Relationship Between Intellectual Capital and Knowledge Management 2012

Decision Sciences
Volume 43 Number 3
June 2012

C 2012 The Authors
C 2012 Decision Sciences Institute
Decision Sciences Journal
Relationship between Intellectual Capital

and Knowledge Management: An Empirical
Investigation
I-Chieh Hsu
Department of Business Administration, National Changhua University of Education, Taiwan,
e-mail: fbhsu@cc.ncue.edu.tw
Rajiv Sabherwal
Sam M. Walton College of Business, University of Arkansas, AR 72701, e-mail:
RSabherwal@walton.uark.edu
ABSTRACT
Two important streams of the literature have examined intellectual capital (IC) and
knowledge management (KM). Surprisingly, they have developed in parallel, without
any empirical research on the relationship between them. This article empirically examines how IC and KM affect each other, and also investigates their consequences, viewing
three intermediate consequences (dynamic capabilities, efficiency, and innovativeness)
to mediate their effects on firm performance. In addition, this article examines the effects
of the organizations culture on IC and KM. To address these issues, a comprehensive
model is developed and tested using a combination of survey and secondary data of 533
companies in Taiwan. The results support the theoretical model. Major findings include
the following: IC affects KM and dynamic capabilities; KM facilitates innovation but
not dynamic capabilities or IC; a learning culture facilitates IC and innovation but not
KM; firm performance depends on efficiency and innovation, but not directly on dynamic capabilities; and efficiency does not depend on any of the other constructs in the
study. The articles implications for research and practice are examined. [Submitted:
November 17, 2010. Revised: June 22, 2011; October 4, 2011. Accepted: November 10,
2011.]
Subject Areas: Firm Performance, Intellectual Capital, Knowledge Acquisition, Knowledge Creation, Knowledge Management, Organizational
Learning, Structural Equation Modeling, and Survey Research.
INTRODUCTION
The current knowledge-based economy has led to the literature emphasizing knowledge management (KM) (Eisenhardt & Santos, 2002; Lee & Choi, 2003; Tanriverdi, 2005) and intellectual capital (IC) (Youndt, Subramaniam, & Snell, 2004;
Subramaniam & Youndt, 2005) as major sources of competitive advantage. KM and
IC are distinct, but conceptually interrelated, concepts (cf. Nahapiet & Ghoshal,
Corresponding author.
489
490
Relationship between IC and KM
1998; Easterby-Smith & Prieto, 2008). Whereas KM in firms has been defined as
doing what is needed to get the most out of knowledge resources, including both
explicit and tacit knowledge (Sabherwal & Becerra-Fernandez, 2003), IC captures
the sum of all knowledge firms utilized for competitive advantage (Subramaniam
& Youndt, 2005, p. 451).
The literature on KM and IC share the same broad objective: understanding
the role of knowledge and its management in firm success and competitiveness
(e.g., Nonaka & Takeuchi, 1995; Grant, 1996a, 1996b; Argote, McEvily, & Reagans, 2003). The literature on IC examines the nature of organizational knowledge
and its different types, and also how they affect firm performance (Roos, Roos, Edvinsson, & Dragonetti, 1998), whereas the KM literature deals with the processes
and practices for managing IC (Alavi & Leidner, 2001; Sabherwal & Sabherwal,
2005).
However, the literature in these two areas has developed in parallel. In the
investigation of the effect of IC, or organizational knowledge, questions related to
the processes through which organizations manage knowledge and appropriate its
value receive less attention (Eisenhardt & Santos, 2002). By contrast, research on
KM has given limited consideration to the pools of different types of knowledge
being managed, including difficult-to-codify knowledge such as social capital
(Nahapiet & Ghoshal, 1998; Easterby-Smith & Prieto, 2008). Thus, the prior
literature has not examined how IC and KM affect firm performance when both
aspects are simultaneously considered. Consequently, some important questions
have not yet been effectively addressed. When IC and KM are simultaneously
examined, how does each aspect affect firm performance? Does KM affect firm
performance through IC? Does IC affect firm performance through KM? Do both
aspects affect firm performance through a common set of mediating variables, or
do they affect firm performance through different mediators?
We believe that questions such as above are important both for practice and for
future research on knowledge-related aspects. Moreover, we believe that answering
such questions requires a simultaneous investigation of both IC and KM and their
effects on firm performance through mediating variables emphasized in the prior
literature. Accordingly, we seek to contribute to a comprehensive understanding
of the impact of knowledge-related aspects on firm performance by addressing the
following two research questions:
(i) When both KM and IC are considered, through which mediating factors
do they affect firm performance?
(ii) How do KM and IC affect each other?
Due to the interrelated nature of IC and KM, examining one aspect while excluding
the other provides a partial view of their effects on firm performance. Addressing
research question (i) should help us understand the effects of KM (i.e., processes for
managing knowledge) and IC (i.e., organizational knowledge) on firm performance
when both aspects are simultaneously examined. Moreover, by including important
intermediate outcomes of IC and KM, we respond to previous calls (e.g., Lee
& Choi, 2003; Subramaniam & Youndt, 2005) to open blackboxes (i.e., the
intermediate outcomes needed to improve firm performance). Based on the prior
Hsu and Sabherwal
491
Figure 1: The broad theoretical model.
literature (e.g., Grant 1996a; OReilly & Tushman, 2008), we incorporate dynamic
capabilities, innovation, and efficiency as three variables mediating the effects of
KM and IC on firm performance.
Addressing research question (ii) should help in gaining an initial insight
into the potential chicken-and-egg relationship between KM and IC. Using the
metaphor of stocks and flows (Dierickx & Cool, 1989), KM (which focuses on
knowledge flows) can enhance the value of knowledge to the organization (Decarolis & Deeds, 1999), whereas the stock of knowledge (i.e., IC) facilitates KM
by providing the inputs needed for the knowledge flows (e.g., Smith, Collins, &
Clark, 2005).
A learning culture has previously been viewed as an important factor that
might facilitate KM (e.g., Gold, Mahortra, & Segars, 2001). Although culture has
received less attention in the literature specific to IC (e.g., Subramaniam & Youndt,
2005), the literature on knowledge (e.g., Grant, 1996a; Nahapiet & Ghoshal, 1998;
Lee and Choi, 2003; Yang & Chen, 2007) implies that an organization with a
learning culture would also have greater IC through the development of organizational knowledge over time. Therefore, questions such as the following arise: Does
a learning culture facilitate organizational knowledge (i.e., IC)? Does a learning
culture facilitate KM directly, or through IC? In order to better understand the
effect of culture on IC and KM, we incorporate learning culturewhich is characterized by management support for the role of knowledge, and encouragement of
employees to interact, explore, and seek assistance (Gold et al., 2001; Nonaka &
Toyama, 2005)in this study. Accordingly, in addition to the two research questions
discussed above, we pursue a third research question in this study:
(iii) How does organizational culture affect IC?
Figure 1 summarizes the broad research model used to pursue the three research
questions.
The rest of the article is organized as follows. The next section develops the
theoretical model for the article. The subsequent two sections present data collection and measures, and analysis and results, respectively. The article concludes
with a discussion of the findings and their implications, and acknowledges the
studys limitations.
492
THEORETICAL DEVELOPMENT
IC and KM
Knowledge has been defined as the set of justified beliefs that can enhance the organizations capability for effective action (Nonaka, 1994; Alavi & Leidner, 2001).
An organizations IC refers to the sum of all its knowledge resources, which exist
in aspects within or outside the organization (Subramaniam & Youndt, 2005). IC
has been viewed in terms of where the knowledge resides. For example, Edvinsson
and Malone (1997) view IC as consisting of human capital, and structural capital,
which includes organizational capital and customer capital. Synthesizing the prior
literature, Youndt, Subramaniam, and Snell (2004) and Subramaniam and Youndt
(2005) categorized IC into three components: human capital, or the knowledge,
skills, and capabilities of individual employees; organizational capital, or the institutionalized knowledge and codified experience residing in databases, manuals,
culture, systems, structures, and processes; and social capital, or the knowledge
embedded in networks of relationships and interactions among individuals. While
adopting these three components of IC in conceptualizing and measuring it, we
focus on the overall IC construct when theoretically examining and empirically
testing its relationship with KM, performance, and a learning culture.
KM has been defined as firms doing what is needed to get the most out
of knowledge resources, including both explicit and tacit knowledge (Sabherwal
& Becerra-Fernandez, 2003). Organizational KM includes three major processes:
acquisition, conversion, and application (Gold et al., 2001). Knowledge acquisition
refers to developing new knowledge from data, information, or knowledge (Gold
et al., 2001). Knowledge conversion refers to making the acquired knowledge
useful for the organization (Gold et al., 2001) by structuring it or transforming
tacit knowledge into explicit knowledge (Nonaka, 1994). Knowledge application
refers to the use of knowledge to perform tasks (Sabherwal & Sabherwal, 2005),
although the party that applies knowledge does not necessarily have to understand
it (Conner & Prahalad, 1996). Thus, KM includes the firms processes of acquiring
new knowledge, converting its knowledge into a form that is usable and easily
accessed, and applying knowledge (Verkasalo & Lappalainen, 1998). In this article,
we focus on the overall KM construct, while recognizing its multi-dimensional
nature.
IC (e.g., Stewart, 1997; Subramaniam & Youndt, 2005) and KM (e.g.,
Nonaka, 1994; Gold et al., 2001) literatures have largely evolved independently of
each other. IC and KM are distinct, and may be viewed as referring to knowledge
stocks and processes, respectively (Haas & Hansen, 2005). The prior literature
has made conceptual connections between IC and KM literatures (Nahapiet &
Ghoshal, 1998), and recognized that they are fundamentally related; organizational knowledge is considered IC and the organizational processes required to
acquire, convert, and apply such knowledge are considered KM (Youndt et al.,
2004). Both IC and KM are important bases for organizational competitiveness,
and neither can be pursued independently of the other (Wiig, 1997). However,
there is little empirical research on the bi-directional relationship between IC and
KM.
Hsu and Sabherwal
493
The Relationship between IC and KM

