低涉入度消費者如何透過中央說服途徑處理產品核心特性

全文

(1)國立臺灣師範大學管理學院全球經營與策略研究所碩士論文 Graduate Institute of Global Business and Strategy College of Management National Taiwan Normal University Master Thesis. 低涉入度消費者如何透過中央說服途徑處理產品核心特性 How May Low Involvement Consumers Process Persuasive Messages via Central Routes. 廖曉彤 Hsiao-Tung Liao. 指導教授：蕭中強博士 Advisor：Chung-Chiang Hsiao, Ph.D. 中華民國100年6月 June, 2011.

(2) ABSTRACT How May Low Involvement Consumers Process Persuasive Messages via Central Routes. Hsiao-Tung Liao,. Advisor: Professor Chung-Chiang Hsiao. Institute of Global Business and Strategy National Taiwan Normal University, Taipei, Taiwan. The Elaboration Likelihood Model (ELM) has been widely used by marketing industry and researchers in explaining advertising persuasion effectiveness since 1980’s. ELM proposed that low elaboration likelihood tends to process peripheral cues rather than arguments. However, in this study we examined the effect of misconceived arguments on ad processing and product assessment. In this study, we define low involvement is the group who are low in objective knowledge regardless their motivation. We found that consumers who are low in objective knowledge (e.g., novices) are capable of processing central arguments as long as they have high motivation and abundant subjective knowledge. The result showed that even without high level of expertise, the attitude certainty of novices is as great as high involvement group (e.g., experts) when they are high in motivation. In I.

(3) contrast, the product attitudes formed by low objective consumers are opposite to high involvement consumers. Key words: Elaboration Likelihood Model; Multiple Role; Attitude Certainty.. II.

(4) 謝辭寫這篇致謝文的同時，意味著學生生涯的結束，接下來即將邁入人生的下一階段。回想起研究所的生活，大概是我人生中最充實的日子，雖然在從事研究的過程遭遇許多挫折，但從中學習到非常多寶貴的知識，最後能順利完成學術論文，我要非常感謝這段時間陪在我身旁的師長與朋友，謝謝你們給予的支持與鼓勵。首先，最要感謝的就是像爸爸一樣親切的蕭中強老師了，即使再忙碌蕭老師總是花很多時間與我們 Meeting，不厭其煩的仔細指導我如何分析資料、釐清邏輯，在閱讀、討論及撰寫論文的過程中，以嚴謹邏輯思維的方式指導我，並在我們遇到困難時，給予許多方向和建議，從中採取開放式的討論，還不時鼓勵我們，給予我很多的信心。準備論文的過程，既充實又開心，特別是在蕭老師的指導下，學習變成一件快樂的事，不僅如此，蕭老師平時就像朋友一般，與我們分享寶貴的人生經驗及做人處事的道理，能夠成為蕭老師的指導學生，真的十分幸福! 再來，還要非常感謝台大的簡怡雯老師，謝謝她總是充滿著細心與耐心，在指導與討論的過程中，給予我許多寶貴的建議。此外也要感謝撥空前來擔任口試委員的台大練奶華老師及政大陳建維老師，他們在兩次口試中給予我許多珍貴的建議，使得這篇論文能夠更完整呈現。除此之外，我也十分感謝台師大同門的佩君、佳霓和台大的貫御學長，在這將近一年的日子裡，不論是討論、發問卷或是最後趕論文的時刻，謝謝你們都在我身邊，給予我最大的鼓勵與建議，希望未來我們都發展順利！謝謝書齊和 Inn 教我這電腦白癡使用 Photoshop；整理樣本沉悶的時候，謝謝有麥麥、穗君、方維、巧婷和張凌嘉陪我一起度過；謝謝和我同樣為論文奮鬥的蔡昌儒及巧涵，在撰寫論文的時刻常常督促我、支持我，感念那一起在圖書館努力的時光；謝謝林韋丞同學三不五時告訴我班上發生的事情，讓糊塗的我不會因此錯過重要的訊息；謝謝廣大的受測者們，花時間幫我填寫問卷；最後，一定要謝謝其他陪在我身邊的國中、高中和大學朋友們，還有我最親愛的家人，在這段時間體諒我、幫助我，讓我能夠專心地完成畢業論文，衷心感謝所有人對我無私的幫助，因為有你們才有現在的我！廖曉彤謹識于台師大全球經營與策略研究所 100 年 7 月. III.

(5) Table of Contents CHAPTER 1 INTRODUCTION ................................................................................ 1 1.1 Motivation and Study Purpose ....................................................................... 1 CHAPTER 2 LITERATURE REVIEW .................................................................... 3 2.1 Elaboration Likelihood Model (ELM) ........................................................... 3 2.2 Multiple Roles................................................................................................ 8 2.3 Attitude Certainty......................................................................................... 11 CHAPTER 3 PROPOSED THEORY AND HYPOTHESES ................................ 13 3.1 Proposed Theory .......................................................................................... 13 3.2 Hypotheses ................................................................................................... 17 3.2.1 Effect of Objective Knowledge .......................................................... 17 3.2.3 Effect of Peripheral Cue...................................................................... 18 3.2.4 Effect of Motivation on Attitude Certainty......................................... 19 CHAPTER 4 RESEARCH METHODOLOGY ...................................................... 20 4.1 Overview ...................................................................................................... 20 4.2 Pretest ........................................................................................................... 20 4.2.1 Participants and Design....................................................................... 20 4.2.2 Result .................................................................................................. 21 4.3 Main Experiment ......................................................................................... 22 4.3.1 Participants and Design....................................................................... 22 4.3.2 Experimental Procedure ...................................................................... 23 4.3.3 Independent Variables ........................................................................ 24 4.3.3.1 Motivation ................................................................................ 24 4.3.3.2 Objective Knowledge............................................................... 25 4.3.3.3 Argument ................................................................................. 26 4.2.3.4 Endorser ................................................................................... 26 4.3.4 Dependent Variables ........................................................................... 27 4.3.4.1 Attitude .................................................................................... 27 4.3.4.2 Attitude Certainty..................................................................... 27 4.3.5 Manipulation Check ............................................................................ 28 4.3.5.1 Motivation ................................................................................ 28 4.3.5.2 Objective Knowledge............................................................... 28 4.3.5.3 Argument ................................................................................. 29 4.3.5.4 Endorser ................................................................................... 29 CHAPTER 5 RESULT .............................................................................................. 31 IV.

(6) 5.1 Manipulation Check ..................................................................................... 31 5.1.1 Motivation ........................................................................................... 31 5.1.2 Objective Knowledge.......................................................................... 32 5.1.3 Argument Quality ............................................................................... 33 5.1.4 Endorser .............................................................................................. 36 5.2 Dependent Measures .................................................................................... 37 5.2.1 Attitude ............................................................................................... 37 5.2.2 Attitude Certainty................................................................................ 41 5.3 Tests for Hypotheses .................................................................................... 43 5.3.1 Effect of Objective Knowledge .......................................................... 44 5.3.2 Effect of Argument Quality ................................................................ 46 5.3.3 Effect of Peripheral Cue...................................................................... 50 5.3.4 Effect of Attitude Certainty ................................................................ 51 CHPATER 6 GENERAL DISCUSSION ................................................................. 54 6.1 Contribution ................................................................................................. 54 6.2 Limitations ................................................................................................... 57 REFERENCES ........................................................................................................... 58 APPENDIX(A) ........................................................................................................... 61 QUESTIONNAIRE OF PRETEST APPENDIX(B)............................................................................................................ 73 QUESTIONNAIRE OF MAIN EXPERIMENT. V.

