• No results found

The Effect of Trust in Feelings on the Customers' Persuasion Knowledge: A Study About Native Advertising.

N/A
N/A
Protected

Academic year: 2021

Share "The Effect of Trust in Feelings on the Customers' Persuasion Knowledge: A Study About Native Advertising."

Copied!
60
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

The Effect of Trust in Feelings on the Customers'

Persuasion Knowledge: A Study About Native

Advertising.

University of Groningen

Faculty of Economics and Business

MSc Marketing Management

Master Thesis

Date of Submission: June the 17

th

2019

Supervisor: Anika Schumacher, MSc.

Co Assesor: prof. dr. B.M. Fennis,

By:

(2)

1

ABSTRACT

In lights of online news site readers inability to detect a persuasion intent contained in the sponsored news article, we build our focus of this study to examine a psychological factor that could make readers (not) aware of a persuasion attempt. We suspect that individuals’ general trust in their feelings could affect their ability in activating their set of persuasion knowledge; even at the presence of different saliency of manipulative intent contained in the message cue of a sponsored news article. That in turns can affect individuals’ attitude formation towards the sponsored news article (i.e., could be positive or negative). To answer our research question, we built three hypotheses of this study. And by the means of a 2x2 between subject design study, we then try to find support to our hypotheses. However, we fail to find enough statistical evidence to support our hypotheses. We found that individuals’ trust in feelings does not significantly affect the attitude formation of an advertisement. Additionally, our study also found that once individuals’ felt that they are being deceived by an advertisement, they will form a more negative attitude towards the respective advertisement.

(3)

2

Preface

Back then, I worked as the operation and marketing manager for an Indonesia’s emerging online news site. During that time, it is still hard for publishers to construct an engaging story for the sponsored news article, while the demand from principals alike are skyrocketing. Thus, it made several emerging and growing publishers hard to achieve the minimum key performance indicators of 20,000 unique readers in a three-day time window. The problem I encountered during that time literally drives my motivation to study about marketing and write my thesis in the sponsored news article literature stream. Thank God that my supervisor, Anika Schumacher, approves my intention to write a thesis about trust in feelings in the context of sponsored news article, which I have personal connections and interest to.

The whole writing process of his master thesis literally sums up my life as master student in the University of Groningen. There are so many ups and downs happened to me during my process in finishing this document. Starting from when I lost my mother that passed away because of a complication illness in Indonesia during the beginning of the writing process of this thesis. Up until the day I finished typing the very last single letter in this document. Thus, I would like to give a thanking note to several people that are involved, helped, and supported me during the writing process of this document.

First, I would like to thank God that always keeping my hope alive through the thick and thin of the writing process of this thesis. Second, I would like to thank my family in Indonesia and friends that continuously giving an endless support to me in finishing this thesis. Subsequently, I would like to give a thank you note to my supervisor, Anika. Thank you for your guidance and putting your believe in me to conduct and complete this master thesis in the native advertising context. Also, I would like to thank Bob Fennis as my co assessor. Thank you for your time and effort in evaluating my thesis.

(4)

3

TABLE OF CONTENTS

ABSTRACT ... 1 PREFACE ... 2 INTRODUCTION ... 5 LITERATURE REVIEW ... 8

NATIVE ADVERTISING AND PERSUASION KNOWLEDGE ... 8

MESSAGE CUE IN NATIVE ADVERTISING AND PERSUASION KNOWLEDGE ... 12

TRUST IN FEELINGS AND PERSUASION KNOWLEDGE ... 13

CONCEPTUAL MODEL ... 19

STUDY 1 ... 19

OVERVIEW ... 19

PARTICIPANTS ... 20

PROCEDURE AND DESIGN ... 21

EXPERIMENTAL MANIPULATIONS ... 22

Trust in Feelings Manipulations ... 22

Triggering the Activation of Persuasion Knowledge ... 23

MEASURES ... 24

Manipulation Check ... 24

Persuasion Knowledge ... 25

Advertisement Attitude... 25

Exploratory Variable: Regulatory Focus ... 26

RESULT ... 27

Manipulation Check ... 27

Hypotheses Tests ... 28

Hypothesis 1 ...28

Hypothesis 2 ...32

Hypothesis 3: Exploratory variable ...33

OVERALL RESULT ... 33

GENERAL DISCUSSION AND CONCLUSION ... 34

IMPLICATIONS ... 36

LIMITATIONS AND FURTHER RESEARCH DIRECTIONS ... 38

LIMITATIONS ... 38

(5)

4 REFERENCE ... 43 APPENDIX ... 50 APPENDIX 1 ... 50 Presented advertisement. ... 50 APPENDIX 2 ... 51

Questionnaire in the study. ... 51

Main Manipulation: High trust in feelings ...51

Main Manipulation: Low trust in feelings ...51

Persuasion Knowledge Measure (Kirmani & Zhu, 2007) ...53

Attitude toward the ad measures (Muehling, 1986) ...54

Manipulations Check ...54

Regulatory focus measure (Lockwood, 2002) ...55

Participants Demographic ...58

APPENDIX 3 ... 59

(6)

5

INTRODUCTION

In recent years, online news media readers are being presented with a new advertising format. This new advertising format is designed to have similar editorial content with the non-commercial content of the online news site (Wojdynski, 2016) and is purposively tailored to engage readers in the manner that the conventional display advertising oftentimes failed to do so (Dréze & Hussherr, 2003). Due to its similarity to the publisher's native editorial content, many researchers termed this type of advertising as "native advertising" (Campbell, Cohen & Ma, 2014; Wojdynski, 2016). As in the online news site, native advertising usually takes form in a text article that resembles the journalistic and editorial style of the non-commercial text article of the respective online news site (Ferrer Conill, 2016); which is later termed as the sponsored news article (Wojdynski, 2016).

Amidst the growing existence of sponsored news article, large proportions of people perceived the sponsored news article as the product of journalistic rather than seeing it as an advertisement (Wojdynski, 2016). Consequently, it makes individuals easily believe the claims made therein (Lazauskas, 2015). In this manner, it could be troublesome for readers as they may create false beliefs from the biasing information contained in the sponsored news article; which is resulted from their inability to objectively judge the true nature of the article and objectively appraise the claims and messages made therein (Olsen & Dover, 1978; Wojdynski, 2016).

(7)

6

intent (Friestad & Wright, 1994) and results in a more reluctance to persuasion that leads customers to form a negative attitude toward the advertising (Kirmani & Zhu, 2007). This important tenet can then determine readers’ ability to critically judge the advertising message contained in the sponsored news article.

Extant empirical studies about native advertisings have demonstrated that readers’ capability in activating their set of persuasion knowledge would rely upon the sponsorship disclosure made before the article content (Wojdynski, 2016; Evans & Park, 2015). However, it is still evident from Wojdysnki & Evans (2016) study that most readers are still incapable of activating their set of persuasion knowledge even when the sponsorship disclosure is being made salient by the publishers. To the best of our knowledge, little is known about why readers are failed to detect the persuasion attempt even when publishers have disclosed the article’s paid and true nature.

