lable at ScienceDirect

Computers in Human Behavior 56 (2016) 225e237

Contents lists avai

Computers in Human Behavior

journal homepage: www.elsevier .com/locate/comphumbeh

Full length article

Social presence, trust, and social commerce purchase intention: Anempirical research

Baozhou Lu a, *, Weiguo Fan b, Mi Zhou b

a School of Economics & Management, China University of Petroleum, 66 West Chang Jiang Rd, Qingdao, 26655, Chinab Pamplin College of Business, Virginia Polytechnic Institute and State University, 3007 Pamplin Hall, Blacksburg, VA 24061, USA

a r t i c l e i n f o

Article history:Received 30 May 2015Received in revised form18 November 2015Accepted 29 November 2015Available online 14 December 2015

Keywords:Social commerceSocial presenceOnline social commerce marketplacesOnline trustPurchase intention

* Corresponding author.E-mail addresses: bzlu@upc.edu.cn (B. Lu), wfan@

edu (M. Zhou).

http://dx.doi.org/10.1016/j.chb.2015.11.0570747-5632/© 2015 Elsevier Ltd. All rights reserved.

a b s t r a c t

Lacking the presence of human and social elements is claimed one major weakness that is hindering thegrowth of e-commerce. The emergence of social commerce might help ameliorate this situation. Socialcommerce is a new evolution of e-commerce that combines the commercial and social activities bydeploying social technologies into e-commerce sites. Social commerce reintroduces the social aspect ofshopping to e-commerce, increasing the degree of social presences in online environment. Drawing uponthe social presence theory, this study theorizes the nature of social aspect in online SC marketplace byproposing a set of three social presence variables. These variables are then hypothesized to have positiveimpacts on trusting beliefs which in turn result in online purchase behaviors. The research model isexamined via data collected from a typical e-commerce site in China. Our findings suggest that socialpresence factors grounded in social technologies contribute significantly to the building of the trust-worthy online exchanging relationships. In doing so, this paper confirms the positive role of social aspectin shaping online purchase behaviors, providing a theoretical evidence for the fusion of social andcommercial activities. Finally, this paper introduces a new perspective of e-commerce and calls moreattention to this new phenomenon of social commerce.

© 2015 Elsevier Ltd. All rights reserved.

1. Introduction

A notable difference between online and offline markets that ishindering the growth of e-commerce is the decreased presence ofhuman and social elements in the online environment (Cyr, Head,Larios, & Pan, 2009; Hassanein, Head, & Chunhua, 2009). Thislean nature of online environment is often mentioned one of themajor drawbacks of e-commerce (Pavlou & Gefen, 2004). It isclaimed to eliminate social cues (e.g., body language), imposeadditional unique risks (Lee, 1998), and consequently impair thebuilding of the trustworthy atmosphere online. However, this sit-uation has been greatly improved recently by incorporating Web2.0 capabilities into the e-commerce website. This new evolution iscommonly referred to as the birth of social commerce (SC) (Huang& Benyoucef, 2013; Shin, 2013; Yadav, De Valck, Hennig-Thurau,Hoffman, & Spann, 2013). New design features built upon socialmedia and Web 2.0 technologies, including recommendation lists,

vt.edu (W. Fan), mizhou@vt.

ratings, comments, social proof, and reciprocity applications(Huang & Benyoucef, 2013; Olbrich & Holsing, 2011), enhancecustomer participation and allow them to collect socially rich in-formation, resulted in a more trustworthy and sociable onlinetransaction environment. Although these positive influences havebeen widely recognized, the social aspect of e-commerce has notbeen completely understood, nor their impacts on purchasingdecision.

In this study, Social Presence Theory (SPT) is employed as thetheoretical lens to understand the impacts of social shopping fea-tures in online SC marketplaces. To account for various aspects ofsocial commerce, a multi-dimensional conceptualization of socialpresence is proposed based on previous studies (Biocca, Harms, &Burgoon, 2003; Caspi & Blau, 2008; Shen & Khalifa, 2009). Thentheir influences on purchasing decision are examined in a researchmodel by using trust as the key mediating variable. In sum, thisstudy tries to offer several potential contributions.

First, it re-conceptualizes social presence as a multi-dimensional construct in social commerce context, and thus,overcomes the limitations of the unidimensional conceptualizationin literature. In doing so it sheds light on the nature of social aspect

B. Lu et al. / Computers in Human Behavior 56 (2016) 225e237226

in SC context by using SPT, contributing a set of social factors thathave impacts on trust. Second, it offers a full understanding ononline purchase behaviors by considering the influences of bothsocial factors and structural factors. Finally, it discloses the impor-tance of the social aspect on online purchase behaviors, callingmore attention to this relatively new but important research area infuture.

The rest of the paper is organized as follows. We beginwith thetheoretical development, including discussing the emergence ofsocial commerce phenomenon and proposing a multi-dimensional model of social presence. Next, we present theresearch model and the hypotheses, followed by a description ofthe research methodology. The paper then presents the researchresults, which is followed by a discussion of key findings andcontributions, as well as the implications for both research andpractice. Finally, we discuss the limitations of this study, thepossible direction of future research, and provide the concludingremarks.

2. Theoretical background and conceptual development

2.1. Social commerce

As a relatively new phenomenon, social commerce has evolvedquickly in practice (Barnes, 2014; Kim& Park, 2013;Wang& Zhang,2012). A recent report byMcKinsey (Chui et al., 2012) estimates thatthe use of social technologies can contribute $900 billion to $1.3trillion in value, and that up to 1/3 of consumer spending is subjectto influence from social commerce. Another report by Barclays(2012) indicates that by 2021 nearly half of the UK consumerpopulation will be engaged in social commerce. The term of socialcommerce was first coined by Yahoo in 2005 to denote onlineplaces where people can share experiences, get advices from oneanother, find goods and services and then purchase them (Mardsen,2010). Its early applications can be found in the late 1990s whenAmazon introduced the rating and review systems. The increasedpopularity of social technologies over last couple of years, includingsocial media, web 2.0 and social networks, has spawned anexpanded range of social commerce tools and opportunities (Liang& Turban, 2011; Mardsen, 2010).

Recently, SC generally refers to as the delivery of e-commerceactivities and transactions via the social media environment (Liang& Turban, 2011). It is viewed as a new evolution of e-commerce(Huang& Benyoucef, 2013;Wang& Zhang, 2012). Liang and Turban(2011) summarized three major attributes of SC: social technolo-gies, community interactions, and commercial activities. Thus, SCcan be considered a subset of e-commerce that involves using socialtechnologies to assist e-commerce transactions and activities(Yadav et al., 2013). In essence, SC is a combination of commercialand social activities (Liang & Turban, 2011; Zhou, Zhang, &Zimmermann, 2013). Traditional e-commerce sites, such asAmazon and Taobao, have added social applications and content tohelp people to connect where they usually buy. Considering thedominance of marketplace-based e-commerce like Amazon andTaobao, we try to uncover how social factors engendered by socialapplications shape the beliefs and behaviors of buyers in online SCmarketplaces.

2.2. Social aspect of online shopping

Shopping has always been a social activity. Consumers tend tobe influenced by their social interactions with others when makingpurchase decisions (Godes et al., 2005). E-commerce focuses moreon maximizing efficiency and the one-way interactions betweencustomers and the system (Huang & Benyoucef, 2013). Online

transactions are usually facilitated and guaranteed by structuralfactors such as escrow services and credit card guarantees (Fanget al., 2014; Pavlou & Gefen, 2004). Social technologies reintro-duce the social side into online purchasing process, making onlinepurchasing a more social experience. They also greatly increasesthe firm ability to directly initiate and manage social interactioneither impossible or too costly in the past (Chen, Wang, & Xie,2011). Thus, while e-business concentrates more on businessgoals, SC is more oriented toward social goals, such as networking,collaborating and information sharing, with a secondary focus onshopping (Wang & Zhang, 2012). Online buyers are able to get ac-cess to social knowledge and experiences to support them in betterunderstanding their purchase purposes, and in making moreinformed and accurate decisions (Dennison, Bourdage-Braun, &Chetuparambil, 2009).

While prior studies offer insights on how social interactionsshape buyer behaviors, such as, word-of-mouth (WOM), obser-vational learning, and social support (Amblee & Bui, 2011; Chenet al., 2011; Trusov, Bucklin, & Pauwels, 2009), they may haveoverlooked the overall effects of the social context (Kreijns,Kirschner, & Jochems, 2003). In SC, buyers are able to get moresocial cues to support their purchasing decisions by collectingmore information from the communities, by observing the actionsof other buyer, or by interacting with online sellers. Huang andBenyoucef (2013) proposed a conceptual model to summarizethe social design features of SC along four layers including theindividual, conversation, community and commerce levels, shownin Table 1. They argued that the key distinction between e-com-merce and social commerce is that the former usually only sees anindividual layer while the latter usually sees a community built onconversation. We argue that the social design features applied inthese four layers enrich social information, make buyers feel moreconnected with others, and finally enhance a social context online.SPT has been indicated a suitable theoretical lens for under-standing the social context in e-commerce. SPT suggests that so-cial presence is built upon signals transmitted in a communicationmedium, such as virtual agents (Hess, Fuller, & Campbell, 2009),IT-enabled human-like interaction (Pavlou, Liang, & Xue, 2007),socially-rich text, personalized greetings (Gefen & Straub, 2004),chat (Qiu & Benbasat, 2005) or message boards (Cyr, Hassanein,Head, & Ivanov, 2007). Thus, the social design features willconvey various types of social presence (seen in Table 1) that willbe discussed in next section.

