A study of the influence of native and commercial online videa formats on consumer avoidance

(1)

A study of the influence of native and commercial online

video formats on consumer avoidance.

Master Thesis

Student: Raquel Prosetiko 10868658 Supervisor: Prof. Ed Peelen

Version: Final

Institution: UVA ABS Executive program in management studies – Marketing track Submission date: 26/03/2017

(2)

Statement of Originality

This document is written by Raquel Prosetiko who declares to take full

responsibility for the contents of this document. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

Abstract

In the context of native advertising little is known about if and to what extent the Persuasion Knowledge Model (PKM), in its current arrangement, can still explain and account for consumers understanding and processingof the content. This thesis employs the PKM to compare consumer perceptions for three different online video formats: one traditional format (a commercial) and two types of native advertising (a vlog and mockumentary). A preliminary study conducted by six in depth interviews finds the PKM still holds in a native advertising context but this hinges on the occurrence of advertising recognition. The quantitative study examined the effect of the online video format on consumer avoidance through selling and persuasion intent, source credibility and advertising recognition. The theoretical framework was tested in an experimental setting by online questionnaires

(N = 278) and analysed by Structural Equation Modelling. Results show higher rates of selling and persuasion intent had an increasing effect on advertising recognition, which in return had a decreasing effect on source credibility and increasing effect on avoidance, whilst a higher level of source credibility reduced avoidance. The mockumentary format had a much larger total reducing effect on avoidance compared to the commercial and vlog. This thesis contributes to a new wave of research on the application of the PKM in a native advertising context and differences in consumer perceptions for three very distinctive online video formats. The findings uncovered the application of a native advertising online video format does not always reduce consumer avoidance. The findings suggest that in order to enhance marketing success, marketers above all should strive to create compelling, engaging, and entertaining formats that reduce consumer avoidance. Furthermore limitations of the research scope and methods are reviewed. Practical implications for the field of native advertising are discussed.

Keywords: native advertising; persuasion knowledge; Persuasion Knowledge Model; PKM;

online video format; avoidance; selling and persuasion intent; advertising recognition; source credibility

Formatted: English (United States)

(5)

1. Introduction

Native advertising increasingly gained popularity over the past couple of years (Joel, 2013) and has become the buzzword in the media landscape (International Advertising Bureau [IBA], 2013). Native advertising has sprung from the essence of a non-disruptive and invited brand content approach and the advertising industry is currently embracing it (Campbell & Marks, 2015). Native advertising is referred to as a form of sponsored content; it basically describes any paid advertising that takes the form and appearance of editorial content of the publisher itself (Wojdynski & Evans, 2016). Yet it is also stated to go beyond sponsored content as it is not only designed to be read, but also designed to be shared and go viral (Mareck, 2014).

It is well known that consumers feel the tendency to avoid advertising for both traditional media (Speck & Elliott, 1998) as the Internet (Cho & Cheon, 2004). Cho and Cheon (2004) found that a perceived goal obstacle by the consumer is actually the most significant antecedent explaining avoidance of online advertising. Research has shown that ads that appear to be distractive lead to more negative attitudes towards the brand (Duff & Faber, 2011). Consumers are also limited in their ability to consume advertising (Malhorta, 1982). Therefore too much of Internet advertising could even reduce the collective effectiveness of it (Cho & Cheon, 2004) . As native advertising blends with its surroundings it can be perceived as a less distractive or disruptive form (Campbell & Marks, 2015). This provides an interesting research angle; would it be possible that the tendency to avoid native advertising will be lower than for commercial advertising?

A great deal of the research in advertising is based on the theory of the Persuasion Knowledge Model (PKM) of Friestad and Wright (1994). This model is based on the assumption that when consumers become aware of marketers their persuasion attempt it may result in a lower level of trust and less favorable attitudes towards the source (Friestad & Wright, 1994). Once consumers recognize a persuasion attempt in advertising, they will activate and carry out strategies to defend themselves against it (Friestad & Wright, 1994). Attaran, Notarantonio and Quigley (2015), whom have applied the PKM in their study, state the valence of selling intent in the message and perceived source credibility are the key pillars of this model. Evans and Park (2015) argue that during a native advertising episode the consumer does actually not recognize the exposure to advertising, therefore does not access priori persuasion knowledge and consequently does not activate an advertising schema. To activate an advertising schema is necessary to consequently activate the PKM. Evans and Park (2015) also argue negative attitudes, which regularly can stem from persuasion knowledge activation, are reduced when exposed to native advertising. Hence consumers don’t develop defending strategies in this case but

(6)

instead activate alternative schemas and associations related to the content and context of the source. There is still a great deal left to research on the supposed persuasion acceptance within the context of native advertising (Tutaj & Reijmersdal, 2012; Evans & Park, 2015). Little is known about consumers’ cooping mechanisms and the differences in effect on avoidance. It’s the question if and to what extent the PKM, in its current arrangement, can still explain and account for consumers understanding and processing when it comes to native advertising. (Evans & Park, 2015).

There is a large wave of research on traditional advertising in print and broadcast media (Speck & Elliott, 1997). Research on native advertising has been conducted for online banners (Duff & Faber, 2011), advergames (Evans & Park, 2015), editorials and advertorials (Attaran et al., 2015) and in the domain of social networks (Kelly, Kerr & Drennan, 2010). Native advertising can come in many rather different forms of information delivery. One of these forms is online video, which is increasingly becoming an important tool for marketers to reach their consumers (Moon, 2014). Since online videos can easily be accessed at websites, video- and social platforms such as YouTube and Facebook (Moon, 2014), and can easily be shared from person to person by convenient ‘share’ and ‘send this to a friend’ buttons, they are an attractive message format for marketers (Lee, Ham, & Kim, 2013).

Most research on video advertising has been conducted in the traditional context of the television experience on for example image quality (Bracken, 2005) and visual display (Lombard & Ditton, 1997). Specifically on online video advertising aspects such as impact of video quality and window size (Moon, 2014), video content (Southgate, Westoby & Page, 2010) and video ad length (Li & Lo, 2015) have been studied. Yet little research has been conducted on different video formats or the differences in a commercial and native context. Finally to the best of my knowledge no research has previously emphasized the differences in associations and schemas between these two contrasting contexts.