The relationship between IC and KM is rooted in the knowledge-based view
(KBV) of the firm, which has been developed from seminal studies such as Simon
(1965), Cyert and March (1963), Levitt and March (1988), and Huber (1991).
KBV also adopted a logic originating from the resource-based view (RBV) of the
firm, which conveys a dual message. On the one hand, organizational resources
are argued to lead to information-based, tangible or intangible processes (i.e.,
intermediate goods), that create firm competitiveness as the final outcome (Amit
& Schoemaker, 1993, p. 35). On the other hand, the same managerial processes
function to enhance and accumulate organizational resources (Amit & Schoemaker,
1993; Hamel & Prahalad, 1993). Thus, one can observe the reciprocal nature of
the relationship between organizational resources and processes.
KBV considers knowledge to be the key source of firm competitiveness and
suggests a similar reciprocity between knowledge and KM. On the one hand,
knowledge serves as the basis for KM. For example, the complementarity of different kinds of specialized knowledge increases the scope of the knowledge being
integrated, and thus facilitates a knowledge-based capability that is inimitable
for competitors (Grant, 1996b). On the other hand, KM improves and strengthens
knowledge resources. For example, Nonaka and Takeuchis (1995) knowledge creation processes that include four modes of knowledge conversion can trigger spirals
of learning at the individual, group, and organizational levels. New knowledge is
created following the three-level learning.
Following the underlying ideology of KBV, our further literature review
provides considerable basis for expecting IC to affect KM. An organizations prior
knowledge, or IC, affects its processes for absorbing new knowledge (Cohen &
Levinthal, 1990), applying knowledge (Grant, 1996b), and converting knowledge
from one form to another (Sabherwal & Becerra-Fernandez, 2003). This is implicit
in the effect of knowledge within a KM system on the use of the system (Kulkarni,
Ravindran, & Freeze, 20062007), and in the notion of organizational learning,
which focuses on using prior experiences to decide future actions (Levitt & March,
1988).
The prior literature provides reasons for expecting IC to affect KM by identifying ways in which each dimension of IC (i.e., social, human, and organizational)
can affect KM. Social capital facilitates KM because interpersonal interactions
enable knowledge integration (Grant, 1996a), within-firm knowledge sharing (Nahapiet & Ghoshal, 1998), interfirm knowledge transfer (Santoro & Bierly, 2006;
Santoro & Saparito, 2006; Chen, Shih, & Yang, 2009), and knowledge creation
(Nonaka, 1994). Human capital enables KM because individuals within the organization can develop appropriate and needed KM processes (Argote et al., 2003),
and can use their knowledge to improve KM (Nonaka & Takeuchi, 1995). Finally,
organizational capital enables KM because the various forms of organizational
capital, including transactive memory systems, organizational structure, and information technologies, can be leveraged in developing KM processes (Alavi &
Leidner, 2001; Olivera, Goodman, & Tan, 2008).
494
H1: Intellectual capital has a positive effect on knowledge management.
Synthesizing KBV, our literature review also suggests that KM affects the
accumulation and development of IC. The discussion of the knowledge-creating
company (Nonaka & Takeuchi, 1995) recognizes the effect of KM on the creation
of new knowledge. Nahapiet and Ghoshal (1998) also view the creation of new IC
to depend on KM (specifically, the processes of combination and exchange).
The prior literature provides reasons for expecting KM to affect IC by identifying ways in which each dimension of KM (i.e., knowledge acquisition, knowledge application, and knowledge conversion) facilitates IC. The effect of KM on
IC can also be seen in terms of the specific dimensions of KM. One-dimensional
knowledge acquisition focuses directly on the processes for creating new knowledge (Gold et al., 2001). Knowledge application processes rely on direction and
routines, which lead to the embedding of the knowledge within the organization (Grant, 1996a). The prior literature also suggests that repeated application of
knowledge in a given task constitutes a learning process that improves the knowledge being utilized (Eisenhardt & Martin, 2000). Finally, knowledge conversion
leads to an increase in the overall organizations IC through processes by which
knowledge is converted into another form, and transferred to others (Nonaka &
Takeuchi, 1995).
H2: Knowledge management has a positive effect on intellectual capital.
Together, H1 and H2 represent the expected reciprocal effects between IC

and KM, which we mentioned in the introduction. They are consistent with arguments in KBV (e.g., Grant, 1996a, 1996b) that organizational knowledge helps
develop organizational capabilities for integrating diverse knowledge bases. These
capabilities, in return, enhance organizational knowledge. H1 and H2 are also captured by Nahapiet and Ghoshals (1998) view of the processes of exchanging and
combining knowledge to both depend on IC (especially social capital), and affect
the creation of new IC. Similarly, Hult, Ketchen, and Slater (2004) view knowledge acquisition and information distribution as depending on achieved memory,
and as affecting shared meaning. Although H1 and H2 seem intuitively appealing,
and may even appear to be obvious in nature, neither of these hypotheses has
been explicitly tested in prior empirical research. By testing these hypotheses, and
moreover by testing both of these reciprocal effects simultaneously, we hope this
study makes a valuable contribution.
Consequences of IC and KM
KBV examines the effects of changes in the content and management of knowledge
on firm performance (Argote et al., 2003). Firm performance may be defined as
the extent to which the firm has achieved its business objectives (Lee & Choi,
2003), and may be evaluated in terms of profitability or other financial benefits
(Simonin, 1997). Although some prior literature examines the effects of IC and KM
on firm performance (e.g., Youndt et al., 2004), it is difficult to attribute better firm
performance to a single factor such as KM or IC (Lee & Choi, 2003). Therefore,
studies examine improvements in specific processes, such as software development
(Slaughter & Kirsch, 2006), or focus on outcomes that are conceptually more
Hsu and Sabherwal
495
proximate to KM, such as satisfaction with KM (Sabherwal & Becerra-Fernandez,

