The current issue and full text archive of this journal is available at

www.emeraldinsight.com/2040-8269.htm

MRR
37,2

Effect of business characteristics

and ERP implementation
on business outcomes

186

An exploratory study of Korean

manufacturing firms
Pairin Katerattanakul
Department of Business Information Systems, Western Michigan University,
Kalamazoo, Michigan, USA

James J. Lee
Department of Management, Seattle University, Seattle,
Washington, USA, and

Soongoo Hong
Department of Management Information Systems,
Dong-A University, Pusan, Korea
Abstract
Purpose This study is an exploratory study aiming to explore whether different groups of
manufacturing firms with similar business characteristics and enterprise resource planning (ERP)
implementation approaches would experience different business outcomes from ERP implementation.
The paper aims to discuss these issues.
Design/methodology/approach Cluster analysis with data collected from 256 Korean
manufacturing firms was employed to identify groups of manufacturing companies having similar
business characteristics and adopting similar ERP implementation approaches. Then, the differences in
business outcomes from implementing ERP systems among these groups of companies were examined.
Findings Company size and production approaches are useful variables for grouping
manufacturing firms into clusters of companies with similar characteristics. Additionally, large
manufacturing firms with make-to-order production approach have significantly higher perceived
benefits from implementing ERP systems regarding external coordination and competitive impact
than other firms do.
Research limitations/implications This study was conducted in only one industry of one
country and used the data collected by self-reporting instrument. Thus, further studies conducted in
other industries and/or other countries and using more objective measures would allow more
generalizability of the findings of this study. It would also be interesting to investigate the effects of
the logistics practices adopted by small manufacturing firms even though these practices may be more
suitable for large manufacturing firms.
Originality/value This study contributes to the literatures on benefits obtained from
implementing ERP systems as none of the previous studies has focused on the relationship among
business characteristics, ERP implementation approaches, and business outcomes from ERP
implementation.
Management Research Review
Vol. 37 No. 2, 2014
pp. 186-206
q Emerald Group Publishing Limited
2040-8269
DOI 10.1108/MRR-10-2012-0218

Keywords Cluster analysis, Organizational performance, ERP, Manufacturing company

Paper type Research paper

Introduction
Enterprise resource planning (ERP) systems have been touted to streamline
organizational functions and processes by integrating enterprise-wide data and
business processes. However, ERP implementation is risky and requires a substantial
amount of resources (Cliffe, 1999). Thus, much academic research has been conducted on
different ERP implementation issues; for example, interactions between contingency,
organizational IT factors, and ERP success (Ifinedo and Nahar, 2009), expectation and
reality in ERP implementation (Helo et al., 2008), perceptions of the ERP system
implementation project (Maguire et al., 2010), impact of organizational support on ERP
implementation (Lee et al., 2010), critical success factors of ERP implementation (Dezdar
and Sulaiman, 2009), and difficulties experienced during ERP adoption (Soja and
Paliwoda-Pekosz, 2009). Additionally, there are a few studies focusing on the benefits
obtained from ERP systems, including effects of ERP systems on organizational
performance (Velcu, 2007), process efficiency and profitability of ERP (Huang et al.,
2009), categories of ERP benefits (Gattiker and Goodhue, 2005), and competitive
advantages of using ERP systems (Beard and Sumner, 2004; Lengnick-Hall et al., 2004).
Factors related to organization and ERP projects are critical for ERP implementation
success (Dezdar and Sulaiman, 2009). However, there are no previous studies focusing
on the relationship among business characteristics, ERP implementation approaches,
and the business outcomes from ERP implementation. Thus, this study is an exploratory
study mainly aiming to explore whether different groups of manufacturing firms with
similar business characteristics and ERP implementation approaches would experience
different business outcomes from ERP implementation.
Exploratory research is appropriate for this study as there have been virtually no
previous findings about the effect of the combination between business characteristics
and ERP implementation approaches on the business outcomes from ERP
implementation. Exploratory studies are loosely structured studies that are suitable
when the areas of investigation are so new or so vague that researchers need to do
some explorations to learn something about the dilemma, to explore some important
variables, or to develop hypotheses or questions for future research (Cooper and
Schindler, 2000).
To accomplish the objective of this exploratory study, we employed cluster analysis
to identify groups of manufacturing firms having similar business characteristics and
adopting similar ERP implementation approaches; then, we examined the differences
in the business outcomes from implementing ERP systems among these different
groups of manufacturing firms.
The current study contributes to the existing literature in two ways. First, this
study identifies significant factors related to ERP implementation approaches and
business characteristics for grouping manufacturing firms. Second, this study
empirically tests whether different groups of manufacturing firms with similar
business characteristics and ERP implementation approaches would experience
different business outcomes from ERP implementation. The findings of this study can
serve as the starting point for future studies that investigate the relationship among
business characteristics, ERP implementation approaches, and business outcomes
from ERP implementation.

Effect of
business
characteristics
187

MRR
37,2

188

ERP implementation approaches, business characteristics, and business

outcomes
ERP implementation approaches
Some previous studies found the dilemma between software customization and using
standardized ERP packages (Helo et al., 2008). That is, it may be convenient for the
adopting company to just adopt the best practice business processes embedded in ERP
system; however, process re-configuration could be costly. On the other hand,
customization of ERP software to better fit the companys existing business processes
is also costly.
In this study, we used two terms to conceptualize these choices: process
re-configuration and software customization. Process re-configuration was defined as
the adoption of the best practice business processes embedded in ERP systems without
modifying the ERP software, but re-configuring the existing business processes
instead. On the other hand, software customization occurs when the adopting company
do not want to (or cannot) change its existing business processes, and instead modifies
ERP software to meet its business requirements.
Additionally, recent research suggested that several system factors could be critical
in determining the success of ERP rollout approach (Dezdar and Sulaiman, 2009).
These factors include number of sites and users, number of ERP modules implemented,
complexity of business processes, level of ERP software customization, and existence
of legacy systems. These factors can lead to varied ERP rollout approaches (Parr and
Shanks, 2000), including:
.
The big bang approach this approach has a single go-live date for all selected
ERP modules. It refers to a total effort to implement all selected ERP modules
together at once.
.
The mini big bang approach this approach has several go-live dates for
different subsets of ERP modules.
.
The phase implementation approach this approach involves incrementally
implementing ERP either module-by-module or site-by-site in a phased manner.
Finally, when implementing ERP systems, companies can employ different ERP
selection approaches including a single ERP package, best-of-breed from several ERP
packages, developing ERP systems in-house, or pursuing a hybrid approach that
includes in-house development and some specialized package functionality
(Katerattanakul et al., 2006).
Business characteristics
For manufacturing companies, the production approach is typically categorized into
two continuums: make-to-order (MTO) and make-to-stock (MTS) (Gupta and
Benjaafar, 2004). Under the MTO approach, a production order is released to the
manufacturing facility only after a purchase order has been received from a customer,
while under the MTS approach, products are manufactured in anticipation of future
orders and stored in the finished goods inventory (Youssef et al., 2004).
The MTO approach is good for customization and volume flexibilities (Yen and
Sheu, 2004); that is, when products are low in volume, but high in variety. While the
MTO approach eliminates finished goods inventory and reduces a firms exposure to
financial risk, it usually spells long customer lead times and large order backlogs

