UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

How positively do they talk? : an investigation on how self-promotion motive induces consumers to give… Chung, Cindy Mann Yien 2000

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2001-610713.pdf [ 4.94MB ]
Metadata
JSON: 831-1.0090600.json
JSON-LD: 831-1.0090600-ld.json
RDF/XML (Pretty): 831-1.0090600-rdf.xml
RDF/JSON: 831-1.0090600-rdf.json
Turtle: 831-1.0090600-turtle.txt
N-Triples: 831-1.0090600-rdf-ntriples.txt
Original Record: 831-1.0090600-source.json
Full Text
831-1.0090600-fulltext.txt
Citation
831-1.0090600.ris

Full Text

H O W P O S I T I V E L Y DO T H E Y T A L K ? A N I N V E S T I G A T I O N O N H O W S E L F - P R O M O T I O N M O T I V E INDUCES C O N S U M E R S T O G I V E POSITIVE W O R D - O F - M O U T H by CINDY M A N N Y I E N C H U N G B.Comm (Hons.), The University of British Columbia, 1995 A THESIS SUBMITTED IN PARTIAL F U L F I L M E N T OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE F A C U L T Y OF G R A D U A T E STUDIES (The Faculty of Commerce and Business Administration) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH C O L U M B I A November 2000 © Cindy Chung, 2000 In presenting this thesis in partial fulfilment of the requirements for an ad-vanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Marketing Division, Faculty of Commerce and Business Administration The University of British Columbia 2053 Main Mall Vancouver, Canada V6T 1Z2 Date: A B S T R A C T This dissertation examines how product self-relatedness may induce positive word-of-mouth (WOM) behaviour. The main prediction was that product self-relatedness would increase the total amount of W O M , and cause consumers to exaggerate their opinions about products they liked. Three studies were conducted to test various hypotheses that related to the main prediction. A pilot study obtained initial evidence that products that reflected consumer self-concept led to more willingness to give W O M , and also more positively valenced W O M . Ownership and expectation of social evaluation on W O M opinion were found to interact with product self-relevance in influencing W O M valence. Using a new methodology, Study 1 replicated the finding in the pilot study that self-relatedness led to more W O M in total. However, W O M seemed more objective in this case, since subjects gave both more positive and more negative product evaluations when products reflected self-concept. Study 2 better established causality between self-relatedness and W O M behaviour. Subjects gave more W O M when they expected social evaluation. This study also demonstrated a positive self-presentational bias in W O M . Given that the subjects liked the product under evaluation, W O M was more positively valenced when advertising copy linked the product to the self than when it did not, or when subjects expected judgement by others than when they did not. Subjects seemed to exaggerate their preferences for the product when giving W O M in service of self-presentational goals. iii T A B L E OF C O N T E N T S Abstract i i Table of Contents i i i List of Tables vi Acknowledgements vii i Dedications ix C H A P T E R S O N E I N T R O D U C T I O N 1 1.1 Research Problem 1 1.2 Focus of Dissertation 3 1.3 Contributions 4 1.4 Organization of Materials 5 T W O E X I S T I N G R E S E A R C H O N W O R D - O F - M O U T H & F R A M E W O R K F O R C U R R E N T R E S E A R C H . 6 2.1 Existing Research on Word-of-Mouth 6 2.1.1 Research on Importance of Word-of-Mouth to Marketers and Consumers 6 2.1.2 Research on Product Attitudes and Word-of-Mouth Behaviour 8 2.1.3 Research on Motivation for Word-of-Mouth :... 9 2.2 A Motivational Framework for Word-of-Mouth 10 2.2.1 Role of Motivation in Driving Effortful Behaviour 10 2.2.2 Personal Relevance in Word-of-Mouth 13 2.2.3 Consequences of Personal Relevance in Word-of-Mouth 14 2.2.4 A Specific Case of Personal Relevance: Product Relevance to Self-Concept 15 2.2.5 Situational Aspects of Personal Relevance: Social Context for Word-of-Mouth 20 2.3 Summary 22 IV T H R E E P I L O T S T U D Y 24 3.1 Overview 24 3.2 Hypotheses to be Tested 24 3.3 Method 26 3.4 Results 29 3.4.1 Manipulation Check 29 3.4.2 Coding Written Word-of-Mouth 30 3.4.3 Total Word-of-Mouth 32 3.4.4 Valence of Word-of-Mouth 32 3.5 Discussion 35 3.5.1 Main Findings and Implications 35 3.5.2 Methodological Concerns and Implications 38 F O U R S T U D Y O N E 49 4.1 Hypotheses to be Tested 49 4.2 Overview 49 4.3 Data Collection in Three Stages 50 4.3.1 Initial Survey 50 4.3.2 Experiment Part I 51 4.3.3 Experiment Part II 54 4.4 Results 55 4.4.1 Data Cleaning and Coding 55 4.4.2 Manipulation Check 56 4.4.3 Total Word-of-Mouth 57 4.4.4 Valence of Word-of-Mouth 58 4.5 Discussion 60 4.5.1 Main Findings and Implications 60 4.5.2 Methodological Implications and Concerns 62 F I V E S T U D Y T W O 72 5.1 Hypotheses to be Tested 72 5.2 Considerations in Choosing a Product 74 5.3 Method 75 5.4 Results 80 5.4.1 Manipulation Check 80 5.4.2 Total Word-of-Mouth 80 5.4.3 Valence of Word-of-Mouth 81 5.5 Discussion 83 SIX F I N A L DISCUSSION 93 6.1 Summary of Main Findings 93 6.2 Implications of Current Research 95 6.2.1 Conceptual Implications 95 6.2.2 Managerial Implications 96 6.2.3 Methodological Implications 99 6.3 Limitations of Current Research ;.... 100 6.3.1 Use of Student Subjects 100 6.3.2 Concerns for Accuracy in Word-of-Mouth 100 6.3.3 Ownership 102 6.4 Future Directions 103 6.4.1 Other Motivational Goals 103 6.4.2 Other Samples 104 6.4.3 Motivations for Negative Word-of-Mouth 105 6.4.4 Consequences of W O M 107 Bibliography 108 Appendix 3-1: Questionnaire for Pilot Study 125 Appendix 4-1: Initial Survey for Study 1 127 Appendix 4-2: Final Questionnaire for Study 1 128 Appendix 5-1: Attitude Measure for Study 2 129 Appendix 5-2: Ads Used in Study 2 130 Appendix 5-3: Final Questionnaire for Study 2 132 VI LIST OF T A B L E S Table 3-1-1: Cell Means for TOTPROMO (Pilot Study) 40 Table 3-1-2: Analysis of Variance on TOTPROMO (Pilot Study) 40 Table 3-2-1: Cell Means for T O T D (Pilot Study) 41 Table 3-2-2: Analysis of Variance on TOT_D (Pilot Study) 41 Table 3-2-3: Analysis of Covariance on T O T D (Pilot Study) 42 Table 3-3-1: Analysis of Variance on POS-NEG (Pilot Study) 43 Table 3-3-2: Cell Means (Product Type by Ownership) for POS-NEG (Pilot Study) 43 Table 3-3-3: Cell Means (Product Type by Ownership by Context) for POS-NEG (Pilot Study) 44 Table 3-4-1: Cell Means for POS_EVAL (Pilot Study) 45 Table 3-4-2: Analysis of Variance on POS_EVAL (Pilot Study) 45 Table 3-5-1: Cell Means for N E G _ E V A L (Pilot Study) 46 Table 3-5-2: Analysis of Variance on N E G _ E V A L (Pilot Study) 46 Table 3-6-1: Cell Means for V A L E N C E (Pilot Study) 47 Table 3-6-2: Analysis of Variance on V A L E N C E (Pilot Study) 47 Table 3-7: Analysis of Covariance on POS-NEG (Pilot Study) 48 Table 4-1-1: Cell Means for TOTPROMO (Study 1) 65 Table 4-1-2: Analysis of Variance on TOTPROMO (Study 1) 65 Table 4-2-1: Cell Means for TOT_D (Study 1) 66 Table 4-2-2: Analysis of Variance on TOT_D (Study 1) 66 Table 4-2-3: Analysis of Covariance on TOT_D (Study 1) 67 Table 4-3-1: Cell Means for POS-NEG (Study 1) 68 Table 4-3-2: Analysis of Variance on POS-NEG (Study 1) 68 Table 4-4-1: Cell Means for P O S _ E V A L (Study 1) 69 Table 4-4-2: Analysis of Variance on POS_EVAL (Study 1) 69 Table 4-5-1: Cell Means for N E G _ E V A L (Study 1) 70 Table 4-5-2: Analysis of Variance on N E G _ E V A L (Study 1) 70 Table 4-6: Analysis of Covariance on POS_EVAL (Study 1) 71 Table 4-7: Analysis of Covariance on N E G _ E V A L (Study 1) 71 Table 5-1: Major Product Categories Mentioned by Subjects in Pilot Study and Study 1 85 Table 5-2: Cell Size for Data Analysis (Study 2) 85 Table 5-3-1: Cell Means for U _ C O N C E R N (Study 2) 86 Table 5-3-2: Analysis of Variance on U _ C O N C E R N (Study 2) 86 Table 5-4-1: Cell Means for O T H E R G O O D (Study 2) 87 Table 5-4-2: Analysis of Variance on O T H E R G O O D (Study 2) 87 Table 5-5-1: Cell Means for TOT_D (Study 2) 88 Table 5-5-2: Analysis of Variance on TOT_D (Study 2) 88 Table 5-6-1: Cell Means for POS-NEG (Study 2) 89 Table 5-6-2: Analysis of Variance on POS-NEG (Study 2) 89 Table 5-7-1: Cell Means for POS_EVAL (Study 2) 90 Table 5-7-2: Analysis of Variance on POS_EVAL (Study 2) 90 Table 5-8-1: Cell Means for N E G _ E V A L (Study 2) 91 Table 5-8-2: Analysis of Variance on N E G _ E V A L (Study 2) Table 5-9: Analysis of Covariance on POS-NEG (Study 2) Table 5-10: Cell Means for Product Attitudes (Study 2) .... viii A C K N O W L E D G E M E N T S First and foremost, I would like to thank my dissertation supervisor, Professor Peter R. Darke for his guidance throughout my preparation of the thesis document. I thank him for advancing my ability in conceptualizing problems, and showing me the essentials of doing good research, and the numerous "tricks" for conducting sound and polished experiment with human subjects. M y sincere thanks also go to my dissertation committee members, Professor Rick Pollay and Professor Darrin Lehman. Their input has constructively challenged and helped to refine what I attempted to establish in my dissertation research. I would also like to thank Professor Dale Griffin and Professor Ralph Hakstian for their advice on data analysis for this research, and Professor Chuck Weinberg for his support at every stage of my pursuit of the Ph.D. degree. Laurence Ashworth and Robin Ritchie have patiently read through this document, drawing my attention to places for improvement. I thank them for doing that. As well, I thank Lorenzo Garlappi, Jason Ho, Freddy Lee, and Yong Liu for their support. Finally, I would like to give my heart-felt thanks to Olaf my best beloved. I thank him for all his support, and for simply being there both in times of joy and stress. To my mother, who has shown me the wonder of knowledge; my stepfather, the silent supporter; and my father, whom I wish had lived to see this. 1 C H A P T E R O N E I N T R O D U C T I O N 1.1 Research Problem Word-of-mouth (WOM) is the informal communication between private parties concerning the evaluation of goods and services (Dichter 1966; Fornell and Bookstein 1982; Singh 1988; Westbrook 1987). It is a key marketing phenomenon that has been found to facilitate the sales of a wide range of products including professional services (Smith and Meyer 1980), movies (Mizerski 1982), automobiles (Swan and Oliver 1989), travel or vacation destinations (Gitelson and Crompton 1983; Nolan 1976), and innovations (Rogers 1983). One survey by a marketing consulting firm found that 60% of respondents bought products solely on the basis of personal recommendation, whereas only 29% suggested they had ever bought a product solely due to a TV advertisement (Globe and Mail 1999). Furthermore, interviews with management from various industries have confirmed that marketers acknowledge the significance of positive W O M for a wide variety of products and services, such as food processing, computers, beer, real estate, prescriptions, and banking services. W O M often has a greater impact on consumers than media-based communications at least partly because it is considered more reliable and trustworthy (Busch and Houston 1985; Day 1971). The chief executive officer of The Henley Centre, a unit of the British advertising agency WPP Group PLC, 2 once said, "[WOM is] probably the most credible means of recommending a product or getting recommendation.. .It's a very powerful medium and a lot of people in areas where they don't know who to trust, where they don't have their own expertise, would much rather use word-of-mouth than anything else" (Globe and Mail 1999). Given the powerful influence of W O M , companies benefit most when consumers engage in positive W O M and are silent about their dissatisfaction. However, this is not an easy goal to achieve. After all, consumers (or potential W O M givers) encounter numerous products and services. They may talk about both products that they like and products that they do not like. In fact, some research has suggested consumers are particularly likely to talk about products with which they are dissatisfied. Even when they do like a product, consumers are not always willing to give W O M on it. Although it has been commonly assumed that product attitudes drive W O M directly, product attitudes do not always lead to W O M behaviour (Swan and Oliver 1989). This dissertation examines how motivation may lead consumers to give positive W O M . Specifically, it investigates how self-promotional motives may induce more positive W O M , given that consumers already hold positive attitudes towards a product. A conceptual framework will first be developed to show how motivation can drive behaviour. The impact of an important source of motivation, personal relevance, on W O M will then be examined. In addition, two specific cases in which personal relevance is expected to induce the self-promotion motive in W O M behaviour will be discussed. 3 1.2 Focus of Dissertation A pilot study and two formal studies were designed to test predictions based on the conceptual framework developed. In general, this framework suggests that personal relevance helps to motivate W O M , and may also lead to a positive bias, especially when self-presentational concerns are high and product attitudes are positive. Findings in the pilot study provided initial support for the notion that products that reflected consumer self-concept (or self-related products) led to more W O M behaviour than products that were not related to their self-concept. Self-related products also induced more positive W O M . Ownership was shown to play some role in influencing W O M valence. The pilot study showed that product self-relatedness and product ownership were more likely to affect W O M valence when the givers expected their opinions to be judged by an expert. Study 1 tested the effect of product-self relatedness (i.e., when self-presentation was particularly important) and ownership on W O M behaviour in a more realistic setting. Finally, Study 2 induced a self-promotion motive through advertising, and used self-promotional concerns in the social context in which W O M was given. Given that subjects had positive attitudes towards the target product, advertising copy that linked "good taste" to the consumption of this product induced more positive W O M than copy that did not include this link. Suggesting that W O M was a means of establishing oneself as an opinion leader also led to more positive W O M . These last findings helped to better establish causality between self-promotion motivation and W O M behaviour, and confirmed the influence of the self-relevant motives on W O M valence. 4 1.3 Contributions The motivational framework developed in this dissertation makes a number of conceptual contributions to W O M research. While this work acknowledges the importance of product attitudes in inducing W O M (a relationship that previous W O M studies have shown), it focuses on additional motivational factors that determine whether consumers will actually be willing to engage in W O M given their attitudes towards a product. Specifically, the current work adds to the existing research by showing that W O M is more likely to occur when it implies personal significance for the W O M giver, in the sense that W O M conveys or promotes his or her self-image. Given that a W O M giver likes a product, positive W O M is more likely to occur when the giver perceives some personal connection with the product, compared to circumstances in which there is no obvious personal connection. Furthermore, W O M may be biased or exaggerated by the giver in an attempt to self-enhance. The current research also provides managerial insights for marketers. It suggests how management can encourage favorable W O M . For instance, firms may induce positive W O M by encouraging consumers to associate products with an important aspect of their self-concept, such as showing their "good taste," eating healthy, or making smart decisions. Finally, this dissertation makes methodological contributions by designing a new methodology (Study 1) for observing actual W O M behaviour at the time it occurs. This methodology introduces a certain amount of realism into the design, while still allowing for experimental manipulation of the independent variables. 1.4 Organization of Materials The dissertation is organized as follows. Chapter 2 reviews existing research on W O M behaviour that is relevant to the current research, and develops a conceptual framework for W O M motivation. This framework is then used to develop several hypotheses. Three studies that test these hypotheses are also described. Chapter 3 reports a pilot study, which sought to test how product relatedness, ownership, and self-presentational concerns may affect total W O M and W O M valence. Chapter 4 presents findings from Study 1, which attempted to replicate the main findings observed in the pilot study, and improved on the W O M data collection by directly and unobtrusively measuring W O M behaviour. Chapter 5 presents Study 2, which better established the causality between self-promotion motives and W O M . Finally, Chapter 6 summarizes the major findings, discusses the conceptual, methodological, and managerial implications of this dissertation, and suggests future research directions based on this research and the proposed framework as a whole. 6 C H A P T E R T W O E X I S T I N G R E S E A R C H O N W O R D - O F - M O U T H A N D F R A M E W O R K F O R C U R R E N T R E S E A R C H In this chapter, existing studies on W O M that are relevant to the current research are reviewed, and opportunities for their integration and extension are identified. A motivational framework for studying W O M is then developed. 2.1 Existing Research on Word-of-Mouth 2.1.1 Research on Importance of Word-of-Mouth to Marketers and Consumers W O M has been defined as the informal communication between private parties concerning the evaluation of goods and services (Dichter 1966; Fornell and Bookstein 1982; Singh 1988; Westbrook 1987). During such communication, very different types of information can be transmitted between the W O M giver and the W O M recipient. For instance, the W O M message may contain objective facts and details about product purchase (e.g., prices, locations and opening hours of stores, etc.), or product consumption (e.g., serving sizes, methods and frequency of use, etc.). Importantly, this information often includes the W O M giver's subjective evaluation and recommendation of the product or service being discussed. Both factual information and more evaluative information may have an important effect on consumer purchases. 7 Many marketers view information conveyed in W O M to be "the most important marketing element that exists" (Alsop 1984), and "the most powerful force in the marketplace" (Silverman 1997). It is also important to consumers. Consumers commonly consider personal sources of information such as W O M to be very important (Katona and Mueller 1954; Katz and Lazarsfeld 1955; LeGrand and Udell 1964; Price and Feick 1984; Robertson 1971; Thorelli 1971). Information from an impartial and knowledgeable third party can significantly affect consumers' choice of products and services (Fferr, Kardes, and Kim 1991; Katona and Mueller 1954; Keaveney 1995; Kiel and Layton 1981; Price and Feick 1984). Consumers tend to rely heavily on personal sources when making high-risk purchases (Arndt 1967, 1968; Cunningham 1964, 1967; Lutz and Reilly 1973; Perry and Hamm 1969; Roselius 1971). For marketers, W O M can be an economical way of making their products known. It is much cheaper than other forms of mass promotion, such as advertising campaigns (Heckman 1999). Person-to-person communications have also been found to be more effective than impersonal media communications in transmitting information and changing opinions (Cantril and Airport 1935; Doob 1948; Knower 1935, 1936; Wilke 1934). Additionally, W O M is particularly effective in the diffusion of new product information, and aiding in the adoption of innovations (Arndt 1967; Coleman, Katz, and Menzel 1966; Engel, Kegerreis, and Blackwell 1969; Feldman and Spencer 1956; Katz and Lazarsfeld 1955; Sheth 1971). Given that W O M has an important impact on 8 consumers, it seems pertinent for marketers to understand what leads consumers to talk about their products. 2.1.2 Research on Product Attitudes and Word-of-Mouth Behaviour Most of the research on W O M acknowledges the significance of product attitude as an antecedent to W O M behaviour. In a series of in-depth interviews, Dichter (1966) found that consumers gave W O M because they had pleasurable experiences with (or positive attitudes towards) certain products and sought to express their joy and release their excitement (p. 149). Holmes and Lett (1977) showed that subjects with favorable attitudes towards the target brand of coffee were more likely to talk positively about the product. Those with unfavorable attitudes said negative things about products. In addition, a TARP report on the Coca-Cola Company confirmed that consumers spread positive W O M about the company after their complaints were handled in a satisfactory manner, and negative W O M when they were dissatisfied (1979). Richins (1983) showed that consumers who were dissatisfied with clothing and appliances also engaged in negative W O M . The satisfaction-WOM relationship was further validated by Swan and Oliver's study on automobile postpurchase communication (1989), Mooradian and Olver's study on satisfaction with automobiles (1997), and Anderson's research with American and Swedish consumers (1998) on a variety of services. Overall, it is clear that product attitudes often have an important impact on W O M valence. Consumers who have good experiences with products or services tend to talk 9 positively about the products or the companies. Conversely, consumers who suffer bad consumption experiences tend to spread negative W O M . Yet, it is conceivable that consumers who are satisfied with a product and have positive attitudes towards it may still fail to recommend it to friends or relatives. In fact, in Holmes and Lett's study (1977), only 38% of respondents with favorable attitudes towards coffee talked to other people about the product. Swan and Oliver (1989) found that satisfaction and perceived fairness explained between 35% and 49% of the variation in W O M behaviour. This meant there was still a lot of unexplained variance to account for. They found that more than half of their respondents who were relatively satisfied with their automobiles did not praise the marketer. One reason that positive attitudes may not lead to positive W O M is that consumers may need to feel motivated to give W O M , in addition to having positive opinions about the product. This is the central idea examined in this dissertation. 2.1.3 Research on Motivation for Word-of-Mouth The existing W O M literature contains little empirical work that explicitly explores the motivation for giving W O M . Those studies that exist tend to be descriptive in nature, with little systematic investigation. Most of these studies use simple interview techniques in which consumers are directly asked to speculate on their reasons for having engaged in W O M behaviour. Some research does not even distinguish between the motivation for contributing W O M , and those factors that influence W O M seeking behaviour (which emphasize the needs of the W O M recipient) (e.g., Bristor 1990). 