Does Gender Matter?

This article was downloaded by: [UAM University Autonoma de Madrid] On: 02 October 2014, At: 05:19 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

International Journal of Human-Computer Interaction Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/hihc20

Gossiping Behavior on Social Networking Sites: Does Gender Matter? a

a

a

Shintaro Okazaki , Natalia Rubio & Sara Campo a

Universidad Autónoma de Madrid, Madrid, Spain Accepted author version posted online: 04 Apr 2014.Published online: 10 Jul 2014.

To cite this article: Shintaro Okazaki, Natalia Rubio & Sara Campo (2014) Gossiping Behavior on Social Networking Sites: Does Gender Matter?, International Journal of Human-Computer Interaction, 30:9, 718-726, DOI: 10.1080/10447318.2014.907710 To link to this article: http://dx.doi.org/10.1080/10447318.2014.907710

PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions

Intl. Journal of Human–Computer Interaction, 30: 718–726, 2014 Copyright © Taylor & Francis Group, LLC ISSN: 1044-7318 print / 1532-7590 online DOI: 10.1080/10447318.2014.907710

Gossiping Behavior on Social Networking Sites: Does Gender Matter? Shintaro Okazaki, Natalia Rubio, and Sara Campo

Downloaded by [UAM University Autonoma de Madrid] at 05:19 02 October 2014

Universidad Autónoma de Madrid, Madrid, Spain

This study examines how gender affects online gossip on social networking sites. Based on gender theories and agency-communion theory, it is posited that achievement value, friendship value, and normative pressure differ according to gender (female vs. male), the level of propensity to gossip (high vs. low), and the interaction between the two. An experimental survey is conducted with 809 general consumers. Between-subjects multivariate analysis of covariance reveals that gender has an impact only on friendship value, whereas propensity to gossip affects achievement value and normative pressure. No interaction effects are observed. However, a subsequent analysis of covariance finds an interaction between gender and propensity to gossip through electronic word-of-mouth for a high-involvement product. In closing, theoretical and managerial implications are discussed while important limitations are recognized.

1. INTRODUCTION Broadly defined, gossip is information about the new, deviant traits or behaviors of other people (Arno, 1980). Gossip usually meets three conditions: It should be (a) of interest and consequence to the listener, (b) difficult to dismiss immediately but also credible, and (c) attention grabbing (Guerin & Miyazaki, 2006). In the literature of evolutionary psychology, gossip has been conceptualized as one of the verbal (or conversational) strategies people use to influence others in some way, and there are probably as many verbal strategies as there are social interactions (Guerin, 2003). Although gossip has long been associated with the communication of negative information about others, more recently researchers have also emphasized the positive aspects of gossip (Foster, 2004; Watson, 2012). Gossip differs from rumor, which is always speculative and sometimes pertains to events rather than people (Foster, 2004). Gossip also differs from electronic word-of-mouth (eWOM), which usually refers to any positive or negative online statement posted or articulated by potential, actual, or former customers Address correspondence to Shintaro Okazaki, c/ Francisco Tomás y Valiente 5, College of Economics and Business Administration, Universidad Autónoma de Madrid, Cantoblanco, Madrid 28049, Spain. E-mail: [email protected] Color versions of one or more of the figures in the article can be found online at www.tandfonline.com/hihc.

about a product or company (Hennig-Thurau, Gwinner, Walsh, & Gremler, 2004). Yet, we argue that online gossip could be a seed of eWOM under certain circumstances. Online gossip can be seen as a verbal strategy for social and personal topics that people use to influence others on the Internet, in particular on social networking sites (SNSs). Consumers often “chitchat” about brands and discuss curious, scandalous, novel, humorous, and unexpected stories about people related to those brands on SNSs. For example, suppose that one Facebook user posts this comment on her wall, “I didn’t know McDonald’s has its own coffee now.” Her friend may write back, “Do you know Jane is a total caffeine addict? She drinks 10 of those McCafés a day.” This statement would involve “small talk” about both the third person not present and the brand. This person may continue and endorse the brand: “I understand why . . . they taste so awesome!” Clearly, this is a brand endorsement or eWOM that is incidentally derived from online gossip. Against this background, the overall objective of this study is to address how gender interacts with online gossip propensity on SNSs. In doing so, we test a series of predictions based on extant gender-difference theories by comparing female and male gossipers’ achievement value, friendship value, and normative pressure, as well as their ultimate eWOM intention to disseminate brand-related information advertised or promoted on an SNS. Our theoretical rationale lies in the agency-communion theory, which explains that agency and communion describe two basic styles of how individuals relate to their social world (Zarbatany, Conley, & Pepper, 2004). Agency refers to an individual’s efforts to master the environment; to assert the self; and to experience competence, achievement, and power. In contrast, communion refers to a person’s desire to closely relate to, cooperate with, and merge with others (Bakan, 1966). The agency-communion dyad seems applicable to explain gender differences in gossiping behavior (Zarbatany et al., 2004), and thus may explain why female and male individuals engage in eWOM differently on SNSs. The site of the study is Spain. Spain seems to be an ideal European country for this research, as its SNS penetration rate (41%) is nearly identical to the average penetration rate of 40% in 40 countries across Europe. With regard to time spent on

