An Investigation of Consumers’ Persuasion Knowledge for Advertorials
and its Influence on Consumer Responses
Master’s Thesis at the Graduate School of Communication, University of Amsterdam by Theresa Maria Mayer (Student ID number: 10652752)
supervised by Dr. A.S. Velthuijsen
Master’s Program Communication Science: Persuasive Communication June, 2014
Abstract
This study investigates whether today’s media consumers have developed persuasion knowledge for advertorials, enabling them to identify, understand and evaluate the advertorial’s disguised commercial character and whether this knowledge alters consumer responses to this advertising format. In an experiment, 134 students were exposed to a magazine excerpt containing either an advertorial or a traditional print advertisement. With respect to persuasion knowledge, the re-sults show that the participants had a very good understanding of the persuasive and selling in-tent of the advertorial’s message, but did not as readily identify advertorials as advertising com-pared to the traditional ad. Furthermore, the advertorial was perceived as a less appropriate ad-vertising strategy while still being evaluated rather neutrally. The persuasion knowledge com-ponent perceived appropriateness seems to be strongly related to the consumer response varia-bles ad attitude and product attitude and enhanced the effects of the advertising format on these two attitudinal consumer responses. Involvement was found to be an additional factor that was positively related to ad attitude and product attitude. This study shows that today’s media con-sumers have developed the necessary level of persuasion knowledge to understand the adver-torial’s commercial character and that the evaluation of its appropriateness plays a role for atti-tudinal consumer responses.
Introduction
Do you know it’s advertising? In today’s media environment, consumers are confronted
with new advertising formats for which this question cannot easily be answered anymore. Ad-vertisers have found new ways to reach consumers by integrating their commercial messages into editorial content or masking them as such, challenging consumers’ ability to identify them as advertising. In order to sidestep consumers’ increasingly negative attitudes towards and re-duced attention to traditional advertising caused by the advertising clutter, these integrated ad-vertising formats aim not to activate persuasion knowledge, a mental construct that enables con-sumers to identify and respond to persuasive attempts such as traditional advertising. Driven by the concern that consumers use their persuasion knowledge for cognitive defenses and resistance against persuasion in advertising, it has been argued that these advertising formats are thus more effective in eliciting favorable consumer responses compared to traditional advertising (Dahlén & Edenius, 2007; de Pelsmacker & Neijens, 2012; van Reijmersdal, Neijens & Smit, 2005).
The advertorial is one of these relatively new advertising formats as it conceals its com-mercial character by mimicking editorial content (Kim, Pasadeos, & Barban, 2001). Research on the effectiveness of advertorials supports the notion that advertorials are a beneficial advertising format compared to traditional advertising to overcome the obstacles of the advertising clutter, as they received higher attention, scored better on message recall and attitude towards the ad (Kim et al., 2001; van Reijmersdal et al., 2005). However, it remains to be investigated whether advertorials elicit these beneficial consumer responses because of the assumed advantage that this advertising format is not as readily identified as advertising, i.e. does not activate persuasion knowledge.
Based on the Persuasion Knowledge Model by Friestad and Wright (1994), this study proposes that today’s media consumers have developed the relevant persuasion knowledge nec-essary to recognize and understand advertorials’ commercial nature due to experience with this advertising format over the past three decades (Donaton, 1992; Friestad & Wright, 1994, 1999). If this proposition holds true, it raises the question whether this knowledge alters the way to-day’s media consumers respond to advertorials. Previous research on the effects of integrated advertising formats revealed that consumers’ responses to these formats can be influenced by their persuasion knowledge and the activation thereof (de Pelsmacker & Neijens, 2012). When studying the effects of integrated advertising formats, such as advertorials, which capitalize on
the assumed advantage of not activating persuasion knowledge, it seems therefore crucial to investigate whether this assumption can be confirmed.
The aim of this study is twofold. First, to investigate to what extent today’s media con-sumers have developed persuasion knowledge for the advertorial format, enabling them to iden-tify, understand and evaluate its commercial nature. Second, whether the level of persuasion knowledge for advertorials influences attitudinal consumer responses to this advertising format. Thus, this study seeks to answer the following research questions: (1) to what extent have con-sumers developed persuasion knowledge for advertorials and (2) does this knowledge influence attitudinal consumer response variables to this advertising format?
Theory and Research on Advertorials and Persuasion Knowledge An Introduction to Advertorials: Definition and Literature Review
Advertorials are a commonly used advertising format in magazines since the 1980s, ini-tiated as a response to the increasingly negative attitudes towards traditional advertising among media consumers caused by the advertisement clutter in magazines and media in general (Dona-ton, 1992). The neologism advertorial, composed of the words advertisement and editorial, em-bodies its two main characteristics: it is an advertisement in the editorial style of the magazine it is placed in. Thus, advertorials are commercial messages that resemble the magazine’s editorial content in typeface and layout (van Reijmersdal et al., 2005), are often composed by the journal-ists of the magazine in cooperation with the advertiser or brand and subsequently aim at deliver-ing the experience of readdeliver-ing a regular article (Prounis & DeSantis, 2004).
Advertorials experienced their first period of prosperity in terms of prevalence and em-pirical research in the 1980s and early 1990s. Between 1986 and 1991, the total revenue from advertorials in American magazines doubled and in 1991, it accounted for 10% of the total mag-azine advertising revenue (Donaton, 1992). This widespread use of advertorials gained consid-erable attention in academic research and in trade publications during this period, investigating the prevalence of advertorials in magazines, the opinions of advertising professionals about vertorials and, to a limited extend, the effectiveness of advertorials (Kim et al., 2001). The ad-vertorial’s way of disguising the sender and commercial character of the message has also prompted a debate, whether the use of this marketing technique is unethical and deceives
consumers (Kim et al., 2001). To protect consumers from deceiving practices in the use of the advertorial format, several guidelines have been developed that require magazines to clearly mark advertorials with the word advertising or similar indications, for example the guidelines for use of advertorials by the American Society of Magazine Editors (ASME). Content analyses showed, however, that not all publications follow these guidelines (Cameron, Ju-Pak, & Kim, 1996). After the mid-1990s, the growth of advertorials in magazines, as well as academic re-search on this new advertising format has somewhat slowed down (Kim et al., 2001) and more recent figures for their prevalence in magazines can only be found for the Netherlands. A con-tent analysis of 26 magazines from van Reijmersdal, Vos and Neijens published in 2002 depict-ed that advertorials accountdepict-ed for 6% of all advertising pages.