2.3. A multi-dimensional conceptualization of social presence insocial commerce context

The concept of social presence is grounded in social presencetheory that elaborates the ability of a communication medium totransmit social cues (Short, Williams, & Christie, 1976). Defined as“the salience of the other in a mediated communication and theconsequent salience of their interpersonal interactions” (Shortet al., 1976), SP is viewed as an inherent quality of a communica-tion medium. From a psychological standpoint, SP also closely re-lates to intimacy and psychological closeness (Short et al., 1976). Inthis perspective, SP is often measured as the perceived warmth,conveying a feeling of human contact, sociability, and sensitivityembodied in a medium (Rice & Case, 1983). Most of prior e-com-merce research has adopted unidimensional model of SP, focusingon the capability of website to convey a sense of human warmthand sociability.

However, this unidimensional conceptualization of SP mightnot be suitable for virtual community, where people not onlyinteract with the computer mediated medium, but also need tocommunicate with other members and immerse themselves into

Table 1Social design features and social presence (adapted from Huang & Benyoucef, 2013).

Design principle Design features Amazon Taobao Dimension of SP

SP of web Perception of others SP of interaction

Individual Personal profile √ √ △ e e

Context profile √ √ △ e e

Activity profile √ √ △ e e

Conversion Social content presentation √ √ △ e e

Topic focus √ √ △ △ e

Notification � � e △ e

Information sharing √ √ △ △ e

Community Community support � √ e △ e

Connection � √ e △ e

Relationship maintenance � � e △ e

Commerce Group purchase � √ e △ e

Social proof √ √ e △ e

Authority √ √ e e

Reciprocity √ √ △ △ e

Participation � � e e e

Social ads and application � � e e e

Messenger tools with sellers � √ e e △

√ ¼ covered, � ¼ uncovered, △ ¼ relevant, e ¼ irrelevant.

B. Lu et al. / Computers in Human Behavior 56 (2016) 225e237 227

the virtual community. Thus, a multidimensional conceptualiza-tion of SP is advocated. Shen and Khalifa (2009) propose a three-dimension model of SP including awareness, affective SP andcognitive SP. Caspi and Blau (2008) indicate three differentconceptualizations of SP for online learning communities, i.e., asperception of other given the subjective quality of a medium, asself-projection onto the group, and as social identification. Tu(2002) suggests SP in online learning communities should havethree dimensions: social context, online communication, andinteractivity. Since social commerce is seen as the combination ofcommercial and community activities (Huang & Benyoucef, 2013;Liang & Turban, 2011), SP in social commerce should also beconceptualized as a multi-dimensional construct.

Based upon prior research, we propose a three-dimensionmodel of SP, named respectively, SP of the web, perception ofothers, and SP of interaction with sellers. Our conceptualization isakin to the SP model of Caspi and Blau (2008), which also containsthree factors: perception of the others, medium's impersonality,and medium as interaction enabling.

SP of the web refers to the capability of a website to convey asense of human warmth and sociability (Gefen & Straub, 2004;Hassanein et al., 2009). It might be the most adopted perspec-tive in prior e-commerce research, reflecting the inherent “sub-jective quality” of website. Most websites do not facilitate directinteraction with another human, but this does not mean that awebsite cannot convey social presence. Awebsite, as the containerof multimedia contents and socially rich text, can convey a senseof personal, sociable and sensitive human contact. Other IT arti-facts embedded into a website, e.g. physically embodied agents(Lee, Jung, Kim, & Kim, 2006), 3D avatar or videos, and Text-To-Speech voice (Qiu & Benbasat, 2005), also help to enhance theperceived SP. The provision of recommendations and consumerreviews in an e-commerce site also increases the SP of the website(Kumar & Benbasat, 2006). Table 1 indicates the social designfeature in SC that build up the perceived SP of a websiteGefe, suchas personal profile, context profile, social content, and informa-tion sharing.

The second dimension is perception of others (Caspi & Blau,2008), also named awareness. Perception of others refers to theextent to which other social actors appear to exist and react to theusers in online communities (Shen & Khalifa, 2009). Awareness inan online community is achieved through status updating, self-

presentation features and continuous participation in online dis-cussion. In SC, a few social applications increase the awareness ofother buyers who might feel interests in the same product or topic.For instance, social proof - a type of social application to resolvecustomer uncertainty about what to do or buy (e.g., the option of“customers who bought this also bought”) - will give buyers a hinton the existence of other buyers and their purchasing interest.WOM (or recommendation and review system) can also increaseperception of other online buyers. WOM valence indicates thepercentage of prior buyers who hold positive or negative opinions,and WOM volume plays an informative role by increasing the de-gree of buyer awareness (Chen et al., 2011). Revealing observationallearning information (Chen et al., 2011), which contains thediscrete signals expressed by the actions of other consumers (e.g.the percentage of adoption, wish list, Facebook “share” and “Like”button), will also let a buyer be aware of others buyers and theiractions. Other social applications relevant for perception of othersare suggested in Table 1, including reciprocity, group purchase,connection, etc. Since buyers in a SC marketplace can rely onmultiple sources to infer the presence of other buyers, socialpresence of others should be measured as the composite latentconstruct that is jointly influenced by the measures (MacKenzie,Podsakoff, & Jarvis, 2005).

The third dimension is SP of interaction with sellers (Caspi &Blau, 2008; Qiu & Benbasat, 2005). In traditional e-commerce,sellers rarely engage direct interactions with buyers. But onlinechat tools make these interactions possible. They can be used as aneffective marketing channel for sales, communication, andcustomer services. While prior studies indicate that SP is moreoften conveyed through the “imaginary interactions” with the webinterface (Hassanein & Head, 2007; Hess et al., 2009), thecomputer-mediated communication (CMC) tools such as customersupport chat (Qiu& Benbasat, 2005) andmessage boards (Cyr et al.,2007) can also convey social presence. This perspective of SP asdirect interactions with other users have been suggested in onlinecommunity research. For instance, Caspi and Blau (2008) pointedout that “medium as interaction enabling” is one factor of SP. Tu(2002) also considered interactivity as one dimension of SP. In SC,CMC tools have been used as one effective way to communicatewith buyers. For instance, Aliwangwang, an instant message tooldeveloped by Alilibaba for its commercial interactions, has ownedmore than 100 million active users. While for JD.com, another

B. Lu et al. / Computers in Human Behavior 56 (2016) 225e237228

leading online marketplace in China, QQ e a third party developedinstant message tool e is used by its members for the daily trans-actions. Considering the extensive usage of CMC tools in practice,we argue that interaction with sellers is one important dimensionof SP.

2.4. Trust in social commerce marketplace

Trust is a central aspect in many economic transactions(Fukuyama, 1995; Mayer, Davis, & Schoorman, 1995). When asocial environment cannot be regulated through rules and cus-toms, people tend to adopt trust as a reducer of social complexity(Luhmann, 1979). This is especially true for online transactionsdue to the absence of effective regulation over the opportunisticbehaviors of e-vendors. Thus, trust is often considered thefoundation of e-commerce (Keen, Ballance, Chan, & Schrump,1999) and the most crucial factor for the success of e-com-merce (Wang & Emurian, 2005). Prior e-commerce research hasfocused on finding key antecedents of trust or disclosing thetrust building mechanism (e.g., Lu, Hirschheim, & Schwarz, 2015;Ou & Sia, 2009). These studies focused more on the impacts offunctionality (e.g. usability, ease of use and usefulness) andinstitutional structures (e.g. structural assurance, situationalnormality and feedback mechanisms), paying very little atten-tion to social factors (Kim & Park, 2013) except social presence ofweb interface (e.g., Gefen & Straub, 2004; Hassanein & Head,2007). Trust, indeed, is built through social interactions withother people and the surrounding environment. Thus, socialcontext should be an important but neglected characteristic oftrust in prior literature.

