Performance Implications of a Retail Purchasing Network: The Role of Social Capital

Journal of Retailing 86 (4, 2010) 310–321

Performance Implications of a Retail Purchasing Network: The Role of Social Capital Matthew T. Seevers a,∗ , Steven J. Skinner b,1 , Robert Dahlstrom b,2 a

Creighton University, College of Business Administration, 2500 California Plaza, Omaha, NE 68178, United States b University of Kentucky, Carol Martin Gatton College of Business and Economics, School of Management 425N, Lexington, KY 40506, United States

Abstract This study employs social capital theory to examine how a retail buyer’s network of industry peers influences retail performance. We propose that performance is enhanced by three network resources – access, referral, and influence – that result from two structural facets of a retail buyer’s network: contact diversity and contact position. We test the model by collecting sociometric data that measures interpersonal ties among 84 retail buyers operating in the same geographic territory in the U.S. golf industry. The results offer evidence that network resources lead to higher levels of performance, even when accounting for differences in human capital and organizational resources. The paper concludes with a discussion of managerial and theoretical implications. © 2010 New York University. Published by Elsevier Inc. All rights reserved. Keywords: Social network analysis; Social capital; Retail purchasing; Performance

The activities that people undertake to enhance firm productivity are conducted within organizational networks (Ganesan et al. 2009). This network perspective suggests that relationships operate through a dynamic process in which discrete transactions are dependent on existing relationships that span firms (Håkansson 1982; Koza and Dant 2007). Scholars have argued for the importance of examining the broader set of relationships among suppliers, vendors, and buyers (Achrol, Reve, and Stern 1983; Wuyts et al. 2004). The subsequent empirical efforts adopt network-based approaches to examine retailer ties to channel partners (e.g., Bradford, Stringfellow, and Weitz 2004; Davis and Mentzer 2008; Mittal, Huppertz, and Khare 2008). In this study, we build on social capital theory to examine how a retail buyer’s network of industry peers influences retail performance. We treat social capital as the structural properties of and resources mobilized through an individual’s relationships in a network (Nahapiet and Ghoshal 1998). Our study makes three contributions. First, social capital theory is employed to offer a

∗

Corresponding author. Tel.: +1 402 280 2093; fax: +1 402 280 5565. E-mail addresses: [email protected] (M.T. Seevers), [email protected] (S.J. Skinner), [email protected] (R. Dahlstrom). 1 Tel.: +1 859 257 1543; fax: +859 257 8031. 2 Tel.: +1 859 257 6717; fax: +859 257 3577.

parsimonious framework that captures multiple strategic influences of extra-organizational ties held by retail agents. Here we focus on social capital derived from external ties that enable individuals to access and mobilize resources beyond the firm (Adler and Kwon 2002). Second, we examine retail purchasing agent effectiveness by adopting an egocentric network perspective, which considers social ties and resources at an individual-level of analysis. Prior channels research of outcomes associated with network ties has emphasized the performance of industrial sellers based on group-level ties (Gu, Hung, and Tse 2008; Palmatier 2008). The current study offers a novel approach not only in its analysis of retail performance, but also in its use of sociometric methods to identify and verify interpersonal ties held by individual retail purchasing agents. Third, we investigate informal ties linking individual peer retail agents. Beyond permitting analysis of relationship-based resources at the same level in a supply chain, the investigation sheds light on emergent ties that form spontaneously among channel participants (Antia and Frazier 2001; Hutt and Reingen 1987). The article is organized as follows: We begin with an overview of social capital theory and present a model of social capital effects on retail buyer performance. The model is tested using a sample of retail buyers from the U.S. golf industry. Finally, we discuss the limitations and the managerial and theoretical implications of the study.

0022-4359/$ – see front matter © 2010 New York University. Published by Elsevier Inc. All rights reserved. doi:10.1016/j.jretai.2010.07.002

M.T. Seevers et al. / Journal of Retailing 86 (4, 2010) 310–321

Theoretical framework and hypotheses The analysis of social capital stems from sociological efforts to examine influences of relationships on social action. Unlike other forms of capital, social capital is not the property of a single actor; instead, it is embedded within relationships, and may be gained or lost as ties are developed or broken (Coleman 1988). Social capital has been examined as the resources available from one’s ties, as well as the ties and structure of a network that yield access to resources (Foley and Edwards 1999). Network scholars also distinguish between internal and external social ties (Borgatti, Jones, and Everett 1998). Internal ties research focuses on relationships among a complete network of actors (e.g., work team), whereas analyses of external ties examine efforts to reach beyond a local clique. The varied use of the social capital concept prompts the need for explicit discussion of its form and dimensionality. Consistent with Nahapiet and Ghoshal (1998), we examine social capital as the structural properties of and resources mobilized through an individual’s relationships in a network. This perspective facilitates independent analyses of an actor’s position in the network as well as the resources associated with the position. In addition, we focus on the external ties established by retailers in their interactions with suppliers and other retailers. External ties enable firms to acquire resources outside the organization (Gu, Hung, and Tse 2008), and examination of these external relationships provides insight into the success of individuals relative to their rivals (Adler and Kwon 2002). Three perspectives recognize that social capital influences an individual’s ability to achieve desirable performance and other outcomes (Seibert, Kraimer, and Liden 2001). Granovetter’s (1973) “strength of ties” and Burt’s (1992) “structural holes” theories focus on the structure of the network. These perspectives maintain that social capital accrues to individuals who reach a variety of contacts within their interpersonal networks. Strong ties evince higher levels of closeness, emotion, and reciprocity than weak ties (Rindfleisch and Moorman 2001), but strong ties are more likely to possess redundant information and constrain resource diversity. By contrast, weak ties characterize relationships between cliques and offer access to more novel information (Granovetter 1973). Similarly, structural holes exist where an individual lacks ties to others and becomes dependent on one or few contacts to gain access to information (Burt 1992). Structural holes arise, for instance, when a retail buyer relies on a limited number of salespeople for learning about market opportunities. Individuals with greater diversity within their networks, however, are more likely to span structural holes in the networks and improve their access to information. A third perspective on social capital shifts the focus from the structure of an individual’s relationships to the resources held by and available to one’s contacts (Lin 2001). Lin, Ensel, and Vaughn (1981) suggest that the structure of the network is secondary to the accessible resources embedded in the network. Social ties are important since they provide access to resources