10 Among the studies that do exist, some work has suggested various goals that might motivate people to initiate W O M conversations. For instance, respondents in a study by Dichter (1966) suggested that W O M was a means to gain attention, show connoisseurship, seek confirmation of judgement, or even assert superiority and power (p. 150). These goals might be broadly conceived as attempts by the consumers to promote their self-concept. Other researchers have suggested that consumers are more likely to initiate topics for conversations that relate to their self-concept (Arndt 1967). Past studies have shown that W O M may also be initiated to facilitate social relationships. For instance, consumers said they sometimes spread W O M to help the recipient make better purchase decisions (Arndt 1967), or to express neighborliness, care, friendship, or love (Dichter 1966, p. 151; see also Arndt 1967; Engel et al. 1969). 2.2 A Motivational Framework for Word-of-Mouth The literature review indicates that W O M is an important aspect of consumer behaviour that is significantly influenced by product attitudes. However, it also seems that attitudes do not always predict W O M behaviour, and some initial research suggests that the attitude-WOM link may be enhanced by the motivational goals of the W O M giver. 2.2.1 Role of Motivation in Driving Effortful Behaviour Research has shown that a large number of factors can determine whether or not attitudes are good predictors of actual behaviour (see Kraus 1995 for a meta-analysis of the 11 existing empirical research on attitudes). Indeed, the question of whether attitudes predict behaviour has been the topic of considerable debate since the 1960s (e.g., Deutscher 1966; Wicker 1969). Some researchers have challenged the predictive power of attitudes on overt behaviour, and argued that attitude-behaviour correlations are generally low (Mischel 1968; Wicker 1969). Other researchers have offered explanations for the low consistency between attitude and behaviour, and have made recommendations on how the attitude-behaviour link may be improved (e.g., Ajzen 1987; Ajzen and Fishbein 1977; Fazio and Zanna 1981; Kelman 1974). Of particular interest to the current research is some evidence that individual differences in terms of people's motives may influence whether or not attitudes are consistent with behaviour. For instance, attitude-behaviour correlations are generally higher among low self-monitors whose behaviour tends to be motivated by their internal states, than among high self-monitors who are usually motivated by situational cues (Snyder and Kendzierski 1982). The theory of reasoned action (Ajzen and Fishbein 1977) acknowledges that aside from personal attitudes, people's motivation to comply with others' beliefs about the attitude object also plays a role in determining behavioural intentions. Subsequent studies have obtained empirical evidence supporting the theory's ability to predict relevant behaviour (Holbrook and Havlena 1988; Shimp and Kavas 1984). Thus, there is initial evidence that individual motivations and goals can moderate the effect that attitudes have on actual behaviour. 12 In fact, it is generally accepted that motives are "an underlying reason for behaviour." They are essential for effortful human activities because people have a finite amount of energy that they can direct towards various goals (Solomon 1996). Depending on the particular needs that people have and the end-states or goals they want to attain, different behaviours may be motivated (Maslow 1962). Needs may be biogenic, motivating people to act to obtain basics for life, such as food, water, air, and shelter. They may also be psychogenic, motivating people to strive for status, power, and affiliation. The affiliation motive, for example, may lead people to such activities as team sports or bar frequenting to alleviate loneliness (Schachter 1959). People high in the power motive may engage in activities that give them control of their environment (Fodor and Smith 1982). In terms of social judgement and decision-making, personal relevance is regarded as one of the most important motivational variables, affecting information processing and reasoning (e.g., Kruglanski and Freund 1983; Kunda 1987; Petty and Cacioppo 1986; Pyszczynski and Greenberg 1987; Sorrentino and Higgins 1986). People who view their current circumstances as relevant to their goals in some way are generally more willing to expend effort in order to achieve their goals. Personal relevance has also been labelled personal involvement (e.g., Apsler and Sears 1968; Sherif, Kelly, Rodgers, Sarup, and Tittler 1973) and outcome relevance (e.g., Chaiken, Liberman, and Eagly 1989; Chaiken and Stangor 1987; Johnson and Eagly 1990; Petty and Caciopppo 1986). It occurs when an issue or behaviour is related to important goals, or when it is expected to bear 13 significant consequences (Apsler and Sears 1968). One goal that may increase personal relevance is the need to make accurate and valid social judgements. Other types of goals may also increase personal relevance and thereby drive behaviour. For instance, the need to present oneself to others in a positive light is often an important goal (e.g., Chaiken et al. 1989; Kruglanski and Ajzen 1983; Kruglanski and Klar 1987; Pyszczynski and Greenberg 1987). 2.2.2 Personal Relevance in Word-of-Mouth W O M will be personally relevant when communication bears significant consequences for the W O M giver. This can occur for various reasons. First, the W O M giver who has consumed the product may have acquired greater product knowledge, and developed specific product attitudes that may be communicated to other consumers. Second, the product may be personally relevant because it conveys important aspects of the W O M giver's self-concept. This means that the W O M giver may be able to reveal these aspects of the self-concept by the very act of giving W O M . Third, W O M may occupy a significant part of daily discourse, and enable the W O M giver to socialize with other consumers and establish friendship. Finally, it may be clear that the W O M recipient will heavily weigh the information given towards a purchase decision, meaning that the W O M giver will be subject to verification by the W O M recipient, or be held accountable for the decision made. 2.2.3 Consequences of Personal Relevance in Word-of-Mouth Personal relevance may have two major effects on W O M behaviour. First, it may increase W O M givers' effort in giving W O M . This could translate into a higher likelihood of giving W O M , and be expressed by greater detail in the information given about product purchase and consumption. Second, personal relevance may also introduce biases in W O M . That is, the W O M giver may tend to be more extreme in his or her expression of opinions than another W O M giver for whom the product is not personally relevant. These biases should affect the valence of W O M (i.e., the positivity or negativity of WOM) depending on the particular motivational goals that W O M givers wish to achieve. For instance, i f the goal is to promote an image of a smart shopper, then the W O M giver may be biased towards giving product information and evaluations, or reporting consumption experiences that enhance this image. In particular, i f the W O M giver knows of a good product, saying good things about it to others may reflect positively on the giver. The resulting W O M would be mainly positive. If, on the other hand, the product is bad, the W O M giver may say negative things about it in order to reveal his or her knowledge and expertise as a smart shopper. The resulting W O M would be mainly negative. If the goal for W O M communication is to facilitate friendship, then the W O M giver may choose to consider his or her friends' opinions. Under this circumstance, the W O M given will more likely be consistent with the others' attitudes. Also, i f the W O M giver is motivated to help a W O M recipient make an important purchase decision, or if the W O M giver is 15 aware that he or she will be held accountable for the decision made, more effort may be put into processing product information and consumption experiences before giving W O M (Tetlock 1983, 1985). W O M is also likely to cover a more balanced account (i.e., including both positive and negative attributes of the product) in an attempt to be complete and objective. The following section will focus on how a self-promotion goal may be induced through emphasizing the personal relevance of some products, or by particular demands of the social context in which W O M is given. 2.2.4 A Specific Case of Personal Relevance: Product Relevance to Self-Concept Effect of Product-Self Relationship on Word-of-Mouth Behaviour Products may serve both utilitarian and symbolic functions for consumers (Holbrook and Hirschman 1982; Levy 1959; Park, Jaworski, and Maclnnis 1986; Shavitt 1990). Indeed, many authors have suggested that the symbolic meaning of a product can often be more important than the practical use that consumers derive from them (Benedict 1934; Form and Stone 1957; Goffman 1951; Richins 1994; Veblen 1899). Symbolic consumption researchers have been studying the relationship between products and consumer self-concept for more than four decades (Gardner and Levy 1955; Levy 1959), and have generally accepted the idea that there is an association between products and some 16 aspects of the self-concept (e.g., Bern and Funder 1978; Biel 1993; Dolich 1969; Landon 1974; Ogilvy 1983; Plummer 1985; Prentice 1987; Snyder and Gangestad 1986, etc.). Self-concept refers to the total organized body of information that people have about themselves (Baumeister 1995). It is a multifaceted collection of self-knowledge, self-conceptions, images, or schemas (e.g., Epstein 1980; Greenwald 1982; Kihlstrom and Cantor 1984; Markus 1977, 1983; Markus and Nurius 1987; Markus and Sentis 1982; Schlenker 1980; Showers 1992; Strykers 1987, etc.). In addition to a person's belief that he or she is independent, intelligent, or overweight, the person's belief that he or she is a smart shopper, or a consumer with "good taste" also seems relevant to the self-concept (Schindler 1998). People are typically very interested in promoting their self-concept (e.g. Arkin and Baumgardner 1986; Baumeister 1982b; Baumeister and Tice 1986; Goffman 1959; Jones and Pittman 1982; Schlenker 1980; Schneider 1981; Tedeschi 1986). They try to convey to others impressions of themselves that are biased in the direction of the identity that they view as positive (Schlenker 1980; Markus and Nurius 1986). They may engage in biased information processing in order to maintain views that are complimentary to their self-concept (e.g., Chaiken, Giner-Sorolla, and Chen 1996; Miller and Ross 1975; Zuckerman 1979), and are generally "self-enhancing, preferring and seeking out positive information about themselves" (Markus & Wurf 1987, pp. 318-319; see also Snyder, Higgins, and Stucky 1983; Taylor and Brown 1988). These biases may also involve reinterpreting, distorting, embellishing, or omitting information in order to avoid any 17 threats to the self-concept (Greenwald 1980). Moreover, there is considerable evidence that people are unrealistically optimistic about various domains in life such as health, academic achievement, and financial stability (e.g., Heine and Lehman 1995; Taylor and Brown 1988; Weinstein 1980, 1982). Furthermore, people's motivation to self-promote is believed to be particularly great when they are self-aware. They try harder to achieve specific goals when the implications of that behaviour for self-evaluation are particularly salient (Wicklund 1975; Wicklund and Duval 1971). There is also evidence that self-awareness makes people more likely to take credit for positive outcomes from tasks. For instance, a study by Federoff and Harvey (1976) showed that subjects were more likely to attribute success in treating a phobic patient to themselves when they were high in self-awareness. Self-promotion can be rewarding to people who attempt to achieve this goal. It may bring influence, power, loyalty, and other highly desirable social and material outcomes (Schlenker 1980; Tedeschi 1981). Importantly, self-promotion helps to maintain or enhance people's self-esteem (Schneider 1969; Leary and Kowalski 1990), reinforcing the belief that they are competent and possess valuable abilities that allow them to function effectively in the world. Self-esteem has been considered one of the core aspects of the self-concept (Greenwald, Bellezza, and Banaji 1988). A large variety of product categories have been examined with respect to the association between product and various important aspects of the self-concept (e.g., Belk 1988; Csikszentmihalyi and Rochberg-Halton 1981; Grubb and Hupp 1968; Grubb and 18 Stern 1971; Landon 1974; Sirgy 1982; Sirgy and Su 2000). These studies suggest that the purchase, display, and use of certain goods can communicate symbolic meaning to both consumers themselves, as well as to other people (Grubb & Grathwohl 1967). In terms of product choice, some consumers seem to prefer products that match their self-concept, i.e., who consumers actually perceive they are (Birdwell 1968; Douglas, Field, and Tarp 1970; Grubb and Hupp 1968; Landon 1974; Maslow 1962; Ross 1971). Self-image congruence models suggest that, given a choice, consumers tend to prefer products that match their self-concept (Onkvisit and Shaw 1987). However, at other times, consumers also like products that help them achieve a self-image that is more of an ideal (Adler 1930; Allport 1955; Fromm 1956). This view is consistent with self-completion theory, which suggests that consumers tend to acquire and display symbols (or products in the context of the current research) that complete a certain aspect or dimension of their personality that may be inadequate or missing from their actual self-concept (Wicklund and Gollwitzer 1982). Either way, existing research acknowledges that some products are more personally relevant to consumers than others because they relate to at least some aspect of consumer self-concept. Given that goals relating to the self-concept can act as an important source of motivation, it might reasonably be expected that consumers who view a product as related to their self-concept in some way would be more likely to engage in W O M about it. Of particular interest to the current research, W O M will aim at maintaining, justifying, or convincing others of the goodness of the product i f the consumers like the product (or 19 have positive attitudes towards it). Furthermore, good products that are believed to reflect an important aspect of the self-concept are not only likely to increase the general level of motivation to engage in W O M , but are also likely to cause consumers to focus on more positive portrayals of these products when giving W O M . Thus, consumers may exaggerate their evaluation of the products they like when products have implications for self-promotion due to their connection to the self-concept. For example, consumers who take pride in leading an active and healthy lifestyle may be more likely to emphasize the positive aspects of health-related products that they like (such as running shoes or low fat foods) to convey the value that they place on such a lifestyle. Therefore, a W O M giver who sees a product as self-related may be expected to give more positive W O M than another W O M giver who likes the product just as much, but is not motivated to self-promote. Effect of Ownership on Word-of-Mouth Behaviour Ownership or possession of a product may also increase its personal relevance (Barone et al. 1999; Belk 1988; Kassarjian 1978; Wallendorf and Arnould 1988). Researchers have suggested that ownership enhances the relationship between the image of the product and some aspects of the owner's self-concept (Barone et al. 1999; Belk 1988; Wallendorf and Arnould 1988). Consumers may even regard some of the products they possess as aspects of their "extended-self." That is, some possessions may actually be self-defining 20 (Belk 1988), and provide a means to express a consumer's personal values (Richins 1994). Products that reflect aspects of a consumer's self-concept are more likely to be taken as part of the extended self than products that serve utilitarian purposes. Ownership is expected to increase effort in giving W O M opinions. W O M for the purpose of self-promotion would seem to be a likely outcome of product ownership because ownership increases the personal relevance of products for consumers. People who own a product (compared to those who do not own it) are more likely to provide W O M to others, at least in part because they are more involved with the product. Product knowledge may also play some part in this effect, since owners may give more W O M simply because they are more knowledgeable. This possibility was therefore also examined. In addition, biases in W O M are more likely to be observed on owned products than those that are not owned. As such, a consumer is expected to give more positive W O M on an owned product than a product that he or she does not own, although he or she may have a positive attitude towards both kinds of products. 2.2.5 Situational Aspects of Personal Relevance: Social Context for Word-of-Mouth Effect of Expected Judgement on WUlingness to Give Word-of-Mouth The personal relevance of W O M may increase if a person gives W O M in an environment that suggests others will judge their opinions. The expectation of social 21 evaluation seems likely to increase a W O M giver's self-awareness, making him or her think about and elaborate on the implications of recent events (in this case, purchase or consumption experience) in relation to the self (Hull and Levy 1979). As a result of greater effort in processing information regarding the consumption event, a W O M giver is expected to be more willing to give W O M and to give more detail regarding consumption information and experiences. Effect of Expected Judgement on Word-of-Mouth Valence A social context that suggests social evaluation may also induce self-promotion attempts that lead to a greater bias in W O M communication. After all, social feedback is an important source of self-esteem (Baumeister and Jones 1978). A person who performs an important task and receives approval from others around him or her will boost his or her self-esteem. There is also evidence that people will self-promote to both friends (Tice, Butler, Muraven, and Stillwell 1995) and strangers (Tedeschi 1986), and are particularly likely to ingratiate themselves with others who are of high status (such as bosses and teachers), and upon whom they depend for valued outcomes (Baumeister and Jones 1978; Bohra and Pandey 1984; Gardner and Martinko 1988; Hendricks and Brickman 1974; Jones, Gergen, Gumpert, and Thibaut 1965; Stires and Jones 1969). In the context of W O M , motivation to self-promote should be higher when the W O M giver expects to be evaluated by others on the basis of what they say, especially when those 22 others are particularly worth impressing due to their status or power (Gergen and Wishnov 1965; Schneider 1969; Sengupta, Dahl, and Gorn, forthcoming). Given that a consumer already holds a positive attitude toward a product, he or she will give more positive W O M when social evaluation is expected versus when it is not expected. 2.3 Summary This chapter has reviewed the existing research that is relevant to the current thesis, and developed a motivational framework for understanding W O M behaviour. W O M is seen as effortful consumer behaviour, and is likely to be induced by various goals of the W O M giver. The personal relevance of W O M is seen as a general source of motivation for expending the effort required to give W O M , as well as a potential source of bias in W O M behaviour. Given an opportunity to self-promote, consumers who have positive product attitudes are expected to exhibit a positive bias in giving W O M , leading them to exaggerate the degree of their preference for the product. This chapter has identified three possible variables - product self-relatedness, product ownership, and the situational context of W O M - that may affect the willingness to give W O M and W O M valence. A pilot study is a preliminary investigation of the effect of these three variables on W O M behaviour. Study 1 formally examines the effects of self-related products and ownership on willingness to give W O M , and the valence of W O M . Study 2 investigates how aspects of the social context in which W O M is given, as well as marketing promotion that links a product to the self-concept, may lead to increases in W O M and positive biases W O M . 24 C H A P T E R T H R E E P I L O T S T U D Y 3.1 Overview The main goal of the pilot study was to obtain preliminary support for the notion that products that reflected some aspect of the self-concept and were owned would separately lead to more willingness to give W O M and more positively valenced W O M . It also examined how the W O M giver's immediate social context might enhance these effects. A survey methodology was used to measure subjects' intended W O M on products that reflected their self-concept in some way. For comparison, W O M behaviour on utilitarian products that subjects consumed mainly for practical purposes was also measured. Before reporting the study in greater detail, various hypotheses will be developed. 3.2 Hypotheses to be Tested Predictions were made on both total amount of W O M (an indication of willingness to give WOM), and the valence of W O M . 