2003).
The considerable prior literature has focused on innovation and efficiency as
two consequences of KM and IC that subsequently affect firm performance. For
example, March (1991) focused on exploration and exploitation of knowledge,
and related them to innovation and efficiency, respectively. Kogut and Zander
(1992) and Majchrzak, Cooper, and Neece (2004) view innovation as a direct
consequence of KM, whereas Grant (1996a) and Conner and Prahalad (1996)
believe that KM enables organizations to be more efficient. Others have linked
dynamic capabilities to KM (e.g., Zollo & Winter, 2002) and IC (Subramaniam
& Youndt, 2005). OReilly and Tushman (2008) consider dynamic capabilities
as representing ambidexterity between efficiency and innovation, and argued that
all three aspects influence firm performance. Although he focused on efficiency
and innovation as important consequences of KM, Grant (1996b) also highlighted
the importance of dynamic capabilities. The inclusion of dynamic capabilities,
innovation, and efficiency as mediating the effects of KM and IC is thus consistent
with KBV. It is also consistent with prior empirical research (Lee & Choi, 2003;
Sabherwal & Sabherwal, 2005; Subramaniam & Youndt, 2005).
Efficiency of an organization reflects the organizations ability to achieve the
same level of output with a lower level of input, or achieve a greater level of output
with the same level of input. Efficiency (McDaniel & Kolari, 1987) and related
notions such as productivity have been considered important determinants of firm
performance. Greater efficiency has been argued to reduce costs and make the firm
more successful (Porter, 1980).
H3a: Firm efficiency has a positive effect on firm performance.
KBV posits that KM processes facilitate efficiency in two distinct ways. First,
knowledge sharing can reduce or eliminate redundancy in knowledge creation or
learning (Grant, 1996a; Sabherwal & Sabherwal, 2005). In other words, KM can
enable an organizations employees to be more productive through the acquisition,
conversion, and application of knowledge possessed by others (Argote & Ingram,
2000), and can thereby enhance organizational efficiency. Second, KM can enhance
efficiency by reducing the need to transfer knowledge (Conner & Prahalad, 1996;
Grant, 1996a). Grant (1996a, p. 114) remarks: If production requires the integration of many peoples specialized knowledge, the key to efficiency is to achieve
effective integration while minimizing knowledge transfer through cross-learning
by organizational members. Conner and Prahalad (1996) label such application
of others knowledge without learning it as knowledge substitution, and consider
it crucial to organizational efficiency.
H3b: Knowledge management has a positive effect on efficiency.
Innovation at the organizational level has been defined as a technology,

strategy, or management practice that a firm is using for the first time, whether
or not other organizations or users have previously adopted it, or as a significant
restructuring or improvement in a process (Li & Atuahene-Gima, 2001, p. 1124).
The prior literature (e.g., Porter, 1980; Brown & Eisenhardt, 1997) has emphasized
the pursuit of innovation as a way for firms to obtain superior profit margins.
496
Innovation enhances firm performance through improved product/service quality,

timely introduction of new products/services, and greater customer responsiveness
(Bae & Lawler, 2000).
H4a: Innovation has a positive effect on firm performance.
KBV emphasizes the role of KM in firm innovation (Grant 1996a, 1996b;

Nonaka & Toyama, 2005). KM contributes to firm innovation by facilitating new
knowledge-based products or enabling improved products that provide a significant
additional value (Argote & Ingram, 2000; Lynn, Reilly, & Akgun, 2000; Eisenhardt
& Santos, 2002). Knowledge conversion facilitates the transfer of knowledge
within the organization, thereby enabling individuals to develop new ideas (Lee
& Choi, 2003), and enables the creation of new knowledge needed for innovation
(Nonaka, 1994; Nonaka & Takeuchi, 1995). Knowledge acquisition is critical to
the development of new ideas and creative solutions (Haas & Hansen, 2005), and
therefore to the development of new products and services (Gold et al., 2001;
Sabherwal & Sabherwal, 2005). Knowledge application enables specialization
within firms (Demsetz, 1991) so that individuals can focus on the development of
new products and services while benefiting from others relevant expertise (Grant,
1996b).
H4b: Knowledge management has a positive effect on innovation.
Dynamic capabilities have been defined as . . . the firms processes that use
resourcesspecifically the processes to integrate, reconfigure, gain and release
resourcesto match and even create market change (Eisenhardt & Martin, 2000,
p. 1107). Dynamic capabilities include specific processes, such as product development, strategic decision-making, and forming collaborations, that contribute to
value creation in firms (Eisenhardt & Martin, 2000). They lead to greater financial
returns and improved market performance (Brown & Eisenhardt, 1997), and are
necessary for competitive advantage (Eisenhardt & Martin, 2000).
H5a: Dynamic capabilities have a positive effect on firm performance.
Dynamic capabilities support the reconfiguration of the firms resources.

They involve seeking process improvements through modifications in operating
processes (Zollo & Winter, 2002) and using governance modes that support change
(Teece, Pisano, & Shuen, 1997). They help firms develop new products and processes in a timely fashion (Wu, 2006), and are essential for a firm engaged in
innovation-based competition (Zollo & Winter, 2002).
H5b: Dynamic capabilities have a positive effect on innovation.
KM supports organizational learning (Zollo & Winter, 2002), as well as

employees learning from each other and external sources (Conner & Prahalad,
1996), and enables the firm to use prior knowledge to develop new ideas (Lee &
Choi, 2003). Through these effects, KM facilitates dynamic capabilities by enabling processes needed in dynamic markets. The prior literature on KM processes
provides further insights into reasons for expecting KM to facilitate dynamic capabilities. Knowledge acquisition helps the firm to learn new operational routines
Hsu and Sabherwal
497
through external knowledge (Zollo & Winter, 2002), knowledge conversion enables dynamic capabilities through collective learning among individuals (Duncan
& Weiss, 1979), and knowledge application enables dynamic capabilities through
the replication of solutions in new contexts (Zollo & Winter, 2002). Thus, KM
enables the development of dynamic capabilities (Eisenhardt & Martin, 2000).
H5c: Knowledge management has a positive effect on dynamic capabilities.
IC, which has been viewed as the sum of the firms knowledge resources, is
also expected to facilitate dynamic capabilities. This is consistent with Kogut and
Zanders (1992, p. 384) emphasis on the need . . .to understand the knowledge
base of a firm as leading to a set of capabilities that enhance the chances for
growth and survival. More skilled individuals (human capital) (Subramaniam
& Youndt, 2005), stronger ties and relationships (social capital) (Blyler & Coff,
2003), and higher levels of institutionalized knowledge (organizational capital)
(Sher & Lee, 2004) contribute to processes that enable dynamic capabilities, such as
product development and strategic decision making (Eisenhardt & Martin, 2000).
Consequently, firms with greater IC have greater dynamic capabilities because they
are better able to enact their environments and respond and adapt to environmental
change (Eisenhardt & Santos, 2002; Wu, 2007).
H5d: Intellectual capital has a positive effect on dynamic capabilities.
The Role of Learning Culture

Barrett (1995, p. 36) defines learning cultures as . . . contexts in which members
can explore, experiment in the margins, extend capabilities, and anticipate customers latent needs. Relevant cultural attributes include management support for
the role of knowledge and encouragement for employees to interact, explore, and
seek assistance (Gold et al., 2001; Nonaka & Toyama, 2005).
The prior literature has revealed the importance of a learning culture in
the development of the stock of organizational knowledge (De Long & Fahey,
2000). Nahapiet and Ghoshal (1998) discuss how norms and other aspects of
culture facilitate IC by motivating individuals to combine knowledge. A learning
culture facilitates organizational capital, as seen in studies (Kankanhalli, Tan,
& Wei, 2005) of how related aspects (e.g., enjoyment in helping others) lead to
greater knowledge in electronic knowledge repositories (i.e., greater organizational
capital). A learning culture facilitates social capital by promoting greater and more
trusting ties among individuals (ODell & Grayson, 1998). A learning culture
also facilitates human capital, because individual knowledge is improved through
learning by doing in a culture that encourages exploration and tolerates mistakes
(Vera & Crossan, 2004), and through learning by observation in a culture that
encourages collaboration (ODell & Grayson, 1998).
H6: Learning culture is positively associated with intellectual capital.
Firm Size as a Control Variable

