國立臺灣大學管理學院商學研究所 碩士論文
Graduate Institute of Business Administration College of Management
National Taiwan University Master Thesis
他圖動機之易取性和消費者動機對說服性訊息處理之 影響-以部落格文章為例
The Effects of Accessibility of Ulterior Motive: the Moderating Role of Motivation to Process the Persuasive
Message of User-Generated Content in Blog Context
洪佳任 Jia-Ren Hung
指導教授：簡怡雯 博士 Advisor: Yi-Wen Chien, Ph.D.
中華民國 107 年 6 月
June, 2018
誌謝
一眨眼，兩年的研究所生活就來到了尾聲，而佔據碩二生活大半歲月的，
莫過於這篇論文了。碩二上開始跟老師meeting，確認題目、討論假設、設計實
驗、檢討等，但因為上學期同時兼職半星期的實習、在學校又當了兩門課程的 TA，時常覺得自己準備不足、進度緩慢。下學期實習暫停了、修的學分也變少 了，開始潛心投入論文當中，雖然老師、同門的支持從未斷過，但是論文的書 寫之路，仍然是也必須是孤獨的，一家又一家的咖啡廳、開滿分頁的網頁、腦 袋卡住的邏輯、海底撈針似的讀文獻...等，一點一滴漸漸堆砌出了成果。紀錄 這些其實是要寫給未來的自己看的，人會成長，但期許未來的自己不要輕看，
或者遺忘那些成長的軌跡。
感謝我的指導教授簡怡雯老師，讓我能夠享受寫論文的過程，聰明且溫暖 地帶著我們，總是給我們滿滿的肯定和鼓勵，讓我很有動力、也很自在地去寫 論文，除了學術上的學習，在老師身上也學到了許多處事為人的榜樣，謝謝老 師，何其幸運可以被您指導!感謝同門的依蒨、又銓，時而彼此激勵，時而互相 取暖，謝謝你們讓論文之路充滿笑聲。感謝我的父母，謝謝您們給我一個慈愛 和真理並俱的成長環境，花費無數的心思在我的身上，才有今天的我。感謝姊 姊、哥哥，你們也都是對論文不陌生的人，謝謝你們一直走在前頭為我帶路。
感謝柯冠州老師、羽均、少明，在行銷研究的課程中，協助我做論文實驗的 pilot study，也給了我許多寶貴的建議。感謝所籃的巨頭們和學長學弟們，一起 度過許多汗水和笑聲交織的時光。感謝商研所的同學們，能在這裡認識你們真 是人生中的美好。感謝教會的弟兄姊妹每個禮拜有聲無聲的代禱和祝福。最 後，感謝天父賜下的一切恩典，使萬事都互相效力，擴張我的境界。
摘要
隨著網路科技、社群平台的蓬勃發展，用戶原創內容不斷地在網路上大量 產生。本研究旨在探討在用戶原創內容的網路環境下，會影響消費者使用說服 知識的影響因素為何，並以部落格文章為例。本研究提出「消費者動機」和
「他圖動機之易取性」會交互作用影響消費者對於目標產品之態度。並證實了 對於高動機的消費者而言，他圖動機之易取性在對目標產品之態度上沒有影 響；而對於低動機的消費者而言，低他圖動機之易取性的情況將會導致對目標 產品之態度較佳於高他圖動機之易取性的情況。本研究也探討了產品「品質模 糊性」對於消費者態度改變之影響，並發現品質模糊性給予了消費者更多的空 間去產生不同的產品態度，在不同的情境之下。本文章將介紹我們所進行的先 導性研究和主要研究。
關鍵字:用戶原創內容、業配文、消費者動機、他圖動機之易取性、說服知識、
產品品質模糊性
ABSTRACT
As the booming of internet technology and social websites/platforms, huge amount of user-generated content has appeared online. Thus, this article aims to examine conditions that influence consumers’ use of persuasion knowledge to user- generated content in an online blog context. We propose that the motivation of
consumer and the accessibility of ulterior motive will interact to affect the consumers’
attitudes toward the target product. When the motivation of consumer is high, the accessibility of ulterior motives has no effect on the attitudes toward the target product; when the motivation of the consumer is low, they will form less favorable attitudes toward the target product under high accessibility than under low
accessibility. This article also introduces the effect of quality ambiguity on attitude change in an online user-generated content context, we found that the quality
ambiguity allows consumers to form diverse attitudes toward the target product under different conditions. A pilot study and a main study are conducted to support the predictions.
Keywords: User-Generated Content, Sponsored Content, Motivation of Consumer, Accessibility of Ulterior Motive, Persuasion Knowledge, Product Quality Ambiguity
CONTENTS
1. Introduction ... 1
2. Literature Review ... 4
2.1 Persuasion Knowledge Model ... 4
2.2 Sponsored Content on Blog ... 7
2.3 Motivation of Consumer... 10
2.4 Accessibility of Ulterior Motive ... 13
3. Hypothesis ... 17
4. Main Study ... 19
4.1 Pilot Study... 19
4.2 Pre-test ... 19
4.3 Procedure and Design ... 21
4.4 Experimental Manipulation ... 23
4.5 Measures ... 26
4.6 Results ... 28
4.7 Discussion... 33
5. General Discussion ... 34
5.1 Accessibility of Ulterior Motive versus Disclosure ... 34
5.2 Contributions... 36
5.3 Managerial Implications ... 39
5.4 Limitations and Future Research ... 40
6. References ... 43
7. Appendix ... 48
7.1 Questionnaire of Pre-test ... 48
7.2 Questionnaire of Main Study ... 51
LIST OF FIGURES
Figure 1 The Proposed Model of How Factors Influence Attitudes ... 18 Figure 2 Average Upper and Lower Bounds of Seven different restaurant types (quality map) ... 21 Figure 3 Attitudes towards the Target Restaurant as a Function of Motivation and Accessibility of Ulterior Motives ... 31 Figure 4 Estimated Probability of Sponsorship as a Function of Motivation and Accessibility of Ulterior Motives ... 35
LIST OF TABLES
Table 1 Average Differences between Upper and Lower Bounds across Different Restaurant Types 21
Table 2 Reliability Test of Measures ... 28
Table 3 Manipulation Checks of Motivation and Accessibility ... 29
Table 4 Manipulation Checks of Positive Messages and Universality of Sponsored Content Knowledge ... 30
Table 5 Descriptive Statistics of Attitudes towards the Target Restaurant ... 31
Table 6 Tests of Between-Subjects Effects ... 32
Table 7 Pairwise Comparisons ... 32
1. Introduction
People use persuasion knowledge every day, from a daily conversation with
friends to an advertisement presented in public. Persuasion Knowledge Model
postulates that consumers develop knowledge about persuasion and then use this
knowledge to respond to persuasion attempts so as to achieve their own goals
(Friestad and Wright 1994). Understanding why, how, and when consumers use
persuasion knowledge has always been of interest in past researches. However, the
explosive increase in marketing scenes resulting from internet popularity exceeds
the speed at which research could keep up. Therefore, this research aims to explore
the consumers use of persuasion knowledge to a user-generated content in a blog
context. The choice of the marketing scene to be conducted in this study were
made by observing the trend of an increasing amount of sponsored blog content in
Taiwan.
