Decision Support Systems

Information quality, trust, and risk perceptions in electronic data exchanges

Andreas I. Nicolaou

_{, Mohammed Ibrahim}

⁎

_{, Eric van Heck}

a

Department of Accounting & IS, Bowling Green State University, OH 43403, USA b

Tilburg University, Department of Information Management, Faculty of Economics and Business Administration, Tilburg, PO Box 90153-5000 LE Tilburg, The Netherlands c_{Erasmus University, Rotterdam School of Management, Department of Decision and Information Sciences (T9-01), Burg. Oudlaan 50, 3062 PA Rotterdam, The Netherlands}

a b s t r a c t

a r t i c l e i n f o

Article history: Received 11 July 2011

Received in revised form 14 May 2012 Accepted 7 October 2012

Available online 23 October 2012 Keywords: Information quality Exchange-risk Performance-risk Competence-trust Goodwill-trust

Electronic transaction performance Intent to use

This study investigates the influence of information quality, trust and risk perceptions on the expected transaction performance of inter-organizational data exchanges and on the user intent to continue using the exchange. This study provides empirical evidence on the distinctive influences of information quality on competence-trust, goodwill-trust, exchange-risk and relationship-risk and how these different dimen-sions influence the intent to use inter-organizational data exchanges. As the performance of a data exchange may vary according to degree of successful completion of a specific transaction on the spot, this study also examines the extent to which expected transaction performance affects the model relationships. A survey is conducted to collect data from 221 business professionals. The study's hypothesized relationships are in general supported by the data and the resulting structural model proved to adequately represent the construct relationships. While thesefindings contribute to information system design theory, they also benefit professionals by providing insights as to how organizations can deal with the different types of uncertainties related to participating in electronic data exchanges. In addition, thesefindings help demon-strate the importance of interventions in the design of electronic data exchanges and the benefits expected by enhancing information quality in those settings.

© 2012 Elsevier B.V. All rights reserved.

1. Introduction

Modern inter-organizational (I-O) partnerships employ electronic data exchange systems to facilitate collaboration and supply-chain per-formance. The various types of business relationships require different types of electronic data exchanges ranging from standard web-based ordering systems to proprietary systems. The benefits of these systems can range from efficient automated processes in ordering systems and cost-effective transactions, to strategic benefits of collaborative plan-ning[44,52].

Electronic data exchanges present risks because they require sharing information with the partner[30]and the exchange of information may be subject to a fear of opportunism by the other[5,73]. Data exchanges also require one to cooperate and to trust that the partner will do like-wise[8,34]. For the customerfirm to use the vendor's exchange system assumes they can rely on both the system and its vendor. The exchange system should thus possess desirable features that ensure transaction performance and reinforce use continuance behavior.

Lack of control over an exchange partner increases uncertainty and the need to devise strategies to manage risk. Past research has argued

that the evolution from traditional electronic data interchanges (EDI) to open electronic commerce I-O exchanges demonstrates the need for real time controls [35, p. 47–49]. Research in risk management (e.g.,[37,68,69]) also proposes a risk model in electronic I-O exchanges which addresses issues of risk at the technical/system, business process, and application/user levels. The present study examines the design of electronic inter-organizational (I-O) systems in an exchange context and the information sharing that occurs in the use of these exchanges. Past studies examined the extent to which integrated information sharing in an I-O exchange enables data standardization, and provides the link necessary to integrate the technical system, business process, and application/user architectural layers over an organization's electron-ic supply chain[26,28,31]. While the extent of integration in information sharing is an important determinant of supply chain performance (e.g.,[57]), concerns over the quality of information exchanged can potentially influence the perceived trustworthiness attributed to the exchange partner, the assessed effectiveness of the data exchange to effectively carry out transactions, and a user's overall intent to continue using an electronic data exchange.

A number of significant research studies have been conducted on what makes data exchanges successful (e.g.,[30,34,52,72]). Several studies have been done from the economics viewpoint,finding that electronic data exchanges provide lower coordination costs, higher exchange quality, reduced inventory costs, and enhanced strategic and operational benefits[5,45,52]. Other studies have examined the broader phenomenon of inter-organizational relationships _— IORs ⁎ Corresponding author. Tel.: +31 13 4662188; fax: +31 13 4663069.

E-mail addresses:[email protected](A.I. Nicolaou),[email protected]

(M. Ibrahim),[email protected](E. van Heck). 1

Tel.: +1 419 372 2932. 2_{Tel.: +31 10 4082032.}

0167-9236/$–see front matter © 2012 Elsevier B.V. All rights reserved.

http://dx.doi.org/10.1016/j.dss.2012.10.024

Contents lists available atSciVerse ScienceDirect

Decision Support Systems

(e.g.,[29,34]). Data exchange relationships form a subset of IORs and their benefits seem to be mainly due to enhanced information sharing in the relationship[2,40,57]. Gulati and Gargiulo [29]suggest that positive cues are needed initially both to overcome “information hurdles”and to help strengthen the exchange relationship. Positive cues like special site features help exchange users feel right about the exchange even when uncertainty about it is high.

Information quality (IQ) is one such cue. IQ means beliefs about the favorability of the characteristics of the exchange information and reflects the user evaluation of information sharing in data exchanges [53,67]. Previous research has examined how information quality of an exchange influences trust and risk perceptions and how such per-ceptions influence the intent to adopt a specific exchange. Trust and risk have been emphasized to play decisive roles in economic ex-changes[16]and to affect the success of inter-organizational exchanges [30]. The complexity of social and economic exchanges has led to the distinction between various types of trust and risk[15,60]. This paper aims at exploring whether information quality has distinctive infl u-ences on two types of trust (competence-trust and goodwill-trust) and two types of risk (exchange-risk and relationship-risk) and how these different types of trust and risk influence the intent for continued use of IOS exchanges. As the performance of a data exchange relies on the successful completion of a specific transaction on the spot, the paper also examines the extent to which transaction performance affects the model relationships.

We believe answering these questions has the potential to benefit data exchange research and practice in significant ways. First, ex-changes are very prevalent in professional practice, and what makes them successful is important. Second, studying how information quality exerts its effects on outcome variables by analyzing its distinc-tive influences on competence-trust and goodwill-trust can influence successful information system design. Previous research has showed that interorganizational relationships can be characterized with spe-cific types of trust [32]. This study contributes to enhancing our knowledge about the precise influences of information quality on the specific types of trust helps to (1) determine the characteristics of information systems that are needed for successful performance in different contexts of interorganizational relationships, (2) establish how information quality can be employed to deal with performance-risk and exchange-performance-risk, and (3) ascertain whether information quality can be utilized to attain the required balance between trust and risk for enhanced transaction performance.

2. Theoretical background, rationale, and research hypotheses

2.1. Theoretical background

Organizations value information quality as they are confronted with increasing uncertainty, market volatility and dynamic customer demands. Decision makers value timely information on market devel-opments, reliable information on customer preferences and accurate information on the latest trends. High information quality gives the system user confidence in the exchange vendor because having quality information suggests that exchange information is reliable, correct, adequate, complete, responsive, and timely[27]. Within an expectation-disconfirmation framework, McKinney et al.[48]use IQ as a construct to predict Internet consumer satisfaction, while DeLone and McLean[19]use information quality to predict user satisfaction and system use. The quality of information sharing, nevertheless, varies widely across different data exchanges. Professional exchanges, includingwww.covisint.comandwww.recycle.net, vary in the degree to which they make available to the user important exchange infor-mation or offer transaction inforinfor-mation.