25 Willingness to Give WOM As discussed in Chapter 2, products that reflect the self in some way are personally relevant to consumers. Given that personal relevance is an important source of motivation, consumers are expected to be more willing to engage in W O M (an effortful activity) on self-related products than on utilitarian products. Also, ownership may increase the personal relevance of a product to consumers. As such, owners may have more motivation to give W O M than non-owners on the same product. HI : The total amount of W O M will be greater on self-related products than on utilitarian products H2: The total amount of W O M will be greater on products that are owned than on products that are not owned Self-relatedness of products and. ownership are therefore regarded as separate sources of personal relevance. No interaction is hypothesized, although this will be examined. Valence of Word-of-Mouth A product that reflects an important aspect of the self-concept may also cause consumers to give more valenced W O M . That is, consumers may tend to polarize positive or negative product attributes when giving W O M . By emphasizing the positive 26 aspects of the product, for example, consumers may be telling others something positive about themselves. Ownership is expected to have a similar effect on W O M valence. H 3 : W O M will be more positively valenced on self-related products than on utilitarian products H 4 : W O M will be more positively valenced on products that are owned than on products that are not owned Self-promotion in terms of more positively valenced W O M as suggested by H 3 and H 4 is expected to be most apparent when the immediate W O M context suggests that subjects will be judged by their W O M . This is because the expectation of social evaluation allows a greater opportunity to self-enhance than contexts in which social evaluation is not salient. H 5 : The effect that self-relatedness and ownership of products have on W O M valence will be greater when social evaluation is expected than when it is not expected 3.3 Method Subjects and Design This study used a 2 (product type: self-related vs. utilitarian) by 2 (ownership: own vs. not own) by 2 (WOM context: judgement vs. no judgement) between-subjects factorial design to examine the effects of these variables on total W O M and W O M valence. A survey methodology was used (Appendix 3-1). One hundred and thirty-nine students in introductory marketing were recruited for a study on product communication for course credit. They were randomly assigned to one of the eight experimental conditions. Procedure Participants completed the study in pairs. The experimenter ensured that there was no communication between subjects, and reminded them of the importance of independent responses. Each subject was first asked to identify a product that matched the experimental condition to which he or she had been assigned. Subjects were asked to either think of a product/service that they had purchased and used (own condition), or never purchased and used (not-own condition). The phrase "purchased and used" rather than "owned" was employed because it better accommodated the discussion of services. Otherwise, it would be confusing to ask subjects to imagine "owning" a service. Those in the self-related condition were further told that the product should tell other people "what kind of a person" they actually were. By contrast, those in the utilitarian condition were told that the product should be used "solely for practical reasons." A l l subjects were asked to be as specific as possible in stating the product/service. 28 Intended W O M was measured by asking subjects to imagine talking about the product/service with another student, and then write down exactly what they would say. Subjects who were assigned to the judgement condition were told that they were talking to another student who was knowledgeable enough about the product to judge their opinions. Those in the no-judgement condition were simply told they were talking to another student. A l l subjects were given sufficient time (3 to 5 minutes) and space to write down what they would say. Subjects were also asked to rate how positive or negative their W O M would be on a scale from -3 (mainly negative) to +3 (mainly positive). Subjects also answered manipulation check items. They rated self-promotion motives on a three-item scale, which included ratings on the extent to which they would try to be "smart," "with-it," and to "look good" when talking about the suggested product/service. A l l items were rated using seven-point scales anchored by not at all (I) and very much so (7). Product type and ownership were manipulated by asking subjects to choose products for each category. This procedure was similar to the idiographic approach that has been successfully used in personality research (Allport 1937; Bern and Allen 1974; Kelly 1955; Mischel 1968). Moreover, this approach fit the idea that the products that are related to the self-concept are likely to be quite different across individuals (Markus and Nurius 1986). As such, the products that a specific individual viewed as relevant to the self-concept were likely to be somewhat idiosyncratic. Using the idiographic 29 approach therefore allowed the subjects to choose products that revealed the aspects of their self-concept that might be important and unique to them. However, by allowing subjects to suggest their own products, the specific products used for each condition were not controlled. Additional measures were included in the present study in order to evaluate the possibility that other aspects of the products that subjects chose might have also varied, in addition to their self-relatedness and ownership. For instance, familiarity with products might allow subjects to give more details during W O M conversations, increasing the total amount of W O M contributed. Familiarity might lead to more liking of the products (Zajonc 1968), hence more positive evaluations. Also, it seemed conceivable that for reasons such as conspicuousness, subjects might be more likely to identify with and give W O M about expensive rather than less expensive products. In that case, total W O M and its valence might be driven by the price of the product. In order to examine these possibilities, subjects rated their general familiarity with the product using a five-point scale ranging from completely unfamiliar (1) to completely familiar (5), and also estimated the price of the product. 3.4 Results 3.4.1 Manipulation Check An index (TOTPROMO) for the self-promotion motive was created for the three self-promotion items (a=.73). A three-way A N O V A on T O T P R O M O showed no significant effects (Tables 3-1-1 and 3-1-2). Hence, the manipulation did not show the expected results. However, this might be because subjects were largely unaware of their own motives when giving W O M (Dichter 1966, p.148). Moreover, subjects might have been unwilling to admit that they tried to self-promote when giving W O M . Therefore, the stronger evidence on a self-promotion motive was expected to show from the written intended W O M . 3.4.2 Coding Written Word-of-Mouth The written W O M was coded to obtain a number of different variables for data analysis. The coder was blind to the experimental conditions. Responses were coded for the total amount of W O M detail as expressed in the number of statements made. Examples of the details that subjects provided included information about products or services (e.g., "It comes in different flavours" for Scope mouthwash), evaluations (e.g., "Guinness is awesome" for beer; "It's very good for us" for Sunlight orange juice), consumption experiences (e.g., "It's a cologne I've purchased recently" for Pleasure cologne for men), or recommendations (e.g., "If you are a business person who needs to call during business hours, another plan will be more suitable" for B C Tel Mobility Cellular Phone Plan). W O M statements were defined as the "smallest units of meaning that had informational or affective value," in accordance with Stafford and Daly (1984, p. 386). The number of words contained in each W O M statement varied across subjects depending on personal style of writing and communication. Hence, multiple idea units might appear in one sentence, or in several related sentences. The number of these 31 statements was summed to derive the total amount of W O M (TOT_D). Among these statements, those that were valenced (i.e., expressing either positive or negative evaluations) were counted separately to derive positive (POS_EVAL) and negative (NEG_EVAL) evaluative statements. The main valence variable (POS-NEG) was then derived by obtaining the difference between them (i.e., POS-NEG = POS_EVAL -NEG_EVAL). Neutral statements such as those providing facts (e.g., "It comes in different flavours") were not counted in these valence variables. Other variables could have been derived from the written intended W O M . For example, besides counting the number of details given by subjects, the frequency of various product attributes mentioned could have also been noted. It is conceivable that some attributes of products were more likely to be mentioned and evaluated than others. Such an account of product attributes might be of interest to marketers who need to decide on the dimensions of a product to emphasize in their marketing campaign. Also, under the current coding scheme, any detail that was repeated by the same subject was treated as redundant, and not counted twice. However, such repetitions could have been intended to indicate an emphasis on the particular detail given during W O M . While interesting, these additional questions go beyond the scope of the current dissertation, which was mainly aimed at better understanding the impact of product self-relatedness, product ownership, and the social context of W O M on the likelihood of W O M and the valence of W O M . 32 3.4.3 Total Word-of-Mouth It was predicted that product type and ownership would both increase T O T _ D separately (HI and H2). A three-way A N O V A on TOT_D showed a product type main effect (F(l,131)=3.92, p=.05) (Tables 3-2-1 and 3-2-2). More W O M was given on self-related than utilitarian products (M seif-reiated =7.36 vs. M utilitarian =6.26). The effect of ownership was not significant. HI was therefore supported, while H2 was not. To rule out alternative explanations that price and familiarity might have driven the above results, a three-way A N C O V A was further conducted on TOT_D with these two variables as covariates. The analysis revealed that both price and familiarity were significant covariates (p's=.053 and .001 respectively), but the product type main effect remained significant (F(l,125)=3.76, p=.055) (Table 3-2-3). 3.4.4 Valence of Word-of-Mouth POS-NEG It was predicted that POS-NEG would be higher for self-related products than utilitarian products (H3), and also higher for owned products than for products that were not owned (H4). Moreover, the effect of product type and ownership on W O M valence was expected to be greater when the W O M context suggested social evaluation on the opinion given (H5). A three-way A N O V A on POS-NEG showed no product type main effect, but an ownership main effect (F(l,131)=4.02, p<05) (Table 3-3-1). Subjects who owned the 33 product gave greater POS-NEG than those who did not (M 0 W n =3.04 vs. M notown =2.01). Thus, H3 was not confirmed and H4 was confirmed. The ownership main effect was qualified by a two-way product type by ownership interaction (F(l,131)=4.70, p<.05). A contrast comparison on this two-way interaction showed that P O S - N E G was significantly lower when products were self-related but not owned, and it was highest when products were self-related and owned (Fy(l,131)=6.85, p=.01) (Table 3-3-2). There was also a marginally significant three-way interaction (F(l,131)=3.17, p=.078). Individual cell means for POS-NEG suggested that the marginal three-way interaction was mainly driven by a two-way interaction between product type and ownership for the judgement group, and a lack of significance from these two independent variables for the no-judgement group (Table 3-3-3). Indeed, a contrast comparison within the judgement group showed that P O S - N E G was higher i f the products were self-related and owned, or surprisingly, i f they were utilitarian and not owned (Fy(l,131)=7.74, p<.01). H3 was therefore partially confirmed in that self-relatedness was associated with more positively valenced W O M , though that was only true when the product was owned. There was a surprising reversal in the findings when the product was not owned. For subjects in the no-judgement group, there was no significant difference across the four experimental conditions. This confirmed H5, which predicted that self-promotion motive would be higher when social evaluation was expected. 34 POS_EVAL The findings for P O S _ E V A L generally mirrored those of the positive valence measure (POS-NEG). A three-way A N O V A on P O S _ E V A L showed a similar pattern as in POS-NEG. There was a significant three-way interaction (F(l,130)=5.25, p<.05) (Tables 3-4-1 and Tables 3-4-2). This seemed to be due to a product type by ownership interaction in the judgement group. A contrast comparison revealed similar evidence as POS-NEG (FM<1,130)=8.98, p<.005). There was no significant effect of product type and ownership in the no-judgement condition. NEG_EVAL A three-way A N O V A on N E G _ E V A L showed no significant effects (Tables 3-5-1 and 3-5-2). Therefore, product type, ownership, and W O M context did not have an influence on subjects' negative evaluations of the products. Ratings for Word-of-Mouth Valence In addition to writing down their intended W O M , subjects were directly asked to rate how positive or negative their intended W O M would be. A three-way A N O V A conducted on these ratings ( V A L E N C E ) revealed a two-way product type by ownership interaction (F(l,131)=4.30, p<.05) (Tables 3-6-1 and 3-6-2). These findings again 35 partially supported the effect of self-relatedness on W O M valence as predicted by H3. There was however no significant W O M context effect in this case. Ruling out Alternative Explanations for POS-NEG It could be argued that the effects observed from the overall valence variable (POS-NEG) were due to the fact that subjects were simply talking more (TOT_D) about self-related and owned products versus utilitarian products, or that these products might be more familiar to them or more expensive. An A N C O V A was therefore conducted on POS-NEG with TOT_D, price, and familiarity as covariates in order to account for these possibilities. T O T J D and familiarity were found to be significant covariates (p's = .001 and .058 respectively), and the three-way interaction dropped in significance (p<.15). However, a two-way product type by ownership interaction effect remained significant (F(l,124)=3.87, p=.05) (Table 3-7). Overall, these results showed that total W O M and subjects' familiarity with the products had some effect on POS-NEG. The impact of product type and ownership on W O M valence remained significant after accounting for them, but the effect of the W O M context was reduced. 3.5 Discussion 3.5.1 Main Findings and Implications This pilot study supported the prediction that more W O M (TOT_D) would be given on self-related rather than utilitarian products. Self-related products seemed to have induced 36 a higher level of involvement with products, leading to relatively high willingness to give W O M and more detail being described. Moreover, it was found that the effect of product type on TOT_D remained significant even after accounting for the effect of prices and subjects' familiarity with the products. More importantly, this study provided partial support for the more general prediction that the personal relevance of a product would lead to more positively valenced W O M (POS-NEG). When they expected to be judged on the W O M given, subjects seemed to self-promote by exaggerating the benefits and giving more positive evaluations of products they thought might reflect some aspect of their self-concept, compared to products that were primarily utilitarian. Of particular interest was the finding that this positivity of W O M was greater when the subjects also owned the product. Surprisingly, self-relevance led to the opposite effect on W O M valence when the product was not owned. Self-related products actually received less positive W O M than utilitarian products that were not owned. Although unexpected, it seems possible to at least make some sense of these findings. It is conceivable that W O M givers may be reluctant to recommend products they see as self-related but do not own, because they do not want others to purchase it. This may be especially true i f there is a close personal relationship between the W O M giver and the recipient. If the recipient bought the product, this might reflect negatively on the giver because of the negative social comparison involved. In order to avoid this situation, the W O M giver may tend to talk less positively about the product. 37 The self-evaluation maintenance model (SEM) (Tesser 1988) supports this explanation. S E M predicts that i f an outcome is relevant to one's self-definition, then good performance by another individual will harm one's self-evaluation through negative social comparison. Furthermore, this negative comparison will have a greater psychological impact the closer the relationship between the two people, and the more relevant the outcome is for self-evaluation. One way to circumvent this negative impact is to undermine the other person's ability to achieve a positive outcome (Tesser and Campbell 1982; Tesser and Smith 1980). In this case, it can be achieved by discouraging a friend from buying a good product that the W O M giver does not own. Hence, subjects in this study might be less inclined to encourage the W O M recipient to buy the self-related product because they did not own it themselves. The results seemed to rule out alternative explanations for the findings on W O M valence in terms of total W O M , or familiarity and price. It was shown that the effect of product type and ownership on W O M valence remained significant even after accounting for these factors as covariates. Another plausible explanation that was not directly examined in this study was that subjects might simply like owned and self-related products more than other products. Such attitudes might lead to a positive focus in W O M in that particular experimental condition. This and other methodological concerns will be discussed next. 38 3.5.2 Methodological Concerns and Implications One oversight of the pilot study was that subjects' attitudes towards the products were not measured. Since subjects in the study were asked to suggest their own products and brands, it could be that they suggested products that were better liked in the self-related condition than those in other conditions. This makes intuitive sense because consumers may be more likely to consider good products as reflective of the self-concept, and more likely to buy products they like than those that they do not like. This might then result in more positive W O M about products that are both self-related and owned. Therefore, it remained to be shown that self-promotion goals contributed to more positive W O M beyond the level that product attitudes alone could explain. The survey research format of the pilot study was considered a useful and efficient way of examining W O M behaviour. Previous research in W O M has commonly adopted this format (Anderson 1998; Blodgett et al. 1993; Mooradian and Olver 1997; Richins 1983; Swan and Oliver 1989; TARP 1979). However, a limitation of this format is that it essentially measures intended W O M behaviour rather than actual W O M behaviour, which may differ. An alternative would have been to ask subjects for reports on actual W O M that occurred in the past. However, self-reports of past W O M involve potential distortions due to memory limitations (Kintsch and Bates 1977; Neisser 1982; Stafford and Daly 1984). For example, Stafford and Daly (1984) found that subjects on average only recalled about 10% of their conversations. More specifically, subjects recalled only around 15% of sentences made for evaluation purposes. Moreover, self-reports on past behaviour in general have been found to be accurate on some but not other occasions (e.g., Jaccard, King, and Pomazal 1977, Study 3; Pomazal and Jaccard 1976 for accuracy; Hyman 1945; Sample and Warland 1973; Tittle and Hi l l 1967 for inaccuracy). The next study focused more on developing a research design that would capture W O M behaviour as it occurred. It was also possible that experimental demand existed in the pilot study. The survey essentially told subjects the specific characteristics of the products (self-related or utilitarian; own or not own) that the experimenter was interested in investigating. Thus, the two main variables were manipulated in a manner that was obvious to the subjects. This might have made it relatively easy for subjects to generate their own hypotheses about what effects the independent variables were expected to have on W O M . In Studies 1 and 2, the main independent variables were more subtly manipulated to eliminate possible demand characteristics. 40 Table 3-1-1 Cell Means for T O T P R O M O (Pilot Study) Judgement No Judgement Own Not Own Own Not Own Self-Related 3.67 4.73 4.35 3.90 (17,1.25)* (17, 1.25) (17, 1.66) (17, 1.30) Utilitarian 3.88 4.02 4.06 4.31 (17, 1.34) (17, 1.53) (17, 1.58) (18, 1.26) (cell size, standard deviation Table 3-1-2 Analysis of Variance on T O T P R O M O (Pilot Study) Source Sum of Degrees of Squares Freedom F-Statistic p -value CONTEXT .24 1 .12 .728 PROD TYPE .30 1 .1.5 .699 OWN 2.15 1 1.09 .299 CONTEXT * PROD TYPE .79 1 .40 .527 C O N T E X T * OWN 4.14 1 2.10 .150 PROD TYPE * OWN .10 1 .05 .823 CONTEXT * PROD T Y P E * OWN 5.68 1 2.88 .092 > Error 254.13 129 Total 267.47 136 41 Table 3-2-1 Cell Means for T O T D (Pilot Study) Judgement No Judgement Own Not Own Self-Related 8.06 6.65 (17,2.59)* (17, 3.20) Utilitarian 5.41 5.56 (17,2.76) (18,2.96) Own Not Own 7.47 (17, 4.20) 7.28 (18,4.18) 6.29 (17, 2.76) 7.72 (18,3.48) (cell size, standard deviation) Table 3-2-2 Analysis of Variance on TOT_D (Pilot Study) Source Sum of Squares Degrees of Freedom F-Statistics p-value CONTEXT 20.74 1 1.88 .173 P R O D T Y P E 43.37 1 3.92 .050 OWN .00 1 .00 .988 CONTEXT * PROD TYPE 19.62 1 1.77 .185 CONTEXT * OWN 13.60 1 1.23 .269 PROD TYPE * OWN 21.90 1 1.98 .162 CONTEXT * PROD TYPE * O W N .01 1 .00 .977 Error 1448.37 131 Total 1567.76 138 42 Table 3-2-3 Analysis of Covariance on T O T D (Pilot Study) Source Sum of Degrees of F-Statistics p-value Squares Freedom P R I C E 37.73 1 3.83 .053 F A M I L I A R 120.80 1 12.25 .001 CONTEXT 12.53 1 1.27 .262 P R O D T Y P E 37.09 1 3.76 .055 OWN 7.00 1 .71 .401 CONTEXT * PROD T Y P E 4.47 1 .45 .502 CONTEXT * OWN 1.35 1 .14 .712 PROD TYPE * OWN 18.69 1 1.90 .171 CONTEXT * PROD T Y P E * OWN 3.60 1 .37 .547 Error 1232.32 125 Total 1492.37 134 43 Table 3-3-1 Analysis of Variance on P O S - N E G (Pilot Study) Source Sum of Squares Degrees of Freedom F-Statistics p-value CONTEXT 4.80 1 .52 .472 PROD TYPE .03 1 .00 .956 O W N 37.09 1 4.02 .047 CONTEXT * PROD T Y P E .00 1 .00 .991 CONTEXT * OWN .40 1 .04 .836 P R O D T Y P E * O W N 43.44 1 4.70 .032 C O N T E X T * P R O D T Y P E * O W N 29.23 1 3.17 .078 Error 1448.37 131 Total 1567.76 138 Table 3-3-2 Cell Means (Product Type by Ownership) for POS-NEG (Pilot Study) Own Not Own Self- 3.62a 1.49c Related (34, 2.50)* (35, 3.67) Utilitarian 2.47b 2.56b (34, 3.10) (36, 2.75) * (cell size, standard deviation) ' b ' c A contrast analysis was conducted in which cell means with subscripts a, b, and c were assigned contrast coefficients +2, +1 and-4 respectively. Cell means with different subscripts were significantly different from each other. 4 4 Table 3-3-3 Cell Means (Product Type by Ownership by Context) for POS-NEG (Pilot Study) Judgement No Judgement Own Not Own Self-Related 3 . 9 4 A . 7 6 B (17, 2.59)* (17, 3.20) Utilitarian 1 . 8 8 B 2 . 7 8 A (17, 2.76) (18, 2.96) Own Not Own 3.29 (17,2.05) 2.17 (18, 4.34) 3.06 (17,2.41) 2.33 (18,2.61) (cell size, standard deviation) a' b A contrast analysis was conducted in which cell means with subscript a and b were assigned contrast coefficients +1 and -1 respectively. Cell means with different subscripts were significantly different from each other. 45 Table 3-4-1 Cell Means for POS_EVAL (Pilot Study) Judgement No Judgement Own Not Own Self-Related 4 . 8 2 A 2 . 0 6 B (17, 2.56)* (16,1.34) Utilitarian 2 . 8 2 B 3 . 