Aim of this thesis is to compare and evaluate consumer perceptions for three different online videos formats: an online video commercial, a mockumentary and a vlog. The first is a more traditional commercial format and the latter two are types of native advertising. The research will be conducted by a preliminary qualitative and a quantitative study. The qualitative study is conducted firstly to examine whether the PKM still holds in a native advertising context, or if adaption is necessary to remain applicable in this context. Aim of this study is to explore the associations and resulting schemas consumers develop when exposed to commercial- and native advertising online video formats. How do these differ for both contexts? Is it possible alternative schemas undermine the activation of the advertising schema and therefore the PKM? What alterations are necessary to adjust

(7)

the PKM to a native advertising context? These findings are employed to build a theoretical framework that enhances a comprehensive construct of the PKM, suitable to apply in both a commercial and a native context. The theoretical framework will be tested in the quantitative study to examine the effect of three online video advertising formats on consumer avoidance.

The objective is to answer the following research questions: Preliminary study:

- Which associations and schemas are activated during a native advertising episode, and how do these influence the activation of persuasion knowledge?

Quantitative study:

- How much is consumer avoidance influenced by the application of different online video formats? And how is this effect influenced by selling and persuasion intent, source credibility and the moderation of advertising recognition?

(8)

2. Literature review

In this chapter there is an elaboration on all topics applicable to answer the research questions. This will be done by means of reviewing key academic titles and management literature. First there is an elaboration on native advertising and the outcome variable avoidance. Second the PKM and its related variables advertising recognition, selling intent and source credibility will be discussed. Consequently schema theory is discussed, as this theory lays the foundation of the qualitative study, followed by an elaboration on the three online video formats.

2.1 Native advertising

The authors Campbell and Marks (2015) state native advertising has sprung from the essence of a non-disruptive and invited brand content approach, a subtler approach in the social media environment. According to them the initial aim of native advertising was to minimize the annoyance of brand content on social media. These authors claim the difference with traditional commercial advertising lies in the fact that native advertisements do not stand out, and are neither interruptive nor distracting. Instead native advertising formats are consistent with the online experience a customer is enjoying, whereas traditional advertising contains a distinct commercial message often with a clearly identifiable source (Tutaj & van Reijmersdal, 2012). Wojdynski and Evans (2015) refer to native advertising as a form of sponsored content. An editorial for example blends sponsored content with news or lifestyle content and is therefore considered as native advertising, whilst a banner ad is clearly commercial advertising. Native advertising is also referred to as secret, disguised or covert advertising. Campbell and Marks (2015) consider the secrecy in terms of both source disclosure and consumer invitation. It is paid advertising that takes the form and appearance of editorial content of the publisher itself. The native advertising playbook defines it as paid ads that are so cohesive with the page content, assimilated into the design, and consistent with the platform behavior that the viewer simply feels that they belong (IAB, 2013). The native ads behave consistently with the native user of the source and function just like natural content. Therefore native advertising is very suitable for viral marketing strategies (Mareck, 2014). Native advertising takes place online. Examples are in feed ads; not displayed at the top, bottom, or full screen like regular banner ads, but integrated into the content (IAB, 2013). The idea of the integration is to make them less annoying for the user. Other forms are paid weblogs (blogs), video weblogs (vlogs) and editorials containing an ad (advertorial). These are mostly featured on social media and platforms. Promoted listings are a form (IAB, 2013), referred to as listicles, that contain for example a top 10 holiday destination which is

(9)

sponsored by a brand. Finally Evans and Park (2015) also classify advergames and in-game advertising as a form of native advertising.

Even though native advertising is embraced by the advertising industry it is also attracting its share of critics (Wasserman, 2013; Campbell & Marks, 2015; Carlson, 2015). Wojdynski and Evans (2015) suggest that the growth in native advertising might not be because customers find it engaging, but because many of them do not recognize it as advertising. Therefore they don’t apply the avoidance and defense strategies they have developed for other types of online ads. Critics state it is a way of deceiving consumers, as the source of the message is often ambiguous or secret. Furthermore, for consumers it is not always apparent as advertising. In the Netherlands the Reclamecode Social Media (RSM) obligates to contain a disclosure in the native advertising form. A disclosure is a message that informs the consumer he/she is dealing with sponsored or paid content. These disclosures should help the consumer to distinguish between commercial and editorial content (Bakker, 2016). A study by Wojdynski & Evans (2015) shows advertising disclosures positioned at the bottom or in the middle mostly increase ad recognition among consumers.

2.2 Avoidance

Consumers are limited in their ability to consume advertising. It is cognitively impossible to pay attention to each and every selling message and besides the capacity of human memory is limited (Malhorta, 1982). Our awareness is selective and consumers reserve a limited amount of psychic space for advertising messages (Rumbo, 2002). Therefore consumers develop cooping mechanisms and ad avoidance strategies (Speck & Elliott, 1997). This will help consumers to guard themselves against being overwhelmed (Rumbo, 2002). Meaning when consumers are being confronted with too many ad messages, the excess stimuli will be filtered out and internal screening criteria will be applied. Only those messages that pass these criteria will receive attention (Rumbo, 2002). Avoidance can include a wide range of behaviors such as ignoring, flipping and skipping. Exposure can even be avoided completely in some cases by turning off a medium or channel (Duff & Faber, 2011). Duff and Faber (2011) argue that ad avoidance can occur for many reasons but that intrusiveness and inference with another activity are the most common explanations. Speck and Elliott (1997) mentioned search hindrance as a significant predictor. Cho and Cheon (2004) researched three antecedents of advertising avoidance online: interruption of task, perceived clutter on Internet sites, and negative past experiences. They found perceived goal obstacles to be the most significant antecedent. These authors particularly researched Internet ads that popped up when respondents were browsing the web. As the respondent was busy with another task it was considered as a

(10)

goal obstacle. This is supported by research of Raymond, Fenske and Tavassoli (2003) that suggests that when one is engaged in a task, distractors that are not part of it will receive negative rather than positive affective ratings. A qualitative exploratory study by Kelly, Kerr and Drennan (2010) argues that advertising in the social networking environment is more likely to be avoided if the user has expectations of a negative experience, the advertising is not relevant to the user, the user is skeptical towards the advertising message, or when the consumer is skeptical towards the advertising medium. Other research on avoidance states that in case consumers perceive an ad to be informative, they find it more enjoyable and appealing (Speck & Elliott, 1997).

The particular goal of the consumer when he or she comes across an ad seems to be an important determinant of whether or not avoidance will occur as a behavior.