(Gupta and Benjaafar, 2004). On the other hand, when there are requests for high
production volume, long set-up times, stable production schedules, a relatively small
number of suppliers, and lower labor skills, it is better for manufacturing firms to
implement MTS approach to obtain immediate reactivity to external demands at the
cost of inventory holding expenses (Yen and Sheu, 2004; Youssef et al., 2004).
Another business characteristic investigated in this exploratory study is company
size due to it being a possible cause explaining why firms have different ERP
implementation experiences (Mabert et al., 2001). This study adopted the definition
provided by the Small & Medium Business Administration of Korea (eng.smba.go.kr).
This definition categorizes companies in Korea, based on the number of employees,
into small enterprises (# 50 employees), medium enterprises (. 50 and # 300
employees), and large enterprises (. 300 employees).
Business outcomes
Having implemented ERP systems, companies experience some improved performance
mainly from the information perspective; that is, information is more easily accessible
and the quality of information is also improved (Mabert et al., 2000; Olhager and
Selldin, 2003). Additionally, issues such as coordination with customers and suppliers
improve as well (Mabert et al., 2000; Olhager and Selldin, 2003). Thus, business
outcomes related to coordination with business partners and improved business
information were investigated in this study.
Furthermore, when this study was conducted, Koreas outward-looking policy was
aimed at responding to the era of global competition and catching opportunities created
by a rapidly changing and globalizing marketplace. Therefore, business outcomes
related to the companys ability to gain strategic advantages and to capture global
opportunities were also investigated in this study.
Research framework
In manufacturing firms, different production approaches need different manufacturing
infrastructure including resource allocation systems and distinctive communication
systems to align the downstream, midstream, and upstream processes (Prasad et al.,
2005). Additionally, without proper manufacturing infrastructure, a firm may not be
able to achieve the full competitive advantage that the ERP system can provide
(Hayes et al., 2005). Furthermore, there is a positive relationship between the success of
ERP system and the alignment of ERP implementation and business strategy (Dezdar
and Sulaiman, 2009; Velcu, 2007). The findings of these previous empirical studies
suggest that a relationship exists among business characteristics, ERP implementation
approaches, and business outcomes from implementing ERP system.
The organizational configuration theory posits that identifying groups different
from others but similar within the group allows better understanding of the
relationship between organizational characteristics and performance (Ketchen et al.,
1993). This organizational configuration is referred to as commonly occurring clusters
of attributes of organizational strategies, structures, and processes (Ketchen et al.,
1993, p. 1278).
The research framework of this study is based on the premise of the organizational
configuration theory and the findings of some previous studies suggesting that a
relationship exists among business characteristics, ERP implementation approaches,

Effect of
business
characteristics
189

MRR
37,2

190

and business outcomes from implementing ERP system. That is, the groups of
manufacturing firms with similar business characteristics and similar ERP
implementation approaches employed could have significant effects on business
outcomes from implementing ERP system.
Unfortunately, there have been virtually no previous studies investigating whether
different groups of companies with similar characteristics experience different
business outcomes from ERP implementation. We argue that the reason for the lack of
studies investigating this relationship may be from the fact that there is no clear
theoretical development or hypotheses relevant to this relationship. As exploratory
studies help to provide initial results for further theoretical development (Cooper and
Schindler, 2000), an exploratory research would be suitable for this study.
In this study, we defined the clusters of manufacturing companies based on two main
attributes: business characteristics (i.e. production approach, company size) and ERP
implementation approaches (i.e. process re-configuration vs software customization,
ERP rollout approach, ERP selection approach). Then, we investigated whether making
clusters of manufacturing companies that have similar business characteristics and
pursue similar ERP implementation approaches could explain the differential business
outcomes from implementing ERP systems. Figure 1 shows the research framework of
this study.
Research methodology
Data collection
The data set used in this study was from an extensive survey conducted on ERP
implementation in Korean manufacturing firms. This survey was administered
by Pollever Research Center, a leading market research company in Korea
(www.pollever.com). The questionnaire used in this survey was adapted from the
instruments used in the similar studies conducted on US and Swedish manufacturing
Manufacturing Firms

Outcomes from
implementing ERP

Combination of Business characteristics and ERP Implementation

Approaches
Business Characteristics:
Production approach
Company size

Figure 1.
Research framework

ERP Implementation Approaches:

Process re-configuration vs.
Software customization
ERP rollout approach
ERP selection approach

firms (Mabert et al., 2000; Olhager and Selldin, 2003). The original version of the
questionnaire was tested in a group interview with five Korean managers in charge of
ERP in their companies. Based on the feedback from this group interview, we modified
the original version of the questionnaire. The final version of the questionnaire was
written in English and included more than 60 questions. In this final version of the
questionnaire, we minimized the threats of common method bias by measuring each
of the constructs via several different questions. The questions measuring the
same construct were placed in random order in the questionnaire. Additionally,
some questions measuring the same construct employed reverse scale to create
counterbalance.
The Pollever Research Center sent e-mail invitations to the managers in charge of
ERP systems in Korean manufacturing firms. This e-mail invitation included the URL
address of the web site hosting the questionnaire. This web site was developed to allow
each Korean manufacturing company participating in this study to submit its response
only once. This would eliminate any bias that may occur due to multiple responses
from the same participant.
Before sending the e-mail invitation, the project team at the Pollever Research
Center contacted the managers in charge of ERP systems in Korean manufacturing
companies and invited them to participate in the study. After one month, the research
center made follow-up phone calls to these managers regarding the study. After
two months, a total of 366 responses were collected. However, 110 of those responses
included incomplete data; thus, only 256 responses were included in this study.
Business characteristics and ERP implementation approaches. From the
256 responding firms, 71.9 percent of them had less than 300 employees (Table I);
thus, according to the Small & Medium Business Administration of Korea
(eng.smba.go.kr), these firms are considered small or medium enterprises. Regarding
the production approach, the percentages shown in Table I indicate the portion of items
being produced in MTS or MTO fashion. Majority of the responding firms (69.9 percent)
are dominated by MTO approach; whereas only 10.2 percent of them are dominated by
MTS approach. The remaining 19.9 percent of them have a more or less equal split
between MTS and MTO. Almost half of the responding firms (43.8 percent) pursue a
single ERP package; while 27.7 percent of them pursue a more multifaceted approach by
selecting best-of-breed from several ERP packages. Interestingly, 28.5 percent of them
indicate that all or parts of their ERP systems were developed in-house (Table I).
When implementing their ERP systems, most of the responding firms (75.0 percent)
follow either the Big bang or the Mini big bang implementation approaches.
Only 25.0 percent of them report following either Phase-in by module or the Phase-in
by site.
Three particular ERP modules distribution/logistics (DLModule), material
management (MMModule), and production planning (PPModule), are directly related to
manufacturing process and widely implemented among Korean manufacturing firms.
Thus, these three ERP modules were included and investigated in this current study. On
average, across the implementations of these three ERP modules among the responding
firms, 47.3 percent of the implementations involve some ERP software customization;
whereas 28.5 percent of the implementations involve some re-configuration of the
existing business processes. Approximately 24.2 percent of the implementations do not
need to conduct either ERP software customization or process re-configuration (Table I).