2 8 A (17, 2.35) (18, 2.14) Own Not Own 3.71 (17, 1.86) 3.33 (18,3.22) 3.53 (17,2.07) 2.89 (18,1.71) (cell size, standard deviation) ' ' b A contrast analysis was conducted in which cell means with subscript a and b were assigned contrast coefficients +1 and -1 respectively. Cell means with different subscripts were significantly different from each other. Table 3-4-2 Analysis of Variance on POS_EVAL (Pilot Study) Source Sum of Squares Degrees of Freedom F-Statistic p-value CONTEXT .48 1 .10 .758 PROD TYPE 4.25 1 .86 .357 OWN 23.73 1 4.77 .031 CONTEXT * PROD TYPE .06 1 .01 .914 CONTEXT * OWN 3.6 1 .72 .396 PROD TYPE * OWN 18.70 1 3.76 .055 CONTEXT * PROD TYPE * OWN 26.12 1 5.25 .024 Error 647.03 130 Total 721.60 137 Table 3-5-1 Cell Means for N E G _ E V A L (Pilot Study) Judgement Own Not Own Self-Related .88 1.18 (17, 1.17)* (17, 1.91) Utilitarian .94 .50 (17,2.16) (18, .1,54) No Judgement Own Not Own .41 (17, .71) 1.17 (18,2.28) .47 (17, .72) .56 (18, 1.34) (cell size, standard deviation) Table 3-5-2 Analysis of Variance on N E G E V A L (Pilot Study) Source Sum of Degrees of F-Statistic p-value Squares Freedom CONTEXT 1.74 1 .69 PROD TYPE 2.97 1 1.18 OWN 1.04 1 .41 CONTEXT * PROD TYPE .01 1 .00 CONTEXT * OWN 2.11 1 .84 PROD TYPE * OWN 4.29 1 1.70 CONTEXT * PROD T Y P E * OWN .01 1 .00 Error 330.97 131 Total 343.17 138 .408 .280 .522 .952 .362 .195 .952 47 Table 3-6-1 Cell Means for V A L E N C E (Pilot Study) Own Not Own 2.03 .94 Self-Related (34, 1.06)* (35, 1.63) 1.56 1.47 Utilitarian (34, 1.40) (36, 1.50) (cell size, standard deviation) Table 3-6-2 Analysis of Variance on V A L E N C E (Pilot Study) Source Sum of Squares Degrees of Freedom F-Statistic p-value CONTEXT 1.21 1 .59 .444 PROD TYPE .04 1 .02 .893 O W N 12.08 1 5.91 .016 CONTEXT * PROD TYPE .61 1 .30 .586 CONTEXT * OWN .56 1 .28 .600 P R O D T Y P E * O W N 8.80 1 4.30 .040 CONTEXT * PROD TYPE * OWN .08 1 .00 .952 Error 267.81 131 Total 290.75 138 Table 3-7 Analysis of Covariance on POS-NEG (Pilot Study) Source Sum of Degrees of F-Statistic p-value Squares Freedom T O T J D 102.08 1 12.12 .001 PRICE 1.61 1 .19 .662 F A M I L I A R 30.72 1 3.65 .058 CONTEXT 1.42 1 .17 .682 PROD TYPE 3.67 1 .44 .510 O W N 45.55 1 5.41 .022 CONTEXT * PROD TYPE .14 1 .02 .899 CONTEXT * OWN .50 1 .06 .809 P R O D T Y P E * O W N 32.56 1 3.87 .051 CONTEXT * PROD TYPE * OWN 20.26 1 2.41 .123 Error 1044.28 124 Total 1267.08 134 49 C H A P T E R F O U R S T U D Y O N E 4.1 Hypotheses to be Tested This study was conducted to further examine the impact of self-relatedness and ownership of products on W O M behaviour, using a new methodology. In particular, a method was devised to measure W O M behaviour as it occurred in a relatively naturalistic setting. The effect of self-related products and product ownership on willingness to give W O M (HI and H2), and W O M valence (H3 and H4) was again tested. Product attitude was also assessed in order to understand what role it might play in W O M behaviour. In general, results of product attitude on W O M were expected to replicate past findings that attitudes predicted W O M . In addition to attitudes, self-relatedness and ownership were expected to significantly influence total W O M and W O M valence. 4.2 Overview Data collection occurred in three stages due to the nature of the experimental design. Two months before the W O M experimental session, an initial survey was conducted to obtain baseline data identifying products that each subject considered to be self-related versus utilitarian, and which he or she owned or did not own. This information was later used to randomly assign subjects to self-relatedness and ownership conditions in the experimental session. The use of an initial survey to identify products that were self-50 related and owned in a separate context was intended to reduce any experimental demand that might have existed in the pilot study. Since the survey was completed well in advance, this should have eliminated the possibility that subjects might relate the survey to their responses in the experimental session. The experimental session conducted later consisted of two parts. First, using information obtained from the survey, a confederate probed subjects for W O M concerning the target product. Second, a one-page survey was administered, containing mainly manipulation check items. 4.3 Data Collection in Three Stages 4.3.1 Initial Survey One hundred and eighteen students completed an initial survey in a classroom setting. They were asked to give examples of products (and brands) for all combinations of self-relatedness and ownership (Appendix 4-1). This information would be used later as the topic of W O M conversation in the first part of the experimental session. One change from the pilot study was that an instruction that asked specifically for products for which subjects held "reasonable to positive" attitudes was added to the general instructions included in the survey. This restriction was imposed to avoid examples of negative products that might have confounded the results in the pilot study. Subjects were also told that they should choose a product about which they felt knowledgeable enough to give recommendations i f somebody asked them for advice. This initial survey was also used to assess attitudes towards the products that subjects mentioned. Subjects rated their 51 product attitudes on a seven-point scale anchored by negative (-3) and positive (+3), and centred by mixed/neutral (0). 4.3.2 Experiment Part I Methods This experiment used a 2 (product type: self-related vs. utilitarian) by 2 (ownership: own vs. not own) between-subjects factorial design. Students from the same subject pool as that used in the initial survey were asked to sign up for an Internet shopping study for course credit. This was an actual study that was conducted between the two parts of the W O M experimental session. Eighty-seven subjects participated in the experiment. They were randomly assigned to one of the four experimental conditions. This determined which of the products that they mentioned in the initial survey would be used for the W O M topic during the experimental session. The experimenter reviewed the products suggested by subjects in the initial survey to determine whether they could be used in this subsequent experimental session. Some items were eliminated because they were too idiosyncratic or peculiar (e.g., pocket protector), or might cause embarrassment for and bias results from discussion (e.g., condom or underwear). Subjects who were assigned to a condition where the product had been eliminated were given a "modal" product for that condition. Modal products were those that were most frequently mentioned for that condition by the other subjects in the initial survey. Some subjects who did not indicate a product for the particular 52 condition to which they were assigned in the initial survey were also given a modal product for discussion. In total, seven subjects were assigned to discuss running shoes (self-related/own), two to discuss cars (self-related/not own), four to discuss toothpaste or toothbrushes (utilitarian/own), and four to discuss laptop computers (utilitarian/not own). Self-relatedness and ownership status of the products assigned to these subjects were later assessed in the final questionnaire, and it was verified that the substituted products matched the conditions to which the subjects had been assigned. Procedure Subjects (Ss) arrived individually at the experiment expecting to participate in an Internet study. There, they met with another subject, who was actually a confederate (C) hired to probe for W O M opinions (as will be explained below). Twelve students (3 males and 9 females) were hired to serve as Cs for this study. They were blind to the experimental conditions and hypotheses. A l l Cs were introduced to Ss as exchange students. Cs were actually mostly recent immigrants or exchange students whose mother tongue was not English. The exchange student story was used in order to convey the fact that C was new to this country, and therefore might be unfamiliar with even well known products in North America. The experimenter (E) showed S and C into a waiting room, and explained that she was still setting up the Internet shopping study. S and C were also told that the second study involved asking them for some general opinions about products. Before leaving the 53 room, E mentioned the particular product for which they would be asked their opinion, and suggested that they think about what they had to say. Although subjects had stated both the product and the brand in the initial survey, some brands might have been too specific to be used for discussion without causing suspicion. Therefore, E only suggested the general product, while C probed for the specific brand. E also told S and C that audiotaping might be needed for the study on general opinions about products, and obtained consent from both of them. After E left the room, C, who was trained to probe S for W O M , first tried to establish some eye contact with S, and waited for approximately 10 seconds before talking. This procedure was adopted from rumour studies (e.g., Walker and Beckerle 1987). C first tried to probe for S's ownership status of the product. Once subjects stated their ownership status, C was no longer blind to the ownership condition. However, C was still blind to the purpose of the experiment, and none of the Cs were aware of the hypotheses. In most cases, C asked S, "Do you own (the product)?" For products that everyone was likely to own (such as toothpaste, toothbrush, or clothes), C instead asked, "You probably own (the product), don't you?" After confirming the ownership status, C told S that he or she was an exchange student and did not know much about the product, and asked whether S could at least provide "some idea about what to say during the study." C also probed for the specific brand that S mentioned two months back in the initial survey, by asking, "Is there a specific brand that you're familiar with?" After S had named the target brand, C asked 54 how S liked that product. If S mentioned any brands other than the target one, C would further ask i f S had heard of the target brand in order to start discussion on it. Most of the conversations between C and S were approximately five to seven minutes long. These conversations were unobtrusively audiotaped, and were later coded to derive measures of total W O M (TOTD), and W O M valence (POS_EVAL, NEG_EVAL, and POS-NEG). E waited outside the room for the duration of the conversation. Approximately 10 seconds after all the main measures had been discussed, she re-entered the room to announce that she had finished the set-up for the Internet study. After the Internet study, S was asked to complete a final questionnaire, which mainly contained a set of manipulation check items. 4.3.3 Experiment Part II This survey asked Ss for their opinions on the specific products that matched the condition to which they had been assigned (Appendix 4-2). In addition, Ss were told to provide the brand for which they felt "comfortable giving someone else advice." These were largely questions asked for the purpose of making the cover story complete, as the main variables had already been measured in the first part of the experiment. Self-promotional motives were measured the same way as in the pilot study. Subjects' familiarity with the products was measured on a seven-point scale (-3=«or at all to +3=very much). In addition, questions on self-relatedness (yes/no) and ownership (yes/no) were included to check for any changes in status from the initial survey. As 55 well, product attitudes were again measured using the same seven-point scale in the initial survey. 4.4 Results 4.4.1 Data Cleaning and Coding Of the 87 participants, 10 cases were discarded. The discarded cases involved suspicion (four), refusal to participate (two), and technical errors in taping (four). Therefore, data from 77 participants remained. Further, when data were matched over the three stages of data collection, it was discovered that some subjects who had been assigned to products according to their responses in the initial survey had changed their responses to the self-relatedness and ownership questions in the final questionnaire. These changes were from utilitarian to self-related (nine subjects or 25% of those originally assigned to talk about utilitarian products), from self-related to utilitarian (5 subjects or 12%), from not-own to own (eight subjects or 22%), and from own to not-own (1 subject or 2.5%). These subjects were therefore re-classified according to their responses on the final questionnaire. However, the majority of the subjects remained in the same self-related (80%) and ownership categories (87%) in both the initial and final survey, indicating that these responses were fairly reliable over time. The final cell sizes were: self-related/own=29, self-related/not own=16, utilitarian/own=18, and utilitarian/not own=14. 56 The taped conversations between S and C were transcribed, and S's responses independently coded by the author who was blind to the experimental conditions, and a coder who was blind to both the hypotheses and experimental conditions. The coding scheme was the same as the one used in the pilot study. Agreement between the author and the coder was at 89% for TOT_D. Disagreement mainly occurred in the recognition of the unit of analysis. For example, the coder counted as detail a statement that was unrelated to the target brand that was under discussion whereas the author did not treat such a statement as contributing toward total W O M in relation to the target product and brand. Agreements on P O S _ E V A L and N E G E V A L were 87% and 82% respectively. Disagreement was somewhat related to that on TOT_D. To continue with the above example, counting the unrelated statement toward TOT_D, the coder would then proceed to coding the valence of it whereas the author would not do that. Analyses were conducted using the author's set of scores. 4.4.2 Manipulation Check As in the pilot study, an index for self-promotion motivation ( T O T P R O M O ) was created (a=.92). A two (product type: self-related vs. utilitarian) by two (ownership: own vs. not own) A N O V A on T O T P R O M O revealed a product type main effect (F(l,73)=5.18, p<.05), and a marginal ownership effect (p=.075) (Tables 4-1-1 and 4-1-2). Subjects indicated greater concern about self-promotion when talking about self-related rather than utilitarian products (M Seif-reiated = 4.55 vs. M utilitarian = 3.72). 57 Surprisingly, this analysis also showed that subjects who did not own the product were more concerned about self-promotion motivation than those who owned the product (M own = 3.80 vs. M not own = 4.46). No other effects were significant. 4.4.3 Total Word-of-Mouth A two-way A N O V A was conducted on TOT_D, with product type and ownership as factors. As in the pilot study, there was a product type main effect (F(l , 73) =5.00, p<.05) (Tables 4-2-1 and 4-2-2). More W O M was given about products that reflected subjects' self-concept than products that were utilitarian (M seif-reiated = 12.66 vs. M utilitarian = 7.78). To rule out the alternative explanation that product familiarity might have driven this main effect, a two-way A N C O V A was conducted on TOT_D with familiarity as a covariate. The product type main effect remained significant (F(l,72)=3.70, p=.059), although familiarity was a significant covariate (p=.001) (Table 4-2-3). Therefore, HI was again supported. However, H2 was again disconfirmed. There was no significant effect of ownership on total W O M . It seemed that owning a product did not lead subjects to talk more about it. 58 4.4.4 Valence of Word-of-Mouth POS-NEG A 2 (self-related vs. utilitarian) by 2 (own vs. not own) A N O V A on this dependent variable revealed no significant effects (Tables 4-3-1 and 4-3-2). However, individual cell means seemed to reveal a pattern of findings that was at least similar to what was expected. Further investigations were therefore made on the secondary variables for W O M valence, namely POS_EVAL and NEG_EVAL. POS_EVAL A two-way A N O V A revealed a product type main effect for POS_EVAL (F(l,73)=9.07, p<.005) (Tables 4-4-1 and 4-4-2). More positive evaluations were given on self-related products versus utilitarian products (M seif-reiated =3.20 vs. M utilitarian =1 -54). However, ownership did not interact with the effect of product type on W O M valence in the present study. There was no ownership main effect, and H4 was not supported. NEG_EVAL A two-way A N O V A on NEG_EVAL showed a product type main effect (F(l,73)=4.80, p<.05) (See Tables 4-5-1 and 4-5-2). Subjects gave more negative evaluations on self-related rather than utilitarian products (M seif-reiated = 1.31 vs. M utilitarian 59 = .43). No other effects were significant. The fact that self-related products also received more negative W O M explained why the product type manipulation was not significant when looking at the overall valence variable (POS-NEG), disconfirming H3. Alternative Explanations for POSJEVAL and NEG_EVAL It could be argued that subjects gave more positive evaluations on self-related rather than utilitarian products because they simply said more (higher TOT_D) when products were relevant to their self-concept, or they were more familiar with these products. It also seemed conceivable that they would have more positive attitudes towards self-related products, which might account for the results concerning P O S _ E V A L 1 . An A N C O V A conducted on P O S _ E V A L with TOT_D, product familiarity, and product attitude as covariates showed that the product type main effect on P O S _ E V A L remained marginally significant (F(l,70)=2.94, p<.10). T O T D was a significant covariate (F(l,70)=52.61, p<.001), but product familiarity and attitudes were not (F's <1) (Table 4-6). Therefore, products that reflected subjects' self-concept seemed ' Product attitudes were measured both in the initial survey and the final questionnaire after W O M conversations had taken place. To meaningfully utilize the data, the products and brands on which attitudes were measured had to be the same as those referred to in the W O M conversations. To ensure this correspondence, the following arrangement was made. If the products and brands that subjects provided in the initial survey matched those discussed in the conversations, then attitudes in the initial survey would be used for analysis. If, however, products and brands matched only between the conversation and the response in the final questionnaire, then subjects' ratings in the final questionnaire would be used. 60 to induce marginally more positive W O M than utilitarian products, even when accounting for total W O M . However, an A N C O V A on N E G E V A L with T O T D, product familiarity and product attitude as covariates revealed that the product type main effect also remained marginally significant (F(l,70)=3.47, p<.10), and TOT_D was a significant covariate (F(l,70)=5.51, p<.05) (Table 4-7). 4.5 Discussion 4.5.1 Main Findings and Implications Study 1 confirmed the prediction that subjects gave more W O M (regardless of valence) on self-related products versus utilitarian ones. The effect of self-relatedness on total W O M remained significant even after accounting for subjects' familiarity with the products. This largely replicated the results on total W O M from the pilot study. The major finding from this study was that there was no significant product self-relatedness effect on overall W O M valence (POS-NEG). To shed light on why that might be, separate analyses were conducted for the number of positive evaluations (POS_EVAL) and the number of negative evaluations ( N E G _ E V A L ) . These analyses showed that subjects gave more positive, as well as more negative evaluations for self-related versus utilitarian products. Overall it seemed that increases in both positive and negative statements about the products cancelled each other out in this study. This also 61 seems to imply that subjects were motivated to be relatively objective in giving W O M , communicating both the positive and negative aspects of their product experiences. It seemed also surprising that product attitudes (which were largely positive as measured) were uncorrelated with positive evaluations, given that the existing research has generally shown a positive relationship between these variables. The lack of correlation in this case was likely due to the fact that the initial survey specifically restricted subjects to suggesting products towards which they had mainly positive attitudes. This instruction was useful in ensuring that all W O M concerned products that subjects actually liked. However, it was also likely that this instruction restricted the range of attitudes, and thereby attenuated their correlations with W O M . The effect of attitudes on W O M valence will be formally examined in the next study. There were no ownership effects shown in any analyses in Study 1, meaning that the findings concerning ownership on W O M valence from the pilot study were not replicated here. However, product type showed greater consistency in its effects on total W O M in the pilot study and Study 1. Also, it had some effect on W O M valence in terms of both positive and negative evaluations in Study 1. Study 2 therefore focused more on better establishing causality between product self-relatedness and W O M behaviour. 