Trust is a complex and multifaceted construct (Gefen,Karahanna, & Straub, 2003). It has been conceptualized in a vari-ety of ways. McKnight, Choudhury, and Kacmar (2002) tried toremove some of the conceptual confusion on trust by separatingbeliefs from intended behavior based on the theory of reasonedaction (Fishbein & Ajzen, 1975). They put trust into two broadcategories: trusting beliefs and trusting intentions and conceptu-alized trust as a set of specific beliefs including integrity, benevo-lence and ability. This conceptualization of trust is akin to that ofother studies adopting SPT (e.g., Hess et al., 2009). Then trustingbeliefs is conceptualized as a second-order construct in this study.Two types of trustees exist for a SC marketplace from a buyerperspective (Lu et al., 2015): marketplace (e.g., Amazon and eBay)and sellers resided in the marketplace. Trust in online sellers isconsidered as the major construct in this study, while trust in

Social Presence of Web

Perception of Others

Social Presence of Interaction with Sellers

Trust

Integrity Ben

H1

H2

H3

Comment Trust D

Fig. 1. Researc

marketplace is taken as control variable.

3. Research model and hypothesis development

The proposed research model is depicted in Fig. 1. In a SCmarketplace, online sellers can overcome the more impersonal,anonymous and automated stigma of online shopping by makingtheir virtual storefront socially rich (Kumar & Benbasat, 2002).Perceived social presences should help buyer shape their trustingbeliefs towards sellers. The model presents that each dimension ofsocial presence has a positive influence on trust in the sellers,which in turn, will shape online purchasing behaviors. The hy-potheses are explained in detail as below.

3.1. Social presence of web and trust

E-commerce is, in essence, a type of information system. Buyersconduct online transactions mainly through interacting with thewebsite. These buyer-web interactions could be viewed similar tointerpersonal interactions (Pavlou et al., 2007) if the website weretreated a social actor (Kumar & Benbasat, 2002). Since humaninteraction is viewed as a precondition of trust (Blau, 1964) thebuyer-web interactions should also contribute to the building oftrust online. A high SPwebsite conveysmore information and socialcues, and thus, is perceived to be more transparent; whereas in amore transparent environment the untrustworthy behaviors willbe inhibited. The SP of a website will also shorten the perceivedsocial distance between buyers and sellers (Pavlou et al., 2007). Andit is easier to form a trustworthy relationship when the perceivedsocial distance is short. Therefore, the SP of a web should enhancebuyers' trust towards online sellers. Prior studies have also sug-gested a positive impact of the SP of a website on trust (Hassaneinet al., 2009). Thus, in a SCmarketplace, the SP of aweb created by anonline seller will make him/her more trustworthy. Hence, we hy-pothesize that:

H1. The SP of web of will have a positive impact on trust in onlinesellers.

3.2. Perception of others and trust

Shopping has always been social activities. Social psychologyresearch indicates that human can learn from and be affected bythe knowledge and experiences of others who they know or trust(Mardsen, 2010). Persuasion can be extremely effective when it

In Sellers

evolent Competence

Purchase Intention H4

isposition Trust In Marketplace

Perceived Price Fairness

h model.

B. Lu et al. / Computers in Human Behavior 56 (2016) 225e237 229

comes from similar others (Cialdini, 2001), even if they arerandom strangers. Marketing literature also suggests that con-sumers' beliefs, attitudes and behaviors tend to be influenced bytheir social interactions with others when they make purchasedecisions (Godes et al., 2005). When shopping online, buyers arehard to validate the information offered by the sellers. However,they can rely on their peers who have visited the same storefrontfor indirect cues (Cialdini, 2001). Buyers not only can interactwith the e-commerce system, but also are able to make sense ofthe existence of other buyers based on various cues resided insocial applications such as customers' reviews (WOM), pick lists,popularity lists and transactional history. If social cues sendpositive signals (e.g. positive WOM valence, most of visitors wishto get this product, most of visitors have purchased this product…), a buyer will feel more confident in the seller's ability,benevolence and integrity in providing good services. Marketingliterature also indicates that when people observe the purchaseactions of all previous individuals, this publicly observed infor-mation outweighs their own private information in shaping theirbeliefs and behaviors (Chen et al., 2011). As a result, people tendto follow the lead of their online predecessors and becomeengaged in a type of “herd behavior” (Chen et al., 2011).Accordingly, we have:

H2. Perception of others will have a positive impact on trust in onlinesellers.

3.3. Social presence of interaction and trust

Online chat tools such as QQ, MSN or chat plug-ins embedded inthe website have been used for sellerebuyer interactions, throughwhich more social information will be revealed to buyers to formtheir trusting beliefs. In order to sustain a good customer rela-tionship, several methods are often employed, such as, choosingspecial words and sending the emotional icons like smile. Forinstance, a special language style, termed “Taobao Style”, is popularin Taobao communities. Buyers are greeted with the word of “Qin”(means “my dear” and rarely used in daily communication in Chi-nese culture but extensively used in online context). The tradedgoods are called “Baobei” (means “treasured objects”). “TaobaoStyle”, in together with other communication methods, makesellers feel friendlier and warmer, and thereby, will help to shortenthe perceived distance between them. Thus, the CMC chat tools arealso able to convey a sense of social presence. Buyers are also able tomake sense of the attitudes, benevolence, and integrity of sellersvia these compute mediated interactions, thereby forming beliefson sellers. The computer mediated interactions, such as, email andteleconference, have been argued to be able to convey SP and inturn to shape user beliefs (Qiu & Benbasat, 2005; Zack, 1993). Inonline learning communities, the text-based online discussion(Medium as interaction enabling) is found to be positively corre-lated with the cognitive aspects of learning (Caspi & Blau, 2008).Hence, we have:

H3. SP of interaction will have a positive impact on trust in onlinesellers.

3.4. Trust and purchase intention

Purchase intention is defined in this study as a buyer's intentionto purchase from a seller resided in a SC marketplace. According tothe Theory of Planned Behavior (Ajzen, 1991), behavioral intentionis the most influential predictor of behavior. Thus, in this study weuse purchase intention to represent purchase behaviors.

Taking part in online transactions requires buyers to deal withthe social complexity associated with the opportunistic behaviorsof sellers. Trust can be viewed as a significant antecedent belief thatcreates a positive attitude toward the transaction behavior(Jarcenpaa, Tractinsky, & Vitale, 2000), which in turn leads totransaction intentions. Trust helps reduce the social complexity andvulnerability that a buyer feels in e-commerce by allowing thebuyer to subjectively rule out undesirable yet possible behaviors ofthe e-vendor. As such, trust can help buyers reduce their risk per-ceptions when dealing with online vendors, thereby encouragingthem to engage in the “trust-related behaviors” with e-vendors,such as sharing information or making purchases (McKnight et al.,2002). Prior studies have shown that trusting beliefs in specificonline vendors are correlated with transaction intentions withthose same vendors (McKnight & Chervany, 2002; Pavlou, 2003).We hypothesize that the same logic can be extended to onlinesellers in a SC marketplace. Accordingly, we have:

H4. Trust in online sellers (trusting beliefs) will increase purchaseintention.

3.5. Control variables

The model incorporates additional control variables known toaffect trust and online behavior that are described below.

Online comment and rating systems provide fine-grained in-formation about a seller's reputation that is likely to engender abuyer's trust in a seller (Pavlou & Dimoka, 2006). Positive com-ments and ratings help buyer to form positive beliefs in sellers.Thus, its effect on trust in seller is controlled.

Trusting disposition is the degree that people in general can betrusted. Since trusting disposition has shown to increase trust, wecontrol for its effect on trust in seller.

Trust transference logic (Stewart, 2003) indicates that trust maybe transferred from a place, an industry association, or an entity toanother individual. In online context, trust of a marketplace cantransfer to the sellers associated with that marketplace. Priorresearch also indicates that trust in marketplace increases buyerintention to engage in online transactions (Lu et al., 2015). Thus, itseffects on both trust in seller and purchase intention need to becontrolled.

Consumers' perception of price fairness has been known tosignificantly affect their reactions toward sellers. Previous studieshave found that perceived price fairness directly affects consumers'purchase intention (e.g., Kahneman, Knetsch, & Thaler, 1986).Therefore, its effect on purchase intention is also controlled.

4. Research methodology

Following Gefen and Straub (2004), the free simulation experi-ment methodology is used in this study. In a free simulationexperiment, subjects are exposed to the research setting thatsimulates or duplicates the real-world situation and respondnaturally to tasks before answering questions about beliefs, atti-tudes, and observation (Fromkin & Streufert, 1976). Since no vari-ables are specifically manipulated, the technique is known as a“free” simulation experiment. The strength of this technique is thatit allows the researcher to observe and measure the subjectsengaged in real world task.

In selecting the sites to be included in this study, we deliber-ately chose online marketplaces that are well known and widelyused. Eventually, Taobao is selected as the real-world settingwhere subjects are required to complete tasks. Taobao is thelargest online marketplace in China. By 2010 it owned over 370million members. Transactions on Taobao totaled over 1 trillion

B. Lu et al. / Computers in Human Behavior 56 (2016) 225e237230

RMB in 2012 and reached 35 billion RMB on the single day ofNovember 11 (double 11) of 2013. More recently Taobao hasevolved into a SC marketplace as indicated in Table 1. Hence,Taobao represents a typical SC marketplace and is qualified as thefree simulation experiment site.

