to no disclosure has a positive, direct effect on cognitive responses such as brand memory. The findings only showed a significant effect on recall and not

recognition, but this can be explained by that there was not enough variance in the recognition measure. The presence of a disclosure did not affect the affective responses, which is in line with the findings of Campbell, Mohr and Verlegh (2013), but not with Nekmat and Gower’s findings (2012). The effects of disclosures in blogs on affective responses still seem to be inconclusive. Therefore future studies about disclosures in blogs need to be conducted in order to gain more knowledge about this process. Perhaps studies which include both explicit and implicit measures.

The third theoretical implication relates to the timing of the disclosure. This study examined how the timing of the disclosure, that is whether it was placed before or after the blog post, affected the brand responses. The findings of the study showed that the timing of the

disclosure did not influence the affective or the cognitive responses. This was not in line with previous studies (Boerman, Reijersdal & Neijens, 2014; Van Reijmersdal, Tutaj & Boerman, 2013), which found that when the disclosure was placed before the blog post as compared to after it gave the consumer more time to process the information, which would in turn have a positive influence on brand memory (Boerman, Van Reijmersdal & Neijens, 2014; Friestad & Wright, 1994). When a disclosure is placed after the blog post it could result in a negative

(29)

effect on affective responses according to the ‘change of meaning’ principle (Friestad & Wright, 1994), thus a disclosure placed before the blog post should have a positive effect on affective responses. Neither of these hypotheses was confirmed. This could be due to the length of the material. The pilot study and the pretest tested successfully that the material was as realistic as possible, but in hindsight it might be argued that because the text and the amount of pictures were quite limited compared to some blog posts the difference between placing the disclosure before or after the blog post were not distinct enough. This might have been prevented if the pretest would have measured more specifically the difference between the two timing-conditions. Future studies, which examine the effects of timing on blogs, should take into consideration the length of the text and amount of pictures to make sure that the difference between the timing of the disclosure is evident.

Practical implications

Despite the fact that none of the hypotheses were supported, there are still some lessons to take away from this study. First of all, it could be argued that the presence of a disclosure and the timing of it simply do not affect brand responses all that much. This means that marketers and advertisers do not need to take into consideration the possible negative effects on for example attitude or credibility, if they choose to use personal bloggers for blog advertising. The findings did show that brand familiarity and blog reading frequency affected recall, thus marketers should have in mind that the more familiar the consumer is with the brand and the less frequent they read blogs the more likely it is that they will recall the brand. Thus maybe disclosure among an interested and involved audience works as an intensifier, because these consumers might be more open to advertisements in blogs.

Secondly, this study has contributed with some practical implicated that relate to the ongoing debate on the content and form of the disclosure. The findings of the pilot study showed that the form and content of disclosures vary. The form and content of the disclosure was made based on the results of the pilot study and its effectiveness in terms it being authentic was further tested in a pretest, thus the disclosure and the material resembles an actual disclosure and blog post. Based on the findings of this study it would seem that a disclosure or the timing of the disclosure only affects brand responses to a very small extent. Cain (2011) raised questions as to how a disclosure should look like and what it should include. Based on the findings of this study it seems that the rules and regulations as to the form and content of

(30)

disclosures should be strengthened not only for disclosures in TV-shows, but for disclosures in blogs as well. This study supports the ongoing debate in that there is a call for a better definition of what a disclosure should communicate and how it can be emphasized.

The studies on disclosures in blogs have so far been limited and this thesis have made some small contributions to the literature and the practical field, but further research is needed in order to be able to reach conclusive results on the effects of disclosures in blogs and how advertisers should deal with the regulations regarding disclosures online (Federal Trade Commission, 2013). Blogs and blog advertising are here to stay and there are still lots to learn on how to use and benefit from it.

(31)

References

Abendroth, L. J., & Heyman, J. E. (2013). Honesty is the best policy: The effects of

disclosure in word-of-mouth marketing. Journal of Marketing Communications, 19(4), 245-257.

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., 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.

Brock, J. (2009). The ad-tracking debate: where should disclosure live? More options are better for consumers – and marketers. Retrieved June 4th

2014 from:

http://adage.com/article/digitalnext/online-advertising-tracking-disclosure-debate/138897/

Bruner, G. C. (1998). Standardization & justification: do AAD scales measure up? Journal of Current Issues & Research in Advertising, 20(1), 1-18.

Cain, R. M. (2011). Embedded advertising on television: disclosure, deception, and free speech rights. Journal of Public Policy & Marketing, 30(2), 226-238.

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.

(32)

Cao, J., Knotts, T., Xu, J. J., & Chau, M. (2009). Word of mouth marketing through online social networks. AMCIS, 291.

Carl, W. J. (2008). The role of disclosure in organized word of mouth marketing programs. Journal of Marketing Communications, 14(3), 225-241.

Cox III, E. P., Wogalter, M. S., Stokes, S. L., & Tipton Murff, E. J. (1997). Do

product warnings increase safe behavior? A meta-analysis. Journal of Public Policy &

Marketing, 195-204.