Effect of
business
characteristics
191

MRR
37,2

192

Table I.
Business orientation and
ERP implementation
approach

%
Number of employees (SIZE)
1. #50 (small businesses)
26.6
2. .50 and # 300 (medium enterprises)
45.3
3. .300 (large enterprises)
28.1
Production approach (MTSMTO)
1. Portions of items produced by MTO
are $ 65%
69.9
2. Portions of items produced by MTO and
by MTS are approximately equal
19.9
Portions of items produced by MTS
are $ 65%
10.2
ERP selection approach (SELECT):
1. A single ERP package
43.8
2. Best-of-breed from several ERP packages
27.7
3. In-house development for all or parts of
ERP system
28.5
ERP vendors
Bizentro (a subsidiary of Samsung
Corporation)
46.1
SAP
26.6
Oracle
16.4
Other domestic vendors
10.9
ERP rollout approach (ROLLOUT):
1. Big bang: a single go-live date for all
ERP modules
47.7
2. Mini big bang: several go-live dates for
different subsets of ERP modules
27.3
3. Phased-in by module or by site:
incrementally implement the ERP system
25.0
Process re-configuration vs software
DLMODULE MMMODULE PPMODULE Average
customization (MODIFY)
(%)
(%)
(%)
(%)
1. Significant numbers of changes are made
to the ERP module to fit the existing
processes
15.6
12.5
12.9
47.3
2. Some changes are made to the ERP module
to fit the existing processes
32.8
33.6
34.4
3. The ERP module and the existing
processes are fit to each other without any
change
25.4
24.6
22.7
24.2
4. Some changes are made to the existing
processes to fit the ERP module
20.3
23.8
20.7
28.5
5. Significant numbers of changes are made
to the existing processes to fit the ERP
module
5.9
5.5
9.4

Outcomes from implementing ERP systems. This study measured six business
outcomes from implementing ERP systems. These six outcomes were also reported in
the previous studies on US and Swedish manufacturing firms (Mabert et al., 2000;
Olhager and Selldin, 2003). Each of these outcomes was administered by using the
five-point Likert scale where the 5-scale represents a great amount of benefit and the

1-scale represents not at all. Table II shows the percentage of the responses for each
outcome. Majority of the responding firms perceive at least some benefit from
implementing ERP systems. The responding firms experience improved performance in
terms of quality and availability of information, coordination with both suppliers and
customers, and competitive impact (i.e. linking to global activities and gaining strategic
advantage).

Effect of
business
characteristics
193

Procedure
Cluster analysis is an analysis tool applied to the data that exhibit natural groupings.
Its objective is to sort through cases or observations (e.g. people, things, events) and
classify them into groups or clusters so that the degree of association is strong between
members of the same cluster and weak between members of different clusters. That is,
a cluster is a group of relatively homogeneous cases or observations. Each cluster
describes the group to which its members belong. Members in a cluster are similar to
each other. They are also dissimilar to members in other clusters.
Cluster analysis is thus a tool of discovery and may reveal associations and structure
in data which, though not previously evident, are sensible and useful once found. The
results of cluster analysis may contribute to the definition of a formal classification
scheme, such as similar customers in each market segment. Researchers have used
cluster analysis technique to identify distinctive groups of supply chain partners in
automobile industry (Trappey et al., 2010).
This study is an exploratory study aiming to explore whether different groups of
manufacturing firms with similar business characteristics and ERP implementation
approaches experience different business outcomes from ERP implementation. Thus,
in this study, we employed cluster analysis to identify groups of manufacturing firms
having similar business characteristics and adopting similar ERP implementation
approaches. Then, multivariate analysis of variance (MANOVA) with post-hoc tests
were employed to examine whether different groups of manufacturing firms with
similar business characteristics and ERP implementation approaches experience
different business outcomes from ERP implementation. However, before performing
cluster analysis and MANOVA, we conducted construct reliability and validity tests.
Construct reliability. The level of process re-configuration vs software customization
was measured across three ERP modules DLModule, MMModule, and PPModule.
Thus, we needed to assess the reliability of this construct. Additionally, among the
six business outcomes, we explored whether similar outcomes could be grouped to form
some constructs. First, we conducted an exploratory factor analysis (EFA) on these nine
Area benefiting from implementing ERP
Improved coordination with customers (CUST)
Improved coordination with suppliers (SUPPLY)
Link to global activities (GLOBAL)
Gain strategic advantages (STRADV)
Quality of information (QUAINFO)
Availability of information (AVAINFO)

5 (%)

4 (%)

3 (%)

2 (%)

1 (%)

8.2
5.5
6.6
4.7
14.1
15.2

46.5
46.5
31.6
31.3
38.7
47.3

32.4
37.5
42.2
49.2
32.0
30.1

9.0
9.0
13.7
9.8
12.5
6.3

3.9
1.6
5.9
5.1
2.7
1.2

Notes: 5 a great amount of benefit; 4 significant amount of benefit; 3 some benefit; 2 only a
little benefit; 1 not at all

Table II.
Outcomes from
implementing ERP
(in percentage)

MRR
37,2

194

measuring items (i.e. level of process re-configuration vs software customization across