Most of the avoidance literature has been about the consciously noted and processed ad. Much less attention has been paid to the unconscious part of processing ads (Duff & Faber, 2011). The latter is a more common and applicable situation in the context of native advertising. For this study the focus will lie on the behavioral construct of avoidance, the tendency for skipping the (native) advertisement.

2.3 Persuasion Knowledge Model (PKM)

The Persuasion Knowledge theory developed by Friestad and Wright (1994) is a conceptual model that is frequently applied for research in the area of advertising. The basic assumption of this model is that when consumers become aware of marketers their persuasion attempts, it may result in a lower level of trust and less favorable attitudes towards the source (Friestad & Wright, 1994). Once the persuasiveness in the message is recognized consumers will activate and carry out strategies and coping skills to defend against it. Wright already writes about the "schemer schema" the predecessor of the PKM in 1986. He argues people have intuitive theories that yield personal insights about marketer’s tactics. The "schemer schema" is required for marketers to design an influence strategy and for consumers to evaluate and interpret the persuasion attempts. Therefore he claims the schemer schema is likely to exist for people on each side of the marketplace game. Friestad and Wright present their conceptual model in 1994. They state that consumers develop persuasion knowledge over time. This knowledge helps consumers to identify when, how, and why they are being influenced by the tactics marketers use in their persuasion attempts. Their persuasion knowledge helps them to adapt and respond properly to these persuasion attempts and especially to pursue their own goals. Friestad and Wright (1994) stress that persuasion knowledge is active during each persuasion episode. In their model they focus on three knowledge structures that interact and determine the outcome of a persuasion attempt. These are: (1) persuasion knowledge; (2)

(11)

agent knowledge, consists of beliefs about the traits, competencies, and goals of the persuasion agent (e.g., an advertiser, a salesperson); and (3) topic knowledge, which consists of beliefs about the topic of the message e.g., a product, service, social cause, or candidate (Friestad & Wright, 1994, p. 4). These authors identify persuasion knowledge as a loose set of beliefs or intuitive theories about persuasion and beliefs that can be accurate or inaccurate. Campbell and Kimani (2000) whom have conducted several experiments on persuasion knowledge in sales settings, describe them as consumers’ mental knowledge structures about marketing and marketers their motives and tactics Persuasion knowledge contains schema like functions, for instance providing inferences about possible background conditions that lead to construct the persuasion attempt, but also to make predictions about the attempt's effects, and evaluate its overall competence. Another schema like function is the consumer’s paid attention to aspects of an advertising campaign or presentation (Friestad & Wright, 1994). Persuasion knowledge leads to the activation of certain advertising schemas in order to resist temptation. The advertising episodes or events are considered as mediators in consumers their conceptions of persuasion. With the help of these mediators consumers might find their emotional reactions are more influenced by ads than their own mental activities (Friestad & Wright, 1994).

There is little research on the activation of the PKM in a native advertising context (Tutaj & van Reijmersdal, 2012; Evans & Park, 2015). In addition the assumptions of these studies are contradicting. Some studies argue adults are fully capable to recognize the persuasion attempt in native advertising (Kinard & Hartman, 2013), whilst others argue that deviated consumer responses are elicited (Wei, Fischer & Main, 2008). Evans and Park (2015) whom studied advergames and in-game advertising argue that in fact other schematic antecedents and associative outcomes apply in the context of native

advertising. These do not lead to the activation of an advertising schema that is necessary to consequently activate the PKM. This activation does not occur since the consumer does not recognize the exposure to advertising and therefore does not access priori persuasion knowledge. Evans and Park (2015) argue the PKM explains persuasion avoidance but does not explain the process of persuasion acceptance. The PKM is based on the assumption that a higher rate of persuasion knowledge increases skepticism and suspicion. Yet, what happens when the consumer does not use his or her priori persuasion knowledge? What happens when other schemas are activated instead of the advertising schema, and what other schemas are those? Could the consumer feel misled when it turns out to be an event of disguised advertising and could this increase skepticism and suspicion even more?

(12)

2.4 Constructs of persuasion knowledge.

Persuasion knowledge has often been applied in studies but authors have studied and applied different constructs for the theory. However literature research yielded some dominant and main similarities that constructs the foundation of the persuasion knowledge theory. An experimental study by Tutaj and van Reijmersdal (2012) has found that different levels of persuasion knowledge dimensions are activated for subtle online ads such as sponsored content versus prominent online advertising. The persuasion knowledge dimensions they distinguished were selling and persuasive intent, advertising recognition and skepticism. They argue persuasion knowledge works differently for subtle and prominent online ad formats. In addition consumers are not only skeptical towards advertising because of the advertiser’s aim to persuade the consumer, but also because of the accredited biased source (Tutaj & van Reijmersdal, 2012). They also argue a higher rate of selling and persuasive intent of the advertiser leads to a more irritated consumer and increases skeptism. Attaran et al., (2015) whom studied different types of information delivery in magazines also consider the valence of selling intent e.g. visible persuasiveness in the message and the accredited source credibility as the key pillars of the PKM. The multilevel construct of persuasion knowledge is supported by other research (Boerman, van Reijmersdal & Neijens, 2012); Rozendaal, Lapierre, van Reijmersdal & Buijzen, 2011). Rozendaal et al., (2011) distinguished two dimensions of persuasion knowledge, a conceptual and attitudinal one. The conceptual dimension of persuasion knowledge is cognitive and concerns advertising recognition, the advertising source and audience, and the understanding of persuasive intent, selling intent and tactics. Attitudinal persuasion knowledge includes the attitudinal mechanisms affective in cooping with advertising. These are critical attitudes that include disliking, and skeptism, disbelief, trustworthiness, honesty and credibility towards the source. Viewers may or may not use attitudinal persuasion knowledge to process a message (Rozendaal et al., 2011). Ham, Nelson and Das (2015) have reviewed academic literature on applied measurements and methods for persuasion knowledge. They define conceptual persuasion knowledge represents if and to what extent consumers recognize the persuasion intent and attitudinal persuasion knowledge refers to if and how much they trust the given persuasion attempt.

Literature research has yielded that persuasion knowledge is a multidimensional construct. The conceptual dimension of its foundation consists of recognition of selling and persuasion intent, intertwined with advertising recognition. The attitudinal dimension consists of the credibility consumers accredit to the source, the cooping mechanism. In the quantitative study the attitudinal dimension is extended with the cooping behavior avoidance. There is an elaboration on these topics in the following paragraphs.