The effects of information quality have been previously examined within the theoretical contexts of system adoption and system success [10]. Applying these insights, this paper highlights the role of IQ and

relationships among exchange partners in the adoption and use of elec-tronic data exchanges. Past research (e.g.,[29]) suggests that initial partners rely on cues (e.g., business credentials) to develop con_fidence in each other. It is likely that the availability of information cues in the exchange serves as a key determinant of the relationship between IQ and exchange outcomes.

Past evidence on initial use of data exchanges establishes the impor-tance of factors that impact the climate of the relationship (i.e., senti-ments of bonding and trust), which Bensaou[6]found to be the most robust predictor of buyer–supplier cooperation. Also, Bensaou and Venkatraman[7]integrate across three theoretical lenses–transaction cost economics, organization theory, and political economy–to model interorganizational relationship (IOR) formation. They show how all three theory bases address exchange uncertainty via joint action, infor-mation sharing, and coordination. Economic theories like transaction cost economics and agency theory suggest that online data exchange use may be fraught with concerns[64]because the vendor–supplier party has the ability to hide information from the customer-user. In the online environment, the ability to appear to be what you are not in-creases the chances for opportunistic behavior[18]. Further, as Singh and Sirdeshmukh[64]explain, the vendor may not deliver goods in a satisfactory manner, resulting in a moral hazard situation the consumer cannot detect. Because each party is dependent on the other, they need and want to cooperate, but suspicion often lurks in their minds due to the possibility of moral hazard or opportunism. Thus, economic ex-change theory suggests that the exex-change must either be supported by control structures such as formal rules, procedures, and policies to monitor and reward desirable performance[16]or informal controls that enforce obligations, promises, and expectations through trust rela-tions[21,25]. While trust has been studied widely in B2C e-commerce [9,23,56], few (e.g.,[30]) have studied trust in settings like electronic data exchanges. In related work, Bensaou and Venkatraman[7]develop a conceptual model for interorganizational partners that include trust as a factor of partnership uncertainty.

In our study, we examine empirical relationships in a spot ex-change (business-to-business) context, and thus consider informal forms of control and cooperation in such business interactions. In their role as boundary spanners, managers need to supervise various horizontal relations with other organizations, including suppliers, competitors, and other entities in their supply chain that can infl u-ence their own organizations. These managers are in fact active in complex networks of social interactions[11]where effective interac-tions are keys for achieving the increased performance and realizing value. An inter-organizational exchange involves partners engaged in high uncertainty and risk[42,70], and such uncertainty and risk are typically high because the behavior of the partner can neither be guaranteed nor monitored[23,58]. Hart and Saunders[30]find trust and risk are both crucial to electronic data exchanges. As a result, an economic exchange perspective on informal exchange processes can theoretically explain the relationships that form during the interac-tion of a user with a data exchange system.

Given the level of uncertainty and volatility that is inherent in business exchanges, trust and risk are pervasive phenomena in such interactions. Trust enables people and organizations to interact with-out fear of getting exploited or taken advantage of[20]. Such trust in the business partner can be founded on competence and demon-strated accomplishments or the consideration of interests and good-will [54]. Although different types of trust are distinguished, the matter of how these types influence the use of data exchange systems has received limited theoretical attention. Perceived risk, on the other hand, appears to be an integral part of people's cognitive processes when dealing with risky situations. As organizations are compelled to creating alliances with numerous parties, including competitors, managers are expected to supervise these alliances even though they are confronted with various risks including opportunistic be-havior, hidden agendas and market volatility. The academic literature

in this field differentiates and measures specific risk perceptions attributing them to competence and relational intentions of the business partner[17]. However, little is known on the distinctive types of perceived risks on information sharing and use in electronic data exchange contexts. The research model that follows attempts to theoretically justify and empirically examine these relationships in a business-to-business data exchange context.

2.2. Research model

Issues of perceived risk and trust are heightened by the electronic commerce environment, especially given the remoteness of buyers and sellers, the lack of face-to-face contact and the complexity of the data exchange technology. While in some e-commerce situations, such issues may be dealt with in developing an ongoing relationship prior to transacting via an exchange, the current study focuses on the spot purchase of products. In this context, because relationships are more transient, the buyer would rely on information provided through the exchange to address issues related to perceptions of risk with regard to the data exchange and trust with regard to the data exchange provider. Trust and perceived risk should thus infl u-ence exchange success, which is defined by the intention to continue using the exchange and the expected transaction performance of the exchange. These concepts were selected because of their importance to the success of an I-O data exchange. Intention to use means the likelihood that a user will continue to employ the exchange in the future, an indicator considered important in the IS success literature [19]. Unless exchange use continues, it cannot be regarded a success. Expected transaction performance means the extent to which the user perceives that the exchange is being completed properly, which also indicates exchange success. This concept is important to an exchange user because it captures expectations that the exchange transaction will progress to the desired conclusion. Drawing from the above theoretical perspectives, we attempt to increase our under-standing on how IQ influences the intent to use and expected transac-tion performance of interorganizatransac-tional exchanges by focusing on the roles of trust and perceived risk.

We argue, nevertheless, that improved information quality has direct influences on competence-trust, goodwill-trust, perceived risks related to the use of the data exchange itself and risks related the performance of the partner. Competence-trust and performance perceived risks are related to the professional accomplishments of the business partner, and hence they are expected to in_fluence expectations regarding transaction performance. These performance expectations are in turn expected to have recursive influences on trust and risk by improving goodwill-trust and diminishing perceived risks related to use of the data exchange. Wefinally expect goodwill-trust, perceived risks related to use of a data exchange and expecta-tion regarding transacexpecta-tion performance to influence the user's intent to continue using the exchange, seeFig. 1.

The strength of this research model is that it enables investigating (1) how IQ influences different types of trust and risk, (2) how managers determine expectations of successful exchange performance, and (3) base their intentions to continue using an information exchange based on trust and risk. To enhance external validity, the conceptual model is tested using a representative sample of 221 business professionals with real-world experience in online data exchange transactions.

2.3. Research hypotheses

In this study trust is determined by trustor perceptions of parti-cular trustee characteristics. Such perceptions can be based on ratio-nal cognition or affect and emotions[47]. This study distinguishes between competence and openness because existing studies in social psychology indicate that cognitive trust‘leaps’and open transparent communication of expectations and activities signi_ficantly in_fluence social interaction[42]. Competence-trust is based on the perceived trustee's abilities, skills and expertise that facilitate performance within a speci_fic domain[32,46,47]. Goodwill-trust is based on the trustee's confidence in its exchange partner's open commitment to supporting and continuing a focal exchange relationship [50,60]. Hence competence-trust and goodwill-trust relate to different di-mensions but are not mutually exclusive.

Improved information quality is expected to have positive infl u-ences on both types of trust. Attainment of accurate information

H3b (-) H4a (+) H4b (-) H5 (+) H1b (+) H2b (-) H3c (+) H1a (+) H3a (+) H2a (-) Not hypothesized Perceived Performance Risk Perceived Exchange Risk Expected Transaction Performance Competence Trust Goodwill Trust Intent for continued Use Information Quality

indicates that the partner is competent in its specialization. The partner's ability to provide accurate and up-to-date information

con-firms its skills, expertise and the fact that it is a source of accurate in-formation to achieve the desired results cooperatively. More particularly, its ability to provide high quality information indicates that it utilizes its expertise to evaluate the focal organization informa-tion needs effectively[49]. Accordingly, the ability to provide correct information is expected to increase perceptions of competence-trust. H1a. Information quality will have a positive influence on competence-trust.

