B2B EC Procurement Info Transfer Analysis Measurement

8/3/2019 B2B EC Procurement Info Transfer Analysis Measurement

1/16

Information transfer in B2B procurement:an empirical analysis and measurement

Kyung Kyu Kima, Narayan S. Umanathb,*

aYonsei University, KoreabCollege of Business Administration, University of Cincinnati,

328 Lindner Hall, Cincinnati, OH 45221-0211, USA

Received 15 July 2003; received in revised form 12 May 2004; accepted 3 August 2004

Available online 8 December 2004

Abstract

Inter-organizational relationships employing IT may be the most important technological breakthrough in B2B partnerships,

since it is likely to alter the competitive landscape of industries radically. Electronic integration (EI) may be defined as the

integration of business processes of two or more independent organizations through the exploitation of the capabilities of

computer and communication technologies. Prior research has primarily used the adoption of electronic data interchange (EDI)

as a surrogate measure for EI. While researchers have called for the assessment of the degreeof EI instead of presence/absence of

EDI between two firms, a measure was still to be developed. Conceptualizing EI as a multi-dimensional construct, our researchfocused on developing a measure for a crucial component: electronic information transfer (EIT). Four dimensions of it (decision

and operation integration (DOI), mutual investment in relationship-specific assets (MIRSA), information sharing (IS), and

monitoring and control (MAC)) were analyzed and an instrument for EIT measurement was developed. Data collected from two

major corporations in the U.S. were used to verify the instruments ability to measure EIT effectively.

# 2004 Elsevier B.V. All rights reserved.

Keywords: Inter-organizational systems; Information sharing; Electronic integration; EDI; Electronic information transfer; Supply chain

management

1. Motivation

The use of the Internet to facilitate B2B commerce

has attracted much attention from both academics and

practitioners due to its potential impact on industry

structure and the way business is conducted today[14]. Internet markets have the potential to widen the

choices available to buyers, provide sellers access to a

larger customer base, and slash transaction costs [17].

The B2B markets take different forms (e.g., spot

markets, electronic hierarchies, cooperative arrange-

ments) depending on the characteristics of the

products being exchanged, market variability repre-

www.elsevier.com/locate/dsw

Information & Management 42 (2005) 813828

* Corresponding author. Tel.: +1 513 556 7195;

fax: +1 513 556 6278.

E-mail addresses: [email protected] (K.K. Kim),

[email protected] (N.S. Umanath).

0378-7206/$ see front matter # 2004 Elsevier B.V. All rights reserved.

doi:10.1016/j.im.2004.08.004


2/16

sented by market fragmentation and market volatility,

and continuity of the business relationship between the

channel partners.

Existence of production economies favors theemergence of specialized firms interacting in spot

markets. However, in some situations, market con-

tracting can become difficult, increasing the transac-

tion costs of managing the interaction. At some point, it

becomes more efficient to administer the interactions

within a long-term cooperative relationship. Often

environmental and technological factors will make it

possible to increase the overall efficiency of production

or exchange through closer integration of decisions

and operations between the trading partners. However,

this increase in coordination necessarily involves

investment to tailor operations to specific interactions.

When the expected benefits from investments in

coordination minus the cost of this investment are

sufficiently large to counterbalance the loss of an out-

side suppliers production economies, firms can make

investments to gain the benefits of such coordination

[9]. Lack of benefits from explicit coordination often

lead to a transaction-oriented spot market.

One of the key differences in various forms of the

B2B market is the level of integration between the

trading partners. In a spot market transaction, since the

buyers goal is to fulfill an immediate need at thelowest possible cost, minimum integration between

the trading partners is sufficient. Meanwhile, in a

system where transactions occur for a long-term

period in negotiated contracts with qualified suppliers,

a high level of integration between the trading partners

should exist to achieve efficiency. Thus, the level of

integration between the channel partners will vary. On

the basis of organizational information processing

theory [12], one can argue that more is not always

better, especially in electronic integration (EI)

between supply channel partners [18]. Electronicmedia may overload decision makers in a supply

channel with too much information [27]. Inability to

cope with such an information overload leads to

organizational dysfunction. Therefore, the fit

between contextual factors and electronic integration

should be examined to seek optimal channel

performance. An investigation of when tight electro-

nic integration is appropriate and when it is not can

generate strategy prescriptions of significant value to

B2B firms in determining their best level of

deployment of electronic integration appropriate for

their specific inter-organizational relationships.

The first step in our research endeavor was to

develop a means of accurately measuring the degree ofelectronic integration between the trading partners in a

B2B relationship. In the past, several researchers have

called for an assessment of the degree of electronic

integration between two firms [15]. EI may be defined

as the integration of business processes of two or

more independent organizations through the exploita-

tion of the capabilities of computers and communica-

tion technologies [32]. So, we conceptualized EI as a

multi-dimensional construct mainly constituting busi-

ness integration and process integration, and focused

our work on developing a measure for a crucial

component of EI: electronic information transfer

(EIT) which serves as the infrastructure for the inter-

organizational business and process integration. We

define electronic information transfer as a regulated

flow of information between trading partners via

electronic linkages.

