Journal of Service Research 2007 Rosenbaum 257 70

http://jsr.sagepub.com/Journal of Service Research

http://jsr.sagepub.com/content/9/3/257The online version of this article can be found at:

DOI: 10.1177/1094670506295851 2007 9: 257Journal of Service Research

Mark S. Rosenbaum and Carolyn A. MassiahSupport on Customer Voluntary Performance

When Customers Receive Support From Other Customers: Exploring the Influence of Intercustomer Social

Published by:

http://www.sagepublications.com

On behalf of:

Center for Excellence in Service, University of Maryland

can be found at:Journal of Service ResearchAdditional services and information for

http://jsr.sagepub.com/cgi/alertsEmail Alerts:

http://jsr.sagepub.com/subscriptionsSubscriptions:

http://www.sagepub.com/journalsReprints.navReprints:

http://www.sagepub.com/journalsPermissions.navPermissions:

http://jsr.sagepub.com/content/9/3/257.refs.htmlCitations:

What is This?

- Jan 19, 2007Version of Record >>

at Alexandru Ioan Cuza on March 11, 2014jsr.sagepub.comDownloaded from at Alexandru Ioan Cuza on March 11, 2014jsr.sagepub.comDownloaded from

Service establishments would relish the opportunity tohave their customers display customer voluntary perfor-mance (CVP) behaviors, which refer to helpful, discre-tionary customer behaviors that support an organizationsservice performance and quality. This article draws onresource exchange theory to offer an explanation as to whysome customers display CVP in the form of customer citi-zenship and customer care behaviors. The data reveal thatcustomers who receive social-emotional support and, to alesser extent, instrumental support from other customersin a service establishment reciprocate by exhibiting CVPtoward the establishment and to customers in the estab-lishment. This article demonstrates that socially support-ive service environments are beneficial for customershealth and for organizational profitability.

Keywords: social support; commercial friendships;intercustomer support; customer voluntaryperformance; empathy

When men are thus knit together, by a love ofsociety, not a spirit of faction, and dont meet tocensure or annoy those who are absent, but to enjoyone another: when they are thus combined for their

own improvement, or for the good of others, or atleast to relax themselves from the Business of theDay, by an innocent and cheerful conversation,there may be something very useful in these littleInstitutions and Establishments.

Joseph Addison(as quoted in Oldenburg and Brissett 1982, p. 268)

It is 4:30 a.m. on a dark and cold January morning inChicago. Six older-aged men each sit in their parked carsoutside a neighborhood diner. Harriet, a morning waitress,arrives a few minutes later. The men follow Harriet intothe restaurant, and they begin their daily routines. Oneof the men fills the saltshakers, and one fills the sugarcontainers. Another man assembles the crayons for eachchilds menu, and another fills the ketchup bottles. By5:30 a.m., Harriet and the men have completed their tasks,and the restaurant opens for business. The men take theirusual positions at the counter, order breakfast, and kibitzamong one another. The men begin to depart at 9 a.m.,knowing that they will meet again the following morning.

Many service organizations would relish the idea ofcustomers not only displaying loyalty to their firms butalso voluntarily performing helping behaviors thatimprove their firms service performance and quality.

Journal of Service Research, Volume 9, No. 3, February 2007 257-270DOI: 10.1177/1094670506295851 2007 Sage Publications

When Customers ReceiveSupport From Other CustomersExploring the Influence of Intercustomer SocialSupport on Customer Voluntary Performance

Mark S. RosenbaumNorthern Illinois University

Carolyn A. MassiahUniversity of Central Florida

at Alexandru Ioan Cuza on March 11, 2014jsr.sagepub.comDownloaded from

Indeed, Bettencourt (1997) deemed such helpful discre-tionary behaviors, which are performed by customers andthat support the firms ability to deliver service quality, ascustomer voluntary performance (CVP; Bailey, Gremler,and McCollough 2001; Hsieh, Yen, and Chin 2004).Bettencourt envisioned CVP as behaviors in which cus-tomers promote a service establishment by exhibitingloyalty (i.e., spreading positive word of mouth; WOM),by offering an establishment suggestion for improvement(i.e., participation) or by displaying courtesy to an estab-lishments employees and customers (i.e., cooperation).

Bettencourt (1997) demonstrated that a customers sat-isfaction with and commitment toward a service firm areantecedents of CVP. Similarly, Bailey, Gremler, andMcCollough (2001) argued that a relationship exists betweenthe emotional value of a service encounter and a customerspropensity to display CVP. Although these findings areinsightful, a question worthy of further exploration pertainsto why some customers display CVP. Although many cus-tomers may indicate their satisfaction with and loyalty to afirm, not all of them display CVP. To investigate why somecustomers demonstrate CVP, we turn to resource exchangetheory (Brinberg and Wood 1983; Foa 1971; Foa and Foa1974; Hirschman 1987; Murgolo-Poore, Pitt, and Berthon2003).