In light of readers' inability to detect the persuasion attempt in the sponsored news article, we propose that one's general trust in feeling is a psychological factor that could affect readers ability in activating their set of persuasion knowledge. Investigating the effect of trust in feelings on readers' ability in activating their persuasion knowledge is becoming increasingly important as oftentimes customers rely on their feelings in making judgments and decisions (Avnet, Pham & Stephen, 2012; Pham, 2009). According to Avnet et al. (2012) individuals that have a high trust in their feeling, oftentimes utilize their feeling (internal information input) as the sole piece of information source in making judgment rather than considering the other sources of information (e.g., external information input) before making judgments and decisions (Avnet et al., 2012). In this study, we propose that individuals’ persuasion knowledge will then act as the external information input for them to construct an appropriate attitude in judging the sponsored news article.

(8)

7

are less motivated in processing information and thus will fully rely on their feelings as the main source of information as they deem that their subjective affective inputs are to be more diagnostic (Avnet et al., 2012). Second, people that rely on their feeling in making judgments and decisions are normally having a promotion focus; which highly value the speed over the accuracy in making judgments when both are in competition (Crowe & Higgins, 1997). Here we define promotion focus as the strategic orientations of an individual, which is rooted from the self-regulatory focus theory about individuals’ goal pursuance orientation (Higgins, 1987). Promotion focus individuals are more of approach-oriented which in the signal detection terms, it is equipped in the direction of maximizing “hits” and minimizing “miss” (Crowe & Higgins, 1987; Pham & Avnet, 2009). This motivational state could then affect individuals’ reliance on the subjective affective input (Pham & Avnet, 2004; Pham & Avnet, 2009) in making judgments.

In this manner, the speed of making judgments is achieved through the reliance to the affect heuristics that utilize feelings to give the subjective affective responses (Pham & Avnet, 2009; Verplanken, Hofstee & Janssen, 1998; Zajonc, 1980). We thus argue that this type of individuals would highly value their feelings and are deem that their subjective input is to be more diagnostic than other external input. Consequently, the negative attitude towards the sponsored news article that goes with the activation of persuasion knowledge would be less influential for readers high on trust in feelings compared to readers low on trust in feelings, resulting in a more positive attitude toward the sponsored news article.

Hence, our proposed research question will be: Are individuals that have a high trust

in their feeling (compared to individuals who have low trust in their feelings) less likely to detect a persuasion attempt?

(9)

8

statistical support to all hypotheses that we constructed for; our result additionally able to extends the generalizability of the persuasion knowledge model. That is, once individuals felt that they are being deceived or manipulated by an advertisement, it will subsequently make them form a more negative attitude toward the advertisement.

This document is constructed as follows. First, we will comprehensively elaborate about the interaction between trust in feelings, persuasion knowledge, and message cues in the native advertising in the literature review section. Subsequently, from the elaborated theories in the literature review, three hypotheses are constructed. Then our hypothesized effects are visualized in the conceptual model section. Next, we describe our experimental setup and summarize our analyses and results. Afterwards, we discuss our findings in the general discussion and conclusion section. At last, we provide several limitations of our study as well as the future research directions.

LITERATURE REVIEW

Native Advertising and Persuasion Knowledge

(10)

9

Researchers in the native advertising has raised their critics upon the sponsored news article. For instance, Campbell & Marks (2015) argued that the manipulative and deceptive nature of the sponsored news article plays a significant role in bringing success for this type of advertisement in the industry. In this manner, manipulative advertising defined as when advertisers are utilizing an inappropriate and unfair means ‒ that frequently deceived the target audiences ‒ in an attempt to persuade customers (Campbel, 1995; Morales, 2005). Utilizing sponsored news article as a means for advertisers to convey their message is oftentimes deemed as a way for advertisers to deceive their target as customers usually tend to think that the advertised message and claims contained in the sponsored news article are made by the respective news site instead of advertisers (van Reijmersdal, Neijens & Smit, 2005). This could happen due to the similarity of the sponsored news article's journalistic style with the non-sponsored news article. In this way, advertisers are claimed to be able to better build their brand or product belief in customers' mind; as suggested by Hastak & Mazis (2011) that customers trust toward the source of the claim (i.e., the respective news site) might degrade the possible negative attitude toward advertising message posited in the sponsored news article's content. Hence, the deceptive nature of native advertising could then be seen from the likelihood of readers being unaware that the article actually contains an advertising message (Lee, Kim & Ham., 2016) and readers unconsciously creates belief from the information and claims posited in the sponsored news article (Hastak & Mazis, 2011; Olsen & Dover, 1978).

(11)

10

might hold a misleading or even false belief about the product claims (Gardner, 1975; Jacoby & Small, 1972; Olsen & Dover, 1978). One crucial takeaway point of studying the negative impact of deceptive advertising and subsequently invoke this concept to the sponsored news article context is that it is crucial for readers to recognize, detect and critically appraise that the editorial content contains advertising messages and claims so that readers might not be deceived and consequently held false beliefs after reading the sponsored news article.

There is an important tenet that can explain readers' ability in detecting, recognizing, and behave accordingly toward the message and claims made in the sponsored news article; that is the activation of persuasive knowledge model (PKM) (Craig, Loureiro, Wood & Vendemia, 2012; Evans & Park, 2015); in which first comprehensively documented in Friestad & Wright (1994) study. Persuasion knowledge refers to “one's understanding of persuasion attempt that involves one's belief about marketers' motives and tactics as well as the appropriateness of the utilization of marketers' persuasion tactics” (Friestad & Wright, 1994, p.6).

(12)

11

Furthermore, Friestad & Wright (1994) suggest that people's knowledge of persuasion attempt will develop over time, and the knowledge acquired from the experience and exposure of persuasion attempt will then be activated and utilized by customer to cope with the persuasion attempt that they are being exposed to; before subsequently forming an appropriate attitude toward marketers' motive and tactics (Campbell & Kirmani, 2000; Friestad & Wright, 1994). In this way, customers' prior exposure to any persuasion episodes could then help them to identify as such persuasion attempt (i.e., in the form of the sponsored news article) before consequently activating their persuasion knowledge in making judgments and decisions (Friestad & Wright, 1994).

However, this knowledge of persuasion attempt is not always easily accessible and retrievable for every customer at any point of time. Rather, it depends on customers’ cognitive capacity (Campbell & Kirmani, 2000); as Fein (1996) and Uleman (1987) argues that when people making inferences about a specific object – that may affect one’s attitude and behavior ‒ it will require them a higher order cognitive process and a substantial amount of cognitive capacity to process it. Thus, individuals would be less likely to access and activate their knowledge of persuasion attempt when they have a high cognitive load; because, individuals will highly devote their mental resources to the other objectives (Campbell & Kirmani, 2000) instead of trying to access and activate their set of persuasion knowledge.