The ability to provide accurate information indicates that the part-ner is committed to a successful transaction. This can provide a good foundation for a‘leap of faith’[12]and positively influences benevo-lence and goodwill-trust as it increases habituation[46,54]. Goodwill-trust focuses on the good intentions expected from the other party. Attainment of complete and correct information will enhance these expectations. This can also be the case when the information indicates a short coming of the partner, such as lack of stock of a particular product. The honesty of the partner can be appreciated in this case as it allows the focal organization to decide whether to adjust plans (order another product) or to wait till the objective can be realized (original product is available). Hence, the high quality in the provided information should enhance coordination and habituation to achieve common goals, which consequently positively influence goodwill-trust. H1b. Information quality will have a positive influence on goodwill-trust.

Risk means the uncertainty involved in achieving desirable out-comes[65]. Perceived risk relates to what could go wrong in an exchange relationship. We distinguish between performance-risk [16]and exchange-risk[29,65]. Performance-risk is related to losses due to the partner not performing as expected. Performance variation can have several reasons including demandfluctuations, intensified rivalry and sheer bad luck [16]. Improved information quality is expected to reduce the uncertainty and risk because the focal organiza-tion can conduct more extensive and elaborate monitoring. Receiving high quality information indicates the partner can measure the perfor-mance of its internal processes and provide transparency. Such trans-parency can enable the focal organization to decide whether partner performance under those particular circumstances is acceptable. Ac-cordingly, improved information quality (IQ) is expected to reduce uncertainties and ambiguities regarding business partner performance and inherently decrease performance-risk.

H2a. Information quality will negatively influence performance-risk. Exchange-risk is specifically related to the use of the electronic exchange. Such risk can be related to threats to existing business models, fear of IT failure[76]and financial losses due to disputes over who is responsible[63]. Acquiring improved IQ reduces uncer-tainties regarding the individual transactions handled through the exchange. The organizations can acquire up-to-date transaction infor-mation that can include order status, manufacturing progress,fi nan-cial payments and delivery status. Such information can significantly decrease uncertainties regarding the electronic exchange and hence decrease the related exchange-risk.

H2b. Information quality will negatively influence exchange-risk. Several empirical studies reviewed in Dirks and Ferrin[20]reported consistent significant effects of trust on attitudinal and cognitive/ perceptual constructs. For example, a main effect of trust on attitudes, perceptions (e.g., perceived accuracy of information exchanged, per-ceived loss), and other cognitive constructs (e.g., satisfaction) has

been consistently identified in prior studies[55,59,75]. A relationship of trust was also important in maintaining a productive relationship that contributes to the degree of con_fidence in interorganizational ex-changes[1]. More particularly, competence-trust can lead to expectations of accurate, efficient and proficient conduct of the partner when such conduct is within the competence domain. The competent partner is expected to perform skillfully in its domain and develop strategies to achieve future goals and realize positive outcome. Competence-trust is founded on the business partner upholding high professional standards and proficiency in its area. The expertise of the partner is expected to produce positive transaction outcome. Furthermore, benefits derived from specialization and concentration on core competencies, provide additional rationalization for expecting positive transaction outcome. Competence-trust, therefore, can have a significant direct effect on expected transaction performance outcomes.

H3a. Competence-trust will positively influence expectations re-garding transaction performance.

Uncertainty regarding the abilities of the partner will increase doubt regarding the partners' professional skills and proficiency of successful completion of the transaction. Such uncertainty is likely to be accompa-nied by negative emotions[3]such as anxiety of the vendor not success-fully fulfilling the transaction. This may include concerns that the partner is not operating adequately to fulfill the transaction or facing operational problems impeding the completion of the transaction. These negative thoughts are related to the specific transaction and can lead to negative expectations of transaction performance.

H3b. Performance-risk will negatively influence expectations re-garding transaction performance.

Transaction cost economic (TCE) theory argues that economic agents have bounded rationality. An immediate consequence of bounded rationality is the inability of organizations to manage com-plex forms of transactions[74]. Attainment of reliable information enables the economic agents, which have bounded rationality, to im-prove their decision making regarding the individual transactions. The organization can thus select the transactions that are more likely to be successful (e.g. only ordering products that are on-stock so they are received on time). The organization can also intervene when the transaction is not progressing as expected (e.g. the wrong product is selected for freight). Hence credible information enables superior deals to be made than could otherwise be supported. Improved infor-mation quality is expected to increase expectations of positive trans-action performance. An information exchange providing improved information quality enables favorable transactions.

High IQ gives the system user confidence in the successful perfor-mance of the transaction because having quality information suggests that exchange information is reliable, correct, responsive, and timely [26]. Information quality (IQ) should thus positively affect system outcomes, as also proposed by DeLone and McLean [19]. This is because a high level of IQ sends a strong signal to the user that the transaction will be performed properly. As a result, the following hypothesis is developed:

H3c.A high level of information quality will positively influence expectations regarding transaction performance.

Goodwill-trust is related to whether the partner has upright inten-tions and is concerned about the focal organization[16]. The partner is expected to be cooperative and not to behave opportunistically even if contractual agreements would allow them to do that[60]. This type of trust can be considered to have elaborate consequences as it assumes the other party will not take advantage, even if an opportunity to do so should arise[51]. This confidence transcends a

Table 1

Questionnaire items.

Items Parcel Standardized

loading

t-value Cronbach's alpha Information quality (IQ)

2. The exchange data is up-to-date enough for my purposes Dropped Dropped 0.833

4. The data this exchange provides is never outdated 11. The exchange maintains the right data for my purposes

3. The exchange provides up-to-date information with regard to transactions Information quality 1 0.81 16.50 0.868 6. I feel satisfied with the data accuracy of the exchange system

7. There are no accuracy problems in the data that I use in this exchange

5. The exchange data that I use is accurate enough for my purposes Information quality 2 0.85 18.63 0.839 8. Data provided by this exchange is completely error free

10. The data maintained by the data exchange is pretty much what I need to carry out my tasks

1. This exchange provides data that is current enough to meet my business needs

Dropped Dropped 0.855

9. The information content of the exchange meets my needs 12. Based on my needs, this exchange has no missing data items Competence-trust

1. The vendor has the necessary skills to manufacture the ordered products

Competence-trust 1 0.87 19.22 0.876

4. The vendor is a very reliable manufacturer

5. The vendor is an excellent source of accurate information

2. The vendor has the necessary abilities to achieve the desired results Competence-trust 2 0.85 18.13 0.890 3. The vendor is very knowledgeable in producing the product I ordered

6. The vendor really knows its business Goodwill-trust

2. The vendor is interested in our well being, not just in its own Goodwill-trust 2 0.89 18.19 0.861 4. The vendor is concerned about our interests

3. The vendor cares about our interests Goodwill-trust 1 0.93 17.58 –

Performance-risk

3. Risk of vendor lacking the abilities to perform as expected Performance-risk 1 0.85 15.82 0.877 5. Risk of vendor not operating well

7. Risk of vendor not having the required knowledge to execute the order

4. Risk of vendor not producing the required products Performance-risk 2 0.93 16.83 0.866

6. Risk of vendor facing performance problems

1. The vendor is committed to accomplish the goals of the relationship Dropped Dropped 0.950 2. The vendor is dedicated to accomplish the goals of the relationship

Exchange-risk

How would you characterize the possibility of using this data exchange to carry out purchasing transactions:

1. Negative situation–Positive situation Exchange-risk 1 0.92 17.34 –

2. Potential for gain–potential for loss Exchange-risk 2 0.94 18.01 –

Considering the case assigned to you, how would you rate the overall risk of carrying out transactions using this data exchange?