More recent research on the effectiveness of advertorials has yielded several advantages of the advertorial format over traditional print advertisements. First, as advertorials are a rela-tively new advertising format, novelty effects have been attributed to enhance attention to their commercial message (Kim et al., 2001). Second, it has been argued that advertorials are appre-ciated1 more compared to traditional advertisements as they are less common (Kim et al., 2001; van Reijmersdal et al., 2005). Third, compared to the short claims and messages that are charac-teristic for traditional print advertisement the extended length of an advertorial’s message can lead to deeper elaboration of its content and subsequently to better message recall (Kim et al., 2001). Finally, the main benefit of advertorials as described in the majority of the literature lies in the similarity of the advertorial with the editorial content of a magazine and the reduced dis-tinctiveness of advertorials as advertising (e.g. Cameron & Ju-Pak, 2000; Donaton, 1992; Kim et al., 2001; Robinson, Ozanne, & Cohen, 2002; van Reijmersdal et al., 2005). The similarity of the advertorial format to the editorial style of its host magazine can lead to a transfer of attitudes towards and perceptions of the editorial material to the advertorials’ commercial message. Pre-vious research has confirmed this notion as it found that the higher the similarity of an adver-torial or other ediadver-torial-like advertisement formats (e.g. theme features), the better they score on attention, appreciation, and believability (Robinson et al., 2002; van Reijmersdal et al., 2005). Furthermore, Kim and colleagues (2001) concluded that advertorials, whether labeled as adver-tising or not, elicit more attention and recall compared to traditional advertisements, as they fool
1 The construct appreciation was also termed as perceived advertising value (Tutaj & van Reijmersdal, 2012) or ad
readers into greater involvement with the advertising message due to their similarity to editorial copy. With respect to these findings on the advantages of advertorials over traditional advertis-ing in elicitadvertis-ing more favorable consumer responses, it seems that advertorials have indeed be-come a successful advertising format to sidestep the increasing negative attitudes towards tradi-tional advertising and consumers’ tendency of avoiding advertising messages (Speck & Elliott, 1997). The reason or underlying mechanism for the advertorial’s advantage in eliciting these favorable consumer responses, however, was not explored within the scope of the mentioned studies and remains to be investigated.
Among others, Dahlén and Edenius (2007) argued that advertorials elicit these favorable consumer responses because the blurred boundaries between editorial and commercial content and the disguised sender of the message impede the identification of advertorials as advertising, thus as a persuasive attempt. In doing so, advertorials deceive the consumer into believing that the message is objective editorial copy instead of biased information that aims to persuade and to sell. Consequently, when exposed to an advertorial, consumers are less likely to activate and use their persuasion knowledge, a mental construct that consumers use to identify and respond to persuasive attempts in advertising (Dahlén & Edenius, 2007; de Pelsmacker & Neijens, 2012; Friestad & Wright, 1994). If this assumption holds true, thus if the underlying mechanism of advertorials’ impact on consumer responses lies in the lack of consumers’ ability to identify this advertising format as advertising by not activating their persuasion knowledge, one must pose the question whether the results from previous studies are still applicable for today’s media con-sumers. Given the widespread use of advertorials since the 1980s, media consumers have gained experience with the advertorial format over the years, possibly enabling them to develop persua-sion knowledge for this specific advertising format. This knowledge could allow consumers to identify advertorials as advertising in the same way as traditional advertisements, altering the way consumers respond to an advertorial’s persuasive message. Figure 1 illustrates these hy-pothesized relations. The explanation of this conceptual model is provided in the following sec-tions. Drawing on persuasion knowledge literature and previous research on integrated advertis-ing formats, the possible development of persuasion knowledge for advertorials is explained, relevant persuasion knowledge components for advertorials are identified and their potential role for consumer responses to the advertorial format are discussed in this context.
H1a, H1b & H2a, H2b Consumer Responses Persuasion Knowledge Advertising Recognition Perceived Appropriateness Advertising Skepticism Understanding the
Persua-sive and Selling Intent
Involvement Exposure to Advertorial or Traditional Print Advertisement
Figure 1. Conceptual Model – Simplified visualization of hypothesized relations between advertising
format, persuasion knowledge and consumer response variables based on previous literature.
Persuasion Knowledge for Advertorials
Definition and delimitation of persuasion knowledge in the context of advertorials. Several conceptualizations of knowledge related to persuasion and advertising can be found in the literature, for example as advertising literacy (Livingstone & Helsper, 2006), understanding
advertising (Lawlor & Prothero, 2008), cognitive advertising competencies (Rozendaal,
Buijzen, & Valenburg, 2010) or persuasion knowledge (Friestad & Wright, 1994). In this re-search, the term consumers’ persuasion knowledge for advertorials is used to describe the set of competences and knowledge related to the persuasive nature of advertorials and the recognition as such. This concept is based on the more general construct persuasion knowledge (Friestad & Wright, 1994), which includes a set of interrelated beliefs and competences related to recogniz-ing and understandrecogniz-ing persuasion situations, for example advertisrecogniz-ing, and guides the evaluation of and subsequent reactions to persuasion attempts (Friestad & Wright, 1994; Tutaj & van Reijmersdal, 2012).
Ad Attitude Product Attitude
Investigating the level of consumers’ persuasion knowledge for advertising formats and its influence on consumer responses is especially important when studying the effectiveness and perception of relatively new advertising formats, such as advertorials, which rely on the as-sumed advantage of not being recognized as advertising, i.e. not activating persuasion knowledge, and challenging the consumers’ way of understanding and processing advertising. De Pelsmacker and Neijens (2012) recently acknowledged the importance of investigating the effect of persuasion knowledge on consumer responses in the study of advertising formats which try to conceal their commercial nature. Within this context the effects of advertising for-mats such as online advertising forfor-mats, advergames and product placements on consumer re-sponses have been studied. The findings revealed an influencing role of persuasion knowledge and awareness of an advertising format’s persuasive intent on consumer responses (cf. de Pelsmacker & Neijens, 2012). This study is to our knowledge the first to investigate this process for the advertorial format by focusing on the identification of relevant persuasion knowledge components that may explain consumer responses to this advertising format.
The Persuasion Knowledge Model(Friestad & Wright, 1994). The Persuasion Knowledge Model (PKM) by Friestad and Wright (1994) is applied in this study to explain a possible de-velopment of persuasion knowledge for the advertorial format and its influence on consumer responses. While several publications for persuasion knowledge and individual differences in persuasibility can be found in the literature (cf. Wright, Friestad, & Boush, 2005), the PKM is one of the most influential in the field. It offers a theoretical framework of how people, more specifically media consumers, develop persuasion knowledge and how this knowledge influ-ences their evaluation of and responses to persuasive attempts.
According to Friestad and Wright (1994), persuasion knowledge continuously develops throughout the lifespan due to experiences with different forms of persuasion episodes, for ex-ample different advertising formats, strategies or sales interactions. Media consumers build up knowledge with regard to marketers’ advertising strategies, their effectiveness and appropriate-ness, as well as their own coping tactics and goals. Friestad and Wright (1994) call this set of knowledge and beliefs persuasion knowledge. Media consumers furthermore develop knowledge concerning the sender of the message, consisting of beliefs about their traits, compe-tences and goals (agent knowledge). The third component of the PKM is topic knowledge,
which relates to the beliefs and knowledge about the topic of the persuasive message. Media consumers can use this combined knowledge to identify how, when and why marketers attempt to influence them and to have control over the outcome of a persuasion episode. Thus, persua-sion knowledge guides consumers’ responses in such a way as to achieve their own goals, which can either lead to resistance against the message or to persuasion (Friestad & Wright, 1994).
