Creative media effects in a real life setting:
A field experiment examining the effects of creative media advertising on attitudinal responses and behavioral intentions through perceived humor, perceived value and perceived
persuasive intent.
Master’s Thesis
Graduate School of Communication
Research Master’s programme Communication Science
Supervisor: mw. dr. M. L. Fransen Date of Completion: 02/02/2017
Paola Remmelswaal
2 Abstract
Despite that the effectiveness of creative media advertising has been subject to an increasing body of research, little is known about its effectiveness in a real life setting. For this reason, the current study examined if creative media advertising is more effective than traditional advertising. To explain why creative media advertising would be more effective than traditional advertising, three mediators were introduced: perceived humor, perceived value and perceived persuasive intent. Drawing from Schema Theory, it is expected that creative media advertisements are perceived as more humorous and valuable than traditional advertisements, which could subsequently lead to more positive attitudinal responses and behavioral intentions. Additionally, this study examined whether the use of creative media advertising leads to a lower or higher perceived persuasive intent and, subsequently, how this affects attitudinal responses and behavioral intentions. Results of a real life experiment (N = 133) showed both direct and indirect effects of advertising type on attitudinal responses and behavioral intentions. More specifically, creative media advertising (versus traditional advertising) enhanced brand attitude, ad attitude, word-of-mouth (WOM) intention and purchase intention. These effects were positively mediated by perceived humor and perceived value, though, not by perceived persuasive intent. An additional serial mediation analysis showed that creative media advertising was perceived as more humorous and subsequently perceived as more valuable, resulting in positive attitudinal responses and behavioral
intentions. Hence, results of this study fill two gaps in the current literature by underlining the effectiveness of creative media advertising (versus traditional advertising) in a real life setting and, accordingly, by clarifying which processes can account for this effectiveness. This knowledge is valuable for practitioners as it indicates the exceptional outcomes in terms of consumer responses when investing in the creation of creative media advertisements.
Keywords: Creative media advertising, non-traditional advertising, guerilla advertising, perceived humor, perceived value, perceived persuasive intent.
3 Introduction
Recently, Xylifresh, a Dutch chewing gum brand, displayed their advertisement on mouth guards belonging to players of a big field hockey club in Amsterdam. Besides gaining more awareness by advertising in a playful way, the brand wants to emphasize that it is important to protect your teeth (Adformatie, 2015). As mouth guards are not usually used as a medium for advertising, consumers might be surprised and start to think about the link
between Xylifresh and mouth guards. Subsequently, consumers understand the association: ‘Xylifresh wants to protect your teeth and so does a mouth guard!’ This advertising campaign
can be defined as a form of creative media advertising.
Nowadays, advertisers are struggling to reach their audience due to the so-called advertising clutter. Advertising clutter is the large amount of excess of commercial messages (Rauwers & Van Noort, 2015). A consumer is exposed to approximately 3000 advertising messages a day, which all try to grasp the consumer’s attention (Jurca & Madberger, 2015;
Kim, Bhargava, & Ramaswami, 2001). This overload may cause irritation and as a result, consumers’ attitudes towards advertising tend to become more negative over time (Speck &
Elliot, 1997). In addition, the persuasive power of traditional forms of advertising to attract consumers’ attention has diminished (Ay, Aytekin, & Nardali, 2010; Darke & Ritchie, 2007).
To overcome this, advertisers are exploring new ways to break through the clutter. A recent development shows that advertisers attempt to do this by making use of creative advertisements (Dahlén & Edenius, 2007). There are several ways to define these types of advertisements. Guerilla advertising can be seen as an overarching term for all unconventional creative advertisements that try to draw attention by evoking surprise and diffusion effects amongst a large number of recipients (Hutter & Hoffmann, 2014; Levinson, 2007). Currently, scholars are increasingly using the term ambient communication to refer to unconventional advertising, such as an advertisement displayed on a bus in which a big snake
4 curls around the bus (Gambetti & Schultz, 2015). Ambient communication puts emphasis on the use of elements in the environment to convey messages to cause customer engagement (Gambetti, 2010).
Another form of creative advertising, creative media advertising, focuses merely on the creative use of a medium and its associative overlap with the brand. Hence, this concept differs from the previously mentioned concepts as it highlights certain features of a brand with the help of a novel and unconventional creative medium (Eelen, Rauwers, Voorveld, & Van Noort, 2016). Whereas guerilla and ambient advertising sometimes use traditional media for distributing advertisements (e.g. billboards, buses, commercials), creative media
advertising requires a new and unconventional medium to advertise on (e.g. banana peels, sheep, benches) (Eelen et al., 2016). Thus, an advertisement can be considered as a creative media advertisement when it is promoted on a media vehicle that is (1) not already used for advertising and (2) has an associative overlap with the advertised brand (Dahlén, 2005). The previous example of Xylifresh is a form of creative media advertising since mouth guards are not previously used as a medium for advertising. Additionally, mouth guards have an
associative overlap with the promoted brand because a mouth guard protects your teeth and so does Xylifresh chewing gum.
The effectiveness of creative media advertising has been subject to an increasing body of research. The majority of these research projects showed that creative media advertising exceeds traditional advertising in improving attitudinal responses and behavioral intentions (Dahlén, 2005; Dahlén, Friberg, & Nilsson, 2009; Dahlén, Granlund, & Grenros, 2009). This study contributes to the current knowledge on creative media advertising effects by focusing on two focal points that were not fully addressed in earlier research. First, this study aims to examine several underlying processes that could explain the positive effects of creative media
5 advertising: perceived humor, perceived value, and perceived persuasive intent. Just a handful of scholars examined the mediating roles of these processes, revealing inconsistent results.
Second, almost all scholars examined these underlying processes in a hypothetical setting instead of a real life setting. In a hypothetical setting, consumers are unable to actually experience the creative medium in its natural context, while this natural context is key in examining creative media advertising effectiveness (Rauwers & Van Noort, 2015). In a hypothetical setting, creative media advertisements are illustrated in a traditional way (e.g. a picture), thereby blocking any potential cues stimulating ad processing (Hutter, 2015; Rauwers, Van Noort, & Dahlén, 2017). A real life setting is thus crucial to convey a real experience to consumers. A real experience triggers consumers to process the advertisement more in depth and allows consumers to use environmental cues to figure out the association between the medium and the brand (Rauwers & Van Noort, 2015). The second aim of this paper is therefore to confirm and extend previous results in a real life setting. In accordance with prior reasoning, the following research question is posed:
RQ: To what extent are the effects of creative media advertising versus traditional advertising, in a real life setting, mediated by perceived humor, perceived value and perceived persuasive intent?
Thus the goal of this study is to examine whether or not creative media advertising is more effective than traditional advertising in a real life setting and to scrutinize whether one or more of the three proposed mediators (e.g. perceived humor, perceived value and perceived persuasive intent) can account for this effect. This study is theoretically relevant as it
incorporates mediators that have yet not been tested in a real life setting (perceived value and perceived persuasive intent) or lead to contradictory results (perceived humor). Several
6 scholars (e.g., Dahlén, 2005; Hutter, 2015; Rauwers & Van Noort, 2015) have therefore called to test these processes in a real life setting. The results of this study do not only cover gaps in current literature, but also provide a guideline for practitioners in marketing
communication by showing which underlying processes are important to take into account while developing an effective marketing communication message.