311

that enable one to achieve objectives. Lin, Vaughn, and Ensel (1981) offer evidence that the resources possessed by an individual influence the ability to reach contacts of higher status, and the status of the contacts influences the outcomes obtained from a relationship. These theoretical perspectives offer complementary insights into the value of relational ties. Tie strength (Granovetter 1973) and structural holes research (Burt 1992) are concerned with how network configuration is associated with reaching diverse contacts, whereas social resource theory (Lin 2001) considers the accessible resources in the network. We suggest that a social capital framework that considers the influences of structural properties and relational resources on organization outcomes enables a more complete picture of how these perspectives are linked (Seibert, Kraimer, and Liden 2001). We implicate contact diversity and contact position as structural facets of social capital, and contact diversity as a determinant of contact position. The network contacts are in turn expected to influence one’s acquisition of three social capital resources (access, referral, and influence). Finally, these network resources are posited to relate positively to a retail buyer’s performance. Fig. 1 provides an illustration of our proposed model. Network configuration and network contacts Contact diversity. In carrying out day-to-day activities, retail buyers seek many sources of information to aid their decisionmaking (Kline and Wagner 1994). Interpersonal sources include internal colleagues as well as external contacts, such as salespeople and industry peers (Hirschman and Mazursky 1982). Contact diversity describes the degree to which an individual’s network includes external contacts that offer variety in terms of their information, knowledge, or experience (Harrison and Klein 2007). Assuming that different firms are themselves heterogeneous bundles of resources, contact diversity grows as connections to individuals outside of one’s own firm increase. Diversity should influence one’s ability to reach prominent contacts, which are ties to individuals that hold high-ranking positions based on prestige, authority, or economic standing (Lin 2001). As the level of diversity within one’s network increases, access to knowledge and information also increases (Hutt and Walker 2006). Establishing ties to diverse others requires substantial effort (Lin 2001); hence, goal-directed individuals are inclined to pursue ties with individuals that can provide opportunities via access to knowledge and resources (Lin, Ensel, and Vaughn 1981). Since prominent industry contacts provide this access, we propose the following: H1. The level of contact diversity in a retail buyer’s interpersonal network is positively related to reaching prominent industry contacts. Network contacts and network resources Contact position. Retail buyers with more prominent contacts should reap greater social capital benefits than counterparts with less prominent contacts. A core premise of the social capital lit-

312

M.T. Seevers et al. / Journal of Retailing 86 (4, 2010) 310–321

Fig. 1. A model of social capital effects on retail buyer performance.

erature suggests that ties to unique clusters of social activity yield timely and unique information (Burt 1992; Granovetter 1973). To the extent that a diverse network includes prominent contacts, the benefits of social capital should be amplified for the retail buyer. Contacts who occupy high levels within a social structure provide enhanced access to marketplace information (Lin 2001). Prestigious contacts are targets of relational efforts from many others and have access to a high volume of information from diverse sources (Burt 1977). Contacts in high-ranking positions have higher levels of decision-making authority and broader access to information that affects the firm (Mintzberg 1979). High performing industry peers are likely to hold valuable market information; hence, ties to such contacts are expected to enhance the overall quality of information transferred. A network of diverse contacts should also ease referral behavior in the form of word-of-mouth information about a retail buyer. Referrals make retail buyers more attractive as trustworthy exchange partners and ease the initiation of new purchasing relationships (Uzzi 1997). Building a network with prominent contacts should enhance this benefit. Prominent channel partners are more centrally located within networks, and this position enables them to provide recommendations to more people (Riitta, Rosenberger, and Eisenhardt 2008). Influence refers to a retail buyer’s effectiveness in the use of compliance-gaining tactics aimed at modifying the behavior of trading partners (Payan and McFarland 2005). Attributions made about an individual are enhanced if they are perceived to be tied to prominent contacts (Kilduff and Krackhardt 1994). In this sense, an individual’s network relationships provide signals about the person to others (Dahlstrom and Ingram 2003). Similarly, Tedeschi and Melburg (1984) suggest that when others believe that an individual holds relationships to prominent contacts, the individual will be seen as more influential. Therefore, the following are proposed:

H2. The prominence of contacts in a retail buyer’s interpersonal network is positively related to the: (a) retail buyer’s access to marketplace information, (b) word-of-mouth referral exhibited on behalf of the retail buyer, and (c) perceptions of the retail buyer’s influence. Network resources and performance Access. Receipt of unique and timely marketplace information through industry contacts should be positively related to a buyer’s performance. Research suggests that when boundary spanners do not have the information necessary to perform their jobs, their performance wanes (Michaels, Day, and Joachimsthaler 1987). Greater access to information through social ties should counteract ambiguity and lead to heightened performance. Information received via interpersonal networks has also been shown to contribute to a buyer’s awareness of potential suppliers (Webster 1970), consideration of larger alternative sets (Ozanne and Churchill 1971), evaluation of purchase options (Czepiel 1974), reduction of perceived risk before a purchase (Webster 1968), and reduction of post-purchase uncertainty (Martilla 1970). Information received from interpersonal networks may also enable buyers to adapt to the specific needs of interaction with vendors (Weitz, Sujan, and Sujan 1986). H3. A retail buyer’s access to information about a marketplace is positively related to his or her performance. Referral. Referral behavior provides a positive force for future endeavors within a network (Burt 1992). Interfirm research suggests that referrals simplify the selection of business partners by signaling reputation and attractiveness (Anderson, Håkansson, and Johanson 1994; Davis and Mentzer 2008).

M.T. Seevers et al. / Journal of Retailing 86 (4, 2010) 310–321

Money et al. (1998), for example, show that firms use wordof-mouth referral networks to find new exchange partners. Dahlstrom and Ingram (2003) posit that network ties may be used to screen potential partners and thus reduce transaction costs associated with adverse selection. Direct endorsements from one’s industry contacts to prospective partners should also ease costs associated with initiating relationships. Hence, referral activity performed on behalf of a retail buyer should enhance performance. H4. Word-of-mouth referral behavior exhibited by a retail buyer’s contacts is positively related to his or her performance. Influence. Influence is manifest in the use of compliancegaining tactics aimed at other market participants (Frazier and Summers 1984). An important application of a buyer’s interpersonal influence is in vendor negotiations. Dwyer and Walker (1981) find that powerful bargainers make more demanding initial bids, yield less to the demands of others, and derive more profitable outcomes. Influence derived from associations to prominent contacts provides a base of social power accessible during negotiations (French and Raven 1959). In addition, control exerted in negotiation derives from an individual’s advantage in controlling the flow of information (Burt 1992). H5. Perceptions of a retail buyer’s influence in business relationships are positively related to his or her performance. Control variables Our model of performance focuses on network properties of retail buyers, but other factors may influence organizational outcomes. Two covariates not based on network properties are included to account for performance determinants beyond the scope of the model. Organizational prestige (Bhattacharya, Rao, and Glynn 1995) concerns the reputation of the firm represented by a buyer, and is expected to relate positively to performance. Another variable of interest is human capital, which refers to an individual’s accumulation of job-related knowledge and skill. We include industry experience to capture the expected positive influence of a buyer’s human capital on performance. Method Sample We tested the proposed research model in the U.S. golf industry. A $76 billion industry (SRI International 2008), retail purchases of golf merchandise account for nearly 10% of industry-wide sales. The purchasing agents in this study are PGA (Professional Golfers’ Association of America) professionals in a single U.S. state, each of whom manages and acts as the buyer for the retail merchandise operations at an independent “on-course” golf facility. The respondents face a highly competitive market in which they vie against one another, as well as “off-course” retailers who sell golf merchandise via brick-andmortar, catalog, Internet, or some combination. Consistent with a focus on extra-organizational ties among industry peers, we