718


GOSSIPING BEHAVIOR ON SNSS

social media, Spain is somewhere in the middle of the European average, at 1.5 hr a day (“We Are Social,” 2014). These figures ensure that our study is based on typical European SNS users, which increases the generalizability of our results. Two issues motivate the current study. First, prior studies investigating marketing and informal communication have focused almost exclusively on eWOM or knowledge sharing (e.g., Cho, Huh, & Faber, 2014; José-Cabezudo & CamareroIzquierdo, 2012; Xu, Li, & Shao, 2012). However, much of the activity on SNSs has been motivated by social grooming— exchanging and browsing social information about friends and acquaintances—through gossip (Tufekci, 2008). Second, gossip as conversation about social and personal topics is often playfully reworded as “women’s talk” or “girl talk” (Foster, 2004), referring specifically to women. The essentially interchangeable equivalents in men’s conversation are “shop talk” or “shooting the breeze,” when it is defined generally as social conversation. Because of such connotations associated with gender, research has long been interested in the connection between gender and gossip (Foster, 2004; Watson, 2012). However, this question has seldom been addressed in a marketing context, in comparison with rumor or eWOM (e.g., Dubois, Rucker, & Tormala, 2011; Kamins, Folkes, & Perner, 1997). Developing an understanding of what role online gossip plays in SNSs in terms of gender behavior in a context of eWOM can make a number of contributions to the literature. First and foremost, a clear distinction between online gossip and eWOM may be a neglected area of research in marketing. Although more and more companies advertise goods and services on SNSs in hopes of eWOM referrals, most users ignore such ads (Rodriguez, 2012). On the other hand, gossip has been one of the most eminent motives for using SNSs, not only because gossip is fueled by the way the news feed system works but also because gossip often provides the social glue that keeps the community alive and interesting (Debatin, Lovejoy, Horn, & Hughes, 2009). This study responds directly to the desire of firms to have their brand names mentioned in such gossip, which may turn into positive eWOM. Second, prior research indicates that women and men differ in their response to the same psychological rewards; thus, female and male consumers require different selling approaches as they respond differently to actions aimed at enhancing customer loyalty. For the purpose of the current study, we exemplify this issue in a context of brand promotion posted on SNSs. Thus, the differences between female and male individuals in response to such promotion would provide useful information to online marketers and advertisers. Third, the prior two issues are universal concerns related to firms’ brand promotion. This study carefully attempts to make a strong case for Spanish female and male individuals, but the results could be tested in other cultural contexts to assess the generalizability of our conclusions. Such a step would truly advance our knowledge in gender theory. The remainder of the article is organized as follows: We first establish our theoretical framework, in terms of the

719

agency-communion theory. On this basis, we formulate a series of hypotheses. Then, we describe the methodology employed in this study in detail and report the statistical results. Finally, both theoretical and managerial implications are drawn, important limitations are recognized, and future research is suggested.

2. AGENCY-COMMUNION THEORY There is a rich literature base in both psychology and marketing suggesting that “talk” may hold different meanings for women and men. Men are more likely to feel comfortable performing public speaking, whereas women tend to feel most comfortable conversing in private. However, Dunbar, Marriott, and Duncan (1997) found that, as far as conversation about “social topics” is concerned, both women and men spent approximately equal time or two thirds of their total conversation time on social matters. Although men are less likely to describe their verbal pastime as gossip, other terms such as networking, shop talk, small talk, schmoozing, shooting the breeze, and killing some time together may be used to describe the same behavior (Foster, 2004). For most women, the language of conversation is primarily a question of rapport or a way of establishing connections and negotiating relationships (Tannen, 1990). Prior research indicates that e-mail and other forms of text messaging tend to attract more female participants, because both offline and online, “talk” makes up the substance of women’s friendships (Weiser, 2000). In addition, advertising research reveals an affinity for verbal ads among female consumers (Putrevu, 2004). Friendship was found to be the primary context of gossip among youth, because small talk with close friends relied on the sense of belonging to a group (Watson, 2012). In this research, gender differences in friendship resulted in how gossip functioned distinctly between women and men. This finding may reflect a fact that, compared with male friendships, female friendships are generally more intimate, because women’s interaction patterns are more socially oriented and focused on relationship building, forming broader, more holistic friendships rather than circumscribed friendships (Johnson, Veltri, & Hornik, 2008; Wright, 2006). Using Bakan’s (1966) original premise, Zarbatany et al. (2004) employed the agency-communion theory to explain gender differences in gossip. Agency-communion theory distinguishes between two types of friendship: communal and agentic (Wiggins, 1992). In the former, power is distributed equally for the most part, and the friendship itself is measured in terms of interpersonal closeness, which includes companionship, intimacy, loyalty, security, emotional support, love and affection, nurturance, and mutual past experiences. Agentic friendship, on the other hand, emphasizes individual differentiation, with the goal of promoting an individual’s social standing through mastery or power. Friends may provide selfworth through acceptance, ego support, and validation, and may contribute to “the development of mastery by providing


720

S. OKAZAKI ET AL.

instruction, instrumental aid, and opportunities to rehearse skills and knowledge” (Zarbatany et al., 2004, p. 300). Although both women and men value the communal aspect of friendship more than agency, men are better able to balance these two facets of friendship. Men achieve their need for status and social prominence through the communal relationships they share with close friends. With men, the relative weight is on agency, as they value the enhanced status that friendship can provide, but they will not give up their communion, or close friendships, in order to increase their social standing (Zarbatany et al., 2004). Gossip may serve the more agentic quality of male friendships, as it focuses on external events, and tends to be more emotionally detached compared to the communication shared by women (Hall, 2011). This agency versus communion distinction in terms of gender gossip tendencies appears to be consistent with the widely accepted theoretical accounts of gender proposed by Chodorow (1978)—male identities are structured based on differentiation, separation, and autonomy, whereas female identities are structured in terms of identification, connectedness, and relationships that are formed.