Theoretical background
The effects of creative media advertising on attitudinal responses and behavioral intentions Over the last ten years, there is a growing body of research showing the effectiveness of creative media advertising (Eelen et al., 2016). More specifically, creative media
advertising repeatedly proved to be more effective than traditional advertising in terms of both attitudinal responses and behavioral intentions, such as brand attitude, purchase intention and word-of-mouth (WOM) intention (Dahlén, 2005; Hutter, 2015; Hutter & Hoffmann, 2014). A creative media choice makes an ad placement appear more relevant than a
traditional advertisement due to an overlap in associations between the medium and the brand (Dahlén, 2005). Consequently, brand and ad attitudes are perceived as more positive when a brand is advertised in a creative media setting as opposed to a traditional media setting (Dahlén, 2005). Moreover, creative media advertisements hold an element of surprise, which strengthens emotions and fosters conative responses such as purchase and WOM intention (Hutter, 2015; Hutter & Hoffmann, 2014).
Only one study has investigated the effectiveness of creative media advertising in a real life setting (Rauwers et al., 2017). In this study the authors mainly focused on the effects of creative media advertising on attitudinal responses (ad attitude and brand attitude). For this reason, the current study does not only focus on these attitudinal responses, but also takes several behavioral intentions (WOM intention, purchase intention and purchase behavior) into account. Behavioral intentions proved to be positively affected by the use of creative media
7 advertising (Dahlén et al., 2009) and are therefore important to include when examining the effects in a real life setting.
A real life setting is vital to catch real feelings of consumers and complementary environmental effects (e.g. interaction with the context of a creative media advertisement) (Hutter, 2015). These environmental effects can serve as cues that can be used by consumers to understand the connection between the brand and the medium (Rauwers & Van Noort, 2015). Further, a more realistic test is provided as people encounter a creative media
advertisement in person rather than in a photograph (Dahlén & Edenius, 2007). Even though the effectiveness of creative media advertising on attitudinal responses (ad attitude and brand attitude) has already been tested in a real life setting, it is important to test behavioral
intentions (WOM intention, purchase intention and purchase behavior) in such a setting as well. This way, a more complete chain of consumer effects can be determined. Thus, considering results of earlier research on the effectiveness of creative media advertising and by taking into account abovementioned consumer effects, the following hypothesis is posed:
H1: When a brand is advertised in a creative media choice setting, as opposed to a traditional media setting, more favorable (a) attitudinal responses and (b) behavioral intentions are evoked.
Underlying processes explaining the effectiveness of creative media advertising
There are two important distinctions between traditional advertising and creative media advertising that may explain the effectiveness of creative media advertising. First, the distinction between traditional and nontraditional media is based on novelty and
exceptionalism; creative media advertising makes use of media vehicles that are not previously used for advertising and can therefore be perceived as original (Dahlén, 2005;
8 Rosengren, Modig, & Dahlén, 2015). Second, in contrast to traditional advertising, creative media advertising uses the medium as a contextual cue to deliver the message (Dahlén, 2005). Dahlén (2005) showed that only using a nontraditional medium is insufficient: a contextual fit is necessary to evoke positive communication effects. For example, promotion of Xylifresh on a bin, a new nontraditional medium, would possibly evoke negative or no effects, as the associations of a bin do not match those of Xylifresh. A mouth guard, however, does have a contextual fit as its associations overlap with the ones of Xylifresh.
According to Schema Theory (Friestad & Wright, 1994), the characteristics of creative media advertising can elicit several mediating processes that could explain why creative media advertisements are more effective than traditional advertisements: perceived humor, perceived value and perceived persuasive intent. Schema Theory (Friestad & Wright, 1994) states that a person’s knowledge is stored and organized into several cognitive frameworks called schemas (Roedder & Whitney, 1986; Warlaumont, 1997). More specifically, schemas are knowledge structures that systematically arrange concepts and information stored in memory (Schmidt & Hitchon, 1999). According to Rauwers and Van Noort (2015), people use two schemata to identify and understand the commercial purpose of an advertising message: an advertising schema and a brand schema. The advertising schema refers to a person’s full understanding about advertising, including the ability to recognize and analyze
its purpose (Dahlén & Edenius, 2007; Friestad & Wright, 1994). The brand schema on the other hand, depicts all opinions and thoughts that a person relates to a brand, resulting from contact with brand-related information such as commercial messages, packaging and samples (Dahlén, Rosengren, Törn, & Öhman, 2008).
Both schemata are required to process any type of advertisement. It is easier to process the communicated information when it is congruent with the information that is already stored in memory (Warlaumont, 1997). Schema congruency can be defined as the extent to which
9 new information fits with prior formed expectations about descriptive characteristics of the stimulus (Schmidt & Hitchon, 1999). Thus, information is congruent when it corresponds with prior expectations based on people’s brand and advertising schemata. When people
encounter a traditional advertisement, schemas are congruent, meaning that people recognize both the brand and the advertising format such as a flyer or poster (Rauwers & Van Noort, 2015). Creative media advertisements, however, are placed on queer objects, such as a bench or an elevator (Dahlén, 2005). As people do not identify such an object as a regular
advertising format, only one’s brand schema is activated, while one’s advertising schema
remains inactivated (Rauwers & Van Noort, 2015).
The misalignment between one’s brand schema and one’s advertising schema can also
be referred to as schema incongruity. Schema incongruity occurs when new information contradicts prior expectations based on schematic knowledge (Schmidt & Hitchon, 1999). Consumers expect to encounter new information in the form of commercial messages on regular advertising formats, and not on queer objects (Rauwers & Van Noort, 2015). Due to schema incongruity, individuals are more driven to find or produce new links in their memory (Eelen et al., 2016). For example, consumers do not expect to see the brand logo of Xylifresh on mouth guards. Consumers therefore try to find the link between the brand and the medium, which can subsequently elicit several processes as perceived humor, perceived value and perceived persuasive intent. These processes can possibly explain the effects of creative media advertising on consumers’ attitudinal responses and behavioral intentions.
Underlying processes affecting advertising effects Perceived Humor
The mediating effect of perceived humor in the relationship between creative media advertising (versus traditional advertising) and consumers’ attitudinal responses and
10 behavioral intentions can potentially be explained by the incongruity-resolution theory
(Speck, 1991). According to this theory, joke-processing is an iterative process in which consumers try to solve the “stimulus-schema incongruity” by testing several semantic and logical explanations (Speck, 1991). Creative media advertisements cause schema incongruity, because of their placement on media vehicles that are not previously used for advertising (Rauwers & Van Noort, 2015). Consequently, consumers try to solve this schema incongruity through the earlier mentioned iterative process by identifying the associative overlap between the media vehicle and the advertised brand (Rauwers et al., 2017). When correspondence is achieved and the connection is understood, people experience joy because of the playful confusion and the fact that they mastered that confusion (Speck, 1991; Strick et al., 2013).