The purpose of this article is to identify and verify the factors that influence
consumers’ use of persuasion knowledge in a user-generated blog content setting.
We would like to know the internal and external factors that are likely to affect
consumers’ attitudes toward products when they browse persuasion messages in
the context of user-generated content on blogs. Two factors are proposed in this
study: the accessibility of ulterior motives and the motivation of consumer. The
basic proposition is that when the motivation of consumer is high, they are more
likely to use persuasion knowledge to infer the ulterior motives behind the
persuasive messages regardless the accessibility of the ulterior motives. In contrast,
when the motivation is low and the ulterior motive is less accessible, persuasion
knowledge will be less likely to be activated and used to interpret the persuasive
content.
This research contributes to both the Persuasion Knowledge Model and the
studies of sponsored content. For the PKM (Friestad and Wright 1994), we applied
the existing research on persuasion knowledge in an online user-generated contents
context, and further proved that the motivation of consumer and the accessibility of
ulterior motives are two influential factors in forming the attitudes toward the
target product in a web-browsing setting. For the studies of sponsored content, we
transferred the focus from “disclosure/non-disclosure” to “level of accessibility of
ulterior motives”, which closed to more real situations in plenty forms and topics
of sponsored content that still favor the non-disclosure practice. The current
research also provides applicable managerial implications that help marketers
better utilize the user-generated content as a promotional tool. They must be aware
of the negative effects caused by the accessibility of ulterior motives, and take the
motivation of consumer into consideration. In addition, we suggest marketers to
consider the product ambiguity beforehand since the proven effect that could have
on attitude change, that is, relatively ambiguous products are more likely to allow
positive influence by user-generated content.
2. Literature Review
2.1 Persuasion Knowledge Model
“One of a consumer's primary tasks is to interpret and cope with marketers'
sales presentations and advertising. Over time consumers develop personal
knowledge about the tactics used in these persuasion attempts. This knowledge
helps them identify how, when, and why marketers try to influence them. It also
helps them adaptively respond to these persuasion attempts so as to achieve their
own goals.” (Friestad and Wright 1994, p1). Based on this model, targets (people
for whom a persuasion attempt is intended) will respond to a persuasion attempt
against agents (who designs and constructs a persuasion attempts) by persuasion
coping behaviors (i.e., to contend or strive) in a persuasion episode. Consumers’
knowledge about persuasion includes beliefs about persuasion goals, in other
words, the motives underlying a persuasion attempt, inclusive of acquiring
information or physical objects, getting permission, changing one’s opinion,
selling something to one, changing an existing relationship, changing one’s
personal habit or characteristic, etc (Rule, Bisanz, and Kohn 1985). Consumers’
knowledge structures also include belief about persuaders’ methods, namely, the
persuasion tactics used on targets, inclusive of indirect asking, invoking role
relationship, informing personal reason, bargaining favor, threatening, invoking
personal expertise, deceiving, etc. In brief, persuasion knowledge includes ideas
about persuasion motives which the influence agent is attempting to achieve, as
well as ideas about persuasion tactics, that is, how the agent tries to achieve it.
Both of persuasion motives and persuasion tactics are of interest to this study.
From the target’s point of view, the Persuasion Knowledge Model presumes
that targets are motivated to use their persuasion knowledge to generate a valid
product and agent attitudes, they will try to allocate cognitive resources efficiently
to this task and use whatever information seems helpful, given the information-
processing constraints they face (Chaiken et al. 1989; Petty and Cacioppo 1986).
Past researches clearly showed that consumers will use cognitive resources only if
they are motivated to, in other words, motivation of the consumer plays a
prerequisite role in determining the amount of the cognitive resources being used.
In prior research, the mental state in which an individual actively entertains
multiple, plausibly rival hypotheses about the motives or sincerity of a person’s
behavior has been defined as “suspicion” (Fein 1996). Fein, Hilton, and Miller
(1990) found that participants were less likely to draw correspondent inferences
when contextual information suggests that multiple rival motives could underlie an
actor’s decision to behave in a particular manner. Jones, Davis, and Gergen (1961),
for example, found that participants who had reason to suspect that a job candidate
may have been affected by ulterior motives when responding to questions about his
personality tended to discount the candidate's statements and refused to draw
correspondent inferences. The same reasoning can be applied to the field of
consumer research. Generally speaking, if a consumer has observed ulterior
motives underling marketing content (which often conveys positive messages
toward the product), s/he is likely to discount, or more precisely, correct the initial
correspondent inference about the target product against the messages received
(Gilbert and Malone 1995; Gilbert, Pelham, and Krull 1988).
In this study, we particularly focus on a setting that consumers are
encountered more and more often than before—user-generated content in blog
context. It starts to prevail as a new marketing tactic in recent year, especially in
Taiwan. However, how Persuasion Knowledge Model is applied to a user-
generated content web-browsing setting is still understudied. Therefore, this study
aims to explore, preliminarily, factors that affect consumers’ use of persuasion
knowledge in a web environment. We propose that two factors—the motivation of
the consumer and the accessibility of ulterior motive—will interact to affect
consumers’ use of persuasion knowledge to make inferences about persuasion
motives underlying the user-generated content. We first look back previous
researches about sponsored content before discussing these two factors.
2.2 Sponsored Content on Blog
Native advertising is a means of presenting consumers with a commercial
persuasive message that resembles the non-third-party content provided by the
same publisher, which can be presented in a variety of formats through different
channels, such as one-off videos, pictures, series of articles, social media posts, or
audio in terms of formats (Faber, Lee, and Nan 2004; Rosin 2015; Wojdynski
2016). Any two examples of native advertising can differ widely in terms of scope,
scale, media employed, nature of the content, immediate goal, and so on. Though
the foci of definitions to native advertising vary by parties, the central to each of
these definition is the notion of relatively seamless integration of paid content with
other non-paid content, namely, there is no distinction between commercial content
and real or authentic opinions, feeling, and experiences of the journalists or senders
(Chia 2012; Pollit 2015). The literature shows that disclosures of native advertising
can activate persuasion knowledge and eventually mitigate persuasion (Boerman,
van Reijmersdal, and Neijens 2012; Nelson, Wood, and Paek 2009; Tessitore and
Geuens 2013; van Reijmersdal, Lammers, Rozendaal, and Buijzen 2015; Wei,
Fischer, and Main 2008). In this study, we focus on a specific form of native
advertising—user-generated article in the context of blog.