2. Existing measures related to electronic

integration

Prior research in EI has used the adoption ofelectronic data interchange (EDI) as a surrogate

measure for EI. Table 1 summarizes the various

definitions of EI and its operationalization in prior

research.

The work asserts that there is a strong- and

mistaken-tendency to equate EI with EDI in existing

research. EI is a broader construct that essentially

subsumes EDI. It caters to two types of integration:

technological interconnectivity issues and business

process interdependence issues [31]. Efforts to

enhance technological interconnectivity have madesignificant strides during the past decade. As EDI

emphasizes technological interconnectivity between

the trading partners, existing measures of it encompass

mainly technological aspects, such as volume,

diversity, breadth, and depth of EDI usage [22].

While such measures serve a definite purpose,

attention to creating interdependent business pro-

cesses is also necessary to allow an organization to

develop a seamless and interoperable technical plat-

form.

K.K. Kim, N.S. Umanath/ Information & Management 42 (2005) 813828814


3/16

EI often involves interaction with the resources

controlled by the partner firm and requires substantial

mutual adaptation. Thus, any meaningful measure-

ment instrument should capture the degree ofelectronic integration in various dimensions of the

business processes between trading partners. In an

explicitly cooperative relationship, decisions are

coordinated between economic activities through

processes and information that are specific to the

exchange. Thus, we view electronic information

transfer as a significant component of EI.

Several researchers have suggested that EI should

not be viewed as a dichotomous variable but rather as

a step in the integration. A simple EDI link that

automates merely the transmission of orders from a

buyer to a seller does not create any electronic

integration; [8] proposed a scheme to classify the

degree of electronic integration between two firms.

For instance, a basic level of electronic integration

may occur when the linked firms develop product

code translation tables so that employees at the

participating firms can place/receive orders using

internal product codes. A higher level of electronic

integration may be possible when the buyers

computer determines a need for a product, based

on preset reorder levels, and automatically transmits

an order to the suppliers order entry system withouthuman intervention. At the highest level of electronic

integration, the firms can create close electronic

coupling among the processes that create or use the

information being exchanged. Researchers have

called for a way to assess the degree of such

integration instead of measuring just the presence or

absence of EI between firms. The fundamental

activity underlying this phenomenon is information

transfer.

While exchanges between trading partners may

entail movement of goods and services, integration ofbusiness processes intrinsic to these exchanges almost

always involves transfer of information between the

trading partners about the products and processes.

While some products and processes are information-

intensive, all have an informational component.

Business process integration is possible only through

transfer of appropriate informationelectronic or

otherwise. Here, we view information transfer through

the lens of decision and operational integration in an

inter-organizational relationship. On the basis of a

K.K. Kim, N.S. Umanath/ Information & Management 42 (2005) 813828 815

Table1

VariousdefinitionsandmeasuresofE

I

Authors

DefinitionofEI

Measures

VenkatramanandZaheer[32]

Integrationofbusinessprocessesoftwoo

rmore

independentorganizationsthroughtheexploitationof

thecapabilitiesofcomputer&communic

ationtechnologies

Adichotomousscalemeasuringwhetheranindependent

insuranceag

entiselectronicallyinterfacedwithaded

icatedsystem

BergeronandRaymond[6]

ThelevelofdiffusionofEDIoutsidethe

organization.

Thelevelofexternalintegrationillustrate

svarious

typesoftradingpartnerswithwhichtheo

rganization

transactsbusinessthroughEDI

Thelevelof

externalintegrationnumberedthediffere

nt

typesoftrad

ingpartnerswithwhichtheorganization

transactsbusinessthroughEDI

ZaheerandVenkatraman[36]

Aspecificformofverticalquasi-integrationachieved

throughthedeploymentofproprietaryinf

ormationsystems

betweenrelevantactorsinadjacentstages

ofthevaluechain

Thepercentageofbusinessdirectedtotheinterfaced

carrier

throughthe

proprietaryelectronicchannel

BensaouandVenkatraman[4]

TheuseofITfunctionalityforfacilitatinginter-organizational

coordination,especiallythenatureandscopeoftheelectronic

linkagesbetweentwomembers

Dichotomou

sitemsmeasuringwhetherdataareexcha

ngedin

electronicfo

rmwiththesupplierin[specific]functions

MassettiandZmud[22]

Thedegreeofelectronicconsolidationthathasbeenestablished

betweenthebusinessprocessesoftwoor

moretradingpartners

Threelevels

of[externalintegration]aredefined:(1)

file-to-fileconnections,(2)application-to-application

exchanges,and(3)coupledworkenvironments


4/16

thesis by [10], we further argue that the electronic

aspect of EIT not only invigorates decision and

operation integration (DOI) but also mitigates the side

effects of decision and operation integration (transac-tion risk)essentially a double-edged (positive)

sword. Therefore, assessment of EIT is an important

issue pertaining to electronic integration in an inter-

organizational relationship.