The PKM applied to advertorials: Persuasion knowledge for advertorials and its role for consumer responses. Considering that advertorials have been applied since the 1980s and constitute an established form of advertising in magazine nowadays (Dix & Phau, 2009), it can be assumed that today’s media consumers have developed some form of persuasion knowledge for advertorials due to experience with this advertising format. Whether this assumption holds true is investigated within the scope of this study. Focusing on adult media consumers between 18 and 30 years of age, allows investigating the level of persuasion knowledge for a group, whose socialization as consumers has taken place at the same time as advertorials have been occurring in magazines. Therefore, there is no variety among members of this group in terms of experience with advertorials caused by this advertising format’s prevalence in magazines. The first research question this study addressed is:
RQ1: To what extent have today’s consumers developed persuasion knowledge for adver-torials?
For the context of advertorials, three persuasion knowledge components have been identified to measure the level of today’s media consumers’ persuasion knowledge for advertorials and to potentially influence subsequent consumer responses. A first component and an indication of the existence of persuasion knowledge for advertorials is recognition of the advertorial format as
advertising, i.e. whether consumers make the distinction between editorial and commercial
con-tent in a magazine and do not mistakenly perceive the advertorial as editorial copy. If consumers possess relevant persuasion knowledge for advertorials, thus if they know that an advertorial is an advertising format that presents its commercial message in an editorial style, this knowledge can be used to identify advertorials in magazines as advertising. In previous literature, recogni-tion of the persuasive nature of an advertising format or strategy was awarded to be the first
level of persuasion knowledge and necessary to activate further persuasion knowledge structures (Friestad & Wright, 1994; John, 1999).
Similarly to previous studies, the understanding the persuasive and selling intent of an advertorial, thus the knowledge that the message is aimed at persuading and selling a product or advertising a brand, is defined as such knowledge structures and was identified as a second component of persuasion knowledge for advertorials (e.g. Dahlén & Edenius, 2007; Tutaj & van Reijmersdal, 2012; Rozendaal et al., 2010). If consumers do not possess this relevant knowledge, they might not be able to understand the commercial intent of the message in an advertorial and risk to be unknowingly persuaded by a biased agent - the brand or advertiser. On the other hand, if consumers recognize advertorials as a persuasive message due to their persua-sion knowledge for this advertising format and understand its persuasive and selling intent, they gain control over the outcome of the exposure to the message. For advertorials, this outcome may still, from the marketers’ viewpoint, be more positive compared to traditional advertise-ments. A previous study by Kim and colleagues (2001) has shown that even though the vast majority of their participants was aware that advertorials are advertising2, advertorials still scored higher on message relevance, attention and recall compared to traditional advertisements.
To understand possible effects on consumer responses for the persuasion knowledge com-ponent understanding of persuasive and selling intent, research investigating similar advertising formats that integrate editorial and commercial content and conceal their persuasive intent is discussed, as research for advertorials for this context is still missing. Tutaj and van Reijmersdal (2012) found in their study that participants’ understanding of persuasive and selling intent for the online advertising formats sponsored content and banner ads negatively affected the attitude towards the advertising’s message in terms that it was perceived as more irritating. It did not, however, influence consumers’ attitudes in terms of the informational and amusement value of the advertising format. Negative effects of the activation of this persuasion knowledge compo-nent can also be found for the integrated advertising format brand placement. The quantitative synthesis of 57 studies on brand placements by van Reijmersdal, Neijens, and Smit (2009) yielded that persuasion knowledge theory can explain the effects of placement prominence.
2 Note: The participants in the study by Kim and colleagues (2001) were 215 communication students, who can be
assumed to be more knowledgeable about new advertising formats, in this case the advertorial format, than the general population. Therefore, a conclusion concerning the level of media consumers’ persuasion knowledge, thus whether a more general population is able to recognize advertorials as advertising, cannot be drawn on the basis of this study.
According to the authors, activated persuasion knowledge due to prominently placed brands leads to the awareness that the placement is aimed to influence, which subsequently leads to cognitive defenses such as skeptical attitudes towards the placement and counterarguing against it (e.g. Cowley & Barron, 2008; Russel, 2002). Further studies on sponsorship and brand place-ment disclosures in television programs yielded that recognition of the placeplace-ment as advertising, i.e. the activation of persuasion knowledge due to disclosures, is a mediating factor for their effectiveness and can cause resistance to persuasion and more critical feelings towards the mes-sage and the communicator (e.g. Boerman, van Reijmersdal, & Neijens, 2014; Campbell & Kirmani, 2000; Cowley & Barron, 2008). These findings indicate that activated persuasion knowledge influences how media consumers perceive and respond to integrated advertising formats such as sponsored content or brand placement. The vast majority of these studies shows that consumers have been socialized to be skeptical towards advertising and that activation of persuasion knowledge causes more skepticism and resistance towards the ad message and the brand or product (e.g. Boerman et al., 2014; Obermiller, Spangenberg, & MacLachlan, 2005).
The discussion of previous research showed that recognition of an advertising format as
ad-vertising can activate subsequent persuasion knowledge structures, defined as understanding of persuasive and selling intent and that this activated knowledge could result in more negative
attitudinal consumer responses. It has been argued that advertorials are therefore a beneficial advertising format, as consumers are less likely to identify them as advertising and understand their persuasive and selling intent, leading to favorable consumer responses in terms of a more positive attitude towards the ad as shown in previous studies (Dahlén & Edenius, 2007; Kim et al., 2001; van Reijmersdal et al., 2005). Whether this assumption holds true was investigated with the test of hypothesis H1a:
H1a: Advertorials elicit a more positive attitude towards the ad compared to traditional print advertisements and this effect is influenced by the level of understanding the persuasive and selling intent (persuasion knowledge).
Previous studies investigating the effectiveness of advertorials only included attitude to-wards the ad as attitudinal consumer response variable. Friestad and Wright (1994) however suggested that persuasion knowledge is used by consumers to refine their product attitude when exposed to advertising. Therefore, attitude towards the product was chosen as second attitudinal
consumer response variable to investigate the influence of the persuasion knowledge component understanding the persuasive and selling intent on the effects of advertorials. It was hypothe-sized in line with hypothesis H1a:
H1b: Advertorials elicit a more positive attitude towards the product compared to traditional print advertisements and this effect is influenced by the level of understanding the persua-sive and selling intent (persuasion knowledge).