62 4.5.2 Methodological Implications and Concerns The methodology used in this study improved upon those of previous W O M studies and the pilot study. The use of a confederate to probe for W O M allowed some control over the topic of conversation, yet gave subjects a great deal of freedom to decide on the specific content (Thomas 1992). This context for giving W O M was likely to be fairly involving, and the W O M behaviour observed was relatively more realistic. Prior studies that used interviews (e.g., Dichter 1966; Engel et al. 1969) or case studies (e.g., Brown and Reingen 1987; Reingen and Kernan 1986) might have induced greater experimental demand because the interviewees were usually asked more direct questions concerning their W O M behaviour. In addition, past methods might have included errors in the report of intended W O M or recollections of W O M that occurred in the past. On the other hand, the present experimental context might have also caused subjects to be objective rather than self-promoting in expressing product opinions. That is, subjects gave more positive as well as negative evaluations when talking about self-related products versus utilitarian products. The fact that the cover story suggested Cs needed the information to answer questions in a study that would be conducted next might have inadvertently led Ss to be more objective in giving W O M . Ss might have felt some responsibility to be balanced or some accountability in their comments considering what the information would be used for. The next study included a more direct manipulation of the self-promotion motive that might lead subjects to give more positive 63 (and fewer negative) evaluations, given that they already had positive attitudes toward the target product. Study 1 did not fully establish causality between self-relatedness and W O M behaviour. This was due to a certain lack of experimental control over the specific products used for the topic of W O M . As in the pilot study, subjects were asked to suggest products for the experimental conditions. Although the new methodology used in this study separated the process of obtaining examples of products from the actual W O M conversations, and might therefore have reduced experimental demand, this procedure did not control for other aspects of products that might have varied. The most important dimensions such as product attitudes, price, and product familiarity were examined, and there was little evidence to support these as viable alternative explanations for the effects of self-relatedness on W O M behaviour. However, there may still be other differences between products that might be confounded with the degree of self-relevance. An alternative to this idiographic approach would have been to identify a modal product (the product that was most frequently mentioned for each of the conditions included in the present study) that fit each category, and assigned subjects to talking about it. However, even with this arrangement, the products chosen could still have differed on other dimensions across conditions. Another concern with this modal product approach was that some subjects assigned to a particular condition might not perceive the product as intended in terms of self-relatedness and ownership. For instance, running shoes might be chosen for the self-related/own condition because they were most 64 frequently mentioned, but some subjects assigned to this condition might not own running shoes, or they might not consider running shoes as relevant to their self-concept. As a result, the use of individual subject-generated products in the pilot study and Study 1 seemed necessary. Study 2 devised a research design that directly manipulated self-relatedness, while controlling for other aspects of the product, and hence better establishing causality. Table 4-1-1 Cell Means for T O T P R O M O (Study 1) Own Not Own Self-Related 4.20 (29, 1.25)* 4.90 (16, 1.49) Utilitarian 3.41 (18, 1.81) 4.02 (14, 1.76) * (cell size, standard deviation) Table 4-1-2 Analysis of Variance on T O T P R O M O (Study 1) Source Sum of Degrees of F-Statistics p-value Squares Freedom P R O D T Y P E 12.30 1 5.18 .026 OWN 7.74 1 3.26 .075 PROD T Y P E * O W N .03 1 .01 .909 Error 173.39 73 Total 192.45 76 Table 4-2-1 Cell Means for TOT_D (Study 1) Own Not Own 12.38 12.94 Self-Related (29, 11.89)* (16, 10.46) 7.78 7.79 Utilitarian (18,3.93) (14, 5.13) (cell size, standard deviation) Table 4-2-2 Analysis of Variance on TOT_D (Study 1) Source Sum of Squares Degrees of Freedom F-Statistics p-value P R O D T Y P E 424.74 1 5.00 .028 O W N 1.43 1 .02 .897 PROD TYPE * O W N 1.35 1 .02 .900 Error 6205.23 73 Total 6638.70 76 Table 4-2-3 Analysis of Covariance on T O T D (Study 1) Source Sum of Degrees of F-Statistics p-value Squares Freedom F A M I L I A R 876.68 1 11.85 .001 P R O D T Y P E 273.50 1 3.70 .059 O W N 69.28 1 .94 .337 PROD TYPE * O W N 15.85 1 .21 .645 Error 5328.56 72 Total 6638.70 76 Table 4-3-1 Cell Means for POS-NEG (Study 1) Own Not Own Self-Related 2.03 (29, 3.21)* 1.75 (16, 2.70) Utilitarian 1.00 (18,2.79) 1.21 (14,1.81) * (cell size, standard deviation) Table 4-3-2 Analysis of Variance on POS-NEG (Study 1) Source Sum of Degrees of F-Statistics p-value Squares Freedom PROD TYPE 11.01 1 1.40 .240 O W N .02 1 .00 .958 PROD T Y P E * O W N 1.11 1 .14 .708 Error 572.32 73 Total 586.70 76 Table 4-4-1 Cell Means for P O S _ E V A L (Study 1) Own Not Own 3.28 3.12 Self-Related (29, 2.59)* (16,2.63) 1.72 1.36 Utilitarian (18,2.02) (14,1.65) (cell size, standard deviation) Table 4-4-2 Analysis of Variance on P O S _ E V A L (Study 1) Source Sum of Degrees of F-Statistics p-value Squares Freedom P R O D T Y P E 49.26 1 9.07 .004 OWN 1.19 1 .22 .641 PROD TYPE * O W N .21 1 .04 .847 Error 396.37 73 Total 449.17 76 Table 4-5-1 Cell Means for N E G E V A L (Study 1) Own Not Own 1.24 1.38 Self-Related (29, 1.77)* (16,2.33) .72 .14 Utilitarian (18, 1.49) (14, .36) (cell size, standard deviation) Table 4-5-2 Analysis of Variance on N E G E V A L (Study 1) Source Sum of Squares Degrees of Freedom F-Statistic p-value P R O D T Y P E 13.69 1 4.80 .032 OWN .89 1 .31 .578 PROD TYPE * OWN 2.27 1 .80 .375 Error 208.39 73 Total 223.79 76 Table 4-6 Analysis of Covariance on POS E V A L (Study 1) Source Sum of Squares Degrees of Freedom F-Statistic p-value TOT_D 151.42 1 52.61 .000 FAMILIAR 1.10 1 .38 .539 ATTITUDE 1.51 1 .525 .471 P R O D T Y P E 8.45 1 2.94 .091 OWN .52 1 .18 .673 PROD TYPE * OWN .24 1 .08 .774 Error 201.46 70 2.88 Total 449.17 76 Table 4-7 Analysis of Covariance on N E G E V A L (Study 1) Source Sum of Degrees of F-Statistic p-value Squares Freedom T O T D 14.19 1 5.51 .022 FAMILIAR 1.32 1 .51 .477 ATTITUDE 8.21 1 3.19 .079 PROD TYPE 8.93 1 3.47 .067 OWN 1.44 1 .56 .458 PROD T Y P E * OWN 1.62 1 .63 .430 Error 180.36 70 Total 293.00 76 72 C H A P T E R F I V E STUDY T W O This study was designed to address some of the remaining concerns about the pilot study and Study 1. In particular, the current study manipulated self-relatedness using the same product in order to better establish causality between this variable and W O M behaviour. Self-relatedness was manipulated using advertising copy that created a link between the chosen product and self-concept. The idea was that consumers would be more willing to engage in W O M about the product to the extent that the ad suggested it was relevant to self-concept. In addition, some subjects were told their W O M would be judged. This manipulation was inspired by previous research that has used similar techniques to increase self-presentational concerns in social interaction (Federoff and Harvey 1976; Schlenker, Miller and Leary 1983). Specific hypotheses are given below. 5.1 Hypotheses to be Tested Study 2 included a more formal examination of the relationship between product attitudes and W O M valence. On the basis of the existing research on W O M , it was hypothesized that: H6: Product attitudes would predict W O M valence 73 More importantly, this study sought to induce self-promotional concerns by using advertising copy that linked a product to the self-concept, and a social context that implied other people would judge the W O M that subjects gave. Consistent with previous predictions, self-relatedness induced through advertising copy and social context was expected to influence W O M as follows: Willingness to Give WOM H7: Advertising copy that links a product to the self-concept will lead to more total W O M than copy that does not make this link H8: A social context that suggests the W O M giver will be judged on the basis of what is said about the product will lead to more total W O M than a context that does not involve social judgement Valence of Word-of-Mouth H9: Advertising copy that links a product to the self-concept will lead to more positively valenced W O M compared to copy that does not link the product to the self-concept H10: A social context that suggests the W O M giver wil l be judged will lead to more positively valenced W O M compared to a context that does not include social judgement 74 Finally, the effect that product attitudes might have on W O M valence had to be accounted for while studying the effects induced by the advertising copy and judgement context. HI 1: The effect of advertising copy and social context on W O M valence will remain significant even after accounting for the effect of product attitudes 5.2 Considerations in Choosing a Product The products that subjects suggested in the surveys for the pilot study and Study 1 were reviewed and categorized. The objective was to first select a product category and then a specific product that would be used in Study 2. Since self-relatedness was to be manipulated this time, product categories that could be perceived as either utilitarian or self-related seemed appropriate. An examination of the product categories showed that the Food & Drinks category met this criterion (Table 5-1). 62% of the participants in the pilot study regarded food and drink items as utilitarian, and almost half the subjects in Study 1 thought the same. The remaining subjects in both studies thought this product category was self-related. This makes intuitive sense because many products in this category are consumed for practical purposes, and consumers may hardly notice anything special about them. Yet, with further thought or some reminder, some of these products may also convey something personal about the consumer. This category was therefore 75 chosen because subjects' opinions seemed fairly mixed with regards to its self-relatedness. The specific product chosen for this study was a set of snack mixes that was sold under the same brand. These included mixed nuts, rice crackers, sweet candies, and sour candies. Most, i f not all, subjects were expected to have neutral to positive attitudes towards these products. 5.3 Method Subjects and Design Study 2 used a 2 (advertising copy: link vs. no link) by 2 (social context: judgement vs. no judgement) between-subjects factorial design with total W O M and W O M valence serving again as the main dependent measures. Eighty-three subjects from introductory marketing participated in a product sampling study for course credit. They were randomly assigned to one of the four experimental conditions. Three subjects opted out of the study due to food allergies. Procedure Subjects participated individually in product sampling sessions. They first tasted the various items from the set of snack mixes. To make the tasting sessions more realistic, they were each provided with a glass of water and advised to take a drink of water between product samples so as to obtain the genuine flavour of the food. Subjects 76 then rated their attitudes towards the snack items (Appendix 5-1). Attitudes were measured at this stage so that these responses would not be influenced by the experimental manipulations, but rather these ratings would reflect subjects' actual preferences for the snack items. The experimenter next told the subjects that the snack mix was available on the Internet, and asked them to read an Internet ad concerning the brand of snack mix they had sampled. Depending on the condition to which subjects were assigned, the ad either contained a self-relevant message or it did not (this manipulation is further described in the section on Independent Variables) (Appendix 5-2). The experimenter then asked the subjects to imagine that a friend was "thinking about this brand for a party or snacking at home," and was interested in the subjects' opinions and recommendations. Subjects were given thirty seconds to think about what exactly they would tell the friend. When ready, they were led into an adjacent room and told that their responses would be audiotaped. The reasons subjects were given for the recording served as the manipulation of the social context variable (also to be elaborated in the section on Independent Variables). W O M was then recorded. Afterwards, subjects completed a final survey that contained manipulation checks and other items that might be related to the major dependent variables in the study (Appendix 5-3). Independent Variables The first independent variable (advertising copy) was manipulated by creating a link between the snack mix and having "good taste," by using slightly different versions 77 of the advertising copy given to subjects. A pretest conducted two months before this experiment confirmed that "good taste" was regarded as an important aspect of the self-concept for these same subjects. Two versions of a bogus Internet ad for the snack mix were created (Appendix 5-2). One version encouraged subjects to think that the choice of this snack mix showed personal taste and connoisseurship. In particular, subjects who were assigned to the link group read a version of the ad with extra advertising copy that said "Your Personal Expression of Good Taste and Discrete Choice" at the top of the ad, and "For the Savvy Connoisseur" at the bottom. Subjects in the no-link group were presented a version of the ad without these claims. A bogus brand name, Fenton & Lee, was used for the snack mix. The extra advertising copy was expected to facilitate subjects' recognition of the product as relevant to a positive self-image. Prior research has shown that the use of words such as "personal," "self," and "own" tends to make the self more salient (Aaker and Williams 1998). Moreover, this kind of advertising design makes a great deal of managerial sense because it is convenient for marketers to use advertising copy to link a product to central aspects of a consumer's self-concept in order to bring out the perceived importance of the product (Bloch and Richins 1983, p.79). The second independent variable (social context) was manipulated by varying the reasons for recording the subjects' W O M . Subjects assigned to the judgement group were told that their W O M would be recorded, and would be judged later to determine whether or not they were opinion leaders. The experimenter defined opinion leaders as 78 those who were "basically smart shoppers. They know what products are good and make smart choices. For this reason, other consumers tend to consult opinion leaders for advice and suggestions. Opinion leaders also tend to be well respected by the individuals they give advice to." Given that the subjects were expected to have positive attitudes toward the snack mix, it was hoped that this instruction would encourage subjects to emphasize their good experiences with tasting the snack mix, and therefore talk more positively about the product in order to convey that they were opinion leaders. To further motivate subjects in this group to self-promote, they were seated in front ofa video camera, and were told that they would be videotaped. It was expected that subjects would become more self-aware talking in front of the video camera, and give more positive W O M on their experiences. Similar techniques such as seating subjects in front of mirrors to make them more self-aware have been used in previous self-awareness research (e.g., Greenwald 1982; Pryor 1980). In fact, subjects in this study were only audiotaped, and the power on the video camera was turned off. Subjects in the no-judgement group were simply told before giving W O M that their opinions would be audiotaped for data coding at a later date. The video camera was not present when they gave W O M . Dependent Variables Once again, the dependent variables were total W O M and W O M valence. The recorded W O M was transcribed and coded by the author and another coder to obtain the 79 same variables as before (i.e., TOT_D, P O S _ E V A L , N E G E V A L , and POS-NEG). Agreement between the author and the coder reached 88%, 77%, and 77% for TOT_D, P O S _ E V A L , and N E G J E V A L (Pearson correlations' = .90, .89, and .84, p's=.01) respectively. The author's set of scores was used for data analysis. Attitudes towards Fenton & Lee snack mix were measured using a four-item seven-point bipolar scale after subjects tasted the snack items, but before they gave W O M (Appendix 5-2). Subjects rated the snack mix from 1 to 7 on items anchored by bad and good, negative and positive, unfavorable and favorable, and disliked and liked. Responses across these items would be averaged to obtain an attitude index. In the final survey (Appendix 5-3), subjects answered yes-no manipulation check items that asked them to identify the specific advertising copy they had read in their assigned version of the ad. One of the four items was a bogus phrase that did not appear in either version of the ad used in the experiment. To check their self-promotional motive, subjects were asked how concerned they were about being savvy, smart, discrete, and good when they were giving W O M opinions on seven-point scales from -3 (Not At All) to +3 (Very). As subjects might be reluctant in admitting to their own self-promotional intentions, they were also asked to rate typical consumers of the snack mix on the same items using the same scales. Two indices, U _ C O N C E R N and O T H E R _ G O O D respectively, were later created by averaging the items for each set of ratings. 80 5.4 Results 5.4.1 Manipulation Check The manipulation checks suggested that 11 of the 80 subjects (link: seven; no-link: four) did not correctly identify the advertising copy, and their data were eliminated from further analysis. Results were based on responses from the remaining 69 subjects (See Table 5-2 for cell size). The U _ C O N C E R N measure did not reveal any significant effects of the experimental manipulations on this measure of self-presentational concerns. A two-way A N O V A on U _ C O N C E R N showed no significant effects (Tables 5-3-1 and 5-3-2). However, a two-way A N O V A on O T H E R _ G O O D revealed a significant copy by context interaction effect (F(l,65)=4.28, p<.05) (Tables 5-4-1 and 5-4-2). A contrast analysis showed that subjects in the no-link/no-judgement condition rated consumers of the snack mix significantly less smart than subjects in the other three conditions (Fy(l,65)=10.20, p<.005; M no-imk/no-judgement = -.36 vs. M's = .34, .39, .46). Therefore, the self-relevant link created by the advertising copy, the social context, or both, seemed to make subjects believe that consumers of the snack mix were relatively savvy when this less direct measure of self-presentational concerns was examined. 5.4.2 Total Word-of-Mouth A two-way A N O V A was conducted on TOT_D, and a marginal context main effect was found (F(l,65)=2.95, p=.09) (Tables 5-5-1 and 5-5-2), marginally supporting H8. 81 Subjects who were told they would be judged on their opinions gave more total W O M than those who were not given this instruction (Mjudgement= 10.82 vs. Mno-judgement= 8.97). No other effects were significant. Hence, there was no effect of the ad copy manipulation, contrary to H7. There was also no additive effect from the copy and context manipulations. 5.4.3 Valence of Word-of-Mouth POS-NEG A two-way A N O V A on the P O S - N E G measure revealed a copy by context interaction (F(l,65)=3.74, p=.058) (Tables 5-6-1 and 5-6-2). An examination of individual cell means suggested that W O M valence was more positive when the advertising copy linked the snack mix to the self-concept, when subjects expected judgement on their W O M opinions, or both. An analysis contrasting the no-link/no-judgement group with the other three conditions partially confirmed this interpretation (Fy(1,65)=3.37, p=.07). Although independent main effects were predicted for the effects of copy and context on W O M valence (H9 and H10), these findings provided at least partial confirmation of these hypotheses. For some reason, the combination of copy and context did not add to the individual effects of these variables on W O M valence. Nevertheless, the copy and context manipulations combined also improved W O M valence compared to the subjects who received neither of the manipulations. 82 POS_EVAL A two-way A N O V A on P O S _ E V A L revealed no significant effects (Tables 5-7-1 and 5-7-2). NEG_EVAL A two-way A N O V A on N E G _ E V A L showed a copy by context interaction effect (F(l,65)=6.27, p<.05) (Tables 5-8-1 and 5-8-2). Examination of the cell means revealed that subjects in the no-link/no-judgement condition and, surprisingly, those in the link/judgement condition gave more negative evaluations than the other two conditions. Post-hoc tests showed that when W O M was to be judged, subjects who received the extra advertising copy gave more negative evaluations than those who did not receive that copy (M link/judgement = 2.81 vs. M no-iink/judgement = 1-50; t(65)=2.03, p<.05). No other effects were significant. Ruling out Alternative Explanations and Testing Effect of Attitude on WOM Valence An A N C O V A was conducted on POS-NEG with T O T D and product attitude included as covariates. The purpose was to again rule out the alternative explanation that subjects scored higher in POS-NEG simply because they talked more overall, and also to replicate existing findings that product attitudes predicted W O M valence (H6). In addition, it was expected that the effect of the copy and context manipulations on W O M 83 valence would remain significant even after accounting for attitudes (HI 1), indicating a positive bias in W O M . The A N C O V A showed that T O T D was a significant covariate (F(l,63)=7.56, p<01). Product attitude was also marginally significant (F(l,63)=3.15, p<.10), partially supporting H6. However, the copy by context interaction effect remained marginally significant (F(l,63)=2.77, p=.10) (Table 5-9). HI 1 was therefore marginally supported. 5.5 Discussion On the whole, the design of the current study more convincingly established the causality between self-relatedness and W O M behaviour. Unlike the pilot study and Study 1, Study 2 used the same product for all subjects, thereby controlling for any idiosyncratic product attributes that might have confounded the results of the previous studies. The social context manipulation led subjects to give marginally more total W O M (regardless of its valence). More importantly, findings showed that a positive bias in W O M could be induced in either of two ways. First, advertising copy that associated Fenton & Lee snack mix with showing good taste and connoisseurship (aspects of the self that subjects regarded as important) led to more valenced evaluations of the brand. Second, subjects were also more likely to talk positively about the snack mix when the context in which they gave W O M suggested their skill would be judged by others. Study 2 also showed that product attitudes predicted W O M valence. This replicated findings from the previous W O M research. More importantly, the effects of 84 the copy and context manipulations on W O M valence remained marginally significant after accounting for attitudes. This confirmed that positive attitudes alone did not completely explain all positive W O M behaviour, and there was a positive bias in W O M . Most of the subjects had fairly positive attitudes towards the snack mixes (Table 5-10), but those who were not given the self-relevant advertising message or did not expect judgement on their W O M opinions were more neutral than others in giving W O M . Rather, given positive attitudes, positive W O M was more likely when either the product under discussion or the context in which W O M occurred suggested some degree of self-concept relevance that motivated self-promotion behaviour. Either of these manipulations proved to be sufficient to produce positive biases. Although subjects did not reveal any attempt at self-promotion when asked directly in manipulation checks, those who received the copy with link, or expected judgement, or both, did rate consumers of Fenton & Lee as more savvy and smarter than those who did not receive either manipulation (Table 5-4-1). 