information. This is almost synonymous to achievement value. We therefore argue that, when users’ general propensity to gossip is high, men’s online verbal strategy will be more utility driven, seeking a greater level of achievement value in chatting or posting on SNSs. For example, prior research found that men evaluated leadership and associated managerial roles more positively than women (Killeen, López-Zafra, & Eagly, 2006), implying men’s greater propensity to seek achievement. In advertising research, men responded more favorably to self-oriented (help-self) charitable ad appeals, as opposed to altruistic (help-others) ads (Brunel & Nelson, 2000). Based on this discussion, we posit the following hypotheses:

3. HYPOTHESES Social exchange theory explains that the main functions of gossip are to facilitate information exchange, provide recreational pastime, and bring groups together through the sharing of norms, thereby creating a group’s influence faction (Stirling, 1956). In this light, this study chose to first focus on three dependent variables: achievement value, friendship value, and normative pressure. We believe that these three variables shape our final dependent variable, eWOM intention. In this section, we explicate these variables and formulate research hypotheses.

3.2. The Role of Gender and Gossip in Friendship Value In light of agency-communion theory, it seems reasonable to assume that, compared with men, women are more prone to establish and maintain friendship value—contact with other people such as social support, friendship, and intimacy (Dholakia et al., 2004)—when they engage in verbal strategy. This is consistent with a general view that women are strongly motivated by affiliation needs (Hoffman, 1972), and are more disposed than men to pursue interpersonal goals and success in interpersonal relationships (Gill et al., 1987; Venkatesh & Morris, 2000). Regardless of whether the communication is online or offline, women tend to seek intimate friendships through gossip, reaffirming their belongingness to a group. This relationship should hold true when users’ general propensity to gossip is high. Thus, we predict the following main and interaction effects of gender on friendship value of SNSs, such that:

3.1. The Role of Gender and Gossip in Achievement Value Agency-communion theory suggests that gossip can be used to increase the status or power of an individual in a group; thus, this type of gossip is more common for men who tend to emphasize agentic friendship (Zarbatany et al., 2004). This is because masculinity essentially concerns “the attainment of goals external to the interaction process” (Gill, Stockard, Johnson, & Williams, 1987, p. 379). Research shows that such purposive or goal-derived communication influences consumers’ search for the meaning of brands, playing an important role in their assessments of brands’ ability to deliver desired benefits (Martin & Gnoth, 2009). This seems to suggest that men’s gossip is purposive or goal directed, rather than random or incidental. Furthermore, prior research indicates that men, compared with women, are more likely to prefer an assertive, direct, and result-oriented style of communication, favoring dogmatic and pragmatic patterns (Kirtley & Weaver, 1999). In conceptualizing social influence through virtual communities, Dholakia, Bagozzi, and Pearo (2004) found that online users were often motivated by purposive value or predetermined instrumental purpose, including giving or receiving

H1a: Men will perceive the achievement value of SNSs more strongly than women. H1b: Gossipers will perceive the achievement value of SNSs more strongly than nongossipers. H1c: Men with a high level of propensity to gossip will perceive the achievement value of SNSs more strongly than their female counterparts.

H2a: Women will perceive the friendship value of SNSs more strongly than men. H2b: Gossipers will perceive the friendship value of SNSs more strongly than nongossipers. H2c: Women with a high level of propensity to gossip will perceive the friendship value of SNSs more strongly than their male counterparts. 3.3. The Role of Gender and Gossip in Normative Pressure Research on online communities generally concerns the overlap between the community’s and the individual’s norms, values, and goals. The extent of such overlap is often examined as normative pressure. This study views normative pressure as the users’ perceptions of the online community’s extrinsic



demands on a person to interact and cooperate within the community (Algesheimer, Dholakia, & Herrmann, 2005). The influence of normative beliefs from group members is an important determinant of the attitude–intention–behavior chain and has been widely documented in works related to theory of planned behavior (Ajzen, 1991). We believe that normative pressure would influence SNS users’ actions related to initiation, engagement, and ongoing interactions through gossip. Prior research shows that women tend to be more compliant, whereas men are more likely to rebel against requests or orders from others (e.g., Minton, Kagan, & Levine, 1971). Similarly, women appear more likely than men to conform with majority opinions (Eagly, 1978). However, more recent empirical evidence suggests the contrary. First, in a study on new software adoption, the perceptions of normative pressure among women were lower than the perceived pressure among men (Venkatesh & Morris, 2000). Second, in a study on drivers’ intentions to break the speed limit, the presence of passengers moderated the impact of normative pressure for male and female participants. The young male drivers were more likely than the young female drivers to report a greater impact of normative beliefs on their intentions in the driving alone condition. Thus, changing the normative pressure on young men to drive fast when driving alone appears to be a potentially important way to change this behavior, whereas such an intervention in women might be less successful (Conner, Smith, & McMillan, 2003). Third, male users’ online activities are influenced by relationship status more strongly than their female counterparts in Facebook (McAndrew & Jeong, 2012) and Second Life (Choi, Chung, & Kim, 2012). In this light, we predict the following hypotheses for normative pressure: H3a: Men will perceive the normative pressure of SNSs more strongly than women. H3b: Gossipers will perceive the normative pressure of SNSs more strongly than nongossipers. H3c: Men with a high level of propensity to gossip will perceive the normative pressure of SNSs more strongly than their female counterparts.