Amidst the increasing concerns over the deceptiveness nature of sponsored news article, Federal Trade Commission (FTC) has stated that it is essential for readers to be presented with cues that allow them to be aware with the intended message contained in the article (Hoy & Andrews, 2004; Wojdynski & Evans, 2016). FTC have strongly recommended that the paid nature of the sponsored news article should be disclosed and must be made salient before the text article so that readers will less likely being deceived of the messages and claims made therein (FTC, 2013) and this recommendation have been validated the Boerman, van Rejmersdal & Neijens (2014) study.

(13)

12

sponsored news article will actually result to a more positive attitude toward the advertising. It is in line with the prior study in the PKM literature which is studied by Friestad & Wright (1994) that suggested when customers can consciously and correctly interpret the advertising, it will decrease the likelihood of customers feeling deceived, and thus they will have a more positive attitude toward the advertising. The argumentation has also been validated by Campbell (1995) and Campbell & Kirmani (1995) as they argued that by extinguishing customers' suspicion against advertisers’ ulterior motives and manipulative intent, it would make customers’ more acceptable toward persuasion and thus having a more positive attitude toward the advertising.

Message Cue in Native Advertising and Persuasion Knowledge

Our knowledge about the effect of disclosure on customers’ attitude toward the native advertising has been enriched lately; however, little is known about how customers’ attitude toward the sponsored news article when they are not aware of the sponsorship disclosure made before the text article. In Wojdysnki & Evans (2016)'s study about the effect of sponsorship disclosure on sponsored news article on customers' recognition of advertising, out of 200 participants in their study, only 40% of the sample recognized that the sponsored news article is an advertising even if the sponsored news article has a "sponsored content" label embedded before the text article. One possible explanation to this phenomenon is that because readers just did not understand the word or even when they understood the words, they did not correctly apply the sponsorship disclosure to the particular sponsored news article's content (Wojdynski, 2016). This finding subsequently made us aware that the likelihood of customers being deceived by the message contained in the sponsored news article remains high.

(14)

13

message cues that use a negative comparison between one brand to the other, using an incomplete comparison, using a biased source in the advertising claim, and delayed sponsor identification will then make the manipulation intent more salient (Campbell, 1995) that could lead customers to form a negative attitude toward the advertisement (Campbell & Kirmani, 2000). The activation of the persuasive knowledge after being exposed to the message cue – that is deemed to be manipulative in nature ‒ will normally entails readers' suspicion against marketer's ulterior motives, skepticism against advertising claims and messages as well as customers' perception about advertisers as manipulative and deceptive (Kirmani & Zhu, 2007). As suggested before, if the suspicion against the sponsored news article is relatively high, readers will be more resistance to the persuasion and consequently will have a more negative attitude toward the sponsored news article (Campbel, 1995; Campbell & Kirmani, 2000; Wojdynski, 2016).

The extent of the manipulativeness and deceptiveness intent of a message cue can vary. For example, published by The Newyork Times, Orange is the New Black (Netflix Series)'s sponsored news article ‒ that talks about the growing numbers of women prisoners and the urge of making a program that meets the woman prisoner needs ‒ use a biased source (In which the Orange is the New Black's author) in giving an information about the experience of being imprisoned in the USA (Deziel, 2013). The instance, as mentioned earlier, deemed to have a biased source as the article stated that the author of the advertised TV series act as the information source for the respective news article's story. The aforementioned instance of biased source phenomenon is the same as instantiated in the Kirmani & Zhu (2007)’s study one in their paper. That is when the advertisers using their brand's study as the information source contained in the advertising message and this will make the manipulative intent more salient. On the other hand, an unbiased source will be instantiated when advertisers using an independent or unknown source of information in the advertising message and this will make the manipulative intent less salient for customers (Kirmani & Zhu, 2007).

Trust in Feelings and Persuasion Knowledge

(15)

14

of judgments and decisions” (Avnet et al., 2012, p.721); and subsequently use their feeling as a piece of information to make the respective judgment at hand (Schwarz & Clore, 1983).

The utilization of feelings as the source of information relies upon two conditions (Greifeneder, Bless & Pham, 2011); namely, the accessibility and the diagnosticity of feeling (Feldman and Lynch, 1988). On the one hand, people tend to have a greater influence over their feelings when they have limited access to processing the information, whether it is because of distractions (Albarracin & Kumkale, 2003), time constraint (Pham, Cohen, Pracejus & Hughes, 2001), or cognitive load (Shiv & Fedorikhin, 1999). On the other hand, the diagnosticity of feeling is the condition when the target could represent one’s momentary feeling (Pham, 2009) and one's representativeness of feelings needs to be seen as relevant with the judgment at hand (Pham, 1998). Additionally, from the recent study of Avnet et al., (2012), they argued that one's general trust in feelings could also be the determinant of the diagnosticity of feelings, which is the extent of one's believe in trusting their feeling when making judgments and decision. As Salovey, Mayer, Goldman, Tuvey & Palfai (1995) and Epstein, Seymour, Pacini, Denes-Raj & Heier (1996) documented in their studies that one's reliance to their feeling in making judgment and decision would vary.

Furthermore, it is important for us to study the process of how individuals are actually utilizing their feelings in making judgment as it will help us to construct the logic on why and when do people actually ignore or access their persuasion knowledge which subsequently could affect their reliance on feelings in making judgments. In this study, we will deliberately look into Avnet et al. (2012)'s framework on how actually a person relies on their feeling in making judgments.

(16)

15

a target - are to be credited to (representative of) the target itself. However, if there is enough and substantial alternative source of information to degrade the undesired influence of non-representative feelings (Albarracin & Kumkale, 2003), individuals would likely to discount their reliance of feelings in making judgments (Greifenender, Bless & Pham, 2011). Furthermore, if the alternative source of information is deemed to be not enough, the non-representative feelings will then act as one factor that affects the judgment at hand (Albarracin & Wyer, 2001). Lastly, to consider that feeling are diagnostic and consequently utilizing feeling as the main source of information, merely looking at the representativeness of feelings with the target would not be enough (Avnet et al., 2012). When the target has already represented feelings, individuals then must either perceive that their momentary feeling is relevant with the judgment at hand or individuals' are generally having a high trust in their respective feeling before they can finally rely on their feelings as a source of information in making judgments.

To be in line with our research question, we will solely focus on the extent of one's trust in their feeling in evaluating a target before finally judging the sponsored news article. Avnet et al. (2012) divided individuals general trust in feelings into two extremes, which are high trust in feelings and low trust in feelings people. For people that have high trust in feelings - regardless of the relevancy between the experienced feeling and the target - would very likely to rely on their feelings and tend to exclude another source of information in making judgments. While on the opposite extreme, people that have a low trust in their feelings, would very likely seek other information and tend to exclude feelings in making judgment. However, Avnet et al. (2012) argue that individuals need a substantial amount and significant process of other resources to degrade the influence of feelings in making judgments. Moreover, if they could not find enough sources to exclude their feelings in making judgments, feelings then also be included and be relied upon in making judgments.