Foas (1971) resource exchange theory posits thatpeople exchange six types of resources with one another:love (i.e., an expression of affectionate regard, warmth,or comfort), status, information, money, goods, andservices. These resources are organized on two dimen-sions: concreteness versus symbolism and particularismversus universalism. Concreteness refers to the degree oftangibility (e.g., concrete goods vs. symbolic informa-tion) related to a resource, and particularism refers to thevalue of the resource in relation to the person who deliv-ers it (e.g., exchanging love with a particular individualvs. exchanging money with an unknown individual). Foacontends that resources that are proximal to one anotherin terms of concreteness and particularism are morelikely to be exchanged than more distance resources. Forexample, a person who receives love is likely to returnlove to that person, as opposed to returning money.

By extending resource exchange theory to the market-place, as some marketing researchers have requested(Murgolo-Poore, Pitt, and Berthon 2003), we propose thatCVP provides a means by which customers can expressan affectionate regard to a service establishment and topeople in the establishment who provide them with lovein the form of social support.

Adelman and Ahuvia (1995) conceptualized social sup-port as a service providers verbal or nonverbal communi-cation to a client that facilitates an exchange. Althoughsocial support is communicated to people, this stance fails

to capture the human needs that social support satisfies orto expose its influence on a persons health (Hawkley et al.2003; Stroebe and Stroebe 1996) and longevity (Gileset al. 2005). For example, psychologists conceptualizesocial support as the perception or experience that one isloved and cared for, esteemed and valued, and part of asocial network of mutual assistance and obligations (Tayloret al. 2004, pp. 354-55). Similarly, health researchers viewsocial support as the social resources that persons per-ceive to be available or that are actually provided to themby nonprofessionals in the context of both formal supportgroups and informal helping relationships (Cohen,Gottlieb, and Underwood 2000, p. 4), such as a customerswho regularly meet in a commercial establishment.

The purpose of this article is to investigate social sup-port as an antecedent of CVP. More specifically, this arti-cle explores the relationship between the frequency withwhich members of a Golds Gym receive social support(i.e., emotional support, companionship, and instrumen-tal support) from other members, or intercustomer socialsupport (Adelman, Ahuva, and Goodwin 1994), and theextent to which they demonstrate CVP in the gym.

Although it is possible for customers to receive socialsupport from employees (Kang and Ridgway 1996) orfrom friends who shop together (Hayko and Baker 2004),we investigate intercustomer support among customerswho primarily interact with one another in a gym. We doso because relatively little is known about intercustomersupport (Clark and Martin 1994; Guenzi and Pelloni2004), and the sample site is ideal for exploring customerrelationships that form in the gym. Thus, we analyze theinfluence of commercial friendships, as opposed to pur-chase pals (Woodside and Sims 1976), on CVP.

FRAMEWORK AND CONTRIBUTION

Figure 1 represents the guiding framework of ourexamination. The framework illustrates that customersmay receive two types of social support from other cus-tomers in a service establishment: social-emotional sup-port (i.e., companionship and emotional support) andinstrumental support. We believe that the frequency withwhich customers receive social support from other cus-tomers in an establishment influences the extent to whichthey display CVP.

Bettencourt (1997) originally viewed CVP as consist-ing of three behaviors: loyalty, participation, and cooper-ation. We concur with Bettencourt but also draw on theorganizational citizenship paradigm (Organ 1988) anddeem loyalty, participation, and cooperation as examplesof customer citizenship. Organ (1988) conceptualizesorganizational citizenship behaviors (OCBs) as individual

258 JOURNAL OF SERVICE RESEARCH / February 2007


(e.g., employee) helping behaviors and gestures that areorganizationally beneficial but are not formally required.Organ defines OCB as behavior that is discretionary, notdirectly or explicitly recognized by the formal reward sys-tem, and that in the aggregate promotes the effective func-tioning of the organization (p. 4; see also Van Dyne,Graham, and Dienesch 1994). Thus, we extend Organsconcept into the services literature by considering thatcustomers may display OCB not only to organizations inwhich they are employed but also to establishments wherethey purchase products and services.