313

examined the social ties among fellow PGA professionals. We also collected peer-report data from the local sales representatives to whom the respondents were tied. Preliminary fieldwork suggested that a buyer’s purchases are spread across multiple sales representatives; similarly, sales reps manage multiple retail accounts. The research context afforded a number of key benefits. First, the research design, which seeks responses from members of a smaller, geographically-bounded network, requires only that respondents report on their own direct relationships. Hence, the context promotes precision of network measurement through cross-validation of the ties reported for each dyad in the network.3 Second, the choice of respondent is also well-suited to an empirical test of our research model. PGA professionals, like other types of retail purchasing agents, often act with autonomy when making purchase decisions (Wagner, Ettenson, and Parrish 1989); hence, the design facilitates an individual-level analysis (John and Reve 1982). Finally, scholars have pointed out that boundaries drawn around social networks are often artificial (Bristor and Ryan 1987). The network investigated here, however, enables a more natural demarcation of network boundaries by focusing on (1) a specific actor role (e.g., retail purchasing) and a specific relational content (e.g., informal communication) that places a contextual boundary on the network; (2) respondents that share a professional affiliation; (3) respondents within a geographic boundary defined by their professional association; and (4) retail establishments within the boundaries of a single U.S. state. Data were collected by mailing surveys to each PGA professional and relevant salesperson in the state. Among the 33 sales representatives servicing the territory, 24 responded to the questionnaire. At the time of data collection, 125 professionals were considered active members in the PGA section and were responsible for retail purchasing activities at their facilities. A total of 84 questionnaires were returned, yielding a 67.2% response rate among the focal respondents. Informant checks confirmed that the PGA professionals held positions of purchasing authority (e.g., “Head Golf Professional”) and served as the principal buyers for their facilities. The respondents had an average age of 40.1 years, were well-educated (69% college graduate), and consistent with the larger membership in the PGA, were largely male (82 of 84). The golf facilities represented by the PGA professionals had on average been established for 41.7 years, and were evenly split between private- and public-access. Retail merchandise operations at the facilities averaged 8.8 employees. Non-response bias was investigated by comparing respondents and nonrespondents on “known values” (Armstrong and Overton 1977). Two comparable values were known for nonrespondents: (1) geographic location and (2) membership class

3

Cross-validated network ties refer to a procedure in which an individual’s network responses are checked against responses from peers. For example, if A claims a relationship to B, we could confirm that B in turn claims a relationship to A. From a technical standpoint, raw network responses were transformed using the Symmetrize Minimum function in UCINET, which sets each tie strength response within a dyad to the minimum tie strength reported by either member for the dyad.

314

M.T. Seevers et al. / Journal of Retailing 86 (4, 2010) 310–321

Table 1 Summary of study measures. Construct

Measurement type

Measurement source

In relation to

Item/item example

Contact diversity

Network ties

Buying agents

Peer buying agents

Contact position

Network ties

Buying agents

Peer buying agents

Access

Self-report perceptive

Buying agent

Self

Referral

Self-report perceptive

Buying agent

Self

Influence

Multi-peer-report perceptive

Local sales reps

Buying Agent

Performance

Self-report objective

Buying agent

Self

“Describe how ‘close’ your relationship is with this golf professional.” “Describe how ‘close’ your relationship is with this golf professional.” “I often hear about unique golf products that my fellow golf professionals are not aware of.” “I have many peers in this PGA section who would serve as a reference for me.” “How influential is this person in working with vendors and sales representatives?” “Expected total sales revenue of golf merchandise for the 200 season?”

of the PGA professionals. An assessment of response rate suggests that neither variable was a significant contributor to nonresponse. We also assessed non-response bias by comparing responses from early versus late responders (Armstrong and Overton 1977). Late and early respondents were found not to be significantly different on any variables. Measures Table 1 provides a summary of measures used in the study. Data used to measure network relationships were collected through response to sociometric questions. Respondents were provided an alphabetical roster containing the names of fellow PGA professionals and prompted to answer questions about the individuals with whom they had discussed retail purchasing matters (e.g., brand assortment, top-selling products) during the previous 12 months. A similar roster was included to capture ties between the PGA professionals and local sales representatives. The 24 local sales representatives had a total of 741 ties to buyers (μ = 30.9), which represented 35.2% of the 2106 (84 × 24) possible buyer-to-seller dyads. By contrast, the 84 PGA professionals shared 186 ties to one another, which represented 5.3% of the 3486 (n*n−1/2) possible buyer-to-buyer dyads. Because traditional reliability tests are not appropriate for sociometric data (Wasserman and Faust 1994), this design incorporated a number of methods to promote accurate measurement of network relationships beyond the use of cross-validation. First, all respondents were provided with a complete list of network participants as a means to aid recall accuracy (Knoke and Kuklinski 1982). Prior to indicating their contacts, respondents were also given a cue to define the relational content of interest (e.g., “individuals with whom you have discussed golf shop purchasing issues within the past 12 months”) (McCallister and Fischer 1978). Finally, network questions focused on longerterm patterns of relationships rather than on particular episodes (Freeman, Romney, and Freeman 1987). Contact diversity was measured by counting the number of external ties a respondent had to peer retail buyers, each of whom represented a different golf facility. Buying agents for these firms offer differing resources and perspectives in terms of informa-

tion, knowledge, and experience (Harrison and Klein 2007). This measurement is also consistent with the study of social capital in an intrafirm setting (Seibert, Kraimer, and Liden 2001). Our measure of contact position captures the degree to which a retail buyer’s network ties include peer buyers at other facilities who are prestigious in the network-at-large. We first measured prestige for each retail buyer in the network using their calculated in-degree centrality, which measures the number of people in the network who indicate a tie with the retail buyer (Freeman 1979).4 Centrality scores were calculated in UCINET 6 for Windows (Borgatti, Everett, and Freeman 2002). To complete our operationalization of this construct, we summed the in-degree centrality of a respondent’s five most prestigious contacts. Restricting the measurement to only the top five contacts was intended to help us distinguish between a personal network comprised of many low-level contacts versus a network comprised of truly prestigious contacts. Access and referral were measured using multi-item selfreport scales that were newly developed for this study (see Measurement Appendix) Creation of both measures followed suggested procedures for multi-item scale development (Churchill 1979; Gerbing and Anderson 1988). The set of items generated to reflect access focused on the quality of marketplace information gathered in terms of timing and uniqueness (Burt 1992). Qualitative and quantitative pre-tests yielded a six-item scale for access. An example item includes, “I often hear about new merchandising programs before other golf professionals.” For the measurement of referral, purification of an initial set of scale items resulted in a five-item scale. An example item is, “My fellow golf professionals speak highly about me to local sales representatives.” Responses to each measure were made using a scale ranging from 1 = “strongly disagree” to 7 = “strongly agree.” Influence was measured by asking sales representatives who cited a focal retail buyer as a contact to rate the buyer’s influence in his or her purchasing relationships (Brass 1984). Respon-

4 In-degree centrality, like other measures of prestige, allows for the possibility that not all relationship efforts from prospective contacts are reciprocated.

M.T. Seevers et al. / Journal of Retailing 86 (4, 2010) 310–321

315

Table 2 Descriptive statistics, correlation matrix, and psychometric properties of multi-item scales. Construct

Mean

S.D.