Large companies have been argued to have economies of scale in research and
development (R&D) activities, and thus can be efficient in achieving innovation
(Shumpeter, 1961). Large companies are more capable of nurturing their new
498
Figure 2: The detailed theoretical model.
products in the market and negotiating with external environments for complimentary assets (Teece, 1986). Therefore, we include firm size as a control variable in
examining the effects on firm performance and interim outcomes.
Figure 2 shows the detailed research model. Consistent with earlier discussion, we do not include some potential hypotheses. First, we do not hypothesize
IC or KM to directly affect firm performance. Instead, we expect their effects to be
mediated through efficiency, innovation, and dynamic capabilities. This disposition
has been supported by KBV and the dynamic capabilities approach. Second, we do
not hypothesize a learning culture to directly facilitate KM. Instead, we argue that
a learning culture directly affects IC, but only indirectly affects KM. KBV argues
that organizational knowledge comes from individuals knowledge (Grant, 1996a,
1996b; Nonaka & Toyama, 2005). Individuals with human capital may form informal networks (social capital) and contribute to organizational capital such as
licenses, documents, and electronic repositories. Organizational culture has a role
to play in these behaviors (e.g., Nonaka & Takeuchi, 1995; Lee & Choi, 2003).
By contrast, KM represents managerial processes and practices to manage knowledge. Therefore, we do not consider culture to directly affect KM. Third, we do
not hypothesize IC to directly affect innovation or efficiency. This stance is consistent with KBV and the dynamic capabilities approach in that knowledge supports
the development of knowledge-based capabilities, which deliver organizational
outcomes such as innovation and efficiency.
DATA COLLECTION AND MEASURES

Sample and Procedures
The studys sample is 1,466 large publicly listed Taiwanese firms, identified in 2006
from the Market Observation Post System maintained by the Taiwan Stock Exchange (http://emops.tse.com.tw/emops_all.htm). Most variables were measured
Hsu and Sabherwal
499
using a questionnaire survey in 2006; secondary data were used to measure efficiency and firm performance.
The questionnaire utilized existing measures, as discussed below. Two academic domain experts with Chinese and English proficiency were invited to translate the questionnaire into English. Consistency of translated and original questionnaires was ensured through back-translation (Bae & Lawler, 2000), using two
other academic experts with similar bilingual capability. The questionnaire was
pretested through meetings, each lasting about an hour, with 12 academic domain
experts and 12 senior managers from publicly listed companies in Taiwan. The
pretest was used to evaluate the content validity of the measures, ensure the clarity
of instructions and items, and make minor refinements (Tanriverdi, 2005).
We conducted an a priori power analysis to determine the required sample size (Baroudi & Orlikowski, 1989) using statistical software G Power 3
(http://www.psycho.uni-duesseldorf.de/abteilungen/aap/gpower3/). The research
model in Figure 2 includes six dependent variables, which are hypothesized to be
affected by one to four independent variables including the control variable (organization size). Assuming a small effect size of .02, an alpha of .05, and a required
statistical power of .80, a dependent variable with one, two, three, and four independent variables requires sample sizes of 395, 485, 550, and 602, respectively.
The sampling frame has a total of 1,466 companies. To obtain a sufficient number
of responses, we mailed the survey to all of these companies.
Following previous KM studies using senior managers (e.g., Gold et al.,
2001; Tanriverdi, 2005) and the advice obtained in the pretest, the questionnaire
was mailed to each firms general manager or vice general manager. We stated
the studys purpose and potential contribution, explained how to complete the
questionnaire, and promised confidentiality of responses. Five steps were taken
to increase the response rate: (i) customization of the cover letter for the specific
firm and top manager; (ii) identification of a contact person from National Science
Council (NSC), Taiwan, to highlight the surveys significance; (iii) the use of phone
calls to request cooperation after mailing the survey; (iv) the provision of a gift
voucher worth US $3, as a token of appreciation; (v) the promise, and subsequent
mailing, of a summary of results. These steps helped obtain usable responses from
533 firms (i.e., a response rate of 36.4%). However, financial information was
available for 2005 and 2006 for only 510 of these firms.
In order to evaluate whether responses from the 510 firms are sufficient for drawing conclusions of this research, we conducted post hoc power
analysis (Baroudi & Orlikowski, 1989) using statistical software G Power 3
(http://www.psycho.uni-duesseldorf.de/abteilungen/aap/gpower3/). With an alpha
of .05 and the sample size of 510, the actual statistical power for the multiple
regression varies from .72 to .89 for one to four independent variables assuming
a small effect size of .02, and is above .80 for one to four independent variables
if the effect size is .024 or above. Therefore, we believe that the sample size is
adequate.
The average age of informants was 45 years, out of which 73% were male.
Seventy-six percent of the informants were top-level managers (general manager, vice general manager, senior vice general manager, CEO, etc.), and the rest
were department or division heads. Comparing the profiles of the informants with
500
previous relevant research (e.g., Tanriverdi, 2005), these informants can be judged
as competent in providing the information needed for the survey. The average age
of the responding firms was 25 years, with an average of 785 employees. Ninetyone percent of the firms operated in manufacturing industries, while the rest were
in service industries. On average, the firms had US $515 million in assets and US
$211 million of sales in 2005.
Measures
In the survey, seven-point Likert-type scales ranging from one (strongly disagree)
to seven (strongly agree) were used to measure learning culture, IC, KM, dynamic
capabilities, and innovation. Appendix A reports all the items for these constructs.
Learning culture was measured using a six-item measure (Gold et al., 2001).
The three dimensions of IChuman capital, social capital, and organizational
capitalwere measured using five, five, and four items, respectively (Youndt et al.,
2004; Subramaniam & Youndt, 2005). The three dimensions of KMknowledge
acquisition, knowledge conversion, and knowledge applicationwere measured
using six, six, and five items, respectively (Gold et al., 2001). Dynamic capabilities
were measured using eight items (Sher & Lee, 2004). Innovation was measured
using a four-item measure (Bae & Lawler, 2000). Financial data, obtained from
the Market Observation Post System, were used to measure efficiency and firm
performance. Following Sabherwal and Sabherwal (2005), efficiency was measured
as the standardized mean of sales-to-total-assets ratio in 2005 and 2006. Firm
performance was computed as the standardized mean of four measures: return on
asset and earnings per share at the end of 2005 and 2006. Firm size was measured
as the logarithm of the number of employees, identified from the survey.
Thus, most constructs were measured using survey data, whereas two constructs in the research modelfirm efficiency and firm performanceand the
control variable (firm size) were measured using objective, secondary data. Our
research that uses survey data collected from key informants is rooted in the prior
relevant literature, including research on the relationships between innovation and
firm performance (e.g., Bae & Lawler, 2000; Damanpour, Walker, & Avellaneda,
2009), and between dynamic capabilities and firm performance (e.g., Song, Di
Benedetto, & Nason, 2007; Pavlou & El Sawy, 2011). Subjective data are capable
of providing meaningful comparisons across firms (Song, Droge, Hanvanich, &
Calantone, 2005). However, our use of firm performance, computed using financial
measures, as the final dependent variable can be unambiguous (Miller, 1988), and
can help reduce common method variance (Malik & Kotabe, 2009; Terziovski,
2010). The value of combining subjective and objective measures has been shown
in the prior empirical literature (e.g., Miller, 1998; Song et al., 2007; Damanpour
et al., 2009; Terziovski, 2010).
Validation of Measures
Learning culture, IC, KM, dynamic capabilities, and innovation were measured
using a total of 49 items. In view of the causal direction being from items to constructs, and the items being interchangeable and correlated, reflective measures
were used (Petter, Straub, & Rai, 2007). Learning culture, dynamic capabilities,
Hsu and Sabherwal
501
and innovation were modeled as first-order constructs. Consistent with the prior
literature, IC (Youndt et al., 2004) and KM (Tanriverdi, 2005) were modeled as
second-order constructs, each including the three first-order constructs identified
earlier. Thus, the measurement model (Model 1) includes two second-order constructs and nine first-order constructs (including the three first-order constructs of
IC and KM, respectively). We tested it using confirmatory factor analysis with
maximum likelihood estimates, using AMOS 4.0. Item scores were standardized
prior to confirmatory factor analysis. Correlations among pairs of item residuals
were freed based on theoretical considerations. We dropped nine items, shown in
Appendix A, due to loadings below .5 or due to cross-loadings. The remaining
40 items loaded on their respective first-order constructs as expected. Appendix A
provides the results.
Several indices were used to assess the fit of the measurement model
(Joreskog, 1978; Medsker, Williams, & Holahan, 1994). We examined the ratio of chi-square to the degrees of freedom; a ratio below three indicates a good fit
for the hypothesized model. We also examined several indices (including goodness
of fit index, or GFI, and adjusted GFI, or AGFI), for which values above .90 are
considered good, and values from .85 to .90 are considered acceptable (Medsker
et al., 1994). We also examined the root mean-square residual (RMR), for which
lower values indicate better fit and a value below .05 is considered desirable (Fulk,
Heino, Flanagin, Monge, & Bar, 2004), and the root mean square error of approximation (RMSEA), which should be below .08 for acceptable fit. As shown in
Appendix A, all the fit indices are either good or acceptable, indicating that the
data have a good fit with the measurement model.
We found support for convergent validity for all nine first-order constructs
and the two second-order constructs. Standardized loadings for the 40 items are
significant (p < .001), with the lowest t-value being 17.61. Standardized loadings
for the six first-order constructs leading to the two second-order constructs are
large and significant (p < .001), with a minimum t-value of 12.86, indicating good
convergent validity for all the focal constructs (Anderson & Gerbing, 1988). Discriminant validity was assessed in three ways (Sabherwal & Becerra-Fernandez,
2003). First, 2 difference between the measurement model and a baseline model
(with one latent construct) was significant (p .001). Second, the average variance extracted for each first-order construct exceeded .5 and the constructs shared
variance with every other construct. Finally, the confidence interval for each con+ two standard errors) with every other
structs pairwise correlation estimate (i.e.,
construct did not include 1 (Anderson & Gerbing, 1988).
Table 1 provides summary information about the measures, reliabilities,
descriptive statistics, correlations, and square roots of average variance extracted.
The standardized alphas for all the measures exceeded .80. For each of the measures
included in the confirmatory factor analysis, the composite reliability exceeded .80
and the average variance extracted exceeded .50. These results further support the
measures. In addition, we conducted further validation tests for the use of secondorder factors to represent IC and KM. These tests, and the associated results, are
described in Appendix B.
.85
(.86)
N.A.
(N.A.)
.97
(N.A.)
.89
(N.A.)
4 (4)
.94
(.94)
.88
(.88)
.83
(.92)
.88
(.93)
8 (7)
17 (13)
14 (10)
Std. Alpha
(Composite
reliability)
5.65
(.92)
2.48
(.53)
.00
(.99)
.01
(.86)
5.01
(.92)
5.60
(.84)
5.14
(.81)
4.98
(.89)
Mean (S.D.)
.07
.04
.05
.09
.12
.06
.02
.59
.59
.60
.03
.01
.78
.81
.91
.64
.89
KM
.84
IC
.64
.69
.74
LC
.11
.00
.14
.59
.82
DC
.14
.04
.12
.77
IN
.09
.02
Size
.11
EF
Notes: N = 533 for LC, IC, KM, DC, IN, and size; N = 510 for EF and FP; p < .5, p < .01, p < .001; the square root of average variance extracted
is given along the diagonal.
Efficiency
(EF)
Firm
performance
(FP)
Learning
culture (LC)
Intellectual
capital (IC)
Knowledge
management
(KM)
Dynamic
capabilities
(DC)
Innovation
(IN)
Size
# of initial
items (# of
retained
items)
Table 1: Measures, reliabilities, descriptive statistics, correlations, and square roots of average variance extracted.
502
Hsu and Sabherwal
503
Other Tests of Appropriateness of Data