85 Table 5-1 Major Product Categories Mentioned by Subjects in Pilot Study and Study 1 Product Category P=Pilot l=Studyl % Self-Related % Utilitarian # Respondents Sports Related & Other Outdoor Leisure P 88 12 8 1 86 14 49 Personal Hygiene & Cosmetics P 41 59 17 1 20 80 61 Vehicles P 55 45 11 1 78 22 68 Clothing & Personal Accessories P 67 33 3 1 62 38 73 Food & Drinks P 38 62 32 1 57 43 14 Table 5-2 Cell Size for Data Analysis (Study 2) Link No L ink Judgement 1 6 1 8 No 1 7 1 8 Judgement Table 5-3-1 Cell Means for U _ C O N C E R N (Study 2) Link No Link -.19 -.15 Judgement (16,2.08)* (18, 1.23) No -.31 -.44 Judgement (17, 1.62) (18,1.71) (cell size, standard deviation) Table 5-3-2 Analysis of Variance on U _ C O N C E R N (Study 2) Source Sum of Degrees of F-Statistics p-value Squares Freedom COPY .04 1 .02 .901 CONTEXT .73 1 .26 .611 COPY*CONTEXT .13 1 .04 .834 Error 182.84 65 Total 183.77 68 Table 5-4-1 Cell Means for O T H E R G O O D (Study 2) Link No Link .34 a .39 a Judgement (16, 1.01)* (18,.65) No .46 a - .36 b Judgement (17, .61) (18, 1.09) (cell size, standard deviation) , b A contrast analysis was conducted in which cell means with subscript a and b were assigned contrast co-efficients 1 and -3 respectively. Cell means with different subscripts were significantly different from each other. Table 5-4-2 Analysis of Variance on O T H E R G O O D (Study 2) Source Sum of Squares Degrees of Freedom F-Statistics p-value C O P Y 2.56 1 3.43 .069 CONTEXT 1.75 1 2.34 .131 C O P Y * C O N T E X T 3.20 1 4.28 .043 Error 48.57 65 Total 56.32 68 88 Table 5-5-1 Cell Means for TOT_D (Study 2) Link No Link 10.75 10.89 Judgement (16, 5.34)* (18, 3.53) No 9.00 8.94 Judgement (17, 4.42) (18, 4.50) (cell size, standard deviation) Table 5-5-2 Analysis of Variance on T O T D (Study 2) • Source Sum of Degrees of F-Statistics p-value Squares Freedom COPY .03 1 .00 .969 C O N T E X T 58.72 1 2.95 .091 COPY*CONTEXT .16 1 .01 .928 Error 1295.72 65 Total 1355.07 68 Table 5-6-1 Cell Means for P O S - N E G (Study 2) Link No Link 1.94a 3.44 a Judgement (16,3.71)* (18,3.91) No 2.76 a .89 b Judgement (17,2.70) (18,4.00) (cell size, standard deviation) a' b A contrast analysis was conducted in which cell means with subscript a and b were assigned contrast co-efficients 1 and -3 respectively. Cell means with different subscripts were significantly different from each other. Table 5-6-2 Analysis of Variance on POS-NEG (Study 2) Source Sum of Squares Degrees of Freedom F-Statistics p-value COPY .59 1 .04 .834 CONTEXT 12.85 1 .98 .327 C O P Y * C O N T E X T 49.23 1 3.74 .058 Error 856.22 65 Total 921.30 68 Table 5-7-1 Cell Means for P O S _ E V A L (Study 2) Link No Link 4.75 4.94 Judgement (16,3.11)* (18,3.72) No 4.47 3.56 Judgement (17,2.58) (18,2.96) * (cell size, standard deviation) Table 5-7-2 Analysis of Variance on P O S _ E V A L (Study 2) Source Sum of Degrees of F-Statistics p-value Squares Freedom COPY 2.23 1 .23 .634 CONTEXT 11.97 1 1.23 .272 COPY*CONTEXT 5.30 1 .54 .464 Error 634.62 65 Total 654.81 68 Table 5-8-1 Cell Means for N E G E V A L (Study 2) Link No Link 2.80 a 1.50 b Judgement (16, 2.26)* (18, 1.50) No 1.71 2.67 Judgement (17, 1.72) (18,2.00) (cell size, standard deviation) a ' b Cell means with different subscripts were significantly different from each other. Table 5-8-2 Analysis of Variance on N E G _ E V A L (Study 2) Source Sum of Degrees of F-Statistics p-value Squares Freedom COPY .53 1 .15 .700 CONTEXT .02 1 .004 .947 C O P Y * C O N T E X T 22.23 1 6.27 .015 Error 230.47 65 Total 253.25 68 Table 5-9 Analysis of Covariance on POS-NEG (Study 2) Source Sum of Degrees of F-Statistics Squares Freedom p-value T O T D 83.54 1 7.56 .008 A T T I T U D E 34.75 1 3.15 .081 COPY .17 1 .02 .902 CONTEXT 2.47 1 .22 .638 C O P Y * C O N T E X T 30.56 1 2.77 .101 Error 695.80 63 Total 921.30 68 Table 5-10 Cell Means for Product Attitudes (Study 2) Link No Link 5.11 5.35 Judgement (16, 1.05)* (18, .80) No 5.47 5.01 Judgement (17, .67) (18, .94) (cell size, standard deviation) 93 C H A P T E R SIX 0 F I N A L DISCUSSION 6.1 Summary of Main Findings This research has shown that personal relevance significantly influences both willingness to give W O M , and W O M valence. It adopted various methods for data collection, and these methods enhanced the systematic replications of the findings. The three studies in this dissertation research used different measures to record W O M behaviour at different settings and times. The pilot study measured intended W O M in a written form, and subjects completed the questionnaire in pairs in experimental sessions. Study 1 audiotaped W O M unobtrusively as it occurred in a waiting room setting. Subjects in Study 2 gave W O M individually with the knowledge that they were being recorded. Subjects from the three studies showed more willingness to give W O M in terms of greater detail on the purchase and consumption of self-related products. Specifically, the pilot study and Study 1 demonstrated that products that were relevant to subjects' self-concept led to more W O M detail than utilitarian products. Study 2 showed that a social context that suggested future judgement on W O M opinions increased the total amount of W O M given. The three studies also sought to demonstrate that personal relevance had an influence on the valence of W O M . It was expected that subjects would talk more positively about self-related products that they liked relative to products that were less self-relevant but that they also liked. This finding would be consistent with the existence of a self-promotion bias, in which W O M givers present themselves in a positive light by exaggerating the positive aspects of the products that they recommended. Since positive attitudes are an important indicator of W O M valence, a positive bias in W O M is demonstrated i f W O M remains overly positive above and beyond what positive attitudes can account for. The pilot study did not reveal this bias because product attitudes were not measured. Thus, the effect of product type and ownership on W O M valence might simply have been induced by attitudes, with subjects giving more positive W O M on self-related and owned products that they also liked better. Study 2 provided the clearest evidence for a positive bias in W O M . Attitudes were shown to be an important indicator of W O M valence. Subjects discussed products towards which they had largely positive attitudes, which predicted positive W O M . The extra advertising copy and the social context presented to subjects in Study 2 both led to more positively valenced W O M separately. The effect of these two independent variables remained somewhat significant even after accounting for attitudes. This showed that personal relevance in the form of ad-link and social context had positive effects on the valence of W O M even when attitudes were accounted for. This showed that W O M went beyond what would be expected on the basis of attitudes alone. Study 1 measured product attitudes, but found no correlation between this variable and W O M valence. Moreover, subjects gave more 95 positive and negative W O M on self-related products versus utilitarian products, and seemed to be objective in W O M . 6.2 Implications of Current Research 6.2.1 Conceptual Implications The conceptual framework developed in this research explains how motivation may affect W O M behaviour, and provides guidance for systematically studying the relationship between personal relevance and W O M behaviour. Although the framework can accommodate different types of motivational goals, this dissertation has mainly focused on investigating the self-promotion motive. The emphasis has been to find evidence for the prediction that consumers who are motivated to self-promote will talk more positively about the products that they like in order to present themselves in a positive light in front of their listeners. This research shows that personal relevance can increase W O M behaviour. Given that consumers already have certain attitudes towards a product, they will be motivated to express their opinions on it i f they have a personally relevant reason for giving W O M . The motivational approach used here extends the understanding of W O M behaviour beyond the idea that W O M simply reflects product attitudes. Past research on W O M seems to show that attitudes are reasonably good predictors of behaviour, and that consumers generally see W O M as a trustworthy and accurate source of product information and opinion. Any distortions or embellishments in W O M will generally be 96 interpreted as cognitive errors due to memory loss, or a lack of knowledge. However, the motivational framework suggests that W O M may be distorted, embellished or polarized due to the self-serving biases of individual consumers when giving W O M . Given the prevalence of such motives in human behaviour, it may be that W O M is much less accurate than consumers often assume. 6.2.2 Managerial Implications It is commonly believed that W O M is hard to control and create despite its importance for marketers. The current research showed that positive W O M could be encouraged through advertising. Specifically, in situations where consumers already like a product, advertising copy that leads them to further associate the product with an important aspect of the self-concept is also likely to lead them to talk more about it, and perhaps even exaggerate the positive aspects of it. By giving positive evaluations of the product, consumers may also tell others something important and positive about themselves, thus portraying a positive image. The advertising copy manipulation used in Study 2 suggests one way to manipulate this kind of self-promotion-motivated W O M behaviour. Marketers might also consider other elements within the advertising context. For example, advertising messages communicated by celebrities, whose image or status the target consumer wishes to attain, may convey self-relevance to the audience and increase the willingness to give positive W O M . 97 In fact, many marketers have already developed successful advertising campaigns by creating links between their products and the consumer's sense of self. The classic Marlboro Man of tobacco fame, or Michael Jordan for Nike are two such examples in which advertisers play on the idea that consumers associate the advertised products with some aspects of the self-concept that are portrayed by the endorsers. In the case of the Marlboro Man, consumers who use the product may view themselves as rugged individualists, whereas Nike users may view themselves as sporty people who lead an active lifestyle. However, existing advertisements that contain such an association mainly aim at changing consumer attitudes towards the ad and the product. The current research demonstrates that these ads may also be beneficial in encouraging W O M for a wide variety of products, provided that marketers manage to identify the aspects of the self-concept that are important to their target consumers, and link them to the products. This research has also shown that positive W O M on favorable products could be obtained in a W O M context where the W O M giver expects his or her recommendations to be judged by others. Marketers might try to create social settings in which consumers' W O M opinions on products would be heard and evaluated. Although this might seem difficult, some marketers have already made an effort in this respect. For example, Swedish mobile phone operator Ericsson has thrown parties to spread brand messages to trendy clubbers in hopes that they would in turn talk about Ericsson products to friends as a signal of trendiness (Lutchford 1997). Companies such as Chevrolet and Saturn 98 Corporation have also invited consumers to give presentations on their opinions regarding the companies' products (Heckman 1999; Silverman 1997). Another implication of this research for marketers is that some products are more likely to be related to the self than others (e.g., running shoes and vehicles vs. computers and toiletry items). The findings suggest that it is easier to generate positive W O M when selling products that appeal to broadly held aspects of the self-concept than when selling products that are more utilitarian. Marketers of self-relevant products can devise appropriate marketing tools to further strengthen the link between their offerings and aspects of the self-concept. Marketers of utilitarian products, on the other hand, may find it hard to convince customers that their products are personally relevant to them for the purpose of encouraging more W O M behaviour. Moreover, the fact that some subjects in Study 1 had to be reclassified from the self-related to the utilitarian conditions, or vice versa, seems to suggest that self-relevance of some products might be temporally unstable, and vary across individuals. Yet, the benefit of this temporal instability is that marketers may have the opportunity to convince consumers that a utilitarian product can nevertheless reflect some important aspect of their self-concept. As Study 2 showed, positive bias in W O M might still occur even for fairly neutral products such as snack mix used there, so long as marketers encouraged consumers to link the product with having "good taste" and showing connoisseurship. Finally, at the same time that positive W O M may boost sales, too much exaggeration may backfire, i f products do not live up to the expectations that positive 99 W O M may create. A company may end up discouraging repurchase behaviour or reduce customer satisfaction i f it offers a product that is not as good as their customers suggest when giving W O M . For example, almost everyone has had the experience of hearing too much about a good movie from others before actually seeing it and then being dissatisfied. In such cases, positive W O M can often detract from the level of satisfaction of the W O M receiver. There is likely to be some optimal level of endorsement that products can receive before long-term satisfaction begins to suffer. This optimum should be related to the actual quality of the product. 6.2.3 Methodological Implications Study 1 used a new approach to record W O M conversations. Unlike some existing W O M studies in which subjects were asked to recall W O M conversations, the experiment in Study 1 was designed so that W O M was recorded as it occurred. The circumstances that the subjects encountered were similar to a real world situation in which one shopper might seek help from another in the process of making a purchase decision. Of course, it could be argued that the W O M recorded was not entirely naturalistic, since the experimental environment surrounding subjects was artificially created. However, the study did include considerably more experimental realism than in previous studies, where subjects were simply asked to imagine what W O M they would give. Subjects in Study 1 thought that the person they gave W O M to would really need the information in order to give their opinions on the products that they would evaluate in 100 the upcoming study. In addition, this experimental design provided better control than studying W O M in an actual consumer environment. 6.3 Limitations of Current Research 6.3.1 Use of Student Subjects Undergraduate students in the business program were the sole source of data in this research. Questions may therefore be raised regarding the generalizability of results to other consumer groups. The subjects for the three studies were fairly homogeneous in terms of their education level, consumption preferences, and perception of what aspects of the self were important to them. This homogeneity might have made them particularly sensitive to the experimental designs or stimuli. 6.3.2 Concerns for Accuracy in Word-of-Mouth This research has not addressed the issue of which factors may limit self-promotion behaviour in W O M . For instance, it is likely that consumers who talk positively about a product in order to present a positive self-image will also be concerned with the accuracy of their opinion. Indeed, researchers have pointed out that people need to be concerned about the plausibility or believability of their self-presentations (Schlenker 1975, 1980; Schlenker and Leary 1982; Tice et al. 1995). Overly favourable self-presentations may be counter-productive and lead to a loss of face and credibility when the speakers do not live up to such impressions (Baumeister, Tice, and Hutton 1989; Tice 1991, 1993; Tice 101 and Baumeister 1990). Moreover, speakers who seek to create overly favourable impressions of themselves may be considered arrogant and conceited, which are both negative personality traits (Schlenker and Leary 1982) that may cause the attempt at self-promotion to backfire. This deters most people from making claims that are grossly inconsistent with their self-concept. Instead, people tend to present themselves more positively (or less accurately) only when they feel they can reasonably expect to get away with it (Baumeister 1982a; Baumeister and Jones 1978; Schlenker 1975), and will not be held strictly accountable for what they say (Lerner and Tetlock 1999; Tetlock and Kim 1987). Otherwise, they tend to strike a balance between favorability and plausibility (Schlenker 1980, 1985, 1986), or between positivity and consistency (Linville and Carlston 1994). In sum, they will try to portray themselves as favorably as possible within the limits of what will be perceived as plausible. In the W O M context, accuracy motives may serve to constrain the degree to which W O M is biased by self-promotional concerns. A W O M giver may be motivated to talk very positively about a good product to cast himself or herself in a more favourable light. Yet, the communicator may also be concerned about giving accurate advice. This situation is particularly likely to arise when the giver knows that the recipient is seriously considering making a purchase, and that the W O M given will be heavily weighted in decision-making. In hindsight, this circumstance is not unlike the experimental context in which subjects in Study 1 gave W O M . There, subjects talked about their product experiences to another person whom they thought would use the given information right 102 away. This expectation might have raised their accuracy concerns, despite the fact that they were also concerned about self-promotion (as shown by their responses to the relevant manipulation check items), resulting in a greater number of both positive and negative evaluations of products. Imminent purchase may therefore serve as a counterbalancing factor to any self-promotion attempts that the W O M giver is likely to display. The intention to give accurate advice may overcome the motive to inflate product experiences. 6.3.3 Ownership The pilot study showed weak ownership effects on W O M behaviour that were not replicated in Study 1. Since the major purpose of Study 2 was to better establish causality between product self-relatedness and W O M behaviour, the ownership variable was not further pursued. However, there may be specific circumstances in which ownership (or the lack of it) matters, and may affect W O M behaviour. Contrary to expectations, the results in Study 1 showed no significant effects of ownership. A possible reason could be that when asked to suggest products that were self-related but not owned, some subjects listed products (e.g., automobiles) that they aspired to own one day. The aspiration to own these products might have induced W O M behaviour that was more similar to that of owned products than products that subjects never planned to own. In asking subjects to suggest products that they did not own, 103 Study 1 did not distinguish W O M responses from subjects who aspired to own the products versus those who never planned to own the products. 6.4 Future Directions 6.4.1 Other Motivational Goals Future research can seek to understand more about how accuracy motive may affect W O M behaviour directly or through interacting with the self-promotion motive. An accuracy-motivated person tends to express appropriate (Kunda 1990), consistent (Baumeister 1995), objective and valid (Chaiken et al. 1996), or precise and complete opinions. There will be more effort spent on processing information, resulting in unbiased, or less biased, judgement. Past research suggests that an accuracy motive is increased if the welfare of a third party is affected (e.g., Freund, Kruglanski, and Shpitzajzen 1985; McAllister, Mitchell, and Beach 1979, Experiment 3). In the W O M context, a consumer may strive for accuracy i f W O M directly contributes to an important purchase decision. This may happen when the W O M recipient is seriously considering buying the product rather than just asking for information for the sake of interest or striking conversation. A n accuracy-motivated consumer may give non-evaluative information on a product (within the limits of time and knowledge), and at the same time be as honest as possible in expressing what he or she really thinks about the product. This may involve talking positively about a product that the W O M giver actually likes, and negatively about a dissatisfactory product. W O M 104 should therefore appropriately reflect the pros and cons of the target product in order to serve accuracy goals. W O M may also be motivated by the goal to socialize with other people. Consumers who give W O M for social facilitation may be particularly likely to agree with their friends on the evaluations of a product. Research has shown that people sometimes "tailor their public images to the perceived values and preferences of significant others" (Leary and Kowalski 1993; see also Gaes and Tedeschi 1978; Mori, Chaiken, and Pliner 1987; von Baeyer, Sherk, and Zanna 1981; Zanna and Pack 1975). High self-monitors are particularly prone to doing this because their behaviours are motivated more by situational cues than internal states. In this case, W O M may correspond closely to the initial attitudes of the recipients with whom the giver wants to facilitate a positive social relationship. A high self-monitor may spread positive W O M rather than negative W O M despite his or her negative attitude if he or she believes that most people like a product. Nevertheless, in relation to the social facilitation goal, consumers may give W O M to help their friends make better purchases. If that is the case, then W O M may tend to be accuracy-driven because consumers are aware that their opinions will weigh significantly in the upcoming purchase. 6.4.2 Other Samples Since W O M behaviour is ubiquitous and not limited to the population of consumers examined in this research, it would be useful to try to generalize the results across other 105 consumer groups. One potentially interesting dimension is to investigate cultural differences in the self-promotion motive (Aaker 1999), and how this affects W O M behaviour. People who are raised in cultures that foster an independent construal of self tend to assert their individuality and uniqueness. Those in cultures that facilitate an interdependent construal of self tend to value interpersonal harmony and strive to fit in with their social groups (Markus and Kitayama 1991; Triandis 1989). The self-serving biases in W O M induced by self-promotion motive seem more in line with an independent self-concept. Consumers in such cultures may try to stand out among each other. They may be more motivated than those in collectivist cultures to give W O M for the purpose of revealing their sense of pride and achievement. This type of self-serving behaviour seems less likely for consumers with an interdependent construal of self. In cultures where collectivistic ideas prevail, acceptance by others is important, and self-effacement in the form of seeing everybody as average is common. People may be motivated to consume to fit in with their groups, and are less likely to self-promote. Instead, they may be more inclined to give W O M to help friends with making purchase decision, or to be agreeable with others. 