4.1. Instrument development

Wherever possible, the measurement items were adapted fromprevious research. All constructs were measured by reflective in-dicators except for the construct of “social presence of others”,which was measured with formative items. Information regardingon others buyers revealed by social applications can be summarizedinto three types, including others who might feel interest with theproduct (e.g., the wish list and the share button), others who haveactually bought the product (e.g., sales record), and others whohave provided positive information about the product (e.g., WOM).Therefore, social presence of others can bemeasured by these threeformative indicators.

As the original reflective items were in English, we used thefollowing procedures to ensure the translation validity. First, aresearcherwhose native language is Chinese forward translated theitems into Chinese. Next, another researcher independently back-ward translated the items into English. Subsequently, the two re-searchers compared and discussed the two English versions todevelop the first Chinese version of the items. The preliminary in-strument was pilot tested and reviewed by two faculty and threedoctoral students in IS field for clarity and face validity. Minor re-visions were made based on their feedbacks. Finally, the two initialtranslators checked this version together and finalized the Chineseinstrument.

We measured all items using seven-point Likert scales rangingfrom strongly disagree (1) to strongly agree (7). A further pretest ofthe survey instrument was conducted with 105 subjects, using thesame data collection method that would be applied in the actualdata collection. The items were modified following the proceduresrecommended by Churchill (1979). Please see the Appendix A forthe scales.

4.2. Experimental procedure

TheMBA and senior undergraduate students in business schoolsfrom two universities of China took part in the experiment. Theclasses in which the experiments were performed were randomlychosen with permission from the instructors. The subjects wereassured that the results would only be reported in aggregate toguarantee their anonymity and confidentiality. The activity wasperformed during class hours in two similar computer labs. In thisway, exogenous variance relating to hardware issues, networkresponse time, browser, purchase activity, and so on, wascontrolled. Subjects were first given a brief introduction of thestudy. After that, they were required to go through the process ofbuying from Taobao without actually completing the transaction byfollowing guidance. To simulate the real world situation as closelyas possible, subjects were allowed to freely select the productsbased on their real needs. Immediately after completing the task,subjects filled out a web-based instrument. The purpose to do thisis to refresh, in the mind of the buyers, the nature of the interactionwith the seller, thereby increasing the content validity by collectingresponses based on a recent activity. Finally, out of the 640 ques-tionnaires distributed, 546 valid responses were received after

removing invalid responses including those containing more thanfive missing values and those with the same answer to all ques-tions, resulting in a valid rate of 85.3%.

4.3. Sample statistics

Table 2 summarizes the descriptive information of the dataset.About 53.1% of the respondents were male, and 46.9% were female.A majority of respondents had online purchasing experiences(96.2%) and were regular Internet consumers (90% of subjectspurchase online at least once a month). As the free simulationexperiment was conducted in universities, a majority (77.3%) of therespondents were students aged between 18 and 30. Even thoughthe subjects are mainly MBA and undergraduate students, theywithout a doubt are the main Internet consumers in China. About60% of Internet consumers in China are aged between 18 and 30years old (iClick, 2014), and more than 60% of Taobao buyers haveundergraduate education or above (Lu, Zhao, & Wang, 2010). Theuse of students sample for e-commerce research was also advo-cated in previous studies (Walczuch & Lundgren, 2004). Thus, webelieve that our sample represents the major segment of Internetconsumers in China and should be appropriate for this study.

Following Armstrong and Overton (1977), nonresponse bias wasassessed by verifying that respondent demographics (age, genderand education) were not significantly distinct from current ChineseInternet consumers (iClick, 2014; iResearch, 2013).

5. Data analysis and result

The partial least square (PLS) approach was used for the dataanalysis considering its ability to deal with the formative constructs(Gefen, Straub, & Boudreau, 2000; Goo, Kishore, & Rao, 2009). PLSis a “second-generation” technique that has gained increasingpopularity in the field of marketing or information systems inrecent years (Becker 1991; Hair, Sarstedt, Ringle,&Mena, 2012). Foraccurate estimation and enough statistical power, minimum sam-ple size check is performed by following Chin (1998). A strong ruleof thumb suggests that the sample size should be at least equal tothe larger of the following: (1) ten times the scale with the largestnumber of formative indicators, or (2) ten times the largest numberof structural paths directed at a particular construct in the struc-tural model (Chin, 1998). The sample size for current study is 552,exceeding the minimum demand of sample size (10*7 ¼ 130) ac-cording to the rule of thumb. Therefore, our sample size isadequate. Next we will discuss the results of measurement andstructural model with the use of PLS.

5.1. Measurement model assessment

We first assessed the second-order construct of trusting beliefsby comparing two separate models by using LISREL (Goo et al.,2009). Model 1 hypothesizes that a unidimensional first-orderfactor accounts for the variance among all measurement items ofthe second-order construct (trust in seller) in this study. Model 2hypothesizes a second-order factor that accounts for the patterns ofcovariance among the first-order factors as conceptualized in thisstudy. Comparison of Model 1 (c2 ¼ 381.4, d.f. ¼ 35, RMR ¼ 0.061)and Model 2 (c2 ¼ 109.6, d.f. ¼ 32, RMR ¼ 0.035), indicates thatModel 2 is a better-fitting model with significant changes in chi-square (Dc2 ¼ 271.8, Dd.f. ¼ 3; p < 0.0001). Items loaded signifi-cantly on their assigned first-order factors, which in turn highly

Table 2Sample statistics (n ¼ 546).

Measure Items Frequency Percentage

Gender Male 290 53.1%Female 256 46.9%

Age 24 or below 246 45.1%>24 and �30 176 32.2%>30 and �35 81 14.8%>35 and �40 39 7.1%41 or above 4 0.7%

Have purchased online Yes 525 96.2%No 21 3.8%

Frequency of purchase online Several times a week 37 6.8%Several times a month 190 34.8%Once a month 267 48.9%Less than once a month 32 5.9%Never 20 3.7%

Types of product purchased online Clothing and footwear 236 43.2%Localized service 73 13.4%Computers and accessories 67 12.3%Food and health products 35 6.4%Books 39 7.1%Skin care & cosmetics 21 3.8%Sporting gear 19 3.5%Others 56 10.3%

Table 3The assessment of measurement model for constructs.

Constructs # of items Composite reliability Cronbach's alpha AVE Std loadings

Integrity 4 .92 .89 .75 IN1 (0.85) IN2 (0.86) IN3 (0.89) IN4 (0.86)Benevolence 4 .90 .85 .69 BEN1 (0.81) BEN2 (0.84) BEN3 (0.86) BEN4 (0.82)Competence 4 .88 .83 .67 COM1 (0.81) COM2 (0.85) COM3 (0.79) COM4 (0.82)Trust in seller 3 .91 .92 .77 IN (0.90) BEN (0.87) COM (0.87)Purchase intention 3 .93 .88 .81 PI1 (0.89) PI2 (0.91) PI3 (0.90)Trust in marketplaces 4 .93 .89 .76 TIM1 (0.88) TIM2 (0.88) TIM3 (0.86) TIM4 (0.86)Social presence of web 4 .93 .90 .77 SPW1 (0.87) SPW2 (0.88) SPW3 (0.90) SPW5 (0.87)Social presence of others 3 e e e SPO1 (0.18) SPO2 (0.64) SPO3 (0.39)Social presence of interaction 4 .89 .84 .68 SPI1 (0.81) SPI2 (0.77) SPI3 (0.87) SPI4 (0.85)Comment 4 .94 .92 .81 C1 (0.85) C2 (0.92) C3 (0.92) C4 (0.90)Trust disposition 5 .92 .89 .70 TD1 (0.72) TD3 (0.87) TD4 (0.87) TD5 (0.87) TD6 (0.85)Perceived price fairness 4 .94 .92 .80 PPF1 (0.85) PPF2 (0.80) PPF3 (0.89) PPF4 (0.86)

Note: All item loadings are significant at 0.01; the composite reliability score is calculated with the formula prescribed by Fornell and Larcker (1981).

B. Lu et al. / Computers in Human Behavior 56 (2016) 225e237 231

loaded on the second-order construct. The LISREL statisticsconfirmed our hypothesis of trusting in sellers as a second-orderconstruct.

A measurement model analysis for the reflective constructs wasconducted via the component part of PLS. We first refined themeasurement model by deleting the items with low loadings andcross loadings. The retained items are shown in Table 4. Internalreliability was verified with all composite reliability scores andCronbach's Alpha of the latent variables exceeding the 0.70threshold, as shown in Table 3. Convergent validity is oftendemonstrated by meeting the following criteria: significant item

Table 4Correlations of latent variables and evidence of discriminant validity.