(13)

2.4.1 Advertising recognition

So in order to properly activate persuasion knowledge, people need to be aware of the persuasion attempt as previous research indicates (Boerman et al., 2012). Once consumers realize the communication is in fact an advertisement, they use their existing persuasion knowledge to make inferences about both the selling and persuasive intent of the communication and the person who communicates it or so called source (Rozendaal et al., 2011). Advertising recognition is a dimension of persuasion knowledge (Friestad & Wright, 1994; Tutaj & van Reijmersdal, 2012). As consumers’ exposure to advertising increases so will the degree to which they recognize advertising. Boerman et al., (2012) formulate advertising recognition more specifically as an aspect of conceptual persuasion knowledge. In literature advertising recognition is often mentioned inseparate from selling and persuasion intent. However there is a difference as the latter does not have to be recognized or present in the message in order to classify it as advertising. On the other hand it is unlikely to assume that advertising recognition stays out when recognition of selling and persuasion intent occurs.

2.4.2 Selling and persuasion intent

Campbell and Kirmani (2000) define the underlying motives of an influence agent or marketer as persuasion intent, the ulterior motive of persuading the customer to buy the product in a selling context. In literature the concept of selling intent is often treated the same as the concept of persuasion intent even though there is a clear difference (Carter, Patterson, Donovan, Ewing & Roberts, 2011). In case of persuasion intent the advertiser is attempting to increase the products desirability by appealing techniques and to make viewers do something they otherwise might not do. Instead selling intent can be defined as the promotion of a product via presentation of its features and qualities in also an unbiased manner (Roberts, 1983; LaPierre, 2015). Recognition of selling intent precedes recognition of an persuasion intent, but does not always trigger it.

In the research area of children and advertising some authors argue that the critical factor signifying children’s capacity for cognitive defense is in fact recognition of

persuasive intent rather than selling intent (Roberts, 1983; Wright, Friestad & Boush, 2005). In addition a quantitative study by Carter et al., (2011) suggests a large majority of children from eleven and twelve years is aware of the selling intent of advertisements, but only a minority is also aware of the persuasive intent. Friestad and Wright (1994) argue persuasion knowledge development is based on personal experience and practice with exposure to advertising. Opportunities for practice may only begin during early or middle adulthood. Besides they assume people’s motivation to learn how to cope with persuasion attempts should increase throughout life. These authors consider persuasion knowledge to

(14)

continue to develop and increase during life. For some advertisement forms the selling intent may be more obvious and most likely the consumer will be more aware of the persuasion attempt. Aligned with their Persuasion Knowledge Model, Friestad and Right (1994) state that when an persuasion attempt is recognized, the message its impact is determined by how consumers interpret the action of the marketer or source. They also state that this interpretation can discredit the message, spokesperson, or brand as they claim that recognition of a selling intent can have a negative impact on the perceived credibility of the source (Friestad & Wright, 1994).

2.4.3 Source credibility

Several studies have shown that consumers perceive commercial sources as less credible than non-commercial information (Cameron & Curtin, 1995; Lord & Putrevu, 1993). This is supported by an empirical study of van Reijmersdal, Neijens, & Smit (2010) that shows commerciality perceptions negatively influence attitudes, via perceptions of credibility and persuasive intent. A meta-analysis conducted by Eisend (2006) finds that when a source is perceived as biased, it is not perceived as trustworthy and therefore not credible. This perception is influenced by variables such as the amount of negative information included, attribute quality, and marketers' voluntariness. When negative information is avoided in the message, people will mostly assume the message is only created for the benefit of the seller. Sternthal, Dholakia & Leavitt (1978) found that trustworthiness along with perceived expertise make up the construct of source

credibility". They also found an effect of source credibility on persuasion. Expertise is the extent to which a communicator is believed to be a capable and appropriate source of valid assertions. Trustworthiness represents the level of confidence in the communicator’s intention to convey the argumentation that he or she considers most valid (Hovland, Janis & Kelley, 1953; Sternthal et al., 1978). Ohanian (1990) has developed a fifteen-item scale to measure source credibility for celebrity endorsers that is based on the models of trustworthiness, attractiveness, and expertise. The model of trustworthiness is built on the constructs of dependence, honesty, reliability, sincerity and trust. The model of expertise is built on the constructs of experience, knowledge, qualification, and skills. Finally the model of attractiveness is based on beauty, elegance, sexiness, classiness and

attractiveness. The latter was based upon the source-attractiveness model that is a component of the source valence model of McGuire (1985) that argues the effectiveness of a message depends on the familiarity, likeability, similarity and attractiveness of the source. A recent quantitative study by Munnukka, Uusitalo and Toivonen (2016) on peer endorsement in advertising has also shown that credibility of a peer endorser is

constructed from trustworthiness, expertise, similarities, and attractiveness dimensions.

(15)

2.5 Schema theory

The schema theory will be discussed as an alternative to explain consumer attitude in the context of native advertising. This theory is the foundation of the preliminary qualitative study.

2.5.1 Schema theory

The PKM seems to hinge on the activation of an advertising schema. In order to study this topic further in depth, it is important to define what schemas are in fact.

According to Bartlett (1932) a classical theorist, a schema is an active organization of past reactions and past experiences, which always operates in an organic response in a regular behavior stated memories are organized around schemas containing summaries of familiar stories or situations (Bartlett, 1932). When a new story is processed a schema recall occurs. In case the new story contains strange elements that do not fit the schemas are changed and adapted so that they fit more to an existing schema. The active organized setting consists of a consolidation all incoming impulses such as visuals, auditory and cutaneous (Bartlett, 1932). A single response is related to other similar responses that have been organized as a series and together form a unitary mass. The mass enables a particular response. In that case experiences from the past form fundaments for future perceptions and responses. Bartlet (1932) defines schemas are formed by past influences. Also schemas are enduring and do not easily change after being established. This classic view on schema theory describes schemas as general knowledge structures that reside in long-term memory. Minsky (1975) is the first author to recognize anticipatory knowledge in the knowledge structure. His theory implies that additional schemas are recalled to add further detail, schemas can be modified, new schemas can be created and the stored memory actually changes as a result of learning.

The classical schema theorist Rumelhart (1984) defines a schema as a data structure stored in memory by which generic concepts are represented. This so-called schema function helps us to interpret situations. We apply schemas for every situation. For example when you go to a hairdresser, you notify the reception upon arrival, sit and wait for your turn, get your hair done, and pay at the end. These are all scripts that belong to the schema of ‘hairdresser’. Variable constraints contain the knowledge we have about the values of variables and the relations between them (Rumlehart, 1984). For the hairdresser example there is knowledge for the values for service, hairdressers, haircuts, coffee, hair products etc. There is also knowledge on the relation between them; you know the hairdresser cuts the hair, and coffee is drunk when you wait. The knowledge on variable constraints and their relations helps us to distinguish important aspects of the situation and to make correct linkages (Rumlehart, 1984).

