ONLINE CONSUMER REVIEWS:
The role of self-‐enhancing vs. self-‐effacing motives and
type of product in word-‐of-‐mouth transmission
Laura van de Stroet
MSc Business Administration, Marketing Management
MSc International Business & Management
University of Groningen
Faculty of Economics and Business
October 2013
ONLINE CONSUMER REVIEWS:
The role of self-‐enhancing vs. self-‐effacing motives and
type of product in word-‐of-‐mouth transmission
Author: Laura van de Stroet S2123819 Rozenstraat 175-‐1 1016NR Amsterdam lauravandestroet@gmail.com +31(0) 6-‐50 74 33 13
1st supervisor: Dr. J.A. Voerman
2nd supervisor: Mr. Drs. H.A. Ritsema
University: University of Groningen
Faculty of Economics and Business
Master Thesis MSc Business Administration, Marketing Management
Master Thesis MSc International Business & Management
ABSTRACT
The purpose of this research is to examine to what extent WOM valence and type of product moderate the probability that eWOM in online consumer reviews (OCR’s) will be transmitted (i.e., experiences that occurred to others). Furthermore, the moderating effects of Western individuals with a self-‐enhancement motive and Eastern individuals with a self-‐effacing motive are tested.
The effects were tested using a 3 (WOM valence: positive/neutral/negative) x 2 (type of product: self-‐relevant/utilitarian) experimental design in which respondents were asked to indicate their eWOM transmission probability after reading an OCR. The results indicate that NWOM has a significantly higher transmission probability than neutral eWOM. Contrary to expectations, both Western and Eastern respondents display self-‐enhancing tendencies. There was no significant effect of a self-‐enhancement vs. self-‐effacement motive on the eWOM transmission probability. Easterners are significantly more likely than Westerners to transmit eWOM, and respondents become more likely to transmit eWOM on utilitarian products when they get older.
Keywords
: Online consumer reviews, word-‐of-‐mouth transmission, word-‐of-‐mouth valence, type of product, self-‐enhancement motive, Eastern vs. Western culture.TABLE OF CONTENTS
CHAPTER 1: INTRODUCTION ... 1
1.1 Why engage in eWOM? ... 4
1.2 Research Question ... 6
1.3 Theoretical and Managerial Relevance ... 7
1.4 Structure of the thesis ... 8
CHAPTER 2: THEORETICAL FRAMEWORK ... 9
2.1 Dependent variable: (e)WOM transmission ... 9
2.2 Independent variable: WOM valence ... 10
2.3 Independent variable: Utilitarian vs. self-‐relevant products ... 13
2.4 Moderator: Self-‐enhancement vs. self-‐effacement ... 14
2.5 Control variables ... 18
2.6 Overview of Hypotheses and Conceptual Framework ... 20
CHAPTER 3: RESEARCH DESIGN ... 23
3.1 Design and participants ... 23
3.2 Procedure ... 24
3.3 Operationalization overview ... 27
3.4 Sample characteristics ... 36
3.5 Plan of analysis ... 37
CHAPTER 4: RESULTS ... 39
4.1 Results for WOM transmission probability ... 39
4.2 Homogeneity of slopes ... 40
4.3 ANCOVA ... 40
4.4 Additional Analyses ... 44
4.5 Summary of Results ... 50
CHAPTER 5: CONCLUSION ... 53
5.1 Discussion of Findings ... 53
5.2 Managerial Implications ... 55
5.3 Limitations and Future Research Directions ... 56
REFERENCES ... 59
APPENDICES
... 66CHAPTER 1: INTRODUCTION
“Advertising cannot replace interpersonal influence” (Dichter, 1966).
Back in 1955, in their book on personal influence, Katz and Lazarsfeld (1955) described that when it comes to influencing consumers to switch brands, word-‐of-‐mouth communication (hereafter WOM) was seven times as effective as newspapers and magazines, four times as effective as personal selling, and twice as effective as radio advertising. WOM communication is commonly defined as ‘the flow of informal communications among consumers about the usage of products and services concerning evaluations of goods and services’ (Westbrook, 1987). It is often the dominant factor in consumer choice, because WOM lacks commercial bias, is swift, and interactive (East, Hammond, & Wright, 2007). Without a doubt, WOM is one of the most influential sources of information transfer by consumers (Khare, Labrecque, & Asare, 2011), and according to McKinsey and Bain & Company, WOM is currently the greatest force driving sales (de Angelis et al., 2012), influencing nearly 70 percent of all buying decisions (Balter, 2008).