Such joy refers to a larger, overarching concept: perceived humor. Although humor is a commonly used concept in the scientific field of communication science, no uniform definition is used (Weinberger & Gulas, 1992). For this reason, the broad definition of Long and Graesser (1988) is used in this study: “Humor is anything done or said, purposely or inadvertently, that is found to be comical or amusing”. Traditional humorous advertisements
proved to engender more positive advertising effects. More specifically, humor does not only elicit positive attitudinal responses, such as brand attitude and ad attitude (Eisend, 2008; Elpers, Mukherjee, & Hoyer, 2004; Zhang & Zinkhan, 1991), but also positive behavioral intentions, like WOM intention, purchase intention and possibly purchase behavior (Eisend, 2008; Keller & Fay, 2012; Mazzarol, Sweeney, & Soutar, 2007; Zhang, 1996).
Yet, the majority of these studies only examined the effects of humor in traditional advertisements. Rauwers and Van Noort (2015) and Rauwers et al. (2017) were the only scholars testing the advertising effects of humor in the context of creative media advertising. Rauwers and Van Noort (2015) did not find a difference in perceived humor between the use of a creative media advertisement or a traditional advertisement. However, in a more recent
11 study, Rauwers et al. (2017) indicated that creative media advertisements are perceived as more humorous, resulting in a more positive brand attitude, ad attitude and a stronger
purchase intention. An explanation for this difference could be that the study of Rauwers and Van Noort (2015) was conducted in a hypothetical setting, whereas the study of Rauwers et al. (2017) took place in a real life setting. Solely showing a picture of a creative media advertisement might be insufficient to motivate participants to process the advertisement in depth (Rauwers & Van Noort, 2015). Subsequently, it is more difficult for participants to find the connection between the brand and the medium and no humorous reaction can be evoked. Hence, due to contradictory results, humor is again taken into consideration in this paper. In accordance with the previous reasoning, the second hypothesis is as follows:
H2: When a brand is advertised in a creative media choice setting, as opposed to a traditional media setting, the advertisement is perceived as more humorous, resulting in more favorable (a) attitudinal responses and (b) behavioral intentions.
Perceived value
The inactivation of one’s advertising schema could further lead to a negotiation
process used to assess the value of a creative media advertisement (Rosengren et al., 2015). Perceived value can, according to Dahlén et al. (2009), be defined as “a cognitive assessment of the value consumers derive from the ad that focuses both on what the advertiser gains (e.g., positive evaluations) and what the consumer receives”. Perceived value is mainly focused on the cognitive effort and time that a consumer offers to attend, comprehend and process an advertisement (Ducoffe, 1995). In exchange for this effort, consumers expect something of value in return (Rosengren et al., 2015).
12 A negotiation process is not necessarily needed when consumers are confronted with traditional media advertisements; consumers already know which valuable exchanges (e.g. entertainment on TV, information in newspapers) they can receive when encountering traditional media advertisements (Dahlén & Edenius, 2007; Rosengren et al., 2015). Thus, in a traditional media context, such valuable exchanges are already linked to consumers’
existing advertising schemas (Rosengren et al., 2015).
Valuable exchanges cannot directly be linked to nontraditional media, because consumers’ advertising schema remain inactivated when encountering a creative media
advertisement (Rosengren et al., 2015). For this reason, consumers start to negotiate whether an advertised message in a nontraditional medium holds equitable exchanges for them. If the medium is indeed appropriate and thus has a contextual fit with the advertised brand,
consumers perceive this as a valuable effort delivered by the company (Rosengren et al., 2015). In return, consumers are willing to attend and comprehend the advertised message (Dahlén et al., 2009). Additionally, creative media advertisements could offer some form of entertainment value due to the clever bridge between the medium and the advertised brand (Dahlén et al., 2009).
This theoretical reasoning is confirmed by several studies. According to these studies, creative media advertisements evoke higher rates of perceived value than traditional media advertisements (Dahlén et al., 2009; Rosengren & Dahlén, 2013; Rosengren et al., 2015). Perceived value proved to be an important antecedent of positive advertising effects focused on traditional advertising (Hansen, Samuelsen & Silseth, 2008). Only two studies (Dahlén et al., 2009; Rosengren et al., 2015) confirmed these results in a creative media advertising context. More specifically, these studies emphasized that perceived value elicits both positive attitudinal responses, like brand attitude and ad attitude (Rosengren et al., 2015) and
13 behavioral intentions, such as WOM intention and purchase intention (Dahlén et al., 2009; Rosengren et al., 2015).
Yet, these studies used photographs rather than real media to draw their conclusions on. Creative media advertisements depicted in photographs are, however, not able to deliver the same experience as creative media advertisements in a real life setting (Dahlén, 2005; Rauwers & Van Noort, 2015). Accordingly, results are expected to be even stronger in a real life setting, subsequently resulting in potentially bigger value exchanges (Dahlén & Edenius, 2007). This study takes into account previous reasoning and results to confirm and extend earlier results in a real life setting. Hence, the third hypothesis is as follows:
H3: When a brand is advertised in a creative media choice setting, as opposed to a traditional media setting, perceived value increases, resulting in more favorable (a) attitudinal responses and (b) behavioral intentions.
Perceived persuasive intent
A third mechanism that could explain the advertising effects of creative media advertising concerns the perceived persuasive intent of advertising, an element of persuasion knowledge (Friestad & Wright, 1995; Eelen et al., 2016). Persuasion knowledge incorporates a consumer’s personal beliefs and knowledge about advertising motives and strategies
(Friestad & Wright, 1995). This cognitive process is described in the Persuasion Knowledge Model (Friestad & Wright, 1994), which focuses on how consumers develop persuasion knowledge and use this knowledge to interpret, evaluate and react to persuasion attempts. The model puts emphasis on the fact that people need to be aware of a persuasion attempt to activate their persuasion knowledge (Boerman, Reijmersdal, & Neijens, 2012). Once people recognize a persuasion attempt, advertising schema are activated (Roedder & Whitney, 1986).
14 Consequently, according to Schema Theory, consumers use their persuasion
knowledge to create an advertising schema to evaluate the persuasive intent of an
advertisement (Friestad & Wright, 1994; Roedder & Whitney, 1986). Persuasive intent can be defined as the degree to which consumers are able to both recognize and comprehend the commercial purpose of an advertisement (Friestad & Wright, 1994). Consumers nowadays have high levels of persuasion knowledge regarding traditional advertisements (De
Pelsmacker & Neijens, 2012). For this reason, advertisers came up with new advertising strategies, such as creative media advertising (Dahlén & Edenius, 2007). As these advertising strategies and motives are generally unclear, consumers hold lower levels of persuasion knowledge towards creative media advertisements (Dahlén & Edenius, 2007). Existing advertising schemas are therefore often incongruent with creative media (Rauwers & Van Noort, 2015). As a result of this, creative media advertisements are to a lesser extent recognized as persuasion attempts (Eelen et al., 2016). Dahlén and Edenius (2009)
correspondingly found that creative media advertisements are less recognized as persuasion attempts than traditional advertisements. However, the latter is not supported by the results of Rauwers and Van Noort (2015), who found no differences in terms of perceived persuasive intent when comparing creative media advertisements and traditional advertisements.
