Unique but integrated: The role of individuation and assimilation processes in teen opinion leadership

Journal of Business Research 67 (2014) 83–91

Contents lists available at ScienceDirect

Journal of Business Research

Unique but integrated: The role of individuation and assimilation processes in teen opinion leadership☆ Elodie Gentina a, d,⁎, Raphaëlle Butori b, Timothy B. Heath c a

Skema Business School, Avenue Willy Brandt, 59 777 Euralille, France ESSEC Business School, Avenue Bernard Hirsch, B.P. 50105, 95021 Cergy-Pontoise, France HEC, 1, rue de la Libération, 78351 Jouy en Josas, France d Skema Business School, University Lille Nord de France, F-59000 Lille, France b c

a r t i c l e

i n f o

Article history: Received 31 March 2012 Received in revised form 12 November 2012 Accepted 17 November 2012 Available online 11 December 2012 Keywords: Opinion leadership Adolescent consumers Social network Need for uniqueness Peer influence

a b s t r a c t Opinion leaders propel the diffusion of innovation and exert a significant influence on the marketplace. This influence is especially pronounced during adolescence, a period marked by increasing reliance on peers and the emergence of a tension between two countervailing needs: assimilation and individuation. A survey of 1142 adolescents reveals that these developmental needs affect adolescent opinion leadership in the critical clothing market. Adolescent opinion leadership relies on a balance between desires for assimilation (i.e., centrality within the peer network) and individuation (i.e., need for uniqueness); adolescents' susceptibility to peers' normative influence and gender moderate these relationships. Adolescents who occupy central positions within their peer network tend to be opinion leaders, though only if they are not susceptible to normative influence. Position within the peer network is a key for girls, whereas need for uniqueness is a key for boys. These differences implicate different approaches for managers targeting adolescent males and females. © 2012 Elsevier Inc. All rights reserved.

1. Introduction Of all the transitions that occur in life, adolescence is one of the most crucial because this is the time when adolescents emerge from the family setting in order to establish an identity separate from parents (Erikson, 1968; Yang & Laroche, 2011). To do so, they increasingly rely on new significant others: their peers (Goodrich & Mangleburg, 2010; Mangleburg, Doney, & Bristol, 2004). Apart from providing emotional support (Buhrmester, 1996), peer group membership helps adolescents reconcile two competing needs essential to the transition from adolescence to adulthood: the need for assimilation and the need for individuation (Brewer, 1991; Bristol & Mangleburg, 2005). The aim of this research is to show that the expression of these two needs translates into a disposition particularly interesting to marketers, that of opinion leaders. Opinion leaders are members of a social system who pass the information they receive from the source on to the other members of the social system (Katz & Lazarsfeld, 1965). Because they are significant contributors to the diffusion and adoption of innovations (Childers, 1986; Watts & Dodds, 2007), they have

☆ The authors thank Samuel K. Bonsu, York University; Arnaud De Bruyn, ESSEC Business School; and Sonja Prokopec, ESSEC Business School, for reading and commenting on an early draft of this article. ⁎ Corresponding author. Tel.: +33 6 88 78 84 47. E-mail addresses: [email protected] (E. Gentina), [email protected] (R. Butori), [email protected] (T.B. Heath). 0148-2963/$ – see front matter © 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jbusres.2012.11.013

inspired a large body of empirical studies. Most of these studies provide a better understanding of the opinion leadership concept and map the positions that opinion leaders occupy within their social networks (Kratzer & Lettl, 2009; Lee, Cotte, & Noseworthy, 2010). However, these studies suffer from two limitations. First, they focus on adults or children, not on adolescents. Therefore, they overlook a segment whose vast purchasing power renders it particularly worthy of investigation. Second, they study assimilation and individuation motives separately (Chan, Berger, & Van Boven, 2012), thereby failing to investigate the question of how these two conflicting motives may combine in the opinion leadership process. In response, the current study develops an integrative perspective that examines how two variables associated with the needs for assimilation and individuation—the adolescent's position in a peer network and her/his need for uniqueness—shape adolescent opinion leadership. In addition, because conformity to a peer group is particularly salient during adolescence, and adolescent boys and girls proceed along different developmental pathways (Chodorow, 1978), the present study also incorporates two moderators whose influence on adolescent opinion leadership has been ignored: susceptibility to peers' normative influence and gender. Results indicate that centrality in a network is associated with opinion leadership, but only among adolescents with low levels of susceptibility to peer influence. In addition, opinion leadership is associated with need for uniqueness but only among teenage boys. These findings extend existing research on opinion leadership and gender differences, as well as provide concrete guidelines for the implementation of efficient marketing strategies.

84

E. Gentina et al. / Journal of Business Research 67 (2014) 83–91

The next section (Section 2) develops theory based on prior assimilation and individuation research. Sections 3 and 4 detail, respectively, the method and results of a study conducted among 1142 adolescents. Because adolescence is a time when individuals are particularly concerned with their appearance (they need to incorporate the changes that their bodies undergo in their views of themselves, Piacentini, 2010; Rose, Boush, & Friestad, 1998), this study examines opinion leadership in the context of clothing. Finally, Section 5 discusses the theoretical and practical implications of this study, as well as provides managerial suggestions for practitioners wishing to target adolescent consumers. 2. Theoretical framework and hypotheses The formation of a strong and coherent sense of identity is the primary task associated with the transition from adolescence to adulthood (Erikson, 1968). The development of the self hinges on two related processes: (1) assimilation with peers who help define the individual within a social context and (2) individuation through the recognition of the self as a distinct entity and corresponding independence in behaviors (Brewer, 1991). Whereas the need for assimilation can be satisfied through one's group membership and identifying with this in-group, the need for individuation can be fulfilled by distinguishing the in-group from out-groups (Brewer, 1991). The proposed framework suggests that both needs for assimilation and individuation affect adolescents' opinion leadership.