1

2

3

4

5

6

7

8

1. Contact diversity 2. Contact position 3. Access 4. Referral 5. Influence 6. Performance 7. Organizational prestige 8. Industry experience

4.48 54.85 3.53 4.76 4.12 157,798 3.90 2.16

4.90 41.50 1.23 1.09 .83 149,630 .72 7.59

1.00 .70** .35** .46** .40** .38** .26* .05

1.00 .32** .52** .42** .47** .14 .01

1.00 .34** .43** .53** .42** −.10

1.00 .33** .45** .38** .21

1.00 .64** .28* −.04

1.00 .39** −.03

1.00 −.11

1.00

Composite reliability Average variance extracted Highest shared variance

– – –

– – –

– – –

– – –

.93 .70 .18

.92 .70 .14

– – –

– – –

* **

.83 .62 .18

– – –

p < .05. p < .01.

dents answered “How influential is this person in working with vendors and sales representatives?” using a scale ranging from 1 = “very little influence” to 7 = “very great influence.” A single measure of influence was obtained for each buyer by calculating the average of ratings received from all sales representatives who cited the buyer as a contact (Brass and Burkhardt 1993).5 The respondents in this study manage and serve as the principal buyers for the retail merchandise operation at their respective facilities. As such, the level of performance of the retail merchandising operation as a whole is synonymous with the level of performance of the individual respondent. Performance of the retail buyer was measured by asking respondents to report their sales volume of golf merchandise for the year recently completed. Finally, measures were included to capture two additional variables that are expected to account for variance in performance. Organizational prestige for each buyer’s facility was measured based on adaptation of Bhattacharya, Rao, and Glynn’s (1995) 3-item scale. An example item is, “People in my community think highly of this golf facility.” Industry experience was measured by asking each buyer to report the number of years passed since entering the industry as a golf professional. Results Table 2 provides descriptive statistics, correlations, and psychometric properties for multi-item scales for the data collected from the PGA professionals. Initial calculations of networkbased responses were performed using UCINET 6 for Windows 5 The peer-report measure of influence follows the procedures used by Brass (1984) and Brass and Burkhardt (1993). We note that this single-item measure differs from the more common “single-item/self-report” measure, which yields a single data point from the respondent. By contrast, the score that results from this peer-report measure is based on an average of multiple responses from one’s contacts. In this sense, the response from each of a respondent’s contacts serves as an item in a multi-item scale. Thus, a more fitting description of our peer-report measure might be “single-item/multi-peer-report”.

(Borgatti, Everett, and Freeman 2002). With input taken from the covariance matrix, LISREL 8.80 (Jöreskog and Sörbom 2006) was then employed to conduct measure validation procedures and assess the structural model. Measurement validity assessment A pre-test was conducted to evaluate the access, referral, and organizational prestige scales. The sample (n = 124) consisted of PGA professionals working outside of the state targeted for the main study. Exploratory factor analysis gauged the dimensionality of the items. Review of the factor loadings, reliabilities, and item-to-total correlations for each scale led to the removal of five items. The reliabilities for access (α = .86) and referral (α = .88) exceeded guidelines for new measures (Nunnally 1978). The organizational prestige scale similarly evinced acceptable reliability (α = .75). Confirmatory factor analysis was then performed using all measures employed in the study (Anderson and Gerbing 1988). The measurement model included 19 indicators representing the six focal variables and two control variables. Each indicator was specified to load onto a single factor corresponding to the construct it was intended to reflect. Furthermore, each construct was represented by only one factor in the model. Factors represented by a single indicator were assumed to be free of measurement error; hence, the error variance was set to zero and its loading was set to 1.00. Variances for the remaining factors were set to 1.00 and covariances among all factors were freely estimated. Taken as a whole, the fit indices suggest that the full measurement model provided an acceptable fit (χ2 = 133.26, d.f. = 132, p = .45; CFI = 1.00). For multi-indicator factors, all standardized loadings of the individual indicators onto their respective factors were significant. For factors with multiple indicators, the composite reliability (CR) and average variance extracted (AVE) were calculated as follows: access (CR = .93; AVE = .70), referral (CR = .92; AVE = .70), and organizational prestige (CR = .83; AVE = .62). Each factor exceeds the guideline of .60 for composite reliability (Bagozzi and Yi 1988) and .50 for AVE (Fornell and Larcker 1981). All indicators have a significant loading on the factor they are posited to reflect, thus

CD → PERF −.12 CD = contact diversity, CP = contact position, ACC = access, REF = referral, INF = influence, PERF = performance.

.65 .44 .24 177.87 (14), p < .001 Nonmediated 4

163.31 (6), p < .001

14.56 (8), p = .07 216.91 (15), p < .001 11.25 (5), p < .05 11.04 (7), p = .13 14.53 (7), p < .05 7.01 (3), p = .07 43.14 (10), p < .001 46.14 (10), p < .001 116.24 (13), p < .001 Hypothesized model Control variables only Partially mediated 1 Partially mediated 2 Partially mediated 3 Partially mediated 4 Nonmediated 1 Nonmediated 2 Nonmediated 3

a

.10

.45

– – CD → ACC, REF, INF CP → PERF CD → PERF All paths added in PM1, PM2, and PM3 CP → PERF CD → PERF CD → PERF .92 −.28 .88 .94 .90 .87 .68 .66 .24 .95 .33 .97 .97 .95 .98 .87 .86 .64 .81 .05 .76 .83 .78 .75 .58 .56 .28 .96 .60 .97 .97 .96 .98 .88 .88 .74 .98 .31 .98 .99 .97 .99 .89 .88 .65 .05 .24 .05 .05 .05 .04 .09 .11 .20

NFI AGFI GFI CFI SRMR χ2 (d.f.) χ2 (d.f.)

Table 3 Nested model comparisons for structural equation model.

Our approach follows Anderson and Gerbing’s (1988) recommendations to assess the merits of alternative and theoretically reasonable models nested within a measurement model. To do so, we employed the framework offered by Kelloway (1998) for generating alternative structural models that include mediated relationships. These models were evaluated based on comparisons of chi-square difference tests and changes in other fit indices. The initial estimation of the structural model included a total of 8 latent variables – six hypothesized variables from the research model, and two control variables. Taken as a whole, the hypothesized model provided an acceptable fit to the data (χ2 = 14.56, d.f. = 8, p = .07; CFI = .98). Given an acceptable fit, the next step was to compare the hypothesized research model to a series of alternative nested models. Kelloway (1998) recommends that proposed mediated relationships be compared to specifications of partial mediation, as well as nonmediation. A total of 9 alternative models were tested, including one model in which only control variables served as predictors of performance, four models that examined partial mediation, and four models that examined nonmediation. Table 3 summarizes the model comparisons. The first alternative model was a control variables-only model in which paths from the two control variables to all other latent variables were freely estimated, whereas the hypothesized paths in the structural model were fixed at zero. The fit of this model was significantly worse (χ2 = 202.35, d.f. = 7, p < .001) than the fit offered by the hypothesized model. Among the nonmediated alternative models, all were found to have worse fit than the hypothesized model across a variety of fit indices. The fit of the partially mediated models is not significantly different than the fit of the hypothesized model; however, each of these alternatives sacrifices parsimony by virtue of adding more structural paths. Based on these results, the hypothesized model was retained as the best-fitting model and used to interpret the hypothesized relationships. The model explains 50% of the variance in performance as measured by the squared multiple correlation (i.e., R2 ); by comparison, the

NNFI

Structural model assessment

– 202.35 (7), p < .001 3.31 (3), p > .05 3.52 (1), p > .05 .03 (1), p > .05 8.55 (5), p > .05 28.58 (2), p < .001 35.18 (2), p < .001 101.68 (5), p < .001

Paths addeda

Paths removeda

offering further evidence of convergent validity (Anderson and Gerbing 1988). Two methods provided evidence of discriminant validity of the measures. First, we compared nested models for each pair of multi-indicator factors whereby the first model constrains the correlation parameter to unity and the second model allows the correlation parameter to be freely estimated (Anderson and Gerbing 1988). Model fit for each of the constrained models was significantly worse (lowest χ2 = 45.00, d.f. = 1, p < .001) (Bagozzi and Phillips 1982). Second, we found that the AVE for each factor with multiple indicators exceeded the shared variance for each pair of multi-indicator factors (Fornell and Larcker 1981). The results of these tests suggest that our measures are robust against the ill effects of multicollinearity (Grewal, Cote, and Baumgartner 2004).