3.4. The Role of Gender and Gossip in eWOM Brown, Broderick, and Lee (2008) found that online community members recognized the social value of product-related gossip because knowledge of such unofficial information made them feel like an “expert.” They demonstrated their knowledge and “insider” status by reporting it to other network members. Such privileged knowledge and status seem more related to agentic, rather than communal, friendship. Furthermore, the value of such information will be higher when consumers’ inherent needs and interests toward that product are high. In this regard, a dichotomy of low- and high-involvement product has been widely examined to explain the variation in the degree of effort a consumer is willing to devote to their

721

consumption (Laaksonen, 1994). Product involvement refers to the degree to which a person perceives a product to be personally relevant (e.g., Zaichkowsky, 1985). High involvement has been connected with consumer behavior in cases such as making an active effort to find product information and making a detailed comparison of product attributes and perceived differences between brands (Zaichkowsky, 1985). In light of agentic friendship, it seems reasonable to assume that the information acquired through gossip will bolster an individual’s prominence within a social group when the perceived values as well as the level of differentiation in that product category are high. In this regard, Williams and Slama (1995) found that men scored high on market mavenism when the products were highly visible and risky, and differentiated in style. In terms of agencycommunion theory, men with greater propensity to gossip may be more likely to participate in eWOM for that kind of product. Accordingly, for a high-involvement product, we propose the following: H4: In high-involvement situations, men with a high level of propensity to gossip will exhibit greater eWOM intention to disseminate information related to the advertised brand promotion, compared with their female counterparts. In contrast, gossipers may be less motivated when the degree of involvement in a product category is low. Prior research has unanimously found that WOM is significantly motivated by product involvement (Sundaram, Mitra, & Webster, 1998) or enduring involvement (Richins & Root-Shaffer, 1988). Thus, it seems empirically as well as theoretically justifiable to predict that, when there is a loss of interest, need, and value associated with the product, neither women nor men may be engaged in eWOM. Accordingly, we contemplate the following: H5: In low-involvement situations, gender will not have an impact on eWOM intention, regardless of the level of propensity to gossip. 4. METHOD 4.1. Procedure To test the hypotheses, a study was conducted in Spain, using a general consumer sample of 809 SNS users. A professional research firm was responsible for recruiting the participants from its online panel database. The panel members were chosen according to age, gender, and geographic location, which approximated the national population. Approximately 60% of the invited panel members agreed to participate in the survey. Of the participants, 54.9% (444) were male and 45.1% (365) were female. In terms of age, the proportions of men and women were very similar, except in two age brackets. The proportion of female respondents ages 30 to 39 was slightly greater than their male counterparts. In contrast, the proportion of male respondents ages 50 to 59 was substantially larger than their


722

S. OKAZAKI ET AL.

female counterparts. With regard to occupation, the proportions of the employed male and female respondents were similar, but the unemployed women outweighed the unemployed male respondents. Clearly, there were substantially more women, compared with men, who had used SNSs for only 1 or 2 years. In contrast, the number of men who had used SNSs for 5 or 6 years or more was significantly greater than that of women. The questionnaire was divided into two parts. The first part included questions related to the constructs of H1a to H3c, that is, achievement value, friendship value, normative pressure, and propensity to gossip. In the second part, a level of product involvement was introduced as an experimental stimulus. The pretest was used to establish two levels of product involvement: The low-involvement product was beer and the high-involvement product was athletic shoes. The respondents were randomly assigned to the two scenarios, according to low and high levels of involvement. The high-involvement scenario was as follows: Imagine that while logging in to your account on a social network today, you find an ad for a popular brand of athletic shoes. This brand is currently promoting a new athletic shoe and offers a 40% discount coupon. In order to obtain this coupon, you simply have to click the link and register your name and e-mail address. This promotion will end in one week, and during that time, the coupon can be redeemed at any store. Based on this scenario, please rate the following questions according to a 7-point scale (1 = completely disagree; 7 = completely agree).

After reading this scenario, the respondents were asked to rate their eWOM intention. 4.2. Measures All questionnaire items are adopted from prior research. Seven-point scales anchored by completely disagree/completely agree were used. Friendship value and achievement value were measured by the purposive value (three items) and interpersonal connectivity (three items) scales suggested by Dholakia et al. (2004). Normative pressure was measured by a three-item scale suggested by Algesheimer et al. (2005), and eWOM intention was measured by a three-item scale adapted from Verhoef, Franses, and Hoekstra (2002). Propensity to gossip was measured by the social value subscale (seven items) of the attitude toward gossip scale developed by Litman and Pezzo (2005). We used a median split to divide the sample into high and low propensity to gossip. 4.3. Manipulation and Realism Check After the data collection, the respondents’ perceived levels of involvement were measured by three questions adapted from Mittal (1989)—“This product means a lot to me,” “I constantly compare the prices and other conditions offered by various companies,” and “I always weigh the pros and cons of my choice” (Cronbach’s α = 0.91). The levels of involvement were compared for two products included in the scenarios (i.e., beer vs.