(7) List of Tables Table 1. Two Argument Statements of Target Product……………………………..…22 Table 2. Experimental Design……………………………………………………..….23 Table 3. Descriptive Statistics for the Manipulation Check of the Level of. Motivation……………………………………………..………………….…31 Table 4. ANOVA Outcome for the Manipulation Check of the Level of Motivation…31 Table 5. Descriptive Statistics for the Manipulation Check of the Level of Objective. Knowledge.......................................................................................................32 Table 6. ANOVA Outcome for the Manipulation Check of the Level of Objective. Knowledge…………………………………………………………………...32 Table 7. Descriptive Statistics for the Manipulation Check of the Argument. Quality..……………………………………………………………………...33 Table 8. ANOVA Outcome for the Manipulation Check of the Argument Quality…...34 Table 9. Pairwise Comparisons for the Manipulation Check of the Argument Quality. under High Motivation....................................................................................35 Table 10. Descriptive Statistics for the Manipulation Check of the Level of. Endorser……………………………………………………………………..36 Table 11. ANOVA Outcome for the Manipulation Check of the Level of Endorser….36 Table 12. Cell Means and Standard deviation of Product Attitude…………………..37 Table 13. The Outcome of Four-way ANOVA (Objective Knowledge, Motivation, Endorser, and Argument Quality) on Evaluations of Target Attitude……….38 Table 14. Cell Means and Standard Deviation of Product Attitude Certainty…..........41 Table 15. The Outcome of Four-way ANOVA (Objective Knowledge, Motivation,. Endorser, and Argument Quality) on Evaluations of Target Attitude Certainty…………………………………………………………………....42 Table 16. Descriptive Statistics of Between-Participants (Objective Knowledge and Argument Quality) …………………………………………………….........43 VI.

(8) Table 17. The Outcome of Two-way ANOVA (Objective Knowledge and Argument Quality) of Evaluations on Shampoo in the High Motivation Condition.....44 Table 18. The Analysis Outcome of Hypothesis 1a......................................................45 Table 19. The Analysis Outcome of Hypothesis 1b......................................................46 Table 20. The Analysis Outcome of Hypothesis 2a.......................................................47. Table 21. The Analysis Outcome of Hypothesis 2b......................................................48 Table 22. Descriptive Statistics of Between-Participants (Objective Knowledge and Endorser) .....................................................................................................49 Table 23. The Outcome of Two-way ANOVA (Objective Knowledge and Endorser) of Evaluations on Shampoo in the Low Motivation Condition........................50 Table 24. Descriptive Statistics of Between-Participants (Objective Knowledge and Motivation) ..................................................................................................52 Table 25. The Outcome of Two-way ANOVA (Objective Knowledge and Motivation) of. Attitude Certainty..........................................................................................53. VII.

(9) List of Figures Figure 1. The Weight of Elaboration Falls along Continuum from High/Low. Involvement…………………………………………………………………6 Figure 2.Multiple Roles of Variable…………………………………………………..8 Figure 3. “Beautiful Scenery” in a Vacation Location/Car Advertisement (Petty &. Wegener, 1999)…………………………………………………………….10 Figure 4.Antecedents and Consequences of Subjective Attitude Certainty (Gross,. Holtz, & Miller, 1995)……………………………………………………..12 Figure 5. Novices and Experts in Central Routes…………………………………….16 Figure 6 .Manipulation Check of the Argument Quality under High Motivation. Condition…………………………………………………………………...35 Figure 7.The Interaction Effect of Motivation and Endorser on Evaluations of Target. Product……………………………………………………………………..40 Figure 8.Target Attitude under the Condition of High Motivation ……………….....49 Figure 9.Attitude toward Products under Low Motivation…………………………..51 Figure 10.The Effect of Target Attitude Certainty…………………………………...53. VIII.

(10) CHAPTER 1 INTRODUCTION 1.1 Motivation and Study Purpose In the highly competitive market nowadays, we are exposed to massive quantity of advertisements. Therefore, consumer psychologists have been interested in developing consumer attitude change, which has become a main issue in consumer behavior research over the past few decades.. Petty and Cacioppo (1980) introduced “dual-route to persuasion” called ELM model to explain the effectiveness of persuasive communication on advertisements. It has been greatly adopted by consumer researchers in the field investigation of advertising persuasion over the past 20 years. Most of the research had discovered that low involvement consumers who are lack of either ability or motivation for processing persuasive message will change their product attitude through peripheral route rather than central route (Petty, Caccioppo, & Schuman, 1983). However, is it true that low involvement people always process persuasive messages via peripheral routes? If the answer is no, what are the factors which influence low involvement people exert in processing rather than using heuristic processing? In this study, we define low involvement are the group who are low in objective knowledge regardless their motivation. Here we discovered that under the condition of high motivation, 1.

(11) consumers who are low in objective knowledge (e.g., novices) are still process central arguments due to novices’ subjective sense of their knowing about target is high. Therefore, as long as novices are high in motivation and high in subjective knowledge, they may scrutinize product relevant arguments rather than peripheral cues. Besides, we also found that the attitude certainty of low involvement consumers is at the same level as high involvement consumers.. In addition, previous research on multiple roles had demonstrated peripheral cues can serve as different role under different elaboration continuum. However, we point out the central arguments, potentially, can also play different roles.. 2.

(12) CHAPTER 2 LITERATURE REVIEW 2.1 Elaboration Likelihood Model (ELM) Over past 20 years, the notion of ELM model has been greatly adopted by la lot of consumer researchers (Petty, Cacioppo, & Schumann, 1983).Especially in the field investigating of advertising persuasion. Generally, ELM theory was about how such classic source (e.g. expertise/celebrity), message (e.g. large/small number of arguments), recipient (e.g. good/bad mood), and contextual (e.g. distraction) variables have impact one’s attitudes toward various objective, issues, and people. According to Elaboration Likelihood Model theory it demonstrated that they are two distinct routes took different impact on persuasion on issue or product judgment. The dual routes-central and peripheral route change attitude depended on how effortful of processing information activity (elaboration likelihood) about a target or product occurs. Central route to persuade meant that people diligent to evaluate the issue/product relevant argument, therefore, relatively extend and effort processing activity help she/he scrutinize and uncover the central merit of the issue or product attitude position. If this information is perceived to be cogent and persuasive, favorable attitudes will result, but if this information is weak and specious, unfavorable attitudes will result. Furthermore, attitude changes induced through the 3.

(13) central route are postulated to be relatively enduring and predictive of behavior (Petty et al., 1983). On the other hand, via peripheral route to change attitude based on person who use relative low effortful and simple inference of positive or negative cues (e.g. good/bad mood, sunny /gloomy weather, endorser) because that simple decision rules need less cognitive effort that prevent people from processing the issue or product relevant information(Petty et al., 1983). Attitude changes through peripheral route are postulated to be relatively temporary and unproductive of behavior. Petty, Cacioppo and Schumann (1983) proposed that involvement as an important moderator to decide the weight of cognitive effort. The ELM showed that as an issue/product increases in personal relevance or consequences, it becomes more important and adaptive to forming a reasoned judgment. Thus, people who are under high involvement are more motivated to do effort for cognitive evaluate the true merits of an issue or product rather than low involvement. Degree of involvement determined by one’s level of motivation and ability to process issue- or productrelevant information. The more motivate, the more likely to devote the cognitive effort. The more knowledge about the issue/product people have, the more they can scrutinize all available issue/product relevant argument and determine how the object really is. When people have higher motivation and ability, they are defined at the high 4.

(14) end of the elaboration continuum, which means they are more likely to engaged in all available object-relevant information. On the contrary, when people lack of either motivation or ability, they are defined at the low end of the elaboration continuum, which means information scrutiny is decreased, thus attitude change results from a low-effort scrutiny of the information available (e.g., examining less number of message arguments than high involvement people or simple decision rules to change issue/product attitude). Hawkins and Hoch (1992) proposed that in the study of decision heuristics, novices are categorized as people in low involvement. For example, Sujan and Mita (1985) found that novices (i.e., low-ability consumers) rely on schema-based category affect. Novices focus more on peripheral cues, such as credibility, than on central cues when cognitive effort requirements are too high (Alba & Hutchinson,1987) People have been misunderstood that the elaboration likelihood model was only refer to two separate endpoint called “central route” and “peripheral route”. In fact, Petty and Wegener (1999) pointed out that the amount of elaboration falls along a continuum anchored at one end by the central route another end by peripheral route rather than being two discrete points. Although people want to hold correct attitudes, the amount and nature of issue-relevant elaboration in which they are willing or able to engage to evaluate a message with individual and situational factors. 5.