6.4.3 Motivations for Negative Word-of-Mouth The current research did not examine motivations for and deterrents of negative W O M . Previous research has shown that negative W O M could be more harmful to companies than positive W O M would be beneficial (Anderson 1998; TARP 1979). At the same 106 time that marketers want to encourage more positive W O M , they may also want to discourage negative words when consumers are dissatisfied. Naturally, marketers have to ensure prompt and satisfactory solutions and compensations in the case of dissatisfaction. However, research has shown that consumers may give negative W O M even after complaining to the company, particularly when they are not satisfied with the company's responses (Blodgett et al. 1993; Richins 1983). Some types of personal goals may affect negatively valenced W O M . For instance, on one hand, a consumer who aims at self-promoting may ignore or attend less carefully to product information that threatens valued aspects of the self. Although he or she may have negative experiences with a product, the W O M given may exclude such experiences. This way, the giver will avoid having the experiences reflect negatively upon himself or herself. However, on the other hand, a W O M giver may be more motivated to talk negatively about a bad product i f he or she does not own it. This is because the W O M giver may benefit from the negative focus in W O M by showing that he or she is smart enough to stay away from buying the product. In addition, some circumstances may require that the W O M givers be skeptical, and skepticism may be reflected in the elaboration on the negative attributes of products. For example, a movie critic may derive professional pride and trust in dissecting a movie in order to expose the negative aspects of it. Also, when motivated to facilitate social relationships, a consumer whose reference group talks positively about a product may follow suit despite his or her 107 negative attitude toward it. Future research might explore the possibility that creating competing motives for W O M might also reduce the amount of negative W O M that dissatisfied customers give. 6.4.4 Consequences of WOM This dissertation research did not examine the possible social effects that W O M might have after it has been given. Some existing studies have looked at W O M in a social network setting (e.g., Brown and Reingen 1987; Reingen and Kernan 1986). There, the focus is on how tie strength (i.e., the closeness of the relationship) facilitates the flow of product information and recommendation. However, it also seems relevant to examine how W O M communications may affect product beliefs and attitudes. For example, what kind of W O M information will be most communicable and likely to be passed on in the W O M communication chain? What kind of product information will be most well received by the W O M recipient? Do W O M givers and recipients follow any rules and norms in exchanging product opinions the way people do in daily conversations (Grice 1975)? Do they in turn influence those norms as a result of their communications? Future research may examine these issues, and attempt to understand why and in what ways W O M concerning the purchase and consumption of products may be similar to or . different from other kinds of social discourse. 108 BIBLIOGRAPHY Aaker, Jennifer L. (1999), "The Malleable Self: The Role of Self-Expression in Persuasion," Journal of Marketing Research, 36(1), 45-57. and patti Williams (1998), "Empathy Versus Pride: The Influence of Emotional Appeals Across Cultures," Journal of Consumer Research, 25(December), 241-261. Adler, Alfred (1930), "Individual Psychology," in Psychologies of1930, ed. C. Murchison, London, H. Milford: Oxford University Press, 395-405. Ajzen, leek (1987), "Attitudes, Traits and Actions: Dispositional Prediction of Behavior in Personality and Social Psychology," in Advances in Experimental Social Psychology, 20, ed. L. Berkowitz, New York: Academic Press, 1-63. — and Martin Fishbein (1977), "Attitude-Behavior Relations: A Theoretical Analysis and Review of Empirical Research," Psychological Bulletin, 84, 888-918. Allport, Gordon W. (1937), Personality: a Psychological Interpretation, New York: Holt. (1955), Becoming: Basic Considerations for a Psychology of Personality, New Haven: Yale University Press. Alsop, R. (1984), "Study of Olympics Ads Casts Doubts on Value of Campaigns," Wall Street Journal, December 6. Anderson, Eugene W. (1998), "Customer Satisfaction and Word-of-Mouth," Journal of Service Research, 1(1), 1-14. Apsler, Robert and David O. Sears (1968), "Warning, Personal Involvement, and Attitude Change," Journal of Personality and Social Psychology, 9, 162-166. Arkin, Robert M . and Ann H . Baumgardner (1986), "Self-Presentation and Self-Evaluation: Processes of Self-Control and Social Control," in Public Self and Private Self, ed. R. F. Baumeister, New York: Springer-Verlag, 75-97. Arndt, Johan (1967), Word of Mouth Advertising. New York: Advertising Research Foundation Inc. 109 (1968), "Word-of-Mouth Advertising and Perceived Risk," in Perspectives in Consumer Behavior, ed. H. Kassarjian and T. Robertson, Glenwood, IL: Scott, Foresman, 330-336. Aronson, Elliot (1968), "Dissonance Theory: Progress and Problems," Theories of Cognitive Consistency: A Sourcebook, eds., R. P. Abelson, E. Aronson, W. J. McGuire, T. M . Newcomb, M . J. Rosenberg, and P. H. Tannenbaum, Chicago: Rand McNally, 5-27. 5 j . Turner, and J. M . Carlsmith (1963), "Communicator Credibility and Communication Discrepancy as Determinants of Opinion Change," Journal of Abnormal and Social Psychology, 67, 31-36. Barone, Michael J., Terence A . Shimp, and David E. Sprott (1999), "Product Ownership as a Moderator of Self-Congruity Effects," Marketing Letters, 10(1), 75-85. Baumeister, Roy F. (1982a), "Self-Esteem, Self-Presentation, and Future Interaction: A Dilemma of Reputation," Journal of Personality, 50, 29-45. - (1982b), " A Self-Presentational View of Social Phenomena," Psychological Bulletin, 91, 3-26. (1995), "Self and Identity: An Introduction," Advanced Social Psychology, ed. Abraham Tesser, New York: McGraw-Hill, Inc., 51-98. — and Edward E. Jones (1978), "When Self-Presentation is Constrained by the Target's Knowledge: Consistency and Compensation," Journal of Personality and Social Psychology, 36(6), 608-618. — and Dianne M . Tice (1986), "Four Selves, Two Motives, and a Substitute Process Self-Regulation Model," in Public Self and Private Self, ed. R. F. Baumeister, New York: Springer-Verlag, 63-74. , and Debra G. Hutton (1989), "Self-Presentational Motivations and Personality Differences in Self-Esteem," Journal of Personality, 57, 547-579. Baumgardner, Ann H. , Elizabeth Lake, and Robert M . Arkin (1985), "Claiming Mood as a Self-Handicap: The Influence of Spoiled and Unspoiled Public Identity," Personality and Social Psychology Bulletin, 11, 349-357. Beggan, James K. (1992), "On the Social Nature of Nonsocial Perception: The Mere Ownership Effect," Journal of Personality and Social Psychology, 62, 229-237. 110 Belk, Russell W. (1987), "Possessions and Extended Sense of Self," Marketing and Semiotics: New Directions in the Study of Signs for Sale, ed. Jean Umiker-Sebeok, Berlin, Germany: Mouton de Gruyter, 151-164. (1988), "Possessions and the Extended Self," Journal of Consumer Research, 15(September), 139-168. Bern, Daryl J. and Andrea Allen (1974), "On Predicting Some of the People Some of the Time: The Search for Cross-Situational Consistencies in Behavior," Psychological Review, 81(6), 506-520. and David C. Funder (1978), "Predicting More of the People More of the Time: Assessing the Personality of Situation," Psychological Review, 85(November), 485-501. Benedict, Ruth (1934), Patterns of Culture. Boston: Houghton Mifflin. Biel, Alexander (1993), "Converting Image into Equity," in Brand Equity and Advertising, eds. David A . Aaker and Alexander Biel, Hillsdale, NJ: Lawrence Erlbaum Associates, 67-83. Birdwell, A . E. (1968), " A Study of the Influence of Image Congruence on Consumer Choice, "Journal of Business, 41(January), 76-88. Bloch, Peter H . and Marsha L. Richins (1983), " A Theoretical Model for the Study of , Product Importance Perceptions," Journal of Marketing, 47(Summer), 69-81. Blodgett, Jeffrey G., Donald H. Granbois, and Rockney G. Walters (1993), "The Effects of Perceived Justice on Complainants' Negative Word-of-Mouth Behaviour and Repatronage Intentions," Journal of Retailing, 69(4), 399-428. Bohra, Kayyum A. and Janak Pandey (1984), "Ingratiation toward Strangers, Friends, and Bosses," Journal of Social Psychology, 22, 217-222. Bristor, Julia M . (1990), "Enhanced Explanations of Word of Mouth Communications: The Power of Relationships," Research in Consumer Behavior, 4, 51-83. Brown, Jacqueline J. and Peter H. Reingen (1987), "Social Ties and Word-of-Mouth Referral Behavior," Journal of Consumer Research, 14(December), 350-362. Busch, Paul S. and Michael J. Houston (1985), Marketing: Strategic Foundation, Homewood, IL: Richard D. Irwin, Inc. I l l Cantril, Hadley and Gordon W. Allport (1935), The Psychology of Radio. New York: Harper. Chaiken, Shelly, Roger Giner-Sorolla, and Serena Chen (1996), "Beyond Accuracy -Defense and Impression Motives in Heuristic and Systematic Information Processing," in The Psychology of Action - Linking Cognition and Motivation to Behavior, eds. P. M . Gollwitzer and J. A . Bargh, New York: Guildford Press, 553-578. 1 Akiva Liberman, and Alice H . Eagly (1989), "Heuristics and Systematic Information Processing Within and Beyond the Persuasion Context," Unintended Thought: Limits of Awareness, Intention, and Control, eds., J. S. Uleman and J. A . Bargh, New York: Guilford Press, 212-252. and Charles Stangor (1987), "Attitudes and Attitude Change," Annual Review of Psychology, 38, 575-630. Coleman, James, Elihu Katz, and Herbert Menzel (1966), Medical Innovation: A Diffusion Study, New York: Bobbs-Merrill. Csikszentmihalyi, Mihaly and Eugene Rochberg-Halton (1981), The Meanings of Things — Domestic Symbols and the Self, Cambridge University Press. Cunningham, Scott (1964), "Perceived Risk as a Factor in Product - Oriented Word-of-Mouth Behavior: A First Step," in Reflections on Progress in Marketing, ed. L . G. Smith, Chicago: American marketing Association, 229-238. . (1967), "Perceived Risk as a Factor in Informal Consumer Communications," in Risk Taking and Information Handling in Consumer Behavior, ed. D. F. Cox, Boston: Harvard University Press, 265-288. Day, George S. (1971), "Attitude Change, Media and Word of Mouth," Journal of Advertising Research, 11(6), 31-40. Deutscher, Irwin (1966), "Words and Deeds: Social Science and Social Policy," Social Problems, 13,235-253. Dichter, Ernest (1966), "How Word-of-Mouth Advertising Works," Harvard Business Review, November-December, 147-166. Dolich, Ira J. (1969), "Congruence Relationships between Self Images and Product Brands," Journal of Marketing Research, 6(February), 80-84. 112 Doob, Leonard W. (1948), Public Opinion and Propaganda, New York: Henry Holt. Douglas, J., Field, G. A. , & Tarp, L. X . (1970), "Human Behavior in Marketing," in Consumer Behavior in Theory and in Action, ed. S. H. Britt, New York: John Wiley and Sons, Inc. Engel, James F., Robert J. Kegerreis, and Roger D. Blackwell (1969), "Word-of-Mouth Communication by the Innovator," Journal of Marketing, 33(July), 15-19. Epstein, Seymour (1980), "The Self-Concept: A Review and the Proposal of an Integrated Theory of Personality," in Personality: Basic Aspects and Current Concerns, ed. E. Staub, Englewood Cliffs, NJ: Prentice-Hall, 81-132. Fazio, Russell H. and Mark P. Zanna (1981), "Direct Experience and Attitude-Behaviour Consistency," in Advances in Experimental Social Psychology, 14, ed. L . Berkowitz, New York: Academic Press, 161-202. Federoff, Nancy A. and John H . Harvey (1976), "Focus of Attention, Self-Esteem, and the Attribution of Causality," Journal of Research in Personality, 10(3), 336-345. Feldman, Sidney and Merlin Spencer (1965), "The Effect of Personal Influence in the Selection of Consumer Services," in Marketing and Economic Development, ed. Peter Bennett, Chicago: American Marketing Association, 440-452. Fodor, Eugene M . and Terry Smith (1982), "The Power Motive as an Influence on Group Decision Making," Journal of Personality and Social Psychology, 42(1), 178-185. Form, William H . and Gregory P. Stone (1957), "Urbanism, Anonymity, and Status Symbolism," American Journal of Sociology, 62, 504-514. Fornell, Claes and Fred L. Bookstein (1982), "Two Structural Equation Models: LISREL and PLS Applied to Exit-Voice Theory," Journal of Marketing Research, 19(November), 440-452. Freund, Tallie, Arie W. Kruglanski, and Avivit Shpitzajzen (1985), "The Freezing and Unfreezing of Impressional Primacy: Effects of the Need for Structure and the Fear of Invalidity," Personality and Social Psychology Bulletin, 11, 479-487. Fromm, Erich (1956), The Act of Loving, New York: Harper. Gaes, Gerald G. and James T. Tedeschi (1978), "An Evaluation of Self-Esteem and Impression Management Theories of Anticipatory Belief Change," Journal of Experimental Social Psychology, 14, 579-587. 113 Gardner, Burleigh B. and Sidney J. Levy (1955), "The Product and the Brand," Harvard Business Review, 33(March-April), 33-39. Gardner, William L. and Mark J. Martinko (1988), "Impression Management: An Observational Study Linking Audience Characteristics with Verbal Self-Presentations," Academy of Management Journal, 31(1), 42-65. Gergen, Kenneth J. and V. Barbara, Wishnov (1965), "Others' Self-Evaluation and Interaction Anticipation as Determinants of Self-Presentation," Journal of Personality and Social Psychology, 2, 348-358. Gitelson, Richard and John Crompton (1983), "The Planning Horizons and Sources of Information used by Pleasure Vacationers," Journal of Travel Research, 21(3), 2-7. Globe and Mail (1999), "Word of Mouth Seen as Key Marketing Tool," May 12, M l . Goffman, Erving (1951), "Symbols of Class Status," British Journal of Sociology, 2, 294-304. (1959), The Presentation of Self in Everyday Life, Garden City, N Y : Doubleday Anchor. Greenwald, Anthony G. (1980), "The Totalitarian Ego: Fabrication and Revision of Personal History," American Psychologist, 35, 603-618. -- (1982), "Ego Task Analysis: An Integration of Research on Ego-Involvement and Self-Awareness," in Cognitive Social Psychology, eds. A . Hastorf and A. Isen, New York: Elsevier. -— , Francis S. Bellezza, and Mahzarin R. Banaji (1988), "Is Self-Esteem a Central Ingredient of the Self-Concept?" Personality and Social Psychology Bulletin, 14, 34-45. Grice, H. P. (1975), "Logic and Conversation," in Syntax and Semantics 3: Speech Acts, eds. P. Cole and J. L . Morgan, New York: Academic Press. Grubb, Edward L. and Gregg Hupp (1968), "Perception of Self, Generalized Stereotypes, and Brand Selection," Journal of Marketing Research, 5(February), 58-63. and Harrison L. Grathwohl (1967), "Consumer Self-Concept, Symbolism, and Market Behaviour: A Theoretical Approach," Journal of Marketing, 31, 22-27'. 114 and Bruce L. Stern (1971), "Self-Concept and Significant Others," Journal of Marketing Research, 8(August), 382-385. Heckman, James (1999), "Word-of-Mouth Advertising: Cheap, Credible, Created," Marketing News, 33(3), 2. Heine, Steven J. and Darrin R. Lehman (1995), "Cultural Variation in Unrealistic Optimism: Does the West Feel More Invulnerable Than the East?" Journal of Personality and Social Psychology, 68(4), 595-607. Hendricks, Michael and Philip Brickman (1974), "Effects of Status and Knowledgeability of Audience on Self-Presentation," Sociometry, 37, 440-449. Herr, Paul M . , Frank R. Kardes, and John Kim (1991), "Effects of Word-of-Mouth and Product-Attribute Information on Persuasion: A n Accessibility-Diagnosticity Perspective," Journal of Consumer Research, 17(March), 454-462. Holbrook, Morris B. and William J. Havlena (1988), "Assessing the Real-to-Artificial Generalizability of Multi-Attribute Attitude Models in Tests of New Product Design," Journal of Marketing Research, 25(February), 25-35. and Elizabeth Hirschman (1982), "The Experiential Aspects of Consumption: Consumer Fantasies, Feelings, and Fun," Journal of Consumer Research, 9(2), 132-140. Holmes, John H . and John D. Lett, Jr. (1977), "Product Sampling and Word of Mouth," Journal of Advertising Research, 17(5), 35-40. Hull, Jay G. and Allan S. Levy (1979), "The Organizational Functions of the Self: An Alternative to the Duval and Wicklund Model of Self-Awareness," Journal of Personality and Social Psychology, 44, 1097-1109. Hyman, H. (1945), "Do They Tell the Truth?" Public Opinion Quarterly, 8, 537-539. Jaccard, James, G. William King, and Richard Pomazal (1977), "Attitudes and Behavior: An Analysis of Specificity of Attitudinal Predictors," Human Relations, 30, 817-824. Johnson, Blair T. and, Alice H . Eagly (1990), "Involvement and Persuasion: Types, Traditions, and the Evidence," Psychological Bulletin, 107(3), 375-384. 115 Jones, Edward E., Kenneth J. Gergen, Peter Gumpert, and John W. Thibaut (1965), "Some Conditions Affecting the Use of Ingratiation to Influence Performance Evaluation," Journal of Personality and Social Psychology, 1, 613-625. a n d Thane S. Pittman (1982), "Toward a General Theory of Strategic Self-Presentation," in Psychological Perspectives on the Self, Vol . 1, ed. J. Suls, Hillsdale, NJ: Erlbaum, 231-262. Kassarjian, Harold H . (1978), "Presidential Address, 1977: Anthropomorphism and Parsimony," in Advances in Consumer Research, Vol . 5, ed. H . K . Hunt, Ann Arbor, MI: Association for Consumer Research, xiii-xiv. Katona, George and Eva Mueller (1954), " A Study of Purchasing Decisions," in Consumer Behavior: The Dynamics of Consumer Reaction, ed. L . H . Clark, New York, New York University Press, 30-87. Katz, Elihu and Paul F. Lazarsfeld (1955), Personal Influence: The Part Played by People in the Flow of Mass Communications, Glencoe, IL: Free Press. Keaveney, Susan M . (1995), "Customer Switching Behavior in Service Industries: An Exploratory Study," Journal of Marketing, 59(April), 71-82. Kelly, George A . (1955), The Psychology of Personal Constructs, New York: Norton. Kelman, Herbert C. (1974), "Attitudes are Alive and Well and Gainfully Employed in the Sphere of Action," American Psychologist, 29, 310-324. Kiel, Geoffrey and Roger Layton (1981), "Dimensions of Consumer Information Seeking Behavior," Journal of Marketing Research, 18 (May), 233-239. Kihlstrom, John F. and Nancy Cantor (1984), "Mental Representations of the Self," Advances in Experimental Social Psychology, 17, 1-47. Kintsch, Walter and Elizabeth Bates (1977), "Recognition memory for statements from a classroom lecture," Journal of Experimental Psychology: Human Learning and Memory, 3, 150-159. Knower, Franklin H. (1935), "Experimental Studies of Changes in Attitude. I. A Study of the Effect of Oral Argument on Changes of Attitude," Journal of Social Psychology, 6, 315-347. 116 (1936), "Experimental Studies of Changes in Attitude. II. A Study of the Effect of Printed Argument in Changes of Attitude," Journal of Abnormal and Social Psychology, 30, 522-532. Kraus, Stephen J. (1995), "Attitudes and the Prediction of Behavior: A Meta-Analysis of the Empirical Literature," Personality and Social Psychology Bulletin, 21(1), 58-75. Kruglanski, Arie W. and leek Ajzen (1983), "Bias and Error in Human Judgment," European Journal of Social Psychology, 13, 1-44. and Tallie Freund (1983), "The Freezing and Unfreezing of Lay-Inferences: Effects on Impressional Primacy, Ethnic Stereotyping, and Numerical Anchoring," Journal of Experimental Social Psychology, 19, 448-468. and Yechiel Klar (1987), " A View from the Bridge: Synthesizing the Consistency and Attribution Paradigms from a Lay Epistemic Perspective," European Journal of Social Psychology, 17, 211 -241. Kunda, Ziva (1987), "Motivation and Inference: Self-Serving Generation and Evaluation of Evidence," Journal of Personality and Social Psychology, 53, 636-647. (1990), "The Case for Motivated Reasoning," Psychological Bulletin, 108(3), 480-498. Landon, E. Laird Jr. (1974), "Self Concept, Ideal Self Concept, and Consumer Purchase Intentions," Journal of Consumer Research, 1 (September), 44-51. Leary, Mark R. and Robin M . Kowalski (1990), "Impression Management: A Literature Review and Two-Component Model," Psychological Bulletin, 107(1), 34-47. LeGrand, Bruce and Jon Udell (1964), "Consumer Behavior in the Market Place," Journal of Retailing, 40 (Fall), 32-48. Lerner, Jennifer S. and Philip E. Tetlock (1999), "Accounting for the Effects of Accountability," Psychological Bulletin, 125(2), 255-275. Levy, Sidney J. (1959), "Symbols for Sale," Harvard Business Review, 37(July-August), 117-124. Linville, Patricia W. and Donal E. Carlston (1994), "Social Cognition of the Self," in Social Cognition: Impact on Social Psychology, eds. Patricia G. Devine and David Lewis Hamilton, San Diego, CA, USA: Academic Press, Inc., 143-193. 117 Lutz, Richard and P. Reilly (1973), "An Exploration of the Effects of Perceived Social and Performance Risk on Consumer Information Acquisition," in Proceedings, Fourth Annual Conference, The Association for Consumer Research, 393-405. Markus, Hazel R. (1977), "Self-Schemata and Processing Information about the Self," Journal of Personality and Social Psychology, 35, 63-78. (1983), "Self-Knowledge: An Expanded View," Journal of Personality, 51, 543-565. — and Shinobu Kitiyama (1991), "Culture and the Self: Implications for Cognition, Emotion, and Motivation," Psychological Review, 98(2), 224-253. and Paula Nurius (1986), "Possible Selves: The Interface between Motivation and the Self-Concept," in Self and Identity, eds. K . Yardley and T. Honess, New York: John Wiley & Sons Ltd., 157-172. and Keith P. Sentis (1982), "The Self in Social Information Processing," in Psychological Perspectives on the Self, Vol . 1, ed. J. Suls, Hillsdale, NJ: Erlbaum, 41-70. . . . . . and Elissa Wurf (1987), "The Dynamic Self-Concept: A Social Psychological Perspective," Annual Review of Psychology, 38, 299-337. Maslow, Abraham H . (1962), "Some Basic Propositions of a Growth and Self-Actualization Psychology," Perceiving, Behaving, Becoming: A New Focus for Education, Washington, D.C.: Yearbook of the Association for Supervision and Curriculum Development, 34-49. McAllister, Daniel W., Terence R. Mitchell, and Lee R. Beach (1979), "The Contingency Model for the Selection of Decision Strategies: An Empirical Test of the Effects of Significance, Accountability, and Reversibility," Organizational Behavior and Human Decision Processes, 24, 228-244. Miller, Dale T. and Michael Ross (1975), "Self-serving Biases in the Attribution of Causality: Fact or Fiction?" Psychological-Bulletin, 82(2), 213-225. Mischel, Walter (1968), Personality and Assessment, New York: Wiley. Mizerski, Richard W. (1982), "An Attribution Explanation of the Disproportionate Influence of Unfavorable Information," Journal of Consumer Research, 9(December), 301-310. 