– All hypothesized – – – – ACC, REF, INF → PERF ACC, REF, INF → PERF CP → ACC, REF, INF ACC, REF, INF → PERF All hypothesized

M.T. Seevers et al. / Journal of Retailing 86 (4, 2010) 310–321

Model

316

M.T. Seevers et al. / Journal of Retailing 86 (4, 2010) 310–321

317

Table 4 Results of hypothesis testinga . Construct

Direction

Construct

Estimate

Standardized estimate

SE

t-Value

p

Hypothesis

Conclusion

Contact diversity Contact Position Contact Position Contact Position Access Referral Influence

→ → → → → → →

Contact Position Access Referral Influence Performance Performance Performance

.71 .27 .45 .40 .23 .20 .45

.71 .27 .47 .40 .23 .20 .45

.08 .10 .09 .10 .09 .09 .08

8.62 2.82 5.52 4.09 2.64 2.31 5.51

.00 .02 .00 .00 .03 .05 .00

1 2a 2b 2c 3 4 5

Supported Supported Supported Supported Supported Supported Supported

a

Structural model fit: χ2 = 14.56, d.f. = 8, p = .07; SRMR = .05; CFI = .98; GFI = .96; AGFI = .81; NFI = .95; NNFI = .92.

control variables-only model explained 15% of the variance in performance. Hypothesis tests Table 4 provides a summary of the standardized path coefficients for all hypothesized relationships. The structural path coefficients reveal that all hypothesized relationships are significant and in the posited direction. H1 posited that a network of diverse contacts is associated with reaching prominent contacts. The standardized path coefficient between contact diversity and contact position was significant and positive (γ = .71, p < .05) offering support for H1. The hypothesized relationships linking contact position and network resources were supported. Specifically, H2 relates the prominence of contacts in a retail buyer’s network to access to marketplace information (2a), word-of-mouth referral from peers (2b), and attaining influence in business relationships (2c). The standardized path coefficients from contact position to access (β = .27, p < .05), referral (β = .47, p < .05), and influence (β = .40, p < .05) were positive and significant. The final three hypotheses in the model posit that performance is positively related to a retail buyer’s access to information (H3), receipt of word-of-mouth referral (H4), and influence in business relationships (H5). The standardized path coefficients support all three hypotheses. A positive and significant linkage was found between access and performance (β = .23, p < .05), referral and performance (β = .20, p < .05), and influence and performance (β = .45, p < .01). Finally, while both organizational prestige and industry experience were positively related to performance in the structural model, neither relationship reached statistical significance. The standardized path coefficients from organizational prestige to access (γ = .38, p < .05) and referral (γ = .34, p < .05) were found to be statistically significant. Industry experience was found to be positively related to referral (γ = .23, p < .05). All other paths linking organizational prestige and industry experience to the latent variables in the hypothesized model were not statistically significant. Discussion and implications The present study contributes to channels research in a retail setting by employing social capital theory as a framework that captures network relationships as strategic resources. Taken as

a whole, the test of the model provides evidence of the value of social capital derived from external ties to industry peers. Retail purchasing agents with external ties to representatives of a wide variety of firms were more likely to reach prominent peers. This finding corroborates the structural foundations of social capital (Burt 1992; Granovetter 1973) that emphasize the value of reaching diverse contacts for goal-directed behavior. In turn, the results support that one’s reach to prominent contacts enhances the mobilization of network-embedded resources, which highlights a linkage between the structural and relational dimensions of social capital (Lin 2001; Nahapiet and Ghoshal 1998). A second contribution is the individual-level of analysis of social structure and resources on the effectiveness of retail purchasing agents. The findings illustrate that a linkage between network configuration and performance is mediated by a reach to prominent contacts who offer three network resources: enhanced access to marketplace information, increased word-of-mouth referral, and greater influence in dealings with representatives of upstream firms. The analysis of retail performance here complements efforts to assess implications of external ties for industrial sellers (Atuahene-Gima and Murray 2007; Gu, Hung, and Tse 2008). In contrast with group- or firm-level studies that capture network ties with psychometric items, this study offers a novel approach by using sociometric methods to capture interpersonal ties. Not only does this approach provide added confidence in network measures (Bernard et al. 1984), but the individuallevel analysis also permits a tighter methodological coupling between network ties and performance. The findings also highlight how individual-level performance gains from social capital may spillover to benefit the firm as a whole. A final contribution of this study is the investigation of informal ties linking individuals operating at the same level in a supply chain. Research examining network ties among channel participants often focuses on prescribed ties that stem from consciously planned contractual relations (e.g., Luo, Rindfleisch, and Tse 2007; Palmatier 2008; Rindfleisch and Moorman 2003). Our analysis contributes to a growing understanding of the informal ties among industry peers that “shadow” more formal relationships (e.g., Gu, Hung, and Tse 2008). Prior research in the retail literature offers evidence of informal ties among retail buyers (Hirschman and Mazursky 1982; Kline and Wagner 1994). The current study extends these findings by offering evidence that informal ties contribute to performance and provide a critical vehicle for firms to mobilize resources within the channel environment.

318

M.T. Seevers et al. / Journal of Retailing 86 (4, 2010) 310–321

Limitations Due in part to the nature of social networks and their measurement, this study was subject to a number of limitations. First, the cross-sectional design places an obvious limit on the ability to make definitive conclusions about causality in our model. Although the results offer support for our hypotheses, it is plausible that a buyer’s performance not only benefits from prominent ties, but also contributes to the formation of new ties to prominent peers. Finding evidence of the latter case would suggest that the quality of one’s contacts may be fueled in part by one’s performance. Longitudinal research that augments our study should provide insight into the causal order of these variables and reveal the influence of social capital developed over time. The results of this study must also be viewed in light of the sample size and response rate. Investigations of small populations are not uncommon due to the practical challenges of gathering network data. In this study, we also made a deliberate effort to seek responses from all members of a defined network. This method enabled us to verify responses from each member of a dyad and to address the known problems of informant accuracy in reporting one’s own ties, as well as ties among his or her contacts (Bernard et al. 1984). Nevertheless, we acknowledge that the resulting sample size limits the statistical power of the analysis. We further recognize that the method used here to gather network data was sensitive to nonresponse, such that each missing respondent is equal to the loss of “N-1 possible relationships involving other network actors” (Knoke and Kuklinski 1982, p. 35). Ties claimed by respondents to those in the network that did not respond could not be verified and were not accounted for in our network measures. Thus, our measurement of network properties may be problematic to the extent that nonrespondents are vital members of the network. Another limitation of this study resulted from bounding the network using a single U.S. state. Although the design facilitated demarcation of the network, buyers residing near bordering states may have informal networks that span beyond the scope of the network we investigated. Research focused on establishing limits of the network should provide insight into this study and other field research attempting to measure interaction across firms. Managerial implications The development of marketplace social networks has long been prescribed as good business for persons involved in sales, purchasing, and other boundary spanning roles. Managerial interest in social networks and their modern implications is on the rise (Üstüner and Godes 2006), but much of the evidence linking marketplace social networks and business performance has been anecdotal. The PGA of America, for example, has promoted to its retail operating members the performance benefits of networking with industry peers to learn about sales and merchandising trends. The results of the current study suggest: (1) interpersonal networks that span organizational boundaries in a purchasing context are linked to performance outcomes; (2) net-