We further expand on Bettencourts (1997) work byconsidering that customers may display CVP by demon-strating feelings of empathetic concern (i.e., helping withthe goal of benefiting the other person; Post et al. 2002)and by having a personal sense of responsibility to helpother customers in a service establishment. This exten-sion is based on the empathy dimension of theSERVQUAL scale (Parasuraman, Zeithaml, and Berry

1988). This dimension suggests that customers evaluate aservice firm on the basis of its employees abilities toextend empathetic concern to them. Furthermore, Leeet al. (2005) point out that empathetic people also possessa sense of moral responsibility and internalized socialresponsibility to help others. Thus, we conceptualize cus-tomer altruism and personal responsibility as examples ofcustomer care. In summary, we suggest that CVP entailsa customer voluntarily displaying helping behaviors bydemonstrating customer citizenship and customer care.

The primary contribution of this article to the servicesliterature is that it offers a theoretical explanation as towhy some customers proactively display CVP behaviorstoward a service establishment and to other customers inthe establishment. As a result, this examination clarifieswhy some customers assume the role of partial employ-ees in service settings and why they display OCB. Bothof these topics are extremely germane to contemporaryservice research (Bove 2006).

Rosenbaum, Massiah / INTERCUSTOMER SOCIAL SUPPORT 259

FIGURE 1Hypothesized Relationships Among Social Support and CVP

CustomerCitizenship:Participation

Social-EmotionalSupportX11

.

.

X1 Y1

Y5

Y6

Y9

Y10

Y11

Y12

Y17

Y18

Y22

InstrumentalSupport

X17

.

.

X12Customer

Citizenship:Cooperation

CustomerCitizenship:

Loyalty

CustomerCare:

Empathy

CustomerCare:

Responsibility

D1

D2

D3

D4

D5

NOTE: CVP = customer voluntary performance. Dashed paths are not significant at p < .05.


Although marketing researchers have explored com-mercial social support in some depth (e.g., Adelmanand Ahuvia 1995; Price and Arnould 1999), the services,marketing, and product/brand community literaturesfocus on the companionship bonds (i.e., friendship) thatdevelop between service providers and their clients or onthe characteristics (e.g., consciousness of kind, sharedrituals, transference of product knowledge) and dynamics(e.g., geography, social context, community characteris-tics) of product/brand communities (Muniz and OGuinn2001) rather than on highlighting the rich, social sup-portive bonds that often form among customers in servicesettings. Indeed, beyond exploring companionship bondsamong otherwise satisfied customers who participatein brand/product communities, few marketing researchershave examined the influence of emotional support oncustomer behavior (cf. Rosenbaum 2006). Ironically,emotional support is more closely related to a personshealth and well-being than any other type of social sup-port (Stroebe and Stroebe 1996).

This article addresses this chasm and offers manager-ial insights into how socially supportive service environ-ments represent win-win situations for both customersand service organizations. All service organizations canbenefit from understanding why some customers opt toact as partial employees, which offers firms financial andservice quality benefits. Finally, this examination heedsKelley, Donnelly, and Skinners (1990) request forresearchers to consider both customers technical andfunctional quality when exploring customers acting aspartial employees. Although customer technical qualityconsists of what a customer provides to the encounter(e.g., customer citizenship), functional quality evaluatesthe humanistic manner in which customers behave inconsumption settings (e.g., customer care).

We organize the remainder of the article as follows:First, we conduct a literature review to define conceptsand to develop hypothesized relationships. Second, weempirically test the hypotheses from data collected frommembers of Golds Gym. The article concludes with adiscussion of the theoretical and managerial implicationsand limitations.

LITERATURE REVIEW

Most service encounters represent social encounters(Berry 1995) in which employee interpersonal skillsaffect customer satisfaction, perceived service quality,and approach behaviors (Baker, Levy, and Grewal 1992;Beatty et al. 1996; Parasuraman, Zeithaml, and Berry1988). However, because marketing researchers tradi-tionally explore the dyadic interface of companionship

between one service provider and one client (Goodwinand Gremler 1996; Gremler and Gwinner 2000; Gwinner,Gremler, and Bitner 1998; Price and Arnould 1999), littleis known about the influence of supportive bonds thatmay form between customers in a setting.

Goodwin (1997) developed a classification schemaregarding customer-customer interactions, suggestingthat customers in service settings form relationships withother customers, whereby they act as strangers, friends,or quasi-family members. It might be considered thatintercustomer support is nonexistent between customersas strangers but present as companionship between cus-tomers who are friends. As customers become quasi-fam-ily members, the social support they receive from eachother is likely to be analogous to the type of support thatpeople often receive from traditional family members.

Customers as Strangers

McGrath and Otnes (1995) conducted research on theinfluence of unacquainted customers who engage in short-lived, fleeting interactions with one another. They exposedthe existence of a Janus effect among strangers in thatalthough customers often help other customers in retailsettings, they also sometimes hinder them. The Januseffect is also evident in Harris, Baron, and Radcliffes(1995) empirical findings. They found that although 25%of IKEA shoppers believed that their short-lived conversa-tions with unacquainted customers enhanced their shop-ping enjoyment, 50% of shoppers believed otherwise.