(17)

16

As for individuals that have high general trust in their feelings, they are expected to be less likely to activate their set of persuasion knowledge even when the message cue made the manipulative intent salient. In this manner, individuals with high general trust in their feeling would very likely to exclude the external and alternative source of information ‒ which is the set of persuasion knowledge ‒ as they will normally rely on their internal input (i.e., their feelings) in making judgments at hand (Avnet et al., 2012). Additionally, it also requires a significant amount of resource (e.g., external information) processing that will allow the exclusion of feelings in making judgments (Avnet et al., 2012). In the other words, individuals with high general trust in their feelings are being less motivated to process the incoming information and thus will instead deemed that their internal affective inputs to be more diagnostic and thus will rely on their feeling in making judgments and decisions. In this way, given the condition of activating the persuasion knowledge required individuals to process the incoming information analytically as well as requiring individuals to utilize a higher order of their cognitive process (Campbell & Kirmani, 2000), it will consequently make people with a high trust in their feelings are unable to access and activate their set of persuasion knowledge in making judgment. Consequently, the negative effect of activating the persuasion knowledge will as well be extinguished. That is, individuals will have less suspicion about advertisers’ ulterior motives and manipulative intent and lead them to have a more positive attitude toward the sponsored news article (Kirmani & Zhu, 2007).

(18)

17

trigger them to access the other source of information and activating their persuasion knowledge. It is important to take note that the activation of persuasion knowledge will as well activate the negative impact that goes with the persuasion knowledge (Friestad & Wirght, 1994; Kirmani & Zhu, 2007). That is readers suspicion against advertisers’ ulterior motives and manipulative intent that lead readers to form a negative attitude toward the sponsored news article (Kirmani & Zhu, 2007) as they become more reluctant to the persuasion attempt (Friestand & Wright, 1994).

However, there are still possibilities, which the low trust in feeling people are not being triggered to activate their persuasion knowledge when being confronted by message cues (i.e., biased source and incomplete brand comparison). Because, they think that the other resource of processing the information simply does not sufficient to extinguish their reliance on feelings in making judgment. It could be because readers are not familiar and never been exposed to the sponsored news article as an advertising format before (Evans & Park, 2015). In this situation, readers will have no prior knowledge about that kind of persuasion attempt (Friestad & Wright, 1994). Hence, we propose that readers attitude toward the sponsored news article will be the same with the attitude formed by the high trust in feeling people, that is having a more positive attitude toward the sponsored news article.

From the constructed logic above, our proposed hypotheses are as follows:

❖ H1: In the presence of different saliency of manipulative intent in the message cue,

the high trust in feelings individuals would less likely to activate their set of persuasion knowledge and thus will have a more positive attitude towards the advertisement.

❖ H2: Trust is feelings moderate the association between the saliency of manipulative

intent on the attitude toward the advertisement.

(19)

self-18

regulatory focus as our exploratory variable to extend our explanation about why high (low) trust in feelings individuals are more (less) likely to detect the persuasion attempt.

We argue that individuals’ trust in their feelings could also be linked to their self-regulation in making judgments and decisions. We linked the high trust in feelings individuals with the promotion focus self-regulation individuals as both are having the same characteristics in their reliance on subjective affective inputs (i.e., their feelings) ‒ compared to the external information source ‒ in making judgments (Pham & Avnet, 2009). Mostly similar with the high trust in feelings individuals, promotion focus individuals are less motivated to process the incoming information analytically. Instead, they highly valued the speed of making judgments and thus will opt to effortlessly judge the information based on their feelings (Friedman & Förster, 2001). Additionally, the promotion focus individuals will deem that their internal affective inputs are to be more diagnostic (Pham & Avnet, 2004). Thus, they will tend to rely more on their feelings in making judgments and decisions. In contrast, the characteristics of the prevention focus self-regulation can also be linked with the low trust in feelings individuals as the prevention focus individuals will also rely more on the external information input (i.e., persuasion knowledge) in making judgment and decisions (Pham & Avnet, 2009). This is because, prevention focus individuals deemed that their affective inputs are less diagnostic, and thus will opt to rely on the external input that they deemed to be more diagnostic (Pham & Avnet, 2004).

However, it is essential to take note that individuals’ regulatory focus in this study would be treated as a mere exploratory variable. We do not include the regulatory focus theory and measures in our construct. Instead, we just suspect and argue that there is a possibility of individuals’ trust in feelings are related to their self-regulatory focus and thus will make us possible to add another explanation on how the reliance on feelings in making judgments and decisions between individuals with high and low trust in feelings differed. Thus, an additional hypothesis is constructed as follows:

❖ H3: There is a significant difference of individuals’ self-regulatory focus across high

(20)

19 CONCEPTUAL MODEL STUDY 1 Overview

The aim of this study is to test our hypotheses stated above. The role of individuals’ general trust in their feelings in judging the sponsored news article is being examined. This study would test how one’s general trust in their feelings could affect the activation of individuals’ persuasion knowledge that would subsequently affect their evaluations of the sponsored news article. Generally, the extent of individuals’ trust in their feelings would depend on the success of their previous episodes in trusting their feeling while making judgments and decisions (Avnet et al., 2012). Individuals with a greater extent of previous success in trusting their feelings while making judgments and decisions would generally believe that their feeling is trustworthy compared to the people that have a lesser extent of success in trusting their feeling while making judgments and decisions. Thus, participants’ successful history in trusting their feelings was manipulated experimentally by utilizing a procedure that is known as the ease of retrieval effect (Schwarz, Bless & Bohner, 1991) and afterwards, participants were asked to judge a one-page advertisement about a mirror-less camera.

‒ ‒ + Message Cue. Detection of Persuasion Intent. Attitude toward Sponsored News Article. Trust in Feelings.

(21)

20

From this study, we predicted that individuals with high trust in feeling conditions would be less likely to access their set of persuasion knowledge compared to the individuals with low trust in feelings. This conditions also applied even when the high trust in feeling individuals are being confronted with the high manipulative intent message cue in the advertisement. Hence, participants would have a more positive attitude toward the advertisement. Conversely, for individuals that are being experimentally manipulated with the low trust in feeling conditions, they would be very likely to access their set of persuasion knowledge. Consequently, the activation of the persuasion knowledge would then also activate the negative effect that goes with the activation of the persuasion knowledge, namely forming a negative attitude toward the advertisement. This condition is specifically expected to have a higher impact for participants that are confronted with the high manipulative intent advertisement as the high manipulative intent would make participants felt that they are being deceived. However, we expect that participants with the low trust in feeling conditions, that are being confronted with the less salient manipulative intent message cue, would have a more positive attitude toward the advertisement as they would be less likely to feel that they are being deceived by the advertisement.

Participants

(22)

21

participants are an internet bot existed in the Amazon Mechanical Turk platform. Through this procedure, we finally found as much as 219 valid participants.