Childers & Rao, 1992; Mascarenhas & Higby, 1993). Friends help adolescents evaluate products, brands, and stores in ways that enhance each teen's sense of belonging while also establishing an identity separate from parents and from out-groups (Goodrich & Mangleburg, 2010). Accordingly, SNI has a primary influence on adolescent consumers' behaviors (Kurt, Inman, & Argo, 2011). Although being central in a network enables teens to gather and update relevant information, it is unlikely that such central teens will be opinion leaders if they fail to transcend the normative influence of their peers and instead seek validation from their peers. It is difficult to be an opinion leader if one's opinions derive from the opinions of others, as is the case with people high on SNI (Mangleburg et al., 2004). In contrast, central adolescents that are low on SNI are more likely to think and act independently and thus share opinions that they have developed on their own. These consumers combine two assets essential to the expression of opinion leadership: access to relevant information (and credibility) in the field and independence from others, which provides enough confidence to engage in product and brand conversations. Accordingly, SNI should moderate the relationship between degree centrality and opinion leadership. H2. The positive relationship between adolescent degree centrality and opinion leadership is stronger for adolescents with low rather than high SNI. 2.2. Need for individuation and opinion leadership

2.1. Need for assimilation and opinion leadership During adolescence, the construction of the self takes place through significant relationships with peers (Gecas & Steff, 1990). As parental influence fades, peer groups become important socialization agents that provide new rules, standards and beliefs about appropriate conducts (Turner, Hogg, Oakes, Reicher, & Wetherell, 1987). Group membership also relieves the pain associated with teens' narcissistic vulnerability, and provides a transitional sense of comfort and security (O'Donnell & Wardlow, 2000). By adopting the rules of the group to which they aspire to belong, adolescents assimilate with the peer group and come to occupy strategic positions within the group (Mangleburg et al., 2004; Valente, Unger, & Johnson, 2005). Therefore, the position that an adolescent occupies within the peer network is a reflection of underlying assimilation concerns. This position, in turn, can be characterized by its degree of centrality, the number of ties that directly link a person within a network to the other members of this network adjusted for the total number of people in the network (Freeman, 1979). High degree centrality in a friendship network means that the individual is popular, such that degree centrality connotes a high level of acceptance within the network (Wasserman & Faust, 1994). It also provides focal individuals with the advantage of accumulating higher quality information (Lee et al., 2010), which reinforces their expertise and credibility in the field (Hinz, Schulze, & Takac, forthcoming). By occupying central positions, adolescents not only develop an ability to influence others, but also have access to a network of peers through which they can exercise this ability. We thus expect to replicate, on an adolescent population, the relationship that Kratzer and Lettl (2009) and Lee et al. (2010) identified on children and adults respectively: H1. Adolescent degree centrality relates positively to opinion leadership. However, because during their adolescence, people depend heavily on peers, this relationship may be moderated by susceptibility to normative influence (SNI), that is, the extent to which people hope to conform to the expectations of others to attain a desired image (Bearden & Etzel, 1982; Bearden, Netemeyer, & Teel, 1989). In a consumer context, several studies show that adolescents' product preferences depend on those of peers (e.g., Bachmann, John, & Rao, 1993;

In addition to the need for social assimilation, people seek to differentiate themselves from others to define the self as a unique entity, distinct from others (White & Argo, 2009). This need for individuation emerges as adolescents break away from parental influence (Nuttall and Tinson, 2008; Yang & Laroche, 2011) as they think and act more independently (Piacentini, 2010). Such individuation balances with assimilation as adolescents evolve into adults, in part by becoming independent thinkers but also by meshing with their cohort and groups. Because the need for individuation involves seeking and recognizing unique features, this study uses consumer need for uniqueness (CNFU) (Tian, Bearden, & Hunter, 2001) to operationalize the need for individuation. Adolescents high in CNFU should desire unique, counternormative, or unconventional options (Tian & McKenzie, 2001; Tian et al., 2001) while avoiding products with significant mass appeal (Berger & Heath, 2007). Therefore, CNFU links closely to the maintenance and display of individuation. By wearing unique products or their combinations, adolescents with high CNFU signal a clear sense of their own personal identity and sufficient confidence to express that identity. That is, a high level of CNFU contributes to the development of desirable identities that others are likely to admire and consult. Consistent with previous research (e.g., Chan & Misra, 1990, Fromkin, 1971; Snyder & Fromkin, 1980), we therefore hypothesize that the willingness to publicly individuate oneself is a necessary trait of an effective opinion leader. H3. Adolescent CNFU relates positively to opinion leadership. SNI, however, should again moderate this relationship. Credibility is a key to the development of opinion leadership, and opinion leaders must send coherent signals. By simultaneously pursuing uniqueness (high CNFU) and detaching from the norms of their peer groups (low SNI), adolescents send such signals and demonstrate confidence in their choices, which further increases the desirability of their opinions and preferences. In contrast, when adolescents pursue uniqueness (high CNFU) but need the approval of peers (high SNI), perceived credibility as an independent thinker wanes, thereby blunting the individual's impact on others. Accordingly: H4. The positive relationship between CNFU and opinion leadership is stronger for adolescents with low rather than high SNI.