However, interaction between unacquainted cus-tomers in settings in which they can engage in conversa-tion for a large amount of time (e.g., trains, Harris andBaron 2004; recreational tours, Arnould and Price 1993;medical waiting areas, Gross 1986) may assuage dissat-isfied customers by lowering their anxiety associatedwith a service exchange, by having other customers facil-itate the service process by acting as partial employees,and by supplying them with social interaction that allevi-ates boredom during their waiting in queue.

Customers as Friends

Customers who enter service establishments or shop-ping centers with friends, or so-called purchase pals(Woodside and Sims 1976), often do so because they useshopping as a means to obtain companionship from theirfriends (Hayko and Baker 2004). Although purchase palsobtain hedonic pleasures from socializing in climate-con-trolled malls, such relationships do not necessarily repre-sent a financial boon to retailers. For example, adolescentmales, or so-called mall rats, often linger in mall corri-dors and food courts (Lewis 1989), and two thirds of



senior citizens practice malingering, which refers togoing to shopping centers without a desire to purchaseanything (Graham 1991). However, many adolescentfemales, who tend to shop with girlfriends, often returnwith their mothers to purchase (Hayko and Baker 2004),and purchase pals often encourage their friends to buy(Woodside and Sims 1976).

Customers as Quasi-Family Members

Some customers may enter the marketplace to ful-fill not only consumption needs but also companionshipor emotional support needs (Rosenbaum 2006). House(1981) states that people typically rely on their family andfriends for support. However, many sociologists contendthat a by-product of modern society, which is character-ized with mobile populations, double-income wage earn-ers, and planned suburban communities, is that people,especially older-aged adults, find it increasingly difficultto obtain sufficient social support from traditional sources(Forman and Sriram 1991; Miller 2002; Putnam 2000).Despite these difficulties, people remain social beingswho require interaction with others; isolated, we shrivelup and die (Vanier 1999, p. 2).

On one hand, many consumers find it difficult to formand maintain supportive relationships in the contemporarymarketplace (Aubert-Gamet and Cova 1999). On the otherhand, Oldenburg (1999) conceptualizes third places aspublic places, such as a diner (Rosenbaum 2006), a beautysalon (Price and Arnould 1999), or a gym (Unger andJohnson 1995), in which some people often gather toengage in sociability. Thus, third places represent localesin which customers may sustain socially supportive rela-tionships with other customers and receive resources thatmost people receive from family, friends, and coworkers.

Receiving Social Supportin Service Establishments

Although researchers have conceptualized the exis-tence of various types of social supportive resources thatpeople exchange (Wills and Shinar 2000), the majority oftaxonomies conclude that three resourcescompanion-ship, emotional support, and instrumental supportareessential to human well-being (Fyrand et al. 2002;Helgeson 2003). Companionship provides people with apartner for activities. Emotional support provides peoplewith outlets for discussing their feelings and expressingtheir concerns and worries. Instrumental support pro-vides people with practical help, assistance with mun-dane activities, or financial aid. Of the three supporttypes, companionship and emotional support are per-ceived as the most essential to a persons health (Sorkin,

Rook, and Lu 2002). Indeed, Suurmeijer et al. (1995)merge companionship and emotional support into a sin-gle dimension: social-emotional support.

Demonstrating Appreciationfor Support Through CVP

By receiving social-emotional and instrumental support,people can remedy some of the negative symptoms associ-ated with loneliness (Rosenbaum 2006) as well as illnessessuch as heart disease, cancer, or other immune-deficiencydiseases (Steptoe et al. 2004; Uchino, Cacioppo, and Kiecolt-Glaser 1996). Thus, we posit that when customers receivesocial-emotional support or instrumental support from oth-ers in a service establishment, they demonstrate their appre-ciation to the establishment by exhibiting CVP.

Our speculation is based on resource exchange theory(Foa and Foa 1974). As mentioned previously, thistheory posits that people exchange six types of resources(i.e., love, status, information, goods, services, and money)according to their degree of similarity. We speculate that ifcustomers receive social support from others in an estab-lishment, which provides them with feelings of concern andlove, they will reciprocate by engaging in CVP behaviorsthat aid the firm in fulfilling its mission and its customers.