From the cleansed data, our participants are consisting of 101 females (MFemale = 39.30,

SDFemale = 11.33) and 118 males (MMale = 35.92, SDMale = 10.02). Next, out of 219 eligible

participants’ responses, 183 participants or amounted for 84% of our sample, are Americans. While the rest of 36 (16%) participants are Indian, Indonesian, Dutch, British, and South African nationals. Additionally, as much 86% participants are aged from 19 to 50, while the rest participants are aged from and above 51 years old, which amounted for 14% of the total eligible participants.

N % Participants 219 Gender: Female 101 46% Male 118 54% Nationality: USA 183 84% Others 36 16% Age Group: 19 - 34 103 47% 35 - 50 86 39% 51 - 68 30 14% Age Range 49

Table 1: Participants demographic.

Procedure and Design

(23)

22

of persuasion knowledge. Afterwards, participants were asked to rate their overall thought and/or impression concerning the previously shown advertisement.

These ratings would then be utilized to construct our mediating variable about the participants’ detection of persuasion intent. Next, participants were asked to rate their overall attitude of the advertisements. These ratings are used to construct our dependent variable, namely attitude towards the sponsored news article. Then, participants were asked to rate on how much they rely on their feelings in judging the advertisement. These ratings were asked to check whether the trust in feelings manipulation had the intended effect. We expected that participants high on trust in feelings should report a greater reliance on feelings when judging the advertisement than participants in the low on trust in feelings condition. In the last part of the study, we measured participants’ self-regulatory focus, namely, promotion and prevention focus (Lockwood, 2002). As stated, these ratings would then be treated as a mere exploratory variable to check the associations of self-regulatory focus with our moderating variable, namely trust in feelings. The full document of our survey is presented in appendix 1 and appendix 2.

Experimental Manipulations

Trust in Feelings Manipulations

(24)

23

In this manner, we expect that participants who were asked to retrieve ten previous episodes in trusting their feelings while making judgments and decisions would infer that those situations are atypical. Hence, it would then create a perception in which participants felt that they have a lack of success in trusting their feelings while making judgments and decisions; consequently, it will degrade participants’ trust in their feelings. Conversely, participants that were asked to only retrieve two previous successful episodes of relying in their feelings while making judgments and decisions would infer that it is common. It then made the success of relying on feelings in making judgments and decisions become salient and in turn will increase their general trust in feelings (Avnet et al., 2012).

The trust in feelings manipulation is based on the ease-of-retrieval phenomenon to diversify participants’ extent of previous success in trusting their feelings while making judgments and decisions (Schwarz et al., 1991). In the original copy of Schwarz et al. (1991) study, it is evident that, compared to the participants who listed lesser situations in which they act based on their self-confident, participants that describe more situations are principally rated themselves as being less assertive. This is because participants who listed only six episodes inferred that the occurrence are relatively common while participants that listed twelve situations inferred in reverse (Schwarz, 2004).

Triggering the Activation of Persuasion Knowledge

(25)

24

source and incomplete brand comparison as untruthful and unbelievable compared to the advertisements that contains unbiased source claims and complete brand comparison. Based on the Kirmani & Zhu (2007) study one, we trigger participants’ set of persuasion knowledge through the exposure to the descriptive advertisement about the newly launched mirror-less camera called Oldig. This one-page advertisement included a headline, the camera picture, and message cues that is purposively made to trigger the activation of participants’ persuasion knowledge (see Appendix 1). We manipulated two different message cues ‒ through the source of consumers claim about the camera and brand comparison ‒ that both will similarly make the manipulative intent salient or not salient. For the experimental manipulation of the salient manipulative intent, we specifically use a biased source claim (i.e., “In a recent study conducted by Oldig…) and an incomplete brand comparison (i.e., “…consumers rated Oldig as producing better quality pictures than the leading brand.”). Additionally, minimum reading time constraint of four seconds is utilized to prevent participants just skim the advertisement instead of reading it carefully. In this way, we expect that the biased source would make the claims untrustworthy compared to the claim that is made by appropriate and independent organization (i.e., Consumer Reports). Furthermore, we also expect that the incomplete brand comparison would as well make the comparison more ambiguous compared to the comparison that specifies the leading brand name (i.e., Canon and Fujifilm).

Measures

Manipulation Check

(26)

25

Persuasion Knowledge

We adopted the measure of persuasion knowledge activation from the Kirmani & Zhu (2007) study. In this way, the activation of persuasion knowledge was measured through two different type of questions. The first one, participants were asked to rate the perceived deceptiveness nature of the advertisement by responding to the three questions of seven items Likert scale: “Unbelievable/Believable,” “Not truthful/Truthful,” “Not deceptive/Deceptive”, that latter we labelled these measures as “ARate1”, “ARate2”, “ARate3” respectively in our dataset. All items are reverse coded, meaning that the greater the value will express greater deceptiveness. We found that these measures are robust (α = .802).

Secondly, the persuasion knowledge activation measure is assessed through the thought protocol (Kirmani & Zhu, 2007). Participants were asked to give their overall thoughts, feelings, and impressions about the previously shown advertisement. Participants’ responses are manually coded as 0 = not feeling deceived/manipulated or 1 = feeling deceived/manipulated. That includes participants skepticism about the advertisement claim (e.g., the advertisement should specify to which leading brand that they are compared themselves to) and firm’s manipulative intent (e.g., the study is not conducted by an independent body). We labelled these coded responses as “Thought” variable in our dataset.

Advertisement Attitude

We adopted the attitude toward the advertisement (Aad) measurement from Muehling

(27)

26

Exploratory Variable: Regulatory Focus

Lastly, to make sure that individuals’ trust in their feelings is associated with the self-regulatory focus, we induced the measurements for participants’ self-self-regulatory focus. As stated, we expect that participants with a high trust in feeling conditions will also have a promotion focus self-regulation, while participants in the low trust in feeling conditions would rather have a prevention focus self-regulation. Again, it is important to take note that the self-regulatory measures are only for our exploratory variable. Hence, this variable will not include in our construct.

With a slight modification (i.e., we changed “school” terms into “life”), we adopted this measurement from Lockwood (2002) study. This regulatory focus measures consist of two different subset that will measure both regulator-focus differently, namely prevention focus and promotion focus. For instance, “I often think about the person I would ideally like to be in the future” statement will be endorsed by people that has a promotion focus while “My major goal in life right now is to avoid becoming a failure” statement will be endorsed by participants with a prevention focus (Lockwood, 2002). Participants were asked to rate to what extent that the eighteen statements of nine items Likert scale are true to themselves (1 = Very true, 7 = Not at all true). Subsequently, in our dataset, we labelled the subset that measures promotion focus with “Pro1” until “Pro9” and the other subset that measures prevention focus were labelled as “Prev1” until “Prev9”. We found that the promotion focus as well as the prevention focus measurement are still highly reliable to be included in our study (αpro = .928; αprev = .920).