In addition to the above validation of research measures, we conducted some
further tests to ascertain the suitability of our data prior to the empirical tests.
First, because data on most of the research constructs were not available
for the nonrespondents, we assessed a nonresponse bias by comparing them
across time-dated waves of respondents (Armstrong & Overton, 1977). The responses received in the first three days were classified as early and the rest as
late. No first-order construct from the measurement model differed significantly
(p > .10) across early and late responses. However, firm size differed significantly across the two groups, indicating that larger firms responded later than
smaller firms. In light of this result, we obtained the number of employees at
the end of 2005 for 925 nonrespondent firms using the Taiwan Economic Data
Bank (http://163.13.35.26:8026/tejcount/) the and 1111 Human Resource Data
Bank (http://www.1111.com.tw/). The respondent and nonrespondent firms do not
differ in size (t-value = 1.46, p > .10). Overall, the above tests do not indicate
non-response bias to be an issue.
Second, a position-bias test and an industry-bias test were conducted by
comparing the initial 49 items across two groups. In the position-bias test, top
managers and division/department heads were compared. Only one of the 49 items
was significantly different (at p .05). In the industry-bias test, manufacturing
and service industries were compared. Three items were significantly different (at
p .05). Based on a chi-square test, the number of significantly different items
does not significantly exceed the number (2.5) expected to be different based on
chance. Thus, there is no reason to believe that position or industry has biased
response patterns.
Finally, we examined whether the common method variance is a serious issue.
This was tested using Harmans one-factor test (Podsakoff & Organ, 1986). We
employed Harmans one-factor test by conducting a confirmatory factor analysis
(Sabherwal & Becerra-Fernandez, 2003). A single-factor model was constructed
by linking all 40 items to a single latent variable, positing a single factor underlying
the focal constructs. This model did not fit the data ( 2 /d.f. = 5.38, GFI = .71,
AGFI = .67, RMSEA = .09). Moreover, an exploratory factor analysis using the
40 items produced five factors. Thus, the results did not suggest the presence of
common method variance.
ANALYSES AND RESULTS

Data analysis was conducted using the detailed theoretical model shown in Figure 2,
with two exogenous latent constructs (learning culture and firm size), and six latent
endogenous constructs (IC, KM, dynamic capabilities, innovation, firm efficiency,
and firm performance). For the variables measured using the questionnaire survey,
the large sample size enabled us to use the individual items in the structural model.
We used a summated score for firm performance (which was measured using two
measures from the secondary data, i.e., return on asset and earnings per share,
and had missing values on one of the measures for 23 firms) and the single item
measure of firm size. To adjust for measurement errors, we set the path from
504
Figure 3: Results for the theoretical model.
Note: p < .05, p < .01, p < .001; one-tailed tests.
the latent variable to each measure to equal the square root of scale reliability
(standardized alpha), and the error variances to the variance of the computed
variable multiplied by 1 minus reliability (Joreskog & Sorbom, 1989). We set the
reliability of firm size, which was evaluated using a single measure, to .90 (Davy
& Shipper, 1993).
We evaluated specific paths in the structural model using t-statistics. We
modified the structural model based on theoretical considerations (Marcoulides
& Heck, 1993) and modification indices (MIs) of 10 or more (Joreskog, 1978).
The fit for the theoretical model was unsatisfactory, with AGFI below .85 and
four nonsignificant paths: KM to IC (H2, t = .82); KM to dynamic capabilities
(H5c, t = .27); KM to efficiency (H3b, t = .56); and dynamic capabilities to firm
performance (H5a, t = .31) (Figure 3). We tried to improve the model in a stepwise
fashion by constraining the path with the lowest absolute value of the t-statistic at
each step. This resulted in all four of the above paths being dropped, indicating
nonsupport for H2, H3b, H5a, and H5c. In the resulting model, one excluded path,
from learning culture to innovation (MI = 11.91), had an MI of 10 or more. This
path had been excluded expecting the effect of learning culture on innovation to
be mediated through IC and KM, but it is consistent with the prior literature on
the importance of organizational culture for innovation (e.g., Cameron & Quinn,
1999). Upon freeing this path, the resulting model performed well, in terms of the
fit indices ( 2 = 1663.24, d.f. = 807, 2 /d.f. = 2.06, GFI = .87, AGFI = .85,
NFI = .91, RFI = .90, CFI = .95, IFI = .95, RMR = .041, RMSEA = .046) and
specific paths, as shown in Figure 4. In all the tested models, the paths from IC and
KM to their three first-order factors were significant (p .001). Some standardized
coefficients were large, approaching 1, but the prior literature indicates that this is
Hsu and Sabherwal
505
Figure 4: The emergent model.
Note: To simplify presentation, we have excluded non-significant hypothesized paths from KM to IC (H2),
KM to efficiency (H2b), KM to dynamic capabilities (H5c), and dynamic capabilities to organizational
performance (H5a).
p < .05, p < .01, p < .001; one-tailed tests.
not a source of concern, and that standardized coefficients above 1 are also feasible
(Deegan, 1978; Joreskog, 1999).
To summarize, the empirical study supports the theoretical model. Seven
of the 11 research hypotheses are supported, with H2, H3b, H5a, and H5c not
supported. There is also one unexpected pathfrom learning culture to innovation.
To further validate the results (e.g., that IC and KM do not directly affect firm
performance), we conducted detailed tests of the mediated effects. Appendix C
reports the procedure and the results. We created a base model that includes all the
direct and indirect paths. Then, to test each mediated effect, we created a model
constraining the corresponding path from independent to dependent construct to
zero, and then compared this constrained model with the base model. The results
suggest that IC and KM do not directly affect firm performance. Instead, the
effects of IC and KM on firm performance could be mediated by variables such as
innovation and dynamic capabilities.
DISCUSSION
Drawing on the literature on KBV (Grant 1996a, 1996b; Nonaka & Toyama, 2005)
and relevant theoretical perspectives (e.g., Teece & Pisano, 1994; Nahapiet &
Ghoshal, 1998), we set out to address three research questions:
(i) When both KM and IC are considered, through which mediating factors
do they affect firm performance?
(ii) How do KM and IC affect each other?
(iii) How does organizational culture affect IC?
These research questions have been addressed using an empirical study involving
survey data from 510 firms.
506
Furthermore, based on tests of statistical power, validity of measures, and