3. Dimensions of electronic information transfer

The first step in developing a measure is to form a

general conceptual understanding of the underlying

construct. A synthesis of existing literature was used to

aid in the conceptual development. Eventually specific

dimensions of the construct were derived and

empirically validated.

3.1. Theoretical foundation

Transaction cost economics (TCE) asserts that

specialized firms interacting in a market lead to

production economies. However, this entails transac-

tion costs. Risk of opportunistic behavior is a cost of

conducting market transactions [35]; however, the

informational aspect of the transaction cost in a marketis the coordination costs. Hierarchies, on the other

hand, sacrifice production economies in order to

reduce transaction costs. According to TCE, the most

efficient governance of such interactions is achieved

by balancing production economies and transaction

costs.

Cooperative relationship has been considered to be

an extension to the usual dichotomous view of markets

and hierarchies of TCE analyses. A move to the

middle between the polarities of markets and

hierarchies is being proposed as opposed to a moveto the market hypothesis. In fact, [21] states that the

emergence of tightly coupled electronic hierarchies

facilitated by IT tends to support a move to the

middle argument better than a move [exclusively]

to the market position. Breaking down the costs of

cooperation as costs of coordination and costs of

transaction risk (a synthesis of views), it can be argued

that IT can reduce both these costs simultaneously.

EIT is the principal conduit through which IT is

harnessed to accomplish this (Fig. 1).

3.2. Coordination of decision and operation

integration (DOI)

Costs of coordination are often explicated in termsof coordinating decisions and operations among

economic activities that occur between partnering

firms. EIT can facilitate coordination of decisions and

operations by reducing the costs of accumulating,

communicating, and processing information [2]. Thus,

decision and operation integration is proposed as a

dimension of EIT.

In order to identify the decision and operational

activities that occur when trading partners exchange

goods, five basic activities in the purchasing cycle

were identified from the literature: order products,

receive/store products, quality assurance, vendor

invoices, and payments. All these can be facilitated

by EIT. Table 2 presents the features of EIT able to

facilitate coordination in inter-organizational business

operations and decisions.

3.3. Management of transaction risk

Transaction risk is involves opportunistic behavior

by a trading partner, leading to uncertainty surround-

ing the level and division of the benefits from the

increased integration of decisions and operations. Thefocus of TCE has historically been on risks generated

by transaction specific capital [34]. However, efforts

to build a greater degree of integration of decisions

and operational activities while reducing costs of

coordination, exacerbate transaction risk. Information

asymmetries and loss of resource control have also

been identified as possible sources of transaction risk

that result from greater integration of decisions and

operations.

3.3.1. Mutual investment in relationship-specificassets (MIRSA)

Relationship-specific assets are investments that

have little or no value other than in the specific

interaction in which it occurs. Reciprocal investments

in the inter-organizational relationship by the parti-

cipating partners offer an alternative to vertical

integration as a means of safeguarding transaction-

specific assets. One form of reciprocity occurs when

each party makes a nonredeployable transaction-

specific investment, signaling a commitment to the



5/16

continuation of the relationship; this will lower the

likelihood of opportunistic behavior. Regarding IT

assets, transaction risk is reduced to the extent that the

two parties invest in specific resources to leverage IT

capabilities and restructure the nature of the relation-

ship. In the context of electronic information transfer,

these refer to investments in the underlying hardware,

software, and communication systems as well as in

providing user training and support [36]. Specialized

investment in inter-organizational IS (IOIS) for a

buyer-seller dyad seems to facilitate joint programs

and activities and provides a more positive transaction

climate in the dyad as long as the balance of power is

not altered significantly [23]. Thus, cooperation rather


Fig. 1. Dimensions of electronic information transfer.


6/16

than competition may make more sense for investment

decisions [9]. Kydd and Jones offered specific

guidelines for creating condusive conditions for

successful implementation of shared information

systems [19]. Relationship-specific skills have also

been documented as a significant investment [29].

While a trading partners investment in relation-

ship-specific assets signals commitment to the

relationship, IT assets possess the unique property

of softening the transaction risk of the trading partners,

viz., switching costs or redeployability. In the current

business environment, firms often may not make

separate investments in computer hardware and

telecommunication equipment for a specific trading

partnership (see items 6 and 7 in Table 3) because the

firms IT infrastrcuture will invariably include such

equipment. Even if a relationship-specific investment

is made by a firm because it is an initial investment,

unless it pertains to specialized equipment, the

durability of the investment as a nonredeployableasset of the specific relationship is questionable.

However, the uniqueness of IT assets in lending some

degrees of freedom in redeployment softens the

transaction risk for both parties. Nonetheless, some or

all such IT investments may also possess a degree of

relationship-specificity in an inter-organizational

relationship. Additionally, the characteristics of

modern software, such as modularity and replicability,

open standards, and intuitive interefaces, render it at

least partially flexible.