E. Gentina et al. / Journal of Business Research 67 (2014) 83–91

2.3. The role of gender in opinion leadership Although gender has been a topic of interest in marketing research, its potential moderating effects on opinion leadership have yet to be explored (Cho & Workman, 2011; Clark & Goldsmith, 2005). In addition, studies investigating the impact of gender on opinion leadership, CNFU, and an individual's position in a network, offer mixed results. For example, while some studies suggest that males are “better” opinion leaders than females (Nieva & Gutek, 1980), others do not find any gender difference in opinion leadership (Kratzer & Lettl, 2009; Vernette, 2004). And whereas Tian et al. (2001) find no evidence that gender is related to CNFU, Soomro, Gilal, and Jatoi (2011) argue that males give more preference to unique products than females. Finally, no empirical study explicitly investigates the relationship between an individual's position in his or her network, though Kratzer and Lettl (2009) argue that gender influences social networks because females have more frequent contact with friends than do males. These inconclusive findings highlight the need for more research that explores these variables and their relationships. A large body of research in psychology suggests that males and females pursue fundamentally different paths to identity formation. Starting from early childhood, socialization processes shape male and female gender identities (Lindsey, 2011) and differentially emphasize communal versus agentic goals (Bakan, 1996). Communal goals refer to a feminine identity, with a focus on social relationships, interpersonal affiliation, and harmony with others. Agentic goals refer to a masculine identity and link to assertiveness, control, and selfassertion (Carlson, 1971; Spence & Helmreich, 1978). Whereas female identity is structured by themes of social belonging and affiliation, male identity reflects differentiation, separation, and autonomy (Chodorow, 1978). These different gender identities provide a conceptual foundation for several marketing studies (McColl-Kennedy & Anderson, 2005; Thompson, 1996; Winterich, Mittal, & Rose, 2009), especially pertaining to clothing consumption. For example, Cox and Dittmar (1995) find that clothes fulfill different social psychological functions for women and men: Men tend to value clothing in a functional and individuating fashion, whereas women have more symbolic and other-oriented perspectives. For men, clothes provide a means to express individuality; for women, clothes offer a way to feel integrated to others. Because men take a more self-oriented approach to clothes, adolescent boys should value the ability to exhibit uniqueness through clothing consumption more than adolescent girls. Because women take an other-oriented approach, adolescent girls should consider clothing consumption a way to be socially integrated (Chafetz, 2006). Therefore, they should be particularly inclined toward holding a central position within the peer social network in order to forge diverse ties with others. Accordingly, we expect gender to moderate the relationships between CNFU and opinion leadership and between degree centrality and opinion leadership. H5. The positive relationship between adolescent degree centrality and opinion leadership is stronger for adolescent girls than for adolescent boys. H6. The positive relationship between CNFU and opinion leadership is stronger for adolescent boys than for adolescent girls. 3. Method 3.1. Data collection Because school is the primary location where adolescents interact every day with their peers (Mascarenhas & Higby, 1993), and schoolmates provide a major source of information for opinion leaders (Hansen & Hansen, 2005), the school class is a relevant unit of analysis to identify peers' networks and study influence processes (Lubbers,

85

2003). A questionnaire was distributed to 1142 adolescents (549 girls, 593 boys) across 37 school classes in an urban region in northern France. The selection of urban area schools ensured adequate representation across various background characteristics (e.g., family structure, living situations). This sample was then randomly divided into two sub-samples. The first sample (n1 = 399) served to validate the measurement scales. The second sample (n2 = 743) provided the input to test the proposed model. It encompassed 24 school classes distributed across seven public and private schools, and consisted of 149 students from grade 8, 185 students from grade 9, 127 students from grade 10, 159 students from grade 11, and 123 students from grade 12 (ages ranged from 13 to 18 years, with an average of 15.2 years). The administration of the survey took place in classrooms, in the presence of one of the researchers, who instructed the students not to discuss their responses with classmates and indicated that the purpose of the research was to “learn what adolescents think about various consumer-related issues.” The respondents first indicated up to five of their closest friends in their classes (when asked to give more than five names, adolescents tend to give names at random, leading to an overestimation of affiliations (Rogers & Kincaid, 1981)), then completed scales measuring the individual variables under consideration. The network data, gathered using Ucinet software (Borgatti, Everett, & Freeman, 2002), produced a matrix of contact frequency among all adolescents in each class. The original square matrices placed all the students in rows and columns, such that a 0 in a cell indicated the absence and a 1 indicated the presence of a friendship tie between two adolescents. All school classes contained approximately 30 students, so there was no need to standardize the degree centrality measurements to account for different group sizes. Fig. 1 displays a sociogram that represents all friendship ties among a group of adolescents in the same school class. Some adolescents (e.g., Thibaut D) are situated on the extreme periphery; others (e.g., Ilias E) have many contacts with subgroup members and thus exhibit a greater degree of centrality. 3.2. Construct measures To measure opinion leadership, two instruments were used: a self-reported measure and a peer-reported measure. Because opinion leadership is most appropriately measured with domain-specific instruments, both instruments were adapted to the specific context of clothing (Flynn, Goldsmith, & Eastman, 1996). The first instrument (self-reported, hereafter OL-a) relied on the scale of Childers (1986), whose internal consistency and validity were demonstrated across countries (Ben Miled & Le Louarn, 1994; Goldsmith & Desborde, 1991; Vernette, 2004). The second instrument (peer-reported, hereafter “OL-b”) was created by asking teenagers to nominate the five adolescents in their school class they considered to be the best dressed. These data enabled to create appearance networks, where the interdependency between each adolescent was not based on friendship (as usually the case in social network studies) but on a judgment of appearance. In such networks, the most central adolescents are those most frequently considered by others as being the best dressed. OL-b was defined as the number of direct contacts that each adolescent had in his or her appearance network. In sum, OL-a and OL-b respectively measure opinion leadership from the focal adolescent's perspective and from others' perspective. To measure CNFU and SNI, we used Tian et al. (2001) and Bearden et al. (1989) scales respectively. Although the questionnaire included all three dimensions of CNFU (creative choice counterconformity, avoidance of similarity and unpopular choice counterconformity), the analysis used only the first two (Unpopular choice counterconformity characterizes consumers who tend to choose products and brands that deviate from group norms and thus risk social disapproval. Because teens place great value on peers' social approval (Piacentini,

86

E. Gentina et al. / Journal of Business Research 67 (2014) 83–91

Fig. 1. A social network of adolescents.