athletic shoes). A t test indicated that the difference between the two products was statistically significant at p < .01 (t = 95.53). A realism check was performed by asking the respondents to rate two questions—“The situation described was realistic” and “You had no difficulty imagining yourself in the situation”— on a scale of 1 (completely disagree) to 7 (completely agree; Dabholkar & Bagozzi, 2002). The results produced a mean rating of 4.51 (Cronbach’s α = 0.78), which was deemed to be sufficiently realistic for the purpose of this study. We conducted a confirmatory factor analysis for achievement value, friendship value, normative pressure, and eWOM intention. The model fit our data reasonably well. The measurement items loaded significantly on the corresponding constructs with high loadings and t values. 5. HYPOTHESES TESTING First, we ran a multivariate analysis of variance with gender as an independent variable. The overall difference between females and males was statistically significant, Wilks’s λ = .98, F(4, 804) = 4.30, p = .002. However, a series of univariate analysis of variance revealed that the significant difference only lay in friendship value, whereas no other variables differed significantly between women and men. We then included age and the number of registered friends as covariates to control the effects. However, the results were practically identical. The hypotheses H1a to H3c were tested using 2 (male/female) × 2 (high/low propensity to gossip) betweensubjects multivariate analysis of covariance (MANCOVA) procedures with gender and propensity to gossip as independent variables and achievement value, friendship value, and normative pressure as dependent variables. Age and the number of registered friends were included as covariates to control the effects external to the agency and communal groupings. The cell means and standard deviations for each cell are shown in Table 1. The MANCOVA results are summarized in Table 2. TABLE 1 Cell Means of Dependent Variables for H1a to H3c Dependent Variables Independent Variables Gossip Gender Propensity Male Male Female Female

Low High Low High

Achievement Value

Friendship Value

Normative Pressure

M

SD

M

SD

M

SD

3.94 4.41 3.92 4.42

1.47 1.31 1.55 1.45

4.94 5.17 5.20 5.51

1.34 1.12 1.45 1.15

2.56 3.69 2.27 3.65

1.35 1.36 1.30 1.53

Note. All constructs were measured by multiple-item, 7-point Likert scales from 1 (completely disagree) to 7 (completely agree), with 4 as an anchoring point.

723


TABLE 2 Multivariate Analysis of Covariance Results for H1a to H3c Univariate Source


Covariates Age Registered friends Gender (G) Propensity to gossip (P) G×P

Multivariate

Achievement Value

Friendship Value

Normative Pressure

F(3, 801) 7.22∗∗∗ 2.46 5.52∗∗ 55.13∗∗∗ .67

F(1, 803) 9.61∗∗ 6.89∗∗ .00 25.70∗∗∗ .01

F(1, 803) 1.75 2.49 10.62∗∗ 6.19∗ .30

F(1, 803) 6.02∗ 2.77 2.41 165.63∗∗∗ 1.47

Note. F ratios are Wilks’s approximation of Fs. ∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001.

H1a to H1c propose that the perception of achievement value differs according to gender and the levels of propensity to gossip (high/low). More specifically, H1a to H1b hypothesize the main effects of gender and propensity to gossip, and H1c contemplates the interaction effects of the two independent variables. For an analysis of covariance (ANCOVA), the effects of the two covariates were statistically significant. Achievement value varied in the direction we predicted, in terms of both gender and propensity to gossip. However, only the latter was statistically significant (M high = 4.44, M low = 3.91), F(1, 804) = 25.70, p < .0001. Thus, H1b was supported by our data, but H1a and H1c were not. Similarly, H2a to H2c address the perception of friendship value in terms of the main effects of gender and propensity to gossip, and the interaction effects between the two variables. Our ANCOVA results indicate that neither covariate was significant. The main effects of independent variables, gender (M male = 5.06, M female = 5.35), F(1, 804) = 25.70, p < .001, and propensity to gossip (M high = 5.32, M low = 5.09), F(1, 804) = 25.70, p < .05, were both significant, but the interaction effects were not. Thus, our data supported both H2a and H2b but not H2c. H3a to H3c posit that normative pressure is perceived in the same way as achievement value. Our predictions were indeed true—our empirical results mirrored those of achievement value. Only propensity to gossip was statistically significant (M high = 3.69, M low = 2.40), F(1, 804) = 165.63, p < .0001, whereas neither gender nor interaction effect was significant. The results suggest that only one covariate, age, was significant. Therefore, H3b was supported, but H3a and H3c were not. H4 and H5 were tested using a 2 (male/female) × 2 (high/low propensity to gossip) between-subjects ANCOVA for each level of involvement separately. Age and the number of registered friends were included as covariates. The cell means and standard deviations for each cell are shown in Table 3. The ANCOVA results are summarized in Table 4.

TABLE 3 Cell Means of Dependent Variables for H4 and H5 Dependent Variable

Independent Variables

Gender

Gossip Propensity

Product Involvement

Male Male Female Female Male Male Female Female

Low High Low High Low High Low High

High High High High Low Low Low Low

eWOM intention M

SD

3.09 4.13 3.47 3.90 3.45 4.11 3.71 3.96

1.77 1.64 1.92 1.76 1.84 1.65 1.87 1.89

Note. All constructs were measured by multiple-item, 7-point Likert scales from 1 (completely disagree) to 7 (completely agree), with 4 as an anchoring point. eWOM = electronic word-of-mouth .

Our H4 expects that, for eWOM intention of a highinvolvement product, there will be an interaction effect of gender and propensity to gossip. On the other hand, H5 predicts no gender effect on eWOM intention for a low-involvement product. Our ANCOVA results supported these predictions. The interaction effects were only observed for athletic shoes or a high-involvement product, F(1, 394) = 3.92, p < .05, but not for beer or a low-involvement product (Figure 1). Thus, both H4 and H5 were supported by our data.

6. DISCUSSION This study provides several important theoretical implications and advances our knowledge on gossip, eWOM, and online promotion from gender perspectives. Our research is

724

S. OKAZAKI ET AL.

TABLE 4 Analysis of Covariance Results for H4 and H5 eWOM Intention Source


Covariates Age Registered friends Gender (G) Propensity to gossip (P) G×P

High Involvement

Low Involvement

F(1, 394) 5.96∗ 17.63∗∗∗ .10 18.53∗∗∗ 3.92∗

F(1, 394) 13.68∗∗∗ 14.97∗∗∗ .21 8.88∗∗ 1.65

Note. eWOM = electronic word-of-mouth. ∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001.