three ERP modules and six business outcomes). The factor loadings (see Table AI in the
Appendix) suggest four constructs: process re-configuration vs software customization
(MODIFY), informational impact (INFO), external coordination (EXTCO), and
competitive impact (COMP). All constructs have their Cronbachs a values (Table AII)
either above or very close to the cutoff point of 0.70 (Nunnally, 1978).
Then, based on the EFA result which suggested a measurement model with four
constructs, we estimated this measurement model by conducting a confirmatory factor
analysis (CFA). The CFA result is presented in Table AIII and the model fit summary
is in Table AIV in the Appendix. These model fit test results suggest that the
measurement model with four constructs fits the sample data fairly well. Additionally,
we computed the composite reliability (CR) and the average variance extracted (AVE)
values of the four constructs (see Table AII in the Appendix). All CR values, except
that of the COMP construct (i.e. 0.676), are above the recommended threshold of 0.70
(Fornell and Larcker, 1981; Hair et al., 2006; Segars, 1997). Similarly, all AVE values are
above the suggested threshold of 0.50 (Hair et al., 2006; Segars, 1997), indicating that
the four constructs have captured a relatively high level of variance. In sum, all the
results of these reliability tests indicate a reasonably high level of instrument
reliability.
Convergent and discriminant validity. All loadings from the CFA results
(see Table AIII in the Appendix) are high and the t-values (ranging from 4 to 14) for
all loadings are above the 2.54 threshold supporting the statistical significance of the
loadings ( p , 0.01). Additionally, all squared multiple correlations (R 2) values are high.
These results support the assertion that the measuring items in this study are good
measures of the constructs (Gefen et al., 2000).
Based on the factor loadings of the EFA results (see Table AI in the Appendix), there
is no cross loading above 0.40. This suggests the discriminant validity of the four
constructs (McKnight et al., 2002). Additionally, the square root of the AVE of each
construct is greater than any of the constructs correlations with other constructs
(see Table AII in the Appendix). This result provides evidence for discriminant validity
of the constructs in the model (Fornell and Larcker, 1981; Segars, 1997). We also
compared the discriminant validity in the original measurement model with four
constructs against other measurement models with only three constructs, which
included every possible combination of collapsing two constructs into one (Gefen et al.,
2000). The x 2-value in the original measurement model is significantly better than the
x 2-value of every reduced measurement model (see Table AV in the Appendix).
Cluster analysis. A three-cluster solution is suggested by result of the two-step
clustering method to be the most parsimonious grouping of the responding firms and
also the solution that best reflects the meaningful pattern of the relationships between
business characteristics and ERP implementation approaches. Frequency of the
responses (in percentage) regarding each of the four categorical cluster variables
(i.e. SIZE, MTSMTO, SELECT, and ROLLOUT) for each cluster is reported in Table III.
Additionally, Table IV shows the mean and standard deviation of all five cluster
variables for each of the three clusters. Figure 2 shows these mean scores (or cluster
centers) with a snake diagram.
The responding firms in Cluster 1 are small businesses focusing more on MTS
small businesses with MTS-oriented. Majority of the responding firms in this cluster

Categorical cluster variables

SIZE
1. Small businesses
2. Medium enterprises
3. Large enterprises
MTSMTO
1. MTO-oriented
2. Equal MTS/MTO focuses
3. MTS-oriented
SELECT
1. A single ERP package
2. Best-of-breed from several ERP packages
3. Some in-house development
ROLLOUT
1. Big bang
2. Mini big bang
3. Phased-in

Cluster 1 (%)

Cluster 2 (%)

100.0

7.5
92.5

57.4
42.6

Cluster 3 (%)

29.4
70.6

84.9
15.1

98.0
2.0

61.3

52.5
34.4
13.1

38.7

22.5
49.0
28.4

41.0
31.1
27.9

77.4
5.4
17.2

24.5
45.1
30.4

Cluster variables

Cluster 1

Cluster 2

Cluster 3

Total

SIZE

1.0000
(0.000)
2.4262
(0.499)
1.6066
(0.714)
1.8689
(0.826)
2.7978
(1.008)
61

1.9247
(0.265)
1.1505
(0.360)
1.7742
(0.979)
1.3978
(0.768)
2.7312
(1.083)
93

2.7059
(0.458)
1.0196
(0.139)
2.0588
(0.715)
2.0588
(0.742)
2.7255
(0.982)
102

2.0156
(0.741)
1.4023
(0.667)
1.8477
(0.838)
1.7734
(0.823)
2.7448
(1.022)
256

MTSMTO
SELECT
ROLLOUT
MODIFY
Cases

Effect of
business
characteristics
195

Table III.
Frequency of each
categorical variable
for each cluster

Table IV.
Mean and standard
deviation of cluster
variables for each cluster

Figure 2.
Snake diagram of the
cluster centers

MRR
37,2

196

select a single ERP package (52.5 percent) and employ big bang implementation
approach (41.0 percent). Virtually all of the responding firms in Cluster 2 are medium
enterprises focusing on MTO medium enterprise with MTO-oriented. More than
half (61.3 percent) of the responding firms in this cluster select a single ERP package
and majority of the responding firms in this cluster (77.4 percent) employ big bang
implementation approach. Finally, majority of the responding firms in Cluster 3
(70.6 percent) are large enterprises focusing on MTO large enterprises with
MTO-oriented. Approximately half (49.0 percent) of the responding firms in this
cluster select best-of-breed from several ERP packages. The responding firms in this
cluster mainly employ mini big bang or phased-in implementation approaches.
Multivariate analysis of variance. The MANOVA results in the following Table V
indicate that the three clusters are significantly different based on the five cluster
variables (i.e. SIZE, MTSMTO, SELECT, ROLLOUT, MODIFY). Then, we conducted a
post-hoc test to check the equality of all five cluster variables across the three clusters.
Researchers suggest that the follow-up analysis in this post-hoc test could employ either
analysis of variance (ANOVA) or discriminant analysis (Bray and Maxwell, 1985).
Results of the post-hoc test employing ANOVA are presented in the following
Table VI (we also conducted the post-hoc test employing discriminant analysis and the
results of this discriminant analysis are similar to the ANOVA results). The three
clusters are significantly different in only four cluster variables SIZE, MTSMTO,
SELECT, and ROLLOUT.
Furthermore, we conducted pairwise comparisons using Bonferroni test. Bonferroni
test is the optimal test for pairwise comparisons because of its reasonable power and
ease of application (Bird, 1975; Ramsey, 1980). From these pairwise comparisons, all
three clusters are significantly different from each other in only two cluster variables
SIZE and MTSMTO. Results of these pairwise comparisons are presented in the
following Table VII.
Based on the mean scores of the three business outcomes across the three clusters
(Table VIII), the responding firms in all three clusters report that they experienced at
least some benefits in all three business outcomes.
Then, we followed similar steps (i.e. using MANOVA followed by post-hoc ANOVA
test including pairwise comparisons with Bonferroni test) to test the difference in the
business outcomes from implementing ERP system across the responding firms in the
three clusters. The MANOVA results (Table IX) suggest that the three clusters are
Effect

Table V.
MANOVA testing
difference across clusters
based on all five cluster
variables

Intercept
Pillais trace
Wilks lambda
Hotellings trace
Roys largest root
Cluster
Pillais trace
Wilks lambda
Hotellings trace
Roys largest root

Value

Hypothesis df

Error df

Sig.

0.983
0.017
56.765
56.765

2,826.911
2,826.911
2,826.911
2,826.911

5
5
5
5

249
249
249
249

, 0.001
, 0.001
, 0.001
, 0.001

1.164
0.081
8.344
7.964

69.572
125.338
206.922
398.192

10
10
10
5

500
498
496
250

, 0.001
, 0.001
, 0.001
, 0.001

Variables
SIZE
Between groups
Within groups
Total
MTSMTO
Between groups
Within groups
Total
SELECT
Between groups
Within groups
Total
ROLLOUT
Between groups
Within groups
Total
MODIFY
Between groups
Within groups
Total

SIZE
Cluster 1 vs Cluster
Cluster 1 vs Cluster
Cluster 2 vs Cluster
MTSMTO
Cluster 1 vs Cluster
Cluster 1 vs Cluster
Cluster 2 vs Cluster
SELECT
Cluster 1 vs Cluster
Cluster 1 vs Cluster
Cluster 2 vs Cluster
ROLLOUT
Cluster 1 vs Cluster
Cluster 1 vs Cluster
Cluster 2 vs Cluster

Sum of squares

df

Mean square

Sig.