Due to these mixed results, perceived persuasive intent is again taken into account in this study. By testing the effects of perceived persuasive intent in a real life setting, in which consumers can use environmental cues to comprehend the commercial message to a greater extent, this study can point out whether creative media advertisements are less or better recognized as persuasion attempts than traditional advertisements. This clarification is of great importance, as the perceptions and attitudes consumers hold towards an advertisement are vital in predicting positive or negative advertising effects (Matthes, Schemer, & Wirth, 2007; Hsieh, Hsieh, & Tang, 2012). Previous research did show that messages that are less
15 recognized as persuasion attempts elicit more favorable attitudinal responses than messages that are better recognized as persuasion attempts (Matthes et al., 2007). Behavioral intentions could also be affected by the persuasive intent of a message. Consumers use their persuasion knowledge as a defense mechanism to counter persuasive messages. Consequently, the persuasive impact of a message is reduced (Panic, Cauberghe, & De Pelsmacker, 2012; Livingstone & Helsper, 2006). This boomerang effect can exert a negative influence on WOM intentions and result in lower purchase intentions (Hsieh et al., 2012; Wicklund, Slattum & Solomon, 1970). Thus, it is highly relevant to re-examine whether the use of creative media advertisements leads to a lower or higher perceived persuasive intent and, subsequently, how this affects attitudinal responses and behavioral intentions. Due to mixed results, the following research question is posed:
RQ1: Does the use of creative media advertising leads to a higher or lower perceived persuasive intent in comparison to traditional advertising and how does this
subsequently affect consumers’ attitudinal responses and behavioral intentions?
The conceptual model and the related variables are visualized in Figure 1.
16 Method
Design and participants
To test the hypotheses of this study in a real life setting, a field experiment is conducted with a one-factor (advertising type: creative vs. traditional) between-subjects design. Participants were recruited in two different stores of a large supermarket chain (Albert Heijn) in the Netherlands. The stores were located in two cities (Katwijk and Oegstgeest) of which one has a population with a rather low socioeconomic status whereas the other has a population with a high socioeconomic status. The inclusion of two cities rather than one ensures that a more diverse public is taken into account, enlarging the external validity of this study. To prevent that participants were aware of the other experimental condition and therefore started to compare the two advertisements with each other, the implementation of the two experimental conditions was spread over two different days in each store. In one store, 31 consumers participated in the creative media condition and 32 consumers participated in the traditional media condition, whereas in the other store 35 consumers participated in the creative media condition and 35 consumers participated in the traditional media condition.
All participants were at least 18 years old and had not heard of the study on
beforehand. Thus, no participants were excluded from the final sample. In total, 133 (74,4% female) consumers participated in the study. Participants’ age ranged from 22 to 83 years with
an average age of 54 years (SD = 14.38). Most participants’ highest level of education was, according to the Dutch education system, University (22,6%), followed by MBO (20,3%).
Procedure
Depending on which of the two days participants visited one of the stores, participants were assigned to either the creative or the traditional media condition. When entering the supermarket with their shopping cart, participants were unaware of the fact that the
17 advertisement on their shopping cart was part of a study. They were exposed to the
advertisement while shopping. Next, only those participants who chose a shopping cart with a manipulated advertisement were approached by one of the researchers, after checking out, to ask whether they wanted to participate in a study about “advertising in supermarkets”. More
specifically, participants were asked to fill in a 5-minute questionnaire (Appendix C). In accordance with the ethics committee of the University of Amsterdam, participants were shortly briefed about the experiment both verbally and per information sheet. The information sheet mentioned that participants’ anonymity is guaranteed, along with the
possibility to refuse to participate or to terminate their participation and the notion that no harm will come to the participants. Participants agreed to participate and confirmed they were sufficiently briefed about the study by signing the informed consent form. Next, the
questionnaire was handed out to the participant. In the first part of the questionnaire,
participants were asked to answer questions measuring brand attitude, purchase intention and purchase behavior. Second, participants were asked to indicate whether or not they saw the advertisement: 85,1% of the participants did see the traditional advertisement and 97,0% of the participants did see the creative media advertisement. To remind participants of the advertisement, they were exposed once again to the advertisement in the form of a photo, so that they were better able to answer the questions thereafter, measuring ad attitude and WOM intention. After this section, questions with regard to several mediators (i.e. perceived humor, perceived value, perceived persuasive intent) needed to be answered. Third, participants were asked to answer several questions regarding the fit between the medium and the brand, which represented the manipulation check. In the final part of the questionnaire, several control variables and demographic characteristics were measured. After filling out the questionnaire, participants were thanked for their participation, debriefed by the researcher and asked if they wanted to compete for one of the three bags of groceries worth 25 euros.
18 Advertising type
Real advertisements, based on the stimulus material of Rauwers et al. (2017), were created for the brand Autodrop. This Dutch licorice brand was chosen as it is a fairly neutral brand: it is suitable for both women and men of all ages. Below, a more detailed description of each advertising type can be found.
Creative media ad. A creative media advertisement requires two key elements: it has to be promoted on a media vehicle that is not previously used for advertising and it has to have an associative overlap with the advertised brand (Dahlén, 2005). For this reason, a shopping cart was selected as a creative medium in this study, because shopping carts themselves are not previously used for advertising and have an associative overlap with the brand “Autodrop”. This brand contains the word “car” in its brand name. Cars and shopping
carts both have similar features, such as their wheels and, consequently, their ability to drive. The Autodrop advertisement consisted out of three elements which all emphasize the car-connection: a steering wheel with the logo and slogan of Autodrop is mounted on the shopping cart, along with breaking pedals and a dashboard (Appendix B, figure 1).
Traditional ad. A poster was used as traditional medium to carry the advertising message of Autodrop. More specifically, these posters were presented as so-called CartBoards in the back of the shopping cart. The placement of these posters represents the key difference between the stimulus material used in this study and that of Rauwers et al. (2017). In the latter, three big posters were placed in the store, whereas in this study, posters were placed in the shopping carts. This way, exposure time between the two experimental conditions would be equal. In addition, to ensure consistency between the two experimental conditions, the poster also contained the logo and slogan of Autodrop alongside a depiction of a car (Appendix B, figure 2).
19 Measures
Attitudinal variables. Two attitudinal variables were measured in this study: brand
attitude and ad attitude.
Brand attitude was measured by asking participants to indicate their attitude towards the brand on a seven-point semantic differential scale based on five items with bipolar ends: “bad/good”, “unappealing/appealing”, “unfavorable/favorable”, “unpleasant/pleasant” and “unlikable/likable” (Bellman et al., 2011). A principal component analysis with varimax
rotation showed that these five items form a one-dimensional scale: only one component is given with an eigenvalue larger than 1 (EV = 4.26), which explains 85,27% of the variance in the five items. The formed scale is proved to be reliable (α = .96, M = 4.75, SD = 1.51). A higher score on this scale reflects a more positive attitude towards the brand.
Ad attitude was measured by a five-item adaption of the brand attitude scale by Bellman et al., (2011). Respondents were asked to rate their attitude towards the ad on a seven-point semantic differential scale based on five items with bipolar ends: “bad/good”, “unappealing/appealing”, “unfavorable/favorable”, “unpleasant/pleasant” and
“unlikable/likable”. A principal component analysis with varimax rotation showed that these
five items form a one-dimensional scale: only one component is given with an eigenvalue larger than 1 (EV = 4.67), which explains 93,35% of the variance in the five items. The formed scale is proved to be reliable (α = .98, M = 4.58, SD = 1.82). A higher score on this scale reflects a more positive attitude towards the ad.
Behavioral variables. Three behavioral variables were measured in this study:
purchase intention, purchase behavior and WOM intention.