from the online availability of other consumers’ evaluations (Chatterjee, 2001). In eWOM, several pieces of positive and negative information from multiple sources are presented simultaneously whereas traditional WOM is only able to communicate a single piece of information that either has a positive or negative valence (Chatterjee, 2001). For businesses, eWOM offers an incredible source of insight to really understand how satisfied their customers are on a real-‐time basis. eWOM is visible and measurable in a cost-‐effective way, whereas previously, information exchanged in private conversations was difficult to observe (Godes & Mayzlin, 2004). eWOM is more influential than traditional, ‘offline’ WOM because of its greater accessibility, higher reach and its exceptionally high diffusion speed (Chatterjee, 2001). Finally, eWOM may enable consumers to obtain higher quality input into a decision, since the Internet provides them access to people with greater expertise on a topic (Constant, Sproull, & Kiesler, 1997). eWOM occurs on various online channels, including blogs, emails, news groups, boycott Web sites, consumer review websites, discussion forums, virtual consumer communities, and social networking sites (e.g. Shu-‐Chuan & Yoojung, 2011; Hennig-‐Thurau et al., 2004). It has several characteristics that can be studied, under which volume, valence, duration, intensity, dispersion and variance (Liu, 2006). In this study, focus will be on valence. eWOM can have a positive, a neutral, or a negative valence, with positive WOM (hereafter PWOM) encouraging brand choice and negative WOM (hereafter NWOM) discouraging brand choice (East, Hammond, & Lomax, 2008).
In 1995, Amazon.com was the first website to offer consumers the option to post their
comments (reviews and ratings) on the products sold through its website. Since then, numerous other sellers in many product categories, under which books, electronics, games, videos, music, beverages, and wine, have adopted similar strategies, making personal product evaluations on sellers’ websites very common nowadays (Yubo & Jinhong, 2008). Platforms using reviews and ratings of products or retailers are nowadays the most prevalent and accessible form of eWOM. Online consumer reviews (hereafter OCR’s) can be defined as ‘peer-‐generated product
that OCR’s have a significant impact on e.g. book sales (Chevalier & Mayzlin, 2006), box office revenues (Liu, 2006) and toiletry products (Moe & Trusov, 2011).
With the emergence of eWOM communicated via OCR’s, a new product information channel was born. OCR’s are a key form of user-‐generated content (Sridhar & Srinivasan, 2012): An increasing amount of consumers is contributing their opinions online, as well as the amount of (potential) consumers relying on the information provided in these OCR’s (Moe & Trusov, 2011). Compared to marketer-‐generated online information, user-‐generated content is considered more relevant, credible, and interest-‐ and empathy-‐generating (Bickart & Schindler, 2001; Schindler & Bickart, 2005. OCR’s and other social sources are a tremendous resource for consumers: 37 percent of shoppers see online social sources as an influential driver when making purchase decisions (up from 19 percent in 2010) (Lecinski, 2012), and 70 percent of Americans look at product reviews before making a purchase1. OCR’s are considered a new element in the marketing communication mix that can work as free ‘sales assistants’ to online retailers because they help consumers to identify the products that best match their
idiosyncratic usage conditions (Chen & Xie, 2008). Allowing consumers the possibility of writing OCR’s on a website improves their perception of both the usefulness and social presence of the website (Kumar and Benbasat, 2006). OCR’s also have the possibility to attract consumer visits, to increase ‘stickiness’ (time spent on a website), and to create a sense of community among frequent shoppers (Mudambi & Schuff, 2010).
In this thesis, eWOM transmission probabilities will be studied. WOM transmissions refer to ‘situations in which consumers pass on information about experiences with products and services they have heard occurred to others’ (De Angelis et al., 2012). Here, it is the probability that an OCR written by an individual (the ‘host’) will be further spread by the reader (‘the receiver’) of that OCR. Several factors that play a role in the probability that eWOM will be transmitted will be identified, including for example product type. Chung and Darke (2006) make a distinction between utilitarian products, which are products serving a functional need, and self-‐relevant products, which are products that are serving as a means of self-‐presentation. They find that product type leads to differences in both the amount and valence of WOM transmitted between Eastern and Western cultures, due to their cultural differences of
individualism vs. collectivism2. This individualism vs. collectivism dimension is important when it comes to understanding cross-‐cultural differences in consumer behaviour (e.g. Triandis et al., 1988). Eastern cultures are considered interdependent and collectivistic whereas Western cultures are considered independent and individualistic (Sedikides et al., 2003). Besides investigating the role of product type, the role of a single psychological motive (namely a self-‐ enhancing vs. a self-‐effacing motive) as a more specific cultural difference between Eastern and Western cultures and its impact on eWOM transmission probability will be studied. Before going into detail on this motive, general motives to engage in WOM will be discussed.