2010), pretests revealed extremely low variance on this dimension. We thus dropped it.). All scales, which were translated into French using a parallel-blind technique, and formatted in five-point Likert scales, are provided in Appendix 1. Finally, to assess adolescent's position in peer network, we chose the in-degree centrality measure. This measure uses the number of incoming arrows to each focal actor in a network to determine his or her popularity. Contrary to out-degree centrality (which rather uses the number of out-going arrows from the focal actor), in-degree centrality therefore measures the extent to which others in the network identified the focal actor as a close friend (Van Den Bulte & Joshi, 2007). As a peer-reported measure of popularity, it is not subject to social desirability bias, and thus provides the most accurate instrument to identify opinion leader with sociometric data (Lee et al., 2010; Van Den Bulte & Joshi, 2007; Wasserman & Faust, 1994). Technically speaking, in-degree centrality is computed as follows: Indegree ¼ ∑ aij ; i≠j

where aij, takes a value of 1 if an arrow moves from point j to point i.

eliminated. This process suggested a four-factor structure (opinion leadership, SNI, creative choice counterconformity, and avoidance of similarity). The Cronbach's alpha values ranged from .73 to .87. Gerbing and Anderson (1988) provided the guidelines for confirming the scales' factor structures and assessing reliabilities and convergent and discriminant validities. The estimations of the purified scales took place simultaneously in a multiple-measurement model. Confirmatory factor analyses (using AMOS) replicated the four-scale structure. The four-factor model also fit the data reasonably well (χ2 =208, df=98, pb .001, root mean square error of approximation [RMSEA]=.05, goodness-of-fit index=.94, comparative fit index [CFI]=.95, adjusted χ2 =2.13). The composite reliability coefficients (Jöreskog rhô) and discriminant validity indicators were satisfactory (see Table 1). 4.2. Test of the measurement model The second sample provided the evaluation of the measurement model (n2 = 743). Each indicator t-value exceeded 12.79 (p b .001), and all standardized factor loadings were greater than .61. The composite reliability (Jöreskog rhô) exceeded .73, and discriminant validity again received support. Table 2 summarizes these analyses.

4. Analyses and results 4.3. Test of the conceptual model 4.1. Measure validity and reliability All items passed the test of univariate quasi-normality (Kline, 2005), which allowed for exploratory and confirmatory factor analyses on the first sample (n1 = 399). The exploratory analysis used principle components with Oblimin rotation. Items that loaded on more than one factor and those with factor loadings below .50 were

Structural equation modeling (SEM) based on maximum likelihood estimation was used to estimate Fig. 2's conceptual model (n2 = 743). The tests for the moderating effects of SNI and gender used multi-group analyses within the SEM framework (Kline, 2005). A stepwise procedure assessed the invariance of the structural paths

E. Gentina et al. / Journal of Business Research 67 (2014) 83–91

for different groups. The first step imposed equality constraints on structural paths (χ 2H). The second step involved running a baseline model with no parameter constraint, resulting in a χ 2 with additional degrees of freedom (χ 2N). The test for the moderator effect took the difference in the two values (χ 2N − χ 2H = χ 2Δ), which produced a χ 2 value with degrees of freedom equal to the difference in degrees of freedom of the (χ 2H) and (χ 2N) values. For the multigroup analyses, a quartile split applied to the adolescents: Low SNI adolescents score in the lower quartile of the SNI distribution; high SNI adolescents score in the upper quartile. 4.4. Hypotheses testing Two sets of analyses were performed. The first one used the selfreported measure of opinion leadership (OL-a) as a dependant variable, the second one its peer-reported measure (OL-b). Except for one hypothesis (H5), these analyses reached extremely consistent results (Table 3). To avoid redundancy, we thus present the results obtained on OL-a, then present the unique difference found between OL-a and OL-b. Contrary to expectations, the analyses reveal a nonsignificant relationship between degree centrality and opinion leadership (γ1a= .04; t = 1.05; p = .29), which rejects H1. However, this outcome does not indicate that an adolescent's position in the network is unrelated to opinion leadership. Consistent with H2, the absence of a significant main effect actually reflects the moderating impact of SNI (χ2Δ(1) = 3.90, p b .05). Degree centrality relates significantly and positively to opinion leadership among adolescents with low SNI (γ2a= .15; t = 2.02; p b .05) but not among those with high SNI (γ2a= −.03; t = −.35; p = .72). Being central in a network is not sufficient to exert a significant influence on others. To qualify as opinion leaders, central teens must be able to detach from the normative influence of peers. As predicted, the analyses also highlight a significant and positive relationship between CNFU and opinion leadership (γ3a=.23; t = 4.62; p b .001), in support of H3. By publicly expressing individuality, adolescents demonstrate a strong and clear sense of their personal identity, which represents an asset for opinion leadership. Although the relationship between CNFU and opinion leadership is slightly stronger for adolescents with low (γ4a= .34; t= 3.83; p b .001) rather than high levels of SNI (γ4a= .31; t = 2.66; p b .05), the χ2 difference values do not

87

reach significance (χ2Δ(1) = .95, p = .32), in only partial support of H4. An expression of uniqueness is an asset, no matter how much adolescents worry about the norms and values imposed by peers. Consistent with H6, gender moderates the relationship between CNFU and opinion leadership (χ 2Δ(1) = 18.19, p b .001). CNFU relates positively to opinion leadership among boys (γ6a = .39; t = 5.14; p b .001) but not among girls (γ6a = − .01; t = − .16; p = .87). Finally, although the χ 2 difference does not reach significance (χ 2Δ(1) = .47, p = .49), the relationship between degree centrality and self-reported opinion leadership is significant and positive for girls (γ5a = .12; t = 1.95; p b .05) but not for boys (γ5a = .06; t = 1.21; p = .22), in partial support for H5. Contrary to girls, the adolescent boys most central in their networks do not necessarily consider themselves opinion leaders in the field of clothes. This finding supports a gender identity perspective, whereby female consumers set greater value on social belonging and relationships, while male consumers are more concerned with the display of uniqueness and autonomy (Chodorow, 1978; McColl-Kennedy & Anderson, 2005). However, when opinion leaders are identified by their peers, the relationship between degree centrality and peer-reported opinion leadership turns out to be significant for both adolescent boys and adolescent girls, meaning that centrality in a network is a cue that both adolescent boys and adolescent girls use to identify opinion leaders. 5. General discussion 5.1. Theoretical and methodological contributions Though the topic of opinion leadership is of interest to consumer researchers, little empirical research addresses opinion leadership during adolescence (Kratzer & Lettl, 2009). Yet research in this area is needed as adolescent consumers are qualitatively different from consumers in other age groups, both in the value they attach to their peer groups (assimilation) and their need to emerge as unique individuals (individuation). Contrary to prior studies (e.g., Kratzer & Lettl, 2009; Lee et al., 2010), the findings presented here highlight that the central position adolescents may occupy within a peer network is not sufficient to exert a significant influence on others. For adolescents to endorse opinion leader roles, they must be both connected to a great number of others and