(16)

Modern schema theories offer more concrete descriptions of schemas. For example the Cognitive Schema Theory of Derry (1996) identifies a hierarchy of schemas: memory objects (specific small units of related characteristics), cognitive fields (activated set of memory objects), and mental models that represent an overall meaning structure of a particular situation or experience. Besides modern schema theory acknowledges there is a difference in the activation of memory sets for individuals based on moral judgment. Individuals with a more complex moral judgment have a better-organized set of memory objects. Their memory objects can also be activated within multiple cognitive fields and become part of a complex mental model, whereas those with lower levels of moral judgment have a limited set of possible activations (Narvaez & Bock, 2002). Also experts have more complex and extensive mental models about their specialism that can be activated in many ways (Narvaez & Bock, 2002).

Schema structures are in fact unconscious and are activated automatically when incoming data matches their pattern (Marcel, 1983). Schemas are constantly present yet they become activated by a stimulus that resembles the stimuli that created the schema in the first place (Marshall, 1995). Elements of schemas are connected on hierarchical level as previously explained, yet also on type level: positive or negative, and relationship degree: strong of weak (Marshall, 1995).

2.5.2 Types of schemas

There are numerous types of schemas that we apply in daily life. As Marshall (1995) stated: many kinds of knowledge can be linked through schemas, conceptual information,

discriminating features, planning mechanisms and procedural skills. Schemas always contain one or more systems that consist of specific knowledge structures, for example procedural knowledge (how to tie your shoe laces) and declarative knowledge such as the meaning of morality (Narvaez & Bock, 2002).

Cheng and Holyoak (1985) stress the importance of pragmatic reasoning schemas in evaluation and problem solving in daily situations. They define them as abstract knowledge structures derived from ordinary life experiences, such as permissions, obligations, and causal connections. They claim these particular schemas act a set of generalized context-specific rules that relate to the goals of an individual. With the help of pragmatic reasoning schemas individuals evaluate a situation. They perform action based on the future predictions about the outcome and how it relates to his or her goals (Cheng & Holyoak, 1985).

Relationship schemas play an important role in how people interpret and respond to the behavior and communication of others (Andersen & Cole 1990). In case a marketers behavior is interpreted as disrespectful or careless it could change the consumers

(17)

perception on how the marketer sees their relationship and therefore also how the consumer sees this (Friestad & Wright, 1994).

Goodstein (1993) has studied schemas of advertising tactics for specific product categories. Activation of this schema guides the way an ad is processed. For example prior affect towards a tactic can transfer to a new ad that is perceived as similar. The similarity can occur in the product category as in ad typicality.

Advertising can be considered as incoming information that is processed by consumers. Therefore Jurca and Madlberger (2015) state schema theory is applicable to consumers’ information processing behavior on advertising messages. Consumers shape an advertising schema to interpret, classify and respond to advertising messages quickly (Friestad & Wright, 1994). Dahlén and Edenius (2007) even describe the advertising schema as a mental shield that protects consumers from persuasive messages. As previously mentioned Friestad and Wright (1994) identify persuasion knowledge as a loose set of beliefs or intuitive theories about persuasion, beliefs that can be accurate or inaccurate. Besides they imply that awareness or recognition of an advertising event or tactic is necessary to instantiate an advertising schema.

2.5.3 an alternative for the PKM

The previously mentioned conceptual model of Evans and Park (2015) agues that

individuals constantly update their mental schemas and there are three routes to process a native advertising episode. They propose that in the absence of native advertising

knowledge, an advertising schema will not initially occur, individuals can also conduct a schema search and instantiate a best-fit schema that is for instance a ’game schema’. They apply a priori knowledge that belongs to the best-fit schema. A second route is that consumers learn from the native advertising episode and instantiate a schema through accretion, tuning and refining by which they access their priori knowledge. This can for instance be an updated/developed version of gaming schema towards advergame schema. They can also restructure by which they create a new schema that can be based on an existing or totally new schema. Finally they propose a third possible route that considers occurrence of associative transfer when activation of an advertising schema occurs. Evans and Park (2015) argue native advertising implies the pairing or embedding of advertising within the context of another media form, the unconditioned response to the content (the game) could transfer to the conditioned stimulus (the advertisement in the game). This process is referred to as ‘associative learning’ or ‘classical conditioning’ (Nord & Peter, 1980). Since the consumer experiences the advertisement in a native context, in this example a game environment, schemas associated with gaming are activated together with the accompanying likely more positive attitudes. In this case persuasion knowledge is still

(18)

activated but due to the associative transfer the positive associations overrule. This means skepticism and negative attitudes that result from persuasion knowledge activation can be reduced. The three described routes from to the conceptual model of Evans and Park (2015) are visualized in figure 1.

Figure 1

Schematic antecedents and associative outcomes of covert advertising persuasion knowledge: Instantiation and non-instantiation processes of advertising schema

Note: retrieved from (Evans & Park, 2015) Rethinking the Persuasion Knowledge Model: Schematic antecedents and associative outcomes

of persuasion knowledge activation for covert advertising.

This theory provides an alternative and plausible view on the application of the PKM in a native advertising context. In order to provide a thorough and complete view on this subject, I will conduct a qualitative study to firstly explore appropriate schemas and associative outcomes for native- and commercial advertising and define whether these undermine the activation of the PKM.

(19)

2.6 Online video formats

Online video advertising has become an important tool for marketers to reach their consumers (Moon, 2014). According to Brown (2008) online video advertising contributes to an immersive brand experience as it combines the branding power of traditional television with the direct response power of the Internet. Brown (2008) considers lower production costs to be an important benefit of online video advertising compared to traditional television broadcasting. Whilst compared to static online content forms videos are more engaging, compelling and lead to a higher brand recall.

Most research on videos has been conducted in the traditional context of the television experience (Bracken, 2005; Lombard & Ditton, 1997) . Research in the video advertising context has focused on video properties such as quality and size (Moon, 2014) or video content (Southgate et al., 2010). Southgate et al., (2010) have conducted a research on viral video viewing and important drivers to stimulate viewing volumes. They found consumer enjoyment, involvement and branding to be creative drivers to positively influence viewing volume. Furthermore they distinguish a few key determinants to positively influence viral video volume such as ad distinctiveness and to include a popular celebrity. In this thesis consumer perceptions on three types of online video formats are compared, all of which were developed for online marketing purposes.