1.1 Why engage in eWOM?
People find it important to find out about other consumers’ experiences before purchasing a good or service. Consumers can use eWOM to reinforce decisions or to increase confidence in the views they already hold (Schindler & Bickart, 2005). eWOM is mainly read to make better buying decisions, to save decision-‐making time, and to serve as a risk-‐reduction strategy: looking at other people’s comments can reduce or eliminate the uncomfortable feeling of risk exposure (Hennig-‐Thurau & Walsh, 2004; Schindler & Bickart, 2005; Buttle, 1998).
Literature defines many motives for consumers to engage in (e)WOM behaviour, most of them building on Dichter’s (1966) seminal framework that included product involvement, other involvement, message involvement, and self-‐involvement. Product involvement occurs because experience with the product (or service) produces a tension that is not eased by simple use of the product. A consumer feels so strongly (either positively or negatively) about a product that he or she builds up pressure and wants to talk about it (Dichter, 1966). In cases where eWOM engagement is driven by self-‐involvement motives, the focus is more on the person than on the product, in which the product is serving as a means through which the consumer can gratify certain emotional needs. So, the speaker has a need to enhance him-‐ or herself in front of his or her audience, which allows the speaker to gain attention, show connoisseurship, suggest status, give the impression of possessing inside information, and assert superiority (Dellarocas et al., 2010). This potentially leads to enhanced self-‐worth of the consumer engaging in (e)WOM
(Hennig-‐Thurau et al., 2004). When a consumer feels a genuine need to help others make a better purchase decision, he or she is said to have concern for other consumers or to engage in altruism. When engaging in (e)WOM because of message involvement, consumers’ discussions are mainly stimulated by the way the product is presented through advertisements,
commercials, or public relations (Dichter, 1966). Later studies slightly refined Dichter’s (1966) initial motivation categories, adding e.g. a social benefits motive (defined as enjoyment from engaging in the social experience of eWOM), an economic incentive motive (incentives offered directly by websites to consumers that post OCR’s) (Hennig-‐Thurau et al., 2004), a motive to exert collective power over companies (Hennig-‐Thurau & Hansen, 2001), and an entertainment motive (Schindler & Bickart, 2005).
Research has produced inconsistent results to the question if consumers are more likely to engage in positive or negative WOM. Some motives to engage in positive WOM (hereafter PWOM) or negative WOM (hereafter NWOM) differ. Drivers of PWOM include psychological motives such as a desire to help the company (Hennig-‐Thurau et al., 2004; Sundaram, Mitra, & Webster, 1998), a desire to signal expertise to others, altruism, and product involvement
(Dichter, 1966). Cheung & Lee (2012) find three significant antecedents of consumers’ intentions to engage in PWOM: reputation, a sense of belonging, and a sense of altruism. People often obtain intrinsic personal enjoyment and satisfaction by helping others through sharing their knowledge, even without an apparent compensation (Wasko & Faraj, 2000). Consumers are driven to engage in NWOM because of psychological motives including (again) altruism, a need to retaliate against the company for a negative experience (e.g. Sundaram et al., 1998), a desire to resolve cognitive dissonance (Engel, Blackwell, & Miniard, 1993), and a desire to seek advice on how to deal with a negative experience (e.g. Hennig-‐Thurau et al., 2004; Sundaram et al., 1998).
In order to resolve the conflicting findings regarding the prevalence of PWOM vs. NWOM, De Angelis et al. (2012) suggest that a single psychological motive, namely a self-‐enhancement motive, leads to systematic differences in when consumers are more likely to engage in PWOM rather than NWOM, and vice versa. Wojnicki and Godes (2008) first introduced self-‐
Baumeister, 1998). De Angelis et al. (2012) make a distinction between two stages of WOM communication: 1) WOM transmission (consumers passing on information about experiences they heard occurred to others) and 2) WOM generation, in which consumers share information about their own experiences. They find that in Western cultures, people tend to generate positive WOM, but to transmit negative WOM. So, they propose that people like to share their own positive experiences, but are more inclined to share negative experiences they have heard occurred to others, simply because individuals have the propensity to reinforce or magnify the positive. However, there is a debate in research on the universal applicability of the self-‐ enhancement motive, with some researchers stating that Eastern cultures might show self-‐ effacing tendencies, in which they do not necessarily try to bolster or improve the self-‐concept (Heine et al., 1999). Thus, differences between Eastern and Western cultures in self-‐effacing vs. self-‐enhancing tendencies might cause differences in when consumers are more likely to engage in PWOM rather than NWOM, and vice versa.