Table 1 Results of the four-factor CFA model (Study 1, n1 = 399). Construct

Opinion leadership

NFU, creative choice

NFU, avoidance of similarity

SNI

Item

y1 y2 y3 y4 y5 y6 x1 x2 x3 x4 x5 x6 x7 x8 x9 x10

Completely standardized factor loading

t-value

.75 .70 .72 .72 .72 .76 .65 .74 .74 .77 .70 .58 .77 .65 .79 .67

Goodness-of-fit indices for confirmatory factor analysis χ2 (ddl) GFI AGFI 208.24 (98) .94 .91 ⁎ Used to provide the scale for the related construct. ⁎⁎ Shared variance among trait factors. ⁎⁎⁎ Correlations among trait factors.

Reliability (Jöreskog Rhô)

Convergent validity (Rhô VC)

⁎

.87

.53

13.45 13.82 13.82 13.86 14.69 ⁎

.75

.51

9.70 9.69 ⁎

.80

.50

12.70 10.65 13.67 ⁎

.75

.50

Discriminant validity SNI

NFU Avoidance of similarity

NFU Creative choice

Opinion leadership X

X

X

.11⁎⁎ (.34)⁎⁎⁎

X

.22 (.47)

.00 (−.00)

.03 (−.19)

.00 (.00)

.21 (.46)

10.23 9.94

RMSEA .05

CFI .95

TLI .94

SRMR .05

χ2/ddl 2.13

88

E. Gentina et al. / Journal of Business Research 67 (2014) 83–91

Table 2 Results of the four-factor CFA model (Study 2, n2 = 743). Construct

Opinion leadership

NFU, creative choice

NFU, avoidance of similarity

SNI

Item

y1 y2 y3 y4 y5 y6 x1 x2 x3 x4 x5 x6 x7 x8 x9 x10

Completely standardized factor loading

t-value

.78 .65 .66 .72 .69 .72 .71 .72 .76 .78 .74 .57 .82 .61 .77 .68

Goodness-of-fit indices for confirmatory factor analysis χ2 (ddl) GFI AGFI 384.12 (98) .94 .91

Reliability (Jöreskog Rhô)

Convergent validity (Rhô VC)

⁎

.86

.50

16.57 16.98 17.72 17.14 18.32 ⁎

.76

.52

15.81 16.23 ⁎

.82

.53

19.49 14.95 21.05 ⁎

.73

.48

Discriminant validity SNI

NFU Avoidance of similarity

NFU Creative choice

Opinion leadership X

X

X

.06⁎⁎ (.24)⁎⁎⁎

X

.35 (.59)

.00 (.04)

.03 (−.18)

.01 (−.11)

.23 (.48)

13.06 12.79

RMSEA .06

CFI .93

TLI .92

SRMR .05

χ2/ddl 3.92

⁎ Used to provide the scale for the related construct. ⁎⁎ Shared variance among trait factors. ⁎⁎⁎ Correlations among trait factors.

ready to deviate from the norms and values expressed by peers (low SNI). That is, opinion leaders are those adolescents who successfully combine a need for assimilation typical of adolescent years (Piacentini, 2010; Yang & Laroche, 2011) as well as a tendency to deviate from norms. By behaving and dressing in accordance with personal standards, adolescents demonstrate the robustness of their personal identities, which increases the likelihood that others admire and consult them. Results also indicate that adolescent opinion leadership relies on different mechanisms among boys and girls. Self-reported opinion leadership is associated with need for uniqueness among boys, but not

among girls, and with degree centrality among girls, but not among boys. These different patterns of results echo the different paths of identity construction highlighted in psychology (e.g., Chodorow, 1978) as well as extend existing findings in marketing (e.g., Winterich et al., 2009). The opposition between needs for social belonging and affiliation, which are typical of the female identity, and needs for differentiation and separation, which are typical of the masculine identity, are evident in the opinion leadership phenomenon during adolescence. Because adolescence is the most crucial stage in the gender identity developmental process (Ge, Conger, & Elder, 2001), and the emergence of this

H3 (γ3) H4 (γ4a, γ4b) H6 (γ6a , γ6b)

H2 (γ2a, 2b)

H5 (γ5a , γ5b) H1 (γ1)

Fig. 2. Structural model results (Study 1, n2 = 743).