Exchange-risk 3—dropped Dropped 0874

4. Low–high

5. Much lower than acceptable level–much higher than acceptable level Expected transaction performance

1. Your degree of confidence in the successful completion of your order (low/high). Expected transaction performance 1 0.91 19.01 0.856 3. Your degree of certainty in assessing whether your order was processed

in a complete, accurate and timely manner (low/high).

2. Your degree of control over the successful completion of your order (low/high). Expected transaction performance 2 0.86 15.06 – Intent for continued use

1. What is the likelihood that you would continue using this exchange in the future to carry out transactions similar to the ones described in your case?

Intent for continued use 1 0.79 17.06 0.843 2. If faced with a similar purchasing decision in the future, I would use this

data exchange again

6. Ido notintend to continue to transact with the vendor of this data exchange

3. I intend to transact for a long time with the vendor of this data exchange Intent for continued use 2 0.90 18.52 0.878 5. I would recommend use of this data exchange to other colleagues who may

be faced with similar ordering needs as the one described in my case

All items were measured on seven-point response scales, using the above stated adjectives as endpoints for each item. Fit statistics. χ2 56= 73.74. RMR = 0.014. RMSEA = 0.036. GFI = 0.956. AGFI = 0.917. NFI = 0.986. CFI = 0.997.

specific transaction in its specific domain. The goodwill-trust gener-ates expectations regarding the cooperative behavior of the partner in general and justi_fies prolonging the interorganizational relation-ship. Hence, goodwill-trust is expected to positively influence the intent for continued use of the electronic exchange.

H4a. Goodwill-trust will positively influence the intent for con-tinued use of a data exchange.

Existence of exchange-risk indicates that the exchange is not perceived as truly cooperative. There is a signi_ficant threat of opportu-nistic behavior and hidden agenda. The partners can feel reluctant to use the exchange because they are anxious that the partner will misuse the information. Risk-taking theory also suggests that risk perception of individuals will affect their willingness to perform a risky behavior [36,65,66]. This reluctance surpasses the current transaction and may have negative influences on future business dealings using the ex-change. Accordingly we expect an inverse relationship between risk perception and intent for continued use of the data exchange.

H4b. Exchange-risk will negatively influence the intent for con-tinued use of a data exchange.

Expectations of improved transaction performance can positively influence the intent to use the interorganizational exchange. Such ex-pectations are critical in forming a positive attitude towards conducting business with the partner. The theory of reasoned action[22]posits that behavioral intention depends on the attitudes about the behavior and subjective norms. Attitude is directly related to convictions about the positive and negative consequences of performing the behavior. Accordingly, a positive attitude towards transaction performance is expected to increase the intentions to continue to use the I-O data exchange.

H5.Transaction performance expectation will positively influence the intent for continued use of a data exchange.

3. Methods

3.1. Construct measurement

The items used to measure the constructs came from a number of sources (seeTable 1). The study used the reflective information quality scale from Nicolaou and McKnight[53]. The items included the curren-cy, accuracurren-cy, relevance, completeness and reliability aspects of the data exchange. We adopted items measuring competence-trust and goodwill-trust from McKnight et al.[49]. We tried to capture more nuance and therefore we complemented them with items from Sengun and Wasti[62]. Performance-risk was measured using items derived from Das and Teng[16]. The items were adapted to coincide with the context of this study. For expected transaction performance, specific transaction performance outcomes were developed in this research and were designed to capture the subjects' assessment that the specific exchange they were assigned to would perform as expected with regard to processing their transactions and ensuring the completeness and validity of transmitted data. Similar items measuring perceptions of performance were also examined in several past studies, including Kaplan and Nieschwietz[33], Roberts and O'Reilly[59], and Zand[75]. Intent for continued use was adapted from the scale used in Nicolaou and McKnight[53]. For the control constructs of risk propensity and dis-position to trust, we used items previously reported in Keil et al.[36] and Lee and Turban[39]. Five different items were used to measure sub-ject perceptions of risk in the exchange they used to carry out their transactions. Thefirst item measured overall risk, as perceived by the subjects, given their condition/case. This item was developed in the Nicolaou and McKnight[53]study and asked the subjects to rate the

overall risk of engaging in transactions with this exchange. The other four items were adapted from Keil et al.[36]. Two of the items assessed the level of perceived risk directly (whether the subjects perceived any risks in continuing to use the data exchange), while the other two items assessed the perceived probability of success of the exchange, which, according to Keil et al.[36]contribute indirectly to risk perception. This measurement corresponds to a conceptualization of risk as a com-posite measure of the perceived probability and magnitude of loss. The four-item scale attained an internal reliability (Cronbach's alpha) coefficient of 0.80 in the Keil et al.[36]study.

3.2. Data collection

Data for this study were collected through a questionnaire adminis-tered in the context of an experiment (the attached Appendix A summarizes the experimental controls employed). The study employed a web exchange designed for this study that simulated a real-world exchange environment. The customized web exchange used experi-mental controls to manipulate an exchange design feature (high versus low control transparency) and two levels of situational exchange char-acteristics (standard product with new business partner, or unique product from a preferred business partner). These manipulations were designed,first, to increase variance among the study variables. Second, along with the simulated data exchange, they increase the study's real-world relevance[4]. We examined a number of actual exchange sites so the manipulations would reflect real conditions.

After being invited to the experiment, thefirst page requested potential participants to provide information regarding their profes-sional experience and procurement responsibilities for screening purposes. Individuals were screened when answering the question “Have you ever had a job responsibility where you purchased goods on behalf of your employer?”A number of 263 potential participants who gave a negative response to this question were excluded from the experiment. The next page described how technology enables web-based data exchanges using extensible markup language (XML). Next, a description of either data exchange A (high control transparency) or data exchange B (low control transparency) was provided with a link to the website of the particular exchange. The website provided instruc-tions how to register as a new customer for thisfictitious vendor and enter transactions using the exchange. The participants were familiar-ized with the exchange by carrying out two practice transactions before placing their“real”order. After the practice session, all participants were asked to assume the role of a purchasing manager in a manufacturing plant conducting a needed raw-materials order. Participants were randomly assigned to one of the two relational characteristics conditions, and also to one of the two control transparency conditions. All treat-ments explained to individuals the features of the product and the status of the relationship with the business partner.

In the end, 221 subjects submitted an actual order on their simu-lated web exchange, and were included in the collection of data on the study's main constructs. Fifty-nine percent of them were male and the average age was forty years. The participants had an average professional experience of 18 years (with a range from 1.2 years to 50 years) and their average length of performing purchasing respon-sibility was 8 years and 3 months (with a range from 3 months to 40 years).

3.3. Questionnaire item quality checks

We examined the measurement invariance of each scale in order to test whether the measurement items for each construct vary and covary in the same manner across the treatments we employed (see Vandenberg and Lance[71]for comprehensive review or Carte and Russell[13] for procedure). Due to violation of measurement invariance assumptions, we removed three items of IQ (items: 1, 9

and 12) and one item of performance-risk (item 1).Table 2shows all items of the model including items that were dropped.