Disclosure (n.d.). In Merriam-Webster’s online dictionary. Retrieved April 3rd 2014 from http://www.merriam-webster.com/dictionary/disclosure

Drummond, R. (2014). Salisbury Steak Meatballs [blog post]. Retrieved April 21st

2014 from:

http://thepioneerwoman.com/cooking/2014/04/salisbury-steak-meatballs/

Edmonds, R. (2013). Federal Trade Commission will put native advertising under the microscope Wednesday. Retrieved June 4th

2014 from: http://www.poynter.org/latest-

news/business-news/the-biz-blog/231868/federal-trade-commission-will-put-native-advertising-under-the-microscope-wednesday/

Federal Trade Commission (2013). How to make effective disclosures in digital advertising. Retrieved March 16th

2014 from:

http://www.ftc.gov/sites/default/files/attachments/press-releases/ftc-staff-revises-online-advertising-disclosure-guidelines/130312dotcomdisclosures.pdf

Friestad, M., & Wright, P. (1994). The persuasion knowledge model: How people cope with persuasion attempts. Journal of Consumer Research, , 1-31.

Fu, J., & Chen, J. H. (2012). An investigation of factors that influence blog

advertising effectiveness. International Journal of Electronic BusinessManagement, 10(3), 194.

(33)

Gaille, B. (2013). How many blogs are on the Internet. Retrieved June 4th 2014 from: http://www.wpvirtuoso.com/how-many-blogs-are-on-the-internet/

Grubbs Hoy, M., & Andrews, J. C. (2004). Adherence of prime-time televised advertising disclosures to the “clear and conspicuous” standard: 1990 versus 2002. Journal of Public Policy and Marketing, 23(2), 170-182.

Hayes, A. F. (2008). Introduction to mediation, moderation, and conditional process

analysis: A regression-based approach. Guilford Press.

Hayes, A.F. (2013). Model templates for SPSS and SAS [PDF-file]. Downloaded May 25th from: http://www.afhayes.com/introduction-to-mediation-moderation-and-conditional-process-analysis.html

Hennig-Thurau, T., Malthouse, E. C., Friege, C., Gensler, S., Lobschat, L.,

Rangaswamy, A., & Skiera, B. (2010). The impact of new media on customer relationships. Journal of Service Research, 13(3), 311-330.

Higgins (2013). Technorati 2013 Digital Influence Report. Retrieved March 16th₂₀₁₄ from: http://technorati.com/business/article/technorati-medias-2013-digital-influence-report/

International Chamber of Commerce (2004). Advertising and marketing

communication practice: Consolidated ICC code (2004). Retrieved April 8th 2014 from:

http://ec.europa.eu/health/archive/ph_determinants/life_style/alcohol/forum/docs/alco

hol_lib5_en.pdf

Johnson, T. J., & Kaye, B. K. (2009). In blog we trust? deciphering credibility of

components of the internet among politically interested internet users. Computers in Human Behavior, 25(1), 175-182.

(34)

knowledge, brand attitude, and purchase intention: the mediating role of presence. Journal of Advertising, 31(3), 43-57.

Mack, R. W., Blose, J. E., & Pan, B. (2008). Believe it or not: Credibility of blogs in tourism. Journal of Vacation Marketing, 14(2), 133-144.

MacKenzie, S. B., Lutz, R. J., & Belch, G. E. (1986). The role of attitude toward the ad as a mediator of advertising effectiveness: A test of competing explanations. Journal of

marketing research, 130-143.

Nekmat, E., & Gower, K. K. (2012). Effects of Disclosure and Message Valence in Online Word-of-Mouth (eWOM) Communication: Implications for Integrated Marketing Communication. International Journal of Integrated Marketing

Communications, 4(1).

Norén, L. (2011). Food blog study descriptive statistics part 1. Retrieved April 15th from: http://thesocietypages.org/graphicsociology/tag/food-blog-study/

Van Noort, G., Voorveld, H. A., & Van Reijmersdal, E. A. (2012). Interactivity in brand web sites: cognitive, affective, and behavioral responses explained by

consumers' online flow experience. Journal of Interactive Marketing, 26(4), 223-234.

Ohanian, R. (1990). Construction and validation of a scale to measure celebrity endorsers' perceived expertise, trustworthiness, and attractiveness. Journal of

advertising, 19(3), 39-52.

Padmanabhan, K. H. (2008). A hierarchy of promotional blogs. Journal of Website

Promotion, 3(3-4), 153-160.

Van Reijmersdal, E., Neijens, P., & Smit, E. (2005). Readers' reactions to mixtures of advertising and editorial content in magazines. Journal of Current Issues & Research

in Advertising, 27(2), 39-53.

I spy with my little eye a disclosure on the blog : a study on how disclosures in blogs affect brand responses and the mediating role of persuasion knowledge

I spy with my little eye a disclosure on the blog

Table of Contents