Blogs represent a new sphere of communication which connects the marketer
and consumers through an online community platform (Palmer and Koenig-Lewis
2009) Therefore, an effective blog marketing requires balancing the benefits of the
blogger, the sponsor, and the blog reader. First, bloggers need to create whatever
content that is useful and reach a certain reader base, and might therefore welcome
company-created content that fits the blogger’s interests. Past studies for sponsored
blog content found that overt marketing has a negative effect on behavioral
intentions, such as future interest in the blogger, intention to engage in word-of-
mouth, and purchase intention. Covert marketing did not affect the intended
behavioral. (Liljander, Gummerus, and Söderlund 2015). Reijmersdal et al. (2016)
found that when readers are exposed to a sponsored blog with disclosure, their
persuasion knowledge is activated to trigger cognitive and/or affective resistance
against the user-generated content, thus lead to more negative brand attitudes and
lower purchase intention. Studies unanimously agree the negative effect of
disclosures showing lower perceived credibility of the blog and the blogger, more
negative attitudes toward the blog (Colliander and Erlandsson 2015). However, we
found that there is a certain kind of sponsored blog article which became prevalent
in Taiwan in recent years—dining brief. Dining brief used to be a kind of
restaurant-visiting note written by ordinary public that truthfully reports the
restaurant features, food quality, price information, service level and so on.
However, restaurant owners came to realize that it can be used as a marketing tool,
thus make deal with well-known bloggers, ask them to write good things about the
restaurant by giving reward in return, for instance, free meals and money. As a
result, it becomes difficult for readers to discern between sponsored and non-
sponsored reviews.
The booming of sponsored articles in a blog context is especially pronounced
in Taiwan; however, it rarely discloses the fact of sponsorship behind. Previous
studies focused on how disclosure can affect persuasion and attitude toward the
blogger and brand (Boerman, van Reijmersdal, and Neijens 2012; Nelson, Wood,
and Paek 2009; Tessitore and Geuens 2013; van Reijmersdal, Lammers,
Rozendaal, and Buijzen 2015; Wei, Fischer, and Main 2008), but in the cases
where there is no disclosure, readers could only make inference on their own. On
the other hand, previous researches focused more on the external, situational
factors, such as disclosure/non-disclosure, duration of the disclosure (Boerman et
al. 2012), disclosure time points (Reijmersdal et al. 2016) and so on. The internal,
individual factors that could consistently influence a reader in a sponsored blog
context are however, less studied. In this study, we propose that two factors—the
consumer’s motivation and the accessibility of ulterior motives—will interact to
affect consumers’ inferences about persuasion motives underlying the user-
generated content, and ultimately affect the judgment of the restaurant.
2.3 Motivation of Consumer
“People are motivated to hold correct attitudes.” (Festinger 1950), since
incorrect attitudes are generally maladaptive and can have damaging behavioral,
affective, and cognitive consequences. Although people want to hold correct
attitudes, the amount and nature of issue-relevant elaboration in which people are
willing or able to engage to evaluate a message vary with individual and situational
factors (Petty and Cacioppo 1986). When conditions foster people’s motivation
and ability to engage in issue-relevant thinking, the “elaboration likelihood” is said
to be high, in other words, the likelihood of elaboration is determined by a person’s
motivation and ability to evaluate the communication presented (Petty and
Cacioppo 1986). Elaboration Likelihood Model clearly stated that motivation and
ability play as two major roles in determining the level of argument scrutiny. As
motivation and/or ability to process arguments is decreased, peripheral cues
become relatively more important determinants of persuasion (peripheral route),
that is, attitudes are determined by positive or negative cues in the persuasion
context which either become directly associated with the message position or
permit a simple inference as to the validity of the message. Conversely, as
argument scrutiny is increased, peripheral cues become relatively less important
determinants of persuasion (central route), attitude changes are instead based on a
thoughtful consideration of issue-relevant information and an integration of that
information into an overall position. Thus, attitude changes prompted by the
central route involve considerably more cognitive work than attitude changes
induced under the peripheral route. However, according to the concept map
presented in ELM (Petty and Cacioppo 1986), when people are encountered with
persuasive communications, the first question to be asked is “Is the consumer
motivated to process?” then the next question “Is the consumer able to process?”.
This shows that motivation plays as a prerequisite role in determining the level of
argument scrutiny, and further affects the attitude changes prompted by
central/peripheral route. In other words, if we control the factor “ability” equally
enough, motivation shall be the only internal factor that can affect the level of
consumer involvement. Therefore, the more motivated consumers are to assess the
central merits of the target product (i.e., determine the true quality), the more likely
they are to effortfully scrutinize all available object-relevant information, thus hold
more accurate attitude toward the attitude object.
In some persuasion communication, there could be biasing factors that affect
consumers; for example, an expert source might bias processing of the verbal
arguments presented (Chaiken and Maheswaran 1994). According to the Flexible
Correction Model (FCM; Wegener and Petty 1997) corrections can proceed in
different directions depending on recipients’ theories of how the biasing event or
stimulus (e.g., an attractive source) is likely to influence their views. FCM clearly
stated that, to make correction occur, people should: (a) be motivated and able to
identify potentially biasing factors, (b) possess or generate a naive theory about the
magnitude and direction of the bias, and (c) be motivated and able to make the
theory-based correction. Both condition (a) and (c) stress the importance of
motivation in the bias correction process. In this study, we choose motivation to be
the manipulation variable rather than ability. The decision is made according to
two reasons. First, the PKM (Friestad and Wright 1994), the ELM (Petty and
Cacioppo 1986), and the FCM (Wegener and Petty 1997) all stress the prerequisite
feature of motivation in a persuasion episode/communication, especially in
determining the amount of cognitive resources, which will further affect level of
scrutiny and use of persuasion knowledge. Second, we assume that it is less likely
for the participants (consumers) to lack ability in the setting of this study, since our
target product is in the food category and participants are assumed to be rather
familiar with this product. Thus, in the current study, we assume that the ability to
process is equal across all conditions, and manipulate high vs. low motivation to
process to examine our current predictions.
In a user-generated blog article web-browsing setting, consumers with high
motivation to process (high involvement) tend to scrutinize all the information, and
are more likely to use persuasion knowledge to cope with the persuasion attempts
(e.g., leave consumers positive impression on the target product), on the other
hand, consumers with low motivation to process (low involvement) tend to rely
more on peripheral cues (e.g., the salient hashtags below the article), and are less
likely to use persuasion knowledge to cope with the persuasion attempts.
According to the reasoning above, it seems that low involvement consumers will
always be deceived by sponsored content in a blog context, however, there is
another factor in this study—the accessibility of ulterior motives—can interact
with consumer’s motivation to affect the consumers’ use of persuasion knowledge
and inferences about persuasion motives underlying the user-generated content,
thus ultimately affect the judgment of the target restaurant.