Purchase intention was measured by asking participants to indicate the chance that they would buy a product of Autodrop in the future. Two items retrieved from the scale of Bearden, Lichtenstein and Teel (1984) were measured on a seven-point semantic differential
20 scale with bipolar ends: “unlikely/likely” and “uncertain/certain” Additionally, one item from the scale of Zhang (1996) is added: “ruled out/possible”. A principal component analysis with
varimax rotation showed that these three items form a one-dimensional scale: only one component is given with an eigenvalue larger than 1 (EV = 2.83), which explains 94,39% of the variance in the three items. The formed scale is proved to be reliable (α = .97, M = 4.38, SD = 2.23). The higher the score on this scale, the higher a participants’ purchase intention. Purchase behavior was measured by asking participants the following question: “Did you just buy a product of Autodrop?” (dummy coded: 0 = no, 1 = yes). However, as less than ten participants bought the product (N = 5), purchase behavior is not normally distributed with a skewness of 4.92 (SE = .21) and a kurtosis of 22.52 (SE = .42). For this reason, purchase behavior is not included in further analyses1.
WOM intention was measured by a two-item scale adapted from Derbaix and
Vanhamme (2003). The scale included one item for offline WOM: “The chance I am going to tell people about this advertisement is high”, and one item for online, electronic word-of-mouth (eWOM): “The chance I am going to tell people about this advertisement online via mail or social media is high”. Both items were measured on seven-point scales, ranging from
1 (Totally disagree) to 7 (Totally agree). A principal component analysis with varimax rotation showed that these two items form a one-dimensional scale: only one component is given with an eigenvalue larger than 1 (EV = 1.48), which explains 74,08% of the variance in the two items. The formed scale is proved to be reliable (r = .48, p < .001, M = 2.35, SD = 1.48). A higher score on this scale reflects a higher WOM intention.
Mediators. Three different processes were measured in this study that could serve as
mediators: perceived humor, perceived value, perceived persuasive intent.
1
All other dependent variables and mediators, except for perceived persuasive intent, had a skew or kurtosis within the range of -2 and 2 and were therefore considered as normally distributed.
21 Perceived humor was measured by four items on a seven-point semantic differential scale developed by Chattopadhyay and Basu (1990). Each item referred to the following statement: “I think that the advertisement is…”. The items had the following bipolar ends: “not humoristic/humoristic”, “not playful/playful”, “dull/not dull” and “boring/not boring”. A
principal component analysis with varimax rotation showed that these four items form a one-dimensional scale: only one component is given with an eigenvalue larger than 1 (EV = 3.60), which explains 90,11% of the variance in the four items. The formed scale is proved to be reliable (α = .96, M = 4.55, SD = 1.88). The higher the score on this scale, the more humorous participants evaluate the ad.
Perceived value was measured by three items on a seven-point Likert-scale, ranging from 1 (Totally disagree) to 7 (Totally agree). The items were: “The advertising is of value to me”, “The advertising is interesting” and “The advertising is worth my attention” (Rosengren
et al., 2015). A principal component analysis with varimax rotation showed that these three items form a one-dimensional scale: only one component is given with an eigenvalue larger than 1 (EV = 2.73), which explains 91,04% of the variance in the three items. The formed scale is proved to be reliable (α = .95, M = 3.85, SD = 1.89). The higher the score on this scale, the more valuable participants evaluate the ad.
Perceived persuasive intent was measured by three items, adapted from Dahlen and Edenius (2007), on a seven-point Likert-scale, ranging from 1 (Totally disagree) to 7 (Totally agree). Participants were asked to indicate how much they agreed with the statements
measuring perceived persuasive intent: “The aim of the advertisement is to sell more Autodrop”, “The aim of the advertisement is to make me buy more Autodrop” and “The
Autodrop advertisement had a commercial purpose”. A principal component analysis with varimax rotation showed that these three items form a one-dimensional scale: only one component is given with an eigenvalue larger than 1 (EV = 2.41), which explains 80,38% of
22 the variance in the three items. The formed scale is proved to be reliable (α = .88, M = 6.14, SD = 1.16). The higher the score on this scale, the better an advertisement is recognized as a persuasion attempt.
Manipulation check. The nontraditional medium needed an associative overlap, and
thus a fit, with the advertised brand in order to be classified as a creative media advertisement (Dahlén, 2005). If participants in the creative media condition perceive a fit between the medium and the brand, while participants in the traditional media condition do not perceive a fit between the medium and the brand, the stimulus material is successfully manipulated. To ensure participants could answer the questions representing the manipulation check properly, the fit between another brand and medium was first explained in terms of their connection. The perceived fit was subsequently measured by two items on a seven-point semantic differential scale. The items both had the same statement: “The brand (Autodrop) and the place (a poster/shopping cart)…”, with the following bipolar ends: “do not match/match” and “do not fit well/fit well” (Dahlén, 2005). A principal component analysis with varimax
rotation showed that these two items form a one-dimensional scale: only one component is given with an eigenvalue larger than 1 (EV = 1.85), which explains 92,69% of the variance in the two items. The formed scale is proved to be reliable (r = .85, p < .001, M = 4.07, SD = 2.16). The higher the score on this scale, the higher the perceived fit is. Additionally, participants in the creative media advertisement condition were asked if they could describe the link between the shopping cart and the advertisement. The answers were coded by the researcher to determine whether respondents understood the connection between the brand and the medium (0 = no, 1 = half, 2 = yes).
Control variables. In addition to the demographics, two control variables were
recorded (i.e. general attitude towards advertising and liking of licorice). First, respondents were asked to indicate their general attitude towards advertising, as previous research showed
23 that consumers’ advertising responses could be affected by this variable (Tan & Chia, 2007). Four items were measured on a seven-point semantic differential scale. The four items were all linked to one statement: “Generally, I think advertising is…”, and had the following bipolar ends: “unlikeable/likable”, “unpleasant/pleasant”, “irritating/not irritating” and “not interesting/interesting” (Tan & Chia, 2007). A principal component analysis with varimax
rotation showed that these four items form a one-dimensional scale: only one component is given with an eigenvalue larger than 1 (EV = 3.53), which explains 88,22% of the variance in the four items. The formed scale is proved to be reliable (α = .95, M = 4.17, SD = 1.59). The higher the score on this scale, the more positive a participants’ general attitude towards advertising is. Second, liking of licorice can possibly affect consumers’ advertising responses
towards Autodrop as well. Therefore, the following question was posed: “Do you like licorice?” (dummy coded: 0 = no, 1 = yes).
Results
Control variables and manipulation check
To check whether participants in the two experimental conditions differed from each other on gender, age, level of education, likeability of licorice, and general attitude towards advertising, randomization checks were performed. Chi-square tests showed that there was an equal distribution across conditions for the level of education, χ2 (6) = 1.96, p = .924,
likeability of licorice, χ2 (1) = .03, p = .857, and gender, χ2 (1) = .72, p = .398. Analyses of variance indicated that participants’ age, F(1, 131) = .23, p = .631), and general attitude
towards advertising, F(1, 131) = .10, p = .758), did not significantly differ between groups. Hence, as the two experimental groups did not significantly differ from each other, no control variables were taken into account as covariate in subsequent analyses.