N of items Cronbach α

Manipulation check (TiF) 2 .803

Persuasion knowledge (ARate) 3 .802

Advertisement attitude (AAd) 3 .956

Regulatory focus:

Prevention focus (Prev) 9 .920

Promotion focus (Pro) 9 .928

(28)

27

Result

Manipulation Check

First, we run a 2 (trust in feelings: High vs. Low) x 2 (saliency of manipulative intent in the message cue: Salient vs. Not Salient) factorial analysis of variance (ANOVA) to test the orthogonality of our independent variables. In other words, we expect that our independent variables are independent with each other. At the 95% confidence interval level, we found that there are no significant interaction effect or main effect (all p = .078 or higher). It means that, consistent with our predictions, our independent variables are independent between each other. The complete statistical result could be find in Appendix 3.

Second, we conduct an independent sample t-test to check whether there are differences in means on participants reported reliance on feelings in making judgments across different trust in feelings conditions (i.e., high trust in feelings vs. low trust in feelings. At the 95% confidence interval level, our result suggests that there are no significant differences in mean on participants’ reported reliance on their feelings in making judgment between the participants assigned in the low trust in feelings condition (MLTF = 5.07, SDLTF

= 1.30) and high trust in feelings condition (MHTF = 4.80, SDHTF = 1.59), t(217) = 1.33, p =

(29)

28

Hypotheses Tests

Hypothesis 1

Our first hypothesis suggests that in the presence of the different saliency of manipulative intent contained in the advertising message, the extent of individuals’ trust in their feelings will affect individuals’ ability in activating their set of persuasion knowledge and thus different attitude towards the advertisement will be formed. To test this hypothesis, we employ the PROCESS macro model 7 (i.e., for the “ARate” variable) and multiple regression analyses (i.e., for the “Thought” variable) as we have two different measure variables for participants’ detection of persuasion intent (Kirmani & Zhu, 2007).

Hypothesis test using “ARate” variable.

First, to test that trust in feelings (i.e., moderating variable) moderates the association between the saliency of manipulative intent in the message cue (i.e., independent variable) on the attitude toward the advertisement (i.e., dependent variable) through individuals’ detection of persuasion intent (i.e., mediating variable), PROCESS macro model 7 test is utilized. For this test, we use the “ARate” variable to represents participants’ reported detection of persuasion intent in the message cue. We employ the 10,000 bootstrap samples at the 95% confidence interval level for this test.

Path d Path b Path a Message Cue. Detection of Persuasion Intent. Attitude toward Sponsored News Article.

Figure 2. Statistical Model H1.

Path c’ (and c)

(30)

29

The test result reveal that neither the saliency of manipulative intent in the message cue (β = .24, SE = .25, t = .97, p = .34, 95% CI: ‒.25, .73) nor the trust in feelings (β = ‒.03,

SE = .25, t = ‒.11, p = .91, 95% CI: ‒.52, .46) variables are significantly predicting

individuals’ detection of persuasion attempt. Furthermore, the interaction effect between message cue and trust in feelings on the participants’ detection of persuasion intent also revealing an insignificant result (β = ‒.44, SE = .35, t = ‒1.27, p = .21, 95% CI: ‒1.22, .24). Considering we found no interaction effect from this test; we can then argue that there is no moderation effect of trust in feelings were observed between the association of message cue and participants’ detection of manipulative intent.

Next, controlling for the mediating variable (i.e., participants’ detection of persuasion intent), our result subsequently shows an insignificant direct effect of message cue in predicting participants’ attitude toward the advertisement (β = .04, SE = .14, t = .26, p =

.80, 95% CI: ‒.24, .32). However, our result reveal that, controlling for the different saliency

of manipulative intent in the message cue, participants’ detection of persuasion attempt could positively predict how participants forming an attitude toward the advertisement (β

= .82, SE = .06, t = 14.75, p < .001, 95% CI: .71, .93). The positive coefficient sign then

suggests that when participants felt that they are being deceived by the advertisement, the more negative attitude towards the advertisement will be formed by the participants. Furthermore, the conditional indirect effect of the saliency of manipulative intent in message cue on the attitude toward the advertisement also shows no sign of significant conditional indirect effect at the values of low trust in feelings (β = .20, SE = .18, 95% CI: -.18, .54) and high trust in feelings (β = ‒.16, SE = ‒.16, 95% CI: -.59, .28). Lastly, the result of the moderated mediation index also shows an insignificant result as the confidence interval did span zero (95% CI: ‒.90, .19). This result then revealed that there is no moderated mediation effect in our construct. Suggesting that the association between the saliency of manipulative intent in the message cue on the attitude toward the advertisement through participants’ ability in detecting persuasion intent does not significantly differs based on participants’ general trust in feelings.

Hypothesis test using “Thought” variable.

(31)

30

persuasion intent. Because the “Thought” variable is a dummy coded variable, we instead use the multiple regression analyses as PROCESS Macro does not allow a dichotomous mediating variable. Subsequently, we follow the assessment made by Baron & Kenny (1986) to check whether there is a moderated mediation effect in our construct.

Before we run the test, we multiply the saliency of manipulative intent and the trust in feelings variables first. This multiplication is done and is used to see if there is an interaction effect between the saliency of manipulative intent and trust in feelings in our simple moderation complete regression model.

First, we conduct a regression analysis to see if there is a direct effect of our independent variable (i.e., the saliency of manipulative intent in the message cue) on our dependent variable (i.e., attitude towards the advertisement). This test is done to later assess if there is a mediating effect in our construct. The result reveal that we found no significant direct effect of the saliency of manipulative intent in the message cue on the attitude towards the advertisement (β = .05, SE = .20, t = .25, p = .80, 95% CI: ‒.35, .44). In this manner, the result then reveal that we fail to fulfil one of Baron & Kenny (1986) conditions of a moderated mediation effect.

(32)

31

different trust in feeling conditions. Hence, we fail to fulfill Baron & Kenny (1986)’s subsequent assumption of moderated mediation construct (i.e., there must be a significant interaction effect between the independent and moderating variable on the mediating variable that later suggesting the presence of a moderation effect).

Subsequently, we conduct the third regression test to see the association between the mediating variable (i.e., “Thought” variable of participants’ detection of persuasion attempt) on the attitude towards the advertisement (I.e., path b in our statistical model) as well as the path c’ in our statistical model (i.e., the association of our independent variable on the attitude towards the advertisement when controlling for mediating variable). We regressed our independent variable (i.e., the saliency of manipulative intent in the message cue) and our mediating variable (i.e., participants’ detection of persuasion intent) on the attitude towards the advertisement. The test result reveal that, when controlling for the saliency of manipulative intent in the message cue, we found a significant direct effect between participants’ detection of persuasion intent on the attitude towards the advertisement (β = 2.11, SE = .17, t = 12.26, p < .001, 95% CI: 1.78, 2.44). The positive coefficient sign then suggests once participants’ felt that they are being deceived by an advertisement, the more negative attitude towards the advertisement will be formed. Conversely, we found no significant direct effect between the saliency of manipulative intent in the message cue on the attitude towards the advertisement when controlling for the mediating variable (β = .06, SE = .15, t = .37, p = .71, 95% CI: ‒.25, .36). In this manner, the test then suggesting that we met two of the Baron & Kenny (1986) assumption of the moderated mediation conditions.