CUSTOMER CITIZENSHIP

Bettencourt (1997) conceives of CVP as voluntary cus-tomer behaviors that assist an organization, such as cus-tomers acting as partial employees (i.e., participation),promoting the organization through WOM (i.e., loyalty),and cooperating with employees (i.e., cooperation). Par-tial employees refer to customers who contribute to thedevelopment and delivery of an organizations servicequality, similar to an organizations employees (Bowen1986; Hsieh,Yen, and Chin 2004; Keh and Teo 2001; Millsand Morris 1986). Gremler and Brown (1999) deem WOMas communication about a service provider that is offeredby a customer who is not trying to obtain monetary gainby doing so. Cooperation pertains to services that aresimultaneously produced and consumed. Consequently,because customers are often present during the serviceproduction process, they may have the ability to assist ina positive way in the service process (Zeithaml, Bitner,and Gremler 2006). For example, customers may bagtheir own groceries or present identification for check-writing purposes without being asked to do so.

We suggest that customers who display CVP throughloyalty, participation, or cooperation are actually exhibit-ing OCBs. Therefore, customer citizenship is a formof CVP in which a customer willingly spreads positive



WOM about a service establishment, acts as the estab-lishments partial employee, and cooperates with the estab-lishments employees.

Hypothesis 1: The frequency with which a customerreceives social-emotional support and instrumentalsupport from other customers in a service estab-lishment positively influences his or her willing-ness to exhibit CVP througha. participationb. cooperation, andc. loyalty.

CUSTOMER CARE

Drawing on resource exchange theory, we posit thatcustomers who receive social support in a service estab-lishment will respond by providing other customers in theestablishment with feelings of love and of genuine con-cern. A key theme in the services marketing literatureis that customers perceptions of service quality are influ-enced by the extent to which employees extend interper-sonal empathy to them. For example, the SERVQUALscale evaluates the importance of this behavior under theempathy dimension (Parasuraman, Zeithaml, and Berry1988; Zeithaml, Bitner, and Gremler 2006). However, cus-tomers may also extend empathy to other customers in aservice setting by exhibiting helping behaviors that are notthe result of a reward or a desire to avoid some punishment(Davis 1996; Post et al. 2002).

To further understand empathetic customers, we drawon the work of Lee et al. (2005), who found that exem-plary altruists also display a sense of internalized socialresponsibility to help other people who are experiencingdifficulties. Lee et al. state that altruists possess a keen abil-ity to recall difficulties that they experienced in situationsand to assist other people in overcoming experienced obsta-cles. Notably, marketing researchers have explored socialresponsibility in terms of corporate behavior (Maignan andFerrell 2004) rather than considering that customers maypossess an internalized prosocial responsibility (Georgeand Bettenhausen 1990) to help other customers.

Although myriad customer care behaviors exist, the lit-erature supports the conclusion that customer care is char-acterized by customers performing at least two behaviors:expressing empathy and expressing a sense of personalresponsibility to other customers in a service setting.

Hypothesis 2: The frequency with which a customerreceives social-emotional support and instrumentalsupport from other customers in a service estab-lishment positively influences his or her willing-ness to exhibit CVP through

a. customer empathy andb. customer responsibility.

METHOD

Sample

We obtained data through self-administered question-naires from a sample of 207 members of a franchisedGolds Gym located in a major metropolitan city. Thegym permitted one of the authors to set up a table near thefront door of the gym over a 1-week period for data collec-tion purposes using convenience sampling. Each respon-dent who volunteered to participate in the study, whichrequired approximately 10 min for completion, received aGolds Gym calendar that was valued at $5. Of the respon-dents, 104 were male, and 99 were female. Four respon-dents failed to indicate their sex on the questionnaire. Theaverage age of the respondents was 35 years, with a min-imum of 18 years and a maximum of 65 years.

Measures

Social support. We assessed the frequency with whichthe members received social support from other nonrela-tional gym members with the Social Support Question-naire for Transactions (SSQT; Doeglas et al. 1996;Suurmeijer et al. 1995). The SSQT comprises 23 ques-tions that ask respondents to indicate the frequency (1 =seldom or never, 2 = now and then, 3 = regularly, and4 = often) with which they receive support from othermembers. The questionnaire asked respondents to con-sider their relationships with other members in the gymwho were not related to them, and they were verbally toldthis before receiving a questionnaire.

Because the SSQT has limited exposure in the mar-keting literature (Rosenbaum 2006), we verified thedimensional structure of the SSQT with factor analysis.We subjected the 23 SSQT items to principal axis factoranalysis (promax rotation, which we employed becauseof expected correlation between the factors; Thompson2004). The first analysis retained all items in a three-fac-tor solution, which accounted for 59% of the variance.After we removed items with low pattern coefficients orcross-loadings, the final analysis retained 17 items in atwo-factor solution, accounting for 65% of the vari-ance. A parallel analysis confirmed a two-factor solution(OConnor 2000); in addition, Rosenbaum (2006) founda two-dimensional structure when employing the SSQTin a commercial setting.