(15) Figure 1. The Weight of Elaboration Falls along Continuum from High/Low Involvement. To sum up, when involvements is high, central route to persuade would be greater impact and more persuasive on issue/product attitude change rather than low. Person who attitude change through central route, not only motivation but also ability are high. Person is engage in considerable cognitive effort. On the other hand, peripheral route to persuasion occur when either motivation or ability is relative low, effort to scrutinize central merit about target will be relative low.. 6.

(16) Misleading vs. Misconceived Russo, Metcalf and Stephens (1981) proposed and test a procedure for identifying misleading advertising. They identify misleading ad based on three criterions, which is consumer beliefs of a false claim, increased belief in a false claim after exposure to an ad and less misleadingly false beliefs for corrected than for original ads. They used a banana ad as an example that claims” there’s only 85 calories “, which is only true for small bananas; on the other hand, a typical medium-sized banana contains 101 calories (Russo et al., 1981). In our study we define misconceived statement as the statements which can make consumers generate incorrect beliefs prior the advertisement exposure. In addition, after receiving a credible knowledge consumers will totally inverse their attitude in an opposite direction.. 7.

(17) 2.2 Multiple Roles When the two routes to persuasion idea were introduce, most people sometimes misunderstood the notion of ELM. For example, ELM holds that variables can take on only one of the postulated roles. Thus, scholars who review the effects of certain variables have sometimes struggled to determine if their variable is a “central” or a “peripheral” one, rather than recognizing the multiple roles for variables. Petty and Cacioppo (1986a:1986b) introduced the notion of multiple roles for persuasion variables. That is, the ELM noted that variables can influence attitudes in four ways: 1) by serving as an argument, 2) by serving as a cue, 3) by determining the extent of elaboration, and 4) by producing a bias in elaboration.. High. Moderate Roles of Variable Figure 2. Multiple Roles of Variable. 8. Low.

(18) Multiple Roles suggests that any one variable can have an impact on attitude change by more than one mechanism. In other words, any given variable which can influence attitudes by take on different roles at different points along the elaboration continuum (Petty & Cacioppo, 1986a; 1986b). It is possible that variables serve as cues under low end of the elaboration continuum may serve as arguments or bias information processing at the high end of the elaboration continuum. Multiple Roles mention that some variables (such as source attractiveness) usually treated as peripheral cues may serve as central arguments under other conditions. For instance, people evaluate an advertisement for vacation location and car, “beautiful scenery” contained in the advertisement would increase or decrease impact on attitude when the elaboration likelihood is increased. Under low elaboration continuum condition, person was not thinking about the advertisement very much, therefore, the beautiful scenery will forming positive attitude on them. Because of at low elaboration people associate the beautiful scenery with location and car in the ads. However, as the elaboration likelihood is increased lead people think more. People serve as a strong argument when the impact of the scenery on attitudes increased in the ad for the vacation location due to its perceived relevance and central merit to the target. Yet, it show decreased impact on attitude in the ad for the car due to its perceived irrelevance for this product merit so people serve the scenery as a weak 9.

(19) argument. The positive impact of “beautiful scenery” could also be reversed if the scenery makes the ad seem more interesting and thus people think about the ad more when the ad contains only weak arguments for the product.. Perceived Product Relevance to Merit (Vacation Location). Favorable (Strong argument). High Elaboration Perceived product irrelevance (Car). Beautiful Scenery (Cue/Argument ). Low Elaboration. Simple Association (Vacation Location & Car). Reduced Argument Impact. Favorable (Cue). Figure 3. “Beautiful Scenery” in a Vacation Location/Car Advertisement (Petty & Wegener, 1999). In brief, the fact that variables can take on different roles at different points along the elaboration continuum implies that the impact of any given variable that serves as a peripheral cue under low elaboration conditions can be enhanced, reduced, reversed, or remain the same as the elaboration likelihood is increased (Petty & Wegener, 1999).. 10.

(20) 2.3Attitude Certainty Attitude certainty is a subjective sense of conviction or validity about one’s attitude or opinion. It refers to the confidence with which an individual holds an attitude: the more confidently an individual holds an attitude to be correct, the more certainty is present (Gross, Holtz, & Miller, 1995). Informational/Cognitive Factors, social/Consensual Factor and Individual Attributes are classified as three different categories of antecedent of attitude certainty. A number of determinants of attitude certainty have been examined. There were many factors affect attitude certainty. People tend to be more certain of their attitude when they based on direct experience (direct experience can provide knowledge), when they come to mind easily, that is, faster recall is associated with an inference of increased validity of one’s thoughts (attitude accessibility is high), when others agree with the attitude (social norm), when they have done much prior thinking about the attitude object (elaboration). Some individual differences have also been related to attitude certainty such as self-monitoring, need for certainty, age et al (Petty, Briñol, Tormala, & Wegener, 2007). Certainty has been associated with number of important attitude-relevant consequence for anyone trying to develop for product advertising. The attitudes held with greater certainty are more persistent over time, resistant to counter argument, and 11.

(21) more predictive of behavior (Petty et al., 2007).. Informational/Cognitive Factors Knowledge Thought Attitude Accessibility Framing. Social/Consensual Factors Social Influence and Support Group Discussion. Persistent over Time Attitude Certainty. Resistant to Counter Argument Predictive of Behavior. Individual Attributes Insecurity Need for Certainty Tolerance of Ambiguity Ambivalence Need for Cognition Self-Monitoring Depression Age. Figure 4. Antecedents and Consequences of Subjective Attitude Certainty (Gross, Holtz, & Miller, 1995). 12.

(22) CHAPTER 3 PROPOSED THEORY AND HYPOTHESES 3.1 Proposed Theory According to the notion of “Multiple Roles”(Petty& Wegener, 1999) which suggests that any given persuasive variable can play different roles in different situations. That is, peripheral variables can serve as either cues under low involvement or arguments when the involvement is high at its continuum. Extending from the view of multiple roles, we postulate that it is possible that central arguments may also take different role to influence low involvement participants. ELM proposed that people under low elaboration likelihood go through peripheral routes to change attitude because they lack of either motivation or ability. Thus, attitude may be changed by simple decision rules or positive/negative cues (Petty, Cacioppo and Schumann, 1983).In general, irrespective of motivation, novices who have no knowledge to process encountered information are categorized in the low involvement group who tend to go through peripheral routes. It is very likely that novices whose objective knowledge that refers to accurate stored information (Moorman, Diehl, Brinberg, & Kidwell, 2010) are low are also low in subjective knowledge which refers to a belief about the state of knowledge (Moorman et al., 13.

(23) 2010). An individual’s perception of low self-sense of knowledge leads her/ him to believe that she/ he is unable to process target-relevant messages. Therefore, novices generally process the persuasive communication through peripheral routes. How about novices who don’t know they are novices? In other words, they believe they know about the information included in the persuasive messages. Yet, their levels of objective knowledge are low. We suggest that when novices they don’t know they are novices encountering misconceived arguments which are paradoxical in nature; they may believe that they know how to explain the arguments in the correct way. In the situation of high motivation, novices who are also high in subjective knowledge may engage in effortful information process without realizing that the subjective knowledge they use to interpret the persuasive messages may be wrong. Therefore, high motivation novices who are high in subjective knowledge may process persuasive messages in central routes but end up evaluating the target differently from high motivation experts. That is, the misconceived arguments interpreted by experts who have objective knowledge to processes misconceived message may be regarded as relatively weak and lead to primarily unfavorable cognitive responses. On the contrary, novices may perceive the same misconceived arguments as relatively strong and elicit primarily favorable cognitive responses. Figure 5 Illustrates in the condition of high motivation, how the “misconceived 14.