2.4 Accessibility of Ulterior Motive
Accessibility denotes the ease or speed a construct is coded in terms of a
given category under varying conditions (Higgins 1989). The accessibility of a
construct is affected by various factors, including expectations, strength of
association, frequency of activation, and recency of activation (Higgins and King
1981) Thus, the accessibility of ulterior motive is likely to be affected by how
strongly associated that motive is with the influence agent. In an interpersonal sales
setting, such as a clothing store interaction between a salesperson (the influence
agent) and a consumer (the influence target), a compliment from a salesperson to
the consumer is very likely to be insincere (Campbell and Kirmani 2000), that is,
an ulterior motive- “trying to make a sale” may exist. Rather than the obvious
meaning of the compliment, which might be “You look perfect in this jacket.”, the
ulterior motive is often much more underlying, implicit, and difficult to detect.
Hence, when the accessible motives come readily to mind, it is likely to
require less cognitive capacity to use persuasion knowledge in order to deal with
the situations, agents, or tactics that are strongly associated with persuasion
motives than that are more weakly associated with persuasion. Research shows that
one factor that affects an individual’s cognitive capacity is the individual’s role or
perspective, the targets are likely to be more cognitively constrained than are
observers because s/he devotes mental resources to the interaction (Gilbert, Jones,
and Pelham 1987; Gilbert et al. 1988). However, the cognitive resources that a
consumer would invest in is determined by the level to which s/he is motivated, for
the purpose of forming an accurate impression of the target object (Petty and
Cacioppo 1986). Furthermore, in a web-browsing setting, which is also the main
concern of this study, the most influential difference is that consumers no longer
need to interact with others, in most cases, they even don’t need to speak or
respond to any information they’ve received. It is unlikely for them to lack
cognitive resources because of their cognitive busyness. Therefore, it is more
reasonable that the involvement level of a consumer is affected by motivation than
individual’s role or perspective. Namely, consumers with high motivation are more
likely to use persuasion knowledge during web-browsing than those with low
motivation. On the other hands, most of dining briefs do not disclose the fact of
sponsorship, however, differences in the accessibility of ulterior motives still exist
between different blog content. The ulterior motives would make consumers
perceive the message source as less sincere (Campbell and Kirmani 2000) and thus
correct the potential bias lies inside the message when forming attitudes toward the
target product. It is predicted that, although consumers with low motivation are less
likely to activate persuasion knowledge, high accessibility of ulterior motives is
able to trigger their spontaneous corrections and makes motivation unnecessary for
persuasion-knowledge application. However, when the ulterior motives are less
accessible, consumers with low motivation would become less likely to apply their
persuasion knowledge, thus form more favorable attitudes toward the target
product than under high accessibility. On the other hand, it’s predicted that
consumers with high motivation would activate persuasion knowledge regardless
of the accessibility of ulterior motives, thus form similar attitudes eventually.
3. Hypothesis
The current research proposes that, both the accessibility of the ulterior
motive and the motivation of consumers will influence the likelihood that a
consumer makes an inference of ulterior motives. When an ulterior motive is not
highly accessible, such as when the sponsored content on the website is not
strongly associated with interests exchanging motive (an exchange of interests
between the bloggers and restaurants), persuasion knowledge is less likely to be
used and activated by low motivation consumers That is, low-motivation
consumers are less likely to draw an inference about the ulterior motive; they tend
to perceive the positive messages conveyed in the sponsored content as sincere and
thus form higher ratings of the target product. In contrast, high motivation
consumers, who are willing to exert more cognitive resources to make accurate
decisions, are more likely to use persuasion knowledge to suspect and infer that the
blogger is motivated to write the positive content in exchange for some kind of
reward from the product supplier, given that more and more contents on blogs have
been found to be sponsored by the suppliers.
When an ulterior motive is highly accessible, such as when the sponsored
content on the website is strongly associated with interests exchanging motive,
motivation is no longer a prerequisite for persuasion-knowledge application. In this
case, both the high motivation consumers and the low motivation consumers are
likely to correct their perception of the bloggers by inferring an ulterior persuasion
motive and thus form lower ratings of the target product. This leads to the
following hypothesis:
H1: For high motivation consumers, they are more likely to use persuasion knowledge to infer ulterior motives of the blogger under both high and low
accessibility. Thus, there is no difference in the product attitudes between low and
high accessibility conditions.
H2: For low motivation consumers, they are less likely to use persuasion knowledge to infer ulterior motives of the blogger under low accessibility than
under high accessibility. Thus, product attitudes will be more favorable under low
accessibility than under high accessibility.
The Proposed Model of How Factors Influence Attitudes Figure 1
4. Main Study
4.1 Pilot Study
We conducted a pilot study and a pre-test before the main study. In the pilot
study, we set the target restaurant, which consumers were going to judge after
browsing the user-generated content on the blog, as a seafood buffet; however, we
found no significant main effect nor interaction effect of motivation and
accessibility on the attitudes towards the target restaurant. Therefore, we infer that
the relatively unambiguous nature of target restaurant neutralized the manipulation
effect, since the seafood buffets in Taiwan are easily associated with images of
high-end restaurants, the accessibility of ulterior motives and the motivation of
consumers become less crucial factors in judging the true merit of the target
restaurant. In other words, even if the reader has already perceived the ulterior
motive underlying and the potential bias existing, it’s still likely for her/him to
hold the anchor provided by prior experience (Helson 1964; Oliver and Linda
1981). Therefore, we conducted a pre-test that directly measured the quality
ambiguity and quality ranges of different types of restaurant in Taiwan.
4.2 Pre-test
The participants were 20 MBA students from National Taiwan University,
who were all between 21-30 years old and had a chance to win a lottery as an
incentive of the test. At the beginning of the test, participants were presented with a
diagram which clearly showed the judgmental quality range of different
smartphone brands, after that, they were asked “According to the above example,
please evaluate the upper/lower bound of ‘overall quality’ in the following types of
restaurant based on your ‘personal life experience in Taiwan’.” Seven different
common restaurant types in Taiwan were evaluated by participants. The average
upper/lower bound and average differences between upper and lower bounds are
presented below. The average differences showed that steak house, shabu-shabu,
and beef noodles were three most ambiguous restaurant types compared to others.
However, steak house was regarded as a high-end restaurant according to the
quality map (which is closer to the right side), hence, shabu-shabu and beef
noodles would be more suitable options to be the target restaurant in the main
study. Considering the popularity in Taiwan, we chose beef noodles as the target
restaurant. We believe that the relatively ambiguous nature and the moderate
quality range would fix the problems we encountered in the pilot study.