Following Suurmeijer et al.s (1995) conceptualizationsfor the SSQT, we labeled the first factor social-emotional



support and the second factor instrumental support. Wecreated two support scales by summing together the itemsthat were representative of each factor. The coefficientalphas were .93 and .90 for the Social-Emotional andInstrumental Support scales, respectively, indicatinginternal consistency (Nunnally 1978). Table 1 illustratesthe SSQT items, pattern coefficients, and alpha coefficients.

Customer citizenship. We performed a principal axisfactor analysis (varimax rotation) on 11 items that assesscustomer citizenship behaviors. We adapted the citi-zenship items from Bettencourt (1997); however, weeliminated 6 of Bettencourts original items that wereapplicable primarily to consumer behaviors at grocerystores (e.g., items regarding shopping carts, bagging, andcheck writing). Similar to Bettencourts results, the finalanalysis retained 79% of the variance in a three-factorsolution (i.e., Participation, a 5-item scale; Coopera-tion, a 4-item scale; and Positive WOM/Loyalty,a 2-item scale). Next, we assessed scale reliability by calculating coefficient alpha and by analyzing item-to-item correlation. Reliability estimates were .90, .77,and .84 for the Participation, Cooperation, and Loyaltyscales, respectively (Nunnally 1978). The scale items,

pattern coefficients, and alpha coefficients appear inTable 2.

Customer care. We created an original 11-item scaleto measure the extent to which customers extend feelingsof empathetic concern to and personal responsibility forother customers. We derived 7 items from Daviss (1996)Empathetic Concern scale. Although the EmpatheticConcern scale evaluates feelings of empathy, concern,and sympathy toward others, it fails to evaluate the extentto which people sense a personal responsibility to protector to aid others. Therefore, we generated 4 new items toevaluate the extent to which customers feel responsiblefor the well-being of other customers in a retail setting.We assessed each item from 1 (strongly disagree) to 7(strongly agree).

To analyze the factor structure of the scale, we subjectedthe items to principal axis factor analysis (promax rotation).The analysis retained 55% of the variance in a two-factorsolution. The analysis of the pattern coefficients resulted inour retaining 11 items. A parallel analysis confirmed atwo-factor solution (OConnor 2000). On the basis of thepattern coefficients, we labeled the first factor customerempathy and the second factor customer responsibility.


TABLE 1SSQT Items, Pattern Coefficients, and Reliability Estimate

F1: Social-Emotional F2: InstrumentalItem (1 = seldom, 2 = now and then, 3 = regularly, 4 = frequently) Support Support

Social-Emotional Support ScaleHow often do members in the gym reassure you about things? .84 .02How often do members in the gym tell you not to lose courage? .82 .13How often do members in the gym perk you up or cheer you up? .79 .02How often do members in the gym give you advice in the right direction? .79 .09How often do members in the gym lend you a friendly ear? .78 .02How often do members in the gym show their understanding to you? .76 .03How often do members in the gym sympathize with you? .74 .06How often do members in the gym give you information or advice? .63 .16How often are members in the gym friendly to you? .63 .06How often do members in the gym make you feel at ease? .62 .06How often can you rely on other members in the gym? .58 .18Mean = 28.57, standard deviation = 6.97, Cronbachs = .93

Instrumental Support ScaleHow often do members from the gym lend you small amounts of money? .07 .88How often do members from the gym drop in to your home for a pleasant visit? .02 .87How often do members from the gym help you do odd jobs such as helping you .01 .86

move furniture, drive you somewhere, etc. How often do members from the gym lend you small things, like tools, or something like that? .00 .80How often do members in the gym lend you valuable things? .05 .80How often do you do things like walking, shopping, going to the movies or sports, etc. .14 .65

together with other members from the gym?Mean = 8.12, standard deviation = 3.31, Cronbachs = .90

SOURCE: Doeglas et al. 1996; Suurmeijer et al. 1995.NOTE: SSQT = Social Support Questionnaire for Transactions. Principal factor analysis (promax rotation). Total variance explained = 63.29%.Correlation between the factors = .49.


We created scales by summing together the items that wereidentified by their pattern coefficient values as being repre-sentative of a specific factor. We assessed scale reliabilityby calculating coefficient alpha. Reliability estimates were.80 and .82 for the Empathy and Responsibility scales,respectively, indicating adequate internal consistency(Nunnally 1978). Table 3 illustrates the items, pattern coef-ficients, and alpha coefficients. It is worth noting that thecustomer citizenship and customer care items were pre-sented to the respondents in a random order to preventrespondent bias and to lessen potential fatigue.