(24) arguments” influence attitude of novices and experts. According to ELM, attitudes based on high amount of thought (elaboration) about an attitude object are proposed to be more persistent over time, resistant to attack, and predictive of behavior than attitudes formed or changed with little though(Petty& Jamie, 2008). Therefore, we conduct the experiment for participants who have high motivation, high subjective but low objective knowledge, we propose that it will lead them to effort fully scrutinize the product relevant arguments. It may have same three consequences of attitude certainty with high involvement people.. 15.

(25) Objective Knowledge (High vs. Low) High Subjective Knowledge. High Motivation. Misconceived Arguments. Central Route Novices. Experts. Perceived Weak Arguments. Perceived Strong Arguments. Unfavorable Product Attitude. Favorable Product Attitude. &. &. High Attitude Certainty. High Attitude Certainty. Figure 5. Novices and Experts in Central Routes. 16.

(26) 3.2 Hypotheses In order to examine the moderating role on central arguments toward the product attitude change based on Figure 5, we developed four hypotheses as follows.. 3.2.1 Effect of Objective Knowledge Hypothesis 1a: Under conditions fostering high motivation and high subjective knowledge, product attitudes tend to be more favorable when novices encounter misconceived arguments and perceive them relatively strong than when experts encounter the same arguments and perceive them relatively weak.. Hypothesis 1b: Under conditions fostering high motivation and high subjective knowledge, product attitudes do not differ between novices encountering un-misconceived and neutral arguments and experts encountering the same arguments.. 17.

(27) 3.2.2 Effect of Argument Quality Hypothesis 2a: Under conditions fostering high motivation and high subjective knowledge, product attitudes tend to be less favorable when experts encounter misconceived arguments and perceive them relatively weak than when they encounter un-misconceived and neutral arguments.. Hypothesis 2b: Under conditions fostering high motivation and high subjective knowledge, product attitudes tend to be more favorable when novices encounter misconceived arguments and perceive them relatively strong than when they encounter un-misconceived and neutral arguments.. 3.2.3 Effect of Peripheral Cue Hypothesis 3: Under conditions fostering low motivation, product attitudes tend to be more favorable when the attractive endorser is exposed than when average endorser is received in the target advertisement.. 18.

(28) 3.2.4 Effect of Motivation on Attitude Certainty Hypothesis 4a: People who are high in motivation and subjective knowledge tend to form attitudes with greater attitude certainty than their counterparts who are low in motivation but high in subjective knowledge.. Hypothesis 4b: The attitude certainty of low objective knowledge (novices) is likely to be indifferent from that of high objective knowledge (experts) when they are high in subjective knowledge and high in motivation.. 19.

(29) CHAPTER 4 RESEARCH METHODOLOGY 4.1Overview. This study was conducted by experimental design and devoted to examine whether or not people under low involvement can process central arguments. In our study, misconceived argument was used to trigger multiple roles of arguments. Therefore, the target judgment in low involvement group who under condition on high motivation, high subjective know ledge but low in objective knowledge tend to process misconceived arguments rather than peripheral cues. Nevertheless, experts and novices both are process encounter information in advertisement via central routes but experts and novices form target attitudes in opposite directions. Neutral argument was used to examine the argument quality effect. Neutral argument was the evidence of process in central routes. Generally, our study is extending ELM, therefore, ELM research method is employed (Petty et.al, 1983).. 4.2 Pretest. 4.2.1 Participants and Design The purpose of pretest is to select a misconceived statement and a neutral statement as the materials for main experiment. The criteria we determinate the paradoxical 20.

(30) statements are whether it can make novices generate wrong or misunderstood beliefs toward product prior to the advertisement exposure. Misconceived arguments we selected had been approved by “Consumer Report of Taiwan” which is published by Consumers' Foundation, Chinese Taipei. We first chose fourteen statements in shampoo advertisement including misconceived and neutral arguments for pretest. According to the result of pretest, we finally selected a statement of misconceived and neutral respectively. A total of 44 undergraduates in College of Management of National Taiwan Normal University in marketing course are recruited to participate in the pretest. The participants were required to rate their overall attitude about the statements on each ad statement by three question, which examined persuasiveness, expression of performance and thought of shampoo on a seven-point scale anchored at -3(very unpersuasive/very negative expression/very negative thought) to 3 points (very persuasive/very positive expression/very positive thought).. 4.2.2 Result First, we took coefficient of Cronbach’s α to measure the reliability of three items in each statement in this questionnaire. The result of each statement’s Cronbach’s α was higher than 0.79 which means the items used in the study are consistent. Thus, we averaged to form an overall index. After, we used the value to continue our analyses. 21.

(31) Second, result of pretest showed in Table 1 display the statement of “富含柑橘精油，並徹底洗淨毛髮“ was be chosen as an misconceived argument(M = 1.363; t=8.042, p = .00) and “熱愛生命，不採動物測試“ as an neutral argument (M = 0.295; t=1.674, p = .10).There were significantly different between two statements (t=4.918, p = .00). Table1. Two Argument Statements of Target Product Argument Quality. Argument Statement. Mean. Std. Dev.. Perceived Absolutely. 富含柑橘精油，並徹底洗淨毛髮. 1.363. 1.12. 熱愛生命，不採動物測試. 0.295. 1.17. Strong Neutral. 4.3Main Experiment. 4.3.1Participants and Design A total of 400 males and females undergraduates participated in the experiment; 25 participants were randomly assign to each of the cells in a 2 (Motivation: high or low) x 2 (Objective knowledge: high or low) x 2 (Argument: misconceived or neutral) x 2 (Endorser: attractive or average) between-participants factorial design. The design of experiment is showed in TABLE2.. 22.

(32) Table 2. Experimental Design Variables. Level. Motivation. High. Low. Objective knowledge Argument Endorser. High. Low. Misconceived. Neutral. Attractive. Average. 4.3.2Experimental Procedure One booklet was prepared for the study which contained the advertising stimuli and dependent measures. We had one booklet which contained one article and three ads in the questionnaire for this study. The first page of the booklet explained that the study purpose of concerned the college students’ reading comprehension and concentration on magazine articles and advertisements. It also an instruction to tell participants through the booklet without reviewing previous page and provide the thought after observing each advertisement. They respond the questionnaire according to their own theories of the product, participants are asked no need to worry whether their answer are correct or not. After the introduction page, following page was a consumer report-manipulation on objective knowledge. Half participants were exposed to a consumer report explaining objective knowledge of the target product. Remaining participants read an article excerpt from magazine irrelevant to the target. After that, before they started to 23.

(33) evaluate the target, participants were read two filler advertisements which were irrelevant to the study are excerpt from magazine. The third ad was target fictitious brand. Before they began to assess target ad, the instruction about manipulate on motivation. Participants were asked to complete evaluations of products ad. We also did suspicion check in the end of the questionnaire. No participants expressed any awareness about true of our study purpose. Upon Completion of the questionnaire, the participants were thoroughly debriefed, thanked for their participation, and dismissed.. 4.3.3Independent Variables 4.3.3.1. Motivation Motivation manipulated was embedded in the instruction before the third target ad. Participants in high motivation condition were told the page that there was a Portugal manufacturer of shampoo named “EUNOMIA” planning to launch their product in northern part of Taiwan around campus this summer .Besides, participants were instruct that “EUNOMIA” target market are in the ages from 19 to 25. Furthermore, only small samples of college students would be collected. Therefore, the manufacturer paid much attention to the appraisal of student’s opinion each individual’s opinion is important. On the other hand, low motivation participants were 24.

