Information & Management 49 (2012) 89–98

Willingness to share information in a supply chain:A partnership-data-process perspective

Timon C. Du a,*, Vincent S. Lai a, Waiman Cheung a, Xiling Cui b

a Department of Decision Sciences and Managerial Economics, Chinese University of Hong Kong, Hong Kongb Business Administration Department, Hong Kong Shue Yan University, Hong Kong

A R T I C L E I N F O

Article history:

Received 21 May 2010

Received in revised form 16 February 2011

Accepted 18 May 2011

Available online 20 October 2011

Keywords:

Computer-mediated communication and

collaboration

Inter-organizational information system

IT-enabled supply chains

Questionnaire surveys

A B S T R A C T

To achieve an efficient and effective supply chain, information needs to be shared. Most current

information-sharing studies address the benefits gained from shared data, but neglect the effect of

willingness to share, in which the benefits of sharing data may be discounted. This study looks into the

factors that affect the extent of the willingness of companies to share information from a partnership-

data-process perspective. To distinguish the mode of sharing, we differentiate information sharing into

template based and proactive. Our results suggest that when partnerships become closer, the willingness

to share template-based information increases and consequently the willingness to proactively share

additional information.

� 2011 Elsevier B.V. All rights reserved.

Contents lists available at SciVerse ScienceDirect

Information & Management

jo u rn al h om ep ag e: ww w.els evier .c o m/lo c ate / im

1. Introduction

A supply chain is a network of material suppliers, productionmanufacturers, and logistics service providers that performdifferent value-added activities together, usually in a sequentialmanner, to produce value for consumers. Information sharing in asupply chain can occur in two ways. Internally, for the effectiveplanning of purchases and company growth, leading to flexibilityand coordination and a sense of ownership, and externally, sharinginformation with supply chain partners to enhance demandplanning, physical flows, and financial work processes [20]. It canalso prevent information distortion, resulting in problems such asthe ‘‘bullwhip effect.’’

Information sharing has garnered greater research attention inrecent years, but most studies have investigated the types ofinformation shared and the gains from sharing [5]. Further, thesestudies make the assumption that the institutions sharinginformation are willing to do so, however, willingness to shareinformation can be predetermined (where the data to be shared arespecified in a contract, with templates used to describe the dataformat) or spontaneous (where the process is voluntary and non-predetermined).

* Corresponding author at: CCS920, Department of Decision Sciences and

Managerial Economics, Chinese University of Hong Kong, Shatin, N.T., Hong Kong.

Tel.: +852 2609 8569; fax: +852 2603 5104.

E-mail address: timon@cuhk.edu.hk (T.C. Du).

0378-7206/$ – see front matter � 2011 Elsevier B.V. All rights reserved.

doi:10.1016/j.im.2011.10.003

Willingness to share reflects the quality of the informationshared, including its timeliness, accuracy, adequacy, completeness,and reliability. These dimensions, combined with the breadth ofinformation shared and level of coordinated knowledge involved,affect the quality of the decisions made by the firm [9].

In the supply chain context, willingness to share information isa trade-off between efficiency and the responsiveness of theinformation resources. What information is shared depends on theeconomics and technology of, while the questions of with whomand when information is shared require that social involvement betaken into account. This suggests that adopting a partnership-data-process (PDP) perspective can increase the partner’s willingness toshare information. In the supply chain context, the partners can besuppliers, buyers, or other service providers, and the partnershipcan extend to different strategic levels, such as causal or long-termalliances.

According to the PDP perspective, partnership and process arethe main determinants of information sharing, chiefly because ofthe uncertainties associated with partnership relationships andcollaborative processes. Successful supply chain collaborationinvolves partnership coordination, commitment, trust, highcommunication quality, participation, and joint problem solving,requiring a willingness to share information. Business processcomplexity is, of course, critically related to partnership successand the extent of information sharing.

Various types of data may be shared to improve theeffectiveness of a supply chain, including inventory level, demandforecasts, sales and order status, and production schedules. Such

T.C. Du et al. / Information & Management 49 (2012) 89–9890

sharing takes place at different levels, including physical, abstract,operational, and strategic. A greater correlation or interdepen-dence between working tasks increases the need for data sharingto facilitate cooperation [13]; hence, complex processes normallyrequire more dynamic data and involve customers or partners thathave a long-term relationship with the firm.

We investigated the extent of information sharing in terms ofthe willingness of a company to share information with itscollaborators. However, in contrast to most previous studies, welooked at the issue from a supply chain perspective, dividinginformation sharing into template based (when, how, what, andwith whom shared) or proactive (sharing of non-predeterminedinformation). We choose three collaborative partners in the supplychain – buyers, suppliers, and logistics service providers – as thetargets and conducted an exploratory empirical study through alarge-scale survey to investigate the partner’s willingness to shareinformation.

More specifically, we attempted to answer three questions:

1. Can information sharing behavior be classified into template-based and proactive behavior? If so, how do these behaviortypes interact?

2. Does the extent of partnership development, data character-istics, and business processes affect the way that information isshared in a supply chain? If so, what are their effects on thesharing behavior?

3. Do suppliers, buyers, and logistics service providers differ intheir information sharing behavior? If so, what are the causes ofthese behavioral differences?

2. Research background

2.1. Role of IT in supply chain collaboration

IT plays a fundamental role in developing workforce agility byproviding speed and flexibility, which are also critical to supplychain agility [26]. An agile supply chain involves close linkageamong supply chain partners in key processes, improvingprocurement, forecasting, supply chain management, and newproduct development. IT enables agility by allowing supply chainpartners to exchange planning and operational data, ranging frominformation on annual contracts and progress reports to real-timedelivery and invoicing data [22].

Recently there have been many studies of the effect of IT onsupply chains (e.g., [4,12]); in these, IT has been reported as havinga critical effect by improving supply chain flexibility andresponsiveness, which in turn improve the form’s competitiveness.Interorganizational systems (IOSs) are of particular importance infacilitating the development of an agile supply chain. These IT-based systems transcend legal enterprise boundaries, offeringconnectivity, cooperation, and coordination among supply chaincollaborators. IOSs generate consistent, timely information withvisibility to all collaborators.

Early efforts focused on the management of demand uncertainty,inventory control, material planning, and reducing cycle time.Typical IOS implementations to achieve supply chain agilityincluded electronic data interchange (EDI), advanced planningsystems (APSs), material requirement planning (MRP), manufactur-ing resource planning (MRP II), enterprise resource planning (ERP),and e-business systems. The deployment of these IOSs not onlyincreased information processing capability and enabled greaterinter-firm cooperation but also nurtured innovation in customerrelationship, manufacturing, procurement, supply chain, and otherkey activities [1]. An IOS can also generate a closer buyer–supplierrelationship and a more cooperative governance relationship [10].