Average Upper and Lower Bounds of Seven different restaurant types (quality map) Figure 2
Average Difference
Ordinary Buffet 3.75
Steak House 4.40
Beef Noodles 4.15
Seafood Buffet 3.65
Traditional Diner 4.10
Shabu-shabu 4.30
Stir-fried Food 3.65
Average Differences between Upper and Lower Bounds across Different Restaurant Types Table 1
4.3 Procedure and Design
The purpose of this study was to test the hypothesis that motivation and
accessibility would interact to affect the judgment toward the target product. The
experiment employed a 2 (motivation of consumer: low vs. high) x 2 (accessibility
of ulterior motives: low vs. high) between-subjects design. Subjects were 80
undergraduate and graduate students from National Taiwan University, who would
have chance to win a lottery (500 NTD voucher of a department store) as an
incentive. Subjects entered the online questionnaire website through the link we
posted on “NTU student forum” fan page on Facebook. The incentive, estimated
time required, and main theme of the research are stated in the instruction.
Participants were randomly assigned across treatments by the time spot they
clicked the link. They all knew that there is no time limit for the entire experiment
before starting.
Once participants entered the online questionnaire, they were told that they
would read a scenario on the next page. After reading the scenario, participants
were presented with a user-generated article in a blog context, which is a dining
brief about a restaurant. The blog was a forged one to control any potential
“particular social website effects” (e.g., Pixnet is known for sponsored dining
brief). Judgments of the restaurant and several questions to check the
manipulations were asked page by page after the scenario and the article. Their
authentic behavioral responses and judgments were requested at the beginning of
the experiment in order to minimize the Hawthorne effect, which makes
individuals modify their behavior in response to their awareness of being observed
(Adair 1984). Participants could not return to previous pages once proceeding
to next one. The scenarios and blog articles are in appendix.
Keep quality ambiguity in mind, we chose beef noodles as the target
restaurant for instead due to its diverse quality across Taiwan, the ambiguous
nature and the moderate quality range renders readers more room to form different
attitudes, and allows accessibility of ulterior motive and motivation of consumer to
interact to affect the attitudes toward the target restaurant.
4.4 Experimental Manipulation
Motivation of consumer was manipulated mostly by the scenario the
participant received. Half of the participants, who were assigned to high motivation
treatment, were gave a scenario, “A very good friend of you is going to study
aboard soon, there will be five years away from Taiwan. In the few days before
s/he left, you two planned to meet for the farewell dinner. S/he also specified that,
s/he wants to eat the traditional food—beef noodles, which is rare in the country
s/he is heading to. Being friends for many years, you really value this relationship,
therefore, you go online to search for beef noodles restaurants, and then you enter
this site…”. The scenario could be easily imagined by NTU students since many
similar events happened around their lives frequently. The other half of participants
were also given an easy-to-imagine scenario, which intend to lower their
motivation, “You are a graduate student and share a research room with five other
students. When you were leaving the room as the last one tonight, you found that
the computer of the one sitting next to you was still on; obviously it’s a careless
mistake. You kindly decided to help turn it off. And you found that her/his screen
showed the following article, you knew this was probably related to where s/he just
decided to go for dinner” This scenario leads to lower personal relevance, which
occurs when people expect the issue “does not have significant consequences for
their own lives” (Apsler and Sears 1968), thus decreases the motivational level
effectively (Petty and Cacioppo 1986). In addition to different scenarios imposed
on participants, we added a statement in red at the beginning of the high motivation
questionnaire, “This study collects a small number of samples. Your opinions will
have a significant impact on the overall experiment”. In this way, we strengthen the
manipulation of motivation further.
Accessibility of ulterior motives was manipulated by the blog content that
presented to the participants, which was right on the next page of the scenario
manipulation. However, to ensure that divergent judgments were not attributed to
different content which participants received, we remain the issue-relevant
information (the article itself) unchanged between treatments (Petty and Cacioppo
1986). Apart from the article, a photo of the beef noodles, sticker and ID of the
blogger are presented in the interface of the fake blog “Flogger”. For participants
who were assigned to the high accessibility of ulterior motive manipulation, a line
of text was placed beside the blogger’s sticker, “Welcome restaurants call to
discuss on business cooperation”, and three hashtags were placed right below,
“#small to medium-size restaurant marketing”, “#new product review”, and
“#service experience promotion”. Furthermore, we placed a digital ID signature of
the blogger “Eat_Jacky” on the beef noodles photo, which is a common practice
for famous bloggers to create personal identification (However, famous bloggers
are more likely to be sponsored for writing content on a blog). We expected that
those embedded cues would raise the accessibility of ulterior motives (e.g.,
interests exchanging) underlies the blogger, comparing to the content without it.
Different from direct disclosure, these cues aroused consumers’ suspicion and thus
elicited the use of persuasion knowledge to infer ulterior motives of the blogger.
Consumers make inferences based on internal conditions and external factors
rather than directly being disclosed of the fact of sponsoring.
To ensure the hypothetical reasoning valid, which assumes that whoever
believes the more the messages conveyed in the user-generated content, the higher
her/his rating of the restaurant would be. We controlled all the messages to be
positive, namely, there is entirely no negative message in the article promoting the
restaurant. In this way, consumers who do not use persuasion knowledge to infer
ulterior motives of the blogger will not correct the potential bias included in the
content and will further form more favorable attitudes toward the restaurant.
4.5 Measures
Attitudes towards the Target Restaurant. Attitudes towards the restaurant
was measured as an average of six seven-point scales, for examples, “What is your
overall assessment of this restaurant?” (1 = very bad, 7 = very good), and “How do
you expect the quality of the restaurant?”. Factor analysis showed that the scale
was unidimensional, and Cronbach’s alpha was .90.
Manipulation Checks. The manipulation check for motivation of consumer
was a set of self-rating items that evaluate the involvement level. Participants
responded to several questions assessing intrinsic importance and personal
relevance of the blog content (Sherif and Hovland 1961), for example, “How
important the information provided in the article is to you?” (1 = very unimportant
to me, 7 = very important to me), or “How relevant it is for you to have a dinner
with an old friend who is going to study abroad?” (1 = very irrelevant to me, 7 =
very relevant to me). Factor analysis showed that the scale was unidimensional,
and Cronbach’s alpha was .90.
Second, the accessibility of ulterior motive was assessed followed by checks
of motivation of consumer. The manipulation check for accessibility was
participants’ level of agreement with a set of statements, for instance, “The author
wrote the online article because there was some kind of interest-exchanging
relationship with the restaurant” (1 = completely disagree, 7 = completely agree),
or reverse item like, “The author wrote the online article because s/he wanted to
share the experience of visiting this restaurant objectively”. Factor analysis showed
that the scale was unidimensional, and Cronbach’s alpha was .89.