(34) informed in the page that the shampoo will launch only in Singapore. The manufacturer is not planning to launch their product to Taiwan market near days. Thus, we expected that participants under high motivation were convinced that the shampoo is going to be available for purchase in the near future. Therefore, they may pay much more attention in evaluating the target advertisement than participants under low motivation.. 4.3.3.2. Objective Knowledge For participants manipulated on high objective knowledge, they were asked to read an article excerpt from “Consumer Report of Taiwan” explaining the issue that most of people are misconceived about central arguments to evaluate the target product (i.e., large orange essence oil contain in shampoo). For participants who hold their own knowledge (low objective knowledge) about the claim in ad is low, they read an article irrelevant to target product. After reading the article, participants were asked to answer three multiple choice questions in order to test their level of comprehension and concentration. In addition, we also ask the participants about how much helpful the information provided by the consumer report in order to choose the right shampoo. Therefore, we expected that participants who expose to the article explaining the 25.

(35) central arguments to evaluate the target product will have more sufficient knowledge to target assessment than do participants who expose to an irrelevant article.. 4.3.3.3. Argument In order to conduct arguments, two augments for target product were identified in the pretest. The purpose of pretest is to select a misconceived statement “ 富含柑橘精油，並徹底洗淨毛髮”(perceived relatively strong by novices and relatively weak by experts) and neutral statement” 熱愛生命，不採動物測試”as the material for the main experiment. The result showed in chapter 4.2.. 4.2.3.4. Endorser According to Petty and Cacioppo (1980) research, we have to pay attention to non-message variables could also play a roles as central variables in the high elaboration continuum. The research conducted a study that the physical attractiveness (e.g. hair) of the endorsers of the shampoo might be relevant to judge the cereal merit of beauty product (e.g. shampoo). There are two endorsers in the target advertisement, one is a well-known golf player Ya-Ni Tseng, the other is an unknown college student. Both of them with short hair, we chose the picture of Ya-Ni Tseng who wears a cap. Both of endorsers are presented in the target advertisement separately, and irrelevant to the shampoo. 26.

(36) 4.3.4. Dependent Variables 4.3.4.1. Attitude Participants were asked to rate their attitude toward EUNOMIA shampoo was assessed by three different items with 7-point Likert scales (bad quality/good quality, unfavorable /favorable and negative/positive). Since ratings on these three items were highly inter-correlated (Cronbach’s α) were .864.These three items were averaged to form an index of overall attitude toward EUNOMIA (Petty & Cacioppo, 1983).. 4.3.4.2. Attitude Certainty Attitude certainty was assessed by three different items with 7-point Likert scales (no confident at all /extreme confident, not certain at all /very certain and not assured at all /very assured). Since ratings on these three scales were highly intercorrelated (Cronbach’s α=.956) these three scales were averaged to form an index of overall attitude certainty (Barden & Petty, 2008).. 27.

(37) 4.3.5 Manipulation Check 4.3.5.1. Motivation Participants’ responses on motivation was assessed by three different items with 7-point Likert scales anchored at -3 and +3(not concentrating/ concentrating, not careful/ careful, not thinking/ thinking) (Petty & Cacioppo,1983)asking participants how careful they are when evaluating the target ad. Since ratings on these three items were highly consistency (Cronbach’s α=.935). These three items were averaged to form an index of overall motivation.. 4.3.5.2. Objective Knowledge Participants’ levels of objective knowledge identified by determining the accuracy of their responses to quiz which utilizing two multiple-choice questions (Sidanius, 1988). First, “According to the consumer report, which one of ingredient’s effectiveness of shampoo has not been approved by any institution and research?” Second, “What kind of Plant Concentrate in the ingredient may have Side effects such as allergies, blackening of the scalp?”When participants answer both of the questions are right, we code “1”, otherwise, code “0”. We also make manipulation check on subjective knowledge. Participants were assessed by self-reports of how knowledgeable feel they are about the target. We 28.

(38) assessed by ask participants to report how well informed about target ad on a 7-point Likert scales anchored at -3 and +3 (completely uninformed / completely informed, unfamiliar/familiar and unknowing /knowing)(Petty& Cacioppo, 1986). Coefficient of Cronbach’s α=.918 showed high consistency. Therefore, three items were averaged to form an index of subjective knowledge.. 4.3.5.3. Argument According to Petty and Cacioppo (1986), favorable thoughts in response to strong arguments, in opposite, weak arguments leads people unfavorable though toward to target. To assess participants’ perceptions of argument quality contained in the advertisement that they perceived. We assessed by ask participants to report how well informed about target ad on a 7-point Likert scales anchored at -3 and +3 (unpersuasive/persuasive, weak reasons/strong reasons and not important at all/very important)(Petty & Cacioppo, 1979b).Coefficient of Cronbach’s α=.918 showed high consistency. The three items were averaged to form an index of argument quality.. 4.3.5.4. Endorser Attitude toward Ya-Ni Tseng is assessed by three7-point Likert scales anchored at -3. and. +3. (unfavorable/favorable,. attractiveness/unattractiveness. and. positive/negative) (Petty & Cacioppo, 1983) asking participants how the target is 29.

(39) perceived. Since ratings on these three items were highly consistency (Cronbach’s α=.922) these three items were averaged to form an index of overall attitude toward Ya-Ni Tseng and average citizen.. 30.

(40) CHAPTER 5 RESULT 5.1Manipulation Check 5.1.1Motivation There were two level of motivation we manipulated in the main experiment. First, we used coefficient of Cronbach’s α to test the reliability of items. Three items in motivation questions showed high consistency (Cronbach’s α=.935). Furthermore, high motivation condition participants (M=.55) were express greater processing activities in reading EUNOMIA shampoo advertisement than low motivation participants (M=.25).The result of one-way ANOVA presented in Table 3 showed that there was a significant different between two groups (M=.40；F= 6.92, p<0.01). Table 3 Descriptive Statistics for the Manipulation Check of the Level of Motivation Level of Motivation. N. Mean. Std. Deviation. High. 176. .55. 1.13. Low. 196. .25. 1.06. Table 4. ANOVA Outcome for the Manipulation Check of the Level of Motivation Source. Sum of Squares. df. Mean Square. F. Sig.. Motivation. 8.28. 1. 8.28. 6.92. .009. Error. 443.16. 370. 1.20. 31.

(41) 5.1.2Objective Knowledge Participants were required assess two shampoo related multiple-choice questions. In the test, participants who were in high objective knowledge condition (M= 1.86) their average scores was a significantly higher than participants in low objective knowledge condition (M= .77). The result is presented in Table 6. (M =1.31；F =312.45, p =.00). Table 5. Descriptive Statistics for the Manipulation Check of the Level of Objective Knowledge Level of Objective. N. Mean. Std. Deviation. High. 207. 1.86. .45. Low. 162. .77. .73. knowledge. Table 6. ANOVA Outcome for the Manipulation Check of the Level of Objective Knowledge Source. Sum of Squares. df. Mean Square. F. Sig.. Objective. 107.90. 1. 107.90. 312.45. .000. 126.74. 367. .35. knowledge Error. 32.

(42) 5.1.3Argument Quality To assess participants’ perceptions of argument quality contained in the target advertisement that they perceived. Coefficient of Cronbach’s α of three items in argument quality showed high consistency (Cronbach’s α=.856).It was confirmed the consistency of the items to measure the argument quality. Whether argument quality we manipulated was successful or not, we used pairwise comparison of 2-way ANOVA analyzed difference between groups. The result showed (see Table 7) experts perceived that the misconceived arguments were less persuasive (M= -.49) than when they encountered the neutral argument (M =.00) (M =-.61；p =.00). However, according to Table 8 we found that persuasiveness of argument had no significant different between misconceived and neutral when they were novices (M=.04, p =.82). Table 7. Descriptive Statistics for the Manipulation Check of the Argument Quality. Objective Knowledge. Argument Quality. Mean. Std. Deviation. N. Low Objective. Neutral Argument. -.23. 1.26. 33. Knowledge. Misconceived Argument. .54. .78. 38. High Objective. Neutral Argument. .00. 1.18. 48. Knowledge. Misconceived Argument. -.49. .94. 57. Total. Neutral Argument. -.094. 1.21. 81. Misconceived Argument. -.077. 1.01. 95. 33.