However, the benefits of an IT-enabled agile supply chainsrequire a concerted effort among supply chain collaborators toalign and integrate their business and IT activities [27]. In thecontext of supply chain collaboration, the role of IT as a platformenabling agility emerges primarily through its complementarityand integration with organizational strategies, designs, structures,and competencies [22]. However, the integration, of strategicaspects across the supply chain involves structural, application,and cultural complexity and thus major supply chain growingpains. Adding to such complexity are the interwoven relationshipsamong the strategic dimensions. For example, IT alignmentrequires that supply chain partners be aligned in four fundamentaldomains of strategic choice: business strategy, IT strategy,organizational infrastructure and processes, and IT infrastructureand processes.

2.2. Studies of information sharing

When sharing information, individuals and organizations havedifferent motives. For individuals, cognitive and relational capital,etc. are important [25], whereas for organizations efficiency due tocost reduction, productivity improvement, and product/marketstrategy is important. In addition, when sharing information,individuals care more about privacy, while organizations are moreconcerned with security. To acquire information, individuals mustdepend on materials provided by strangers, while organizationsuse contracts to guarantee sharing. However, for both individualsand organizations, trust, coordination, and interdependence arecentral to the success of information sharing.

Studies of information sharing at the individual and organiza-tional level are typically rooted in social exchange and transactioncost theory, respectively. Sharing knowledge within departmentsis an internalization process that is closely related to theorganizational culture [16]. Individuals share information primar-ily because of common interest, generalized reciprocity, and pro-social behavior. Thus, members of a successful community act outof community rather than self-interest.

Studies of information sharing among organizations are com-monly based on a transaction cost approach. For example, Son et al.[23] developed a theory of power and reciprocal investment toinvestigate the effect of relationships and channel climate on EDIadoption, and found that investment and cooperation from suppliersincreased both EDI volume and diversity.

2.3. Supply chain partnerships and information sharing

Traditional supply chain management research has focused onoperational aspect by stressing the efficient flow of products andservices but recently, more emphasis has been placed on supplychain strategy in terms of relationship building for betterpartnership performance. Studies generally treat supply chainpartnerships as a continuum ranging from independent partner-ships to strategic partnerships based on the degree of itsinterdependence, its exclusivity, and its strategic goals. Informa-tion integrates supply chain partners to enhance partnershipsuccess and performance [3]; information sharing is obligatory andmay be split into four levels: order exchange, operationalinformation sharing, strategic information sharing, and strategicand competitive information sharing. However, strategic partnersshare both strategic and operational information, whereasoperational partners share only operational information. Inaddition, the more strategic the partnership, the greater thedegree of information sharing needed for real-time, dynamic,integrated business operations.

Supply chain partnerships take many forms including vendor-managed inventory (VMI) partnerships, supplier–managed

Data Dynamism

Partnership Extent

Process Complexity

Proactive Information Sharing (PIS)

Template-based Information Sharing

(TIS)

H2A

H2B

H3A

H3B

H4A

H4B

H7

H1

H5 H6

Fig. 1. Model of proactive information sharing.

T.C. Du et al. / Information & Management 49 (2012) 89–98 91

inventory partnerships, collaborative forecasting, and replenish-ment partnerships, and collaborative, planning, forecasting andreplenishment (CPFR) partnerships, all of which are seen incurrent businesses. In a VMI partnership, the supplier isresponsible for replenishing the customer’s inventory. Allretailers transmit their point-of-sale data to the vendor’s centraldatabase to facilitate centralized control and management. SMIpartnerships resemble VMI partnerships but the supplier takesresponsibility for managing the supply chain inventory. Althoughboth VMI and SMI partnerships are beneficial, they may fail ifretailers do not warn the central vendor or supplier of theirpromotion-related or local information [2]. CFAR partnerships,however, enable retailers and manufacturers to collaborate inforecasting demand and scheduling production, allowing partnersto collaborate and coordinate their scheduling and replenishmentdecisions by exchanging complex decision support models andstrategies in a manner that can only work if there is obligatoryinformation sharing. Adopters of this approach such as Wal-Mart,Warnier-Lambert, Procter and Gamble, and Levi Strauss, havereported that its implementation delivers promising resultsincluding increased sales, reduced transportation and logisticscosts, and lower levels of inventory investment. The CFAR modelhas been renamed CPFR (collaborative planning, forecasting andreplenishment) to emphasize the role of planning in thecollaborative process; it not only emphasizes demand forecastingand inventory replenishment, but also focuses on coordinatingproduction and purchase planning.

To be effective, partnerships should be designed to address twoimportant issues: the mechanism for integrating business pro-cesses and the mechanism for sharing real-time information.Although CFAR and CPFR partnerships provide better inventoryperformance, they draw substantially on organizational resourcesand require a willingness to share more information. In practice,effective partnerships inevitably involve the systematic sharing ofinformation to allow the collaborators to complete their tasksmore effectively. In addition, research findings have suggested thatsustainable collaboration, effective information sharing filling, andstrong partnership performance, require collaborators to focus onresources and commitment, intra-organizational support, corpo-rate focus [21], the use of technology, external and internal trust,and mutual benefits.

2.4. The partnership-data-process perspective

The focus of our study was organizational information sharingwhich can significantly reduce the cost of data entry and integrity.The relationship among partners is a major factor in determiningthe degree of sharing and the better the relationships, the morewilling they will be to share information, and trust is key to therelationship. In a study by Patnayakuni et al. [18], it was noted thatan e-business platform allows an organization to share informa-tion; but, a decision may be not to share in the short term asinformation asymmetry can be used as a source of rent. Also, ahigher level process via a formal mechanism positively affects thesharing of operational, tactical, and strategic information in thecollaborative relationships between firms and their suppliers.Subramani [24] argued that IT is used to improve firm performancein such areas as cost reduction and process efficiency. Higher IT usecan enhance the information exchange but IT used for explorationadopts unstructured interorganizational processes for new capa-bilities such as sharing knowledge and understanding theoperating environment. When information is shared voluntarily,it is positively related to profitability and productivity andnegatively associated with labor costs.

The relationship between information seekers and givers candetermine the degree of sharing, although information may

sometimes be given to weakly tied partners to benefit thecommunity. It has also been noted that the process structureand routine and domain knowledge can determine the extent andquality of sharing.