1.2 Research Question
Overall, OCR’s have become an important information source for consumers’ judgments (Chatterjee, 2001) and for consumer purchase decisions (Chen & Xie, 2008). It is obvious that OCR’s are gaining in both importance and popularity, explaining why this relatively new phenomenon has been receiving growing attention in the marketing literature (e.g. Hennig-‐ Thurau et al., 2004). However, there are many researchers who call for more academic research in the field of OCR’s (e.g. Yubo & Jinhong, 2008). To my knowledge, hardly any research has taken into account the possible moderating effects of cross-‐country cultural differences in WOM communications. In this research, I will investigate how self-‐effacing vs. self-‐enhancing
tendencies (as a single psychological, cultural motive) moderate the probability of WOM transmission when WOM valence is positive, negative, and neutral. I will also investigate how product type (utilitarian vs. self-‐relevant) moderates the probability of WOM transmission via OCR’s between Western individuals with a self-‐enhancement motive and Eastern individuals with a self-‐effacing motive. Therefore, the research question of this thesis will be:
‘How do self-‐effacing vs. self-‐enhancing tendencies and the type of product moderate the
1.3 Theoretical and Managerial Relevance
1.3.1 Theoretical Relevance
Yubo & Jinhong (2008) call for more academic research on OCR’s, because the new information channel plays a fundamental role in the marketplace and has many strategic implications to marketers regarding consumer review information. Culture has an impact on consumers’ WOM behaviour and therefore, the universal applicability of WOM research conducted in Western economies was already questioned back in 1998 (Buttle, 1998). In particular, Sridhar and Srinivasan (2012) call for a study that investigates cross-‐country differences in the evolution of OCR’s. Most research has focused on the aggregate outcomes of WOM (e.g. sales), but the process of what drives or motivates individuals to engage in (e)WOM transmission has barely been studied yet (Stephen & Lehmann, 2009). It is interesting to study these individual
processes because new technologies make it easier for consumers to share product-‐ and brand-‐ related information (e.g. Godes & Mayzlin, 2004). De Angelis et al. (2012) do not consider cultural differences in their study on the role of a self-‐enhancement motive in WOM transmission vs. WOM generation. According to the authors themselves, accounting for important cultural differences (e.g. self-‐effacement vs. self-‐enhancement tendencies) might produce less pronounced or even contradictory results.
1.3.2 Managerial Relevance
OCR’s are likely to continue to increase in importance and popularity, since advances in
information technology continue to reach an increasing amount of people in the world. In 2013 there are approximately 2.7 billion Internet users around the globe, up from 1 billion users back in 20053. According to the Dutch Central Bureau of Statistics (2012), 96% of Dutch households have Internet access (up from 83% in 2005); of which 84% indicate to have made online
purchases and 87% indicate to use or have used the Internet to look up information about goods and services4. Consumers increasingly use the Internet as a vehicle for pre-‐purchase information gathering (Adjei, Noble, & Noble, 2010). In a study on social shopping behaviour, Freedman (2008) states that 98 percent of shoppers read reviews on retailers’ websites, with 22 percent of shoppers stating to ‘always’, and 43 percent stating to ‘most of the time’ read OCR’s prior to making a product purchase decision. It can thus be concluded that people find it important to
find out about other consumers’ experiences before purchasing a good or service, which explains the growing importance and popularity of OCR’s.
Whereas buzz, defined as “explosive self-‐generating demand”, is often thought of to be largely serendipitous, it is becoming increasingly clear that buzz does not just happen. Instead, it is more and more the result of carefully managed marketing programs (Dye, 2000). Buzz happens because different groups of consumers interact and influence each other, and will only become more important because of trends like globalization, advancements in information technology, the increasing proliferation of brands, and increasing disposable incomes across the globe. So, in order to exploit the full power of customer WOM, it is of vital importance that companies understand how buzz happens. Voight (2007) states that the growth in OCR’s has benefitted online retailers with increased customer loyalty and decreased costs by e.g. savings in returns. However, so far, “merchants have done little traditional measurement of OCR’s other than to understand their customer’s passion for the tool and their subsequent adoption” (Freedman, 2008). At the same time, Verhoef and Leeflang (2009) state that accountability is one of the key antecedents of the marketing department’s influence within a firm. It is thus becoming
increasingly important for marketing departments to justify their expenditures in terms of direct return on investment. In order to become more accountable, it is necessary for (marketing) managers to gain more insights into the world of (e)WOM, including OCR’s.
1.4 Structure of the thesis
In the next chapter, important concepts will be introduced, leading to hypotheses that will be tested in this study. Furthermore, a conceptual model will be presented. Chapter 3 will outline the research design. Research results will be given in Chapter 4, which will be discussed in Chapter 5, in which managerial implications, limitations of the study, and directions for future research will also be presented.
CHAPTER 2: THEORETICAL FRAMEWORK
Chapter 1 provided a general introduction to OCR’s, the concept of (e)WOM and motives for people to use and engage in e(WOM). It also described the managerial and theoretical relevance of this research and presented a research question. Chapter 2 will go into more detail on the main concepts of this thesis: (e)WOM transmission, WOM valence, self-‐relevant and utilitarian products, and a self-‐enhancement vs. a self-‐effacement motive. Hypotheses will be developed based on existing literature. This chapter will also present the conceptual framework and an overview of all hypotheses.