(43) Table 8. ANOVA Outcome for the Manipulation Check of the Argument Quality Objective. (I). Argument. (J) Argument. Knowledge Low. Mean Difference Std. Error. Sig.. (I-J) Misconceived. Neutral Argument. .042. .185. .820. Neutral Argument. -.612. .163. .000. Argument High. Misconceived Argument. According to ELM model we know that under low motivation conditions, attitudes appear to be affected by simple acceptance and rejection peripheral cues in the persuasion context and are less affected by argument quality (Patty and Cacioppo, 1980).Therefore, we did argument quality manipulation check under the condition of high motivation. The result of difference between neutral and misconceived argument in high motivation participants were in line with ELM showed in Table 9. Misconceived arguments perceived by novices assess it more persuasive than the neutral argument (M =-.77, p<01).These findings were consisted with our hypotheses that participants would feel more persuasive toward to the misconceived argument under novices rather than experts. On the contrary, Table 9 presented under the condition of high motivation, experts perceived that the misconceived arguments were less persuasive than when they encounter the neutral argument (M =-.49, p <.05).. 34.

(44) Table 9. Pairwise Comparisons for the Manipulation Check of the Argument Quality Under High Motivation Objective Knowledge. (I) Argument. (J) Argument. Mean Difference(I-J). Std. Error. Sig.. Low. Misconceived. Neutral. .776*. .249. .002. Argument. Argument. Misconceived. Neutral. -.491*. .205. .018. Argument. Argument. High. Dependent Variable: Argument Quality *p < .05.. .600 .400 .200 Neutral Argument .000 -.200. Low Objective Knowledge. High Objective Knowledge. Misconceived Argument. -.400 -.600. Figure 6. Manipulation Check of the Argument Quality under High Motivation Condition. 35.

(45) 5.1.4. Endorser The Cronbach’s α of three items about endorser attractiveness showed high consistency (Cronbach’s α=.922).Table 10 showed that the well-known golf player Ya-Ni Tseng was rated as more attractiveness (M=.35) than average citizen (M= -.95). It also showed a significant difference between attractive celebrity and average citizen (M=-.304；F= 103.006, p=.00).. Table 10. Descriptive Statistics for the Manipulation Check of the Level of Endorser Level of Endorser. N. Mean. Std. Deviation. Attractive Celebrity. 187. .35. 1.35. Average Citizen. 184. -.95. 1.11. Table 11. ANOVA Outcome for the Manipulation Check of the Level of Endorser Source. Sum of Squares. df. Mean Square. F. Sig.. Endorser. 157.29. 1. 157.29. 103.01. .000. Error. 563.47. 369. 1.53. 36.

(46) 5.2 Dependent Measures 5.2.1Attitude One of dependent measures in our study is target product attitude. Participants rated overall impression of the target product in four items included: quality, favorability, appraisals and purchase intention. These items to form attitude index were high inter-correlations (Cronbach’s α=.864).Each cell means and standard deviation of 16 conditions were summarized in the table.. Table 12. Cell Means and Standard deviation of Product Attitude High Subjective Knowledge High. Motivation. Low. Low. Objective. High. Low. High. knowledge Argument. Misconceived. Neutral. Misconceived. Neutral. Misconceived. Neutral. Misconceived. Neutral. quality. Argument. argument. argument. argument. argument. argument. argument. argument. Attractive celebrity. .22 (.65). .06 (.46). -.21 (.69). .34 (.64). .00 (.48). .45 (1.02). .06 (.91). .20 (.84). Average. .51. .01. -.32. .15. -0.43. -.04. -.25. -.15. citizen. (.61). (.51). (.71). (.89). (.62). (.79). (.72). (.94). Note. Dependent Variable: Target Attitude. The result of four-way analysis of variance was presented in Table12.There were some simple main effects revealed, that were, endorser attractiveness (F=6.36, p<.05),. 37.

(47) argument quality (F=5.00, p<.05).Except the main effects, there it was significant two-way interaction effect: Motivation. Endorser (F=5.53, p<.05).This interaction. reflected that level of endorser attractiveness had greater impact on attitudes under low motivation conditions than under high motivation (See Table 13). Furthermore, three way interaction effect: Objective Knowledge. Motivation. Argument. Quality (F=12.157, p=.001).We were give a further discuss about the potential effectiveness later in hypotheses. Table 13. The Outcome of Four-way ANOVA (Objective Knowledge, Motivation, Endorser, and Argument Quality) on Evaluations of Target Attitude. Source. Type III Sum. df. Mean Square. F. Sig.. 25.856a. 15. 1.724. 2.984. .000. Intercept. .521. 1. .521. .902. .343. Objective Knowledge. 1.212. 1. 1.212. 2.098. .148. Motivation. 1.147. 1. 1.147. 1.986. .160. Endorser. 3.671. 1. 3.671. 6.356. .012. Argument Quality. 2.888. 1. 2.888. 5.000. .026. Objective Knowledge. .720. 1. .720. 1.246. .265. .109. 1. .109. .189. .664. of Squares Corrected Model. * Motivation Objective Knowledge * Endorser. 38.

(48) 1.540. 1. 1.540. 2.666. .103. Motivation * Endorser. 3.196. 1. 3.196. 5.533. .019. Motivation *. .694. 1. .694. 1.201. .274. Endorser * Argument. .352. 1. .352. .609. .436. Objective Knowledge. .890. 1. .890. 1.541. .215. 7.022. 1. 7.022. 12.157. .001. .117. 1. .117. .202. .654. .138. 1. .138. .238. .626. .081. 1. .081. .141. .708. Error. 205.038. 355. .578. Total. 231.563. 371. Corrected Total. 230.894. 370. Objective Knowledge * Argument Quality. Argument Quality. * Motivation * Endorser Objective Knowledge * Motivation * Argument Quality Objective Knowledge * Endorser * Argument Motivation * Endorser * Argument Quality Objective Knowledge * Motivation * Endorser * Argument Quality. 39.

(49) Figure 7. The Interaction Effect of Motivation and Endorser on Evaluations of Target Product. 40.

(50) 5.2.2Attitude Certainty Another dependent measure in our study was target product attitude certainty. As a result, the Cronbach’s α of three items of attitude certainty index were high inter-correlations (Cronbach’s α=.956).Each cell means and standard deviation of 16 conditions were show as follow: Table 14. Cell Means and Standard Deviation of Product Attitude Certainty High Subjective knowledge High. Motivation. Low. Low. Objective. High. Low. High. knowledge Argument. Misconceived. Neutral. Misconceived. Neutral. Misconceived. Neutral. Misconceived. Neutral. quality. Argument. argument. argument. argument. argument. argument. argument. argument. Attractive. 1.13 (1.72). .21 (1.39). .70 (.95). .47 (1.13). .49 (1.09). .74 (1.18). .52 (1.23). .41 (1.50). .96 (.86). .24 (1.10). .65 (1.21). .92 (1.06). .25 (.97). .28 (1.44). .26 (1.47). .16 (1.17). celebrity Average citizen. Note. Dependent Variable: Target Attitude Certainty. The result of four-way analysis of variances were presented in Table14.First,There was an simple main effect which was Motivation (F=4.246, p<.05).Second, a significant of three-way interaction effect : Objective Knowledge Argument Quality (F=4.228, p<.05).. 41. Motivation.