112.288
27.650
139.938

2
253
255

56.144
0.109

513.730

, 0.001

84.787
28.771
113.559

2
253
255

42.394
0.114

372.788

, 0.001

8.596
170.463
179.059

2
253
255

4.298
0.674

6.379

0.002

21.982
150.877
172.859

2
253
255

10.991
0.596

18.430

, 0.001

0.227
266.211
266.438

2
253
255

0.113
1.052

0.108

0.898

Mean difference (I J)

SE

Sig.

2
3
3

2 0.9247
2 1.7059
2 0.7811

0.0545
0.0535
0.0474

, 0.001
, 0.001
, 0.001

2
3
3

1.2757
1.4066
0.1309

0.0556
0.0546
0.0484

, 0.001
, 0.001
0.022

2
3
3

2 0.1676
2 0.4523
2 0.2846

0.1352
0.1329
0.1177

0.649
0.002
0.049

2
3
3

0.4710
2 0.1900
2 0.6610

0.1272
0.1250
0.1107

0.001
0.389
, 0.001

significantly different because of the three business outcomes from implementing ERP
system. Then, results of the post-hoc test using ANOVA in Table X indicate that
manufacturing firms in the three clusters experienced significant differences in only
EXTCO and COMP outcomes (we also conducted this post-hoc test by using
discriminant analysis and results from this discriminant analysis are similar to the
results from ANOVA). Finally, results of the pairwise comparisons (Table XI) show
that the manufacturing firms in Cluster 3 (large enterprises with MTO-oriented) report
significantly higher benefits in both EXTCO and COMP outcomes than the

Effect of
business
characteristics
197

Table VI.
Post-hoc ANOVA testing
equality of all five
cluster variables

Table VII.
Pairwise comparisons
testing equality of four
cluster variables across
three clusters

MRR
37,2

198

manufacturing firms in Cluster 1 (small businesses with MTS-oriented) and Cluster 2

(medium enterprises with MTO-oriented) do.
Discussion
When implementing ERP systems, the responding firms in all three clusters seemed to
have approximately the same amount of process re-configuration vs software
customization. The responding firms reported that, on average, they made only some
changes to the ERP modules; that is, the average score for MODIFY was approximately
2.75 (Table IV) from the five-point dichotomous scale (Table I). This may be the result

Variables

Cluster 1

Cluster 2

Cluster 3

Total

EXTCO

3.1885
(0.690)
3.2541
(0.789)
3.1148
(0.733)
61

3.2419
(0.686)
3.4570
(0.765)
3.0000
(0.818)
93

3.5245
(0.772)
3.4167
(0.731)
3.4363
(0.757)
102

3.3418
(0.735)
3.3926
(0.759)
3.2012
(0.796)
256

INFO
Table VIII.
Mean and standard
deviation of the outcomes
for each cluster

COMP
Cases

Effect

Table IX.
MANOVA testing
difference across clusters
based on all three
business outcomes

Intercept
Pillais trace
Wilks lambda
Hotellings trace
Roys largest root
Cluster
Pillais trace
Wilks lambda
Hotellings trace
Roys largest root

ERP outcomes

Table X.
Post-hoc ANOVA testing
equality of all three
business outcomes

EXTCO
Between groups
Within groups
Total
INFO
Between groups
Within groups
Total
COMP
Between groups
Within groups
Total

Value

Hypothesis df

Error df

Sig.

0.973
0.027
36.255
36.255

3,033.355
3,033.355
3,033.355
3,033.355

3
3
3
3

251
251
251
251

, 0.001
, 0.001
, 0.001
, 0.001

0.115
0.887
0.126
0.109

5.124
5.188
5.250
9.163

6
6
6
3

504
502
500
252

, 0.001
, 0.001
, 0.001
, 0.001

Sum of squares

df

Mean square

5.766
132.077
137.843

2
253
255

2.883
0.522

5.522

0.004

1.615
145.181
146.796

2
253
255

0.807
0.574

1.407

0.247

9.857
151.533
161.390

2
253
255

4.929

8.229

, 0.001

0.599

Sig.

from the finding that the major ERP vendors among the responding firms were domestic
companies (Table I). Similarly, the domestic ERP vendors dominated the Chinese ERP
market as well (Liang et al., 2004).
Foreign ERP systems may not meet the requirements of the domestic companies in
terms of language, reporting format and content, and system flexibility (Liang et al.,
2004). On the other hand, the ERP systems from domestic vendors may better meet the
requirements of the domestic companies in Asia; thus, these domestic companies do not
need to extensively modify either their existing business processes or the ERP software.
However, a firm explanation can only be supported with additional investigations.
For ERP selection approach, large enterprises tend to select best-of-breed from
several ERP packages; on the other hand, small businesses and medium enterprises
tend to select a single ERP package (Table III). Unlike large enterprises, small and
medium businesses usually have scarce financial resources and do not have the same
ease of hiring qualified IS/IT experts (Caldeira and Ward, 2003). Thus, it is more
feasible for large enterprises to adopt several different ERP packages that best meet
their business requirements and then integrate these packages together. In contrast, for
small businesses and medium enterprises, this approach is less feasible because of
their lack of IS/IT expertise; thus, it is more reasonable for small businesses and
medium enterprises to adopt a single ERP package and avoid the integration
challenges of implementing several different ERP packages.
For small businesses and medium enterprises, the most common ROLLOUT approach
employed is the big bang approach; in contrast, large enterprises tend to adopt either
mini big bang or phase-in approaches (Table III). For small businesses, their business
functions and processes are simple and less complicated, implying somewhat easy ERP
implementation (Shin, 2006). Thus, it is feasible to rollout the whole ERP project in small
businesses by using the big bang approach. However, for the bigger ERP project in large
enterprises, it is necessary to employ either mini big bang or phase-in approaches.
Results of the analysis suggest that none of the three cluster variables regarding ERP
implementation approaches MODIFY, SELECT, and ROLLOUT, are significantly
different across the responding firms in all three clusters. On the other hand, the two cluster
variables related to business characteristics SIZE and MTSMTO, are significantly
different among the manufacturing firms across the three clusters. These findings suggest
that business characteristics, not ERP implementation approaches, are significant factors
for grouping manufacturing firms into clusters of companies with similar characteristics.
The findings also showed that small businesses tend to be more MTS-oriented;
while medium and large enterprises tend to focus more on MTO orientation.
I
EXTCO
Cluster 1
Cluster 1
Cluster 2
COMP
Cluster 1
Cluster 1
Cluster 2