2.5. Research model and hypotheses

We developed an exploratory empirical model, as illustrated inFig. 1, to investigate the determinants of information sharingamong supply chain partners. Our focus was on factors that affectthe willingness to share: partnership extent (a measure of therelationship between a company and its supply chain partners,which may affect the degree to which information is shared andfrequency of change in its value over time), data dynamism, andprocess complexity (a measure of the business processes that takeplace in a supply chain, representing the degree of variability,difficulty, and interdependency of business tasks that must beresolved in the collaboration).

In the model, willingness to share information, our dependentvariable, was divided into two types – template-based information(TIS) and proactive information (PIS) sharing. In our study, thesetwo types were investigated independently to determine how eachtype would be affected by the extent of partnership development,data dynamism, and process properties.

2.6. Willingness to share information

Information sharing at the organizational level requires therelease of confidential and closely guarded financial and strategicinformation to partners who might have been or may later becompetitors. Therefore, organizations are cautious about theinformation they share and to what extent and to whom it shouldbe divulged. When determining the potential options for sharing,regularly or spontaneously, and via formal or informal channels,organizations may devise their strategy by deciding on theirwillingness to share, which depends on trust and commitment inthe partnership. Raban and Rafaeli investigated informationownership and willingness to share by classifying sharingwillingness into share and not share.

Our study was based on the assumption that classifyingwillingness to share into template-based and proactive levels mayallow organizations to evolve their collaboration from explorationto strategic. Relational exchange theorists have confirmed that amulti-level sharing strategy establishes key relational norms foreffective inter-organizational relationships and that it can be basedon informal arrangements and self-reinforcement rather thanstrict formal contracts and third-party control.

Firms involved in sharing data need to evaluate the informationshared on an ongoing basis and update its content. Hence, firms

T.C. Du et al. / Information & Management 49 (2012) 89–9892

may prefer to extend the scope of their information sharing fromcontract-based order and operational information to dynamicexchange of tactical and strategic information once they havesuccessfully cultivated partnership trust, not only as a means tosave on transaction costs, but also to optimize partnershipperformance and benefits [17]. This is particularly true whenthe extent of contract-based information sharing is greater. Wetherefore proposed:

Hypothesis 1. The greater is the extent of template-based infor-mation sharing, the greater is the extent of proactive informationsharing.

2.7. Partnership extent and information sharing

Partnership involves behavioral characteristics that include thebuilding of a relationship to achieve shared goals. Partnershipextent, depends on the reasons for collaboration; alliances havedifferent levels, such as casual (where the shared information islimited to orders and historical data), long-term (involving a closerrelationship with more extensive sharing), and outsourcing orbusiness (with a higher level of cooperation and thus a greaterneed for trust). Consequently, more intensive communication,supported by a wider spectrum of data between strategic partnersshould lead to their being better informed and thus more confidentin the relationship and willing to make it successful.

When information sharing is confined to template-based data,the relationship between partners is likely to be causal and theneed for collaboration limited. However, as partnerships grow andthe acquaintance and understanding of the partners improves,motivation to work to a common goal is enhanced [14] and a largervolume and range of information is needed to sustain thepartnership. Further, the shared understanding of partners isenhanced and further sharing of knowledge results. We proposed:

Hypothesis 2A. The greater is the partnership extent, the greater isthe extent of template-based information sharing.

Hypothesis 2B. The greater is the partnership extent, the greater isthe extent of proactive information sharing.

2.8. Data dynamism and information sharing

Data dynamism is the extent to which information is updatedasynchronously as updates become available. Dynamic dataevolves over time. Data dynamism allows partnerships to leverageupdated information to achieve greater competitiveness andagility. Hence, the efficient sharing of dynamic data and knowledgeis a prerequisite for a successful partnership [15]. Technicalsupport for the exchange of dynamic data involves data migrationand integration between the heterogeneous distributed reposito-ries and databases of partners. From a business perspective, thepartners in a relationship may have to supplement informationsharing by using ad hoc media such as fax or phone. We thereforehypothesized:

Hypothesis 3A. The greater is the data dynamism, the greater isthe extent of template-based information sharing.

Hypothesis 3B. The greater is the data dynamism, the greater isthe extent of proactive data sharing.

2.9. Process complexity and information sharing

Business processes are composed of tasks and sets of conditionsthat determine their order and the needs and direction of the

processes. Their complexity can be described in terms of theirroutineness and dependence on other processes. A non-routineprocess has a high level of variety and difficulty. Processdependence measures the degree of inter-company collaborationthat a process requires; when task variability, interdependence,and complication are low, the work performed tends to be morestructured and predictable and information processing require-ments tend to be minimal but as task variability, uncertainty,interdependence, and complexity increase, problems are moreunstructured, and the assessment of events and specification ofappropriate rules and procedures becomes difficult. Whencomplex and interdependent tasks are chained to form businessprocesses, they will require more coordination among supplychain partners. This leads to two hypotheses.

Hypothesis 4A. The greater is the process complexity, the greateris the extent of template-based information sharing.

Hypothesis 4B. The greater is the process complexity, the greateris the extent of proactive information sharing.

2.10. Partnership extent, data dynamism, and process complexity

When the extent of a partnership increases, business processesbecome more intertwined; this forces the partners to use dynamicdata to reduce business uncertainty. Similarly, strategic partner-ships involve more heterogeneous task environments, whichresults in more non-routine and interdependent tasks to establishmutual goals. We therefore hypothesized:

Hypothesis 5. The greater is the partnership extent, the greater isthe data dynamism.

When the extent of a partnership increases in a dynamicbusiness environment, more coordination is needed betweenpartners. They must develop a collaborative system and commoncodes for products, customers, and data communications; definestandards for systems integration, development, and architecture;and agree on the terms of the partnership and the procedures forsolving conflicts. Such agreements and procedures may result in aset of process-related constructs that help in guiding the complexflow of information between partners and the management of themultifaceted interaction of business processes across organiza-tions. Hence, we hypothesized:

Hypothesis 6. The greater is the partnership extent, the greater isthe process complexity.

When business tasks and processes are complex, informationacquisition for decision making is more continuous, varied, andwide ranging. Therefore, we expected the correlation betweenprocess complexity and data dynamism to be positive.

Hypothesis 7. The greater is the process complexity, the greater isthe data dynamism.