FIG. 1. Effect of gender and propensity to gossip on electronic word-of-mouth (eWOM) Intention for a High-Involvement Product (Athletic Shoes).

driven by agency-communion theory, which has been empirically tested less frequently, compared with other gender theories, in a marketing context. Our hypotheses testing results for H1a to H3c suggest that the interaction between gender and propensity to gossip is rather limited in terms of achievement value, friendship value, and normative pressure—none of the interactions were statistically significant. Our results indicate that both gender and propensity to gossip had significant main effects on friendship value. Through the use of SNSs, friendships among women and gossipers are facilitated more than friendships among men and nongossipers, respectively. However, in light of the statistical significance, the effects driven by gender seem to clearly exceed the effects driven by gossip. Thus, these findings are consistent with traditional communal interpersonal relationships among females (Wright, 2006; Zarbatany et al., 2004).

A lack of interaction between gender and propensity to gossip in terms of friendship value seems somewhat contradictory to Watson’s (2012) findings on a stronger relationship between gossip and friendship quality in men, compared with women. One possible explanation for this is that Watson measured friendship quality from the perspectives of “best friends” (e.g., “I have one or two particular best friends,” “I have several friends who I would call best friends”). In our study, friendship value was defined in a broader sense, which might have caused such perceptual differences. On the other hand, our findings show that propensity to gossip exhibited significant main effects on all three dependent variables, whereas gender did so only on friendship value. Statistical significance indicates that the impact of propensity to gossip on achievement value was solid, and the effect on normative pressure was rather strong. Furthermore, multivariate effects of propensity to gossip were greater than those of gender. These results seem to collectively indicate that gossipers generally seek agentic friendship, individual differentiation, and promotion of higher social standings through mastery or power, which is fairly consistent with prior research on gossip (Zarbatany et al., 2004). The issue of gender and agency-communion theory becomes more relevant when we look into the results of H4 and H5. In both hypotheses, propensity to gossip was a statistically significant predictor, whereas gender was not. Nonetheless, the outcome of H4 indicates that there is modest but clear interaction between gender and propensity to gossip for highinvolvement product. More specifically, male gossipers are more prone to disseminate promotional information through eWOM, compared with female gossipers. This interaction seems consistent with Watson’s (2012) view. That is, men connected with closer friendships (i.e., “friendship quality”) may have been more motivated to engage in eWOM due to higher involvement. In fact, this interaction was not observed for lowinvolvement product. Because Watson did not address eWOM, this finding clearly adds incremental value to the literature on gossip and gender. Managerially, online marketing managers could learn some lessons from this research. One obvious implication is the targeting issue in brand campaigns on SNSs. We argue that eWOM or viral marketing campaigns can be more effectively executed if we better understand the gossiping behavior of SNS users. Although industry journals occasionally report online gossip on SNSs as an anecdotal story, much more attention needs to be paid, because small talk could be a powerful initiator of subsequent product or brand recommendations. In terms of friendship value, firms may be interested in men’s gossiping behavior when they are connected with stronger ties. As our results indicate, gossipers—in particular the male segment—are more likely to seek agentic friendship, where a series of eWOM behaviors may be executed. This aspect is especially important for high-involvement products, such as personal computers, automobiles, or home appliances. Because



the majority of SNS conversations are motivated by social grooming (Tufekci, 2008), the use of gossip for marketing purposes could generate unexpected but significant results. In this light, one of the most popular social grooming tools is probably Twitter. Twitter allows us to “tag,” “tweet,” or “retweet” our friends over the Internet. Yet, according to Pear Analytics, most tweets are found to be “pointless babble” or conversational (40.6% and 37.6%, respectively), whereas only 3.6% are news (Boyd, 2009). Perhaps it is in this pointless babble where most gossip is buried. Close monitoring of these tweets may reveal the role of hidden talk in information dissemination. In fact, more and more firms are trying to develop capabilities for data mining on “big data.” Data mining is “the process of discovering meaningful new correlations, patterns and trends by sifting through large amounts of data stored in repositories, using pattern recognition technologies as well as statistical and mathematical techniques” (Larose, 2005). These data, which are collected as a by-product of normal consumption of interactive media, are referred to as big data. Our recommendation is for firms to employ these techniques to extract hidden content, in relation to our findings on gendered gossiping behavior. Such data mining could help firms effectively identify their target gender segment, leading to more successful SNS marketing strategies.

7. LIMITATIONS AND FUTURE RESEARCH SUGGESTIONS Despite these findings derived from our theories on online gossip and gender, it is important to recognize the exploratory nature of this research and its attendant limitations. In particular, two potential limitations of this research surround the procedure used in the scenario method. First, despite the rigorous pretest of product involvement, the two products used in this study—beer and athletic shoes—may reflect some masculine preference. We are aware that some may raise questions about the possible confounding of gender and product choices. Future research should address this issue by using more product variety to increase the study’s generalizability. Second, due to the nature of the questions, the respondents may have anticipated the study purpose and rated the eWOM intention in a socially desirable way. A possible solution to this bias may be the use of in situ experiments where we ask the respondents to use their actual SNS to see ads, assess their feelings and thoughts, and post their reactions on the site. Such experimentation may more accurately capture consumers’ verbal strategy related to brand promotion on SNSs. Another limitation of this research is the fact that the study was conducted in Spain and online behaviors may be culturally dependent. Any generalization on the applicability of the study should be treated with caution. In this regard, future extension in other countries could provide some evidence for the existence of a global tendency in online social grooming. Whereas prior research argues that gender differences in cognition and