4. Results

Structural equation modeling (SEM) is used for data analysis as it allows validating the entire model in a single and systematic comphensive analysis. This is realized by using LISREL for modeling the re-lationships among multiple related equations simultaneously [23]. LISREL consists of two parts: the measurement model and the structural equation model. The measurement model identifies the relations between the observed measures, i.e. items, and their underlying latent constructs. The structural equation model calculates the causal relations between the constructs as put forward by the underlying theory. LISREL offers the opportunity to calculate the maximum likelihood estimates for both the measurement model and the structural equation model si-multaneously. However, it is recommended to follow a two-staged ap-proach whereby the measurement model is calculated and fixed before the structural model is estimated[23,24].

Psychology literature recommends the development of parcels using multiple items, which are used to measure the same construct [14]. Parceling is a measurement practice that is used most commonly in multivariate approaches to psychometrics, particularly for use with latent-variable analysis techniques[41]. A parcel can be defined as an aggregate-level indicator comprised of the sum (or average) of two or more items. Parceling is theoretically justified in cases where con-structs vary in their fundamental composition, as measured using unidimensional or multidimensional items. The measurement of our constructs is represented by a heterogeneous set of items, meant to capture a representative assessment of all facets or nuances of the be-havioral domain underlying each construct. While there is a debate in the psychometrics literature[41], given these reasons, it is theoreti-cally justified to employ the parceling method. We constructed unidi-mensional parcels by assigning items to parcels with the objective of attaining a small number of parcels for each construct[38]. Parcels have a higher reliability than single items and the smaller order of the correlation matrix enables improved modelfit[14]. Parcels can also reduce sampling error, originating from the lack of exact corre-spondence between a sample and a population[22]. For the previous-ly mentioned reasons, parcels of items were constructed and used in this study. Table 1 shows the items included in each parcel and Table 2shows descriptive statistics of all parcels.

The measurement properties of the constructed parcels have been further examined and tested for convergent and discriminant validity using LISREL confirmatory factor analysis (CFA). The measurement model is revised by dropping parcels that shared a high degree of re-sidual variance with other items[23,24,43].Table 2includes all rele-vant statistics (loadings, t-values and Cronbach's alpha) of the measurement model. One parcel related to information quality and

one relating to exchange-risk were dropped due to high residual var-iance.3_{The measurement model showed acceptable model}

fit. The_χ2 of 73.74 with 56 degrees of freedom is aχ2_{to df ratio of less than the} recommended 1:3. The GFI at 0.956, AGFI at 0.917, NFI at 0.986, CFI at 0.997, RMR at 0.016, and RMSEA at 0.036 are all within acceptable limits for CFA.

Discriminant validity of the measurement model was also verified using the guidelines of Segars[61]. All modification indices in the lambda X matrix were below the threshold of 5. This indicates that adding any path from any measurement item to other latent variables, other than the one it is assigned to, does not cause significant change in theχ2 sta-tistic. Hence, this indicates there is no cross loading of any measurement item on other constructs. Finally, the composite construct reliabilities of all constructs were validated to have values higher that the recom-mended threshold of 0.8 (seeTable 3).

Next, all hypothesized propositions are simultaneously tested by means of examining the structural model. Thefit measures are accept-able:χ2

to degrees of freedom ratio of 1:2.065 (χ2

71= 187.16), The GFI at 0.90, AGFI at 0.87, CFI at 0.98, NFI at 0.97, and RMSEA at 0.069 are within acceptable limits; only the RMR at 0.10 is slightly above the recommended threshold.Fig. 2shows the standardized LISREL path coefficients.

In terms of hypotheses testing, the results support hypothesis 1A as IQ positively influences competence-trust (t= 8.18). Hypothesis 1B is also supported as the influence of IQ on goodwill-trust is significant (t= 7.63). Furthermore, we found IQ does not have a significant negative influence on performance-risk (t=−1.07), hence hypothesis 2A is not supported. However IQ does have a signi_ficant in_fluence on exchange-risk (t=−8.22), thus supporting hypothesis 2B. We also found expected transaction performance to be significantly positively influenced by competence-trust (t= 4.60) and IQ (t= 6.80), and signif-icantly negatively influenced by performance-risk (t=−4.46). Hence, hypotheses H3a, H3b and H3care supported. As expected, intention to continue using the exchange was significantly positively influenced by goodwill-trust (t= 2.71) and significantly negatively influenced by exchange-risk (t=−6.89), which supporthypotheses H4a and H4b. Finally,H5is also supported as transaction performance expectation positively influences the intention to continue using the exchange (t= 7.96).Table 4shows the results of each research hypothesis and summarizes the results for the overall structural model.

5. Discussion

This study builds on existing studies on the effects of IQ on system adoption and system success and contributes to IS theory in two ways. First, we have succeeded in providing empirical evidence on

Table 2

Descriptive statistics and correlations between the parcels.

M SD SK KU 1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 Information quality 1 4.23 1.42 −.27 −.23 – 2 Information quality 2 4.11 1.38 −.27 −.14 .89 – 3 Competence-trust 1 4.3 1.05 −.28 1.54 .43 .49 – 4 Competence-trust 2 4.39 1.05 −.29 1.44 .40 .44 .92 – 5 Goodwill-trust 1 4.19 1.06 −.10 1.07 .38 .45 .65 .64 – 6 Goodwill-trust 2 4.17 1.11 −.07 .57 .38 .42 .66 .63 .88 – 7 Performance-risk 1 4.13 1.19 .11 .59 −.05 −.07 −.11 −.10 −.14 −.08 – 8 Performance-risk 2 4.19 1.18 .11 .33 −.05 −.07 −.11 −.10 −.16 −.11 .93 – 9 Exchange-risk 1 3.83 1.36 .24 .01 −.41 −.46 −.45 −.44 −.43 −.40 .28 .28 – 10 Exchange-risk 2 3.99 1.36 .20 −.18 −.40 −.45 −.45 −.43 −.49 −.48 .29 .30 .86 – 11 Transaction performance 4.25 1.39 −.36 .07 .52 .54 .49 .50 .48 .48 −.28 −.29 −.70 −.71 – 12 Transaction performance 3.99 1.59 −.31 −.62 .51 .50 .44 .43 .39 .42 −.20 −.22 −.60 −.62 .84 – 13 Intent for continued use 1 4.11 1.45 −.37 .05 .47 .49 .44 .42 .48 .45 −.26 −.24 −.67 −.67 .70 .59 – 14 Intent for continued use 2 4.07 1.50 −.38 −.09 .44 .50 .49 .47 .52 .48 −.22 −.23 −.67 −.68 .76 .68 .86 –

3

Two items measuring exchange risk remained in the model. These items were used in the measurement model directly.

distinctive roles of two types of trust and two types of risk. These dis-tinctions are beneficial in explaining how organizations can use IQ to utilize informal forms of cooperation with the objective of enhancing participation in electronic exchanges. We provide empirical evidence that IQ has distinctive in_fluences on competence-trust and goodwill-trust. On the one hand, Information quality impacts rational cognition by influencing competence-trust. High quality information is perceived as an indicator of the pro_ficiency of the partner. By providing the required information, the partner organization shows it is willing to share its expertise to realize a successful business transaction. On the other hand, information quality impacts affect by influencing goodwill-trust. High quality information indicates that the partner has good intentions. These intentions can increase the positive beliefs and creates a leap of faith boasting willingness to utilize the exchange. The honesty and good information provided by the partner are thus essential for the existence of benevolence. Hence, the combination of insights on different types of trust[47,60]with insights on the effects of trust[30,34] enables us to have more detailed understanding of the complex trust-related mechanisms interfering with the usage of inter-organizational systems. The second contribution to IS theory relies on exploring the mechanisms related to risk. We provide empirical evidence that IQ has distinctive influences on two types of risk. High information quality improves the transparency. Transparency is utilized to monitor the transaction and detect any abnormalities such asfl uctu-ations in stock and order status changes. The tight monitoring and detection of abnormalities enables the organization to monitor partner

performance. This is utilized to reduce performance-risks and decrease moral hazards of opportunism[64]. Besides, the high quality informa-tion can also be used to observe the performance of the electronic ex-change. The exchange may rely on internal and external systems for calculating product availability and prices. The high quality information can be utilized to ensure the exchange is working correctly and exclude exchange-risk. Thesefindings support economic exchange theory as the exchanges are supported by controls to monitor and reward desir-able performance[16]and information that enforce obligations and strengthen trust relationships[21,25].