(33)

32

reveals that the we could not find an enough evidence to state that the detection of persuasion intent mediates the relationship between the saliency of manipulative intent in the message cue on the attitude towards the advertisement at the values of different trust in feeling conditions.

Hence, from the serial tests that we conducted by using two different measures of participants’ detection of persuasion attempt (i.e., “ARate” and “Thought”), we can now fully reject our third hypothesis and instead accept the null hypothesis. That is, the association between the saliency of manipulative intent and the attitude towards the advertisement through individuals’ detection of persuasion intent does not significantly differs across all conditions of participants’ general trust in feelings. This latter suggests that trust in feelings has no effect in determining one’s activation of their persuasion knowledge before forming an appropriate (i.e., could be negative or positive) attitude towards the advertisement.

Hypothesis 2

Our second hypothesis argue that the different trust in feelings conditions (i.e., high vs. low trust in feelings) moderates the association between the saliency of manipulative intent in the message cue on the individuals’ attitude towards the advertisement. To test this hypothesis, we utilized the PROCESS macro model 1 of two-way simple moderation test.

(34)

33

This test latter suggests that the association between the different saliency of manipulative intent in the message cue on the attitude towards the advertisement is not buffered by the different trust in feeling conditions. In other words, individuals’ general trust in their feelings has no effect in determining participants’ attitude towards the advertisement in the presence of the different saliency of manipulative intent in the message cue. Hence, we can now fully reject our second hypothesis and instead accept the null hypothesis.

Hypothesis 3: Exploratory variable

As previously stated, we expected that regulatory focus will have a significant association with individual’s trust in feelings. To check our premise, a separate independent sample t-test of trust in feeling conditions (i.e., high vs low) on both prevention and promotion focus measures is conducted.

First, we want to check if there is a difference in means across the different trust in feelings conditions on the prevention focus measure. The independent sample t-test reveal that there are no significant differences between participants in the high trust in feeling conditions (MHTF = 4.96, SDHTF = 2.09) and participants in the low trust in feeling conditions

(MLTF = 5.20, SDLTF = 2.03) which can support our indications that participants in the low

trust in feeling conditions have a prevention focus (t(219) = .89, p = .97, d = .12). The result reveal that there are no difference between participants in the high trust in feeling conditions (MHTF = 6.84, SDHTF = 1.49) and in the low trust in feeling conditions (MLTF =

6.90, SDLTF = 1.54) which can support our indications that participants in the high trust in

feeling conditions have a promotion focus (t(217) = .30, p = .95, d = .04). Hence, we can now fully reject our third hypothesis. Instead, we can now state that there is no correlation on both promotion and prevention focus across all conditions of trust in feelings. In other words, there is no significant correlation between individuals’ trust in feelings and their self-regulatory focus.

Overall Result

(35)

34

By the means of serial tests that we conducted to check the hypotheses of this study, we can now say that we found not enough statistical justification to accept all of our hypotheses. However, considering the insignificant result that consequently showing an indication of our hypotheses are not supported, we found that there is a significant direct effect between participants detection of persuasion intent on the attitude towards the advertisement when controlling for the saliency of manipulative intent in the message cue. This result then suggests once participants felt that they are being deceived by an advertisement, they will form a more negative attitude towards the advertisement.

GENERAL DISCUSSION AND CONCLUSION

Considering many readers in the online news sites have failed to detect the true and paid nature of the sponsored news article – and subsequently failed to form an appropriate attitude toward the sponsored news article (Wojdynski & Evans, 2016) ‒, we built our focus of this study to analyze specific factors that could affect readers’ capability in activating their set of persuasion. We believe that our construct in this study is novel and thus It is important to take note of how our study differs from the previous study in the sponsored news article literature stream. Several reasons on how our study is different from the previous studies in the sponsored news article literature stream are listed below.

First, instead of utilizing the popular sponsorship disclosure made before the text article to trigger individuals’ activation of their persuasion knowledge (like what Lee et al., 2016; Van Reijmersdal et al., 2005; Wojdynski, 2016; Wojdysnki & Evans, 2016 used in their studies), we utilized the other means of when individuals could be triggered to activate the persuasion knowledge, namely message cues (Kirmani & Zhu, 2007) in the native advertising (e.g., Deziel, 2013). Second, while research that considers the psychological factors in the sponsored news article literature stream is still in its infancy (Wang, Xiong & Yang, 2018), this study is meant to answer that call for research. Hence, we consider individuals’ trust in their feelings as the psychological factor that could buffer individuals’ ability to activate their set of persuasion knowledge.

(36)

35

readers are mostly incapable of detecting advertisers’ persuasion intent and subsequently activate their persuasion knowledge to form the right attitude towards the sponsored news article. Without statistical evidence, Wojdynski & Evans (2016) suspected that readers inability to activate their set of persuasion knowledge ‒ even after the sponsorship disclosure is being made salient before the text article ‒ is simply because of readers are not paying attention or not fully understood the sponsorship disclosure made before the text article. Unlike Wojdynski & Evans (2016), we consider a psychological reason and suspecting that readers’ general trust in feelings as a factor that could better explain the why question in regards of the Wojdynski & Evans (2016) study result. We utilize the trust in feelings tenet as it is evident that nowadays, customers oftentimes utilize their feelings in making judgments and decisions (Avnet et al., 2012). Additionally, we aim to empirically proof our premise so that we have a better ground in explaining why readers are unable to detect a persuasion attempt in the sponsored news article.

In order for us to answer our research question, we built three hypotheses that were constructed from previous studies within the trust in feelings, persuasion knowledge, and native advertising (specifically in the sponsored news article) literature stream. We tested our hypotheses empirically by means of an experimental study. Our experimental study involved as much as 219 participants that were pooled from the Amazon Mechanical Turk platform. After we tested our hypotheses, in this section, we aim to answer our research question in general. That is “Are individuals that have a high trust in their feeling

(compared to individuals who have low trust in their feelings) are less likely to detect the persuasion attempts?”.

(37)

36

To begin with, our result reveals a shred of contrasting evidence to our hypothesis that we constructed earlier. Suggesting that, trust in feelings does not buffer the association between the different saliency of manipulative intent in the message cue on the attitude toward the sponsored news article through the detection of persuasion intent. Furthermore, we also found no evidence of the potential buffering effect of trust in feelings in the association between different saliency of manipulative intent contained in the message cue on how individuals formed their attitude towards the advertisement. In all, this suggest that the answer of our research question is that the different trust in feeling conditions (i.e., low and high trust in feelings conditions) does not affect individuals’ attitude formation in the presence of the different saliency of manipulative intent in the message cue.

Last, as we aim to explore an additional explanation of why the high trust in feelings individuals are less motivated in processing the incoming information, we consider and suspect that self-regulatory focus theory could aid the explanation. However, we found no statistical evidence to state that one’s general trust in their feelings is associated with their self-regulatory focus. Thus, our third hypothesis is rejected and instead accept the null hypothesis. That is, there is no significant difference among individuals’ self-regulatory focus across the different states of general trust in feelings.