FINDINGS

Path Analysis

To prepare the ordinal data for structural equationmodeling, we normalized the data in PRELIS (du Toit etal. 2006), which enabled better estimates for maximumlikelihood estimation. We used structural equation mod-eling to estimate the theoretical model depicted in Figure1. After we removed two nonsignificant paths, the fitfor the corrected model was acceptable (root mean squareerror of approximation = .08, comparative fit index = .94,standardized root mean square residual = .09; Hu andBentler 1999), especially given the small sample size.Table 4 shows the parameter estimates for the relation-ships depicted in Figure 1.

Relationships

Customer citizenship. In terms of participation, bothsocial-emotional support ( = .45, p < .01) and instru-mental support ( = .29, p < .01) emerge as significantpredictors of the extent to which gym members exhibitparticipatory or partial employee behaviors, with social-emotional support being a more powerful predictor.Thus, the results provide support for Hypothesis 1a. Forcooperation, both social-emotional support ( = .59,p < .01) and instrumental support ( = .17, p < .05) rep-resent significant predictors of a members propensityto cooperate with the gyms employees; however,social-emotional support again is a more significantpredictor than instrumental support. Thus, Hypothesis1b is supported. Given the predictive strength of social-emotional support frequency over instrumental supportfrequency, the finding that social-emotional support issolely predictive of a respondents WOM/loyalty ( =.55, p < .01) is not surprising; Hypothesis 1c is partiallysupported.

Customer care. The relationship between social-emotional support and the extent to which a memberdisplays empathic concern toward other members wassignificant ( = .54, p < .01), but the relationship betweeninstrumental support and empathic concern was not signif-icant. As a result, Hypothesis 2a is partially supported. Interms of people feeling responsible for other customers,


TABLE 2Customer Citizenship Scalesa

F1: Customer F2: Customer F3: CustomerParticipation Cooperation WOM/Loyalty

Customer Participation ScaleIf I notice a problem at the gym, I inform an employee, or customer, even if it doesnt affect me. .92 .20 .15I make constructive suggestions to the gyms employees on how to improve its service. .91 .21 .14I let the gyms employees know of ways that they can better serve my needs. .90 .17 .13I have shared a useful idea on how to improve the gym with its employees. .88 .18 .15When I experience a problem at the gym, I let an employee know. .85 .14 .13

M = 17.26, SD = 8.28, Cronbachs = .90

Customer Cooperation ScaleI try to help keep this gym clean (e.g., pick up towels that may not be yours, wipe equipment). .26 .80 .23I give the gyms employees my full cooperation. .26 .79 .19I go out of my way to treat the gyms personnel with kindness and respect. .15 .78 .09I carefully observe the rules and policies of this gym. .06 .77 .08

M = 16.49, SD = 5.87, Cronbachs = .77

Customer WOM/Loyalty ScaleI encourage my friends and relatives to work out at this gym. .21 .14 .90I say positive things about this gym to others. .17 .28 .87

M = 10.88, SD = 8.79, Cronbachs = .84

NOTE: Principal axis factor analysis (varimax rotation). Total variance explained = 79.36%. WOM = word of mouth.a. 1 = strongly disagree and 7 = strongly agree.


both social-emotional support ( = .54, p < .01) and instru-mental support ( = .22, p < .01) emerge as significant pre-dictors, supporting Hypothesis 2b. However, consistentwith the preceding statistical results, social-emotional sup-port emerges as the more dominant predictor.

These findings illustrate that social support is indeedan antecedent of CVP; however, social-emotional support,not instrumental support, represents the primary driver

regarding the extent to which customers willingly displayCVP. Emotional support has been found to have moreinfluence on a persons health than any other type ofsocial support (Stroebe and Stroebe 1996), and compan-ionship is also essential to human well-being. As a result,the finding that the combined influence of emotional andcompanionship support has a greater influence on CVPthan does instrumental support is understandable.


TABLE 3Customer Care Scalesa

F1: Customer F2: CustomerEmpathy Responsibility

Customer Empathy ScaleI would describe myself as a pretty soft-hearted person to other members in the gym. .85 .03I am touched by things that I see happen to other members in the gym. .69 .08Other members problems in the gym do not usually disturb me a great deal.RC .65 .02When I see a member in the gym being taken advantage of, I feel kind of protective toward them. .64 .08I dont feel very sorry for other members in the gym when they have problems (e.g., using equipment).RC .54 .08

When I see a member in the gym being treated unfairly, I sometimes dont feel much pity for them.RC .51 .11M = 29.01, SD = 6.47, Cronbachs = .80

Customer Responsibility ScaleI often feel that I have a special responsibility to assist other members in the gym when they need help.b .04 .79Other members have done things for me that I feel I am responsible for repaying them.b .17 .73I frequently look for opportunities to help other members in the gym.b .17 .65I often have tender, concerned feelings for customers in the gym who are less fortunate then me. .04 .64I try to stay aware of when other members in the gym are having difficulties at the gym.b .03 .63

M = 29.01, SD = 6.47, Cronbachs = .82

NOTE: Principal axis factor analysis (promax rotation). RC = reverse coded. Total variance explained = 51.30%. Correlation between the two factors = .56.a. 1 = strongly disagree and 7 = strongly agree.b. A new item; all other items are from Davis (1996).