tests of potential biases, the data seem to be appropriate for addressing our three
research questions. Therefore, the research findings should have implications for
research as well as practice. We now discuss the results in the light of the three
research questions addressed in this article and examine their implications for
research.
Implications for Research

Research question 1: When both KM and IC are considered, through which
mediating factors do they affect firm performance?
We argued that IC and KM do not directly affect performance, but instead they
affect it indirectly through dynamic capabilities, innovation, and efficiency. The
results indicate that neither IC nor KM affects performance directly, as expected.
The results for one interim outcome, innovation, are as expected: innovation
affects firm performance (support for H4a) and depends on KM (support for H4b).
However, the results for the other two interim outcomes, efficiency and dynamic
capabilities, differ from expectations. Efficiency facilitates performance (support
for H3a) but is affected by neither KM (nonsupport for H3b) nor IC. Finally, dynamic capabilities depend on IC (support for H5d) but not KM (nonsupport for
H5c). Dynamic capabilities do not directly affect firm performance (nonsupport
for H5a), but they indirectly facilitate it through innovation (support for H5b).
The article presents only an early attempt to understand the potential causal
mechanisms by which IC improves firm performance. Readers are advised to interpret our research findings with caution because the causal mechanisms cannot
be fully investigated due to a lack of time dimension in the research design. Our
findings suggest two distinct paths by which IC improves firm performance. IC
can enable dynamic capabilities, which result in innovation, which improves firm
performance. This finding supports general assertions of dynamic capabilities literature. Dynamic capabilities build on knowledge resources of the firm (Teece &
Pisano, 1994; Eisenhardt & Martin, 2000; Zollo & Winter, 2002), and enable the integration, building, and reconfiguration of resources and competencies, which can
be best represented as innovation (Teece & Pisano, 1994; Helfat, 1997). Financial
performance follows.
The second path through which IC seems to enable firm performance is by
leading to better KM, followed by greater innovation. The effect of IC on KM,
which has been rooted in KBV and the theories of absorptive capacity and social
capital, has been discussed above. Moreover, consistent with the stance adopted
by the KM literature, KM should affect firm performance through intermediate
outcomes (Lee & Choi, 2003). The effect of IC on firm performance through
innovation (either through dynamic capabilities or through KM) is also consistent
with the KBV literature (Nonaka & Takeuchi, 1995; Eisenhardt & Santos, 2002).
However, our results depart from prior arguments concerning the role of
KM in developing dynamic capabilities (e.g., Eisenhardt & Martin, 2000; Zollo
& Winter, 2002). The adequacy of statistical power analysis leads to further confidence in this nonsignificant finding. Indeed, both IC and KM can be important
bases of dynamic capabilities, and individually, IC and KM are highly correlated
Hsu and Sabherwal
507
with dynamic capabilities. However, when their effects on dynamic capabilities are
tested at the same time, this study has suggested that the effect of KM on dynamic
capabilities is less important than that of IC.
Thus, the results indicate that when IC is explicitly included in the empirical
study, improvements in dynamic capabilities result from IC and not from KM
(which seems to facilitate innovation but not dynamic capabilities). This provides
support for the view of KM as an organizational capability (Gold et al., 2001)
that might parallel but not facilitate dynamic capabilities (Esterby-Smith & Prieto,
2008). Both KM and dynamic capabilities are supported by the organizations
IC. In other words, both KM and dynamic capabilities depend on whether the
employees are highly talented and skilled (human capital), are capable of forming informal networks for knowledge exchanges and collaboration (social capital), and have a sophisticated system of organizational memories (organizational
capital).
The view of KM as a capability also extends Subramaniam and Youndts
(2005) findings about IC positively affecting innovative capabilities. The article
establishes that IC contributes to KM capabilities as well as dynamic capabilities,
both of which lead to innovation. The acquisition of new knowledge, the conversion
of absorbed knowledge, and the application of the converted knowledge seem to
produce capabilities of the firm that produce innovation.
KM processes have not been found to improve firm efficiency, thereby differing from the claims of improved efficiency as one of the benefits of KM processes, such as knowledge conversion (Grant, 1996a; Davenport & Klahr, 1998)
and knowledge utilization through directions and rules (Demsetz, 1991; Conner
& Prahalad, 1996). There are three possible explanations of this nonsignificant
result. First, according to Porter (1980), the pursuits of innovation and efficiency
may be disparate undertakings. Our results may simply suggest that the sampled
companies placed innovation with a higher priority over efficiency. Second, this
nonsignificant result may be due to the financial nature of the measure used for
efficiency, which may depend on a number of other factors, such as age and quality
of manufacturing equipment, that are beyond the scope of this study. Third, this
nonsignificant result may be due to the empirical context of this research. Over
the years the Taiwanese government has been enforcing a policy that encourages
and funds companies in their innovative activities. The policy aims to improve
the competitiveness of Taiwanese companies in the global market. This may have
caused a firms innovation, but not efficiency, to depend on the quality of the firms
KM efforts.
Research question 2: How do KM and IC affect each other?

The reciprocal effects of IC and KM have been argued by KBV (Grant, 1996a,
1996b) and relevant theories (e.g., Nahapiet & Ghoshal, 1998). We explicitly
included and empirically tested both causal directions of the relationship between
IC and KM. However, the results are somewhat surprising: IC contributes to KM
(support for H1), but KM does not affect IC (nonsupport for H2). It seems that, at
least in the short term, the effects of knowledge on KM may be more important
than the effects of KM on knowledge.
508
The results support the view that existing organizational knowledge facilitates
KM processes (e.g., Grant, 1996a, 1996b; Nonaka & Toyama, 2005). Such effect of
knowledge on KM processes is also consistent with prior arguments in the theory
of absorptive capacity that existing knowledge facilitates knowledge absorption,
assimilation, and application (Cohen & Levinthal, 1990), and prior arguments in
theory of social capital that existing knowledge facilitates knowledge exchange
and combination (Nahapiet & Ghoshal, 1998). Thus, the article contributes to
KBV, which recognizes the firm as a unique bundle of knowledge resources and
capabilities (e.g., Grant, 1996a; Eisenhardt & Santos, 2002), but does not explicitly
examine the effects of knowledge on KM processes that enable organizational
resources. The results suggest that it would be useful to explicitly incorporate IC
in future research on KM and its consequences.
Research question 3: How does organizational culture affect intellectual