Perceived appropriateness of the advertising format was identified as the third persuasion
knowledge component for advertorials that potentially influences consumer responses to adver-torials. A consumer’s perception of an advertising format’s appropriateness can become relevant when persuasion knowledge guides a consumer’s coping behavior in regard to the persuasive attempt of the advertorial. As Friestad and Wright (1994) note, the perceived appropriateness (e.g. fairness, manipulative nature) of the advertising strategy influences the evaluation of the persuasion episode and the subsequent coping behavior. While previous studies have not includ-ed the variable perceivinclud-ed appropriateness as part of persuasion knowlinclud-edge (e.g. Tutaj & van Reijmersdal, 2012), it seems all the more important to include this aspect in the concept of per-suasion knowledge for advertorials due to the format’s hidden attempt to influence consumers by blurring the boundaries between editorial and commercial content. In previous research the concern was expressed that consumers perceive advertorials as misleading and not appropriate due to this disguised commercial character of advertorials, some say deceiving way of influenc-ing consumers (Cameron et al., 1996; Hausknecht, Wilkinson, & Prough, 1991; Kim et al., 2001). Van Reijmersdal and colleagues (2005) on the other hand found in their research on mixed advertising formats that advertorials are perceived as a rather appropriate3 advertising format and concluded that this might lead to positive consumer responses. Thus, whether con-sumers evaluate this strategy to be appropriate or rather manipulative could influence their re-sponse to advertorials.
A study by Wei, Fisher, and Main (2008) on branded product placements in radio pro-grams supports the notion that perceived appropriateness of new, integrated advertising formats influences consumer responses. In this study, the consumers’ perception of the advertising
3 Note: The construct perceived appropriateness was conceptualized as acceptance in the study by van Reijmersdal
strategy as appropriate diminished the negative effects of sponsorship disclosures, i.e. awareness of the persuasive intent, on brand evaluations. The same effect was shown by Yoo (2009) with regard to consumers’ click-through behavior and intentions to visit a website for the integrated online advertising practice keyword search ads. These findings point out shortcomings of any assumptions that the consumers’ use of persuasion knowledge inevitable results in resistance (Yoo, 2009) and show that the perceived appropriateness of integrated advertising formats in-fluences consumer responses. Wei and colleagues (2008) as well as Yoo (2009) did not, howev-er, consider perceived appropriateness as a part of consumers’ persuasion knowledge but as a separate aspect of consumers’ processing of advertising, failing to adequately reflect the PKM by Friestad and Wright (1994). This study therefore attempts to investigate consumers’ use of persuasion knowledge in evaluating and responding to integrated advertising formats in a more comprehensive way by acknowledging perceived appropriateness of the advertising format to be one component of persuasion knowledge for advertorials.
It is concluded that if persuasion knowledge is used to respond to advertorials and if this advertising format is perceived as an appropriate advertising format, they are being evaluated more positively. In this study, perceived appropriateness was therefore deemed to be a crucial element for persuasion knowledge for advertorials and as an underlying mechanism that could explain the consumer response variable attitude towards the ad. Similarly to hypothesis H1a, it is assumed that advertorials elicit a more positive ad attitude compared to traditional print adver-tisements and that this effect can be influenced by the persuasion knowledge component per-ceived appropriateness of the advertising format. The following hypothesis was formulated:
H2a: Advertorials elicit a more positive attitude towards the ad compared to traditional print advertisements and this effect is influenced by the level of perceived appropriateness of the advertising format (persuasion knowledge).
While the study from Wei and colleagues (2008) showed that perceived appropriateness furthermore is positively related to brand evaluations, it was hypothesized that the same holds true for product evaluations. As previously mentioned, Friestad and Wright (1994) identified product attitude as one of the consumer responses which may be refined by the use of persua-sion knowledge. Furthermore, various content analyses showed that the vast majority of adver-torials in US magazines are used to advertise products or services instead of brands (Cameron et
al., 1996; Cameron & Ju-Pak, 2000). Therefore, product attitude was deemed to be a more ap-propriate measure as a consumer response variable for advertorials than brand attitude and it was hypothesized in line with hypothesis H1b and the findings from Wei and colleagues (2008) on brand evaluations:
H2b: Advertorials elicit a more positive attitude towards the product compared to traditional print advertisements and this effect is influenced by the level of perceived appropriateness of the advertising format (persuasion knowledge).
To summarize, three components were identified to measure persuasion knowledge for ad-vertorials, namely recognition of the advertorial as advertising, understanding the persuasive
and selling intent of its message and perceived appropriateness of the advertising format.
Fur-thermore, it was hypothesized that the two latter components influence the effects of advertori-als on attitude towards the ad and product. The second aim of this research was to investigate these hypothesized relations by answering research question 2:
RQ2: Do the persuasion knowledge components perceived appropriateness of the advertis-ing format and understandadvertis-ing the persuasive and selladvertis-ing intent of its message influence atti-tudinal consumer responses to advertorials?
The role of advertising skepticism and involvement. The study by Tutaj and van Reijmersdal’s (2012) was previously mentioned as it yields interesting findings on how persua-sion knowledge for new, integrated advertising formats affects consumer responses and was used to make predictions for the context of this research. In the study by Tutaj and van Reijmersdal (2012) the variable advertising skepticism was found to be the most important con-tributor to the perception of the two online advertising formats, the integrated advertising format
sponsored content and banner ads. High skepticism led to a more negative attitude towards the
ad in terms of its informational and amusement value and to more irritation of the advertising format. Next to recognition as advertising and understanding the persuasive and selling intent, Tutaj and van Reijmersdal (2012) included advertising skepticism, as a third component of per-suasion knowledge for new online advertising formats. This approach, however, has to be ques-tioned due to several reasons. First, their results showed that advertising skepticism was not re-lated to the two other components of the construct persuasion knowledge in the study, which in
turn were both positively correlated. This result indicates that advertising skepticism is not a component of persuasion knowledge for specific new advertising formats. This argument is fur-thermore supported by the PKM, which does not include advertising skepticism as a component, but rather as a possible outcome of consumers’ experiences with advertising (Friestad & Wright, 1994). In addition, advertising skepticism might differ for different advertisement formats. In-cluding a general belief about advertising, as done by Tutaj and van Reijmersdal (2012), into the construct of persuasion knowledge for a specific advertising format seems not to adequately measure this specific knowledge. Nevertheless, advertising skepticism as a general, attitudinal predisposition should be considered, next to persuasion knowledge, as an influencing factor on consumer responses to integrated advertising formats. Not only Tutaj and van Reijmersdal (2012) demonstrated that advertisement skepticism influences the perception of the integrated advertising format sponsored content. Studies for traditional advertising also showed that the extent of advertising skepticism is a determinant of consumers’ responses to advertising (John, 1999; Obermiller & Spangenberg 1998, 2000; Obermiller et al., 2005).
One can conclude that next to persuasion knowledge, the level of media consumers’ general advertising skepticism may influence consumer responses to advertorials. It was therefore con-sidered in this study as an additional variable with the following predictions regarding its influ-ence on the identified attitudinal consumer responses:
H3: General advertising skepticism is negatively correlated with (a) ad attitude and (b) product attitude.