E. Gentina et al. / Journal of Business Research 67 (2014) 83–91

89

Table 3 Main results of hypotheses testing (Study 2, n2 = 743). Self-reported measure of opinion leadership (OL-a) Standardized parameter estimates

Chi square difference (OL-a)

Peer-reported measure of opinion leadership (OL-b)

Chi square difference (OL-b)

p

Standardized parameter estimates

t-value

p

1.05 4.62

.29 b.001

γ1b = −.05 γ3b = .10

1.25 2.19

.21 b.05

γ5b = .15 γ6b = .28 γ6b = .02 γ6b = .16

2.57 5.14 .31 2.62

b.05 b.001 .75 b.05

χ2Δ(1) = 1.96, p = .16

t-value

Degree centrality→ OL CNFU → Opinion leadership

γ1a = .04 γ3a = .23

Moderator effect of gender Girls: Degree centrality →OL Boys: Degree centrality → OL Girls: CNFU →OL Boys: CNFU →OL

γ5a = .12 γ5a = .06 γ6a = −.01 γ6a = .39

1.95 1.21 −.16 5.14

b.05 .22 .87 b.001

γ2a = .15 γ2a = −.03

2.02 −.35

b.05 .72

χ2Δ(1) = 3.90, p b .05

γ2b = .28 γ2b = .22

4.05 2.81

b.001 b.05

χ2Δ(1) = .23, p = .63

3.83 2.66

b.001 b.05

χ2Δ(1) = .95, p = .32

γ4b = .11 γ4b = .08

1.41 .92

b.10 .35

χ2Δ(1) = .01, p = .92

Moderator effect of SNI Low level of SNI: Degree centrality →OL High level of SNI: Degree centrality →OL Low level of SNI: CNFU → OL High level of SNI: CNFU →OL

γ4a = .34 γ4a = .31

identity exacerbates gender differences (Archer, 1989), we have reason to believe that the moderating effects identified in our research might not be transferred to adults. Finally, this research uses two tools of different natures to identify opinion leaders: a self-reported measure (Childers' scale), and a peerreported measure, with extremely consistent results. Whereas the former has been extensively used in the literature, the latter has been specifically designed for the study. Its reliance on network data makes it an innovative and interesting tool to identify opinion leaders from the perspective of others. This issue is of particular importance as the identification of opinion leaders is one of the most crucial phases in the study of opinion leaders (Kratzer & Lettl, 2009). 5.2. Managerial implications As objects of targeted communications, adolescent consumers are more valuable than others, particularly those who influence peers (Feick & Price, 1987). Adolescent opinion leaders represent a particularly attractive target because of their tendency to spread information quickly throughout the market (Piacentini, 2010). Therefore, this research points out the need to understand the motives that drive adolescents to be opinion leaders and provides a set of cues they might use to attract this growing segment more effectively. Specifically, the moderator role of gender adds a refinement that merits managerial consideration. To encourage strong relationships with male opinion leaders, while satisfying their desire for unique products, businesses might launch special models in limited series. Sellers could encourage male consumers to customize products online; a group customization option could allow boys to design a logo or verbal expression shared across their group but modified in unique ways by each member. Invitations to private events, fashion shows, or precollection launches also might satisfy male opinion leaders' need for uniqueness. Beyond stimulating the feeling of being special, these events enable the targeted adolescents to gain access to new and trendy information, which reinforces their credibility as opinion leaders. In contrast, female opinion leaders seek to occupy central positions in their peer network and thus are more likely to respond to marketing efforts that communicate peer acceptance associated with a product's consumption. In terms of distribution, retailers could enlarge and adapt changing rooms to create a convivial atmosphere (e.g., spotlights, music, sofas) and encourage female adolescents to try on clothes together. Clothing stores should be more than collections of clothing items; they should represent a teenaged

χ2Δ(1) = .47, p = .49 χ2Δ(1) = 18.19, p b .001

χ2Δ(1) = 3.99, p b .05

consumption experience, with entertainment, social interaction, and socialization aspects (Bloch, Ridgway, & Dawson, 1994). Finally, marketers should turn to viral marketing on Internet sites such as Facebook or My Space to promote new products or brands among female opinion leaders, which enables them to purchase products online while remaining constantly virtually connected with friends. 5.3. Limitations and further research The present research assessed only one product category, clothing. Although clothing is particularly relevant to adolescents, who greatly value clothing and appearance (Martin & Gentry, 1997), additional research should assess other product categories to improve the generalizability of the findings. In particular, need for uniqueness effects may be greater in the visible clothing product category than in less public categories (Tian & McKenzie, 2001). The present study was also constrained by the limited number of adolescents that each respondent could cite when asked to mention his or her closest friends (5), as well as by the social network's unit of analysis (the school class). Since high school students may have more than 5 close friends outside the classroom, future research is needed to examine wider networks, with friendship ties at the grade or entire school level. This research took place in France, an independence-oriented culture. More interdependent cultures might provide interesting insights. For example, in collectivistic Asian cultures, parents often exert more control over their children for more time (Rose, 1999), and children feel motivation to perform duties imposed by the collective, prioritize collective goals, and emphasize connectedness (Triandis, 1995). Researchers might investigate how the combined needs for social assimilation and individuation shape adolescents' opinion leadership in collectivist cultures. Individuation may fall by the wayside, as might opinion leadership altogether, if the culture sufficiently discounts individual orientations. References Archer, S. L. (1989). Gender differences in identity development: Issues of process, domain and timing. Journal of Adolescence, 12, 117–138. Bachmann, G. R., John, D. R., & Rao, A. R. (1993). In L. McAlister, & M. L. Rothschild (Eds.), Children's susceptibility to peer group purchase influence: An exploratory investigation. Advances in consumer research, 20. (pp. 463–468) Provo, UT: Association for Consumer Research. Bakan, D. (1996). The duality of human existence. Chicago: Rand McNally.

90

E. Gentina et al. / Journal of Business Research 67 (2014) 83–91

Appendix 1. Measurement scales (n2 = 743) Measure

Items

Mean

SD

Need for uniqueness, creative choice counterconformity

I have sometimes purchased unusual clothes as a way to create a more distinctive personal image. I often look for one-of-a-kind clothes or brands so that I create a style that is all my own. I actively seek to develop my personal uniqueness by buying special clothes. The more commonplace clothes or brands are among the general population, the less interested I am in buying it. As a rule, I dislike clothes or brands that are customarily purchased by everyone. When clothes or brands I like become extremely popular, I lose interest in them. I avoid clothes or brands that have already been accepted and purchased by the average consumer. If my friends can see me using clothes, I often purchase the clothing brand they expect me to buy. It is important that my friends like the clothes and brands I buy. I like to know what clothes or brands make a good impression on my friends. I talk to my friends about clothes very often. In discussions of clothes, what happens most often is that I tell my friends about clothes. Overall in all of my discussions with friends, I am often used as a source of advice. During the past six months, I have told a number of people about clothes. When I talk to my friends about clothes, I give a great deal of information. Compared with my circle of friends, I am very likely to be asked about clothes.