In the end of the questionnaire, participants were asked “According to your
impression, the messages conveyed in the article are generally…?” (1 = very
negative, 7 = very positive). In addition, to make sure the universality of the
knowledge about sponsored content, the understanding of sponsored article was
assessed, “How is your understanding of sponsored articles?” (1 = very little, 7 =
very well).
Measure Cronbach’s Alpha N of Items Items
Attitudes .902 6
(1) 請問您對於這家餐廳的整體評價為何?
(2) 請問您對於這家餐廳的喜好度為何?
(3) 請問您對於這家餐廳的觀感為何?
(4) 請問您認為這家餐廳的品質如何?
(5) 請問您對於這家餐廳的造訪意願為何?
(6) 請問您預期這家店所賣的牛肉麵會如何?
Motivation .897 5
(1) 請問您認為「與即將出國留學的同學吃飯」和您的相關程度？
(2) 請問您認為「與即將出國留學的同學吃飯」對您的重要程度？
(3) 請問您認為「為您朋友選到一家好的牛肉麵」與您的相關程度?
(4) 請問您認為「為您朋友選到一家好的牛肉麵」與您的重要程度?
(5) 請問您認為「這篇文章的資訊」對您的重要程度?
Accessibility .891 4
(1) 作者寫這篇網路文章是因為和店家有某種利益交換的關係。
(2) 作者寫這篇網路文章是為了讓自己得到好處。
(3) 作者寫這篇網路文章是為了履行對店家的某種承諾。
(4) 作者寫這篇網路文章是為了客觀地分享造訪這間餐廳的心得。
Reliability Test of Measures Table 2
4.6 Results
Manipulation Checks. A 2 x 2 AVONA revealed a significant main effect of
consumers’ motivation on consumer’s involvement (F(1,76) = 80.96, p < .05). As
expected, consumers who read a high motivation scenario rated the scenario and
the blog content as more important and relevant to them (Mhigh = 5.295) than those
who read a low motivation scenario (Mlow = 3.07). There were no other significant
effects on consumer’s involvement.
Also, a significant main effect of accessibility appeared on the perceived
interests exchanging motive (F(1,76) = 14.92, p < .000). The blogger was
perceived as writing the blog in order to get reward in return when the motive was
accessible (cues were embedded) than when the motive was less accessible (no
cues were embedded) (Mhigh = 4.706, Mlow = 3.731). There were no other
significant treatment effects, indicating that the embedded cues successfully
influenced the accessibility of ulterior motives.
In addition, participants responded that the messages conveyed in the article
were positive in general (M = 5.04; t(79) = 7.98, p < .05, compared to the scale
midpoint of 4), indicating that the positive messages manipulation was effective.
Moreover, the universality of the knowledge about sponsored content/article is
proved by the participants’ high understanding level (M = 4.75; t(79) = 5.17, p
< .05, compared to the scale midpoint of 4).
High Low F-statistic p-value
Motivation 5.295 3.070 F(1,76) = 80.96 <.000
Accessibility 4.706 3.731 F(1,76) = 14.92 <.000
Manipulation Checks of Motivation and Accessibility Table 3
Mean Test Value t-statistic p-value
Positive Messages 5.040 4 t(79) = 7.98 <.000
Understanding 4.750 4 t(79) = 5.17 <.000
Manipulation Checks of Positive Messages and Universality of Sponsored Content Knowledge Table 4
Attitudes of the Restaurant. It was hypothesized that when the motivation of
consumers is low, they are less likely to use persuasion knowledge to infer ulterior
motives of the blogger under low accessibility, thus form more favorable attitudes
than under high accessibility. In contrast, when the motivation of consumers is
high, they are likely to use persuasion knowledge to infer ulterior motives of the
blogger under both high and low accessibility, thus causes no difference in the
attitudes toward the restaurant. Supporting this, a 2 x 2 ANOVA revealed a
significant interaction effect on attitudes of the restaurant (F(1,76) = 5.10, p < .03)
and a significant main effect of accessibility (F(1,76) = 7.13, p < .01). Planned
comparisons showed that, as predicted, when the motivation of consumer was low,
they formed more favorable attitudes under low accessibility than under high
accessibility. (Mhigh = 4.142, Mlow = 5.042; F(1,76) = 12.14, p < .01). When the
motivation of consumer was high, no difference occurred between high and low
accessibility situations on the attitudes of the target restaurant (Mhigh = 4.425, Mlow
= 4.500; F(1,76) = .08, p = .77 (NS)). There were no other significant main or
interaction effects.
Attitudes towards the Target Restaurant as a Function of Motivation and Accessibility of Ulterior Motives
Figure 3
Motivation Accessibility Mean Std. Deviation N
H
H 4.4250 .57348 20
L 4.5000 .81470 20
Total 4.4625 .69644 40
L
H 4.1417 .98122 20
L 5.0417 .84444 20
Total 4.5917 1.01200 40
Total
H 4.2833 .80614 40
L 4.7708 .86371 40
Total 4.5271 .86560 80
Descriptive Statistics of Attitudes towards the Target Restaurant Table 5
Source Type III Sum of Squares
df Mean Square F Sig.
Corrected Model Intercept Motivation Accessibility Motivation*Accessibility
Error Total Corrected Total
8.490a 1639.559
.334 4.753 3.403 50.701 1698.750
59.191
3 1 1 1 1 76 80 79
2.830 1639.559
.334 4.753 3.403 .667
4.242 2457.654
.500 7.125 5.101
.008 .000 .482 .009 .027
*Dependent Variable: Attitudes toward the Target Restaurant a. R Squared = .143 (Adjusted R Squared = .110)
Tests of Between-Subjects Effects Table 6
Motivation (I)Accessibility (J)Accessibility
Mean Difference
(I-J)
Std.
Error F Sig.
95% Confidence Interval for
Difference
Lower Bound
Upper Bound
H H L -.075 .258 .084 .772 -.589 .439
L H L -.900 .258 12.142 .001 -1.414 -.386
*Dependent Variable: Attitudes toward the Target Restaurant Pairwise Comparisons
Table 7
4.7 Discussion
The results of the main study support the two hypotheses we proposed. When
the motivation level of consumers is high, the accessibility of ulterior motives is
ineffective in affecting the attitudes toward the restaurant, and cause no difference
in it, H1 is proven. Since highly motivated consumers possess more cognitive
capacity, thus are more willing to scrutinize all available information and to use
persuasion knowledge to infer the ulterior motives underlay no matter the
accessibility is high or low. On the other hand, when the motivation of consumers
is low, they are less willing to scrutinize the arguments presented in the blog and to
use persuasion knowledge, and thus, are less likely to infer the ulterior motives
underlay. However, as the motivation to process arguments decreases, peripheral
cues play a more important role in forming the attitude corresponding to the target
restaurant, accessibility of ulterior motives thus becomes more effective in
influencing consumers’ attitudes; low motivation consumers will form less
favorable attitudes under high accessibility than under low accessibility, H2 is
proven.