2.1 Dependent variable: (e)WOM transmission
Post-‐purchase (e)WOM transmissions by consumers are a significant, non-‐marketer-‐dominated purchase influence to other consumers (Arndt, 1967; Dichter, 1966). Opportunities for (e)WOM transmission are rising because consumers are becoming increasingly ‘connected’ through social networks like Facebook and Twitter (Stephen & Lehmann, 2009). When consumers transmit WOM, they mainly care about self-‐presentation, self-‐enhancement, and image building (e.g. De Angelis et al., 2012; Toubia & Stephen, 2013; Wojnicki & Godes, 2008). For example, Toubia and Stephen (2013) found that eWOM in the form of Twitter posting (‘tweeting’) can be driven by image-‐related motivations. Stephan and Lehmann (2009) also found that WOM transmitters are largely self-‐focused, and found no evidence to suggest that altruistic intentions play a role in WOM transmission.
In the process of eWOM transmission via OCR’s, a reader (the receiver) reads a message written by an individual (the host) and becomes influenced by that message. The receiver then decides to either further spread the exact message contained in the OCR, to reword it in his/her own words (possibly at a later time), or to not take any action (Kumar et al., 2013). People generally prefer to transmit information that is congruent with the valence of the topic. When a topic is emotionally negative, people pass along exaggeratedly bad news, as opposed to passing along exaggeratedly good news when the topic is emotionally positive. People also prefer to pass on central rather than extreme information, and more believable rather than surprising
information (Heath, 1996).
Research has indicated a broad range of antecedents affecting how likely receivers are to further spread a host’s message. Most of these antecedents can be shared under 1) product, 2) consumer, and 3) WOM message characteristics. First, product factors driving WOM
whether the product is a good or a service, perceived product risk (Moldovan, Goldenberg, & Chattopadhyay, 2011), interestingness, accessibility (Berger & Schwartz, 2011), (perceptions of) product availability and popularity (Dellarocas et al., 2010), and type of product (Zhu & Zhang, 2010). Second, consumer characteristics driving WOM transmission are for example the receiver’s talkativeness (Kumar et al., 2013), Internet experience, education attainment (Zhu & Zhang, 2010), pre-‐commitment, attitude toward WOM (Khare et al., 2011), consumer product experience, media exposure through print, television, and the Internet (Yang et al., 2012) need for uniqueness (Cheema & Kaikati, 2010) and the level of connectivity (Toubia & Stephen, 2013). Third, WOM message characteristics include for example the message stickiness and the host’s perceived expertise (Kumar et al., 2013).
Several dimensions can be used to study and measure WOM transmission, e.g. the amount of WOM (how much people talk) and the WOM transmission probability. The amount of WOM can be measured by looking at the contribution of each individual, or the total buzz in the market (Moldovan et al., 2011). Stephen and Lehmann (2013) study WOM transmission probabilities by asking participants how likely they would be to forward a particular offer to a friend on a scale of 0-‐100%. When looking at the particular ways in which WOM can be transmitted, a distinction can be made between offline and online communication. 75% of WOM is estimated to be transmitted offline, i.e., face-‐to-‐face (Keller & Libai, 2009). Online transmission can for example take place by sharing it on someone’s Facebook wall, tweeting about it, publishing it on a personal blog, rewording the content in another OCR, etc. Berger and Schwartz (2011) suggest there may be different drivers for offline vs. online WOM transmission. The threshold for discussion is often higher in eWOM, since there is no need to fill conversational space (e.g. awkward silences during dinner with a friend). Instead, there is a belief that there is something useful or interesting to be passed along. Another difference between offline and eWOM transmission is the timing of the decision what to share and whom to share it with: in offline WOM, someone first determines whom to share it with; whereas in eWOM what to share is decided before one decides whom to share it with (Berger & Schwartz, 2011).