2.77 2.89 2.97 3.14

1.04 1.09 1.05 1.23

2.71 2.46 2.93 2.61 3.03 3.15 2.89 2.44 3.07 2.60 2.65 2.56

1.16 1.12 1.16 1.23 1.18 1.09 1.18 1.00 .09 1.24 1.03 .98

Need for uniqueness, avoidance of similarity

Susceptibility to the normative influence of peers

Opinion leadership

Bearden, W. O., & Etzel, M. J. (1982). Reference group influence on product and brand purchase decisions. Journal of Consumer Research, 9(2), 183–194. Bearden, W. O., Netemeyer, R. G., & Teel, J. E. (1989). Measurement of consumer susceptibility to interpersonal influence. Journal of Consumer Research, 15(4), 473–481. Ben Miled, H., & Le Louarn, P. (1994). Analyse comparative de deux échelles de mesures du leadership d'opinion: Validité et interprétation. Recherche et Applications en Marketing, 9(4), 23–51. Berger, J., & Heath, C. (2007). Where consumers diverge from others: Identity-signaling and product domains. Journal of Consumer Research, 34(2), 121–134. Bloch, P. H., Ridgway, N. M., & Dawson, S. A. (1994). The shopping mall as consumer habitat. Journal of Retailing, 70(1), 23–42. Borgatti, S. P., Everett, M. G., & Freeman, L. C. (2002). UCINET 6 for Windows. Harvard, MA: Analytic Technologies. Brewer, M. B. (1991). The social self: On being the same and different at the same time. Personality and Psychology Bulletin, 17(5), 475–482. Bristol, T., & Mangleburg, T. F. (2005). Not telling the whole story: Teen deception in purchasing. Journal of the Academy of Marketing Science, 33, 79–95. Buhrmester, D. (1996). Need fulfillment, interpersonal competence, and the developmental contexts of early adolescent friendship. In W. M. Buhrmester, A. F. Newcomb, & W. W. Hartup (Eds.), The company they keep: Friendship in childhood and adolescence (pp. 158–185). Cambridge: Cambridge University Press. Carlson, R. (1971). Sex differences in ego functioning: Exploratory studies of agency and communion. Journal of Consulting and Clinical Psychology, 37, 267–277 (October). Chafetz, J. S. (2006). Handbook of the sociology of gender. New-York: Springer. Chan, C., Berger, J., & Van Boven, L. (2012). Combining social identity and uniqueness motives in choice. Journal of Consumer Research, 39(3), 561–573. Chan, K. S., & Misra, S. (1990). Characteristics of the opinion leader: A new dimension. Journal of Advertising, 19(3), 53–61. Childers, T. L. (1986). Assessment of the psychometric properties of an opinion leadership scale. Journal of Marketing Research, 23(2), 184–188. Childers, T. L., & Rao, A. R. (1992). The influence of familial and peer-based reference groups on consumer decisions. Journal of Consumer Research, 19(2), 198–211. Cho, S., & Workman, J. (2011). Gender, fashion innovativeness and opinion leadership, and need for touch — Effects on multi-channel choice and touch/non-touch preference in clothing shopping. Emerald Group Publishing Limited, 15(3), 363–382. Chodorow, N. J. (1978). The reproduction of mothering. Berkeley: University of California Press. Clark, R., & Goldsmith, R. (2005). Market mavens: Psychological influences. Journal of Psychology and Marketing, 22(4), 289–312. Cox, J., & Dittmar, H. (1995). The functions of clothes and clothing (dis)satisfaction: A gender analysis among British students. Journal of Consumer Policy, 18, 237–265. Erikson, E. H. (1968). Identity: Youth and crisis. New-York: Norton. Feick, L. F., & Price, L. L. (1987). The market maven: A diffuser of marketplace information. Journal of Marketing, 51, 83–97 (January). Flynn, L. R., Goldsmith, R. E., & Eastman, J. K. (1996). Opinion leaders and opinion seekers: Two new measurement scales. Journal of the Academy of Marketing Sciences, 24(2), 137–148. Freeman, L. C. (1979). Centrality in social networks: Conceptual clarification. Social Networks, 1(3), 215–239. Fromkin, H. L. (1971). In M. G. David (Ed.), A social psychological analysis of the adoption and diffusion of new products and practices from a uniqueness motivation perspective. Advances in consumer research, 2. (pp. 464–469) Ann Arbor, MI: Association for Consumer Research. Ge, X., Conger, R. D., & Elder, G. H. (2001). Pubertal transition, stressful life events, and the emergence of gender differences in adolescent depressive symptoms. Development Psychology, 37(3), 404–417. Gecas, V., & Steff, M. (1990). Families and adolescents: A review of the 1980s. Journal of Marriage and the Family, 52(4), 941–958.