work ties contribute to relationship-based resources, including access to marketplace information, word-of-mouth referral, and interpersonal influence; (3) not all network ties are equally valuable; and (4) network ties to prestigious contacts are particularly valuable in the pursuit of relationship-based resources. For persons in boundary spanning roles that depend upon access to high quality market information, our results suggest that network ties offer a valuable source. Advancements in communications and information technology have made public sources of information increasingly accessible, but this very accessibility makes it difficult to leverage the information for competitive advantage. By contrast, Uzzi and Dunlap have highlighted the benefits of private information, which “is gathered from personal contacts who can offer something unique that cannot be found in the public domain, such as the release date of a new product” (2005, p. 54). Indeed, we find that personal contacts are linked to the receipt of high quality information about the market. The results do not suggest that boundary spanners receive these benefits merely from casting a wide net; instead, our findings encourage selectivity in the development of an interpersonal network. Acquiring relationship-based resources appears to depend upon building a network of contacts that are well-positioned to locate resources and decipher the social landscape in the market. Theoretical implications This study has several important implications, the first being that social capital derived from ties external to a retail buyer’s firm appear to make a positive contribution to performance. Many studies have demonstrated how emergent social ties contribute to personal gain (e.g., Podolny and Baron 1997). Other studies have shown that emergent firm-level ties contribute to organizational performance (Atuahene-Gima and Murray 2007; Gu, Hung, and Tse 2008). The present study, however, provides initial evidence that social networks that span organizational boundaries are associated with achievement of performance goals. This finding complements prior research by showing that personal returns from social capital may contribute to an organization’s bottom-line. Burt (1992) proposed that the linkage between structural properties of an interpersonal network and performance is based on mediating resources. Likewise, Nahapiet and Ghoshal (1998) posit that relational facets contribute to the outcomes associated with social capital. This study empirically tests these arguments with its inclusion of three mediating resources of social capital: access, referral, and influence. The model offered here should be instructive for scholars seeking to apply the Nahapiet and Ghoshal (1998) framework to the study of individual-level social capital. The results also help to generalize the work of Seibert, Kraimer, and Liden (2001) by showing how individual-level social capital that results from extra-organizational ties may lead to unique resources in a channel setting. Scholars that endorse either the political economy framework or the IMP Group’s interaction approach acknowledge the broad base of literature upon which these paradigms draw (Arndt 1983; Cunningham 1980). We contend that social

M.T. Seevers et al. / Journal of Retailing 86 (4, 2010) 310–321

capital theory may also contribute to explanation of the phenomena bounded by these paradigms. One issue on which both paradigms converge is the relevance of networks as a means to understand observed interaction in interfirm relationships. Within this context, scholars have also considered the normative implications of network relationships (e.g., Achrol and Kotler 1999). Social capital theory appears promising for this line of investigation because it offers a framework that captures the idea of networks as valued resources. Another implication from this study concerns the inclusion of peer-report measurement that reflects a retail buyer’s influence within the network. In general, the use of multiple measures eases the assessment of construct validity. Collecting both self-report and network-based peer-report measures may, however, be particularly valuable in the assessment of social capital since it is embodied in relationships and is not strictly the property of a single actor (Coleman 1988). It is also plausible for some relationship-based constructs that the level of agreement among measures from multiple informants may itself contribute to the construct’s effects. Hence, future research of social capital seems likely to benefit from investigation of perceptual agreement among network actors. Executive summary The development, maintenance, and leverage of social ties has long been prescribed as sound business practice for persons involved in sales and purchasing. Managerial interest in social networks and their implications has grown sharply in recent years, yet the body of evidence linking marketplace social networks and business performance is still emerging. It is perhaps natural that industrial salespeople have been a common focus of studies that investigate social ties among buying and selling firms. This study takes a different perspective. We adopt the focus of the retail buyer, and seek to shed light on the value of a retail buyer’s informal social ties to industry peers, i.e., retail buyers working for other companies. Using this approach, our study seeks to understand how such ties might enable a retail buyer to perform more effectively according to the resources attributable to his or her social network. The conceptual model developed in this paper begins with an understanding of social capital. We treat social capital as the structural properties of resources mobilized through an individual’s relationships in a network. Multiple research perspectives recognize that social capital influences an individual’s ability to achieve desirable performance, over and above organizational resources (e.g., company reputation) and human capital (e.g., work experience). Our model implicates the diversity of ties within one’s network as a precursor of reaching prominent contacts. In turn, we argue that the prominence of a retail buyer’s contacts is positively linked to three key resources: (1) access to quality marketplace information; (2) word-of-mouth referral that signals partnering attractiveness; and (3) perceptions of the retail buyer’s influence in trading relationships. Finally, these network-based resources are posited to enhance a retail buyer’s performance.

319

We tested our model in the U.S. golf industry by surveying PGA (Professional Golfers’ Association of America) professionals in a single U.S. state. Each respondent manages and acts as the buyer for the retail merchandise operations at an independent golf facility. Network measures were collected using an alphabetic roster in which buyers answered questions about those peers with whom they discussed purchasing matters during the previous 12 months. Our findings provide strong support for social capital theory as a framework that captures network relationships as strategic resources. The results suggest that interpersonal networks contribute to performance outcomes when a retail buyer is connected to other buyers in the same industry. We find that interpersonal networks serve as a conduit through which valued, performanceenhancing resources are made available. In particular, our results link the properties of a retail buyer’s network to receipt of unique and timely marketplace information, positive word-of-mouth referrals from industry peers, and perceptions of one’s influence. The findings imply that retail buyers do not receive these benefits merely from casting a wide net; instead, our results encourage selectivity in the development of one’s interpersonal network. Acquiring relationship-based resources appears to depend upon building a network of prominent contacts who are themselves well-positioned to locate valued resources and decipher the social landscape within the market. Appendix A. Measurement appendix: scale items and response format

Access (Seven-Point Likert-Type Scale: Strongly Disagree – Strongly Agree) I often hear about unique golf products that my fellow golf professionals are not aware of. The information that I hear about the golf industry is typically better than that of the average golf professional. I am exposed to many unique purchasing opportunities. I know about merchandising opportunities well before my fellow golf professionals. I often hear about new merchandising programs before other golf professionals. I am often one of the first to hear about new golf products before they are introduced. Referral (Seven-Point Likert-Type Scale: Strongly Disagree – Strongly Agree) I am well-known among my fellow golf professionals. I have many peers in this PGA section that would vouch for my credibility to others. I have many peers in this PGA section who would serve as a reference for me. My peers have helped me promote my business to potential customers. My fellow golf professionals speak highly about me to local sales representatives. Organizational Prestige (Five-Point Likert-Type Scale: Strongly Disagree – Strongly Agree) People in my community think highly of this golf facility. Our members, guests, and other customers consider this golf facility to be prestigious. This golf facility does not have an outstanding reputation in my community (reverse-coded).