Despite the fact that we could not find enough statistical evidence to support our hypotheses, we found support to confirm past studies inference. Our result reveals that once individuals felt that they are being deceived by an advertisement ‒ after they activate their set of persuasion knowledge ‒, it will lead them to form a more negative attitude towards the advertisement. This result then confirms the findings from several past studies (e.g., Lee et al., 2016; Wojdynski, 2016; Wojdysnki & Evans, 2016) in the sponsored news article literature stream. However, in this study, we fail to acknowledge what factor(s) that in turns trigger individuals’ activation of persuasion knowledge as we have no statistical support to infer our premise made in this study earlier.

Implications

(38)

37

advertising (specifically sponsored news article) literature stream, researchers that consider and study about the role of a psychological factor in the sponsored news article are still in its infancy (Wang et al., 2018). In other words, our study could be used as the basis for researchers in the native advertising (specifically sponsored news article) literature stream that aims to involve consumer psychological factors in their study. Second, as we fail to acknowledge sound reasoning on what factors are actually triggering individuals’ ability in detecting a persuasion intent, we found enough support to question Kirmani & Zhu (2007) argumentations. Namely, is the message cue in the Kirmani & Zhu (2007) study one’s instance (i.e., incomplete brand comparison, biased source claim (Campbell, 1995)) has been outdated that in turns not giving the intended effect to the persuasion knowledge activation? This research then could as well trigger advancement of the Kirmani & Zhu (2007) study.

Third, our study extends the generalizability of past studies inferences which conducted either in the persuasion knowledge literature stream (e,g., Friestad & Wright, 1994; Kirmani & Zhu, 2007, etc.) and native advertising literature stream (e.g., Lee et al., 2016, Wojdysnki, 2016). We found enough evidence to state that the activation of persuasion knowledge ‒ that usually entails individuals’ skepticism about the claims made in the advertisement ‒ made individuals perceived that they are being deceived, thus in turns will make them in forming a more negative attitude toward the advertisement.

Fourth, for the researchers in the trust in feelings literature stream, we have tried to extend the reach of this tenet by infusing the trust in feeling concept to the native advertising and persuasion theory. Our research could then be used as the antecedent study for researchers in this literature stream if they are interested in studying the effect of trust in feelings in an advertisement or in the persuasion knowledge context in general.

(39)

38

Second, the study result could then suggest advertisers ignore the potential effect of different message cue that might trigger the activation of persuasion knowledge (Kirmani & Zhu, 2007). As we found no evidence that the message cue could actually trigger the activation of persuasion knowledge, the subtle use of biased source claims ‒ as instantiated in the The Newyork Times branded content of Orange is the New Black sponsored news article (Deziel, 2013) ‒ could still be a safe way for advertisers to convey their messages in the branded content (i.e., sponsored news article).

LIMITATIONS AND FURTHER RESEARCH DIRECTIONS

We acknowledge that our study contains limitations in one aspect or another. In this section, we would like to specify the limitations of this study and thus specify the further research directions for this topic.

Limitations

We acknowledge that the participants pooled from the Amazon Mechanical Turk platform could affect our result of this study in general. Specifically, many respondents in the Amazon Mechanical Turk are a dedicated survey participant, thus it will make them determined to complete as many surveys as they can in a smallest time window possible. Consequently, participants might not carefully read the questions and instructions of the study that leads them to also not thoughtfully answer and follows the study instruction (Fleishcer, Mead & Huang, 2015). Additionally, it can also add an additional explanation of why our manipulation check is unsuccessful. That is, because participants might already been involved in the several other studies before completing the survey from this study, it will make them harder to be experimentally manipulated. Thus, it can impact our expected result of the study in general.

(40)

39

or even not familiar with camera at all. This could also affect our result as they might only giving a neutral answer in the study.

Next, we acknowledged that our main manipulation of trust in feelings only gave participants with 4 minutes to think of their previous success episode in trusting their feelings. This is originally different from what Avnet et al. (2012) did in their study. Specifically, Avnet et al. (2012) gave participants 7 minutes time span for participants to recall their previous success episodes in trusting their feelings. In this manner it might explain on why our manipulation check does not works that in turns might affect our result of this study.

Furthermore, due to the limited availability of time, writer individually coded participants’ responses (I.e., for the participants’ thought protocol measure, or as we labelled as “Thought” in our dataset). Therefore, it increases subjectivity of the coded responses that could also affect the result of this study.

Our study format could as well contribute to the limitations of this study. Our study is conducted in an online study format. This is fundamentally different from what Avnet et al. (2012) and Kirmani & Zhu (2007) did on their original study; which are conducted in a lab study design. While Avnet et al. (2012) and Kirmani & Zhu (2007) could better control for environmental confounds in their study setting, we could not control for environmental confounds in the online survey setting. This could also affect participants in being successfully manipulated or not. Therefore, this condition could affect our result in general.

(41)

40

Next, considering that the saliency of manipulative intent is subjective in nature (Hoch, 2002) and due to the fact that individual’s skepticism of advertisement claims might vary (Ford, Smith & Swasy, 1990), it could also affect our result. Specifically, it could affect our participants’ activation of persuasion knowledge and the effectiveness of our intended message cue to bring the saliency of manipulative intent. For instance, participant “A” might deem that a complete brand comparison is deceptive in nature, while for participant “B” the complete brand comparison is not deceptive in nature. This subjectivity difference between participants could then lead to an unintended result from participants after they are reading and processing the message cue presented in the study. Consequently, participants might infer differently from what it is previously studied by Campbell (1995) and Kirmani & Zhu (2007). Hence, it could affect our result in general.

Referenties

GERELATEERDE DOCUMENTEN

Until now this was done by conventional imaging (liver ultrasonography, chest X-ray and bone scintigraphy), but evidence favoring the use of FDG-PET/CT is accumulating. Methods:

Such profitability predictions of each of the actors involved, are determined taking into account given levels of uncertainty (expressed as probability distributions) at

De ondernemingsraad heeft instemmingsrecht op de spelregels bij het individueel roosteren maar niet op de roosters zelf, omdat deze individueel zijn – en dus niet betrekking

The Messianic Kingdom will come about in all three dimensions, viz., the spiritual (religious), the political, and the natural. Considering the natural aspect, we

Tijdige en adequate signalering, diagnosestelling en eventuele verwijzing naar een medisch specialist en/of kinderfysiotherapeut van zuige lingen met een voorkeurshouding en/of

• “The tendency of individuals to bias their evaluation of new information in favour of a preferred option by cohering new information to that preferred option.”. • Feelings

Congruent with this line of reasoning, the current study explores whether the knowledge reported by the members of one party - about the other party’s project team

literature, through increased communication, cooperation and effective dispute resolution, control as coordination mechanism will increase goodwill trust in an