Gerbing, D. W., & Anderson, J. C. (1988). An update paradigm for scale development incorporating unidimensionality and its assessment. Journal of Marketing Research, 25(2), 186–192. Goldsmith, R., & Desborde, R. (1991). A validity study of a measure of opinion leadership. Journal of Business Research, 22(1), 11–19. Goodrich, K., & Mangleburg, T. F. (2010). Adolescent perceptions of parent and peer influences on teen purchase: An application of social power theory. Journal of Business Research, 63(12), 1328–1335. Hansen, F., & Hansen, M. H. (2005). Children as innovators and opinion leaders. Young Consumers, 6(2), 44–59. Hinz, O., Schulze, C., & Takac, C. (2012). New product adoption in social networks: Why direction matters. Journal of Business Research. Katz, E., & Lazarsfeld, P. F. (1965). Personal influence. Glencoe, IL: Free Press. Kline, R. B. (2005). Principles and practice of structural equation modeling. New York: Guilford Press. Kratzer, J., & Lettl, C. (2009). Distinctive roles of lead users and opinion leaders in the social networks of schoolchildren. Journal of Consumer Research, 36(4), 646–659. Kurt, D., Inman, J., & Argo, J. J. (2011). The influence of friends on consumer spending: The role of agency—Communion orientation and self-monitoring. Journal of Marketing Research, 58(4), 741–754. Lee, S. H., Cotte, J., & Noseworthy, T. J. (2010). The role of network centrality in the flow of consumer influence. Journal of Consumer Psychology, 20(1), 66–77. Lindsey, L. L. (2011). Gender roles. A sociological perspective. Upper Saddle River, NJ: Prentice Hall. Lubbers, M. J. (2003). Group composition and network structure in school classes: A multilevel application of the p* model. Social Networks, 25(4), 309–332. Mangleburg, T. F., Doney, P. M., & Bristol, T. (2004). Shopping with friends and teens' susceptibility to peer influence. Journal of Retailing, 80(2), 101–116. Martin, M. C., & Gentry, J. W. (1997). Stuck in the model trap: The effects of beautiful. models in ads on female pre-adolescents and adolescents. Journal of Advertising, 26(2), 19–34. Mascarenhas, O. A. J., & Higby, M. A. (1993). Peer, parent, and media influences in teen apparel shopping. Journal of the Academy of Marketing Science, 21(1), 53–58. McColl-Kennedy, J. R., & Anderson, R. D. (2005). Subordinate-manager gender combination and perceived leadership style influence on emotions, self-esteem and organizational commitment. Journal of Business Research, 58(2), 115–125. Nieva, V. F., & Gutek, B. A. (1980). Sex effects on evaluation. Academy of Management Review, 5, 267–276. Nuttall, P., & Tinson, J. (2008). Keeping it in the family: how teenagers use music to bond, build bridges and seek autonomy. In A. Y. Lee, & D. Soman (Eds.), Advances in Consumer Research, Vol. 35 (pp. 450–456). Duluth, MN: Association for Consumer Research. O'Donnell, K. A., & Wardlow, D. L. (2000). In S. J. Hoch, & R. J. Meyer (Eds.), A theory on the origins of coolness. Advances in consumer research, 27. (pp. 13–18) : Association for Consumer Research. Piacentini, M. (2010). Children and fashion. In D. Marshal (Ed.), Understanding children and consumers (pp. 202–217). London: Sage Publication. Rogers, E. M., & Kincaid, L. (1981). Communication networks: Towards a new paradigm for research. New York: Free Press. Rose, G. M. (1999). Consumer socialization, parental style, and developmental timetables in the United States and Japan. Journal of Marketing, 63(3), 105–119. Rose, G. M., Boush, D., & Friestad, M. (1998). In B. G. Englis, & A. Olofsson (Eds.), Self-esteem, susceptibility to peer interpersonal influence, and fashion attribute preference in early adolescence. European advances in consumer research, 3. (pp. 197–203) Provo, UT: Association for Consumer Research. Snyder, C. R., & Fromkin, H. C. (1980). Uniqueness: The human pursuit of difference. New York: Plenum. Soomro, M. I., Gilal, R. G., & Jatoi, M. M. (2011). A measure of uniqueness and social character to predict shopping behavior: A comparative study on gender difference. Interdisciplinary Journal of Contemporary Research in Business, 2(12), 405–414.

E. Gentina et al. / Journal of Business Research 67 (2014) 83–91 Spence, J., & Helmreich, R. L. (1978). Masculinity and femininity: Their psychological dimensions, correlates, and antecedents. Austin: University of Texas Press. Thompson, C. J. (1996). Caring consumers: Gendered consumption meanings and the juggling lifestyle. Journal of Consumer Research, 22(4), 388–407. Tian, K. T., Bearden, W. O., & Hunter, G. L. (2001). Consumers' need for uniqueness: Scale development and validation. Journal of Consumer Research, 28(1), 50–66. Tian, K. L., & McKenzie, K. (2001). The long-term predictive validity of the consumers' need for uniqueness. Journal of Consumer Psychology, 10(3), 171–193. Triandis, H. C. (1995). Individualism and collectivism. Boulder: Westview Press. Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D., & Wetherell, M. S. (1987). Rediscovering the social group: A self-categorization theory. Oxford, UK: Blackwell. Valente, T. W., Unger, J. B., & Johnson, A. C. (2005). Do popular students smoke? The association between popularity and smoking among middle school students. Journal of Adolescent Health, 37(4), 323–329. Van Den Bulte, C., & Joshi, Y. V. (2007). New product diffusion with influentials and imitators. Marketing Science, 26(3), 400–421.

91

Vernette, E. (2004). Targeting women's clothing fashion opinion leaders in media planning: An application for magazines. Journal of Advertising Research, 44(1), 90–107. Wasserman, S., & Faust, K. (1994). Social network analysis: Methods and applications. Cambridge, MA: Cambridge University Press. Watts, D. J., & Dodds, P. S. (2007). Influentials, networks, and public opinion formation. Journal of Consumer Research, 34(4), 441–458. White, K., & Argo, J. J. (2009). Social identity threat and consumer preferences. Journal of Consumer Psychology, 19(3), 313–325. Winterich, K., Mittal, V., & Rose, W. T. (2009). Who is more generous?: Moral identity moderates the effect of gender identity on donations. Journal of Consumer Research, 36(2), 199–214. Yang, Z., & Laroche, M. (2011). Parental responsiveness and adolescent susceptibility to peer influence: A cross-cultural investigation. Journal of Business Research, 64(9), 979–987.