320

M.T. Seevers et al. / Journal of Retailing 86 (4, 2010) 310–321

References Achrol, Ravi Singh, Torger Reve and Louis W. Stern (1983), “The Environment of Marketing Channel Dyads: A Framework for Comparative Analysis,” Journal of Marketing, 47 (Fall), 55–67. Achrol, Ravi S. and Philip Kotler (1999), “Marketing in the Network Economy,” Journal of Marketing, 63 (Special Issue), 146–63. Adler, Paul S. and Seok-Woo Kwon (2002), “Social Capital: Prospects for a New Concept,” Academy of Management Review, 27 (January), 17–40. Anderson, James C. and David W. Gerbing (1988), “Structural Equation Modeling in Practice: A Review and Recommended Two-Step Approach,” Psychological Bulletin, 103 (3), 411–23. Anderson, James C., Håkan Håkansson and Jan Johanson (1994), “Dyadic Business Relationships Within a Business Network Context,” Journal of Marketing, 58 (October), 1–15. Antia, Kersi D. and Gary L. Frazier (2001), “The Severity of Contract Enforcement in Interfirm Channel Relationships,” Journal of Marketing, 65 (October), 67–81. Armstrong, J. Scott and Terry S. Overton (1977), “Estimating Nonresponse Bias in Mail Surveys,” Journal of Marketing Research, 14 (August), 396–402. Arndt, Johan (1983), “The Political Economy Paradigm: Foundation for Theory Building in Marketing,” Journal of Marketing, 47 (Fall), 44–5. Atuahene-Gima, Kwaku and Janet Y. Murray (2007), “Exploratory and Exploitative Learning in New Product Development: A Social Capital Perspective on New Technology Ventures in China,” Journal of International Marketing, 15 (2), 1–29. Bagozzi, Richard P. and Lynn W. Phillips (1982), “Representing and Testing Organizational Theories: A Holistic Construal,” Administrative Science Quarterly, 27, 459–8. Bagozzi, Richard P. and Youjae Yi (1988), “On the Evaluation of Structural Equation Models,” Journal of the Academy of Marketing Science, 16 (Spring), 74–9. Bernard, H. Russell, Peter Killworth, David Kronenfeld and Lee Sailer (1984), “The Problem of Informant Accuracy: The Validity of Retrospective Data,” Annual Review of Anthropology, 13, 495–517. Bhattacharya, C.B., Hayagreeva Rao and Mary Ann Glynn (1995), “Understanding the Bond of Identification: An Investigation of Its Correlates Among Art Museum Members,” Journal of Marketing, 59 (October), 46–57. Borgatti, Stephen P., Martin G. Everett and Linton C. Freeman (2002), “Ucinet for Windows: Software for Social Network Analysis,” Harvard: Analytic Technologies. Borgatti, Stephen P., Candace Jones and Martin G. Everett (1998), “Network Measures of Social Capital,” Connections, 21 (2), 1–36. Bradford, Kevin D., Anne Stringfellow and Barton A. Weitz (2004), “Managing Conflict to Improve the Effectiveness of Retail Networks,” Journal of Retailing, 80 (Fall), 181–95. Brass, Daniel J. (1984), “Being in the Right Place: A Structural Analysis of Individual Influence in an Organization,” Administrative Science Quarterly, 29 (December), 518–39. Brass, Daniel J. and Marlene E. Burkhardt (1993), “Potential Power and Power Use: An Investigation of Structure and Behavior,” Academy of Management Journal, 36 (June), 441–70. Bristor, Julia M. and Michael J. Ryan (1987), “The Buying Center is Dead, Long Live the Buying Center,” in Advances in Consumer Research, Anderson Paul F. and Wallendorf Melanie, eds. Provo, UT: Association for Consumer Research. Burt, Ronald S. (1977), “Positions in Multiple Network Systems, Part One: A General Conception of Stratification and Prestige in a System of Actors Cast as a Social Topology,” Social Forces, 56 (September), 106–31. (1992), “Structural Holes: The Social Structure of Competition,” Cambridge, MA: Harvard University Press. Churchill, Gilbert A. Jr. (1979), “A Paradigm for Developing Better Measures of Marketing Constructs,” Journal of Marketing Research, 16 (February), 64–73. Coleman, James S. (1988), “Social Capital in the Creation of Human Capital,” The American Journal of Sociology, 94 (Supplement), S95–S120.

Cunningham, M.T. (1980), “International Marketing and Purchasing of Industrial Goods – Features of a European Research Project,” European Journal of Marketing, 14 (5/6), 322–38. Czepiel, John A. (1974), “Word-of-Mouth Processes in the Diffusion of a Major Technological Innovation,” Journal of Marketing Research, 11 (May), 172–80. Dahlstrom, Robert and Rhea Ingram (2003), “Social Networks and the Adverse Selection Problem in Agency Relationships,” Journal of Business Research, 56, 767–75. Davis, Donna F. and John T. Mentzer (2008), “Relational Resources in Interorganizational Exchange: The Effects of Trade Equity and Brand Equity,” Journal of Retailing, 84 (December), 435–48. Dwyer, F. Robert and Orville C. Walker Jr. (1981), “Bargaining in an Asymmetrical Power Structure,” Journal of Marketing, 45 (Winter), 104–15. Foley, Michael W. and Bob Edwards (1999), “Is it Time to Disinvest in Social Capital?,” Journal of Public Policy, 19 (2), 141–73. Fornell, Claes and David F. Larcker (1981), “Evaluating Structural Equation Models with Unobservable Variables and Measurement Error,” Journal of Marketing Research, 18 (February), 39–50. Frazier, Gary L. and John O. Summers (1984), “Interfirm Influence Strategies and Their Application Within Distribution Channels,” Journal of Marketing, 48 (Summer), 43–55. Freeman, Linton (1979), “Centrality in Social Networks: Conceptual Clarification,” Social Networks, 1 (February), 215–39. Freeman, Linton, A. Kimball Romney and Sue C. Freeman (1987), “Cognitive Structure and Informant Accuracy,” American Anthropologist, 89 (June), 310–25. French, John R.P. and Bertram Raven (1959), “The Bases of Social Power,” in Studies in Social Power, Dorwin Cartwright ed. Ann Arbor, MI: University of Michigan Press. Ganesan, Shankar, Morris George, Sandy Jap, Robert W. Palmatier and Barton Weitz (2009), “Supply Chain Management and Retailer Performance: Emerging Trends, Issues, and Implications for Research and Practice,” Journal of Retailing, 85 (1), 84–9. Gerbing, David W and James C. Anderson (1988), “An Updated Paradigm for Scale Development Incorporating Unidimensionality and Its Assessment,” Journal of Marketing Research, 25 (May), 186–92. Granovetter, Mark S. (1973), “The Strength of Weak Ties,” American Journal of Sociology, 78, 1360–8. Grewal, Rajdeep, Joseph A. Cote and Hans Baumgartner (2004), “Multicollinearity and Measurement Error in Structural Equation Models: Implications for Theory Testing,” Marketing Science, 23 (Fall), 519–2. Gu, Flora F., Kineta Hung and David K. Tse (2008), “When Does Guanxi Matter? Issues of Capitalization and Its Dark Side,” Journal of Marketing, 72 (July), 12–28. Håkansson, Håkan (1982), “International Marketing and Purchasing of Industrial Goods,” New York, NY: Wiley. Harrison, David A. and Katherine J. Klein (2007), “What’s the Difference? Diversity Constructs as Separation, Variety or Disparity in Organizations,” Academy of Management Review, 32 (October), 1199–228. Hirschman, Elizabeth C. and David Mazursky (1982). A Trans-Organizational Investigation of Retail Buyers’ Criteria and Information Sources, New York University Institute of Retail Management Working Paper No. 82-8, New York University Graduate School of Business Administration Working Paper No. 82-17. Hutt, Michael D. and Peter H. Reingen (1987), “Social Network Analysis: Emergent Versus Prescribed Patterns in Organizational Buying Behavior,” In Advances in Consumer Research, 14, Paul F., Anderson and Melanie Wallendorf eds. Provo, UT: Association for Consumer Research. Hutt, Michael D. and Beth A. Walker (2006), “A Network Perspective of Account Manager Performance,” Journal of Business & Industrial Marketing, 21 (7), 466–73. John, George and Torger Reve (1982), “The Reliability and Validity of Key Informant Data from Dyadic Relationships in Marketing,” Journal of Marketing Research, 19 (November), 517–24. Jöreskog, Karl and Dag Sörbom (2006). LISREL 8.80. Scientific Software International, Inc.