Thus, IT investments in EIT affects transaction risk

of both partners in the inter-organizational relation-

ship, but in different ways.

3.3.2. Reduce information asymmetries

Information asymmetry occurs when either partner

in the relationship has privy to information specific to

the relationship that the other do not. Such asymme-

tries increase transaction risk when integrating

decisions and operation in the trading partnership.

A view based on markets and hierarchies emphasizes

the informational aspects of information asymmetry,


Table 2

EIT in coordination of decision and operation integration

Purchasing Cycle Facilitating EIT features

Order products 1. Use common product codes2. Place purchase order electronically

Quality assurance 3. Trace product failures back to the

offending component(s) electronically

Vendor invoices 4. Receive suppliers invoices electronically

Payment 5. Make electronic payments

Table 3

Role of EIT in managing transaction risk

Management of transaction risk using EIT Facilitating EIT features

Mutual investment in relationship-specific assets 6. Reciprocal investment in communication network

7. Reciprocal investment in hardware

8. Provide IT training

9. Provide technical support

10. Provide customized support

11. Help supplier develop their own software

Information Sharing 12. Exchange production (or sales) data with the supplier electronically

13. Use the data, electronically transferred from the trading partner,

in business decisions14. Vendor-managed inventory (VMI)

15. Share promotion plans with the trading partner electronically

Monitoring and control 16. Access the suppliers shipping/delivery schedule

17. Access the suppliers production schedule

18. Access the suppliers inventory levels of finished products

19. Access the suppliers inventory levels of raw materials

20. Provide performance feedback

21. Search for alternative suppliers for the products

22. Monitor the order status

23. Monitor the suppliers production capacities

24. Monitor the quality of the products being produced


7/16

while economic theories (e.g., TCE, agency theory)

focus on the opportunistic aspect of the phenomenon.

3.3.2.1. Information sharing. Information asymme-try can be reduced when partners freely share

information relevant to the relationship, e.g., electro-

nic exchange of production/sales data, sharing

promotion plans, vendor-managed inventory (VMI),

etc. EIT, by virtue of its inherent capabilities, is a

powerful agent in facilitating information sharing (IS).

Characteristics were derived from this and are

included in Table 3.

3.3.2.2. Monitoring and control. Information asym-

metry can prompt opportunistic behavior (e.g.,

shirking) by trading partners. The difficulty in

measuring the specific contribution of inputs in

generating outputs creates an opportunity for perfor-

mance shirking by the supplier. If the buyer lacks the

ability to monitor the status of the suppliers

production process (the production capacities, inven-

tory levels, shipping/delivery schedule, quality of the

products being produced, etc.), the supplier can reduce

its effort level. Opportunism can also occur in the

presence of a limited number of suppliers, since this

increases the dependency of a buyer on a specific

supplier. The consequent difficulties in performancemonitoring entail transaction risk. While the tradi-

tional economic theories tend to ascribe opportunistic

behavior to the supplier, the transaction risk in a

cooperative relationship can be bidirectional. Once

again, IT is capable of monitoring and controling the

transactions as a byproduct of normal operations in the

EIT environment. Features that pertain to monitoring

and control (MAC) were derived from this rationale

and are also included.

4. Research method

Based on these constructs, an instrument to

measure EIT was developed. The initial questionnaire

contained 24 items representing the four conceptua-

lized constructs: decision and operation integration,

mutual investment in relationship-specific assets

(MIRSA), information sharing, and monitoring and

control. A pilot test was performed to fine-tune the

instrument and to establish face validity. A compre-

hensive instrument validation procedure followed to

validate the instrument empirically.

4.1. Sample

Data were collected from two corporations in the

U.S., each a major player in its industry segmentone

is in the retail grocery industry and the other is a

machine tool manufacturer.

A set of interviews were initially conducted with

senior managers responsible for purchasing since this

is the boundary spanning function considered to be

most critical in supplier relations. The interviews:

(1) provided a preliminary corroboration of the

applicability and appropriateness of the EIT

construct; and

(2) ensured that we had an adequate base to sample

the relationships covering the vast array of

suppliers and products.

Through these interviews, we were directed to appr-

opriate buyers.

The unit of analysis in our study pertains to the

inter-organizational relationship-more specifically p-

airs of product categories and suppliers. A product

category refers to a group of products with similar

characteristics and it already existed in the organiza-tions sampled. Typically most buyers were responsible

for multiple product categories and often used multi-

ple suppliers for each category. Each buyer was asked

to answer a questionnaire that measured the dimen-

sions of EIT. A data point in this study consisted of a

dyad made up of a product category and a supplier.

Altogether, 39 buyers from eight offices throughout

the U.S. participated in this study and all handled

multiple product categories. The final sample had 160

data points, i.e., 160 product category-supplier dyads.

4.2. Pilot test

Preliminary testing involved structured interviews

and a pretest carried out using responses from six

buyers. The multiple structured interviews conducted

with managers of purchasing divisions were intended

to test the face validity of the instrument. Based on the

feedback from these interviews, a few questions were

rephrased to reflect industry specific situations better.