24 Analyses of variance further showed that the manipulation of advertising type is successful, as participants in the creative media condition perceived a stronger fit between the medium and the brand (M = 5.68, SD = 1.23) than participants in the traditional media
condition (M = 2.48, SD = 1.63), F(1, 131) = 163.48, p < .001.
Main analyses
To test whether creative media advertising leads to more favorable attitudinal
responses than traditional advertising, a MANOVA was conducted with ad attitude and brand attitude as dependent variables and advertising type (creative vs. traditional) as the
independent variable. The analysis yielded significant effects (Wilks’ Lamda = .91, F(1,131) = 6.76, p = .002) for advertising type on brand attitude(F(1, 131) = 13.53, p < .001, ηₚ2 = .094) and ad attitude (F(1, 131) = 4.11, p = .045, ηₚ2 = .030). A second MANOVA was conducted to test the effects of advertising type on behavioral intentions. In this model, WOM intention and purchase intention were included as dependent variables and advertising type (creative vs. traditional) as the independent variable. The analysis yielded significant main effects (Wilks’ Lamda = .83, F(1,131) = 13.73, p < .001) for advertising type on WOM intention (F(1, 131) = 23.34, p < .001, ηₚ2 = .151) and purchase intention (F(1, 131) = 14.96, p < .001, ηₚ2 = .102)23. The means in table 2 demonstrate that the creative media
advertisement did indeed elicit a more favorable brand attitude, ad attitude, WOM intention and purchase intention than the traditional advertisement.
2
These analyses were repeated by comparing participants in the traditional condition with those participants in the creative media condition who fully understood the connection between the medium and the brand (72,7%). All results remained significant when participants in the creative condition who did not understand the link at all (21,2%) or only half of it (6,1%), were excluded from the sample, but the effect on ad attitude turned out to be weaker (p = .079). The same results came forth when comprehension was included as a covariate. However, results might be biased as some participants did not have any time to fill in the open question.
3
These analyses were repeated with an additional covariate accounting for the fact whether or not people saw the advertisement. The effects of advertising type on all consumer responses remained the same, however, the effect on ad attitude turned out weaker (p = .100). Nevertheless, it should be noted that only 12 participants did not see the advertisement.
25 Table 2. Average scores on attitudinal and behavioral outcomes for each type of advertising
Dependent variable Advertising type
Creative Traditional
Brand attitude 5.22 (1.25)*** 4.29 (1.61)***
Ad attitude 4.90 (1.72)* 4.27 (1.86)*
WOM intention 2.93 (1.59)*** 1.78 (1.11)***
Purchase intention 5.10 (1.92)*** 3.67 (2.30)***
Notes. Mean with standard deviation between brackets. Significance levels are indicated with p<0.05* p<0.01** p<0.001***.
Mediation effects
To test the predictions that perceived humor, perceived value and perceived persuasive intent mediate the relationship between advertising type and consumers’ attitudinal responses and behavioral intentions, mediation models (Hayes’ model 4) were estimated (Hayes, 2013) for each dependent variable. The mediators were tested separately. For all models, 10.000 bootstrap samples were used with biased corrected confidence intervals of 95%.
Perceived humor
Hypothesis 2 states that creative media advertising is perceived as more humorous than traditional advertising, which subsequently leads to more favorable attitudinal responses and behavioral intentions. These responses were separated and treated as individual dependent variables in the mediation model in PROCESS. The results indicate a significant mediation effect of perceived humor on brand attitude (indirect = .51, boot SE = .19, 95% BCI [.19, .92]), ad attitude (indirect = .84, boot SE = .25, 95% BCI [.36, 1.35]), WOM intention
(indirect = .39, boot SE = .14, 95% BCI [.16, .69]) and purchase intention (indirect = .63, boot SE = .22, 95% BCI [.25, 1.13]). More specifically, as reflected by the means in table 3, the creative media advertisement is perceived as more humorous than the traditional
26 brand attitude (b = .49, SE = .06, p < .001), ad attitude (b = .81, SE = .05, p < .001), a higher WOM intention (b = .37, SE = .06, p < .001) and a higher purchase intention (b = .60, SE = .09, p < .001). These results are in line with the results of the Sobel test (table 4). In line with the predictions, hypothesis 2 is accepted, with the exception of purchase behavior.
Perceived value
To test whether creative media advertising is perceived as more valuable than traditional advertising and subsequently leads to more favorable attitudinal responses and behavioral intentions, as proposed in hypothesis 3, mediation models were run in PROCESS. Again, all attitudinal and behavioral variables were individually tested. The results indicated a significant mediation effect of perceived value on brand attitude (indirect = .48, boot SE = .17, 95% BCI [.19, .86]), ad attitude (indirect = .74, boot SE = .23, 95% BCI [.32, 1.21]), WOM intention (indirect = .43, boot SE = .15, BCI [.16, .75]) and purchase intention (indirect = .65, boot SE = .23, 95% BCI [.25, 1.15]). More specifically, as reflected by the means in table 3, the creative media advertisement is perceived as more valuable than the traditional advertisement (b = 1.03, SE = .32, p < .05). This effect subsequently evoked a more positive brand attitude (b = .47, SE = .06, p < .001), ad attitude (b = .72, SE = .06, p < .001), a higher WOM intention attitude (b = .42, SE = .05, p < .001) and a higher purchase intention (b = .63, SE = .09, p < .001). These results are in line with the results of the Sobel test (table 4).
Therefore, with the exception of purchase behavior, hypothesis 3 is accepted.
Perceived persuasive intent
To answer Research Question 1, mediation models were run in PROCESS to
determine whether perceived persuasive intent mediates the relationship between advertising type and attitudinal responses and behavioral intentions. Individual analyses showed no
27 significant mediation effect of perceived persuasive intent on brand attitude (indirect = -.04, boot SE = .06, 95% BCI .20, .04]), ad attitude (indirect = -.03, boot SE = .05, 95% BCI [-.23, .03]), WOM intention (indirect = -.04, boot SE = .05, 95% BCI [-.19, .03]) or purchase intention (indirect = -.08, boot SE = .10, 95% BCI [-.32, .07])4. These results are in line with the results of the Sobel test (table 4). In other words, the type of medium does not
significantly influence attitudinal responses and behavioral intentions through perceived persuasive intent.
Table 3. Average scores on humor, value and persuasive intent for each type of advertising
Mediator Advertising type
Creative Traditional
Perceived humor 5.07(1.64)* 4.03 (1.96)*
Perceived value 4.37 (1.96)* 3.34 (1.69)*
Perceived persuasive intent 6.05 (1.19) 6.22 (1.13)
Notes. Mean with standard deviation between brackets; Significance levels are indicated with p<0.05* p<0.01** p<0.001***.
Table 4. Z-values (Sobel test)
Dependent variable Mediator
Perceived humor Perceived value Perceived persuasive intent
Z Z Z
Brand attitude 3.10** 3.00** -.77
Ad attitude 3.26** 3.12** -.59
WOM intention 2.93** 2.95** -.74
Purchase intention 2.97** 2.94** -.80
Notes. Each dependent variable was tested separately; Significance levels are indicated with p<0.05* p<0.01** p<0.001***.