5. General Discussion
In general, the objective of this research was to investigate the consumer’s use
of persuasion knowledge in an online blog context. Pilot study and pre-test
demonstrated that the quality unambiguity of the target restaurant will neutralize
the effects of the manipulations on ultimate attitudes forming, hence suggested us
on the target restaurant type being used in the experiment. Main study verified the
two hypotheses that we proposed, which proved that the effect of accessibility will
be moderated by the motivation of consumer. When consumer motivation is low,
the accessibility of ulterior motive will negatively affect the attitudes toward the
target restaurant; when the consumer motivation is high, the accessibility of
ulterior motives has no significant effect on the attitudes toward the target
restaurant.
5.1 Accessibility of Ulterior Motive versus Disclosure
One possible question to be asked is that if the effects of accessibility of
ulterior motives equal to the direct disclosure of sponsorship. To further clarify this
point, we conducted an extra check. The item “What’s the probability you think
that this article is sponsored?” was used as a dependent variable. The results
showed a nearly significant positive effect that the accessibility of ulterior motives
had on the estimated probability of sponsorship regardless of motivation levels
(high motivation: Mhigh = 5.300, Mlow = 4.650; F(1,76) = 3.140, p = .08; low
motivation: Mhigh = 5.150, Mlow = 4.450; F(1,76) = 3.642, p = .06).
Estimated Probability of Sponsorship as a Function of Motivation and Accessibility of Ulterior Motives
Figure 4
It’s clear that high accessibility of ulterior motives made participants perceive
the user-generated content more like a sponsored content. However, according to
the probability rating, participants were still not completely sure about whether the
content was truly sponsored or not even in the high accessibility setting (Mhigh =
5.225, the upper bound was 7), which means that it was still different from the
direct disclosure. Furthermore, though the estimated probability of sponsorship
rose as accessibility, the attitudes toward the target restaurant remained the same
for high motivation consumers (high motivation: Mhigh = 4.425, Mlow = 4.500;
F(1,76) = .08, p = .77 (NS)), which was not consistent with the negative effects of
disclosure found in past studies. Therefore, we could reasonably conclude that
accessibility of ulterior motives are different from disclosure. Although similarity
exists between them, the accessibility of ulterior motives is a more implicit, subtle,
and general factor that could affect consumers’ attitudes. Past research manipulated
the accessibility of ulterior motives in an offline clothing sales setting by the time
point of ingratiation (Campbell and Kirmani 2000), in this study, we manipulated it
in an online blog by the extra ad-text embedded. We believe that ulterior motives
would generally raise suspicion, and eventually lead to negative reactions of
consumers in most cases. Other than direct disclosure, the negative effect of
ulterior motives in a blog context is proven in this study.
5.2 Contributions
Past researches in field of native advertising focused on the effects of
disclosure, and unanimously showed that disclosures of native advertising can
activate persuasion knowledge and ultimately mitigate persuasion (Boerman, van
Reijmersdal, and Neijens 2012; Nelson, Wood, and Paek 2009; Tessitore and
Geuens 2013; van Reijmersdal, Lammers, Rozendaal, and Buijzen 2015; Wei,
Fischer, and Main 2008). In more specific blog study, negative effect of disclosure
shows lower perceived credibility of the blog and the blogger (Colliander and
Erlandsson 2015) However, there are plenty of different forms and topics of
sponsored contents which are not disclosed with sponsorship in real marketing
world. In this study, the manipulation was changed from the disclosure/non-
disclosure to high/low accessibility of ulterior motives; we believed such an
experiment could close to more real situations in some sponsored content settings.
Take online dining brief—a certain kind of sponsored blog content that became
prevalent in Taiwan past few years—for example, the non-disclosure phenomenon
is still the mainstream. The transferred focus from “disclosure/non-disclosure” to
“level of accessibility of ulterior motives” and the proven effects of that built up
the field of sponsored content research, especially for the types of sponsored
contents that do not treat disclosure as an option.
This research also contributes to the Persuasion Knowledge Model, the
general theory about how consumers responds to marketers’ attempts at persuasion.
We have further developed one portion of the model, the use of persuasion
knowledge in an online blog context. We apply the existing research on persuasion
knowledge in an online user-generated contents context, and introduce motivation
of consumer and accessibility of ulterior motives as influential factors in the use of
persuasion knowledge. As past research had already verified that cognitive
capacity will affect the use of persuasion knowledge, which requires higher-order
reasoning (Campbell and Kirmani 2000), we provide another internal factor that
have, to some extent, more prerequisite impact on it, the motivation of consumer.
The PKM (Friestad and Wright 1994), the ELM (Petty and Cacioppo 1986), and
the FCM (Wegener and Petty 1997) all stressed the prerequisite feature of
motivation in determining the amount of cognitive resources. In addition, one of
the major differences between online blog context and traditional interpersonal
persuasion context is that consumers do not have to interact with others anymore,
they even don’t have to give back any kind of response (e.g., speaking, facial
expression, changing postures), in other words, cognitive constraint resulting from
cognitive busyness could rarely happen in a blog context, or some other similar
online user-generated content like videos, pictures, social media posts, audio, etc.
This research identified the moderating role of motivation to process persuasive
messages in an online marketing setting, and provides a different perspective that
extends the attention from external factors (e.g., disclosure/non-disclosure,
duration of the disclosure, disclosure time points) to internal conditions (i.e.,
motivation of consumer) in the field of native advertising/sponsored content. We
believe that an integrated viewpoint that includes both external factors and internal
conditions will help us understand the persuasion episode to the fullest.
Contribution is also made to the sponsored content studies, we found that
when the target restaurant type is relatively unambiguous in nature, which is to say,
consumers regard it to have a certain level of quality, kind of image, and features, it
is more likely for accessibility of ulterior motives and motivation of consumer to
become non-influential to the attitudes toward the restaurant. Anchors are still hold
by consumers in this situation (Helson 1964; Oliver and Linda 1981). This finding
could be generalized to other types of product and other formats of user-generated
content (e.g., user-generated video that promotes a brand-new drink).
5.3 Managerial Implications
For marketers who aim to leverage user-generated content in a blog context to
attain marketing goals, they must be aware of the negative effects caused by the
accessibility of ulterior motives, which will happen even without disclosure. In
addition, the motivation of consumer also plays as an important role during the
persuasion episode. High motivation consumers may use their persuasion
knowledge in whatever condition and thus there is no significant effects of the
accessibility; marketers should put more attention on the issue-relevant information
to attract those consumers, rather than hiding the fact of sponsorship; Low
motivation consumers tend to use their persuasion knowledge more when the
accessibility is high, thus lead to less favorable attitudes about the target product,
marketers should try to avoid them from inferring the ulterior motives underlay,
especially be conscious of cues embedded in the content which would raise
consumers’ suspicion.