3. Research methodology

3.1. Measurement development

A survey was developed based on a seven-point Likert scalewith anchor points ranging from strongly disagree (1) to stronglyagree (7). The measures used to operationalize the factors wereadapted from relevant studies, when available, with changes inwording to fit the target context.

The measures of our dependent variable, the extent ofinformation sharing (TIS and PIS), were modified from Monczka

Table 1Respondent profiles.

Number of

respondents

Percentage

Role in supplier chain (can have more than one role)

Suppliers 333 81.8%

Buyers 325 79.9%

Logistics providers 292 71.7%

Business nature

Importers/exporters 133 32.7%

Manufacturers 116 28.5%

Retailers 17 4.2%

Wholesalers 50 12%

Others 20 4.8%

No response 71 17.4%

Firm size (employee number)

<10 148 36.3%

10–49 145 35.6%

50–99 20 4.9%

100–499 40 9.8%

>500 35 8.6%

No response 19 4.6%

Respondent position

Top management (chief executives,

owners, directors etc.)

56 13.8%

Senior management (senior managers and

departmental managers)

208 51.0%

Middle management (assistant managers,

officers, etc.)

29 7.2%

Administration and supporting staff

(accountant, secretary, engineers, production

coordinator, logistic planner, etc.)

81 19.8%

Others 33 8.1%

T.C. Du et al. / Information & Management 49 (2012) 89–98 93

et al. In their study, seven information sharing questions weredeveloped to evaluate their impact on the alliances of suppliersand buyers. These questions were modified and used to developour measures.

To assess process complexity, we adapted the measure of thenon-routineness and interdependence of task characteristics fromthe studies of Goodhue and Thompson [8] and Karimi et al. [13]. Ina supply chain, tasks are not independent, but integrated to formbusiness processes. Therefore, the complexity of a business processis the measure of the complexity and interdependence of itsrelated tasks.

The questions on partnership extent were modified from thestudy of Guimaraes et al. [11].

Finally, we developed a set of questions based on the study of[7], which addressed data properties in the adoption of RFIDs. Inour measure, dynamic data was defined as the data that grew orchanged over the lifetime of a physical object.

3.2. Questionnaire translation

The original questionnaire was developed in English. However,as most executives in China are not proficient in English, thequestionnaire was first translated into Chinese and then translatedback into English; the final English version was then comparedwith the original by two Chinese IS professors from the School ofBusiness to check the translation. The feedback from this phase ofinstrument development resulted in significant refinement of thesurvey, which improved its content validity.

3.3. Pretest and pilot test

Before its distribution, the survey was subjected to severalrounds of pretests with four IS professors and ten supply chainexecutives to test its wording, the ease with which the question-naire could be answered, the appropriateness of the questionsequence, and the consistency of meaning. The pretests also helpedto validate the scales for the survey items. A pilot test was thenconducted in December 2006 with 20 part-time e-Businessstudents in full-time employment. The reliability of the instrumentwas assessed using Cronbach’s alpha. The resulting values rangedfrom 0.85 to 0.97, which are acceptable for pilot tests. Factoranalysis was also conducted, and, based on these results, onequestion relating to data properties was dropped. The finalquestionnaire items are shown in Appendix A. The first part ofthe survey included questions relating to the extent of datasharing, partnership extent, data dynamism, and process com-plexity, and the second part of the questionnaire solicited generaldemographic information.

3.4. Data collection

The sample was recruited from the database of an internationalorganization that specializes in supply chain services. Thisorganization also organized the survey distribution and datacollection from its database of its members (3951 members locatedin the Pearl River Delta, the main manufacturing base in China).The members were mainly from four target industries: import/export, manufacturing, retail, and wholesale. The majority weresmall or medium-sized companies. To improve the response rate,the chosen members were invited to fill out and return the surveyin return for a Starbucks gift certificate; also those whosuccessfully completed the survey were given the opportunityto win admission to a famous theme park. The research instrumentmeasured the willingness of buyers, suppliers, and logistics serviceproviders to engage in data sharing with their partners. Thevalidity of the measurement was increased by the fact that the

subjects were currently engaged in business with their partnersand could reflect on existing issues and concerns about datasharing in the business environment.

The survey was conducted over a period of two weeks.Questionnaires were sent by the international organization tothe chosen subjects by either e-mail or fax with a covering letterthat explained the purpose of the study, solicited their participa-tion, and assured the confidentiality of the responses. Therespondents were asked to answer questions about informationsharing with their suppliers, buyers, and logistics providers. Theywere also asked to answer the questions honestly. Participationwas voluntary. A total of 3397 questionnaires were sent either bye-mail or fax to participating members. Completed questionnaireswere collected by e-mail or fax by the organization, and were thenhanded over or e-mailed to the investigators.

4. Results and analyses

4.1. Profile of the respondents

A total of 407 valid questionnaires were collected by e-mail andfax, representing a response rate of about 12%. Non-response biaswas assessed by comparing the early with the late quartile ofrespondents using two-tailed t-tests on scale variables of thecompany characteristics (employee number, total fixed assets, andtotal sales) and all PDP constructs. In addition, using the sameprocedure, we compared the fax responses with those from the e-mail for the same variables. The respondents and non-respondentswere also compared using the same technique. The comparisonsshowed no significant difference, suggesting that non-responsebias was not present in our data.

The majority of our respondents, as shown in Table 1, camefrom import/export companies, manufacturers, retailers, andwholesalers. The companies operated mainly in the Pearl RiverDelta region, which included Hong Kong. Of the respondents, 62%were at the top or senior managerial level.

Table 2Factor loadings of the items in the EFA.

Factor loadings

Partnership extent Data dynamism Process complexity PIS TIS

PIS1 0.136 0.096 0.048 0.709 �0.241

PIS2 0.194 0.205 0.046 0.592 0.296

PIS3 0.140 0.265 �0.004 0.715 0.223

PIS4 0.123 �0.043 0.231 0.725 �0.007

TIS1 0.169 0.263 �0.020 0.401 0.579TIS2 0.240 0.030 0.109 �0.124 0.711TIS3 0.357 0.163 0.157 0.203 0.592EP1 0.653 �0.020 �0.021 0.274 0.304