Ourfindings are also interesting for practitioners that are confronted with increasing uncertainty and market volatility. This study shows that improving IQ facilitates competence-trust and goodwill-trust besides decreasing exchange-risk. Decision makers can utilize IQ to effectively increase business partner's perceptions of the expertise and professional reputation of the focal organization. Decision makers can also improve IQ to demonstrate their benevolence and goodwill towards the business partner. The combination of increased expertise perceptions, increased goodwill and decreased exchange-risk will increase IOS adoption in the short term and is likely to enhance the business relationship in the long term.

6. Limitations

This study has two main limitations due to the experimental set-ting. Firstly, the generalizability offindings is limited to the domain of automated interorganizational data exchanges. Because the ex-change is automated, users cannot conduct any interpersonal interac-tion, e.g. face to face discussions or via a chat context. Future research can expand the domain by conducting case studies or surveys on real-world exchanges. Such exchanges can utilize more advanced and rich communication features. Secondly, the design of the study was limited to single transactions. Future research can treat this issue by adopting longitudinal experimental designs or conducting

field studies on real-world exchanges to incorporate the influences of past interactions.

Table 3

Composite construct reliabilities.

Construct Cronbach's alpha

IQ 0.943 Competence-trust 0.957 Goodwill-trust 0.938 Performance-risk 0.966 Exchange-risk 0.927 Transaction performance 0.908 Intent to use 0.925 -0,23* 0,13* -0,34* 0,48* 0,48* -0,52* 0,44* 0,51* 0,28* -0,07 0,03 Perceived Performance Risk Perceived Exchange Risk Expected Transaction Performance Competence Trust Goodwill Trust Intent for continued Use Information Quality

7. Conclusion

Overall, the framework presented in this study produced two main findings regarding the influences of information quality on intent to use electronic exchanges. First, the study was able to inves-tigate the distinctive mediating roles of two types of trust and two types of risk. This study extends prior research by providing empirical evidence to illustrate the different influences of IQ on each type. Second, the study illustrates the complex influences of trust and risk by utilizing the intermediating role of users' expectations from trans-action performance.

Appendix A. Experimental manipulations control

Participants were provided an experimental experience meant to simulate real world exchanges and then their reactions were elicited via a questionnaire. Participants were recruited through a commercial agency that has contact information of professionals within different

fields. Our objective is to acquire a random sample by selecting a sub group of professionals from a larger population. Participants are com-pensated with a minor monetary reward. The participants are profes-sional managers with current or prior procurement responsibilities in the United Kingdom. Individual prospective participants are screened at the beginning of the survey to ensure that allfinal subjects hold appropriate qualifications and are able to relate to the purchasing management role required by the simulated exchange. A total num-ber of 221 participants were secured for this study.

The task was to perform the role of a plant purchasing manager ordering raw materials (aluminum sheets) required for their plant's production (all 221 of the subjects reported having actual purchasing experience). We manipulated control transparency and product novelty features of the data exchange. Since the real world exchanges we saw employ different design features, we experimentally manipu-late the control transparency and product novelty system design interventions.

Prior to their use of the experimental transaction, participants received training in the web-based data exchange and were then assigned to their particular treatment. The study design was a 2 × 2 fully-crossed between subjects, with high/low levels of control

transparency and two levels of product novely (new or familiar prod-uct). We randomly assigned participants to treatments by varying the web-based data exchange model used (resulting in four different de-sign variations, corresponding to our 2 × 2 study dede-sign). For example, after a subject input an order, the validation screen for high control transparency showed_“Valid Data_”next to each item entered correctly and determined to match the content included in an underlying working database. The product novelty treatment was materialized by presenting different instructions to subjects in a case scenario, directing subjects to deal with a novel or familiar product from their supplier. All other information and instructions were the same across groups except the part embodying the treatment.

We measured a number of manipulation check items. These were included at the end of the measurement instrument in order to avoid in-ducing demand effects. Items used a seven-point, strongly agree to strongly disagree scale. Control transparency was assessed using the four items: (1) the data exchange displayed an error message when I entered data that were invalid; (2) the data exchange applied controls over the format of data I input on forms (e.g., all required data needed to be present); (3) the data exchange applied controls over the content of data I input on forms (e.g., it checked to verify I input the correct item number); and (4) the data exchange displayed an error message when I entered data that were invalid. The results of F-tests on all four control transparency checks show that manipulations worked: Item 1 means: A—4.93, B—3.09, F = 60.5; Item 2 means: A—4.92, B—3.44, F = 49.9; Item 3 means: A—4.99, B—3.30, F= 60.1; Item 4 means: A— 5.17, B—2.93, F = 96.5. Product novelty was assessed by asking respon-dents about their degree of agreement with this item:“the ordered product had specifications that were unusual to this supplier.” As expected, F-tests on the product novelty manipulation check revealed that there were significant differences in the novel versus familiar product groups, in the direction expected by the manipulation. These results demonstrate that each of our experimental manipulations worked as intended. We were thus assured of the efficacy of the exper-imental manipulations employed.

References

[1] D.K. Allen, D. Colligan, A. Finnie, T. Kern, Trust, power and interorganizational information systems: the case of the electronic trading community translease, Information Systems Journal 10 (1) (2000) 21–40.

[2] V. Arnold, T. Benford, C. Hampton, S. Sutton, Competing pressures of risk and absorptive capacity potential on commitment and information sharing in global supply chains, European Journal of Information Systems 19 (2010) 134–152. [3] L. Asveld, S. Roeser, The Ethics of Technological Risk, Earthscan, London, 2009. [4] P. Attewell, J.B. Rule, Survey and other methodologies applied to IT impact research:

experiences from a comparative study of business computing, in: K.L. Kraemer (Ed.), The Information Systems Research Challenge: Survey Research Methods vol. 3, Harvard Business School Press, Cambridge, MA, 1991, pp. 229–316.

[5] J.Y. Bakos, E. Brynjolfsson, Information technology, incentives and the optimal num-ber of suppliers, Journal of Management Information Systems 10 (2) (1993) 37. [6] M. Bensaou, Interorganizational cooperation: the role of information technology

—an empirical comparison of U.S. and Japanese supplier relations, Information Systems Research 8 (June 1997) 107–124.

[7] M. Bensaou, N. Venkatraman, Configurations of interorganizational relationships: a comparison between U.S. and Japanese automakers, Management Science 41 (September 1995) 1471–1492.