(51) Table 15.. The Outcome of Four-way ANOVA (Objective Knowledge, Motivation, Endorser, and Argument Quality) on Evaluations of Target Attitude Certainty. Source. Type III Sum of. df. Squares. Mean. F. Sig.. Square. Corrected Model. 27.422a. 15. 1.828. 1.201. .268. Intercept. 95.272. 1. 95.272. 62.594. .000. Objective Knowledge. .056. 1. .056. .037. .847. Motivation. 6.462. 1. 6.462. 4.246. .040. Endorser. 1.218. 1. 1.218. .800. .372. Argument Quality. 3.224. 1. 3.224. 2.118. .146. Objective Knowledge *. .546. 1. .546. .358. .550. .767. 1. .767. .504. .478. 1.913. 1. 1.913. 1.257. .263. Motivation * Endorser. 2.898. 1. 2.898. 1.904. .168. Motivation * Argument. 3.814. 1. 3.814. 2.506. .114. Endorser * Argument. .322. 1. .322. .211. .646. Objective Knowledge *. .177. 1. .177. .116. .733. 6.436. 1. 6.436. 4.228. .040. .373. 1. .373. .245. .621. 1.148. 1. 1.148. .754. .386. Motivation Objective Knowledge * Endorser Objective Knowledge * Argument Quality. Quality. Motivation * Endorser Objective Knowledge * Motivation * Argument Quality Objective Knowledge * Endorser * Argument Motivation * Endorser * Argument Quality. 42.

(52) .006. 1. .006. Error. 541.857. 356. 1.522. Total. 674.667. 372. Corrected Total. 569.280. 371. Objective Knowledge *. .004. .949. Motivation * Endorser * Argument Quality. 5.3 Tests for Hypotheses Hypotheses on the condition of high motivation To conduct hypothesis one and hypothesis two, under high motivation condition, given high motivation as a control variable. Other two between-participants (Objective Knowledge: High vs. Low). Argument Quality (Misconceived vs.. Neutral) analyses of variance were conducted. Table 16. Descriptive Statistics of Between-Participants (Objective Knowledge and Argument Quality). Objective Knowledge. Argument Quality. Mean. Std. Deviation. N. Low Objective. Neutral Argument. .0379. .48056. 33. Knowledge. Misconceived Argument. .3947. .63574. 38. High Objective. Neutral Argument. .2396. .78204. 48. Knowledge. Misconceived Argument. -.2632. .69665. 57. Total. Neutral Argument. .1574. .67944. 81. Misconceived Argument. .0000. .74377. 95. Dependent Variable: Target Attitude. 43.

(53) Table 17. The Outcome of Two-way ANOVA (Objective Knowledge and Argument Quality) of Evaluations on Shampoo in the High Motivation Condition. Source. Type III Sum of. df. Squares. Mean. F. Sig.. Square. 11.747a. 3. 3.916. 8.605. .000. Intercept. 1.761. 1. 1.761. 3.871. .051. Objective Knowledge. 2.191. 1. 2.191. 4.814. .030. Argument Quality. .224. 1. .224. .492. .484. Objective Knowledge *. 7.778. 1. 7.778. 17.094. .000. Error. 78.267. 172. .455. Total. 90.938. 176. Corrected Total. 90.014. 175. Corrected Model. Argument Quality. Dependent Variable: Target Attitude Certainty. 5.3.1. Effect of Objective Knowledge Hypothesis 1a: A-B>0 Under conditions fostering high motivation and high subjective knowledge, product attitudes tend to be more favorable when novices encounter misconceived arguments and perceive them relatively strong than when experts encounter the same arguments and perceive them relatively weak. Hypothesis 1a, we assume that when people in high subjective knowledge may engage in effortful target relevant information as long as they are high in motivation. Even though novices interpreted the message was wrong, however, in our study, their 44.

(54) attitude toward to target product were more favorable compare to experts when they encounter misconceived argument. Simple main effect was used to test the influence of objective knowledge showed in Table 18.. Table 18. The Analysis Outcome of Hypothesis 1a Argument Quality. Mean Difference. Std. Error. F. Sig. .14. 21.687. .00. (Low Objective Knowledge High Objective Knowledge) Misconceived. .66. Argument. Note. Dependent Variable: Target Attitude. Simply expressed the result showed in Table18, two mean groups of attitude on condition A (M=.40) were significantly higher than B (M =-.26). As a result, Hypothesis 1a was confirmed(A-B=.66；p=00).. Hypothesis 1b: C-D=0 or D-C=0 Under conditions fostering high motivation and high subjective knowledge, product attitudes do not differ between novices encountering un-misconceived and neutral arguments and experts encountering the same arguments. In hypothesis1b, people under high motivation would assess the target judgment. 45.

(55) by scrutinizing the central merits of the target product. However, when people encounter neutral argument which needless expertise, we presume that neutral argument leads participant had neutral attitude. Thus, attitude about target product was no different between novices and experts when they processing neutral augment under high motivation. Table 19. The Analysis Outcome of Hypothesis 1b Argument Quality. Mean Difference. Std. Error. F. Sig. .15. 1.748. .19. (Low Objective Knowledge High Objective Knowledge). -.20. Neutral Argument. Note. Dependent Variable: Target Attitude. According to Table19, we simply expressed that two mean groups of attitude were no different between C (M=.03) and D (M =-.24). As a result, Hypothesis 1b was confirmed(C-D=-.20；p=.19).. 5.3.2 Effect of Argument Quality Hypothesis 2a: B-D<0 Under conditions fostering high motivation and high subjective knowledge, product attitudes tend to be less favorable when experts encounter misconceived arguments and perceive them relatively weak than when they encounter misconceived and neutral arguments.. 46.

(56) When high motivation group are able to process the argument, their attitude are influenced by argument quality. Therefore, when they encounter misconceived argument, experts can accurate interpreted the persuasive message then regard it is as absolutely weak argument elicit unfavorable attitude. Experts were perceived misconceived argument less favorable than when they encounter to neutral argument. Simple main effect analysis showed as follow. Table 20. The Analysis Outcome of Hypothesis 2a Objective knowledge. Mean Difference. Std. Error. F. Sig. .132. 14.47. .00. (Misconceived Argument Neutral Argument) High Objective. -.50. Knowledge. Note. Dependent Variable: Target Attitude. According to Table 20, we simply expressed that two mean groups of attitude measure of B(M=-.26) were significant less than group of D (M =-.24). As a result, Hypothesis 2a was confirmed(B-D=-.5；p=00).. Hypothesis 2b: A-C>0 Under conditions fostering high motivation and high subjective knowledge, product attitudes tend to be more favorable when novices encounter misconceived arguments and perceive them relatively strong than when they encounter un-misconceived and neutral arguments. 47.

(57) In hypothesis 2b, we assume that novice’s perception of high self-sense of knowledge leaded them believe that they are able to process products relevant information. However, novices are low in objective knowledge. Therefore, they perceived the misconceived argument is relatively strong than when they encounter the neutral argument. Simple main effect analysis showed as follow.. Table 21. The Analysis Outcome of Hypothesis 2b Objective knowledge. Mean Difference. Std. Error. F. Sig. .16. 4.94. .03. (Misconceived Argument Neutral Argument) Low Objective. .36. Knowledge. Note. Dependent Variable: Target Attitude. From Table 21, we simply expressed that two mean groups of attitude measure of A(M=.40) were significant more favorable than group of C (M =.03). It was Consistent with our hypothesis 2b was confirmed(A-C=.36；p<.05).. 48.

(58) A (.40). D (.24). C (.04). B (-.26) Figure 8. Target Attitude under the Condition of High Motivation. Hypotheses on the condition of low motivation Table 22. Descriptive Statistics of Between-Participants (Objective Knowledge and Endorser). Objective Knowledge. Endorser. Mean. Std. Deviation. N. Low Objective. Average Endorser. -.1900. .75010. 50. Knowledge. Attractive Endorser. .2917. .89018. 42. High Objective. Average Endorser. -.2063. .81036. 40. Knowledge. Attractive Endorser. .1349. .86815. 63. Total. Average Endorser. -.1972. .77305. 90. Attractive Endorser. .1976. .87616. 105. Note. Dependent Variable: Target Attitude. 49.