Then, the six buyers were asked to indicate their



8/16

responses to each of the 24 questions, which had a

seven-point Likert scale with values from Not at all to

Very much so. The pretest yielded 32 data points on the

{product categorysupplier} dyad from the sixbuyers. Based on this pretest, some questions were

modified to enhance clarity.

4.3. Exploratory factor analysis

Principal components analysis provided prelimin-

ary verification of the four constructs. Table 4 provides

the standardized parameter estimates (factor loadings)

of the items over the four dimensions. We used the rule

of thumb of 0.50 as the cut off value for factor

loadings. These four factors accounted for 60.4% of

the variance in the data set.Four out of the five items theoretically developed to

reflect the construct of decision and operation

integration loaded on a single factor. Item 3 did not

load on any factor and was discarded from further

consideration. All six items expected to measure the

mutual investment in relationship-specific assets

loaded on a single unique factor. Only two out of

the four items expected to represent the information


Table 4

Factor loadings of the EIT measure

Defined items Derived

Factor 1 Factor 2 Factor 3 Factor 4

Decision and operation integration (DOI)

Item 1 0.155 0.185 0.600* 0.071

Item 2 0.126 0.150 0.640* 0.073

Item 3 0.202 0.092 0.021 0.221

Item 4 0.078 0.074 0.739* 0.136

Item 5 0.336 0.039 0.516* 0.175

Mutual investment in relationship-specific assets (MIRSA)

Item 6 0.767* 0.004 0.108 0.185

Item 7 0.810* 0.110 0.030 0.070

Item 8 0.779* 0.054 0.082 0.359

Item 9 0.832* 0.037 0.163 0.224

Item 10 0.744* 0.125 0.146 0.341

Item 11 0.668* 0.023 0.089 0.337

Information sharing (IS)

Item 12 0.084 0.105 0.128 0.679*

Item 13 0.341 0.260 0.312 0.017

Item 14 0.238 0.016 0.066 0.555*

Item 15 0.593* 0.226 0.030 0.211

Monitoring and control (MAC)

Item 16 0.047 0.746* 0.013 0.124

Item 17 0.038 0.327 0.330 0.291

Item 18

0.004 0.779*

0.187 0.077

Item 19 0.091 0.812* 0.120 0.068

Item 20 0.389 0.510* 0.131 0.220

Item 21 0.238 0.082 0.262 0.408

Item 22 0.115 0.834* 0.138 0.061

Item 23 0.027 0.520* 0.366 0.087

Item 24 0.077 0.437 0.425 0.038

Eigenvalue 5.5 4.26 2.63 1.51

Percentage of variance explained 25.83 20.38 8.05 6.18

Cronbachs Alpha 0.93 0.85 0.78 0.67

Note: Extraction method = principal component analysis; rotation method = varimax rotation. Cutoff eigenvalue: 1.0.* Indicates the highest loadings greater than 0.5.


9/16

sharing construct loaded on a single factor. One (item

13) did not load on any of the four factors and the other

(item 15) exhibited a significant loading on the

MIRSA dimension. We discarded this item asspurious, perhaps this was a specific result for this

data set, since we could not find any reason that

explained such behavior.

Finally, six out of the nine items defining the

construct of monitoring and control loaded together.

The other three (items 17, 21, and 24) did not exhibit

any relationship with any of the other factors and were

removed from further analysis. Thus, the final

instrument consisted of a four-factor structure with

18 scale items. A copy of this instrument appears in

the Appendix A.

4.4. Instrument validation

A comprehensive instrument validation procedure

suggested by [11] was then followed. Structural

equation modeling (SEM) was used to test construct

validity, which may be defined as the correspondence

between a construct and the operational procedure to

measure or manipulate that construct [25]. Confirma-

tory factor analysis (CFA) is considered well suited to

investigate constructs can be distinguished from one

another [20]. Thus, the CFA procedure was used todetermine whether the EIT items in the instrument

adequately represented the hypothesized dimensions.

The analyses used are summarized in Table 5. The

structural equation modeling tool, EQS, was used to

conduct the analysis.

Sample size is an important consideration in

determining the appropriateness of estimating a

CFA model [16]. Based on the sample size to

parameter estimate ratio of 5 suggested in [5] as the

minimum sample size, we concluded that our sample

size was adequate.

4.5. Results

The pattern of factor condensation is the first

indicator of the degree of convergence. The factorloadings of the measurement items depicting the

theoretically derived constructs strongly ratify con-

vergent validity of the instrument. In CFA, it is

essential first to examine the overall fit of the model. If

a model does not fit the data, the hypothesis that the

model accurately represents the data is rejected.

However, because the underlying population distribu-

tions for these statistics are unknown and because

there is no clear consensus in what constitutes an

appropriate fit, assessment of overall model fit is still a

subjective process [30].