[8] P. Blau, Exchange and Power in Social Life, Wiley, New York, 1964.

[9] G.W. Bock, J. Lee, H.H. Kuan, J.H. Kim, The progression of online trust in the multi-channel retailer context and the role of product uncertainty, Decision Sup-port Systems 53 (2012) 97–107.

[10] M.V. Bovee, Information Quality: A Conceptual Framework and Empirical Validation, Ph.D. Dissertation, University of Kansas (2004).

[11] J.L. Bradach, R.G. Eccles, Price, authority, and trust: from ideal types to plural forms, Annual Review of Sociology 15 (1989) 97–118.

[12] J. Brownlie, A. Howson, Leaps of faith and MMR: an empirical study, Sociology 39 (2) (2005) 221–239.

[13] T.A. Carte, C.J. Rusel, In pursuit of moderation: nine common errors and their solutions, MIS Quarterly 27 (3) (2003) 479–501.

[14] D.L. Coffman, R.C. MacCallum, Using parcels to convert path analysis models into latent variable models, Multivariate Behavioral Research 40 (2) (2005) 235–259.

Table 4

Parameter estimates for the structural model.

Hypotheses Hypothesized sign Standardized solution t-value 1a. IQ→Competence-trust + 0.52 8.18 1b. IQ→Goodwill-trust + 0.53 7.63 2a. IQ→Performance-risk − −0.078 −1.07 2b. IQ→Exchange-risk − −0.59 −8.22

3a. Competence-trust→Transaction performance + 0.30 4.60 3b. Performance-risk→Transaction performance − −0.24 −4.46 3c. IQ→Transaction performance + 0.46 6.80 4a. Goodwill-trust→Intent to use + 0.13 2.71 4b. Exchange-risk→Intent to use − −0.32 −6.80 5. Transaction performance→Intent

to use + 0.43 7.96 Fit statistics. χ2 71= 187.16. RMR = 0.10. RMSEA = 0.08. GFI = 0.90. AGFI = 0.85. NFI = 0.97. CFI = 0.98.

[15] T.K. Das, B.S. Teng, Between trust and control: developing confidence in partner cooperation in alliances, Academic Management Review 23 (3) (1998) 491–512. [16] T.K. Das, B.S. Teng, Trust, control, and risk in strategic alliances: an integrated

framework, Organization Studies 22 (2) (2001) 251–283.

[17] T.K. Das, B.S. Teng, Relational risk and its personal correlates in strategic alliances, Journal Of Business And Psychology 15 (3) (2001) 449–465.

[18] C. Dellarocas, Building Trust On-Line; the Design of Reliable Reputation Reporting Mechanisms for Online Trading Communities, Center for EBusiness working paper 101, MIT Sloan School of Management, Cambridge, MA, 2001. (Pub-lished (as of 2/01/02) athttp://ebusiness.mit.edu/research/papers-number.html

#101).

[19] W.H. DeLone, E.R. McLean, The DeLone and McLean model of information systems success: a ten-year update, Journal of Management Information Systems 19 (4) (2003) 9–30.

[20] K.T. Dirks, D.L. Ferrin, The role of trust in organizational settings, Organization Science 12 (4) (2001) 450–467.

[21] J.H. Dyer, H. Singh, The relational view: cooperative strategy and sources of interorganizational competitive advantage, Academy of Management Review 23 (4) (1998) 660–679.

[22] M. Fishbein, I. Ajzen, Belief, Attitude, Intention, and Behavior: An Introduction to Theory and Research, Addison-Wesley, Reading, 1975.

[23] D. Gefen, E. Karahanna, D.W. Straub, Trust and TAM in online shopping: an integrated model, MIS Quarterly 27 (March 2003) 51–90.

[24] D.W. Gerbing, J.C. Anderson, An updated paradigm for scale development incor-porating unidimensionality and its assessment, Journal Of Marketing Research 25 (1988) 186–192.

[25] J. Goo, R. Kishore, H.R. Rao, K. Nam, The role of service level agreements in rela-tional management of information technology outsourcing: an empirical study, MIS Quarterly 33 (1) (2009) 119–145.

[26] D.L. Goodhue, Task-technologyfit and individual performance, MIS Quarterly 19 (June 1995) 213–236.

[27] D.L. Goodhue, Understanding user evaluations of information systems, Manage-ment Science 41 (December 1995) 1827–1844.

[28] D.L. Goodhue, M.D. Wybo, L.J. Kirsch, The impact of data integration on the costs and benefits of information systems, MIS Quarterly 16 (3) (1992) 293–308. [29] R. Gulati, M. Gargiulo, Where do interorganizational networks come from? American

Journal of Sociology 104 (1999) 1439–1493.

[30] P. Hart, C. Saunders, Power and trust: critical factors in the adoption and use of electronic data interchange, Organization Science 8 (1) (1997) 23–42. [31] W. Hasselbring, Information system integration, Communications of the ACM 43

(6) (June 2000) 32–38.

[32] M. Ibrahim, P.M. Ribbers, The impacts of competence-trust and openness-trust on interorganizational systems, European Journal of Information Systems 18 (3) (2009) 223–234.

[33] S.E. Kaplan, R.J. Nieschwietz, An examination of the effects of webtrust and company type on consumers' purchase intentions, International Journal of Accounting Infor-mation Systems (7) (2003) 155–168.

[34] W. Ke, H. Liu, K.K. Wei, J. Gu, H. Chen, How do mediated and non-mediated power affect electronic supply chain management system adoption? The mediating effects of trust and institutional pressures, Decision Support Systems 46 (2009) 839–851.

[35] P.G.W. Keen, C. Ballance, S. Chan, S. Schrump, Electronic Commerce Relationships: Trust by Design, Prentice Hall PTR, Upper Saddle River, NJ, 2000.

[36] M. Keil, B.C.Y. Tan, K.K. Wei, T. Saarinen, A cross-cultural study on escalation of commitment behavior in software projects, MIS Quarterly 24 (2) (2000) 299–325.

[37] D. Khazanchi, S.G. Sutton, Assurance services for business-to-business electronic commerce: a framework and implications, Journal of the Association for Information Systems 1 (January 2001).

[38] J.M. Kishton, K.F. Widaman, Unidimensional versus domain representative parcelling of questionnaire items: an empirical analysis, Educational And Psychological Mea-surement 54 (1994) 757–765.

[39] M.K.O. Lee, E. Turban, A trust model for consumer internet shopping, International Journal Of Electronic Commerce 6 (1) (2001) 75–91.

[40] S.C. Lee, B. Pak, H. Lee, Business value of b2b electronic commerce: the critical role of inter-firm collaboration, Electronic Commerce Research and Applications 2 (4) (2003) 350–361.

[41] T.D. Little, W.A. Cunningham, G. Shahar, K.F. Widaman, To parcel or not to parcel: exploring the question, weighing the merits, Structural Equation Modeling 9 (2) (2002) 151–173.

[42] N. Luhmann, Trust and Power, John Wiley, New York, 1979.

[43] R.C. Maccallum, K.F. Widaman, S. Zhang, S. Hong, Sample size in factor analysis, Psychological Methods 4 (1999) 84–99.

[44] A. Malhotra, S. Gosain, O.A.E. Sawy, Absorptive capacity configurations in supply chains: gearing for partner-enabled market knowledge creation, MIS Quarterly 29 (1) (2005) 145–187.

[45] T.W. Malone, J. Yates, R.I. Benjamin, Electronic markets and electronic hierarchies, Communications of the ACM 30 (1987) 484–497.