(59) 5.3.3Effect of Peripheral Cue Hypothesis 3: Under conditions fostering low motivation, product attitudes tend to be more favorable when the attractive endorser is exposed than when average endorser is received in the target ad. People who lack of motivation process the product relevant arguments are low involvement group. Regardless participants have ability to process central argument or not, their attitude change based on people who were simple inference of positive or negative peripheral cues. Therefore, the impact on attitude will influence by endorser attractiveness. According to the table 23, the outcome of low involvement individuals is consisted with ELM. Product attitude was more favorable when the attractive celebrity rather than average was presented (F = 11.534, p=.001). Table 23. The Outcome of Two-way ANOVA (Objective Knowledge and Endorser) of Evaluations on Shampoo in the Low Motivation Condition Source. Type III Sum of. df. Squares. Mean. F. Sig.. Square. Corrected Model. 8.180. 3. 2.727. 3.934. .009. Intercept. .011. 1. .011. .016. .900. Objective Knowledge. .353. 1. .353. .510. .476. Endorser. 7.995. 1. 7.995. 11.534. .001. Objective Knowledge *. .233. 1. .233. .336. .563. Endorser 50.

(60) Error. 132.399. 191. Total. 140.625. 195. Corrected Total. 140.579. 194. .693. Note. Dependent Variable: Target Attitude 0.4 0.3 0.2 0.1. Low Objective Knowledge. 0. High Objective Knowledge. Average Endorser. Attractive Endorser. -0.1 -0.2 -0.3. Figure9. Attitude toward Products under Low Motivation. 5.3.4Effect of Attitude Certainty Hypothesis 4a: People who are high in motivation and subjective knowledge tend to form attitudes with greater attitude certainty than their counterparts who are low in motivation but high in subjective knowledge. Hypothesis 4b: The attitude certainty of low objective knowledge (novices) is likely to be indifferent from that of high objective knowledge (experts) when they are high in subjective knowledge and high in motivation. 51.

(61) In the Hypotheses 4a, as long as people are high subjective knowledge, no matter they are experts or novices, participants in high motivation condition have significantly more attitude certainty than participants in low motivation condition (F=4.35, p<.05). Attitude change base on high elaboration about the target product and perception of high self-sense of knowing tent to be high attitude certainty. In the Hypotheses 4b, the result showed that there is no difference between experts and novices when both of subjective knowledge and motivation are high (F=.016, p = 0.90). As a result, Hypothesis 4a and 4b are consistent with our expected. Table 24. Descriptive Statistics of Between-Participants (Objective Knowledge and Motivation). Motivation. Objective Knowledge. Low Motivation. High Motivation. Total. Mean. Std. Deviation. N. Low Objective Knowledge. .44. 1.22. 93. High Objective Knowledge. .65. 1.30. 71. Low Objective Knowledge. .37. 1.35. 103. High Objective Knowledge. .70. 1.08. 105. Low Objective Knowledge. .40. 1.29. 196. High Objective Knowledge. .68. 1.17. 176. Note. Dependent Variable: Target Attitude Certainty. 52.

(62) Table 25. The Outcome of Two-way ANOVA (Objective Knowledge and Motivation) of Attitude Certainty Source. Type III Sum of. df. Squares. Mean. F. Sig.. Square. Corrected Model. 7.451a. 3. 2.484. 1.627. .183. Intercept. 105.318. 1. 105.318. 68.984. .000. Objective Knowledge. .024. 1. .024. .016. .900. Motivation. 6.648. 1. 6.648. 4.354. .038. Objective Knowledge *. .315. 1. .315. .206. .650. Error. 561.828. 368. 1.527. Total. 674.667. 372. Corrected Total. 569.280. 371. Motivation. Note. Dependent Variable: Target Attitude Certainty. 0.7 0.65 0.6 0.55. Low Objective Knowledge. 0.5. High Objective Knowledge. 0.45 0.4 0.35 0.3 Low Motivation. High Motivation. Figure10. The Effect of Target Attitude Certainty. 53.

(63) CHPATER 6 GENERAL DISCUSSION 6.1 Contribution Academic and managerial contributions of this research are as follows. The current research further extends the notion of multiple roles in ELM proposed by Petty and Wegener (1999). The past research in multiple roles mainly concentrates either on how the peripheral cues might serve as central arguments or biases or on the intrinsic and extrinsic factors which may influence the elaboration resources. However, in the current research, we suggest that central arguments may play different roles for novices (i.e., low involvement people without sufficient knowledge or capability to process product related information). Such a potential new role of central arguments has never been examined before. Particularly, we propose that the differential role of central arguments is likely to occur when novices’ subjective senses of knowledge toward the target are as considerable as those of experts. Given the same persuasive arguments, high motivation novices who are high in subjective knowledge may encode and process the persuasive messages totally differently from experts. Therefore, novices are likely to evaluate the target in a way significantly different from or even opposite to the evaluation of experts. In addition, the majority of researches in ELM holds that people low in 54.

(64) involvement (i.e., either low in motivation or low in ability to scrutinize information) tend to process peripheral cues rather than central arguments. However, the current research proposes and examines the potential scenario in which consumers low in ability (i.e., low in objective knowledge) are still likely to process central arguments as long as they are high in motivation and subjective knowledge. In recent years, the market has become highly competitive while products are getting low in diversity. Hence, marketers use merit of product to differentiate among competitors. However, in order to increase sales volume, some companies promote their products disregard the arguments are misconceived or not. The result of this study provides marketers more information on misconceived arguments affecting product judgment of low involvement consumers who have no objective knowledge but high motivation. Consequently, they might be influenced by misconceived argument. Misconceived argument can be combined with product position for high involvement consumers to identify the product then make poor judgment. Comparatively, low involvement consumers will hold positive attitudes towards the product. However, regardless which group, the attitudes certainty is high. This confident not only results from participants’ self-sense of knowledge and high motivation but also misleading information. Breaking the misconceived argument will bring novices with high motivation group a negative attitude but heighten their 55.

(65) attitude certainty. If we want to promote a product successfully, we have to avoid this mistake. Finally, in the previous attitude studies, involvement has generally been acknowledged as an important moderator leading to differential attitudinal processes and corresponding consequences of attitude (such as attitude strength and attitude certainty). In this study, however, we suggest that consumers low in involvement (e.g., novices) are likely to share the same attitude consequences (e.g., attitude certainty) as high involvement counterparts (e.g., experts), even when attitudes formed by low involvement consumers are in the opposite direction to those formed by high involvement consumers.. 56.

(66) 6.2 Limitations There are two limitations in our study. 1. In the proposal, we do not propose the occurrence of correction (cite). However, when the neutral arguments are appeared in the target advertisement, high involvement people maybe rely on irrelevant attractive endorser to evaluate the central merit of product. When attractive endorser is irrelevant to the target, the high involvement people will tend to decrease attitude toward to target. The weight of cues for high involvement people forming their attitude may be high; therefore, the bias may be easier to be detected. High objective knowledge people may try to bias away. So, correction is likely to happen. 2. There was some literature showing that attitudes based on high amounts of thought are higher attitude certainty (Petty& Jamie, 2008). For the impact on attitude certainty consequences such as resistance, persistence and predict behavior .Even though we postulated that processing information via central routes will lead high attitude certainty. However, we do not check that experts and novices attitude certainty consequences are same or not. Such as the level of resistance counter arguments, it may be different between experts and novices. For example, novices may be decrease attitude if a strong counterargument attacks from someone who has more knowledge. 57.