Pursuant to the conceptual development, we

formulated the measurement model as a four-factor

structure: DOI, MIRSA, IS, and MAC. The fit

statistics of this model were: comparative fit

index = 0.80, BentlerBonnet normed fit index = 0.75,

nonnormed fit index = 0.77. While rule of thumb value

of 0.90 for the indices has been suggested [24], we

decided to retain the model because the pattern of

factor condensation was robust (see Table 4). Further

the fit indices are reasonably close to the rule of thumb

value. These fit statistics and parameter estimates

provided further evidence of the convergent validity ofthe items measuring the four dimensions of EIT.

Discriminant validity was assessed using two

different methods. The first was to see whether the

scale items were capable of differentiating the

multiple dimensions of EI (Table 5, item 3). This

was achieved by comparing the fit of the hypothesized

four-factor model to a model with a single EIT

construct. The discriminant validity would be estab-

lished when the fit of a single factor model was

significantly different from the hypothesized model. In

such a case, a single factor model would beinsufficient to capture the multidimensionality of

the EIT. Our data indicated that the fit of the single

factor model was significantly different from the

hypothesized four-factor model (p < 0.001). This

result provided support for the discriminant validity

of the proposed dimensions of EIT.

It can also be assessed by examining whether

measures of purportedly different constructs display

differential patterns of correlations with other external

construct(s). We used EDI as the external construct


Table 5

Summary of analysis

Analysis Research objective

Significance of factor loadings Convergent validity

Fit of the four dimensional model Convergent validity

Alternative (single factor)

measurement models

Discriminant validity

Differential correlations between EIT

dimensions and EDI

Discriminant validity


10/16

(Table 5, item 4) and considered EDI as appropriate to

test the discriminant validity of the EIT instrument,

since it has often been used as the surrogate for EI in

prior research. If EDI relates similarly with all the fourdimensions of EIT, it calls into question the utility of

distinguishing among the four dimensions.

The test procedure entailed estimating two models

and comparing them. One allowed the correlations

between EDI and each of the four EIT dimensions to

be freely estimated; the other constrained the

correlations between EDI and the dimensions of

EIT to be equal. If the fits of these two models were not

significantly different, then EDI would be construed to

have a similar association with all the dimensions,

essentially refuting the utility of distinguishing among

the four dimensions. We developed a four-item scale

to measure EDI. It pertained to EDI tasks in document

exchanges in EIT. The reliability of the EDI scale, as

indicated by Cronbachs coefficient alpha (0.86), was

quite robust.

The results showed that constraining EDI to have

the same relationship with each dimension resulted in

a significant difference from those of the unrestricted

model (p < 0.001). This indicated that the dimensions

of EIT indeed exhibited different relationships from

EDI, thus, supporting discriminant validity of the

derived dimensions.

5. Discussion

While several authors have argued that EDI is not

EI and have called for a comprehensive measure that

assesses the degree of EI between firms, none

developed an instrument to measure EI per se. A

major contribution of our research was, thus, the

development of an instrument to measure EIT, the

information transfer/flow infrastructure for EI. Usingthe triangulation of multiple theoretical perspectives

as the foundation, we culled electronic information

transfer, the underlying infrastructure for EI based on

coordination of decisions and operation and transac-

tion risk. We then identified relationship-specific

assets, information sharing, and monitoring and

control as the dimensions of transaction risk.

The four dimensions of the EIT construct were

empirically proved in our study. Convergent validity

of the scale was indicated by the factor condensation

pattern matching the conceptualized dimensions.

Also, results from the CFA procedure further ratified

convergent validity. In terms of discriminant validity,

the hypothesized four-factor model represented thedata significantly better than the aggregated single

factor model. The four dimensions also exhibited

differential patterns of relationships from those of

EDI.

Although our sample size was adequate for the

stability of statistical analyses, the data were collected

from only two companies. The interpretation of the

results is therefore subject to the constraints of

organizational and business characteristics of these

two organizations. Since the task context in this

research is procurement, we sought generalizability

over product categories. The two firms in our study are

major players in their respective industry segment in

the U.S. Therefore, we were able to collect rich data on

product categories.

6. Implications for practice

An example of research incorporating EIT as a

dependent variable is to identify and evaluate the

antecedents of EIT. There has been a stream of

research based on the premise that EI corresponds to ashift away from the market-based exchange toward

more bilateral and cooperative governance. In order to

reduce transaction costs, most retail chains have

traditionally used captive distributors, vertically

integrating to combine retail stores and distribution,

while some manufacturers have vertically integrated

into distribution by providing direct store delivery for

their products [1]. However, recent information

interchange using electronic linkages between firms

has been transforming the nature of the B2B

relationships [26]. Antecedents of EIT cited in theliterature include organizational trust [13], governance

structure of the relationship [3], environmental

uncertainty, and product complexity. This stream of

research can benefit from the instrument developed in

our study.