Mean difference (I J)

SE

Sig.

vs Cluster 2
vs Cluster 3
vs Cluster 3

2 0.0534
2 0.3360
2 0.2826

0.1190
0.1170
0.1036

1.000
0.013
0.020

vs Cluster 2
vs Cluster 3
vs Cluster 3

0.1148
2 0.3215
2 0.4363

0.1275
0.1253
0.1110

1.000
0.033
, 0.001

Effect of
business
characteristics
199

Table XI.
Pairwise comparisons
testing equality of two
business outcomes across
three clusters

MRR
37,2

200

These results may be understood based on the relationship between the number of
product offerings and the production approach. That is, firm size affects the product
varieties at which the firm can produce internally (Ono and Stango, 2005) and the
increase in number of product offerings goes hand-in-hand with a shift from MTS
approach to MTO approach (Gupta and Benjaafar, 2004).
Regarding business outcomes from implementing ERP systems, the analysis results
suggest that the manufacturing firms in all three groups experience similar benefits from
implementing ERP systems regarding the quality and availability of information. However,
the manufacturing firms in the large enterprises with MTO-oriented group perceive
significantly higher benefits in coordination with their business partners and in competitive
impact than the manufacturing firms in the other two clusters do. This may be explained by
the previous findings that small firms are subject to contradictory pressures forcing them to
provide better logistics contributions and to develop and maintain closer relationships with
their trading partners despite their limited resources (Bagchi and Virum, 1998).
To be successful in the current competitive environment, it is important for both
large and small manufacturing companies to coordinate with their business partners
and adapt to the demand of logistics chain integration. Large manufacturing
enterprises usually find themselves at the top of large networks of suppliers which are
mostly small manufacturing firms. It is normal for large manufacturing enterprises to
place strong pressure on small manufacturing firms (i.e. their suppliers) to adopt the
management practices of large manufacturing enterprises (De Toni et al., 1995). These
management practices include controls, procedures and systems, etc. which are
established by the ERP systems adopted by the large manufacturing enterprises.
On the other hand, small manufacturing firms are often conditioned to behave
according to large manufacturing enterprises expectations and management practices.
The logistics and operational systems put in place in small manufacturing firms are very
dependent on their large manufacturing customers. Similarly, how the ERP is planned is
not the sole decision of these small manufacturing firms (Huin et al., 2002).
Unfortunately, the ERP systems adopted by large manufacturing enterprises may
implement the logistics chain integration that is not compatible with the features and
the intrinsic characteristics of small manufacturing firms. As a result, these small
manufacturing firms often face risk of quasi-integration (Gelinas and Bigras, 2004).
Additionally, the ongoing unilateral relationship between a large manufacturing
enterprise and its small suppliers (i.e. small manufacturing firms) may inhibit the small
manufacturing firms relationships with other partners as well (Kasouf and Celuch,
1997). It was found that only 61 percent of small manufacturing firms said they were
involved in at least one partnership, compared to nearly 88 percent of large
manufacturing enterprises (Gelinas and Bigras, 2004).
Thus, when compared to large manufacturing enterprises, small manufacturing firms
may perceive that they gain limited benefits from implementing ERP systems to participate
and coordinate in logistics chain activities, and eventually limited competitive impact.
Conclusion
In this study, we employed cluster analysis to identify groups of manufacturing firms
having similar business characteristics and adopting similar ERP implementation
approaches. Results of the cluster analysis suggest three clusters or groups of
manufacturing firms. However, none of the three ERP implementation approaches

investigated in this study (i.e. MODIFY, SELECT, and ROLLOUT) are significant
factors for grouping manufacturing firms. On the other hand, the two business
characteristics included in this study (i.e. SIZE and MTSMTO) are significant criteria
for grouping manufacturing companies. This result is consistent with the notion that
large enterprises offer higher product variety than small businesses do (Ono and
Stango, 2005) and the increase in number of product offerings goes hand-in-hand with
a shift from MTS approach to MTO approach (Gupta and Benjaafar, 2004).
One of the major implications of this study lies in these findings. The findings
suggest that company size and production approaches are useful variables for
grouping manufacturing firms into clusters of companies. Thus, any future studies
could refer to these findings and use company size and production approaches in
grouping manufacturing companies.
Then, we examined the differences in three business outcomes from implementing ERP
systems among the three groups of manufacturing firms. The manufacturing companies
in all three clusters report that they experience at least some benefits in all of the three
business outcomes. However, the results of MANOVA and post-hoc tests suggest that the
manufacturing companies in the large firms with MTO-oriented group have
significantly higher perceived benefits in coordination with their business partners and
in competitive impact than the smaller manufacturing firms in the other two clusters do.
Small manufacturing firms often are forced by their customers (which are usually large
manufacturing firms at the top of the networks of suppliers) to adopt the logistics practices
established by ERP systems that may be more suitable for large manufacturing firms.
This leads to some disadvantages for small manufacturing firms and eventually these
small manufacturing firms gain only limited benefits from adopting ERP systems,
participating in logistics integration, and coordinating with large manufacturing firms.
The findings of this study suggest several other implications for future research.
Large manufacturing firms may tend to select best-of-breed from several ERP
packages because these large firms have IS/IT capability to integrate the ERP
packages together. However, additional investigations would be necessary to confirm
this argument. It would also be interesting to investigate the effects of the logistics
practices established by ERP systems and adopted by small manufacturing firms even
though these logistics practices may be more suitable for large manufacturing firms.
Additionally, the practical implication of this study is for the managers of small
manufacturing firms. A small manufacturing firm may need to adopt an ERP system in
response to the request from its business partners who are large manufacturing
companies. The manager of this small manufacturing company needs to be aware that the
ERP system suggested or requested by his business partners (i.e. large manufacturing
firms) may not be compatible with the features and the intrinsic characteristics of his small
manufacturing firm. In this case, the manager should identify the requirements of his
business partners and the incompatibilities that the suggested ERP system could cause to
his small manufacturing firm. Then, the manager may negotiate to implement another
ERP system that could fulfill the same requirements and cause the least incompatibilities.
Alternatively, the manager may need to prepare for a project to rework his companys
existing incompatibilities or to modify the suggested ERP system.
This study, like any other studies, is not free of limitations. First, it was conducted in
only one industry of one country (i.e. Korean manufacturing firms). Second, the study
used perception data collected by self-reporting instrument. Thus, further studies