Additional analyses
4
This analysis was repeated by testing all three mediators simultaneously for each dependent variable. Results remained the same, except for the indirect effect of advertising type on WOM intention via perceived humor, which turned out to be weaker (p < .10). Specific results can be found in Appendix A table 1.
28 Next, the two mediators that significantly predicted the relationship between
advertising type and consumer responses (i.e. perceived humor and perceived value) were detangled in a sequential mediation analysis. When consumers perceive a creative
advertisement as more humorous, the advertisement could hold value for consumers and subsequently enhance positive consumer responses. This reasoning can be theoretically supported since literature on perceived value states that creative media advertisements often hold some entertainment value (Dahlén et al., 2009). This form of entertainment could be emanated from the joy that occurs when consumers figure out the link between the medium and the advertised brand (Dahlén et al., 2009; Strick et al., 2013). The link is thus perceived as humorous and could therefore hold a certain entertainment value for consumers.
Hence, to test whether perceived humor and perceived value sequentially mediate the influence of advertising type on attitudinal responses and behavioral intentions (PROCESS model: advertising type → perceived humor → perceived value → attitudinal/behavioral variables) serial mediation analyses are conducted for each dependent variable. In this procedure, an ordinary-least squares path analyses is used to estimate the coefficients in the model (Hayes’ model 6). Subsequently, direct and indirect effects of advertising type on
attitudinal responses and behavioral intentions can be determined (Preacher & Hayes, 2008). All attitudinal and behavioral variables were separately tested in serial mediation models using 10.000 bootstrap samples with biased corrected confidence intervals of 95%. The paths for the full serial mediation model are illustrated in figure 2 and the coefficients
corresponding to these paths are provided in table 5.
In line with previous results, the total effects (c1) of advertising type on all included attitudinal and behavioral outcome variables were significant. However, when removing the effect of the mediators, only the direct effects (c1’) of advertising type on WOM intention and purchase intention remained significant. Interestingly, some paths turned insignificant when
29 two mediators, instead of one, were included. The effect of advertising type on perceived value (a2) as well as the effect of perceived humor on WOM intention (b1) were insignificant. Notwithstanding, the analyses yielded overall significant indirect effects. The indirect effect through perceived humor was significant for brand attitude (a1b1 = .32, SE = .14, 95% BCBCI [.10, .66]), ad attitude (a1b1 = .61, SE = .20, 95% BCBCI [.26, 1.05]) and purchase intention (a1b1 = .31, SE = .17, 95% BCBCI [.07, .76]), but was insignificant for WOM intention (a1b1 = .15, SE = .11, 95% BCBCI [-.02, .42]). The indirect effect through the other mediator, perceived value, was insignificant for brand attitude (a1b1 = .07, SE = .07, 95% BCI [-.04, .22]), ad attitude (a1b1 = .09, SE = .08, 95% BCBCI [-.04, .28]), WOM intention (a1b1 = .09, SE = .07, 95% BCBCI [-.05, .24]) and purchase intention (a1b1 = .11, SE = .11, 95% BCBCI [-.06, .41]).
However, serial multiple mediation models showed significant effects for the indirect effect of advertising type on all attitudinal and behavioral outcome variables through both perceived humor and perceived value: brand attitude (a1a3b2 = .19, SE = 08, 95% BCBCI [.07, .41]), ad attitude (a1a3b2 = .24, SE = .10, 95% BCBCI [.09, .50]), WOM intention (a1a3b2 = .24, SE = .12, 95% BCI [.07, .54]) and purchase intention (a1a3b2 = .31, SE =.13 , 95% BCBCI [.11, .64]). Thus, a creative media advertising (compared to traditional
advertising) is perceived as more humorous and is subsequently perceived as more valuable, eliciting more favorable attitudinal responses and behavioral intentions
30 Figure 2. Serial multiple mediation model with perceived humor and perceived value as
proposed mediators of advertising type on several attitudinal and behavioral outcomes.
Table 5. Estimated path coefficients for main effects
Results for mediation effects
Dependent variable Path b SE t p
All outcomes a1 1.05 .31 3.34 .001 a2 .27 .23 1.17 .245 a3 .72 .06 11.55 .000 Brand attitude b1 .31 .08 4.04 .000 b2 .25 .08 3.38 .001 c1 .92 .25 3.68 .000 c1’ .34 .20 1.70 .092 Ad attitude b1 .58 .07 8.67 .000 b2 .31 .07 4.72 .000 c1 .63 .31 2.03 .045 c1’ -.30 .18 -1.68 .095 WOM intention b1 .14 .08 1.85 .067 b2 .31 .08 4.07 .000 c1 1.15 .24 4.83 .000 c1’ .68 .21 3.28 .001 Purchase intention b1 .30 .12 2.49 .014 b2 .42 .12 3.50 .001 c1 1.42 .37 3.87 .000 c1’ .68 .32 2.14 .035
31 General conclusion and discussion
This study aimed to fill two gaps in the current literature by 1) examining the
effectiveness of creative media advertising versus traditional advertising in a real life setting, and 2) by clarifying the underlying processes (i.e. perceived humor, perceived value and perceived persuasive intent) accounting for the effects of advertising type on attitudinal responses and behavioral intentions. First, results of the field experiment showed that creative media advertising, compared to traditional advertising, leads to a more favorable brand
attitude, ad attitude and a higher purchase intention and WOM intention. However, against the expectation, no effect was found for purchase behavior. This is probably due to the fact that, even though Autodrop is a well-known licorice brand in the Netherlands, only 5 participants bought a product of Autodrop. Samples smaller than 20 per cell are simply not powerful enough to detect most effects (Simmons, Nelson & Simohnsohn, 2011). Future research should therefore incorporate brands that are bought more frequently or daily, such as milk, energy drinks and coke (Het Parool, 2016). In addition, products could be placed more prominently in the store, for example by offering the same product again at the counter to encourage consumers to buy the product.
Second, to explain the effectiveness of creative media advertising, perceived humor, perceived value and perceived persuasive intent were considered as mediators. The results show that two of these processes elicited more positive advertising effects when consumers were exposed to a creative media advertisement. In line with the expectations, creative media advertisements are perceived as more humorous and more valuable than traditional
advertisements, resulting in more favorable attitudinal responses and behavioral intentions. However, no effect was found for the third mediator: perceived persuasive intent. The
remaining significant mediators, perceived humor and perceived value, were further detangled in a multiple mediation model. The model indicated that perceived humor and perceived value
32 sequentially mediate the influence of advertising type on attitudinal responses and behavioral intentions. More precisely, creative media advertising, compared to traditional advertising, is perceived as more humorous and is subsequently perceived as more valuable, resulting in favorable attitudinal responses and behavioral intentions.
The results of this study add to the literature in several ways. To begin with, this study is one of the first to examine the effectiveness of creative media advertising in a real life setting and contributes to the current body of research by indicating if and why, creative media advertising is more effective than traditional advertising. Previous research on the effects of creative media advertising, conducted in a hypothetical setting, has demonstrated that creative advertising is more effective than traditional advertising in terms of both attitudinal responses and behavioral intentions (Dahlén, 2005; Hutter, 2015; Hutter & Hoffmann, 2014). Rauwers et al. (2017) were the only scholars who showed that primarily attitudinal responses are enhanced when creative media advertising is used. The results presented in this study add to these findings by demonstrating that not only attitudinal responses, but behavioral intentions are positively affected by creative media advertising as well. Specifically, when an advertisement is placed in a creative media context instead of a traditional media context, consumers do not only like a brand or advertisement more, but are also more inclined to talk about the advertisement and buy the advertised brand. Hence, a more complete chain of attitudinal and behavioral effects is confirmed in this study.