1 2

Trust in seller 1 .88Purchase intention 2 .59 .90Trust in marketplaces 3 .46 .46Social presence of web 4 .57 .40Social presence of interaction 5 .64 .47Comment 6 .50 .50Trust disposition 7 .52 .42Perceived price fairness 8 .44 .53

Note: (1) Bolded diagonal elements are the square root of average variance extracted (AVEadequate discriminant validity; (2) Social Presence of Others is excluded since it is meas

loadings (>0.70) for each construct, adequate internal reliabilities(>0.70), and high AVE (average variance extracted) scores (>0.50).Evidence from Table 3 shows that all these criteria have beensatisfied for each construct, indicating convergent validity wasestablished. As shown in Table 4, the square root of the AVE for eachconstruct (diagonal term) exceeds the correlations between theconstruct and other constructs (off-diagonal terms), providingstrong evidence of discriminant validity. A low to moderate corre-lation coefficient (<0.60) (in Table 4) is also often considered evi-dence of discriminant validity.

The formative items might suffer from multicollinearity

3 4 5 6 7 8

.87

.32 .88

.42 .65 .83

.45 .30 .41 .90

.42 .43 .49 .37 .84

.35 .35 .42 .30 .28 .89

).These values should exceed inter-construct correlations (off-diagonal elements) forured by formative indicators.

B. Lu et al. / Computers in Human Behavior 56 (2016) 225e237232

(Diamantopoulos, 2011). The degree of multicollinearity for theformative items is measured by the variance inflation factor (VIF),which was calculated via SPSS. All the VIFs (1.99, 1.31, 1.83,respectively) are below the rule of thumb of 10 (or even the moreconservative value of 5) (Hair, Black, Babin, Anderson, & Tatham,2006), indicating that multicollinearity is not a serious concernfor the study.

The extent of common method bias was assessed with threetests. First, Harman's one-factor test (Podsakoff, MacKenzie, Lee,& Podsakoff, 2003) was performed by including all reflectiveitems with both principal axis factoring and principal componentfactoring. In both cases more than one factor emerged. Thevariance explained by the largest factor was 36.2% for principalaxis factoring, and 37% for principal component factoring. Bothresults are below the critical value of 50%. Second, the markervariable approach was performed by first identifying one makeritems that are not included in the research model and do nothave an explicit theoretical influence on the construct in theresearch model. The correlations between the marker item andthe items of interest have to be caused by the method (R€onkk€o &Ylitalo, 2011). The mean correlation coefficients values for themarker item is 0.035, indicating insignificant influence of theCMV. Results of the marker variable analysis are omitted forsimplicity. Third, the correlation matrix (Table 4) does not indi-cate any highly correlated factors (highest correlation is r ¼ 0.65),whereas evidence of common method bias should have resultedin extremely high correlations (r > 0.90) (Pavlou et al., 2007).These results indicate that the CMV is unlikely to distort theresults of this study.

5.2. Structural model and hypothesis testing

The results of testing the structural model are shown in Fig. 2.The t-values for the path coefficients were estimated via thebootstrapping procedure (Chin, 1998). The R2 numbers representthe percentage of variance explained for the dependent variables.All three dimensions of the social presences in online marketplaceare found to contribute positively and significantly to Trust inSellers, and respectively they are: Social Presence of Web (b ¼ 0.21,P < 0.01); Social Presence of Interaction (b ¼ 0.27, P < 0.01) andSocial Presence of Others (b ¼ 0.09, P < 0.01). Hence, hypothesesH1e3 are supported. Collectively, these three social presence con-structs account for almost 48 percent of the variance explained inTrust in Sellers. As hypothesized, Purchase Intention is found to besignificantly influenced by Trust in Sellers (b ¼ 0.38, P < 0.01),supporting H4. Trust in Sellers by itself can account for about 35

Web

Trust IR2

Integrity R2 = .81

BenevoR2 =

.21**

.09**

.27**

.90

.19**.1

Comment Trust Di

Fig. 2. PLS results of

percent of the variance explained in Purchase Intention. As ex-pected, all the controlled variables, except Trust in Platform(b ¼ 0.10, P < 0.10), have significant impacts on their assigneddependent variables. The results show that even when evenallowing for the effects of the control variables, social presenceconstructs still significantly contribute to buyer trusting beliefstowards online sellers and Trust in Seller is still a strong predicatorfor buyer purchase intention.

6. Discussion

6.1. Key findings and contributions

This study has several key findings. First, the results confirmthe significant positive impact of trusting beliefs on purchaseintention in the context of online SC marketplace even allowingthe effects of controlling variables. A substantial degree of trustingbeliefs in sellers' traits of integrity, benevolence and competenceis needed by online buyers before making purchase decisions.Thus, this study upholds the claim that trust is important in e-commerce (Fang et al., 2014; McKnight et al., 2002; Pavlou &Gefen, 2004) and extends it to the context of social commerce.Second, the results support our hypothesis of trusting beliefs as asecond-order construct composing of integrity, benevolence andcompetence, providing a more parsimonious conceptualization oftrusting beliefs in the context of both e-commerce and socialcommerce. Although trusting belies has been suggested a multi-dimensional construct (McKnight et al., 2002), it was oftentreated as a set of separate variables, increasing model complexityor making results suffer from high multicollinearity because ofhigh correlations among the dimensions of trust. Third, drawingupon the perspective of social presence theory, this study shedslight on the nature of social atmosphere of the social commercemarketplace by proposing and testing a set of three antecedents oftrusting beliefs: SP of web, perception of others, and SP of inter-action with sellers. All the three proposed social presence con-structs are found to be strong predicators of trust in seller, jointlyaccounting for a substantial degree of the variance in trust inseller. The results, thereby, suggest the vital role of social atmo-sphere in building buyer trusting beliefs in sellers in an online SCmarketplaces. Finally, the entire structural model with thecontrolled variables helps delineate the full picture in which bothsocial and structural factors shape buyer beliefs and purchasebehaviors in a SC marketplace.

The primary contribution of this study is to introduce a newset of social antecedent factors of trusting beliefs into a model

n Sellers = .55

lence .75

Competence R2 = .75

Purchase Intention R2 = .47.38**

.87 .87

6** .10 .18** .30**

** Significant level .01

* Significant level .05

Variance explained in bold

sposition Trust In Marketplace

Perceived Price Fairness

structural model.

B. Lu et al. / Computers in Human Behavior 56 (2016) 225e237 233

that explains the buyer purchase behavior in the online SCmarketplaces. By identifying the underlying social sources of e-trust in addition to the institutional and individual sources inliterature (Gefen & Pavlou, 2012; McKnight et al., 2002), thisstudy provide a more full understanding on online purchasebehaviors in SC marketplaces. Prior research concentrates moreon the impacts of institutional context and technological factorsfor trust building in e-commerce (e.g., Fang et al., 2014),neglecting the influences of the social context. However, socialcontext is an important source of trust (Luhmann, 1979). Thus,this study extends current trust research by introducing a newset of social factors relying social presence theory. Second,drawing upon the social presence theory, this study theorizes thenature of social aspect of shopping by surfacing three SPT-basedconstructs: SP of web, perception of others and SP of interactionwith online sellers. While social presence was often treated asone-dimensional construct that represents the static character-istic of web interface (Shen & Khalifa, 2009); this study hasproposed and validated it as a multi-dimensional concept in theSC context. Third, via the social presence constructs, we disclosehow the use of social applications and functions in e-commercewebsites helps build the trustworthy online environment andshape buyer perceptions and behaviors. Fourth, this study con-tributes to trust literature by separating the individual-level trust(trust in seller) with the institution-based trust (e.g., trust inmarketplace, structural assurance). It reflects the nature of re-lationships in online marketplace (Lu et al., 2015): buyers need totrust both the marketplace and sellers before making purchasedecision. Thus, two types of relationship exist in an onlinemarketplace: the dyadic buy-seller relationship reflected by theconstruct of trust in seller and the buyer-institution relationshipreflected by construct of trust in marketplace. Finally, by lookinginto buyer behaviors in online SC marketplaces, we introduce anew phenomenon e social commerce e that deserves moreattention in future study. Social influences in e-commerce havebeen overlooked or simplified in prior literature, however, themushrooming of SC call more attention to the social perspectiveof online shopping.

6.2. Implications for research

Online trust is viewed as a central concept of e-commerce inliterature (Keen et al., 1999) because of its substantial roles inmitigating negative perceptions (uncertainty and risks) and inenabling online transactions. Most e-commerce studies focused ona single website and viewed trust as a dyadic (one-to-one) rela-tionship between a buyer and a specific vendor. Since the domi-nance of online marketplace, this study examines trust in onlinemarketplace and extends it to the SC context. From a relationshipperspective, two types of relationships exist in online market-places: the online buyereseller relationship (trust in seller) and thebuyer-marketplace relationship (trust in marketplace). Thoughtrust in marketplace is controlled for this study, the results supportits significant positive effects on purchase intention. Thus, thebuyer-marketplace relationship also helps encourage onlinetransactions (Lu et al., 2015). This study contributes to e-commerceliterature in delineating the relationships in online marketplaces. Italso suggests that both types of relationships deserve specialattention in future studies.