118 Mooradian, Todd A. and James M . Olver (1997), ""I Can't Get No Satisfaction:" The Impact of Personality and Emotion on Postpurchase Processes," Psychology & Marketing, 14(4), 379-393. Mori, DeAnna, Shelly Chaiken, and Patricia Pliner (1987), "Eating Lightly and the Self-Presentation of Femininity," Journal of Personality and Social Psychology, 53, 693-702. Neisser, Ulric (1982), "John Dean's Memory: A Case Study," in Memory Observed: Remembering in Natural Contexts, ed. U . Neisser, San Francisco: W. H . Freeman Company, 139-159. Ogilvy, David (1983), Confessions of an Advertising Man, New York: Dell. Onkvisit, Sak and John Shaw (1987), "Self-Concept and Image Congruence: Some Research and Managerial Implications," Journal of Consumer Marketing, 4(Winter), 13-24. Park, C. Whan, Bernard J. Jaworski, and Deborah J. Maclnnis (1986), "Strategic Brand Concept-Image Management," Journal of Marketing, 50(October), 135-145. Perry, Michael and B. Curtis Hamm (1969), "Canonical Analysis of Relations Between Socioeconomic Risk and Personal Influences in Purchase Decisions," Journal of Marketing Research, 6 (August), 351-354. Petty, Richard E. and John T. Cacioppo (1986), "The Elaboration Likelihood Model of Persuasion," in Advances in Experimental Social Psychology, Vol . 19, ed. L. Berkowitz. New York: Academic Press, 123-205. Plummer, Joseph T. (1985), "How Personality Makes a Difference," Journal of Advertising Research, 24(December), 27-31. Pomazal, Richard and James Jaccard (1976), "An Informational Approach to Altruistic Behavior," Journal of Personality and Social Psychology, 33, 317-326. Prentice, Deborah A. (1987), "Psychological Correspondence of Possessions, Attitudes and Values," Journal of Personality and Social Psychology, 53(6), 993-1003. Price, Linda and Lawrence Feick (1984), "The Role of Interpersonal Sources in External Search: An Informational Perspective," Advances in Consumer Research, 10, 250-255. 119 Pryor, John B. (1980), "Self-Reports and Behaviour," in The Self in Social Psychology, eds. Daniel M . Wegner and Robin R. Vallacher, New York: Oxford University Press. Pyszczynski, Tom and Jeff Greenberg (1987), "Toward an Integration of Cognitive and Motivational Perspectives on Social Inference: A Biased Hypothesis-Testing Model," Advances in Experimental Social Psychology, Vol . 20, ed. L . Berkowitz, New York: Academic Press, 297-340. Reingen, Peter H . and Jerome B. Kernan (1986), "Analysis of Referral Networks in Marketing: Methods and Illustration," Journal of Marketing Research, 23(November), 370-378. Richins, Marsha L. (1983), "Negative Word-of-Mouth by Dissatisfied Consumers: A Pilot Study," Journal of Marketing, 47(Winter), 68-78. (1984), "Word of Mouth Communication as Negative Information," in Advances in Consumer Research, ed. T. Kinnear, Provo, Utah: Association for Consumer Research, 697-702. (1994), "Special Possessions and the Expression of Material Values," Journal of Consumer Research, 21(December), 522-533. Robertson, Thomas (1971), Innovative Behavior and Communication, New York: Holt, Rinehart & Winston. Roselius, Ted (1971), "Consumer Rankings of Risk Reduction Methods," Journal of Marketing, 35 (January), 56-61. Ross, Ivan (1971), "Self-Concept and Brand Preference," Journal of Business, 44, 38-50. Sample, John and Rex Warland (1973), "Attitude and Prediction of Behavior," Social Forces, 51,292-304. Sartre, Jean P. (1956), Being and Nothingness, Philosophical Library, New York. Schachter, Stanley (1959), The Psychology of Affiliation, Stanford, C A : Stanford University Press. Schindler, Robert M . (1998), "Consequences of Perceiving Oneself as Responsible for Obtaining a Discount: Evidence for Smart-Shopper Feelings," Journal of Consumer Psychology, 7(4), 371-392. 120 Schlenker, Barry R. (1975), "Self-Presentation: Managing the Impression of Consistency when Reality Interferes with Self-Enhancement," Journal of Personality and Social Psychology, 32, 1030-1037. (1980), Impression Management: The Self Concept, Social Identity, and Interpersonal Relations, Monterey, C A : Brooks/Cole. (1985), "Identity and Self-Identification," in The Self and Social Life, ed. B. R. Schlenker, New York: McGraw-Hill, 65-99. (1986), "Self-Identification: Toward an Integration of the Private and Public Self," in Public Self and Private Self ed. R. Baumeister, New York: Springer-Verlag, 21-62. and Mark Leary (1982), "Audiences' Reactions to Self-Enhancing, Self-Denigrating, and Accurate Self-Presentations," Journal of Experimental Social Psychology, 18, 89-104. . s Rowland S. Miller, and.Mark R. Leary (1983), "Self-Presentation as a Function of Self-Monitoring and the Validity and Quality of Past Performance," Representative Research in Social Psychology, 13(1), 2-14. Schneider, David J. (1969), Tactical Self-Presentation after Success and Failure," Journal of Personality and Social Psychology, 13, 262-268. (1981), "Tactical Self-Presentations: Toward a Broader Conception," in Impression Management Theory and Social Psychological Research, ed. J. T. Tedeschi, New York: Academic Press, 23-40. Sengupta, Jaideep, Darren W. Dahl, and Gerald J. Gorn (forthcoming), "Misrepresentation in the Consumer Context," Journal of Consumer Psychology. Shavitt, Sharon (1990), "The Role of Attitude Objects in Attitude Functions," Journal of Experimental Social Psychology, 26(2), 124-148. Sherif, Carolyn W., Merrilea Kelly, H. Lewis Rodgers Jr., Gian Sarup, and Bennett I. Tittler (1973), "Personal Involvement, Social Judgement, and Action," Journal of Personality and Social Psychology, 27(3), 311-328. Sheth, Jagdish (1971), "Word-of-Mouth in Low-Risk Innovations," Journal of Advertising Research, 11, 15-18. 121 Shimp, Terence A. and Alican Kavas (1984), "The Theory of Reasoned Action Applied to Coupon Usage," Journal of Consumer Research, 11 (December), 795-809. Showers, Carolin (1992), "Compartmentalization of Postive and Negative Self-Knowledge: Keeping Bad Apples Out of the Bunch," Journal of Personality and Social Psychology, 62(6), 1036-1049. Singh, Jagdip (1988), "Consumer Complaint Intentions and Behavior: A Review and Prospect," Journal of Marketing, 52(January), 93-107. Silverman, George (1997), "How to Harness the Awesome Power of Word of Mouth," Direct Marketing, 60(7), 32-37'. Sirgy, M . Joseph (1982), "Self-Concept in Consumer Behavior: A Critical Review," Journal of Consumer Research, 9(December), 287-300. - and Chenting Su (2000), "Destination Image, Self-Congruity, and Travel Behaviour: Toward an Integrative Model," Journal of Travel Research, 38(4), 340-352. Smith, Robert E. and Tiffany S. Meyer (1980), "Attorney Advertising: A Consumer Perspective," Journal of Marketing, 44(2), 56-64. Snyder, C. R., Raymond L. Higgins, and Rita J. Stucky (1983), Excuses: Masquerades in Search of Grace, New York: Wiley. Snyder, Mark and Steve Gangestad (1986), "On the Nature of Self-Monitoring: Matters of Assessment, Matters of Validity," Journal of Personality and Social Psychology, 51(1), 125-139. and D . Kendzierski (1982), "Choosing Social Situations: Investigating the Origins of Correspondence between Attitudes and Behavior," Journal of Personality, 50, 280-295. Solomon, Michael R. (1996), Consumer Behavior, 3th Edition, Prentice Hall, Upper Saddle River, New Jersey, Sorrentino, Richard M . and Edward T. Higgins (1986), "Motivation and Cognition: Warming Up to Synergism," Variability in Social Behavior: The Ontario Symposium, Vol . 2, eds. M . P. Zanna, E. T. Higgins, and C. P. Herman, Hillsdale, NJ: Erlbaum, 105-130. 122 Stafford, Laura and John A. Daly (1984), "The Effects of Recall Mode and Memory Expectancies on Remembrances of Natural Conversations," Human Communication Research, lO(Spring), 379-402. Stires, Lloyd D. and Edward E. Jones (1969), "Modesty vs. Self-Enhancement as Alternative Forms of Ingratiation," Journal of Experimental Social Psychology, 5, 172-188. Stryker, Sheldon (1968), "Identity Theory: Developments and Extensions," in Self and Identity, eds. K. Yardley and T. Honess, New York: John Wiley & Sons Ltd., 89-104. Swan, John E., & Oliver, Richard L. (1989), "Postpurchase Communications by Consumers," Journal of Retailing, 65(4), 516-533. TARP (1979), "Measuring the Grapevine: Consumer Response and Word-of-Mouth," Atlanta, GA: The Coca-Cola Co. Taylor, Shelley E. and Jonathan D. Brown (1988), "Illusion and Well-Being: A Social Psychological Perspective on Mental Health," Psychological Bulletin, 103, 193-210. Tedeschi, James T. (1981), Impression Management Theory and Social Psychological Research, New York: Academic Press. (1986), "Private and Public Experiences and the Self," in Public Self and Private Self, ed. R. F. Baumeister, New York: Springer-Verlag, 1-20. Tesser, Abraham (1988), "Toward a Self-Evaluation Maintenance Model of Social Behavior," Advances in Experimental Social Psychology, 21, 181-227. a n ( j Jennifer Campbell (1982), "Self-Evaluation Maintenance and the Perception of Friends and Strangers," Journal of Personality, 59, 261-279. and Jonathan Smith (1980), "Some Effects of Friendship and Task Relevance on Helping: You Don't Always Help the One You Like," Journal of Experimental Social Psychology, 16, 582-590. Tetlock, Philip E. (1983), "Accountability and the Perseverance of First Impressions," Social Psychology Quarterly, 46, 285-292. (1985), "Accountability: A Social Check on the Fundamental Attribution Error," Social Psychology Quarterly, 48, 227-236. 123 ? and j , i . K im (1987), "Accountability and Judgement Processes in a Personality Prediction Task," Journal of Personality and Social Psychology, 52, 700-709. Thomas, Gloria Perm (1992), "The Influence of Processing Conversational Information on Inference, Argument Elaboration, and Memory," Journal of Consumer Research, 19(June), 83-92. Thorelli, Hans B. (1971), "Concentration of Information Power Among Consumer," Journal of Marketing Research, 8(November), 427-432. Tice, Dianne M . (1991), "Esteem Protection or Enhancement? Self-Handicapping Motives and Attributions Differ by Trait Self-Esteem," Journal of Personality and Social Psychology, 60, 711 -725. (1993), "The Social Motivations of People with Low Self-Esteem," in Self-Esteem: The Puzzle of Low Self-Regard, ed. R. Baumeister, New York: Plenum, 37-53. and Roy F. Baumeister (1990), "Self-Esteem, Self-Handicapping, and Self-Presentation: The Strategy of Inadequate Practice," Journal of Personality, 58, 443-464. 5 Jennifer L . Butler, Mark B. Muraven, and Arlene M . Stillwell (1995), "When Modesty Prevails: Differential Favorabiility of Self-Presentation to Friends and Stranger," Journal of Personality and Social Psychology, 69(6), 1120-1138. Tittle, Charles R. and Richard J. Hi l l (1967), "The Accuracy of Self-Reported Data and Prediction of Political Activity," Public Opinion Quarterly, 31, 103-106. Triandis, Harry C. (1989), "The Self and Social Behavior in Differing Cultural Contexts," Psychological Review, 96(3), 506-520. Veblen, Thorstein (1899), The Theory of the Leisure Class, New York: The MacMillan Co. Von Baeyer, Carl L. , Debbie L. Sherk, and Mark P. Zanna (1981), "Impression Management in the Job Interview: When the Female Applicant Meets the Male (Chauvinist) Interviewer," Personality and Social Pyschology Bulletin, 7, 45-51. Walker, Charles J. and Carol A. Beckerle (1987), "The Effect of State Anxiety on Rumor Transmission," Journal of Social Behavior and Personality, 2(3), 353-360. 124 Wallendorf, Melanie and Eric J. Arnould (1988), ' " M y Favorite Things': A Cross-Cultural Inquiry into Object Attachment, Possessiveness, and Social Linkage," Journal of Consumer Research, 14(March), 531-547. Weinstein, Neil D. (1980), "Unrealistic Optimism about Future Life Events," Journal of Personality and Social Psychology, 39, 806-820. (1982), "Unrealistic Optimism about Susceptibility to Health Problems," Journal of Behavioral Medicine, 5, 441-460. Westbrook, Robert A. (1987), "Product/Consumption-Based Affective Responses and Postpurchase Processes," Journal of Marketing Research, 24( August), 258-270. Wicker, Allan W. (1969), "Attitudes versus Actions: The Relationship of Verbal and Overt Behavioral Responses to Attitude Objects," Journal of Social Issues, 25, 41-78. Wicklund, Robert A. (1975), "Objective Self-Awareness," in Advances in Experimental Social Psychology, 8, ed. L . Berkowitz, New York: Academic Press, 233-275. and Shelley Duval (1971), "Opinion Change and Performance Facilitation as a Result of Objective Self-Awareness," Journal of Experimental Social Psychology, 7, 319-342. — and Peter M . Gollwitzer (1982), Symbolic Self Completion, Hillsdale, NJ: Erlbaum. Wilke, W. H. (1934), "An Experimental Comparison of the Speech, the Radio and the Printed Page as Propaganda Devices," Archives of Psychology, Columbia University, No. 169: 32. Zanna, Mark P. and Susan J. Pack (1975), "On the Self-Fulfilling Nature of Apparent Sex Differences in Behavior," Journal of Experimental Social Psychology, 11, 583-591. Zajonc, Robert B. (1968), "Attitudinal Effects of Mere Exposure," monograph, Journal of Personality and Social Psychology, 8, 1-29. Zuckerman, Miron (1979), "Attribution of Success and Failure Revisited: The Motivational Bias is Alive and Well in Attribution Theory, " Journal of Personality, 47(2), 245-287. 125 Appendix 3-1 Questionnaire for Pilot Study S E C T I O N 1: This section asks you questions on a specific product/service. 1. Think of a product or service that you have P U R C H A S E D and USED. Also, this should be a product/service that you believe tells other people what kind of a person you actually are. What is this product/service? (Please be specific in answering this question. E.g. Yoplait Set-Style Strawberry Yogurt rather than simply yogurt) 2. Imagine talking about this product/service with another student who knows a lot about it. In other words, this person is knowledgeable enough to judge your opinion. What exactly will you tell this person? 3. Rate how positive or negative the information you gave above is about the product/service? Mainly Mixed/ Mainly Negative Neutral Positive -3 -2 -1 0 +1 +2 +3 126 Appendix 3-1 (Continued) 4. Rate each of the items below from l=not at all to 7=very much so. In talking about this product/service, to what extent were you trying to be: (a) accurate (b) complete (c) precise (d) smart (e) with-it (f) looking good 5. Rate your overall familiarity with this product/service (Circle One): (a) Completely unfamiliar (b) Unfamiliar (c) neither familiar nor unfamiliar (d) Familiar (e) Completely familiar 6. What is the approximate price of this product/service? 127 Appendix 4-1 Initial Survey for Study 1 In what follows, we would like you to provide us with 4 different products according to the various conditions stated. In each case, you should have a reasonable to positive attitude toward the product, and also be knowledgeable enough to give recommendations on the product if somebody asked you for advice. Finally, please be specific in stating the products A N D their brand names. Think of a product that you own or have owned. Also, this should be a product that you believe tells other people what kind of a person you actually are. a. What is this product (and its brand)? b. Rate your attitude toward this product. Negative -3 -2 Mixed/ Neutral -1 0 +1 Positive +2 +3 Think of a product that you have never owned. Also, this should be a product that you believe tells other people what kind of a person you actually are. a. What is this product (and its brand)? b. Rate your attitude toward this product. Positive +1 +2 +3 Negative -3 -2 Mixed/ Neutral -1 0 Think of a product that you own or have owned solely for practical reasons. Make sure that this is not a product that would tell other people anything about what kind of a person you actually are. a. What is this product (and its brand)? b. Rate your attitude toward this product. Negative -3 -2 Mixed/ Neutral -1 0 +1 +2 Positive +3 Think of a product that you have never owned. Also, this should be a product that would be used solely for practical reasons. Make sure that this is not a product/service that would tell other people what kind of a person you actually are. a. What is this product (and its brand)? b. Rate your attitude toward this product. Negative -3 -2 Mixed/ Neutral -1 0 +1 Positive +2 +3 128 Appendix 4-2 Final Questionnaire for Study 1 Subject:_ We would like to ask your opinion on About what brand would you feel comfortable giving someone else advice? 1. Rate how positive or negative your advice about the product would be? Mainly Negative -3 -2 -1 Mixed/ Neutral 0 +1 +2 Mainly Positive +3 2. How important would it be to be accurate, complete, and precise when giving advice about a product like this (l=not at all to 7=very much so)? (b) accurate (b) complete (c) precise 3. To what extent do you think the advice you give will make you feel smart, "with-it," and look good (l=not at all to 7=very much so)? (a) smart (b) "with-it" (c) look good 4. Have you ever owned this product? Yes or No 5. How familiar are you with this product? Not At All -3 -2 -1 ( +1 +2 Very Much +3 6. What is your attitude toward this product? Mainly Negative -3 -1 Mixed/ Neutral 0 +1 Mainly Positive +2 +3 7. Would this product tell other people something important about the kind of person you are? Yes or No THANK YOU FOR YOUR HELP WITH THIS STUDY! 129 Appendix 5-1 Attitude Measure for Study 2 Subject: Rate your attitude toward Fenton & Lee Snack Mixes by circling a number for each of the following scales. a ) B a d 1 2 3 4 5 6 7 G o o d b ) N e g a t i v e 1 2 3 4 5 6 7 P o s i t i v e c ) U n f a v o r a b l e 1 2 3 4 5 6 7 F a v o r a b l e d ) D i s l i k e d 1 2 3 4 5 6 7 L i k e d 130 Appendix 5-2 Ads Used in Study 2 AD WITH LINK Meat Food and Ormtt Dairy Products . Condiments international Foods Snack Foods Baked Goods Sweets Drinks Seafood Prepared Meals Specialty Gift Baskets and Samptcfs Food and Wine Books Pasta and Noodles Herbs and Specs Vegetarian Your Personal Expression of Good Taste and Discrete Choice Mouthwatering mixes and party snacks include our House nuts, Matsui rice crackers, European gummies, Jazzy sours, and many other munchies. Ideal for social gathering, or just for snacking at home. For the Savvy Connoisseur Click for larger Image Item #MLGF016 Quantity:! Buy Now | Place in Cart Add to Shopping List * The word "Discrete" in the advertising copy was intended to mean "clearly distinct from other alternatives." However, it is conceivable that subjects had interpreted it as discreet, or being prudent and reserved. This interpretation might have introduced some ambiguity to the advertising copy. However, no subject had raised questions regarding this word during the experiment, or expressed any suspicion in the suspicion probe in the final questionnaire. Moreover, the claims that consuming the snack mix showed "Good Taste" and "Savvy Connoisseurship" were the focus of the ad copy, conveying the thrust of the experimental manipulation. 131 Appendix 5-2 (Continued) A D W I T H O U T L I N K Meal Food and Drink Dairy Products Condtfncnts International Foods Snack Foods Baked Goods Sweets Drinks Seafood Prepared Meals SpociaAy Gift Baskets and Samplers Food and Wine Books Pas la and Noodles Herbs and Specs Vegetarian Jenton <J!ee Stxaclz WjlxeA Mouthwatering mixes and party snacks include our House nuts, Matsui rice crackers, European gummies, Jazzy sours, and many other munchies. Ideal for social gathering, or just for snacking at home. Click for larger Image Item #MLGF016 Quantity:! Buy Now | Place in Cart Add to Shopping List Appendix 5-3 Final Questionnaire for Study 2 Did you see any of the following phrases in the ad for the snack mixes? (PLEASE CIRCLE YES OR NO FOR EACH OF THE FOLLOWING) a) Your Personal Expression of Good Taste and Discrete Choice Yes or No b) For the Savvy Connoisseur Yes or No c) The care we give to both the art and science of snacks preparation results in products of superior quality. Yes or No d) Ideal for social gathering, or just for snacking at home. Yes or No 133 Appendix 5-3 (Continued) Subject: Section 1: This section contains questions that directly relate to the materials and products you read about and tried. Please circle a number for each of the following scales. 1. Rate your attitude toward Fenton & Lee Snack Mixes for each of the following: a) Bad 1 2 3 4 5 6 7 Good b) Negative 1 2 3 4 5 6 7 Positive c) Unfavorable 1 2 3 4 5 6 7 Favorable d) Disliked 1 2 3 4 5 6 7 Liked 2. How knowledgeable were you about Fenton & Lee Snack Mixes before participating in this study? Not At A l l - 3 - 2 - 1 0 1 2 3 Very 3. How often do you eat snack mixes? a) Every day b) About once a week c) About once a month d) Several times a year e) Never 134 Appendix 5-3 (Continued) Section 2: This section contains items that deal with whether you share information about snacks with your friends. 1. In general, do you like to talk about snacks with your friends? Y or N 2. Would you say you give: (a) Very little information about snacks to your friends? (b) An average amount of information about snacks to your friends? (c) A great deal of information about snacks to your friends? 3. During the past six months, have you told anyone about some snacks? Y or N 4. Compared with your circle of friends, are you (a) Less likely to be asked for advice about snacks? (b) About as likely to be asked for advice about snacks? (c) More likely to be asked for advice about snacks? 5. If you and your friends were to discuss snacks, would you (a) Mainly listen to your friends' ideas? (b) Try to convince them of your ideas? 6. Which of the following happens more often? (a) You tell your friends about some snacks? (b) They tell you about some snacks? 7. Do you have the feeling that you are generally regarded by your friends and neighbours as a good source of advice about snacks? Y or N 135 Appendix 5-3 (Continued) Section 3: This section contains items regarding products in general. Strongly Strongly Disagree Agree 1.1 like introducing new brands and products to my friends. 1 2 3 4 5 6 7 2.1 like helping people by providing them with information 1 2 3 4 5 6 7 about many kinds of products. 3. People ask me for information about products, places to 1 2 3 4 5 6 7 shop, or sales. 4. If someone asked where to get the best buy on several types 1 2 3 4 5 6 7 of products, I could tell him or her where to shop. 5. M y friends think of me as a good source of information 1 2 3 4 5 6 7 when it comes to new products or sales. 6. Think about a person who has information about a variety 1 2 3 4 5 6 7 of products and likes to share this information with others. This person knows about new products, sales, stores, and so on, but does not necessarily feel he or she is an expert on one particular product. Would you say that this description fits you? 136 Appendix 5-3 (continued) Section 4: This section contains some final items on which we would like your opinion. 1. 3. B a s e d o n t h e a d , a r e p e o p l e w h o c h o o s e F e n t o n & L e e S n a c k M i x e s : Not At All Very a ) S a w y ? - 3 - 2 - 1 0 1 2 3 b ) S m a r t ? - 3 - 2 - 1 0 1 2 3 c ) D i s c r e t e ? - 3 - 2 - 1 0 1 2 3 d ) G o o d ? - 3 - 2 - 1 0 1 2 3 I n y o u r o p i n i o n . a r e p e o p l e Not At All w h o c h o o s e F e n t o n & L e e S n a c k M i x e s : Very a ) S a v v y ? - 3 - 2 -1 0 1 2 3 b ) S m a r t ? - 3 - 2 - 1 0 1 2 3 c ) D i s c r e t e ? - 3 - 2 - 1 0 1 2 3 d ) G o o d ? - 3 - 2 -1 0 1 2 3 W h e n v o u w e r e g i v i n g o p i n i o n s a b o u t t h e s n a c k m i x e s , w e r e y o u c o n c e r n e d a b o u t l o o k i n g : Not At All Very a ) S a v v y ? - 3 - 2 - 1 0 1 2 3 b ) S m a r t ? - 3 - 2 - 1 0 1 2 3 c ) D i s c r e t e ? - 3 - 2 -1 0 1 2 3 d ) G o o d ? - 3 - 2 - 1 0 1 2 3 I n y o u r o w n w o r d s , d e s c r i b e t h e p u r p o s e o f recording y o u r o p i n i o n : 5 . I n y o u r o w n w o r d s , d e s c r i b e t h e p u r p o s e o f t h e s t u d y as a whole: THANK YOU FOR YOUR PARTICIPATION I 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0090600/manifest

Comment

Related Items