M.T. Seevers et al. / Journal of Retailing 86 (4, 2010) 310–321 Kelloway, E. Kevin (1998), “Using LISREL for Structural Equation Modeling: A Researcher’s Guide,” Thousand Oaks, CA: Sage Publications. Kilduff, Martin and David Krackhardt (1994), “Bringing the Individual Back In: A Structural Analysis of the Internal Market for Reputation in Organizations,” Academy of Management Journal, 37 (February), 87–108. Kline, Barbara and Janet Wagner (1994), “Information Sources and Retail Buying Decision-Making: The Effect of Product-Specific Buying Experience,” Journal of Retailing, 70 (Spring), 75–88. Knoke, David and James H. Kuklinski (1982), “Network Analysis,” Beverly Hills, CA: Sage Publications. Koza, Karen L. and Rajiv P. Dant (2007), “Effects of Relationship Climate, Control Mechanism, and Communications on Conflict Resolution Behavior and Performance Outcomes,” Journal of Retailing, 83 (3), 279–96. Lin, Nan (2001), “Social Capital: A Theory of Social Structure and Action,” New York: Cambridge University Press. Lin, Nan, Walter M. Ensel and John C. Vaughn (1981a), “Social Resources and Strength of Ties: Structural Factors in Occupational Status Attainment,” American Sociological Review, 46 (August), 393–405. , John C. Vaughn and Walter M. Ensel (1981b), “Social Resources and Occupational Status Attainment,” Social Forces, 59 (June), 1163–81. Luo, Xueming, Aric Rindfleisch and David K. Tse (2007), “Working with Rivals: The Impact of Competitor Alliances on Financial Performance,” Journal of Marketing Research, 44 (February), 73–8. Martilla, John A. (1970), “Word-of-Mouth Communication in the Industrial Adoption Process,” Journal of Marketing Research, 8 (May), 173–8. McCallister, L. and C.S. Fischer (1978), “A Procedure for Surveying Personal Networks,” Sociological Methods and Research, 7, 131–48. Michaels, Ronald E., Ralph L. Day and Erich A. Joachimsthaler (1987), “Role Stress Among Industrial Buyers: An Integrative Model,” Journal of Marketing, 51 (April), 28–45. Mintzberg, Henry (1979), “The Structuring of Organizations,” Englewood Cliffs, NJ: Prentice-Hall. Mittal, Vikas, John W. Huppertz and Adwait Khare (2008), “Customer Complaining: The Role of Tie Strength and Information Control,” Journal of Retailing, 84 (2), 195–204. Money, R. Bruce, Mary C. Gilly and John L. Graham (1998), “Explorations of National Culture and Word-of-Mouth Referral Behavior in the Purchase of Industrial Services in the United States and Japan,” Journal of Marketing, 62 (October), 76–87. Nahapiet, Janine and Sumantra Ghoshal (1998), “Social Capital, Intellectual Capital, and the Organizational Advantage,” Academy of Management Review, 23 (April), 242–66. Nunnally, Jum (1978), “Psychometric Theory,” New York: McGraw-Hill. Ozanne, Urban B. and Gilbert A. Churchill Jr. (1971), “Five Dimensions of the Industrial Adoption Process,” Journal of Marketing Research, 8 (August), 322–8. Palmatier, Robert W. (2008), “Interfirm Relational Drivers of Customer Value,” Journal of Marketing, 72 (July), 76–89.

321

Payan, Janice M. and Richard G. McFarland (2005), “Decomposing Influence Strategies: Argument Structure and Dependence as Determinants of the Effectiveness of Influence Strategies in Gaining Channel Member Compliance,” Journal of Marketing, 69 (July), 66–79. Podolny, Joel M. and James N. Baron (1997), “Resources and Relationships: Social Networks and Mobility in the Workplace,” American Sociological Review, 62 (October), 673–9. Riitta, Katila, Jeff D. Rosenberger and Kathleen M. Eisenhardt (2008), “Swimming with Sharks: Technology Ventures, Defense Mechanisms and Corporate Relationships,” Administrative Science Quarterly, 53 (2), 295–332. Rindfleisch, Aric and Christine Moorman (2001), “The Acquisition and Utilization of Information in New Product Alliances: A Strength-of-Ties Perspective,” Journal of Marketing, 65 (April), 1–18. and (2003), “Interfirm Cooperation and Customer Orientation,” Journal of Marketing Research, 40 (November), 421–36. Seibert, Scott E., Maria L. Kraimer and Robert C. Liden (2001), “A Social Capital Theory of Career Success,” Academy of Management Journal, 44 (2), 219–37. SRI International (2008). The 2005 Golf Economy Report. January 17. Tedeschi, James T. and Valerie Melburg (1984), “Impression Management and Influence in the Organization,” in Research in the Sociology of Organizations, Bacharach Samuel B. and Lawler Edward J., eds. Greenwich, CT: JAI Press. Ustüner, Tuba and David Godes (2006). “Better Sales Networks,” Harvard Business Review, July–August: 102–12. Uzzi, Brian (1997), “Social Structure and Competition in Interfirm Networks: The Paradox of Embeddedness,” Administrative Science Quarterly, 42 (March), 35–67. Uzzi, Brian and Shannon Dunlap (2005), “How to Build Your Network,” Harvard Business Review, 83 (December), 53–60. Wagner, Janet, Richard Ettenson and Jean Parrish (1989), “Vendor Selection Among Retail Buyers: An Analysis by Merchandise Division,” Journal of Retailing, 65 (Spring), 58–79. Wasserman, Stanley and Katherine Faust (1994), “Social Network Analysis: Methods and Applications,” New York: Cambridge University Press. Webster, Frederick E. Jr. (1968), “On the Applicability of Communication Theory to Industrial Markets,” Journal of Marketing Research, 5 (November), 426–8. (1970), “Informal Communication in Industrial Markets,” Journal of Marketing Research, 7 (May), 186–9. Weitz, Barton A., Harish Sujan and Mita Sujan (1986), “Knowledge, Motivation, and Adaptive Behavior: A Framework for Improving Selling Effectiveness,” Journal of Marketing, 50 (October), 174–91. Wuyts, Stefan, Stefan Stremersch, Christophe Van den Bulte and Philip Hans Franses (2004), “Vertical Marketing Systems for Complex Products: A Triadic Perspective,” Journal of Marketing Research, 41 (November), 479–87.