Since trust (as a set of beliefs) is a key mediating variable inonline relationship models (Morgan & Hunt, 1994), a lot effort hasbeen put into finding the underlying sources of online trust in e-commerce literature. Trust are found to root in institutionalmechanisms (McKnight et al., 2002; Pavlou & Gefen, 2004), in-dividual characteristics such as trust disposition and familiarity

(Gefen & Straub, 2004), and website design features (Kim, Jin, &Swinney, 2009; Ou & Sia, 2009). There have been very fewstudies (e.g., Liang & Turban, 2011) that tried to identify ante-cedents of trust from the social perspective. However, the fastdevelopment of social commerce reintroduces the social aspect ofshopping to e-commerce, making it possible for online buyers torely on social cues in assisting their purchase decisions. This studyconfirms the positive impacts of social factors on trust (beliefs) bytheorizing the nature of social aspect in online marketplacesbased on the SPT perspective. In doing so, this study not onlysuggests that SPT is a relevant theoretical perspective, but alsoindicates that the social aspect of e-commerce deserve furtherexploration with a variety of theoretical lenses such as socialcapital theory.

To overcome the limitations of viewing as a one-dimensionalconstruct (Shen & Khalifa, 2009) and to better conceptualize thenature of social aspect of e-commerce, this study proposes andvalidates social presence as a multi-dimensional constructcomposing a set of three variables. Jointly they account for a sub-stantial degree of variance explained in trusting beliefs. A multi-dimensional conceptualization of social presence offers at leasttwo strengths. First, it can show which dimensions of SP havesignificant influences on online behaviors, providing a deep un-derstanding on how the social factors impact user perceptions,beliefs and behaviors. Second, it helps us uncover antecedent fac-tors for each dimension of SP. In online context, perceptions of SPare grounded in the use of social artifacts. A multi-dimensionalconceptualization of SP can disclose how various social artifactsaffect different SP dimensions (Shen & Khalifa, 2009). Online buyercommunities have been found to increase engagement in e-com-merce adoption (Liang & Turban, 2011; Lu et al., 2010). Therefore,we contribute a set of three SPT-based variables as social ante-cedent factors of trust that can be explored in future e-commerceand online community research.

As for the formative SP construct e social presence of others,the indicator of SPO1, reflecting the information about otherbuyers who might feel interest with the product, seems a weakitem (b¼ 0.18, P > 0.10). This suggests that in comparisonwith theinformation about who might want to buy, the social cuesregarding to others who have actually bought and providedcomments are more persuasive for an online buyer to trust aspecific seller. Thus, the findings of this study is consistent withthe study of Chen et al. (2011), in which the positive observationallearning information was found to have a significant impact onbuyers' purchase actions. Since this is the first endeavor toconceptualize social presence of others in terms of social com-merce context, our specification of this construct might be notcomprehensive (Diamantopoulos, 2011), we encourage futurestudies to continue to explore this construct.

It should be noticed that the above findings are based on thedata collected from China, which is significant different fromWestern culture. Although previous research revealed that therewere no significant differences for the effects of social presence ontrust between Chinese and Western culture (Hassanein et al.,2009), further examination of our model in the context of West-ern culture might been necessary before the generalization theresearch results.

Finally, from a descriptive perspective, this study describes theprocess by which a set of social factors facilitate online exchangerelationships through the key mediating role of trusting beliefsafter controlling the effects of a few key variables. This studysuggests that in the context of social commerce buyer trustingbeliefs depend on both structural sources and social sources.Research models that examine online buyer behaviors in socialcommerce without considering either one of their effects may

B. Lu et al. / Computers in Human Behavior 56 (2016) 225e237234

result in problems of “errors of exclusion” (Benbasat & Zmud,2003).

6.3. Implications for practice

Our research also has several practical implications. First,although embedding social technologies into e-commerce sites hasbeen widely adopted by practitioners, its effectiveness is rarelyexamined and confirmed in literature. This study shows that theperceived social presences conveyed by the use of social technol-ogies can increase buyer trusting beliefs towards sellers, which is anecessary precursor of online purchase. Thus, to some extent, thecombination of social technologies with e-commerce applicationsis effective, though the effectiveness of some social applicationsneed to be further examined. This encourages the practitioners tocontinue to include social applications into their site but with amore careful examination on their effectiveness.

Second, our results indicate that the social presence conveyedby the web content (text, pictures, videos, etc.) and the socialpresence conveyed by interaction with sellers can significantlyincrease buyer trusting beliefs. The results of this study, combinedwith previous research (Hassanein et al., 2009), suggest that on-line sellers might wish to invest in creating and maintaining thesetwo effective social presence channels with potential buyers.Furthermore, this study indicates that perception of other buyersalso has a positive impact on buyers' trusting beliefs towards aspecific online seller. But not all information revealed about otherbuyers contribute significantly. Only information regarding withothers who have actually bought and who have provided posi-tively information is useful in shaping online buyers' trustingbeliefs. This indicates that online platform managers shouldclosely monitor the effects of the information revealed aboutother online buyers.

Finally, our results uphold the claims that trust is critical for thesuccess of electronic commerce in literature (Lu et al., 2010) andextends it to the SC context. Our findings suggest that both socialfactors and structural factors are necessary for trust building inonline SC marketplaces. This implies that building social environ-ment based on social technologies is as important as buildinginstitutional environment with the IT-enabled mechanisms foronline marketplace providers. Since most e-commerce websiteshave completed the building of reliable institutional environment,it is more urgent and important for e-commerce companies andmarketplaces providers to establish a more social environment forthe next step.

6.4. Limitations and suggestions for future research

This paper has a number of limitations. First, MBA and under-graduate students sample was used in this study. Althoughnonresponse sample bias was assessed and verified, the samplemay not represent actual online buyers of online marketplace.Second, a single website (Taobao) data was used for this study. Inspite of that Taobao is the largest SC marketplace in China,including more sites for data collection will increase the externalvalidity of the results. Additional research can examine the modelby using actual online buyer data with the inclusion of more sites.Third, the model was tested only with the data collected fromChinese culture. The results might not be generalized to othercultural context. Future study is also needed for testing the modelin different cultures. Fourth, the common method bias wasassessed and the results showed that common method variancewas not a major concern statistically, however, common methodbias cannot be absolutely ruled out. Finally, we have controlled theeffects of some important variables on trust; however, this paper

cannot address all these possible antecedents of trust, such as,reputation, branding, and previous purchase experiences. Addi-tional research is also needed to sort out the effects of thesevariables.

In addition to the aforementioned future research suggestions,this paper suggests several more research directions as follows.First, we have proposed and tested social presence as a multi-dimensional construct in this study. Future research can furtherexamine our conceptualization in different cultural context or byproposing new dimensions of social presence. Second, socialpresences are conveyed by the use of social applications andtechnologies. Then future research can study how various socialartifacts influence different types of social presence. Third, onlinebuyer community has been found to positively affect e-commerceadoption (Lu et al., 2010). Social presence theory is also a suitableperspective for online community research (Shen & Khalifa,2009). Thus, future research can explore the user behaviors inonline buyer communities by using the SPT perspective. Fourth,the focus of this study is to examine the effects of different typesof social presence on trusting beliefs. We use trust in marketplace(intentions) as the surrogate for institutional mechanism andother factors associated with marketplace. A more completemodel that includes both social factors and structural factors willbe more interesting for future research. Finally, we have shownthat the use of social technologies is effective for online trustingbuilding via a set of variables based on SPT. Future research cantheorize the nature of social aspect by employing social theoret-ical perspectives.

7. Conclusion

Drawing upon the social presence theory, this paper looks intothe nature of social aspect in online social commerce marketplacesby proposing a set of three social presence variables, SP of the web,perception of others and SP of interaction, respectively. A researchmodel that describes buyer purchase behaviors in online socialmarketplaces is proposed by including the three social presencevariables as social antecedents of trusting beliefs. In sum, this studysheds new light on both social commerce research and socialpresence theory as follows. First, it re-conceptualizes and validatesthe concept of social presence as a multi-dimensional construct insocial commerce, overcoming the limitations of one-dimensionalconceptualization in previous e-commerce research. Next, it con-tributes a new set of social presence variables that can be exploredin future research. Then this paper discloses the positive effects ofsocial presence variables on trusting beliefs, suggesting an effectiveway for trust building in online social commerce marketplaces, thatis, the combination of IT-enabled structural factors and the socialfactors based on the use of social technologies. Finally, we introducea new perspective of e-commerce that needs special attention forfuture investigations, that is, a new phenomenon of socialcommerce.

Fund sources

This work is supported by National Key Technology R&D Pro-gram of the Ministry of Science and Technology (Grant No.2013BAH54F03), the Ministry of education of Humanities and So-cial Science project (Grant No. 13YJC630105), the Doctoral Fund ofMinistry of Education of China (Grant No. 20120133120005), theSoft Science Fund of Shandong (Grant No. 2014RKE28002) andFundamental Research Funds for the Central Universities(13CX04039B).