EP2 0.694 0.167 0.061 0.194 0.224

EP3 0.707 0.088 0.142 0.084 0.084

EP4 0.694 0.206 0.109 0.116 0.135

EP5 0.700 0.199 0.078 0.060 0.178

EP6 0.738 0.240 �0.037 0.160 �0.124

EP7 0.515 0.077 �0.012 0.053 0.009

EP8 0.508 0.289 0.196 �0.093 0.140

EP9 0.678 0.360 0.066 0.003 0.295

EP10 0.686 0.165 0.032 0.132 0.039

EP11 (discarded) 0.289 0.191 0.239 0.135 �0.248

PR1 0.211 0.220 0.703 0.000 0.063

PR2 0.017 0.146 0.773 0.057 �0.005

PR3 �0.015 0.035 0.789 0.167 0.006

PR4 0.048 0.264 0.704 0.043 0.174

DA1 0.301 0.581 0.128 0.250 0.039

DA2 0.155 0.712 0.198 0.168 0.070

DA3 0.236 0.731 0.079 0.061 0.051

DD1 0.229 0.745 0.033 0.170 0.207

DD2 0.214 0.691 0.265 �0.021 0.040

DD3 0.149 0.612 0.419 0.056 �0.034

Eigenvalue 5.104 3.605 2.756 2.477 1.861

Cronbach’s alpha 0.885 0.849 0.789 0.720 0.669

Total variance explained 56%

The Cronbach’s alpha values higher than 0.5 are in bold font.

T.C. Du et al. / Information & Management 49 (2012) 89–9894

4.2. Measurement model analysis

Confirmatory factor analysis was conducted to test themeasurement model. The overall model fit was assessed usingsix goodness-of-fit indices: the chi-square/degree of freedom (x2/df), normalized fit index (NFI), non-normalized fit index (NNFI),comparative fit index (CFI), root mean square residual (RMR), androot mean square error of approximation (RMSEA). The results forthese indices were x2/df = 2.05 (�3.00), RMSEA = 0.062 (�0.08),NNFI = 0.96 (�0.95), CFI = 0.96 (�0.95), and standardizedRMR = 0.061 (�0.08). All of the indices exceeded the minimumrecommended values, which suggests that the measurementmodel was an adequate fit.

The measurement model was further assessed for constructreliability and validity. Construct reliability was first evaluatedusing Cronbach’s alpha reliability test. The Cronbach’s alpha valuesof all of the variables were all above 0.65, which is close to or abovethe 0.7 threshold level. The convergent validity of the instrumentwas assessed by examining the factor loadings. Exploratory factoranalysis (EFA) was applied to ensure the unidimensionality of thescales, and principal component analysis and varimax rotation inSPSS were chosen to identify the factors. The number of factors wasnot specified in advance. Multi-item factors with eigenvaluesgreater than 1.0 were required to be listed in the output, which isshown in Table 2. Six factors were extracted, but a drop ineigenvalue between the fifth and sixth factors was noticed andthus only five were retained as constructs in the model. All of theitems loaded well, except for EP11, which displayed a small loadingon the intended construct Partnership Extent. The item wastherefore removed from later analysis.

A series of constrained models was also developed forcomparison with the original CFA model – termed the uncon-strained model – to assess discriminant validity [29]. In each

constrained model, the correlation between one pair of constructswas constrained to 1. This resulted in changes to the chi-squarevalues, which were then used as the statistics to examine whetherthe constraint had caused a significant change. If the difference inthe chi-square between a constrained and the unconstrainedmodel was significant, then constraining the two constructs to 1resulted in significantly different results from those obtained whenthe constructs were freely estimated in the unconstrained model.This in turn indicated that the correlation between the twoconstructs was significantly different from 1. If all of thecorrelations between the constructs were significantly differentfrom 1, then the discriminant validity of the instrument would besupported. Table 3 shows that all of the differences between theoriginal model and the constrained ones were significant at the0.05 level, and thus the discriminant validity of the instrument wasestablished.

Although common method variance (CMV) is not as serious in ISas in other disciplines, it is a noteworthy issue for studiesconducted in a Chinese context because of Chinese cultural factors[28]. Hence, we conducted two analyses to detect whether CMVwas present in our study. First, we conducted the commonly usedHarmon’s single factor test, in which all items are included in anexploratory factor analysis (EFA). CMV is present if a single factoremerges from EFA or the first factor accounts for most of thecovariance. Second, we checked the correlation matrix. CMV isunlikely if there are no extremely high correlations of 0.90 or above[19]. Our analysis showed that neither of these two conditionsheld. Thus CMV was not a major problem in our study.

4.3. Structural model analysis

LISREL 8.5 and the maximum likelihood estimation (MLE)method were used to test the main effect of the research model.

Table 3Discriminant analysis.

Chi-square df Difference in chi-square Difference in df P-value

Original model without constraints 642.5 314

Constrained model 1 (for PIS and TIS) 661.4 315 18.9 1 0.0

Constrained model 2 (for PIS and EP) 680.0 315 37.5 1 0.0

Constrained model 3 (for PIS and PC) 684.8 315 42.3 1 0.0

Constrained model 4 (for PIS and DD) 671.9 315 29.4 1 0.0

Constrained model 5 (for TIS and EP) 656.3 315 13.8 1 0.0

Constrained model 6 (for TIS and PC) 671.7 315 29.1 1 0.0

Constrained model 7 (for TIS and DD) 660.2 315 17.7 1 0.0

Constrained model 8 (for EP and PC) 685.1 315 42.6 1 0.0

Constrained model 9 (for EP and DD) 662.4 315 19.9 1 0.0

Constrained model 10 (for PC and DD) 656.6 315 14.1 1 0.0

T.C. Du et al. / Information & Management 49 (2012) 89–98 95

The model fit indices of x2/df = 2.05, RMSEA = 0.06 (�0.08),NNFI = 0.96 (�0.95), CFI = 0.96 (�0.95), and standardizedRMR = 0.06 (�0.08) clearly exceeded the minimum recommendedvalues for a good model fit, suggesting the adequacy of ourresearch models for further analysis, including causal linkevaluation. The overall explanatory power of the research modelwas examined using the R2 for each dependent construct. Themodel had an R2 of 56.4%, and thus explained 56.4% of the variancein willingness to share information. Fig. 2 shows the pathcoefficients of the model, along with their respective significancelevels. One path (data dynamism ! PIS) was found to besignificant at the 0.05 level, and five paths (partnership exten-t ! process complexity, partnership extent ! TIS, partnershipextent ! data dynamism, process complexity ! data dynamism,and TIS ! PIS) were found to be significant at the 0.01 level.

These results showed that Hypothesis 1 was fully supported.Regarding Hypothesis 2, partnership extent showed a strongpower to explain TIS at the 0.01 level (2A), but had no significantdirect impact on PIS (2B). Hypothesis 3B was supported.Hypotheses 5–7 were all supported. The only hypotheses thatwere found to be insignificant were 2B, 3A, 4A, and 4B.