725

behavior are universally applicable due to evolutionary and biological forces (e.g., Ellis, 2006), little research has tested such a thesis in a context of gossip or eWOM on SNSs. This is beyond the scope of this article, and thus should be explored in the future. FUNDING This research was funded by a grant from the Spanish Ministry of Science and Innovation (National Plan for Research, Development and Innovation ECO2011-30105). REFERENCES Ajzen, I. (1991). Theory of planned behavior. Organizational Behavior and Human Decision Processes, 50, 179–211. Algesheimer, R., Dholakia, U. M., & Herrmann, A. (2005). The social influence of brand community: Evidence from European car clubs. Journal of Marketing, 69(7), 19–34. Arno, A. (1980). Fijian gossip as adjudication: A communicative model of informal social control. Journal of Anthropological Research, 36, 343–360. Bakan, D. (1966). The duality of human existence: Isolation and communion in Western man. Chicago, IL: Rand McNally. Boyd, D. (2009, August 16). Twitter: “Pointless babble” or peripheral awareness + social grooming? Apophenia. Retrieved from http://www.zephoria.org/ thoughts/archives/2009/08/16/twitter_pointle.html Brown, J., Broderick, A. J., & Lee, N. (2008). Word of mouth communication within online communities: Conceptualizing the online social network. Journal of Interactive Marketing, 21(3), 2–20. Brunel, F. F., & Nelson, M. R. (2000). Explaining gendered responses to “helpself” and “help-others” charity ad appeals: The mediating role of worldviews. Journal of Advertising, 29(3), 15–28. Cho, S., Huh, J., & Faber, R. J. (2014). The influence of sender trust and advertiser trust on multistage effects of viral advertising. Journal of Advertising, 43(1), 100–114. Chodorow, N. J. (1978). The reproduction of mothering. Berkeley: University of California Press. Choi, G., Chung, H., & Kim, Y. (2012). Are stereotypes relative to gender usage applicable to virtual worlds? International Journal of Human–Computer Interaction, 28, 399–405. Conner, M., Smith, N., & McMillan, B. (2003). Examining normative pressure in the theory of planned behavior: Impact of gender and passengers on intentions to break the speed limit. Current Psychology, 22, 252–263. Dabholkar, P. A., & Bagozzi, R. P. (2002). An attitudinal model of technologybased self-service: Moderating effects of consumer traits and situational factors. Journal of the Academy of Marketing Science, 30, 184–201. Debatin, B., Lovejoy, J. P., Horn, A. K., & Hughes, B. N. (2009). Facebook and online privacy: Attitudes, behaviors, and unintended consequences. Journal of Computer-Mediated Communication, 15, 83–108. Dholakia, U. M., Bagozzi, R. P., & Pearo, L. K. (2004). A social influence model of consumer participation in network- and small-group-based virtual communities. International Journal of Research in Marketing, 21, 241–263. Dubois, D., Rucker, D. D., & Tormala, Z. I. (2011). From rumors to facts, and facts to rumors: The role of certainty decay in consumer communications. Journal of Marketing Research, 48, 1020–1032. Dunbar, R. I. M., Marriott, A., & Duncan, N. D. C. (1997). Human conversational behavior. Human Nature, 8, 231–246. Eagly, A. H. (1978). Sex differences in influenceability. Psychological Bulletin, 85, 86–116. Ellis, L. (2006). Gender differences in smiling: An evolutionary neuroandrogenic theory. Physiology & Behavior, 88, 303–308. Foster, E. K. (2004). Research on gossip: Taxonomy, methods, and future directions. Review of General Psychology, 8, 78–99. Gill, S., Stockard, J., Johnson, M., & Williams, S. (1987). Measuring gender differences: The expressive dimension and critique of androgyny scales. Sex Roles, 17, 375–400.