These formats are described in the following paragraphs. 2.6.1 Mockumentary

Mockumentaries or mock-documentaries are fiction, but look like a documentary production. Over the past years the mockumentary has made its way into mainstream television. A mockumentary is a humoristic variant of a documentary or fake documentary (Mast, 2009). According to Campbell (2007) mockumentaries undermine the documentary’s claim to objectively tell the truth. It is a genre that contains irony. This can be either in the parody of the form of the documentary or in the satirical treatment or critique of an issue. Rhodes and Springer (2006) define mockumentaries as a documentary form that contains fictional content, by which they differ from docudramas that contain a fictional form and documentary content, or the classic documentaries that contain documentary form and documentary content. Mockumentaries and docudramas can be classified as ‘docufiction’ (Mast, 2009). Mockumentaries are set-up with the style and techniques that apply to the classic documentary setting such as a voice-over, commentary, interviews, and a hand held camera (Mast, 2009).

The mockumentary has made its entrance on ‘reality television’ and entertainment and has become an aspect of contemporary television and popular culture (Mast, 2009). However the application of this format in advertising seems to be relatively new and only

(20)

little. The research field lacks empirical research on consumer perceptions of

mockumentaries as a format, but especially on mockumentaries that contain an advertising or brand message.

2.6.2 Vlog

Nowadays online consumers have also become content creators, by composing, editing and modifying data (Ho, Chiu, Chen & Papazafeiropoulou, 2015) Vlogs have become a very popular medium for online video users (Biel, Aran & Gatica-Perez, 2011). A vlog is basically a video blog; the term blog is deducted from weblog (Ho & al., 2015). Vlogs can be considered as a unique medium for self-representation and interpersonal perception in social media, one that goes beyond the usage of text or photo’s (Biel et al., 2011), as it is the case for blogs. Basically vlogs are the evolved form of blogs that contain video content. Vlogs are a genre of computer-mediated communication. A vlog involves mostly a single speaker, the so-called vlogger, shooting video footage of him- or herself (Frobenius, 2011). This video footage is then uploaded on the Internet (Frobenius, 2011). Anyone can make and upload a vlog and it is the opposite of a professional presented show or news report. Some vlogs show signs of hesitation and nervousness, since the speakers are often not professional presenters (Frobenius, 2011). The topics of vlogs tend to be very various and there are no restrictions or limitations for this. Vlogs are significantly shaped by the social context, and vloggers are mostly independent, untrained, private and unpaid (Frobenius, 2011). The reach level of vlogs varies as some vlogs only reach about a 100 views, yet others reach a million views (Frobenius, 2011).

There is little research on the effectiveness of vlogs as an advertising tool, but blogs have been researched to a certain extent. Bansal and Voyer (2000) state that consumers whom frequently read and exchange information on blogs, develop strong bonds and close interpersonal relations. This also increases the likelihood of trust and credibility they assign to the information published on blogs. Blogs serve as an important information source to gain product advice and suggestions and they can impact consumer’s evaluations of products and the decision-making process (Ho & al., 2015). Ho et al., (2015) conducted a study on celebrity blogs, expert blogs and consumer blogs. They point out that

consumers have most confidence in information on typical consumer blogs, yet expert blogs do a better job in facilitating the brand message. Therefore expert blogs can enhance more advertising effectiveness. Besides they state that consumers tend to discount the credibility of a blog if they perceive it as commercialized.

(21)

2.6.3 Commercial

Television used to be one of the most cost-effective media. Its advantages are impact and large reach. Therefore TV commercials have been a popular advertising form. Yet it is also highly expensive to create, produce and air a TV commercial (Duan, Wang, Zheng, Jin, Lu, & Xu, 2006). Nowadays video commercials do not only appear on TV, but also on mobile and online (Kong, Zhao, Hu, Vecchiato, Babiloni, 2013). Online video ads merge the branding power of traditional broadcast advertising and the direct response power of the web (Ham et al., 2015). Online video ads bring advertising content that is more compelling to consumers than static forms such as banners or popup ads (Southgate et al., 2010). Nowadays many brands use online commercials for specific promotions or smaller scale activations. An online commercial will be used for this study, as it contradicts the native nature of the mockumentary and vlog, and also applies to the social media context.

(22)

3. Preliminary study

To answer the central research questions in this thesis a sequential embedded mixed method research was pursued, as one methodology was meant to support the other (Saunders, Lewis & Thornhill, 2012). The first research was conducted by a grounded theory strategy and focused on the identification of the schemas and associations that are activated during a native advertising episode; this is a qualitative study and will be discussed in this chapter. In chapter 4 the quantitative study will be discussed. This chapter contains the theoretical framework built with the output of the qualitative study. This framework is tested in an experimental setting by a survey strategy.

3.1 Research question qualitative study

Aim of this study is to identify which alternative associations and schemas occur and possibly undermine the activation of the advertising schema and therefore the PKM. This will help to identify whether the PKM is still applicable or how it should be adjusted in order to remain applicable in both a commercial and a native advertising context?

The findings of this study are utilized to build a theoretical framework to apply for the quantitative study.

Research Question 1:

- Which associations and schemas are activated during a native advertising episode and how do these influence the activation of persuasion knowledge?

Figure 2

Theoretical framework preliminary study

3.2 Method

The preliminary study was a small-scale exploratory research to gain insight in the occurring schemas and associations for the three video formats and to define how these relate to the activation of the PKM. This study contributes to explain whether or not and how the PKM will preserve itself in a native advertising context. Furthermore it exposes the alterations that need to be made to the PKM in order to make it applicable for a native advertising context.

(23)

3.2.1 Research design

The qualitative study was preliminary and meant to identify the appropriate variables and add insights to the quantitative research model. Aim was to apply a schema theory to the PKM and highlight the associations that come up during exposure to a commercial and native advertising episode. These associations will lead to the activation of certain schema and cooping mechanisms. Associations that are not directly linked to advertising might be activated, their effect could be elicited and influence the final consumer attitude. Aim of this study was to find out ‘what is happening’ and to seek new insights in order to map associations and schemas. This set the purpose to exploratory as the study may provide more background and contextual material. The research approach was mainly inductive and slightly deductive as a research question was formulated based on theory. Qualitative research is based on meanings expressed through words. A grounded theory research strategy was applied, as the aim of this strategy was to develop theory inductively from a set of data. Grounded theory is usually referred to when an inductive approach is maintained (Saunders et al., 2012). Individuals construct meanings to make sense of everyday experiences in specific situations and grounded theory is used to analyze and explain these meanings (Charmaz, 2014; Suddaby, 2006).