Effect of
business
characteristics
201

MRR
37,2

202

conducted in other industries and/or other countries and using more objective measures
would allow more generalization of the findings of this study. Third, the data provided
by the research center conducting the survey were not sufficient to either assess response
rate or analyze any nonresponse bias. Finally, although the results of several reliability
tests indicate a reasonably high level of instrument reliability, Cronbachs a values of
some constructs in the measurement model are not higher than the cutoff point of
0.70 (Nunnally, 1978) and, for one construct, its CR is not higher than the recommended
threshold of 0.70 (Fornell and Larcker, 1981; Hair et al., 2006; Segars, 1997).
References
Bagchi, P. and Virum, H. (1998), Logistical competencies of SMEs from Norway, Logistiqueet
Management, Vol. 6 No. 2, pp. 9-111.
Beard, J.W. and Sumner, M. (2004), Seeking strategic advantage in the post-net era: viewing
ERP systems from the resource-based perspective, Journal of Strategic Information
Systems, Vol. 13 No. 2, pp. 129-150.
Bird, K.D. (1975), Simultaneous contrast testing procedures for multivariate experiments,
Multivariate Behavioral Research, Vol. 10 No. 3, pp. 343-351.
Bray, J.H. and Maxwell, S.E. (1985), Multivariate Analysis of Variance, Sage, Beverly Hills, CA.
Caldeira, M.M. and Ward, J. (2003), Using resource-based theory to interpret the successful adoption
and use of information systems and technology in manufacturing small and medium-sized
enterprises, European Journal of Information Systems, Vol. 12 No. 2, pp. 127-141.
Cliffe, S. (1999), ERP implementation how to avoid $100 million write-offs, Harvard Business
Review, Vol. 77 No. 1, pp. 16-17.
Cooper, D.R. and Schindler, P.S. (2000), Business Research Methods, 7th ed., McGraw-Hill,
New York, NY.
De Toni, A., Nassimbeni, G. and Tonchia, S. (1995), Small local firms inside the supply chain:
challenges and perspective, Small Business Economics, Vol. 7 No. 3, pp. 241-249.
Dezdar, S. and Sulaiman, A. (2009), Successful enterprise resource planning implementation:
taxonomy of critical factors, Industrial Management & Data Systems, Vol. 109 Nos 8/9,
pp. 1037-1052.
Fornell, C. and Larcker, D. (1981), Evaluating structural equation models with unobservable
variables and measurement error, Journal of Marketing Research, Vol. 18 No. 1, pp. 39-50.
Gattiker, T.F. and Goodhue, D.L. (2005), What happens after ERP implementation:
understanding the impact of interdependence and differentiation on plant-level
outcomes, MIS Quarterly, Vol. 29 No. 3, pp. 559-585.
Gefen, D., Straub, D.W. and Boudreau, M. (2000), Structural equation modeling and regression:
guidelines for research practice, Communications of the Association for Information
Systems, Vol. 4, Article 7, pp. 1-79.
Gelinas, R. and Bigras, Y. (2004), The characteristics and features of SMEs: favorable or unfavorable
to logistics integration?, Journal of Small Business Management, Vol. 42 No. 3, pp. 263-278.
Gupta, D. and Benjaafar, S. (2004), Make-to-order, make-to-stock, or delay product
differentiation? A common framework for modeling and analysis, IIE Transactions,
Vol. 36 No. 6, pp. 529-546.
Hair, J.F., Black, W.C., Babin, B.J., Anderson, R.E. and Tatham, R.L. (2006), Multivariate Data
Analysis, 6th ed., Prentice-Hall, Upper Saddle River, NJ.

Hayes, R., Pisano, G., Upton, D. and Wheelwright, S. (2005), Operations Strategy, and Technology,
Wiley, Danvers, MA.
Helo, P., Anussornnitisarn, P. and Phusavat, K. (2008), Expectation and reality in ERP
implementation: consultant and solution provider perspective, Industrial Management
& Data Systems, Vol. 108 No. 8, pp. 1045-1059.
Huang, S.Y., Huang, S.M., Wu, T.H. and Lin, W.K. (2009), Process efficiency of the enterprise resource
planning adoption, Industrial Management & Data Systems, Vol. 109 Nos 8/9, pp. 1085-1100.
Huin, S.F., Luong, L.H.S. and Abhary, K. (2002), Internal supply chain planning determinants in
small and medium-sized manufacturers, International Journal of Physical Distribution
& Logistics Management, Vol. 3 No. 9, pp. 771-782.
Ifinedo, P. and Nahar, N. (2009), Interactions between contingency, organizational IT factors, and ERP
success, Industrial Management & Data Systems, Vol. 109 Nos 1/2, pp. 118-137.
Kasouf, C.J. and Celuch, K.G. (1997), Interfirm relationships in the supply chain: the small
suppliers view, Industrial Marketing Management, Vol. 26 No. 6, pp. 475-486.
Katerattanakul, P., Hong, S. and Lee, J. (2006), Enterprise resource planning survey of Korean
manufacturing firms, Management Research News, Vol. 29 No. 12, pp. 820-837.
Ketchen, D.J., Thomas, J.B. and Snow, C.C. (1993), Organizational configurations
and performance: a comparison, Academy of Management Journal, Vol. 36 No. 6,
pp. 1278-1313.
Lee, D.H., Lee, S.M., Olson, D.L. and Chung, S.H. (2010), The effect of organizational support on ERP
implementation, Industrial Management & Data Systems, Vol. 110 Nos 1/2, pp. 269-283.
Lengnick-Hall, C., Lengnick-Hall, M. and Abdinnour-Helm, S. (2004), The role of social and
intellectual capital in achieving competitive advantage through enterprise resource
planning (ERP) systems, Journal of Engineering and Technology Management, Vol. 21
No. 4, pp. 307-330.
Liang, H.G., Xue, Y.J., Boulton, W.R. and Byrd, T.A. (2004), Why western vendors dont
dominate Chinas ERP market, Communications of the ACM, Vol. 47 No. 7, pp. 69-72.
McKnight, D.H., Choudhury, V. and Kacmar, C. (2002), Developing and validating trust
measures for e-commerce: an integrative typology, Information Systems Research, Vol. 13
No. 3, pp. 334-359.
Mabert, V.A., Soni, A. and Venkatramanan, M.A. (2000), Enterprise resource planning survey
of US manufacturing firms, Production and Inventory Management Journal, Vol. 41 No. 2,
pp. 52-58.
Maguire, S., Ojiako, U. and Said, A. (2010), ERP implementation in Omantel: a case study,
Industrial Management & Data Systems, Vol. 110 Nos 1/2, pp. 78-92.
Nunnally, J. (1978), Psychometric Theory, McGraw-Hill, New York, NY.
Olhager, J. and Selldin, E. (2003), Enterprise resource planning survey of Swedish manufacturing
firms, European Journal of Operational Research, Vol. 46 No. 2, pp. 365-373.
Ono, Y. and Stango, V. (2005), Outsourcing, firm size, and product complexity: evidence from
credit unions, Economic Perspective, Vol. 29 No. 1, pp. 2-11.
Parr, A. and Shanks, G. (2000), A model of ERP project implementation, Journal of Information
Technology, Vol. 15 No. 4, pp. 289-303.
Prasad, S., Tata, J. and Madan, M. (2005), Build to order supply chains in developed and
developing countries, Journal of Operations Management, Vol. 23 No. 5, pp. 551-568.
Ramsey, P.H. (1980), Choosing the most powerful pairwise multiple comparison procedure in
multivariate analysis of variance, Journal of Applied Psychology, Vol. 65 No. 3, pp. 317-326.