capital?
We expected a learning culture to directly facilitate IC but not KM. The results
are as expected: a learning culture facilitates IC (support for H6) but not KM.
Moreover, when we added a path from learning culture to KM in the revised
model shown in Figure 4, it was not significant (.068; t = 1.372; p > .05, onetailed). The results are consistent with the argument that a learning culture can
help accumulate organizational knowledge (Nahapiet & Ghoshal, 1998; De Long
& Fahey, 2000). The results also indicate that the effect of learning culture on
KM is through IC. Thus, IC seems to be the key mediator between organizational
culture and KM and should be included in further research into the relationship
between organizational culture and KM (Lee & Choi, 2003).
The results indicate that the learning culture also directly affects innovation.
This path was not included in the theoretical model as we expected the effect
of learning culture on innovation to be mediated through IC and KM. However,
the prior literature does suggest that a learning culture improves organizational
innovativeness (e.g., Cameron & Quinn, 1999). This emergent path indicates that
a learning culture leads to greater innovation, over and above its indirect effect
through IC. Thus, a learning culture contributes to both IC and innovation.
Limitations
The above results need to be considered in view of the studys limitations. First,
this study is cross-sectional in nature, which may have reduced the importance
of relationships that might take time to occur (perhaps, for example, the effect of
KM on IC). Although there is precedence for using cross-sectional data to test
reciprocal effects (e.g., Sabherwal, 1999), a longitudinal study would have been
more effective for investigating the reciprocal relationship between KM and IC.
Second, the study was conducted using a sample of firms in Taiwan, which
have distinct cultural traits that have been identified to inhibit knowledge sharing
(Hsu, 2006). Further research is needed to examine whether our findings generalize
to other cultural contexts with less salient impediments to knowledge sharing.
Third, theoretical considerations, concerns for parsimony, and limitations of
secondary data in Taiwan led us to construct a theoretical model that excludes
Hsu and Sabherwal
509
moderating effects (e.g., firm size on the relationship between KM and innovation)
and other potentially important variables, such as organizational structure. Inclusion of moderating effects and additional antecedent variables may have provided
somewhat different results. We also acknowledge that firm performance can be
affected by multiple factors that could not all be included in this study. This is
not unusual in field studies where some of the variables affecting a dependent
variable have to be excluded to focus on the constructs of greatest relevance to the
study. Consistent with prior field studies using survey data, we develop and test
our research model from a theoretical perspective in order to test the focal theory.
Finally, the limitations of secondary data further prevented us from devising
indicator-type measures for innovation (e.g., Hess & Rothaermel, 2011), or even
dynamic capabilities (e.g., Sirmon & Hitt, 2009) that will offer a complementary
approach in testing the two constructs as interim outcomes of IC and KM, as
has been attempted for efficiency here in the research. Inclusion of the measures
developed from secondary data may improve the robustness of our findings.
Implications for Practice

Despite the above limitations, the results of this study offer some potentially valuable implications for practice. First, efficiency does not depend either directly or
indirectly on any of the other constructs, including IC, KM, and learning culture. Therefore, managers of firms emphasizing efficiency should be cautious in
investing in these three knowledge-related aspects, and should specifically seek
opportunities (such as systems and routines that enable knowledge reuse) that can
enhance efficiency.
Second, our study highlights the importance of IC and a learning culture,
especially for firms that are not competing based on efficiency. IC benefits the
organization through improvement in dynamic capabilities and KM, both of which
benefit the organization by enabling innovation. The emergent results also highlight
the importance of learning culture in attaining innovation. Therefore, the results
should encourage managers of firms that are seeking to innovate through investing
in IC and a learning culture.
Third, our study indicates that the development of IC and a learning culture
may be more important than investments in KM processes. IC affects KM and dynamic capabilities, but KM affects neither IC nor dynamic capabilities. A learning
culture affects IC and innovation. Therefore, managers should use mechanisms
(e.g., manuals, systems) to store individual knowledge (organizational capital),
hire and develop high-quality employees (human capital), who are willing to collaborate with each other (social capital), and promote risk taking and trust (a
learning culture). These actions would also indirectly facilitate innovation, KM,
and dynamic capabilities.
Finally, even though KM depends on IC, the importance of KM is worth
advising here. KM facilitates innovation, which improves firm performance. An
organization seeking to establish competitive advantage based on innovation should
realize that KM can help achieve this goal. This research has proposed that KM
should include the following processes for it to result in outcomes of innovation:
knowledge acquisition, conversion, and application.
510
Implications for Future Research

Future research on knowledge, or IC, and its management (i.e., KM) could benefit
from this study in at least two important ways. First, our study shows the benefits
of simultaneously examining both IC and KM, and future studies in this area
would benefit from similarly incorporating both aspects. Doing so not only helps
in examining the relationships among these two dimensions but also helps in
reducing the likelihood of potentially misleading results that may be obtained if
either IC or KM is excluded. Simultaneous incorporation of IC and KM in this
study enabled us to conclude that IC facilitates KM (support for H1) but KM does
not facilitate IC (nonsupport for H2). The results suggest that the current state
of IC affects current KM processes, but current KM processes might not affect
current IC, possibly because current KM processes might affect future IC over
time. However, these findings may be due to the cross-sectional nature of this
study and how IC and KM might evolve; a longitudinal study, incorporating the
potential lagged effects of KM, would help in further examining the relationship
between IC and KM. Simultaneously examining IC and KM also enabled us to
conclude that a learning culture facilitates IC (support for H6) but, as expected,
not KM.
Second, future studies on knowledge and its management would benefit from
adopting a more comprehensive view of the various aspects of performance consequences. Instead of examining either the effects on overall firm performance or
the effects on intermediate outcomes such as innovation, efficiency, and dynamic
capabilities, this study incorporated them both. This produced results that question
some conclusions reached in prior studies where firm performance and intermediate effects were not studied simultaneously. More specifically, the results of this
study do not support the expected effect of KM on efficiency (H3b) or the two relationships involving dynamic capabilities: its expected effect on firm performance
(H5a), and its expected dependence on KM (H5c). Moreover, the simulatenous
incorporation of firm performance and intermediate effects revealed the intuitively
appealing but unexpected effect of learning culture on innovation. Further research
is needed to better understand whether these results can be generalized, or are due
to the research methods, measures, and samples. Nevertheless, this study indicates
that future research would benefit from similar simultaneous consideration of firm
performance and intermediate effects.
CONCLUSION
We have proposed and empirically tested a comprehensive model of the mutual
relationship between IC and KM; their effects on firm performance, mediated by
efficiency, innovation, and dynamic capabilities; and their dependence on the organizations culture. We have provided some results that are consistent with the prior
literature: IC facilitates KM and dynamic capabilities; KM, a learning culture, and
dynamic capabilities facilitate innovation; and innovation and efficiency facilitate
firm performance. However, our study also provides some surprises: neither IC
nor KM affects efficiency; KM also does not affect IC or dynamic capabilities;
and dynamic capabilities do not directly affect firm performance. Thus, this study
contributes to the literature by providing insights into the relationship between IC
Hsu and Sabherwal
511
and KM, as well as their consequences. It also raises some interesting questions for
future research on how organizations benefit from the management of knowledge
and IC.
REFERENCES
Alavi, M., & Leidner, D. E. (2001). Knowledge management and knowledge management systems: Conceptual foundations and research issues. MIS Quarterly, 25(1), 107136.
Amit R., & Schoemaker, P. J. H. (1993). Strategic assets and organizational rents.
Strategic Management Journal, 14(1), 3346.
Anderson, J. C. (1987). An approach for confirmatory measurement and structural
equation modeling of organizational properties. Management Science, 33(4),
525541.
Anderson, J. C., & Gerbing, D. W. (1988). Structural equation modeling in practice: A review and recommended two-step approach. Psychological Bulletin,
103(3), 411423.
Argote, L., & Ingram, P. (2000). Knowledge transfer: A basis for competitive
advantage in firms. Organizational Behavior and Human Decision Processes,
82(1), 150169.
Argote, L., McEvily, B., & Reagans, R. (2003). Managing knowledge in organizations: An integrative framework and review of emerging themes. Management Science, 49(4), 571582.
Armstrong, J. S., & Overton, T. (1977). Estimating nonresponse bias in mail
surveys. Journal of Marketing Research, 14(3), 396402.
Bae, J., & Lawler, J. J. (2000). Organizational and HRM strategies in Korea: Impact
on firm performance in an emerging economy. Academy of Management
Journal, 43(3), 502517.
Bagozzi, R. P., Yi, Y., & Phillips, L. W. (1991). Assessing construct validity in
organizational research. Administrative Science Quarterly, 36(3), 421458.
Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction
in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51(6), 11731182.
Baroudi, J. J., & Orlikowski, W. J. (1989). The problem of statistical power in MIS
research. MIS Quarterly, 13(1), 87106.
Barrett, F. J. (1995). Creating appreciative learning cultures. Organizational Dynamics, 24(2), 3649.
Blyler, M., & Coff, R. W. (2003). Dynamic capabilities, social capital, and rent
appropriation: Ties that split pies. Strategic Management Journal, 24(7),
677686.
Brown, S. L., & Eisenhardt, K. M. (1997). The art of continuous change: Linking
complexity theory and time-paced evolution in relentlessly shifting organizations. Administrative Science Quarterly, 42(1), 134.
512
Cameron, K. S., & Quinn, R. E. (1999). Diagnosing and changing organizational