2.2 Independent variable: WOM valence
(East, Hammond, & Lomax, 2008; Liu, 2006). There are mixed findings regarding the prevalence of PWOM vs. NWOM. For example, looking at the number of incidences (i.e. occurrences in a period), East, Hammond, and Wright (2007) state that PWOM is more common than NWOM with a ratio of 3:1. Alternatively, it is often assumed that dissatisfied consumers are more inclined to engage in NWOM and communicate their disliking to more consumers than satisfied consumers are to engage in and communicate their PWOM. Global PR and communications firm Burson-‐Marsteller (2009) proposes that negative experiences people have heard occurred to others is passed on to an average of 17 people, as opposed to 11 people that information about others’ positive experiences are communicated to. However, TRND research5 finds that the widespread belief that negative experiences are shared with more people than positive ones is not true, simply because people enjoy speaking about positive news and remember good news clearly. There is also not a single answer to the question whether PWOM or NWOM has a greater impact on for example consumers’ judgment, attitude, and brand purchase probability. Some studies have found the impact of NWOM to be greater than PWOM since people tend to weigh negative information more heavily, however, other studies state this is not the case (e.g. Chevalier & Mayzlin, 2006; East, Hammond, & Lomax, 2008; Arndt, 1967). Several studies (e.g. Mangold, Miller, and Brockway, 1999) have found that factors other than (dis)satisfaction affect the production of WOM, which implies the existence of moderators in the relationship between WOM valence and its impact on consumers’ purchasing decisions.
Since literature presents mixed findings regarding both the prevalence of PWOM vs. NWOM and the impact of WOM valence on consumers’ purchasing decisions, research on information transmission is used to set up directional hypotheses on the PWOM and NWOM transmission probabilities compared to that of neutral WOM. Generally, people prefer to pass on central rather than extreme information, and more believable rather than surprising information (Heath, 1996). This suggests that neutral WOM might have a higher transmission probability than
PWOM and NWOM, since it is very central and believable rather than surprising. However, in comparison with PWOM and NWOM, neutral WOM can be perceived as the least useful since it balances positive and negative cues and therefore contains less diagnostic information of the product or service than PWOM or NWOM. Cues are considered ambiguous when the
information has multiple possible interpretations, whereas information that is diagnostic helps
consumers to assign a product to one (and only one) cognitive category (i.e., ‘good’ or ‘bad’ product) (Herr, Kardes, & Lim, 1991). The mixed cues in neutral WOM cause ambiguousness and this in turn might lead consumers to perceive the neutral message as less valuable. Therefore, transmission probability for neutral (ambiguous) messages might be lower than positive or negative (diagnostic) messages, leading to the following hypotheses:
H1a: Compared to neutral WOM, PWOM has a greater probability to be transmitted.
H1b: Compared to neutral WOM, NWOM has a greater probability to be transmitted.
Rozin and Royzman (2001) find that humans and animals have a general bias, based on both innate predispositions and experience, to give greater weight to negative entities. They find evidence for what they call the negativity bias: in most situations, negative events are more salient, potent, dominant in combinations, and generally more effective than positive events. Although not universal, negative entities are also found to be more contagious than positive entities. Additionally, Baumeister et al. (2001) find that bad emotions and bad feedback have more impact than good ones. Their findings also suggest that bad is stronger than good, as a general principle across a broad range of psychological phenomena.
Extremely negative cues are generally less ambiguous than positive or neutral cues, especially in product-‐judgment contexts. Whereas positive and neutral features are associated with many high-‐, medium-‐, and low-‐quality products and services, negative features have stronger implications for categorization (i.e., low quality). Even when many positive features are exhibited, a single extremely negative feature can be highly informative and will be weighed heavily in judgment (Herr et al., 1991). This is consistent with later findings that many good events can overcome the psychological effects of a single bad ones; but when equal measures of good and bad are present, the psychological effects of bad ones outweigh those of the good ones (Baumeister et al., 2001).
desirable, and/or beneficial) experiences. So, consumers will be more likely to transmit NWOM than PWOM, leading to the following hypothesis:
H1c: NWOM has a greater effect than PWOM on the probability to be transmitted.
2.3 Independent variable: Utilitarian vs. self-‐relevant products
Utilitarian goods are those whose consumption is more cognitively driven, instrumental, and goal oriented. These products mainly satisfy a functional need or accomplish a practical task (Strahilevitz & Myers, 1998). Utilitarian purchases (e.g. microwaves and minivans) are primarily motivated by functional product aspects (Dhar & Wertenbroch, 2000). On the other hand, hedonic goods provide for experiential consumption, fun, pleasure, and excitement (Holbrook & Hirschman, 1982). Focus in this study will be on those products that can be seen as a specific type of hedonic products, namely self-‐relevant products. Some products are more likely to convey information about their owners than others: they might convey social impressions or serve a self-‐esteem function (Chung & Darke, 2006). Sometimes consumers identify with products or brands in a way that it actually becomes a part of their extended self (Belk, 1988). Products can, to some extent, be valued because they are symbolic of the consumer’s
personality or identity in some way (Holbrook & Hirschman, 1982). In this study, self-‐relevant products are defined as those products that, compared to more utilitarian products, seem to ‘offer consumers an opportunity to communicate something important about themselves to
others by providing WOM’ (consistent with Chung & Darke, 2006). Examples are designer
clothes and luxury watches (Dhar & Wertenbroch, 2000).