Since the two companies in the sample belong to

different parts of a supply chain, that is, the retail

grocery company in the downstream chain and the

machine tool manufacturing company in the upstream,

different inter-organizational information system



11/16

(IOIS) capabilities may be needed to realize appro-

priate supply chain coordination [7]. Since an IOIS

provides the fundamental infrastructure for linking

members of a supply chain, proper alignment betweenthe developmental stage of the supply chain and the

integrating capabilities of IOIS is critical [28,33].

From a management perspective, EIT, if properly

deployed, offers the unusual capability of providing

significant benefits to members on both sides of the

partnership. The firm can not only deliver cost savings

to its trading partners, but also enhance the service it

provides while reducing its own costs of operation.

When the firm can reduce its trading partners

transaction costs while simultaneously reducing its

own, the entire supply channel can perform more

effectively. The linchpin in this complex set of

relationships is the extent of EIT between the

participating firms. This research shows why man-

agers ought to invest time and effort in assessing

current and future information exchange with their

supply chain partners and align their IOIS capabilities

accordingly.

Also, electronic linkage between a firm and its

suppliers has opened up new avenues for business

integration between the participants. As a conse-

quence, the very nature of the relationship between the

firm and its suppliers is drastically affected, necessi-

tating a reassessment of the firms business strategy.Managers have been limited to assessing EI via

through technological surrogate, EDI. The instrument

developed here provides practicing managers addi-

tional capabilities in the assessment of EI. Managers

can now validate current industry practices and

generate specific recommendations regarding the

information flow aspect of their supply chain logistics.

The instrument also enables managers to isolate and

examine decision and operation integration, invest-

ments in relation-specific assets, information sharing

and/or monitoring and control aspects of information

transfer/flow infrastructure of the firm and its partners

and strengthens the weaker links.

Acknowledgements

We thank the three anonymous reviewers for their

comments. This research was partially funded by

Yonsei University, Korea.



12/16


Appendix A

A.1. EI QuestionnairePart I

Please pick two major product categories you manage that add most value to your company.

For each major product category, list the major suppliers. If the type/level of electronic integration (EI) with any

two suppliers is about the same, you may use only one of them.

Level of EI is a continuum from minimal integration to complete integration. For example, simply sending

purchase orders electronically can be considered as minimal integration. Other forms of EI include vendor managed

inventory (VMI) (you supply data elements to vendors, and you create purchase orders), VMI-Advanced (you

supply data elements to a supplier and the supplier creates purchase orders), VMI-Advanced plus electronic

payments, etc.

Where possible, choose suppliers for a product category with differing degrees of EI.


13/16


A.2. EI QuestionnairePart II {SupplierProduct

Category Dyad}


14/16



15/16

References

[1] L. Applegate, W. McFarlan, J. McKenney, Corporate Informa-tion Systems Management: Text and Cases, fifth ed., Irwin/

McGraw-Hill, New York, 1999.

[2] Y. Bakos, Reducing buyer search costs: implications for

electronic marketplaces, Management Science 43(12),

1997.

[3] M. Bensaou, Inter-organizational cooperation: the role of

information technology, An empirical comparison of U.S.

and Japanese supplier relations, Information Systems Research

8(2), 1997, pp. 107124.

[4] M. Bensaou, N. Venkatraman, Configurations of inter-organi-

zational relationships: a comparison between U.S. and Japa-

nese automakers, Management Science 41(9), 1995, pp. 1471

1492.

[5] P. Bentler, EQS Structural Equations Program Manual, Multi-variate Software, Encino, CA, 1995.

[6] F. Bergeron, L. Raymond, The advantage of electronic data

interchange, Database 23(4), 1992, pp. 1931.

[7] D. Bowersox, D. Closs, B. Cooper, Supply Chain Logistics

Management, McGraw-Hill, 2002.

[8] V. Choudhury, Strategic choices in the development of inter-

organizational information systems, Information Systems

Research 8(1), 1997, pp. 124.

[9] E. Clemons, M. Knez, Competition and cooperation in infor-

mation systems innovation, Information and Management (15)

1988, pp. 2535.

[10] E. Clemons, M. Row, Information technology and industrial

cooperation: the changing economics of coordination and

ownership, Journal of Management Information Systems

9(2), 1992, pp. 928.

[11] S. Currall, T. Judge, Measuring trust between organizational

boundary role persons, Organizational Behavior and Human

Decision Processes 64(2), 1995, pp. 151170.

[12] J. Galbraith, Designing Complex Organizations, Addison-

Wesley, Reading, MA, 1973.

[13] P. Hart, C. Saunders, Power and trust: critical factors in the

adoption and use of electronic data interchange, Organization

Science 8(1), 1997, pp. 2342.

[14] I. Hong, A new framework for inter-organizational systems

based on the linkage of participants roles, Information and

Management (39) 2002, pp. 261270.

[15] C.L. Iacovou, I. Benbasat, A.S. Dexter, Electronic data inter-

change and small organizations: adoption and impact oftechnology, MIS Quarterly 19(4), 1995, pp. 465485.