TABLE 4Standardized Parameter Estimates for CVP Measures

Relationship Unstandardized Coefficient Standardized Coefficient t

Social-emotional support Participation 2.92 .45 5.19**Instrumental support Participation 0.62 .29 3.83**Social-emotional support Cooperation 3.63 .59 5.66**Instrumental support Cooperation 0.34 .17 1.99*Social-emotional support WOM/loyalty 7.51 .55 6.01**Instrumental support WOM/loyalty ns nsSocial-emotional support Customer empathy 1.94 .54 5.11*Instrumental support Customer empathy ns nsSocial-emotional support Customer responsibility 2.09 .54 5.37**Instrumental support Customer responsibility 0.28 .22 2.68**R2

Customer participation .45Customer cooperation .49Customer WOM/loyalty .30Customer empathy .30Customer responsibility .48

NOTE: CVP = customer voluntary performance; WOM = word of mouth.*p < .05. **p < .01.


Gender Analysis and ClusterAnalysis Procedures

To assess the impact of social support on a membersCVP in more depth, we performed two additional pro-cedures. First, we conducted a one-way multivariateanalysis of variance (MANOVA) to determine theimpact of gender on five dependent CVP variables (i.e.,Participation, Cooperation, Loyalty, Empathy, and Partic-ipation scales). We found significant differences amongthe member types on the measures (Wilkss = .94), F(5,181) = 2.37, p < .05. The multivariate 2 based onWilkss lambda indicted a medium effect at .06. Table 5contains the means and standard deviations on the depen-dent variables. The follow-up analyses of variance(ANOVAs) on each variable indicated that gender differ-ences existed only on one measure, namely, customerempathy, F(1, 185) = 7.40, p < .01, 2 = .04. Overall,female members (M = 5.03) are more likely than malemembers (M = 4.63) to demonstrate empathy to othermembers; however, the effect of gender on empathy isquite mild.

Second, we performed a two-step cluster analysis(Statistical Package for the Social Sciences 2005) toevaluate the frequency with which gym membersreceive intercustomer social support, regardless of thesupport type, and its effect on CVP. The data set com-prised gym members responses to the 17 purifiedSSQT items. On the basis of the Bayesian informationcriterion, we classified the respondents into two groups.Of the respondents, 88 (42.5%) were placed in the firstcluster, which we labeled infrequent support, and 92(44.4%) were placed in the second cluster, which welabeled frequent support. A total of 27 (13.0%) respon-dents were unclassified.

Mean Analysis

We conducted a one-way MANOVA to determine theeffect of the frequency with which respondents receivesocial support from other customers on the aforemen-tioned five dependent CVP variables. We found signifi-cant differences among the member types on themeasures (Wilkss = .74), F(5, 153) = 11.01, p < .001.The multivariate 2 based on Wilkss lambda was strongat .27. Table 5 contains the means and standard devia-tions of the dependent variables.

We conducted ANOVAs on each variable as follow-uptests. To minimize Type I errors, we tested each ANOVA atthe .01 level. The ANOVAs on participation, F(1, 157) =76.42, p < .001, 2 = .18, cooperation, F(1, 157) = 19.35,p < .001, 2 = .16, WOM/loyalty, F(1, 157) = 37.71, p /JPEG2000ColorImageDict > /AntiAliasGrayImages false /DownsampleGrayImages true /GrayImageDownsampleType /Bicubic /GrayImageResolution 300 /GrayImageDepth -1 /GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true /GrayImageFilter /DCTEncode /AutoFilterGrayImages true /GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict > /GrayImageDict > /JPEG2000GrayACSImageDict > /JPEG2000GrayImageDict > /AntiAliasMonoImages false /DownsampleMonoImages true /MonoImageDownsampleType /Bicubic /MonoImageResolution 1200 /MonoImageDepth -1 /MonoImageDownsampleThreshold 1.50000 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict > /AllowPSXObjects false /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox false /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile (U.S. Web Coated \050SWOP\051 v2) /PDFXOutputCondition () /PDFXRegistryName (http://www.color.org) /PDFXTrapped /Unknown

/SyntheticBoldness 1.000000 /Description >>> setdistillerparams> setpagedevice

Journal of Service Research 2007 Rosenbaum 257 70

Documents

Transcript of Journal of Service Research 2007 Rosenbaum 257 70