Chung and Darke (2006) find that consumers are more likely to provide WOM for products that are self-‐relevant, than for products that are utilitarian. WOM concerning self-‐relevant products serves as a means of self-‐presentation, whereas WOM about utilitarian products does not provide the same social benefits. Impression goals and a need to self-‐promote also cause WOM of individualistic consumers to be biased in a way that consumers exaggerate the benefits of self-‐relevant products, compared to utilitarian products (Schlenker, 1980). Consistent with the findings that consumers mainly care about self-‐presentation, self-‐enhancement, and image building in WOM transmission (De Angelis et al., 2012; Toubia & Stephen, 2013; Wojnicki & Godes, 2008), it is hypothesized that:
H2: The probability that eWOM will be transmitted is higher for self-‐relevant products than for utilitarian products.
2.4 Moderator: Self-‐enhancement vs. self-‐effacement
Research has found that even though there are several motives that can independently drive either PWOM or NWOM, the single motive of self-‐enhancement leads to systematic differences in WOM valence, because in order to fulfil this basic motive, consumers can use both PWOM and NWOM (Wojnicki & Godes, 2008; De Angelis et al., 2012). Here, the self-‐enhancement motive will be described in depth and a self-‐ effacement motive will be introduced.
The assumption that people have a need to view themselves positively is one of the most common and unquestioned assumptions in research on the self (e.g. Heine et al., 1999). This self-‐enhancement motive refers to an individual’s propensity to reinforce or magnify the positive, and to reduce the negative of the self-‐concept, in which the self-‐concept is defined as ‘the cognitive representation of one’s own attributes’ (Sedikides & Strube, 1997; Markus & Wurf, 1987). People tend to enhance the self because of for example a basic human need to feel good about oneself, one’s desire for positive recognition from others, and a need to enhance, verify, assess, or improve the self (or self-‐concept) (e.g. Baumeister, 1998; Engel et al., 1993; Sedikides & Strube, 1997). Individuals’ self-‐perceptions are often heavily skewed towards the positive end, because people (sometimes unconsciously) use several ‘techniques’ to positively bias their self-‐ concepts. Most individuals are very aware of their strengths and assets, but considerably less aware of their weaknesses and faults (Taylor & Brown, 1988). Also, people tend to consider themselves to be ‘above-‐average’ (also known as the ‘better than average (BTA) effect’) to a logically unfeasible degree, known as the false uniqueness bias (Alicke, 1985).
destiny, causing people to make relatively fewer person attributions and more situation attributions (e.g. Miller, 1984).
Even though researchers do seem to agree to the fact that at an abstract level, there is a universal need for positive self-‐regard (e.g. Heine, 2005; Heine et al., 1999), there is an intense debate in research between the universalist and relativist perspective. They disagree on whether or not tendencies to elaborate on the positive compared with the negative about the self are universal or culturally relative (e.g. Sedikides, Gaertner, & Toguchi, 2003; Heine et al., 1999). In the following section, both perspectives will be described in more detail.
Universalist perspective
Researchers in this stream believe that people in both Eastern and Western cultural contexts are motivated to evaluate themselves favourably (e.g. Kim et al., 2010). This argument is mainly based on the general assumption that humans have a basic motivation to self-‐bolster or self-‐ protect (Sedikides & Strube, 1997). Universalists also believe that satisfaction of the self-‐ enhancement motive promotes psychological well-‐being, regardless of culture (O’Mara et al., 2012). Sedikides, Gaertner and Vevea (2005) find that both Westerners and Easterners self-‐ enhance on attributes they find important, i.e. those that imply culturally prescribed role fulfilment. They state that Westerners self-‐enhance on individualistic attributes, whereas Easterners self-‐enhance on attributes that are relevant to the cultural ideal of collectivism. So, the universalist perspective states that self-‐enhancement is universal, but manifests differently across cultures because it is being regulated by cultural norms, e.g. the norm of modesty (Alicke & Sedikides, 2009; Kim et al., 2010).
Relativist perspective
Construals (interpretations) of the self, others, and the interdependence of the two vary greatly across people in different cultures. Generally, Western cultures stress attending to the
order to facilitate this process of accommodation, East Asians are focusing on their shortcomings (Heine et al., 1999), i.e. they have self-‐effacing or self-‐criticizing tendencies. According to e.g. Heine et al. (1999), there is less need for self-‐esteem in Eastern cultures. Self-‐ effacement (i.e. self-‐criticism) serves as a means to promote and maintain social connections among self and others (Heine et al., 1999) and is defined as “an orientation in the direction of attending, elaborating, and emphasizing negatively valenced aspects of the self” (Kitayama et al., 1997). So, the relativist perspective states that the internalization of the Western mandate for individualism gives rise to the self-‐enhancement motive, whereas internalization of the East Asian mandate for collectivism gives rise to the self-‐effacement (e.g. criticism, averageness) and self-‐improvement motives (Gaertner et al., 2012). Therefore, the self-‐enhancement motive does not occur in Eastern cultures and is not universal (Heine et al., 1999).