5. Discussion

Our results confirmed the presence of both template-based(TIS) and proactive information (PIS) sharing – among supply chainmembers including buyers, suppliers, and logistics service

Data Dynamism

Partnership Extent

Process Complexity

0.40(t=5.65)**

0.13 (t=1.16)

0.61 (t=5

-0.00 (t

0.08 (t=0.96)

0.24 (t=2.07)*

-0.01 (t=-0.

0.30

(t=4.03)**

0.51 (t=6.9)**

Fig. 2. Results of the

providers. The correlation between TIS and PIS was also validated;this implied that companies with a greater extent of TIS were morelikely to have a greater extent of PIS.

As anticipated, partnership extent was a determinant of notonly TIS but also process complexity and data dynamism. However,it did not have an impact on PIS. Partnership extent may influencePIS only if the technical and political issues associated with PIS areresolved.

Data dynamism, although found to be critical to PIS, did nothave a significant impact on TIS. This implied that when data wasmore dynamic, the need for PI, but not TIS became greater. In mostcircumstances, TIS focuses on the exchange of structured andpersistent information, which allows supply chain partners toconfigure a common data model to facilitate information sharing.However, as data become more dynamic, a prescribed TIS datamodel becomes inadequate to address the proactive informationneeds of supply chain fulfillment, forecasting, and management.For example, a prescribed data model may not be able to updatesupply chain partners on asynchronously updated information asfurther updates to the information become available. In addition, itmay be unable to provide rapid, flexible, collaborative informationsharing as data become more dynamic.

Surprisingly, process complexity was not found to have a directimpact on either TIS or PIS. However, the effect of processcomplexity on PIS was found to be mediated by data dynamism, asdemonstrated by significant links between process complexity,data dynamism, and PIS. From the logistics service providers’ point

Proactive Information Sharing

(PIS)

Template-based Information Sharing

(TIS)

0.50 (t=3.27)**

.66)**

=-0.05)

07)

*: at the 0.05 level.

**: at the 0.01 level.

model testing.

Table 4Summary of company type comparison.

Full model Company type Pairwise comparison

Buyers Suppliers Logistics service providers

(H1) TIS ! PIS 0.50 (t = 3.27)** 0.56 (t = 3.42)** 0.39 (t = 3.11)** 0.54 (t = 3.84)** No

(H2a) partnership extent ! TIS 0.61 (t = 5.66)** 0.57 (t = 5.79)** 0.57 (t = 5.33)** 0.55 (t = 5.74)** No

(H2b) partnership extent ! PIS �0.01 (t = �0.07) �0.07 (t = �0.57) 0.14 (t = 1.25) 0.00 (t = �0.04) No

(H3a) Data dynamism ! TIS 0.13 (t = 1.16) 0.17 (t = 1.73) 0.07 (t = 0.61) 0.17 (t = 1.64) No

(H3b) Data dynamism ! PIS 0.24 (t = 2.07)* 0.30 (t = 2.83)** 0.15 (t = 1.40) 0.24 (t = 2.10)* Yesa,b

(H4a) Process complexity ! TIS 0.08 (t = 0.96) 0.14 (t = 1.66) 0.09 (t = 1.01) 0.08 (t = 1.00) No

(H4b) Process complexity ! PIS �0.00 (t = �0.05) �0.06 (t = �0.71) 0.06 (t = �0.71) �0.02 (t = �0.28) No

(H5) Partnership extent ! data dynamism 0.51 (t = 6.90)** 0.46 (t = 6.58)** 0.46 (t = 6.57)** 0.49 (t = �7.10)** No

(H6) Partnership extent ! process complexity 0.30 (t = 4.03)** 0.28 (t = 4.02)** 0.35 (t = 4.99)** 0.41 (t = 5.95)** No

(H7) Process complexity ! data dynamism 0.40 (t = 5.65)** 0.41 (t = 6.62)** 0.44 (t = 6.37)** 0.36 (t = 5.06)** No

*Significant at the 0.05 level.**Significant at the 0.01 level.aDifference between suppliers (0.13, t = 1.16) and buyers (0.26, t = 1.16) (chi-square is 1215.8 (df = 682), RMSEA = 0.05, NNFI = 0.97, CFI = 0.97, standardized RMR = 0.06).bDifference between suppliers (0.14, t = 1.20) and logistics service providers (0.24, t = 2.08) (chi-square is 1233.5 (df = 682), RMSEA = 0.06, NNFI = 0.97, CFI = 0.97, standardized

RMR = 0.07).

T.C. Du et al. / Information & Management 49 (2012) 89–9896

of view, information sharing with buyers or suppliers is arequirement of their business. Traditional freight forwardingbusiness requires information to be shared in the form ofinstructions and status reports which are usually detailed in theoriginal quotation request. As described by the executive of anInternational Integrator in our pretest panel, his company wouldvery much like to move from a freight forwarding business tothat of a high value-added supply chain management business.While the trend is moving in that direction, the pace is far fromsatisfactory. The difficulty is often in the prescription of whatneeds to be shared in response to an RFP. TIS and complexprocess are just prerequisites for SCM business which reallyrequires PIS in the form of changing SC strategy, new businessmodel and corresponding innovative logistics solutions.

5.1. Post hoc analysis

Information sharing among various supply chain partners wasinvestigated in greater depth by conducting a comparison of thesupply chain companies sampled. We applied multi-groupstructural equation modeling (SEM) to see whether there wereany distinctions in the extent of willingness to share informationwith different partners. Cross comparison was conducted of thecompanies to determine whether there were differences in theevaluation of the research variables. In each comparison, thecompanies were analyzed in pairs to determine whether thesecompanies perceived the hypothesized information-sharing rela-tionships to be similar or different.

The results, which are presented in Table 4, suggested that thedifferent types of supply chain companies perceived the PDP, TIS,and PIS relationships similarly (except for the perception of therelationship between data dynamism and PIS).

The only difference in information-sharing perception wasfound in the link between data dynamism and PIS. Suppliers andbuyers, along with suppliers and logistics service providers, werefound to have very different perceptions of the way data dynamismcould impact PIS. As shown in Table 3, data dynamics was found tohave a significant impact on PIS for buyers and logistics serviceproviders. For suppliers, however, data dynamism extent was notfound to have any impact on PIS extent. Most buyers in our surveywere from the US or Europe and this could explain inconsistentfindings. They have higher purchasing power and real-timeinformation sharing demand, which drives their need for a greaterextent of PIS for dynamic data such as shipment and payment. Incontrast, most suppliers were local companies with less negotiat-ing power, and thus a reduced need for real-time or updatedinformation.