[16] J. Idaszak, W. Bottom, F. Drasgow, A test of the measurement

equivalenceof therevised job diagnostic survey: past problems

and current solutions, Journal of Applied Psychology 73(4),

1988, pp. 647656.

[17] S. Kaplan, M. Sawhney, E-Hubs: the new B2B marketplaces,

Harvard Business Review 2000, pp. 98103.

[18] K. Kim, N.S. Umanath, An empirical investigation of electro-

nic integration in supply channel relationship, in: Proceedings

of the 20th International Conference on Information Systems,

Charlotte, North Carolina, 1999, pp. 546551.


A.3. EDI Questionnaire


16/16

[19] C.T. Kydd, L.H. Jones, Corporate productivity and shared

information technologies, Information and Management (17)

1988, pp. 277282.

[20] J.S. Long, Covariance Structure Models: An Introduction to

LISREL, Sage, Beverly Hills, 1983.[21] T. Malone, J. Yates, R. Benjamin, Electronic markets and

electronic hierarchies, Communications of the ACM 30(6),

1987, pp. 484497.

[22] B. Massetti, R. Zmud, Measuring the extent of EDI usage in

complex organizations: strategies and illustrative examples,

MIS Quarterly 20(3), 1996, pp. 331347.

[23] S.R. Nidumolu, Interorganizational information systems and

the structure and climate of seller-buyer relationships, Infor-

mation and Management (28) 1995, pp. 277282.

[24] R. Nidumolu, G.P. Knotts, The effects of customizability and

reusability on the perceived performance of software firms,

MIS Quarterly 22(2), 1998, pp. 105137.

[25] J. Nunnally, I. Bernstein, Psychometric Theory, third ed.,

McGraw-Hill, NY, 1994.

[26] R. OCallaghan, P. Kaufmann, B. Konsynski, Adoption corre-

lates and share effects of electronic data interchange systems in

marketing channels, Journal of Marketing 56(2), 1992, pp. 45

56.

[27] C. Saunders, S. Miranda, Information acquisition in group

decision making, Information and Management 34(2), 1998,

pp. 5574.

[28] R. Shah, S. Goldstein, P. Ward, Aligning supply chain manage-

ment characteristics and interorganizational information sys-

tem types: an exploratory study, IEEE Transactions on

Engineering Management 49(3), 2002, pp. 282292.

[29] M. Subramani, How do suppliers benefit from information

technology use in supply chain relationships? MIS Quarterly28(1), 2004, pp. 4573.

[30] J. Tanaka, How big is enough? Sample size and goodness offit

in structural equation models with latent variables Child

Development (58) 1987, pp. 134146.

[31] N. Venkatraman, IT-Enabled business transformation: from

automation to business scope redefinition, Sloan Management

Review (35) 1994, pp. 7387.

[32] N. Venkatraman, A. Zaheer,Electronic integration and strategic

advantage: a quasi-experimental study in the insurance indus-

try, Information Systems Research 1(4), 1990, pp. 377393.

[33] P. Weill, M. Vitale, What IT infrastructure capabilities are

needed to implement e-business models? MIS Quarterly

Executive 1(1), 2002, pp. 1734.

[34] O. Williamson, Comparative economic organization: the ana-

lysis of discrete structural alternatives, Administrative ScienceQuarterly 36(2), 1991, pp. 269296.

[35] O. Williamson, Transaction cost economics, in: R. Schmalen-

see, R. Willig (Eds.), Handbook of Industrial Organization,

Amsterdam, North-Holland, 1989, pp. 135182.

[36] A. Zaheer, N. Venkatraman, Determinants of electronic inte-

gration in the insurance industry: an empirical test, Manage-

ment Science 40(5), 1994, pp. 549566.

Kyung Kyu Kim is a Professor of Infor-

mation Systems at Yonsei University,

Korea. His research interests are in the

areas of B2B e-commerce, supply chain

management, and trust in e-commerce

adoption. He has been a faculty memberat the University of Cincinnati, Pennsyl-

vania State University, and Inha Univer-

sity in Korea. He has published his

research works in Accounting Review,

MIS Quarterly, Decision Sciences, Journal of MIS, Information

and Management, Database and Journal of Information Systems.

Narayan S. Umanath is Professor of

Information Systems at the University

of Cincinnati, Ohio. Entering academia

after fifteen years of technical and man-

agerial experience in software develop-

ment, Umanath received his Ph.D. in

Business Administration from the Uni-versity of Houston in 1987. His under-

graduate and graduate educations are in

mechanical engineering and industrial

engineering respectively. His current research interests include

electronic integration in supply chain relationships, organizational

issues pertaining to Information Systems, and data modeling & data

warehousing. His research publications have appeared in Commu-

nications of the ACM, Decision Sciences, Information & Manage-

ment, Information Resources Management Journal, Journal of MIS,

Journal of Managerial Issues and Management Science.


Documents

B2B EC Procurement Info Transfer Analysis Measurement