A second additional factor to influence consumer responses to advertorials could be in-volvement. Previous research on format effects and stimulus processing of magazine advertise-ments revealed that involvement affects several consumer responses such as attention to the message, recall and the type and amount of information processing (elaboration) (Celsi & Ol-son, 1988; Lord & Putrevu 1998; Petty, Cacioppo, & Schumann, 1983). The Elaboration Likeli-hood Model (ELM) from Petty and Cacioppo (1986) is an established model to explain the role of involvement in the process of how the attitudes of consumers are formed and changed due to exposure to advertising. According to the ELM, involvement, conceptualized as personal rele-vance, increases a consumer’s motivation to engage in a thorough consideration of information that is relevant for the topic or product. If involvement is high, the presence of arguments and
their quality influence a consumer’s attitude towards the product or topic at hand. This form of information processing is called the central route of processing, while the peripheral route of processing guides low involved consumers’ attitude formation. Here, heuristic cues such as vis-uals or emotional appeals are awarded to influence a consumer’s attitude (Petty et al., 1983).
In the context of advertorials, involvement can play a role in guiding consumers’ attention towards the detailed product information provided in an advertorial, as consumers’ motivation to attend to information increases with the level of involvement. Hence, consumers who are low involved with a product or topic of an advertorial’s message may generally lack the interest to assess this information and are therefore less likely to attend to the advertorial’s message. High-ly involved consumers on the other hand should be motivated to do so, as well as to thoroughHigh-ly process the provided information. If this assumption holds true, involvement should be positive-ly related with the recall of an advertorial’s information due to more attention to and higher elaboration of the message.
A more thorough processing of an advertorials message due to higher involvement could additionally lead to a more positive product attitude if arguments in the advertorial’s message are convincing (Petty et al., 1983). However, involvement could also explain possible negative responses as a result of exposure to an advertorial. Highly involved consumers tend to be more skeptical towards advertised products (Petty et al., 1983) and are subsequently more likely to counterargue a message or argument compared to lowly involved consumers (Lord & Putrevu, 1998). The previously mentioned controversy surrounding advertorials because of their possible deception may reinforce the emergence of negative cognitive responses when consumers are exposed to an advertorial. Kim and colleagues (2001) support this notion as they argued that their finding, advertorials elicit higher message elaboration, could translate into less favorable attitudes towards the brand or product if consumers were suspicious of the message’s source, i.e. the advertiser of the advertorial format. In addition, deeper elaboration of the text may lead to a better understanding of the persuasive and selling intent of the advertorial and to a reduced like-lihood of being mistakenly perceived as editorial material.
Based on the ELM and previous research on advertorials, it was hypothesized that higher in-volvement is related to better recall of the information of an advertorial’s message and to a bet-ter understanding of its persuasive and selling intent, as well as recognition as advertising.
H4: Involvement is positively correlated with (b) recognition as advertising and (c) under-standing of persuasive and selling intent.
As mentioned, the level of involvement could lead to either a more positive or a more nega-tive product attitude. Therefore, a relation between consumers’ involvement and this attitudinal consumer response was expected, but no hypotheses could be formulated for the direction of these relations. Within the scope of this study, the relation between involvement and product attitude were therefore investigated to identify additional factors that may, next to the predicted influence of persuasion knowledge, explain consumer response variables to advertorials. Addi-tionally, the same was explored for ad attitude, the second attitudinal consumer response varia-ble in this study.
Method Design and Participants
A one factorial between-subjects design was used to test the predicted relations between the advertising format, persuasion knowledge, consumer response variables and the factors in-volvement and advertising skepticism. In this design, advertising format (advertorial vs. tradi-tional print advertisement) was the independent variable and recall, ad attitude and product atti-tude the dependent variables. In the analysis, persuasion knowledge was considered a mediator.
134 students who were approached in four different university cafeterias of two large German universities participated in this study (nadvertorial = 68, nadvertisement = 66) and received a thank you in the form of a sweet snack. The participants were between 18 and 30 years of age (M = 22.61, SD = 2.52) and 60.4% were female.
Experimental Stimuli Development
Each participant was exposed to one magazine excerpt of the German magazine NEON, which was composed for this study. The magazine excerpt contained a cover page, six pages with editorial content and the experimental stimulus (see Appendix A). The magazine NEON was chosen due to the similarity between the magazine’s main audience and the research sam-ple, namely students and young professionals between 20 and 29 years of age. The advertorial
or traditional print advertisement stimuli was embedded in the magazine excerpt on page four, following three pages of short articles and it preceded a one three-page article, in order to mimic the regular composition of the magazine.
A technological fitness gadget was identified as a suitable product for the experimental stimuli, as fitness and sports, as well as electronics, are among the most frequently used catego-ries in advertorials (Cameron et al., 1996; Stout, Wilcox, & Greer, 1989). This product and topic choice furthermore permitted a not manipulated variance of involvement between the partici-pants. The pretest of the questionnaire (N = 8) confirmed that the topic running and product choice technological sports gadgets allowed participants to be assigned to low and high in-volvement groups. The German sporting goods brand Adidas was chosen to ensure the same level of brand familiarity for all participants, as it is a highly familiar brand in Germany. The product choice for the stimuli fell on the Adidas miCoach, which is a training optimizing system for non-professionals and professionals, comprised of a small heart rate monitor, a speed tracker and a mobile app.
The advertorial and the traditional print advertisement were written in German and de-signed to be as similar as possible, only differing in the characteristics of the advertising format (see Figure 2 or Appendix B). The same visual was used as the background of the advertisement and as the main visual of the advertorial. The claim in the traditional advertisement “Train smart. Set your goals. Build a plan. See your results.” was likewise used as the headline of the advertorial. A visual of the product and the Adidas miCoach logo was shown in both stimuli. The design of the advertorial was based on previously used advertorials of the magazine NEON and identically labeled as advertisement on the top right of the page, complying with the ASME guidelines. The text of the advertorial was written from the viewpoint of a young woman, who was training for the marathon, telling about her positive experience with the Adidas miCoach and explaining how to use it. Similarly to regular articles in the NEON magazine, which show a small portrait of the author with a descriptive subline, a small portrait of the woman was shown next to the text, with the subline “Lina Schrag, 28, got fit for the marathon with miCoach”.
A pretest was conducted for the advertorial (n = 9) and the advertisement (n = 8), com-paring the perceptions of the designed stimuli with an existing advertorial and advertisement for a similar product by Nike. Two separate groups of students, who did not participate in the study, evaluated the stimuli and confirmed them to be suitable for the study.
Procedure
After agreeing to participate, the participants received the paper-and-pencil questionnaire and the magazine excerpt containing the stimulus for the experiment. The questionnaire con-tained a short explanation of the study’s procedure and background, as well as the informed consent. To avoid any bias during the lecture of the magazine excerpt, the participants were told that the study investigated their opinion on the thematic composition of magazines. Two filler questions were used to fit the cover story. After elicitation of some demographic and control variables, the participants were instructed to read and flip through the pages in the same way as they would normally do when reading a magazine. The information that the participants could read articles they found interesting, but that it was also sufficient for this study if they only
scanned the articles they were not interested in was added after pretesting the questionnaire, because participants thought they were expected to read every article in detail. For those partici-pants who were not familiar with the magazine, a short explanation about the frequency of pub-lication, the publisher and the target group was provided. At the end of the questionnaire, the participants were debriefed and thanked for their participation.