B. Lu et al. / Computers in Human Behavior 56 (2016) 225e237 235

Appendix A. Instrument and measurement properties

Mean Std dev Loading

Principle variablesSocial presence of web (Gefen & Straub, 2004)SPW1 There is a sense of human contact in the web of the seller. 4.38 1.13 0.87SPW2 There is a sense of personalness in the web of the seller. 4.25 1.12 0.8SPW3 There is a sense of sociability in the web of the seller. 4.42 4.12 0.90SPW4* There is a sense of human warmth in the web of the seller.SPW5 There is a sense of human sensitivity in the web of the seller. 4.14 1.17 0.87Social presence of others (composite latent construct)SPO1 There are many others buyers feel interested with the product. 4.87 1.24 0.18SPO2 There are many others buyers sharing information regarding with the product 4.59 1.12 0.64SPO3 There are many others who have bought the product. 5.14 1.24 0.39Social presence of interaction (Caspi & Blau 2008; Hess et al., 2009)SPI1 I can make sense of the attitude of sellers by interacting with them via Aliwangwang. 4.71 1.10 0.81SPI2 I can imagine how they may look like by interacting with them via Aliwangwang. 5.02 1.10 0.77SPI3 There is a sense of human touch to communicate with sellers via Aliwangwang. 4.49 1.06 0.87SPI4 Communication via Aliwangwang was warm. 4.35 1.09 0.85

Trusting beliefs (Gefen & Straub, 2004; McKnight et al., 2002)BenevolenceBEN1 I believe that the seller would act in my best interest. 4.05 1.15 0.81BEN2 If I required help, the seller would do his/her best to help me. 4.51 1.12 0.84BEN3 The seller is interested in my well-being. 4.04 1.19 0.86BEN4 I expect that the seller's intentions are benevolent 4.59 1.03 0.82IntegrityIN1 Promises made by the vendor are likely to be reliable. 4.53 1.02 0.85IN2 I do not doubt the honesty of the vendor. 4.58 1.07 0.86IN3 The vendor is sincere and genuine. 4.60 0.97 0.89IN4 I expect that the vendor will keep promises he/she make. 4.80 0.95 0.86CompetenceCOM1 The vendor is competent and effective. 4.71 0.90 0.81COM2 The vendor performs his/her role very well. 4.90 0.91 0.85COM3 The vendor knows about the product. 5.11 1.04 0.79COM4 The vendor knows how to provide excellent service. 4.82 1.00 0.82

Purchase intention (Gefen & Straub, 2004)PI1 I am very likely to buy the product from seller. 4.95 1.14 0.89PI2 I would consider buying the product from the seller in the future. 4.91 1.07 0.91PI3 I intend to buy the product from the seller. 4.81 1.15 0.90

Control variablesComment (Pavlou & Dimoka, 2006)C1 The average rating of the vendor is high. 5.08 1.12 0.85C2 The comments on the product are positive. 5.05 1.05 0.92C3 The comments on the service are positive. 5.10 0.97 0.91C4 Overall, the comments on the vendor are positive. 5.21 0.99 0.90Trust disposition (Gefen & Straub, 2004)TD1 I generally trust other people 4.49 1.22 0.72TD2* I tend to count upon other peopleTD3 I generally have faith in humanity 4.52 1.10 0.87TD4 I feel that people are generally well meaning 4.79 1.03 0.87TD5 I feel that people are generally trustworthy 4.52 1.09 0.87TD6 I feel that people are generally reliable 4.42 1.12 0.85Trust in marketplaces (Pavlou & Gefen, 2004)TIM1 As the online marketplace, Taobao can be trusted at all times. 4.79 0.95 0.88TIM2 As the online marketplace Taobao can be counted on to do what is right. 4.53 0.98 0.88TIM3 As the online marketplace, Taobao has high integrity. 4.43 1.00 0.86TIM4 Taobao is a competent and knowledgeable online transaction platform. 4.77 0.97 0.86Perceived price fairnessPPF1 It is worth buying the product from this vendor with this price. 4.80 1.24 0.90PPF2 The price of the product is competitive. 4.75 1.21 0.86PPF3 My choice of this product with this price will be a wise one. 4.69 1.20 0.91PPF4 I am satisfied with the price. 4.75 1.27 0.90

*Item dropped from final analysis. All scales anchored by: 1-strongly disagree, 7-strongly agree.

References

Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and HumanDecision Processes, 50(2), 179e211.

Amblee, N., & Bui, T. (2011). Harnessing the influence of social proof in onlineshopping: the effect of electronic word of mouth on sales of digital micro-products. International Journal of Electronic Commerce, 16(2), 91e113.

Armstrong, J. S., & Overton, T. S. (1977). Estimating nonresponse bias in mail

surveys. Journal of Marketing Research, 19, 396e402.Barclays. (2012). Social commerce: The next generation of retail. Available at http://

www.barclayscorporate.com/content/dam/corppublic/corporate/Documents/research/S-Commerce-report.pdf.

Barnes, N. G. (2014). Social commerce emerges as big brands position themselves toturn “follows”, “likes” and “pins” into sales. American Journal of Management,14(4), 11e18.

Becker, G. (1991). A note on restaurant pricing and other examples of social

B. Lu et al. / Computers in Human Behavior 56 (2016) 225e237236

influences on price. Journal of Political Economy, 99(5), 1109e1116.Benbasat, I., & Zmud, R. W. (2003). The identity crisis within the IS discipline:

defining and communicating the discipline's core properties. MIS Quarterly,183e194.

Biocca, F., Harms, C., & Burgoon, J. K. (2003). Toward a more robust theory andmeasure of social presence: review and suggested criteria. Presence, 12(5),456e480.

Blau, P. M. (1964). Exchange and power in social life. New York: Wiley.Caspi, A., & Blau, I. (2008). Social presence in online discussion groups: testing three

conceptions and their relations to perceived learning. Social Psychology of Ed-ucation, 11(3), 323e346.

Chen, Y., Wang, Q., & Xie, J. (2011). Online social interactions: a natural experimenton word of mouth versus observational learning. Journal of Marketing Research,48(2), 238e254.

Chin, W. W. (1998). The partial least squares approach to structural equationmodeling. Modern Methods for Business Research, 295(2), 295e336.

Chui, M., Manyika, J., Bughin, J., Dobbs, R., Roxburgh, C., Sarazin, H., et al. (2012). Thesocial economy: Unlocking value and productivity through social technologies.McKinsey Global Institute Report.

Churchill, G. (1979). A paradigm for developing better measures of marketingconstructs. Journal of Marketing Research, 16(1), 64e73.

Cialdini, R. B. (2001). Harnessing the science of persuasion. Harvard Business Review,79(9), 72e81.

Cyr, D., Hassanein, K., Head, M., & Ivanov, A. (2007). The role of social presence inestablishing loyalty in e-service environments. Interacting with Computers,19(1), 43e56.

Cyr, D., Head, M., Larios, H., & Pan, B. (2009). Exploring human images in websitedesign: a multi-method approach. MIS Quarterly, 33(3), 539e566.

Dennison, G., Bourdage-Braun, S., & Chetuparambil, M. (2009). Social commercedefined. White paper #23747. NC: IBM Corporation, Research Triangle Park.

Diamantopoulos, A. (2011). Incorporating formative measures into covariance-based structural equation models. MIS Quarterly, 35(2), 335e358.

Fang, Y., Qureshi, I., Sun, H., McCole, P., Ramsey, E., & Lim, K. H. (2014). Trust,satisfaction, and online repurchase intention: the moderating role of perceivedeffectiveness of e-commerce institutional mechanisms. MIS Quarterly, 38(2),407e427.

Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention and behavior: An intro-duction to theory and research. Reading, MA: Addison-Wesley.

Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation models with un-observable variables and measurement error. Journal of Marketing Research,18(1), 39e50.

Fromkin, H. L., & Streufert, S. (1976). Laboratory experimentation. In B. Dunnette(Ed.), Handbook of industrial and organizational psychology (pp. 415e465). Chi-cago: Rand McNally College Publishing Company.

Fukuyama, F. (1995). Trust: The social virtues and the creation of prosperity. NewYork: The Free Press.

Gefen, D., Karahanna, E., & Straub, D. W. (2003). Trust and TAM in online shopping:an integrated model. MIS Quarterly, 27(1), 51e90.

Gefen, D., & Pavlou, P. A. (2012). The boundaries of trust and risk: the quadraticmoderating role of institutional structures. Information Systems Research,23(3e2), 940e959.

Gefen, D., & Straub, D. W. (2004). Consumer trust in B2C e-commerce and theimportance of social presence: experiments in e-products and e-services.Omega, 32(6), 407e424.

Gefen, D., Straub, D. W., & Boudreau, M. C. (2000). Structural equation modeling andregression: guidelines for research practice. Communications of the AssociationInformation Systems, 4(7), 1e70.

Godes, D., Mayzlin, D., Chen, Y., Das, S., Dellarocas, C., Pfeiffer, B., et al. (2005). Thefirm's management of social interactions. Marketing Letters, 16(3/4), 415e428.