[46] R.C. Mayer, J.H. Davis, F.D. Schoorman, An integrative model of organizational trust, Academic Management Review 20 (1995) 709–734.

[47] D.J. McAllister, Affect- and cognition-based trust as foundations for interpersonal cooperation in organizations, Academy Of Management Journal 38 (1) (1995) 24–59.

[48] V. McKinney, K. Yoon, F.M. Zahedi, The measurement of web-customer satisfaction: an expectation and disconfirmation approach, Information Systems Research 13 (3) (2002) 296–315.

[49] D.H. McKnight, V. Choudhury, C. Kacmar, Developing and validating trust measures for e-commerce: an integrative typology, Information Systems Research 13 (3) (2002) 334.

[50] T. Miyamoto, R. Nexhmi, Determinants of three facets of customer trust: a marketing model of Japanese buyer–supplier relationship, Journal of Business Research 57 (3) (2004) 312–319.

[51] R.M. Morgan, S.D. Hunt, The commitment-trust theory of relationship marketing, Journal Of Marketing 58 (1) (1994) 20–38.

[52] T. Mukhopadhyay, S. Kekre, Strategic and operational benefits of electronic inte-gration in B2B procurement processes, Management Science 48 (10) (2002) 1301–1313.

[53] A.I. Nicolaou, D.H. McKnight, Perceived information quality in data exchanges: effects on risk, trust, and intention to use, Information Systems Research 17 (4) (2006) 332–351.

[54] B. Nooteboom, Trust: Forms, Foundations, Functions, Failures and Figures, Edward Elgar, Cheltenham, UK, 2002.

[55] B. Nooteboom, H. Berger, N.G. Noorderhaven, Effects of trust and governance on relational risk, Academy Of Management Journal 40 (2) (1997) 308–338. [56] P.A. Pavlou, D. Gefen, Building effective online marketplaces with institution-based

trust, Information Systems Research 15 (1) (2004) 35–53.

[57] A. Rai, R. Patnayakuni, N. Seth, Firm performance impacts of digitally enabled supply chain integration capabilities, MIS Quarterly 30 (2) (2006) 225–246. [58] F.F. Reichheld, P. Schefter, E-loyalty: your secret weapon on the web, Harvard

Business Review 78 (4) (2000) 105–113.

[59] K.H. Roberts, C.A. O'Reilly, Some correlations of communication roles in organizations, Academy Of Management Journal 22 (1) (1979) 42–57.

[60] M. Sako, Does trust improve business performance, in: C. Lane, R. Buchmann (Eds.), Trust Within And Between Organizations: Conceptual Issues And Empirical Applica-tions, Oxford University Press, Oxford, 1998.

[61] A.H. Segars, Assessing the unidimensionality of measurement: a paradigm and illustration within the context of information systems research, Omega 1 (25) (February 1997) 107–121.

[62] A.E. Sengun, S.N. Wasti, Trust, control, and risk, a test of Das and Teng's conceptual framework for pharmaceutical buyer–supplier relationships, Group And Organiza-tion Management 32 (4) (2007) 430–464.

[63] J.A. Sharp, Evaluating the effect of sourcing decisions on e-commerce process operations risk, European Journal Of Operational Research 180 (2) (2007) 815–833. [64] J. Singh, D. Sirdeshmukh, Agency and trust mechanisms in consumer satisfaction and loyalty judgments, Journal of the Academy of Marketing Science 28 (1) (2000) 150–167.

[65] S.B. Sitkin, A.L. Pablo, Reconceptualizing the determinants of risk behavior, Academy Of Management Review 17 (1) (1992) 9.

[66] S. Sitkin, L.R. Weingart, Determinants of risky decision-making behavior: a test of the mediating role of risk perceptions and propensity, Academy Of Management Journal 38 (6) (1995) 1573.

[67] J. Song, F.M. Zahedi, A theoretical approach to web design in e-commerce: a belief reinforcement model, Management Science 51 (8) (2007) 1219–1235. [68] S.G. Sutton, C. Hampton, Risk assessment in an extended enterprise environment:

redefining the audit model, International Journal of Accounting Information Systems 1 (4) (2003) 57–73.

[69] S.G. Sutton, D. Khazanchi, C. Hampton, V. Arnold, Identification of critical risk factors in B2B e-commerce relationships, Journal of the Association for Informa-tion Systems 9 (3/4) (2008) 151–174.

[70] A.H. Van de Ven, P.S. Ring, Relying on trust in cooperative inter-organizational relationships, in: R. Bachmann, A. Zaheer (Eds.), Handbook of Trust Research, Edward Elgar, Cheltenham, UK, 2006, pp. 144–164.

[71] R.J. Vandenberg, C.E. Lance, A review and synthesis off the measurement invariance literature: suggestions, practices and recommendations for organizatonal research, Organizational Research Methods 3 (January 2000) 4–69.

[72] N. Venkatraman, A. Zaheer, Electronic integration and strategic advantage: a quasi-experimental study in the insurance industry, Information Systems Research 1 (4) (1990) 377–393.

[73] O.E. Williamson, Transaction-cost economics: the governance of contractual rela-tions, Journal of Law and Economics 22 (2) (Oct. 1979) 233–261.

[74] O.E. Williamson, Calculativeness, trust, and economic organization, Journal of Law and Economics 36 (1 Part 2) (1993) 453–486.

[75] D.E. Zand, Trust and managerial problem solving, Administrative Science Quarterly 17 (1972) 229–239.

[76] J. Zheng, E. Bakker, L. Knight, H. Gilhespy, C. Harland, H. Walker, A strategic case for e-adoption in healthcare supply chains, International Journal Of Information Management 26 (4) (2006) 290–301.

Andreas I. Nicolaouis the Owens-Illinois Endowed Professor at the College of Business, Bowling Green State University, USA, and the Editor-in-Chief of theInternational Journal of Accounting Information Systems. He serves as editor of IJAIS since 2010. He earned his Ph.D. from Southern Illinois University-Carbondale, and his baccalaureate from the Athens University of Economics and Business, Greece. He adopts both a quantitative as well as qualitative focus in his research as he examines relational issues in inter-organizational exchanges and management control, and issues on the use of integrated information sys-tems and their relations to the information environment of business organizations. He has published over 50 research articles, and his work has appeared in such journals asContem-porary Accounting Research,Information Systems Research,Journal of Management Informa-tion Systems,Journal of Information Systems,Electronic Markets,Information Technology& People,and European Accounting Review, among others. He is also the author of two books. He can be reached at [email protected].

Mohammed Ibrahimreceived the M.Sc. and Ph.D. degrees in Information Manage-ment from CentER, Center of Economic Research, at Tilburg University. He became an assistant professor at the Rotterdam School of Management in 2006 before assuming his current position as an architect of SOA-enabled enterprise systems in 2009. His current work and research interests focus on service-oriented architecture, interorganizational systems, trust and social psychology. He can be reached at [email protected].

Eric van Heckholds the Chair of Information Management and Markets at Rotterdam School of Management, Erasmus University, where he is conducting research and is teach-ing on the strategic and operational use of information technologies for companies and markets. He was ERIM's Director of Doctoral Education (2007–2009), a Visiting Professor at the Helsinki School of Economics (2002–2005) and the Ludwig-Maximilians University in Munich (Summer 2006), and a Visiting Scholar at MIT Sloan School of Management (Summer 2009). He can be reached at [email protected].