Effect of
business
characteristics
203

MRR
37,2

204

Segars, A.H. (1997), Assessing the unidimensionality of measurement: a paradigm and illustration
within the context of information systems research, Omega, Vol. 25 No. 1, pp. 107-121.
Shin, I. (2006), Adoption of enterprise application software and firm performance, Small
Business Economics, Vol. 26 No. 3, pp. 241-256.
Soja, P. and Paliwoda-Pekosz, G. (2009), What are real problems in enterprise system
adoption?, Industrial Management & Data Systems, Vol. 109 Nos 5/6, pp. 610-627.
Trappey, C.V., Trappey, A.J.C., Chang, A.C. and Huang, A.Y.L. (2010), Clustering analysis
prioritization of automobile logistics services, Industrial Management & Data Systems,
Vol. 110 Nos 5/6, pp. 731-743.
Velcu, O. (2007), Exploring the effects of ERP systems on organizational performance: evidence from
Finnish companies, Industrial Management & Data Systems, Vol. 107 No. 9, pp. 1316-1334.
Yen, H. and Sheu, C. (2004), Aligning ERP implementation with competitive priorities of
manufacturing firms: an exploratory study, International Journal of Production
Economics, Vol. 92 No. 3, pp. 207-220.
Youssef, K., Van Delft, C. and Dallery, Y. (2004), Efficient scheduling rules in a combined
make-to-stock and make-to-order manufacturing system, Annals of Operations Research,
Vol. 126 Nos 1-4, pp. 103-134.

Appendix. Construct reliability and validity tests

Factors
EXTCO
CUST
SUPPLY
MMMODULE
PPMODULE
DLMODULE
QUAINFO
AVAINFO
GLOBAL
STRADV
Table A1.
Exploratory factor
analysis

Table A2.
Construct correlation
and reliability

MODIFY

INFO

COMP

0.929
0.927
2 0.929
2 0.884
2 0.863
0.890
0.860
0.844
0.886

Note: Extraction method principal component analysis; rotation method oblimin with kaiser
normalization; rotation converged in six iterations; total variance extracted 79.79 percent; any loadings
smaller than 0.40 are not reported

EXTCO
MODIFY
INFO
COMP

CR

AVE

0.843
0.875
0.701
0.676

0.729
0.701
0.540
0.512

(0.838)
0.312
0.375
0.504

(0.872)
0.149
0.128

(0.693)
0.402

(0.673)

Notes: n 256 cases; / Cronbachs a (reported on the diagonal); CR composite reliability;

AVE average variance extracted

Constructs

Items

External coordination (EXTCO)

Improved coordination with

customers (CUST)
Improved coordination with
suppliers (SUPPLY)
Link to global activities (GLOBAL)
Gain strategic advantages
(STRADV)
Quality of information (QUAINFO)
Availability of information
(AVAINFO)
Distribution/logistics (DLModule)
Material management (MMModule)
Production planning (PPModule)

Competitive impact (COMP)

Informational impact (INFO)
Process re-configuration vs software
customization (MODIFY)

Loading t-value

R2

1.118

9.625 0.710

1.000

0.746

1.000
0.830

0.559
5.844 0.465

1.019
1.000

4.930 0.474
0.605

0.990
1.088
1.000

13.654 0.658
14.411 0.828
0.618

Notes: Loading is the non-standardized regression weight; R 2 is the squared multiple correlations

x2
Model
Suggested value

24.5
NA

x 2/d.f.
d.f. ratio
21
NA

1.168
,3.0

RMSEA
0.026
, 0.06

GFI

AGFI

CFI

NFI

IFI

TLI

0.979
. 0.95

0.956
. 0.95

0.996
.0.95

0.972
.0.95

0.996
.0.95

0.993
. 0.95

Notes: x 2/d.f. ratio of x 2 to degrees of freedom; RMSEA root mean square error of
approximation; GFI goodness-of-fit index; AGFI adjusted goodness-of-fit index; CFI
comparative fit index; NFI normed fit index; IFI incremental index of fit; TLI Tucker-Lewis
index

Models
[1]
[2]
[3]
[4]
[5]
[6]
[7]

Original measurement model

Model with COMP and MODIFY combined
Model with COMP and EXTCO combined
Model with INFO and MODIFY combined
Model with COMP and INFO combined
Model with INFO and EXTCO combined
Model with EXTCO and MODIFY combined

x2

d.f.

Compared to [1]
x2diff , d.f. ( p-value)

24.5
111.9
102.6
123.8
112.4
135.1
105.4

21
22
22
22
22
22
22

87.4,
78.1,
99.3,
87.9,
110.6,
80.9,

1
1
1
1
1
1

(, 0.001)
(, 0.001)
(, 0.001)
(, 0.001)
(, 0.001)
(, 0.001)

About the authors

Pairin Katerattanakul is a Professor and the Telecommunications and Information Management
Program Coordinator at the Department of Business Information Systems, Western Michigan
University. He has published his research in several journals such as European Journal of Information
Systems, Journal of the American Society for Information Science and Technology, Communications of
the ACM, Communications of the AIS, Management Research News, Journal of Information
Technology Theory and Application. He also has longer than ten years of industry experience as a
System Engineer and a Product Manager at several multinational companies. Pairin Katerattanakul
is the corresponding author and can be contacted at: p.katerattanakul@wmich.edu

Effect of
business
characteristics
205

Table A3.
CFA result

Table A4.
Model fit summary

Table A5.
Original vs reduced
measurement models

MRR
37,2

206

James J. Lee is a Genevieve Albers Professor of Albers School of Business and Economics.
Prior to that he had served as an Assistant Professor at State University of New York at
Binghamton. He teaches primarily in the areas of E-Commerce & Information Systems,
E-Business, and Web Applications. His research is in the areas of enterprise integration in
social context, information technology conceptualization, and communication structure.
His publications have appeared in such journals as Communications of the ACM, Journal of
Information Technology, Industrial Management & Data Systems, and many more.
Soongoo Hong is an Associate Professor at the Department of Management Information Systems
in Dong-A University, Korea. He received his PhD in management information systems in 2000 and
Master of arts in management in 1995 from the University of Nebraska Lincoln. He has published
his research in several journals such as Communications of the ACM, Communications of the AIS,
International Journal of Information and Decision Making. His research interests include
web accessibility, ERP implementation, data warehousing, and IT impacts on organizations.

To purchase reprints of this article please e-mail: reprints@emeraldinsight.com

Or visit our web site for further details: www.emeraldinsight.com/reprints

Reproduced with permission of the copyright owner. Further reproduction prohibited without
permission.