726

S. OKAZAKI ET AL.

Guerin, B. (2003). Language use as social strategy: A review and an analytic framework for the social sciences. Review of General Psychology, 7, 251–298. Guerin, B., & Miyazaki, Y. (2006). Analyzing rumors, gossip, and urban legends through their conversational properties. Psychological Record, 56, 23–33. Hall, J. A. (2011). Sex differences in friendship expectations: A meta-analysis. Journal of Social and Personal Relationships, 28, 723–747. Hennig-Thurau, T., Gwinner, K. P., Walsh, G., & Gremler, D. D. (2004). Electronic word-of-mouth via consumer-opinion platforms: What motivates consumers to articulate themselves on the Internet? Journal of Interactive Marketing, 18, 38–52. Hoffman, L. W. (1972). Early childhood experiences and women’s achievement motives. Journal of Social Issues, 28, 129–155. Johnson, R. D., Veltri, N. F., & Hornik, S. (2008). Attributions of responsibility toward computing technology: The role of interface social cues and user gender. International Journal of Human–Computer Interaction, 24, 595–612. José-Cabezudo, R. S., & Camarero-Izquierdo, C. (2012). Determinants of opening-forwarding e-mail messages. Journal of Advertising, 41, 97–112. Kamins, M. A., Folkes, V. S., & Perner, L. (1997). Consumer responses to rumors: Good news, bad news. Journal of Consumer Psychology, 6, 165–187. Killeen, L. A., López-Zafra, E., & Eagly, A. H. (2006). Envisioning oneself as a leader: Comparisons of women and men in Spain and the United States. Psychology of Women Quarterly, 30, 312–322. Kirtley, M. D., & Weaver, J. B. (1999). Exploring the impact of gender role selfperception on communication style. Women’s Studies in Communication, 22, 190–209. Laaksonen, P. (1994). Consumer involvement: Concepts and research. London, UK: Routledge. Larose, D. T. (2005). Discovering knowledge in data: An introduction to data mining. New York, NY: Wiley & Sons. Litman, J. A., & Pezzo, M. V. (2005). Individual differences in attitudes towards gossip. Personality and Individual Differences, 38, 963–980. Martin, B. A. S., & Gnoth, J. (2009). Is the Marlboro man the only alternative? The role of gender identity and self-construal salience in evaluations of male models. Marketing Letters, 20, 353–367. McAndrew, F. T., & Jeong, H. S. (2012). Who does what on Facebook? Age, sex, and relationship status as predictors of Facebook use. Computers in Human Behavior, 28, 2359–2365. Minton, C., Kagan, J., & Levine, J. A. (1971). Maternal control and obedience in the two-year-old. Child Development, 42, 1873–1894. Mittal, B. (1989). Measuring purchase-decision involvement. Psychology and Marketing, 6, 147–162. Putrevu, S. (2004). Communicating with the sexes: Male and female responses to print advertisements. Journal of Advertising, 33(3), 51–62. Richins, M. L., & Root-Shaffer, T. (1988). The role of involvement and opinion leadership in consumer word-of-mouth: An implicit model made explicit. Advances in Consumer Research, 15, 32–36. Rodriguez, S. (2012, June 6). Facebook users ignore ads and spend less time on site, poll says. The Los Angeles Times. Retrieved from http://articles.latimes. com/ Stirling, R. B. (1956). Some psychological mechanisms operative in gossip. Social Forces, 34, 262–267. Sundaram, D. S., Mitra, K., & Webster, C. (1998). Word-of-mouth communications: A motivational analysis. Advances in Consumer Research, 25, 527–531. Tannen, D. (1990). You just don’t understand: Women and men in conversation. New York, NY: William Morrow. Tufekci, Z. (2008). Grooming, gossip, Facebook and MySpace: What can we learn about these sites from those who won’t assimilate? Information Communication and Society, 11, 544–564. Venkatesh, V., & Morris, M. G. (2000). Why don’t men ever stop to ask for directions? Gender, social influence, and their role in technology acceptance and usage behavior. MIS Quarterly, 24, 115–139.

Verhoef, P. C., Franses, P. H., & Hoekstra, J. C. (2002). The effect of relational constructs on customer referrals and number of services purchased from a multiservice provider: Does age of relationship matter? Journal of the Academy of Marketing Science, 30, 202–216. Watson, D. C. (2012). Gender differences in gossip and friendship. Sex Roles, 67, 494–502. We are social. (2014, February 5). Social, digital & mobile in Europe in 2014. Retrieved from http://wearesocial.net/blog/2014/02/social-digital-mobileeurope-2014/ Weiser, E. B. (2000). Gender differences in Internet use patterns and Internet application preferences: A two-sample comparison. CyberPsychology & Behavior, 3, 167–178. Wiggins, J. S. (1992). Agency and communion as conceptual coordinates for the understanding and measurement of interpersonal behavior. In W. M. Grove & D. Cicchetti (Eds.), Thinking clearly about psychology (pp. 89–113). Minneapolis: University of Minnesota Press. Williams, T. G., & Slama, M. E. (1995). Market mavens’ purchase decision evaluative criteria: Implications for brand and store promotion efforts. Journal of Consumer Marketing, 12(3), 4–21. Wright, P. H. (2006). Towards an expanded orientation to the comparative study of women’s and men’s same-sex friendships. In K. Dindia & D. J. Canary (Eds.), Sex differences and similarities in communication (2nd ed., pp. 37–57). Mahwah, NJ: Erbaum. Xu, B., Li, D., & Shao, B. (2012). Knowledge sharing in virtual communities: A study of citizenship behavior and its social-relational antecedents. International Journal of Human–Computer Interaction, 28, 347–359. Zaichkowsky, J. L. (1985). Measuring the involvement construct. Journal of Consumer Research, 12, 341–352. Zarbatany, L., Conley, R., & Pepper, S. (2004). Personality and gender differences in friendship needs and experiences in preadolescence and young adulthood. International Journal of Behavioral Development, 28, 299–310.

ABOUT THE AUTHORS Shintaro Okazaki is Associate Professor of Marketing at Universidad Autónoma de Madrid, Spain. His work has appeared in leading marketing and advertising journals. He serves on the editorial review boards of International Journal of Advertising, Journal of Advertising Research, and Journal of Public Policy & Marketing, among others. He is currently Editor-in-Chief of Journal of Advertising. Natalia Rubio is Associate Professor of Marketing at Universidad Autónoma de Madrid, Spain. Her research interests focus on brand management and retailing. She has been published in leading academic journals, such as International Journal of Market Research; Journal of Business-to-Business Marketing; Food Quality and Preference; British Food Journal; and Cyberpsychology, Behavior, and Social Networking, among others. Sara Campo is Associate Professor of Marketing at Universidad Autónoma de Madrid, Spain. Her research interests lie in tourism marketing and service marketing. She has been published in leading academic journals, such as International Journal of Market Research; Cyberpsychology, Behavior, and Social Networking; Journal of Electronic Commerce Research; Journal of Business-to-Business Marketing; and Tourism Management, among others.