Measures
Persuasion knowledge. Since no existing scale to measure persuasion knowledge for advertorials exists, three relevant components were identified in the discussion of the theoretical background. Previous literature was reviewed and scales to measure the three identified persua-sion knowledge components were adopted for the specific context of advertorials. The question-naire (see Appendix C) was translated into German with the help of a professional translator.
Advertising recognition. Recognition of the two advertising formats as advertising was measured similarly to Tutaj and van Reijmersdal (2012). As second question after exposure to the magazine excerpt, participants were asked to indicate whether they had encountered adver-tising in the magazine excerpt (yes = 1, no = 0). This question was preceded by a filler question that fit the cover story and created a transition to the questions related to the two advertising formats (“In your opinion, was there too much advertising in this magazine?”) and followed by a second filler question that was only answered by those who recognized an advertisement (“In your opinion, does the advertisement fit thematically the editorial content in the magazine?”).
Understanding the persuasive and selling intent. Adapted from Tutaj and van
Reijmersdal (2012) and Rozendaal and colleagues (2010), the participants’ understanding of the
advertising format’s persuasive and selling intent was measured with six items. The scale
fol-lowed the second filler question and was placed after the participants were instructed to look again at the page with the stimuli to ensure exposure. For this scale, as for all other scales in the questionnaire, the advertorial was referred to as text and the advertisement was referred to as
advertisement. All six items were measured on a seven-point bipolar scale ranging from 1
(strongly disagree) to 7 (strongly agree). Two items referred to the selling intent: “The aim of this text (advertisement) is…” (1) “… to sell the product” and (2) “… to stimulate the sales of the product”. Two items referred to the persuasive intent: “The aim of this text (advertisement) is…” (3) “… to influence your opinion“ and (4) “… to make people like certain products“. Two
filler items referred to the informational intent of the two advertising formats: “The aim of this text (advertisement) is…” (5) “… to give information about the product” and (6) “…to let peo-ple know more about the product”. The factor analysis yielded that the four items covering the persuasive and selling intent loaded on one factor (EV = 2.84 , R² = .47) and the two filler items covering the informational intent on a second factor (EV = 1.46, R² = .24) as expected. After reliability analysis, a single measure understanding the persuasive and selling intent was creat-ed, using the mean scores of the four items referring to the persuasive and selling intent of the advertorial and advertisement (α = .81, M = 5.24, SD = 1.34).
A study by van Reijmersdal and colleagues (2005) revealed that not all consumers are aware of the term advertorial, but do know what an advertorial is and remember seeing it in magazines after being shown a picture of this ad format. With this finding in mind, a short de-scription of the advertorial format was provided after elicitation of the consumer response varia-bles, followed by the question, whether the participant has ever seen this advertisement format in magazines (yes; no). This question served as a filter question to only ask those participants about perceived appropriateness of the advertorial format that were familiar with the advertising format and had a preformed opinion (n = 66). Participants in the traditional advertisement condi-tion did not receive an explanacondi-tion for the advertisement format and no filter quescondi-tion.
Perceived appropriateness. Perceived appropriateness of the advertising format was
then measured with the two items adapted by van Reijmersdal and colleagues (2005) who termed this construct as acceptance in their study: “Mentioning brands in advertorials (print advertisements) is acceptable” and ”Mentioning brands in advertorials (print advertisements) is misleading”. The two items were measured on seven-point bipolar scales ranging from 1 (strongly disagree) to 7 (strongly agree). The values of the second item were inversed for the data analysis, so that a high value represented a positive perception in both items. After reliabil-ity analysis (r = .66, p < .001), the two items were averaged to create a single scale of perceived
appropriateness (M = 4.6, SD = 1.63).
Consumer response variables. Following the scale understanding the persuasive and
selling intent, the consumer response variables product attitude and ad attitude were measured.
Product attitude. Attitude towards the advertised product was measured through four items on a seven-point bipolar scale based on Bruner’s (2009) attitude toward the product and
product miCoach with four items: “is functional (7) - is not functional (1)”, “is innovative (7) - is not innovative (1)” and “can help me to improve my sport performance (7) - cannot help me improve my sport performance (1)”. Finally participants were asked to indicate an overall eval-uation of miCoach measured with “very positive (7) - very negative (1)”. After reliability analy-sis, a single scale for product attitude was created, using the mean scores of the four items (α = .73, M = 4.49, SD = 1.15).
Ad attitude. Attitude towards the ad, also termed as perceived advertising value (Tutaj &
van Reijmersdal, 2012) or appreciation (van Reijmersdal et al., 2005) in former studies, was measured with three items covering the three beliefs informational, amusement and irritation
value, adapted from van Reijmersdal and colleagues (2005). Following the scale for product
attitude, participants were asked to evaluate the text or advertisement, by indicating how strong-ly they agreed with the statements covering the three beliefs. The seven-point bipolar scales ranged from 1 (strongly disagree) to 7 (strongly agree). Informational value was measured with the statement “It gives me useful information about a product”, amusement value with “I like to read it” and irritation value with “There are too many texts (advertisements) like this in maga-zines”. The values of the item for the belief irritation value were inversed, so that high values for all three items indicated a positive ad attitude. Although reliability analysis yielded a Cronbach’s alpha (.55) that was not as high as hoped, it was satisfying for the purpose of this study. A single measure of ad attitude was created, using the mean scores of all three beliefs (M = 3.49; SD = 1.23).
Recall. The consumer response variables recall of the brand and product were measured after the question whether the participant had encountered advertising in the magazine excerpt (advertising recognition). Those participants who had answered this question with yes answered two subquestions eliciting unaided brand and product recall, similarly to Gupta and Lord (1998). First, “Can you remember the brand that was advertised? Please indicate the brand name if you can remember it.” (1 = Adidas mentioned, 0 = Adidas not mentioned) and second, “Can you remember the product which was advertised? Please indicate the name of the product or the product category if you can remember it.” (2 = miCoach or mi mentioned, 1 = product category mentioned, 0 = nothing or wrongly mentioned). To facilitate the data analysis, this variable was transformed in a dichotomous variable with 1 = product recall (name or category mentioned) and 0 = no recall (nothing or wrongly mentioned).
Additional Factors. Advertising skepticism and involvement.