Conclusion on self-‐enhancing vs. self-‐effacing effect
So, whereas the universalist perspective states that both Easterners and Westerners have a need to make favourable self-‐evaluations and thereby, have a need to self-‐enhance, the relativist perspective states that this is only the case in Western cultures. In a meta-‐analysis of cross-‐cultural studies on self-‐enhancement, 88 of 91 studies revealed that East Asians self-‐ enhanced less than Westerners, with large effect sizes for 58 percent of those comparisons (Heine & Hamamura, 2007). Mezulis et al. (2004) conducted a meta-‐analysis of 266 studies on the self-‐serving attributional bias (one of the ways to self-‐enhance) and conclude that this bias is pervasive in the general population. However, there is significant variability between cultures, with Asian samples displaying significantly smaller self-‐serving biases than U.S. or Western samples. For those reasons, this research will be conducted from a relativist perspective. This leads to the following hypotheses:
H3a: Individuals from Western cultures display a self-‐enhancing motive.
H3b: Individuals from Eastern cultures display a self-‐effacing motive.
generation. They state that a self-‐enhancement motive leads consumers to mainly transmit NWOM (i.e. share negative information about others), because when others perform poorly or have a negative experience, highlighting these negative aspects makes people feel better about themselves. It helps them to maintain a positive self-‐view and thus offers a means to self-‐ enhance. Conversely, people’s evaluation of the self might be negatively affected when others perform well or have a positive experience, which is another motivation to only highlight the negative aspects when transmitting WOM (e.g. Tesser, 1988; Krueger, 1998). However, de Angelis et al. (2012) did not take into account the Eastern self-‐effacing motive that might cause changes in (e)WOM valence between the WOM stages. Research conducted among Eastern and Western instead of just Western respondents might cause less pronounced or even reversed findings. Therefore, it is hypothesized that:
H4a: Because of a self-‐enhancing motive, there is a higher probability that individuals in Western cultures transmit negative eWOM, rather than positive eWOM.
H4b: Because of a self-‐effacing motive, there is a higher probability that individuals in East Asian cultures transmit positive eWOM, rather than negative eWOM.
Considering the type of product, consumers are more likely to provide WOM on self-‐relevant products than on utilitarian products. However, the exaggeration of benefits of self-‐relevant products is found to cause differences in WOM valence between cultures (Chung & Darke, 2006). Self-‐relevant products have a greater impact on WOM in individualistic (Western) than collectivist (Eastern) cultures, consistent with the differences in the way self-‐concept is typically construed by these groups (Chung & Darke, 2006). Furthermore, collectivistic consumers generally tended to give less WOM than individualistic consumers. Here, the cultural variable of individualism vs. collectivism used by Chung and Darke (2006) will be studied into more detail by looking at the influence of a self-‐enhancement vs. a self-‐effacing motive on the probability of WOM transmission by Western self-‐enhancing vs. Eastern self-‐effacing consumers. It is hypothesized that:
H5a: (Western) individuals with a self-‐enhancing motive, will have a higher probability to transmit eWOM on self-‐relevant products than on utilitarian products.
H5b: (Eastern) individuals with a self-‐effacing motive will show no significant differences in the probability of transmitting eWOM on self-‐relevant vs. utilitarian products.
2.5 Control variables
Certain variables might have an effect on WOM transmission probability, including
characteristics of the subjects and the stimulus (Park et al., 2007). To control for the effects of possible confounding variables, three control variables will be included: attitude toward OCR’s, Internet usage frequency and four individualism-‐collectivism dimensions. Furthermore, a validation for the moderator will be included: the above-‐average effect.
1) Attitude toward OCR’s
One’s attitude toward OCR’s can impact consumer attitudes and behaviours (Park et al., 2007). This means that one’s attitude toward WOM in general and OCR’s can affect the eWOM transmission probability and therefore, attitude toward OCR’s is included as a control variable. 2) Internet usage frequency
Internet usage frequency is used as an indicator of Internet experience. OCR’s are found to be more influential when consumers have relatively greater Internet experience (Zhu & Zhang, 2010). Internet experience can also lead to greater confidence in using the Internet (Bart et al., 2005) and benefits of reduced search costs (Bryanjolfsson & Smith, 2000). Furthermore, consumers’ online experience moderates their trust in different information sources (Klein & Ford, 2003). People also become more capable of evaluating different kinds of reviews when they use OCR’s on a regular basis (Park et al., 2007).
3) Individualism vs. collectivism dimensions