Since the product ambiguity will influence the effect of user-generated
content on attitude change, the marketing goals marketers aim to achieve through
the user-generated content should be clarified beforehand. If the ambiguity of the
target product or brand is low in nature (e.g., mug, Taylor’s guitar), the user-
generated content would not be valid to enhance the positive attitudes toward it,
however, if the goal is the other (e.g., creating awareness, maintaining loyalty), it’s
still a good way to leverage user-generated content since the information
abundance it could deliver. If the target product or brand marketers are promoting
is ambiguous in nature (e.g., laptop, Samsung), user-generated content would be
effective in positively affecting the attitudes toward it. In conclusion, ambiguity
should be considered regardless of what object is being promoted (i.e., product,
brand, service, etc.), then marketers can utilize the user-generated content in the
reasonable way.
5.4 Limitations and Future Research
This research aimed to study factors that influencing consumers’ use of
persuasion knowledge by user-generated content in a blog context, and we have
presented how accessibility of ulterior motives and motivation of consumer can
interact to affect the use of persuasion knowledge thus influence the attitudes
toward the target restaurant. Future research should explore more factors that can
influence the consumers’ use of persuasion knowledge from both perspectives of
internal and external. We especially emphasize on further investigation in internal
factors that could affect consumers’ use of persuasion knowledge in context of
online user-generated content. Internal factors like internet familiarity could be an
instance, the participants of this study were all students, who could be imagined
that being more familiar with internet world. The effects of internet familiarity
could be further probed, sub-factors of it like knowledge about different
websites/platforms, ability to integrate digital information, understanding of
internet slang and more could all become a single study in the future. On the other
hand, external factors that can affect consumers’ use of persuasion knowledge are
worth exploring as well. This study focused on a single format of user-generated
content (i.e., blog article) and a single type of product (i.e., beef noodle restaurant).
Whether the effects proved in this study could be generalized to other formats and
product types still needs more consideration. Moreover, we suggest future research
to delve into different product types, to see if the product type itself could affect
the use of persuasion knowledge, take an example in Taiwan, healthy food being
promoted in user-generated content might easily arouse consumers’ suspicion and
use of persuasion knowledge, however, sports shoes might not.
The moderating role of motivation was presented in this article, which
moderated the effect that accessibility had on attitudes. We suggest that future
research to include motivation as well to see if it could also moderate other effects.
Lastly, the effects of product ambiguity are also an interesting topic to examine.
Past research has investigated on the effects that ambiguity has on consequences of
priming (Herr, Sherman, Fazio 1983). Now we suggest future research to
investigate effects that product ambiguity would have on user-generated content
among different product types (e.g., 3C product, daily necessity, garment, etc.). We
had already found that product unambiguity could neutralize the effects of
accessibility and motivation on attitudes, however, the mechanism that behind it is
still worth to explore. For example, either it is because consumers do not use
persuasion knowledge due to product unambiguity, or it’s because the anchoring
effect makes consumers hold the anchor provided by prior experience even if the
persuasion knowledge is used. We believe that clarifying the mechanism will help
marketer better utilize the user-generated content across different products.
6. References
Adair, J. G (1984), “The Hawthorne Effect: A Reconsideration of the Methodological Artifact,” Journal of Applied Psychology, 69 (2), 334-345.
Boerman, S. C., van Reijmersdal, E. A, and Neijens, P. C (2012),
“Sponsorship Disclosure: Effects of Duration on Persuasion Knowledge and Brand Responses,” Journal of Communication, 62, 1047-1064.
Chaiken, S., Liberman, A, and Eagly, A. H (1989), “Heuristic and Systematic Processing within and Beyond the Persuasion Context,” in J. S. Uleman, and J. A.
Bargh (Ed.), Unintended thought, 212-252, New York: Guilford Press.
———, and Maheswaran, D (1994), “Heuristic Processing Can Bias Systematic Processing: Effects of Source Credibility, Argument Ambiguity, and Task Importance on Attitude Judgment,” Journal of Personality and Social Psychology, 66 (3), 460-473.
Chia, A (2012), “Welcome to me-mart: The Politics of User-Generated Content in Personal Blogs,” American Behavioral Scientist, 56, 421-438.
Colliander, J., and Erlandsson, S (2015), “The Blog and the Bountiful:
Exploring the Effects of Disguised Product Placement on Blogs that are Revealed by a Third Party”, Journal of Marketing Communication, 21 (2), 110-124.
Campbell, M. C., and Kirmani, A (2000), “Consumers' Use of Persuasion Knowledge: The Effects of Accessibility and Cognitive Capacity on Perceptions of an Influence Agent,” Journal of Consumer Research, 27 (1), 69–83.
Duane T. Wegener., and Richard E. Petty (1997), “The Flexible Correction Model: The Role of Naive Theories of Bias in Bias Correction,” Advances in Experimental Social Psychology, Academic Press, 29, 141-208
Fein, Steven (1996), “Effects of Suspicion on Attributional Thinking and the
Correspondence Bias,” Journal of Personality and Social Psychology, 70 (June), 1164–1184.
———, James L. Hilton, and Dale T. Miller (1990), “Suspicion of Ulterior Motivation and the Correspondence Bias,” Journal of Personality and Social Psychology, 58 (May), 753–764
Festinger, L (1950), “Informal Social Communication,” Psychological Review, 57, 271-282.
Friestad., Marian., and Peter Wright (1994), “The Persuasion Knowledge Model: How People Cope with Persuasion Attempts,” Journal of Consumer Research, 21 (June), 1–31.
Faber, Ronald J., Lee, Mira, and Nan, Xiaoli (2004), “Advertising and the Consumer Information Environment Online,” American Behavioral Scientist, 48 (4), 447-466.
Gilbert, D. T., and Malone, P. S (1995), “The Correspondence Bias,” Psychological Bulletin, 117 (1), 21-38.
———, D. T., Pelham, B. W., and Krull, D. S (1988), “On Cognitive Busyness: When Person Perceivers Meet Persons Perceived,” Journal of Personality and Social Psychology, 54 (5), 733-740.
———, Daniel T, Edward E. Jones, and Brett W. Pelham (1987), “Influence and Inference: What the Active Perceiver Overlooks,” Journal of Personality and Social Psychology, 52 (5), 861–870.
Higgins, E. Tory (1989), “Knowledge Accessibility and Activation:
Subjectivity and Suffering from Unconscious Sources,” in Unintended Thought, ed. J. S. Uleman and J. A. Bargh, New York: Guilford, 75–123.
Higgins, E. Tory, and Gillian King (1981), “Accessibility of Social