6. Conclusion and implications

Many studies of supply chain management have addressedinformation sharing by focusing on the issues and benefits ofsharing. It is probable that most companies in a supply chain arewilling to share information to enhance the conduct of theirbusiness. However, their collaboration may be constrained byseveral determinants that directly or indirectly affect the needs,tradeoffs, and consequences of their business operations. Weproposed a partnership-data-process (PDP) model to investigatewillingness to share information in the supply chain context. Thiswas found to be useful in explaining the variation in the extent ofwillingness to share information, as indicated by the significance ofthe proposed determinants and relationships in the model.

A number of implications for practitioners can be drawn fromour results. First, supply chain managers should remember thatinformation sharing may TIS or PIS. At a fundamental level, TIS isaffected directly only by partnership extent. Only when supplychain companies have TIS agreements can they advance to a higherlevel of sharing. This suggests that a successful strategicpartnership is needed before a strong will to share informationarises. Hence, supply chain managers need to devise strategies toenhance commitment, trust, coordination, and interdependencewith their partners.

Supply chain managers must also recognize that a higher levelof partnership is a necessary but not sufficient condition for PIS.Thus, to achieve more extensive PIS, managers need to establish asolid foundation through formal contracts in which the rules aboutinformation exchange are well specified and provision is made forthe resolution of any failure to support information exchangethrough specified clauses in the contract. They also need to provideevidence of the timeliness and dynamism of the required data tojustify PIS, provided that a strategic partnership has already beenestablished.

Supply chain managers should also investigate the reason forinsignificance of data dynamism and process complexity ininformation sharing – is it managerial, architectural, or operation-al? Although these are complex issues, supply chain collaboratorsneed to establish a consensus if TIS and PIS are to be furthersupported.

Our study had, of course, some limitations. First, it wasconducted in China using Chinese supply chain companies for ourpopulation, thus we are not able to generalize the findings to othergeographical regions having different cultures. Second, our studyfocused on the extent of willingness to share information. Thisissue is somewhat technical, as it concerns the security, control,and standards of information sharing. In a supply chain,

T.C. Du et al. / Information & Management 49 (2012) 89–98 97

information sharing must consider access control issues and policiesto reduce risks, especially for activities and resources that involvebusiness processes within and across supply chain participants.Information sharing also requires a way to store and presentinformation and knowledge in a standardized format; these issueshave been explored previously (e.g., [6]), but need furtherinvestigation. Although we evaluated willingness to share informa-tion using a classification of TIS and PIS, we did not adopt a moresophisticated conceptualization or classification of partnership,data, and process in the PDP framework. For example, partnershipcould have been classified by type, which could affect information

Appendix A. Questionnaire items

Proactive information sharing (PIS1) When needed, we shall share our

(PIS2) When our business information re

business information with us

(PIS3) When needed, we shall provide an

(PIS4) When needed, our partners will sh

Template-based information sharing (TIS1) Our partners and our company ke

(TIS2) Our partners and our company on

(TIS3) Our partners and our company se

Process complexity (PC1) We frequently deal with ad hoc, n

(PC2) We frequently deal with ill-defined

(PC3) We frequently deal with business

(PC4) We frequently deal with business

Data properties (DD1) We frequently update our partner

(DD2) We frequently deal with constant

(DD3) We frequently deal with changing

Partnership extent (EP1) We have a high degree of confiden

(EP2) We have a high degree of understa

(EP3) We have a high degree of agreeme

(EP4) We have a high degree of compati

(EP5) We have a mutual commitment w

(EP6) We have a high degree of similar v

(EP7) Our company and our partners hav

(EP8) We have a high degree of willingn

(EP9) We communicate with a high degr

(EP10) We have a high degree of smooth

(EP11) We and our partners are able to i

sharing differently in various social situations. Our study did not takea longitudinal approach when evaluating willingness to shareinformation. Although our PDP model explained some of thevariation in willingness to share, it says little about how or whetherwillingness to share will evolve over time.

Acknowledgment

This study is supported in part by the Li & Fung Institute ofSupply Chain Management and Logistics, CUHK.

company’s internal business information with our partners

quirements changed, our partners will still share the required

y business information that might help our partners

are their internal business information with us

ep each other informed of data changes, but limited to those agreed for sharing

ly provide information according to pre-specified agreements.

ldom exchange data other than those pre-determined

on-routine business processes with our partners

business processes with our partners

processes that involve more than one business function

processes that involve more than one partner.

s with new business information

ly changing business information with our partners

business information at data item level with our partners

ce in our partners

nding with our partners about protecting exchanged business information

nt with our partners on matters of benefit and risk.

bility in business activities with our partners

ith our partners to continue the partnership

alues to our partners

e a high degree of freedom not to be restricted by organizational boundaries

ess to cooperate in business activities with our partners

ee of timeliness with our partners

ly coordinated business activity with our partners

nfluence each other’s business decisions

References

[1] R. Agarwal, V. Sambamurthy, Principles and models for organizing the IT function,MIS Quarterly Executive 1 (1), 2002, pp. 1–16.

[2] Y. Aviv, Gaining benefits from joint forecasting and replenishment processes: thecase of auto-correlated demand, Manufacturing & Service Operations Management4 (1), 2002, pp. 55–74.

[3] M. Barratt, Understanding the meaning of collaboration in the supply chain, SupplyChain Management 9 (1), 2004, pp. 30–42.

[4] B. Chae, H.R. Yen, C. Sheu, Information technology and supply chain collaboration:moderating effects of existing relationships between partners, IEEE Transactionson Engineering Management 52 (4), 2005, pp. 440–448.

[5] R. Croson, K. Donohue, Behavioral causes of the bullwhip effect and the observedvalue of inventory information, Management Science 52 (3), 2006, pp. 323–337.

[6] F. D’Aubeterre, R. Singh, L. Iyer, Secure activity resource coordination: empiricalevidence of enhanced security awareness in designing secure business processes,European Journal of Information Systems 17 (5), 2008, pp. 528–542.

[7] EPCglobal ‘‘EPCglobal Network Final Architecture,’’, 2005.[8] D.L. Goodhue, R.L. Thompson, Task-technology fit and individual-performance, MIS

Quarterly 19 (2), 1995, pp. 213–236.[9] S. Gosain, A. Malhotra, O.A.E. Sawy, Coordinating for flexibility in e-business supply

chains, Journal of Management Information Systems 21 (3), 2004–2005, pp. 7–45.[10] V. Grover, J. Teng, K. Fiedler, Investigating the role of information technology in

building buyer–supplier relationships, Journal of the Association of InformationSystems 3 (2), 2002, pp. 217–245.