Goo, J., Kishore, R., & Rao, H. R. (2009). The role of service level agreements inrelational management of information technology outsourcing: an empiricalstudy. MIS Quarterly, 33(1), 119e145.

Hair, J. F., Black, W. C., Babin, B. J., Anderson, R. E., & Tatham, R. L. (2006). Multi-variate data analysis (6th ed.). Upper Saddle River, NJ: Pearson Prentice Hall.

Hair, J., Sarstedt, M., Ringle, C. M., & Mena, J. (2012). An assessment of the use ofpartial least squares structural equation modeling in marketing research.Journal of the Academy of Marketing Science, 40(3), 414e433.

Hassanein, K., & Head, M. (2007). Manipulating perceived social presence throughthe web interface and its impact on attitude towards online shopping. Inter-national Journal of Human-Computer Studies, 65(8), 689e708.

Hassanein, K., Head, M., & Chunhua, J. (2009). A cross-cultural comparison of theimpact of social presence on website trust, usefulness and enjoyment. Inter-national Journal of Electronic Business, 7(6), 625e641.

Hess, T. J., Fuller, M., & Campbell, D. E. (2009). Designing interfaces with socialpresence: using vividness and extraversion to create social recommendationagents. Journal of the Association for Information Systems, 10(12), 237e272.

Huang, Z., & Benyoucef, M. (2013). From e-commerce to social commerce: a closelook at design features. Electronic Commerce Research and Applications, 12(4),246e259.

iClick. (2014). China eCommerce market analysis final report 2013. Available fromhttp://www.i-click.asia/insights/.

iResearch. (2013). 2012e2013 China Internet consumer behavior report. Availablefrom http://report.iresearch.cn/1901.html.

Jarcenpaa, S., Tractinsky, N., & Vitale, M. (2000). Consumer trust in an Internet store.Information Technology and Management, 1(2), 45e71.

Kahneman, D., Knetsch, J. L., & Thaler, R. H. (1986). Fairness and the assumptions ofeconomics. Journal of Business, S285eS300.

Keen, P., Ballance, G., Chan, S., & Schrump, S. (1999). Electronic commerce relation-ships: Trust by design. Englewood Cliffs, NJ: Prentice Hall.

Kim, J., Jin, B., & Swinney, J. L. (2009). The role of etail quality, e-satisfaction and e-trust in online loyalty development process. Journal of Retailing and Consumer,16(4), 239e247.

Kim, S., & Park, H. (2013). Effects of various characteristics of social commerce (s-commerce) on consumers' trust and trust performance. International Journal ofInformation Management, 33(2), 318e332.

Kreijns, K., Kirschner, P. A., & Jochems, W. (2003). Identifying the pitfalls for socialinteraction in computer-supported collaborative learning environments: a re-view of the research. Computers in Human Behavior, 19(3), 335e353.

Kumar, N., & Benbasat, I. (2002). Para-social presence and communication capa-bilities of a web site: a theoretical perspective. E-service Journal, 1(3), 5e24.

Kumar, N., & Benbasat, I. (2006). The influence of recommendations and consumerreviews on evaluations of websites. Information Systems Research, 4(17),425e441.

Lee, H. G. (1998). Do electronic marketplaces lower the price of goods? Communi-cations of the ACM, 41(1), 73e80.

Lee, K. M., Jung, Y., Kim, J., & Kim, S. R. (2006). Are physically embodied social agentsbetter than disembodied social agents?: the effects of physical embodiment,tactile interaction, and people's loneliness in humanerobot interaction. Inter-national Journal of Human-Computer Studies, 64(10), 962e973.

Liang, T. P., & Turban, E. (2011). Introduction to the special issue social commerce: aresearch framework for social commerce. International Journal of ElectronicCommerce, 16(2), 5e14.

Lu, B., Hirschheim, R., & Schwarz, A. (2015). Examining the antecedent factors ofonline microsourcing. Information Systems Frontiers, 17(3), 601e617.

Luhmann, N. (1979). Trust and power. London: Wiley.Lu, Y., Zhao, L., & Wang, B. (2010). From virtual community members to C2C e-

commerce buyers: trust in virtual communities and its effect on consumers'purchase intention. Electronic Commerce Research and Applications, 9(4),346e360.

MacKenzie, S. B., Podsakoff, P. M., & Jarvis, C. B. (2005). The problem of measure-ment model misspecification in behavioral and organizational research andsome recommended solutions. Journal of Applied Psychology, 90(4), 710e730.

Mardsen, P. (2010). Social commerce: Monetizing social media. Hamburg, Germany:Syzygy Deutschland Gmbh.

Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995). An integrative model of orga-nizational trust. Academy of Management Review, 20(3), 709e734.

McKnight, D., & Chervany, N. (2002). What trust means in e-commerce customerrelationships: an interdisciplinary conceptual typology. International Journal ofElectronic Commerce, 6(2), 35e59.

McKnight, D. H., Choudhury, V., & Kacmar, C. (2002). Developing and validatingtrust measures for e-commerce: an integrative typology. Information SystemsResearch, 13(3), 334e359.

Morgan, R. M., & Hunt, S. D. (1994). The commitment-trust theory of relationshipmarketing. Journal of Marketing, 58(3), 20e38.

Olbrich, R., & Holsing, C. (2011). Modeling consumer purchasing behavior in socialshopping communities with clickstream data. International Journal of ElectronicCommerce, 16(2), 15e40.

Ou, X., & Sia, C. L. (2009). To trust or to distrust, that is the question investigating thetrust-distrust paradox. Communications of the ACM, 52(5), 135e139.

Pavlou, P. A. (2003). Consumer acceptance of electronic commerce: integrating trustand risk with the technology acceptance model. International Journal of Elec-tronic Commerce, 7(3), 101e134.

Pavlou, P. A., & Dimoka, A. (2006). The nature and role of feedback text comments inonline marketplaces: implications for trust building, price premiums, and sellerdifferentiation. Information Systems Research, 17(4), 392e414.

Pavlou, P. A., & Gefen, D. (2004). Building effective online marketplaces withinstitution-based trust. Information Systems Research, 15(1), 37e59.

Pavlou, P. A., Liang, H., & Xue, Y. (2007). Understanding and mitigating uncertaintyin online environments: a principal-agent perspective. MIS Quarterly, 31(1),105e136.

Podsakoff, S. B., MacKenzie, B., Lee, J. Y., & Podsakoff, N. P. (2003). Common methodbiases in behavioral research: a critical review of the literature and recom-mended remedies. Journal of Applied Psychology, 88(5), 879e903.

Qiu, L., & Benbasat, I. (2005). An investigation into the effects of text-to-speechvoice and 3D avatars on the perception of presence and flow of live help inelectronic commerce. ACM Transactions on Computer-Human Interaction, 12(4),329e355.

Rice, R. E., & Case, D. (1983). Electronic message systems in the university: adescription of use and utility. Journal of Communication, 33(1), 131e152.

R€onkk€o, M., & Ylitalo, Y. (2011). PLS marker variable approach to diagnosing andcontrolling for method variance. In Proceedings of the international conference oninformation systems (ICIS 2011).

Shen, K., & Khalifa, M. (2009). Design for social presence in online communities: amultidimensional approach. AIS Transactions on Human-Computer Interaction,1(2), 33e54.

Shin, D. H. (2013). User experience in social commerce: in friends we trust. Behavior& Information Technology, 32(1), 52e67.

Short, J., Williams, E., & Christie, B. (1976). The social psychology of telecommuni-cations. London: Wiley.

Stewart, K. J. (2003). Trust transfer on the world wide web. Organization Science,

B. Lu et al. / Computers in Human Behavior 56 (2016) 225e237 237

14(1), 5e17.Trusov, M., Bucklin, R. E., & Pauwels, K. (2009). Effects of word-of-mouth versus

traditional marketing: findings from an internet social networking site. Journalof Marketing, 73(5), 90e102.

Tu, C. H. (2002). The relationship between social presence and online privacy.Internet and Higher Education, 5(4), 293e318.

Walczuch, R., & Lundgren, H. (2004). Psychological antecedents of institution-basedconsumer trust in e-retailing. Information & Management, 42(1), 159e177.

Wang, Y. D., & Emurian, H. H. (2005). An overview of online trust: concepts, ele-ments, and implications. Computers in Human Behavior, 21(1), 105e125.

Wang, C., & Zhang, P. (2012). The evolution of social commerce: the people, man-agement, technology, and information dimensions. Communications of the As-sociation for Information Systems, 31(5), 105e127.

Yadav, M. S., De Valck, K., Hennig-Thurau, T., Hoffman, D. L., & Spann, M. (2013).Social commerce: a contingency framework for assessing marketing potential.Journal of Interactive Marketing, 27(4), 311e323.

Zack, M. H. (1993). Interactivity and communication mode choice in ongoingmanagement groups. Information Systems Research, 4(3), 207e239.

Zhou, L., Zhang, P., & Zimmermann, H. D. (2013). Social commerce research: anintegrated view. Electronic Commerce Research and Applications, 12(2), 61e68.