The data was gathered through interviews. According to Saunders et al., (2011) the use of interviews helps to gather valid and reliable data relevant for the research

question(s) and objectives. First the participants were exposed to the stimuli and after this they were asked to mention as many associations as they had experienced. The interviewer would write all these topics on cards. The participants were then asked to divide them into senseful groups in order to discuss these further. The interviews took place with six respondents in total, two persons allocated to each of the three conditions.

Table 1

Qualitative research design Type of information delivery

Treatment 1 Commercial Treatment 2 Mockumenary Treatment 3 Vlog Product subjects 2 participants 2 participants 2 participants

(24)

3.2.2 Stimuli

The following formats of online video advertisements are included for both studies: -A commercial that contains a very prominent activation message and the Samsung brand name is mentioned with extra emphasis and accent. This video can be classified as a traditional commercial advertising. The commercial is about a promotion called ‘Own The Twilight’. Consumers can win a free trip to Bali when they upload a photo on the Samsung website made during a ‘Twilight’ moment.

-A mockumentary that only features the brand name in the final animation logo that appears at the end. In advertising this final tagline referred to as the ‘supers’. The mockumentary contains a storyline about a fisherman in Iceland who always posts blurry photos on his Instagram account as his camera cannot deal with the limited light conditions there. Then he received a brand new Samsung smartphone. With this phone he can make beautiful photos even in situations with little light. At the end the fisherman is reunited with his friend Radek who once saved his life. At the time he could not make a proper photo of the two of them but this is now possible with the Samsung smartphone.

-A vlog featuring a review of a vlogger from the techforum Tweakers that is considerably known for their technical reviews. It is a talk about the features of an edition of a Samsung telephone. The vlog contains a desk setting with a presenter and close-ups of the smartphone.

The vlog and mockumentary can both be classified as forms of native advertising. So a comparison can be made between a native advertising and commercial context. All three videos feature the same smartphone product of the brand Samsung.

3.2.3 Participants and procedure

The target group for both the qualitative and quantitative study consisted of consumers that are exposed to Internet, online video’s and advertising. Besides the targets should not work for, or have a professional relation with the brand that is featured in the videos. In order to gather sufficient information it was important to select ‘information rich participants’. Krueger and Casey (2015) define ‘information rich participants’ as those from whom you can learn a great deal. Participants were chosen by a non-probability convenience sampling technique. This technique is often applied for interviews as there is mostly a specific purpose for selection (Saunders et al,. 2012).

The qualitative data was acquired from my personal network. Participants were randomly assigned to one of the three conditions. All participants were informed that they would watch a video and had to answer some questions about this. The interviewer sat across them during the video in order to prevent distraction or tension for the participant. The interviews were mostly conducted in home settings with the exception of two

(25)

interview settings in lunch places. The interviews were conducted in a one on one and face-to-face situation. By doing so the participant would have all freedom to speak freely and furthermore participants could not influence each other. The interviews were non-standardized and semi-structured in order to allow in depth exploration and probing (see Appendix A). These characteristics are often met for qualitative research interviews (King, 2004). The interviewer would have a list of themes and key questions to be covered. The order of these questions could vary, and additional questions could be asked to explore the research question further.

The first part of the interview would contain basic questions such as age and gender. After this the respondent was exposed to the video. The middle part right after the video was unstructured in order to map the associations and schemas that are at the top of mind of the respondent. The final part of the interview was structured and would contain open-ended questions. The interviews were audio recorded and the length was between ten and twenty-five minutes.

3.3 Data analysis and results

The results of the interviews were analyzed by Kings’ (2012) template analysis, a form of a thematic analysis. This is considered to be a flexible approach unlike other grounded theory approaches that are more prescriptive (Brooks, McCluskey, Turley, & King, 2015). A coding template was developed in order to further revise and define the data (Brooks et al., 2015). This technique encourages to develop themes as extensive as their richness of data and emphasis lies in the hierarchy of coding. The initial coding template can either be developed before or after the collection of the first set of data (Brooks et al., 2015).

I chose to develop the initial coding template after collection and transcription of the data in order to increase familiarity with it. The initial template as shown in table 2 contained the main themes and topics that were covered in the interview setup. During this process constant comparison was made, new codes were created and existing codes were re-analyzed. According to Saunders et al., (2012) these are higher levels of analytical coding since it combines both inductive and deductive thinking. Due to the small sample size, demographic data was disregarded in this analysis. Consequently the data was organized by NVivo software according to these themes.

Answers from probing questions were used as input to tweak the coding template further. The revised template did not contain the category ‘entertainment’ as it turned out to be redundant for the analysis and the category ‘neutral’ was added to the topic of attitude.

(26)

Table 2

Initial and revised coding template

Initial coding template Revised coding template 1. Schemas • Associations 1. Schemas • Associations 2. Advertising recognition • Advertising cues 2. Advertising recognition • Advertising cues 3. Brand recognition 4. Brand recognition 5. Entertainment 3. Attitude • Neutral • Positive • Negative 6. Attitude • Positive • Negative

The schemas were analyzed according to both the coding categories in NVivo and the data as written on the cards. The participants would often group together multiple associations under one topic (schema). After analyzing the probing questions and exact wordings in the transcripts more schemas could be identified. The schemas were further identified during the coding of the transcripts and analysis of the cards, which resulted in the final coding template.

Table 3:

Final coding template 1. Advertising recognition • Contains activation message • Features brand

• Focus on product • Made by professionals • Not objective

• Similar to other commercials • Too positive 2. Brand recognition 3. Attitude • Neutral • Positive • Negative

(27)

Table 3:

Final coding template (continued) 4. Schemas > Elicited per condition

Commercial Mockumentary Vlog

Advertising Product and features Product and features

Samsung phone Iceland Samsung products in possesion

Samsung (brand) Personalities in video Own phone

Unrealistic Photos Opinion about video

Camera Storyline

• Story sequence • Negative causes • Positive consequences

Associated people

Connect with others Contrast Associated products (covers) Desirable aspects Advertising Presentation style

Social media Criticism

Take action Samsung (brand)

Advertising

3.4 Discussion

During all three conditions the videos were considered to be advertising, with the exception of only one participant in the vlog condition. Yet only for the commercial, advertising recognition was explicitly mentioned on the own initiative of the participants. For the mockumentary and vlog the interviewer had to ask the participants if this was considered advertising. It seems like the advertising schema is primarily activated and more prominent for the commercial instead of the native advertising conditions. One participant in the commercial condition does even hardly experience any other schemas besides the advertising schema. However for the other participant in this condition schemas also arise for desired aspects of the video such as leisure time, special moments, traveling and social media.