[11] T. Guimaraes, D. Cook, N. Natarajan, Exploring the importance of business clock-speed as a moderator for determinants of supplier network performance, Deci-sion Sciences 33 (4), 2002, pp. 629–644.

[12] A. Gunasekaran, E.W.T. Ngai, Information systems in supply chain integration andmanagement, European Journal of Operational Research 159 (2), 2004, pp. 269–295.

[13] J. Karimi, T.M. Somers, Y.P. Gupta, Impact of environmental uncertainty and taskcharacteristics on user satisfaction with data, Information Systems Research 15(2), 2004, pp. 175–193.

[14] D.-G. Ko, L.J. Kirsch, W.R. King, Antecedents of knowledge transfer from con-sultants to clients in enterprise systems implementations, MIS Quarterly 29 (1),2005, pp. 59–85.

[15] C.K.M. Lee, H.C.W. Lau, K.M. Yu, R.Y.K. Fung, Development of a dynamic dataexchange scheme to support product design in agile manufacturing, InternationalJournal of Production Economics 87 (3), 2004, pp. 295–308.

[16] H. Lee, B. Choi, Knowledge management enablers, processes, and organizationalperformance: an integrative view and empirical examination, Journal of Man-agement Information Systems 20 (1), 2003, pp. 179–228.

[17] F.R. Lin, S.H. Huang, S.C. Lin, Effects of information sharing on supply chainperformance in electronic commerce, IEEE Transactions on Engineering Manage-ment 49 (3), 2002, pp. 258–268.

[18] R. Patnayakuni, A. Rai, N. Seth, Relational antecedents of information flowintegration for supply chain coordination, Journal of Management InformationSystems 23 (1), 2006, pp. 13–49.

[19] P.A. Pavlou, H. Liang, Y. Xue, Understanding and mitigating uncertainty in onlineexchange relationships: a principal-agent perspective, MIS Quarterly 31 (1), 2007,pp. 105–136.

[20] A. Rai, R. Patnayakuni, N. Patnayakuni, Firm performance impacts of digitallyenabled supply chain integration capabilities, MIS Quarterly 30 (2), 2006, pp.225–246.

[21] R. Sabath, J. Fontanella, The unfulfilled promise of supply chain collaboration,Supply Chain Management Review 6 (4), 2002, pp. 24–29.

[22] V. Sambamurthy, A. Bharadwaj, V. Grover, Shaping agility through digital options:reconceptualizing the role of information technology in contemporary firms, MISQuarterly 27 (2), 2003, pp. 237–263.

[23] J.-Y. Son, S. Narasimhan, F.J. Riggins, Effects of relational factors and channelclimate on EDI usage in the customer–supplier relationship, Journal of Manage-ment Information Systems 22 (1), 2005, pp. 321–353.

[24] M. Subramani, How do suppliers benefit from information technology use insupply chain relationships? MIS Quarterly 28 (1), 2004, pp. 45–73.

[25] M.M. Wasko, S. Faraj, Why should I share? Examining social capital and knowl-edge contribution in electronic communities of practice MIS Quarterly 29 (1),2005, pp. 35–57.

T.C. Du et al. / Information & Management 49 (2012) 89–9898

[26] A.E.D.M. White, M. Mohdzain, The role of emergent information technologies andsystems in enabling supply chain agility, International Journal of InformationManagement 25 (5), 2005, pp. 396–410.

[27] F. Wu, S. Yeniyurt, D. Kim, S.T. Cavusgil, The impact of information technology onsupply chain capabilities and firm performance: a resource-based view, IndustrialMarketing Management 35 (4), 2006, pp. 493–504.

[28] X. Zhao, B.B. Flynn, A.V. Roth, Decision sciences research in china: a critical reviewand research agenda – foundations and overview, Decision Sciences 37 (4), 2006, pp.451–496.

[29] X. Zhao, B. Huo, B.B. Flynn, J.H.Y. Yeung, The impact of power and relationshipcommitment on the integration between manufacturers and customers in asupply chain, Journal of Operations Management 26 (3), 2008, pp. 368–388.

Timon C. Du obtained his MS and PhD degrees in

Industrial Engineering from the Arizona State University,

USA. Dr. Du is a Professor at the Chinese University of Hong

Kong. His research interests include e-business, IS strategy,

collaborative commerce, and the semantic web. He has

published papers in many leading journals, including

Information and Management, Decision Support Systems,

IEEE Transactions on Knowledge and Data Engineering,

Communications of the ACM, and IIE Transactions.

Vincent S. Lai is a professor of MIS at the Chinese

University of Hong Kong. His current research focuses on

ERP assimilation, online auctions, virtual collaboration,

and global IS strategy. Dr. Lai has published extensively in

IS journals, including Journal of MIS, Information and

Management, Communications of the ACM, Decision

Support Systems, European Journal of Information

Systems, European Journal of Operational Research, and

IEEE Transactions on Engineering Management.

Waiman Cheung, Director of Li & Fung Institute of

Supply Chain Management & Logistics and Director of

Center of Cyber Logistics, is also the Chairman of the

Department of Decision Sciences and Managerial

Economics in the Faculty of Business Administration,

The Chinese University of Hong Kong. Dr. Cheung is

keen on working closely with local industries for

knowledge/technology transfer and has conducted

studies and consulting works for DHL, Airport Authority

Hong Kong, Dragonair, Accenture, SML Group Ltd. and

the Innovation and Technology Commission. His

research interests are mainly in applying IT on logistics

and supply chain management. Dr. Cheung has

contributed articles to ACM Transactions on Informa-

tion Systems, Decision Sciences, IEEE Transactions on

Systems, Man and Cybernetics, Annals of Operations

Research, European Journal of Operational Research,

Decision Support Systems, and Information & Manage-

ment, etc.

Xiling Cui is a lecturer in Business Administration at

the Hong Kong Shue Yan University. Her research

interest focuses on electronic commerce, online

auction, knowledge sharing, and IT investment analy-

sis. She has paper published in Electronic Markets,

International Journal of Electronic Business, and some

proceedings of conferences, such as Pacific Asia

Conference of Information Systems 2007, and 9th

International DSI Conference, 11th Asia Pacific DSI

Conference, etc.