Advertising skepticism. To elicit the participants’ general skepticism towards
advertis-ing, the well-established seven-point bipolar scale to measure consumer skepticism toward ad-vertising (SKEP) by Obermiller and Spangenberg (1998) was deemed to be appropriate and for the purpose of this study reduced to four items. The four items covered the scale’s aspects of
truthfulness, believability and informational value of advertising in general. The four items
were: “We can depend on getting the truth from most advertising”, “Advertising’s aim is to in-form the consumer”, “Advertising is a reliable source of inin-formation about the quality and per-formance of products” and “In general, advertising presents a true picture of the product being advertised” with answers ranging from 1 (strongly disagree) to 7 (strongly agree). To facilitate the interpretation of the data, the scale was reversed for all four items, thus higher scores in this scale mean being more skeptical. After a first reliability analysis, the item “Advertising’s aim is to inform the consumer” was deleted. A single measure advertising skepticism was then created, using the mean scores of the three remaining items. Reliability analysis yielded a Cronbach’s alpha that was acceptable for the purpose of this study (α = .65, M = 5.93, SD = .93).
Involvement. The personal involvement inventory (PII) from Zaichkowsky (1985, 1994) was used to measure the participants’ involvement concerning the topic running and the product category technological sports gadgets. For the purpose of this study, the original scale was re-duced to one item for topic involvement and to two items for product involvement. Topic
in-volvement was measured with the question “How important is running/jogging to you?” on a
seven-point bipolar scale with answers ranging from “very important (7)” to “not important at all (1)” (M = 4.24, SD = 1.99). Product involvement was measured with the question “What is your stance on technological gadgets such as watches that monitor your heart rate and pulse and thereby help you to optimize your training?” on a seven-point bipolar scale with the items “For me, technological gadgets like this are…”: “very relevant (7) - not relevant at all (1)” and “very important (7) - not important at all (1)”. Both items were significantly correlated (r =.89, p < .001), therefore a single measure for product involvement was created using their mean scores (M = 2.71, SD = 1.74).
Control Variables. Similarly to the study by Tutaj and van Reijmersdal (2012), investi-gating the role of persuasion knowledge for online advertising formats, the following control variables were included in the questionnaire. Before exposure to the stimuli, the basic
de-mographics age (in years), gender (male = 1, female = 2) and whether the participant is a
stu-dent (yes = 1, no = 2) were elicited. Subsequently, familiarity with the magazine (yes = 1, no =
2) and frequency of magazine use with the response categories (1) “Never (I have never read NEON before)”, (2) “Occasionally (I do not read NEON regularly, but read it once in a while)”, (3) “Regularly (I read NEON on a regular basis, e.g. every issue or every other issue)” were measured. Prior brand attitude was elicited following the scale for product attitude. Adapted from Wei and colleagues (2008), brand attitude was measured with two items on a seven-point bipolar scale: “likeable (7) - dislikeable (1)” and “good (7) - bad (1)” The two items (r = .76, p < .001) were averaged to one single scale for brand attitude (M = 5.26, SD = 1.25).
The last questions of the questionnaire comprised the check for brand familiarity with the brand Adidas (1 = familiar with Adidas, 2 = unfamiliar with Adidas), product possession of the Adidas miCoach (yes = 1, no = 2) and brand experience (possession of Adidas sports cloth-ing and other products) (yes = 1, no = 2). Finally, attention paid to the magazine was measured by asking the participants how attentively they looked at the excerpt on a seven-point bipolar scale with “very attentive (7) - not attentive (1)” (M = 4.31, SD = 1.67). Lastly, participants were asked whether they could imagine what the goal of the study was. The open answers were coded with “correctly indicated (1)” and “not correctly indicated (2)”.
Results Design Checks
Statistics showed that the two conditions did not differ significantly with respect to the demographic variables age (F(1, 132) = 1.64, p = .203) and gender (χ2 (1, N = 134) = .45, p = .503), nor with respect to the control variables familiarity with the magazine (χ2 (1, N = 134) = 1.45, p = .221), frequency of magazine use (z = 1.54, p = .551), brand experience (χ2Yates (1, N = 134) = 3.28, p = .070), brand attitude (F(1, 134) = .58, p = .449) and attention paid to the maga-zine (F(1, 134) = .29, p = .592). All 134 participants were students, were familiar with the brand Adidas and none of the participants owned an Adidas miCoach. These results imply that differ-ences in the groups regarding the measured variables (persuasion knowledge and consumer re-sponse variables) cannot be attributed to differences in these background and control variables, but are likely the result of the manipulation. 14 out of the 134 participants guessed correctly that
the aim of the study was to investigate whether the advertorial or advertisement were perceived as advertising and how this affected the impact of the advertising format. To investigate whether the participants’ awareness of the research topic influenced the outcomes of the three measured consumer variables and three persuasion knowledge components, t-test and Chi square analyses were conducted for all variables (see Appendix D, Table D1). No significant difference between those participants who were aware and those who were not aware of the research topic could be found for any of the variables. It can therefore be assumed that awareness of the research topic did not bias the participants’ responses in this study. Thus, all 134 participants were included in the data analysis.
Effects of advertising format
Effects of advertising format on consumer response variables. To preliminary exam-ine the effects of the ad format on the consumer response variables ad attitude and product atti-tude, two independent samples t-tests were performed with ad attitude and product attitude as test variables. Both test variables were normally distributed. The t-test yielded a significant dif-ference between the respondents’ ad attitude for the advertorial and for the traditional print ad-vertisement (t(132) = 2.39, p = .018, 95% CI [.09, .91], η2 = .04). The mean scores depicted in Table 1 show that exposure to the advertorial resulted in a more positive ad attitude compared to exposure to the advertisement. The effect size was small with .04 (Cohen, 1988), indicating that 4% of the variance in the variable ad attitude was explained by the advertising format.
For the test variable product attitude, in contrast, exposure to the two advertising formats did not lead to significant differences concerning the respondents’ attitude towards the product (see Table 1), t (132) = .83, p = .409, 95% CI [-.23, .56].
Finally, the effect of the advertising format on the consumer response variables brand
recall and product recall was tested using chi square analyses for independence. Only those
participants who recognized an advertisement in the magazine excerpt were included in the analysis to test the effect of the advertising format only, without interference of the persuasion knowledge component advertising recognition. The analysis for brand recall did not yield a sig-nificant difference between the two ad formats concerning the percentage of respondents who recalled the brand name correctly when exposed to an advertorial (82.4%) compared to those exposed to the advertisement (87.5%), (χ2Yate(1, N = 57) = .009, p = .924). The same counted for
product recall. No significant difference was found between the two ad formats concerning the percentage of respondents who recalled the product name or product category correctly adver-tised in the advertorial (41.2%) and the traditional advertisement (22.5%) (χ2Yate(1, N = 57) = 1.24, p = .266).