The advertising schema, treated indifferently from advertising recognition, was activated by several cues. Further analysis has yielded seven different categories of cues for advertising recognition: in case the video contained a clear activation message, stimulated action, contained a brand, when there was a clear product focus, a professional appeal, it was unobjective, too positive, and when participants recognized elements from other commercials. The latter category is completely aligned with Evans and Park (2015) their view as consumers make an attempt to access a priori persuasion knowledge to evaluate whether they are dealing with advertising. Selling and persuasion intent seems to precede advertising recognition. Selling and persuasion intent overlaps with all seven cue

(28)

categories of advertising recognition. As these are all cues that can possibly lead to the consumer’s recognition that someone is trying to sell them something or persuade them. For instance the cues of ‘to contain an activation message’, ‘to stimulate action’, ‘unobjectivity’ and ‘too positive’ directly relate to recognition of selling and persuasion intent. The other cues relate indirectly since ‘professional appeal’, ‘contains a brand’, ‘clear product focus’, and ‘to recognize elements from other commercials’ insinuate they are dealing with a professional company and the video might have a commercial purpose. Brand recognition was a cue that could lead to advertising recognition but not per se. Remarkably enough all participants in the mockumentary and vlog condition recognized the brand but did not classify the video as advertising on their own initiative nor did they point out Samsung as the responsible marketer. Thus in the native advertising conditions the assumed responsible marketer was not identified as the brand.

In all conditions the product and feature schema was activated. Remarkably enough in the vlog condition participants also developed schemas about their own devices, its features and associated products.

Furthermore the storyline of the mockumentary was an important schema as participants made efforts to recall the chronological order, the causes and consequences. Both participants experienced the unhappy wife and the blurry photos as negative causes or antecedents in the storyline. It was considered to be a positive consequence when the fisherman received the phone and could make more beautiful photos. Also contrasts were explicitly mentioned such as: the difference between an old-fashioned fisherman and advanced telephone, or the fact that the fisherman knew perfectly how to use the phone from the start. Besides the participants developed schemas for the personalities

(fisherman and wife) and setting (Iceland, dark and cold).

The participants in the vlog condition formed schemas regarding their own opinion and evaluation of the vlog more prominently (boring, objective etc.). Possibly these occurred since the vlog only contained little visual aspects; only the presenter and the telephone. Participants couldn’t develop schemas about the setting of storyline. Also there was little distraction from the message as the talk of the presenter was the main stimulus. It is notable that only the criticism of the vlog was recalled instead of all other positive highlighted features. The valence of the message recall seemed to lie on the negative information presented by the vlog.

In all conditions participants developed pragmatic reasoning schemas to evaluate source credibility as schemas arose for unrealistic/contrasting aspects, incongruency (commercial and mockumentary), their own evaluation of the source, applied presentation style and negative arguments presented (vlog). Also the commercial was claimed to be

(29)

‘’unrealistically positive’’. This is aligned with the view of Eisend (2006) that when negative information is avoided, people will believe the message is only created for the benefit of the seller.

Unfortunately no conclusions could be drawn on consumer attitude as the neutral, positive and negative responses accounted for all conditions. No clear distinguishments could be made.

The findings indicate numerous schemas were activated during both the commercial as the native advertising conditions. These alternative schemas contained storytelling, novelty aspects and desired aspects from the videos. Most striking finding is the advertising schema arose primarily in the commercial condition and occasionally prevented the development of all other schemas. In the native advertising conditions more alternative schemas were developed, seemingly since the advertising schema was less leading and persuasion knowledge was only remotely activated. This is the key difference between the native advertising and commercial condition. This finding is consistent with Friestad & Wright’s PKM (1994) as the activation of advertising schemas makes consumers aware of the persuasion attempt and stimulates them to resist temptation. It can be concluded that the PKM also holds in the context of native advertising. However its activation does not seem to hinge on the occurrence of alternative schemas, but on whether an advertising schema is developed. As the research question implied it was expected alternative schemas would undermine the activation of the advertising schema and therefore the PKM. Instead this depends on the prescence or absence of cues for that trigger advertising recognition.

The research has yielded several cues that might activate the association network for advertising recognition and might activate the coping mechanisms. These cues seemed to be overlapped by selling and persuasion intent. As all participants developed pragmatic reasoning schemas in order to evaluate source credibility, it seemed like participants attempted to activate the attitudinal dimension of persuasion knowledge. Participants first attempted to activate the conceptual persuasion knowledge and consequently did or did not trigger attitudinal persuasion knowledge. Based on these findings it was concluded the variables of selling and persuasion intent, advertising recognition and source credibility should be included for the quantitative study.

Considering the research method there are a few limitations to take into account. An advantage of the interview technique is the high involvement of the participant and the lack of response dropout. In total no more than six interviews were conducted and theoretical saturation occurred. On one hand the validity was relatively high since the interviewer did have proper access to the participant’s knowledge and experience. Since

(30)

the nature of the interview was semi-structured, the interviewer was able to infer meanings from the language of the participant by probing. On the other hand as one single interviewer conducted all interviews, the results are very much depending on the

interpretation and approach of the interviewer, an aspect that decreases the validity of this study as interviewer bias might occur. There lies a limitation in the generalizability of this study as the sample was relatively small and conducted in different settings and the findings might not be replicable (Saunders et al., 2012). According to Saunders et al., (2012) semi-structured interviews reflect reality at the time of collection and are not indented to be repeated. Hence the limited ability for replication affects the reliability of this study. However the approach was adjusted to the objective of this study: preliminary and exploratory with the intention of data induction and to develop a theoretical framework for the quantitative study. To increase reliability a lab setting might level out any side effects of the environment. To prevent interviewer bias the interviews should be conducted by more interviewers and the coding should be done by a different person.

A study of the influence of native and commercial online videa formats on consumer avoidance