Furthermore, two processes (perceived humor and perceived value) demonstrated why creative media advertising is more effective than traditional advertising. Whereas Rauwers and Van Noort (2015) indicated that creative media advertisements are not perceived as more humorous than traditional advertisements, this study disputes this finding: creative media advertisements are perceived as more humorous and subsequently even lead to more positive consumer responses. The findings of this study are in line with those of Rauwers et al. (2017)
33 who also tested the effects of creative media advertising in a real life setting. Hence, it seems that a real life setting offers consumers indeed the opportunity to elaborate more extensively on the connection between the medium and the brand. Several scholars argued that a real life setting would be more appropriate than a hypothetical setting to test the effectiveness of creative media advertising, however, no empirical evidence is available to support this notion (Dahlén & Edenius, 2007; Rauwers & Van Noort, 2015). Future research could therefore examine the difference in creative media advertising effects by directly comparing the results of a field experiment with those of a scenario-based experiment. Results could indicate whether consumers are better able to find the connection between the medium brand in a real life setting as opposed to a hypothetical setting (e.g. a photo). The outcomes could
subsequently determine which method is more convenient to use in future research on creative media advertising.
A second mediator that explains the effectiveness of creative media advertising is perceived value. Earlier work on perceived value regarding traditional advertising
demonstrated that valuable advertisements elicit positive advertising effects (Hansen et al., 2008). Notwithstanding, findings of this study indicate that creative media advertisements also hold valuable exchanges for consumers, resulting in positive attitudinal responses and behavioral intentions. Both Dahlén et al. (2009) and Rosengren et al. (2015) found the same results, however, in a hypothetical setting. This study is the first to demonstrate the mediation effect of perceived value in a real life setting. Even in a natural setting, which incorporates many variables that may influence the advertising processing (e.g. crowdedness of the
supermarket), results from previous studies hold up. In sum, the findings partially support the purport of Schema Theory. Consumers who solve their schema incongruity by finding the connection between the brand and the medium, perceived the creative media advertisement as more humorous and more valuable. However, as only differences between traditional and
34 creative media advertising have been examined, this study cannot draw any conclusions on the assumption that creative media advertising only works for those who solve the incongruity (Eelen et al., 2016). Future research could deliver more support for Schema Theory, by testing whether or not processes as perceived value and perceived humor are only activated when consumers solve the incongruity.
In addition, Schema Theory could not explain the effectiveness of the third mediator tested in this study, perceived persuasive intent, as it did not explain the effects of creative media advertising on attitudinal responses and behavioral intentions. This result contradicts the findings of Dahlén and Edenius (2009), who found that creative media advertisements are less recognized as persuasion attempts than traditional advertisements, but it is in accordance with the findings of Rauwers and Van Noort (2015), who found no differences in terms of perceived persuasive intent when comparing creative media advertisements and traditional advertisements. These conflicting findings, and the insignificant result in this study, could be explained by the fact that perceived persuasive intent is often measured by using only one aspect of persuasion knowledge. In general, advertising theories merely focus on the cognitive aspect of persuasion knowledge rather than on the attitudinal aspect of persuasion knowledge (Rozendaal, Lapierre, Van Reijmersdal & Buijzen, 2011). Conceptual persuasion knowledge is the cognitive element that comprises the recognition of commercial messages, its source and public, and the apprehension of the persuasive intent, selling intent and tactics (Rozendaal et al., 2011). This and previous studies mainly measured perceived persuasive intent by focusing on conceptual persuasion knowledge (e.g., Dahlén & Edenius, 2007; Rauwers & Van Noort, 2015). Attitudinal persuasion knowledge on the other hand, states that attitudinal mechanisms can also be used to cope with persuasion attempts. It embraces mainly critical attitudes, including disliking and skepticism, towards a persuasion attempt (Roozendaal et al., 2011). The use of creative media advertising may affect not only conceptual persuasion
35 knowledge, but also attitudinal persuasion knowledge. For example, conceptual persuasion knowledge can be high, meaning that consumers recognize that they are being persuaded, while attitudinal persuasion knowledge is low, meaning that consumers do not mind being persuaded via creative media advertising. This would also explain why in some cases perceived persuasive intent does not negatively affect brand attitude. Hence, future research could elucidate contradictory results by measuring perceived persuasive intent in two ways: on a cognitive and an attitudinal level.
Furthermore, future research could draw special attention to the development of the stimulus material. In this study, a pink candy bag was illustrated in the traditional
advertisement, but not in the creative media advertisement. Due to the bright pink color of the bag, the traditional advertisement could stand out to consumers, which could elicit more brand awareness. A higher brand awareness could subsequently lead to a higher purchase intention (Oh, 2000). However, even though the traditional advertisement contained an additional prominent feature, the creative media advertisement proved to be more effective in terms of attitudinal responses and behavioral intentions than traditional advertising. Either way, future research should strive for equal elements in both advertisements to rule out any other possible explanations that could cause the effects.
It should also be noted that consumers were exposed once again to the advertisement in the form of a photo half way through the questionnaire. By reminding participants of the advertisement, they were better able to answer the questions that followed. These questions, focused on ad attitude and WOM intention, might therefore not be completely based on the real life setting. Still, most participants tended to look again at their shopping cart, rather than looking at the photograph, while filling in the questionnaire. Additional analyses which included an additional covariate accounting for the fact whether or not people saw the advertisement, showed creative media advertising elicited a more positive WOM intention,
36 however, no effects were found for ad attitude. So, in general, results regarding ad attitude and WOM intention should be interpreted with some caution. To exclude the influence of a photograph, future research could test the effects of creative media advertising on ad attitude and WOM intention without incorporating extra photographs of the advertisement in the questionnaire.
Furthermore, on a more abstract level, it would be interesting to take into account various theories that are able to explain how creative media advertising works. Based on Schema Theory (Friestad & Wright, 1994), creative media advertising is only effective when consumers figure out the link between the medium and the advertised brand (Eelen et al., 2016). However, if this is the case, consumers have to be able and motivated to solve the puzzle. The elaboration likelihood model (Petty & Cacioppo, 1986) can be used to analyze through which route (central or peripheral) consumers process a creative media
advertisement. When consumers are motivated, capable and able to solve the puzzle, creative media advertisements are more likely to be processed via the central route (Petty & Cacioppo, 1986). Vice versa, when consumers are not motivated, capable or able to solve the puzzle, creative media advertisements are be processed via the peripheral route. To find out whether or not creative media advertising only elicits positive effects (e.g. attitudinal responses) when consumers solve the puzzle via the central route as opposed to the peripheral route, several variables from the elaboration likelihood model (e.g. motivation, opportunity and ability) need to be measured or manipulated. The results could consequently point out whether there are differences in the effects of creative media advertising on attitudinal responses between the central route and the peripheral route. In other words, the results would indicate if creative media advertising is only effective when consumers solve the puzzle.
Lastly, this study is mainly focused on the persuasive side of communication. However, the findings of this study could also be taken into consideration in other fields of