Table 1
Consumer response variables and persuasion knowledge components as a function of ad format
Variables Ad format
Advertorial Advertisement
Consumer response variables:
Ad attitude1 3.74 (1.13)* 3.24 (1.28)* Product attitude2 4.57 (1.15) 4.40 (1.15) Brand recall3 .82 (.39)a .88 (.33)b Product recall4 .41 (.51)a .23 (.42)b
Persuasion knowledge components:
Advertising recognition5 .25 (.44)*** .61 (.49)*** Understanding the persuasive and selling intent6 5.53 (1.21)** 4.94 (1.41)** Perceived appropriateness7 3.97 (1.67)c*** 5.23 (1.33)***
Note: n = 68, except if indicated with a n = 17; b n = 40; c n = 66. Mean scores are provided with standard deviations within parenthe-ses. Means in the same row differ significantly from each other, if indicated with: *p < .05, **p < .01, ***p < .001
1,2
Scale ranging from 1 (negative attitude) to 7 (positive attitude), 3,4 Dichotomous variable with 0 (no recall) and 1 (recall),
5
Dichotomous variable with 0 (no recognition) and 1 (recognition), 6Scale ranging from 1 (no understanding) to 7 (very good
under-standing), 7Scale ranging from 1 (negative perception) to 7 (positive perception)
Effects of advertising format on persuasion knowledge components. To answer re-search question 1, the participants’ level of persuasion knowledge for advertorials was investi-gated in comparison with the level of persuasion knowledge for traditional print advertising.
For the persuasion knowledge component advertising recognition, the chi square analy-sis showed a significant difference between the proportion of participants who were exposed to an advertorial format and who recognized the format as advertising (25%), compared to those
(χ2(1, N = 134) = 17.37, p < .001). A logistic regression analysis with the two ad formats as cat-egorical predictor was performed to investigate the impact of the ad format on the persuasion knowledge component advertising recognition. The analysis yielded that the ad format had a significant effect on recognition (-2LL = 164.98, Nagelkerke R² = .17, Chi²(1) =17.79, p < .0005), explaining around 17 % of the variance in recognition of the ad format as advertising. The mean scores in Table 1 show that advertising recognition, i.e. the distinction between the magazine’s commercial and editorial content, was significantly higher for the traditional adver-tisement compared to the advertorial.
Secondly, two independent sample t-tests were conducted to investigate the effect of the ad format on the two persuasion knowledge components understanding the persuasive and
sell-ing intent and perceived appropriateness. The results of the analysis with the dimension
under-standing of persuasive and selling intent as test variable showed significant higher scores for the advertorial compared to the traditional advertisement (see Table 1) (t(132) = 2.60, p = .010, , 95% CI [.14, 1.05], η2 = .05). In this study, the advertorial induced a higher level of understand-ing of persuasive and sellunderstand-ing intent for the commercial message, compared to the traditional advertisement. The effect size however was small with .05 (Cohen, 1988). 5% of the variance in the variable understanding of persuasive and selling intent is explained by the ad format.
The analysis for the persuasion knowledge component perceived appropriateness yielded significantly higher scores for the advertisement than for the advertorial (t(123.7) = -4.79, p < .001, 95% CI [-1.78, -.74], η2 = .15). Thus, the participants in this study perceive traditional print advertisements to be more appropriate than advertorials. The effect size of .15 indicates a large effect of advertising format on perceived appropriateness (Cohen, 1988), meaning that 15% of the variance in this variable can be explained by the ad format.
Taken together, the analyses found differing levels of persuasion knowledge for the ad-vertorial and traditional print advertisement in this study along all three persuasion knowledge components. Compared to the participants’ persuasion knowledge for the traditional advertise-ment, recognition as advertising was lower for advertorials, understanding the persuasive and selling intent of the message higher and perceived appropriateness of the ad format more nega-tive.
The relations between advertising format, persuasion knowledge and consumer response variables
To answer research question 2, the four-step mediation analysis as proposed by Baron and Ken-ny (1986) was used to investigate whether persuasion knowledge for advertorials could be an underlying mechanism for the identified effect of exposure to an advertorial or traditional tisement and the consumer response variable ad attitude. For the analyses, ad format (1 = adver-torial; 0 = advertisement) was established as predictor, ad attitude as the outcome variable and the two persuasion knowledge components perceived appropriateness and understanding the
persuasive and selling intent as hypothesized mediators based on previous theoretical
considera-tions.
For the test of hypothesis H2a, the mediation analysis visualized in Figure 3 was run with perceived appropriateness as hypothesized mediator (see also Appendix D, Table D2). The predictor ad format (as independent variable) and outcome variable ad attitude (as dependent variable) were regressed in the first step. The analysis showed that the unstandardized regression coefficient between the two variables was statistically significant (B = .498, p = .018). Path c was therefore significant and the requirement for mediation in Step 1 was met. The second re-gression analyses included ad format (as independent variable) and perceived appropriateness (as dependent variable) for Step 2 of the mediation analysis. The unstandardized regression co-efficient associated with this relation was also significant (B = -1.26, p < .001), hence the re-quirement for Step 2 was met and path a significant.
.23** b Perceived Appropriateness Ad Format .20* Ad Attitude c Ad Format -.39*** Ad Attitude a .34*** c‘
Figure 3. Standardized regression coefficients yielded by four step mediation analysis to test
Whether perceived appropriateness (as independent variable) was related to ad attitude (as de-pendent variable) was tested in a third regression analysis, which yielded a significant coeffi-cient associated with this relation (B = .17, p = .009). Thus, path b was significant and the condi-tion of Step 3 met. The fourth regression analysis, run with all three variables, provided an esti-mate of the relation between ad format and ad attitude, when controlling for perceived appropri-ateness. The results yielded a coefficient which was still highly significant (B = .85, p < .001).
According to the conditions of Baron and Kenny’s (1986) mediation analysis, the per-suasion knowledge component perceived appropriateness can therefore not be assumed to me-diate the relation between ad format and ad attitude. For mediation, the magnitude of the rela-tion between ad format and ad attitude had to be reduced when perceived appropriateness was introduced to the model. In this model however, introducing perceived appropriateness in-creased the magnitude of the relation between ad format and ad attitude, which indicates sup-pression (MacKinnon, Krull, & Lockwood, 2000). Supsup-pression can be defined as “a variable which increases the predictive validity of another variable (or set of variables) by its inclusion in a regression equation. This variable is a suppressor only for those variables whose regression weights are increased.” (Conger, 1974, p. 36-37). Predictive validity in this case is assessed by the magnitude of the regression coefficient (MacKinnon et al., 2000). The Sobel test used to investigate the significance of the suppressor variable perceived appropriateness (McKinnon et al., 2000) yielded that the effect of the advertising format on ad attitude significantly increased when perceived appropriateness was added to the equation (z = 3.07; p = .002), leading to the support of hypothesis H2a.
The notion that perceived appropriateness served as a suppressor, specifically as a recip-rocal suppressor, in this study can be explained by the opposing signs of path c and path a. Thus, the model in this study is a so-called inconsistent mediation model (MacKinnon et al., 2000). The direct effect of the ad format on ad attitude was positive, since the advertorial elicited a more positive ad attitude compared to the traditional advertisement. However, the advertorial was related to a more negative perceived appropriateness. A positive perception in appropriate-ness in turn was related to a more positive ad attitude. Thus, when perceived appropriateappropriate-ness is included in the model as a third variable, it leads to an increase of the magnitude of the relation between ad format and ad attitude. This effect occurs because the variables advertising format and perceived appropriateness share irrelevant information on ad attitude, which is suppressed