culture. New York: Addison-Wesley.
Carmines, E. G., & McIver, S. P. (1981). Analyzing models with unobserved
variables: Analysis of covariance structures. In G. W. Bohrnstedt & E. F.
Borgatta (Eds.), Social measurement: Current issues. Beverly Hills, CA:
Sage, 65115.
Chen, C.-J., Shih, H.-A., & Yang, S.-Y. (2009). The role of intellectual capital in
knowledge transfer. IEEE Transactions on Engineering Management, 56(3),
402411.
Cohen, W., & Levinthal, D. (1990). Absorptive capability: A new perspective
on learning and innovation. Administrative Science Quarterly, 35(1), 128
152.
Conner, K. R., & Prahalad, C. K. (1996). A resource-based theory of the firm:
Knowledge versus opportunism. Organization Science, 7(5), 477501.
Cyert, R. M., & March, J. G. (1963). A behavioral theory of the firm. Englewood
Cliffs, NJ: Prentice-Hall.
Damanpour, F., Walker, R. M., & Avellaneda C. N. (2009). Combinative effects
of innovation types and organizational performance: A longitudinal study of
service organizations. Journal of Management Studies, 46(4), 650675.
Davenport, T. H., & Klahr, P. (1998). Managing customer support knowledge.
California Management Review, 40(3), 195208.
Davy, J. A., & Shipper, F. (1993). Voter behavior in union certification elections:
A longitudinal study. Academy of Management Journal, 36(1), 187199.
Decarolis, D. M., & Deeds, D. L. (1999). The impact of stocks and flows of
organizational knowledge on firm performance: An empirical investigation
of the biotechnology industry. Strategic Management Journal, 20(10), 953
968.
Deegan, Jr., J. (1978). On the occurrence of standardized regression coefficients
greater than one. Education and Psychological Measurement, 38(4), 873
888.
De Long, D. W., & L. Fahey, L. (2000). Diagnosing cultural barriers to knowledge
management. Academy of Management Executive, 14(4), 113127.
Demsetz, H. (1991). The theory of the firm revisited. In O. E. Williamson & S.
G. Winter (Eds.), The nature of the firm. New York, NY: Oxford University
Press, 159178.
Dierickx, I., & Cool, K. (1989). Asset stock accumulation and sustainability of
competitive advantage. Management Science, 35(12), 15041513.
Duncan, R., & Weiss, A. (1979). Organizational learning: Implications for organizational design. In B. Staw (Ed.), Research in organizational behavior, vol.
1. Greenwich, CT: JAI Press, 75123.
Edvinsson, L., & Malone, M. S. (1997). Intellectual capital: Realizing your companys true value by finding its hidden brainpower. New York, NY: Harper
Business.
Hsu and Sabherwal
513
Easterby-Smith, M., & Prieto, I. M. (2008). Dynamic capabilities and knowledge

management: An integrative role for learning? British Journal of Management, 19(3), 235249.
Eisenhardt, K. M., & Martin, J. A. (2000). Dynamic capabilities: What are they?
Eisenhardt, K. M., & Santos, F. M. (2002). Knowledge-based view: A new theory
of strategy? In A. Pettigrew, H. Thomas, & R. Whittington (Eds.), Handbook
of strategy and management. London, UK: Sage, 139164.
Fulk, J., Heino, R., Flanagin, A. J., Monge, P. R., & Bar, F. (2004). A test of the individual action model for organizational information commons. Organization
Science, 15(5), 569585.
Gold, A. H., Malhotra, A., & Segars, A. H. (2001). Knowledge management: An
organizational capabilities perspective. Journal of Management Information
Systems, 18(1), 185214.
Grant, R. M. (1996a). Toward a knowledge-based theory of the firm. Strategic
Management Journal, 17(Winter Special Issue), 109122.
Grant, R. M. (1996b). Prospering in dynamically-competitive environments: Organizational capability as knowledge integration. Organization Science, 7(4),
375387.
Haas, M. R., & Hansen, M. T. (2005). When using knowledge can hurt performance: The value of organizational capabilities in a management consulting
company. Strategic Management Journal, 26(1), 124.
Hamel, G., & Prahalad, C. K. (1993). Strategy as stretch and leverage. Harvard
Business Review, 71(2), 7584.
Helfat, C. E. (1997). Know-how and asset complementarity and dynamic capability
accumulation: The case of R&D. Strategic Management Journal, 18(5), 339
360.
Hess, A. M., & Rothaermel, F. T. (2011). When are assets complementary? Star
scientists, strategic alliances, and innovation in the pharmaceutical industry.
Hsu, I.-C. (2006). Enhancing employee tendencies to share knowledge: Case studies of nine company cases in Taiwan. International Journal of Information
Management, 26(4), 326338.
Huber, G. P. (1991). Organizational learning: The contributing processes and literatures. Organization Science, 2(1), 7187.
Hult, G. T. M., Ketchen, D. J., & Slater, S. F. (2004). Information processing,
knowledge development, and strategic supply chain performance. Academy
Management Journal, 47(2), 241253.
Joreskog, K. G. (1978). Structural analysis of covariance and correlation matrices.
Psychometrika, 43(4), 443477.
Joreskog, K. G. (1999). How large can a standardized coefficient be? Accessed May 21, 2008, available at http://www.ssicentral.com/lisrel/techdocs/
HowLargeCanaStandardizedCoefﬁcientbe.pdf.
514
Joreskog, K. G., & Sorbom, D. (1989). LISREL 7: A guide to the program and
applications. Chicago, IL: SPSSX.
Kankanhalli, A., Tan, B. C. Y., & Wei, K.-K. (2005). Contributing knowledge to
electronic knowledge repositories: An empirical investigation. MIS Quarterly, 29(1), 113143.
Kline, R. B. (2005). Principles and practice of structural equation modeling (2nd
ed.). New York, NY: Guilford Press.
Kogut, B., & Zander, U. (1992). Knowledge of the firm, combinative capabilities, and the replication of technology. Organization Science, 3(3), 383
397.
Kulkarni, U., Ravindran, S., & Freeze, R. (20062007). A knowledge management
success model: Theoretical development and empirical validation. Journal
of Management Information Systems, 23(3), 309347.
Lee, H., & Choi, B. (2003). Knowledge management enablers, processes, and
organizational performance: An integrative view and empirical examination.
Journal of Management Information Systems, 20(1), 179228.
Levitt, B., & March, J. G. (1988). Organizational learning. Annual Review of
Sociology, 14(2), 319340.
Li, H., & Atuahene-Gima, K. (2001). Product innovation strategy and the performance of new technology ventures in China. Academy of Management
Journal, 44(6), 11231134.
Lynn, G. S., Reilly, R. R., & Akgun, A. E. (2000). Knowledge management in new
product teams: Practices and outcomes. IEEE Transactions on Engineering
Management, 47(2), 221231.
Mackinnon, D. P., Lockwood, C. M., & Williams, J. (2004). Confidence limits
for the indirect effect: Distribution of the product and resampling methods.
Multivariate Behavioral Research, 39(1), 99128.
Majchrzak, A., Cooper, L. P., & Neece, O. E. (2004). Knowledge reuse for innovation. Management Science, 50(2), 174188.
Malik, O. R., & Kotabe, M. (2009). Dynamic capabilities, government policies,
and performance in firms from emerging economies: Evidence from India
and Pakistan. Journal of Management Studies, 46(3), 421450.
March, J. (1991). Exploration and exploitation in organizational learning. Organization Science, 2(1), 7187.
Marcoulides, G. A., & Heck, R. H. (1993). Organizational culture and performance:
Proposing and testing a model. Organization Science, 4(2), 209225.
Marsh, H. W., & Hocevar, D. (1985). Application of confirmatory factor analysis
to the study of self concept: First and higher order factors models and their
invariance across groups. Psychological Bulletin, 97(3), 562582.
McDaniel, S. W., & Kolari, J. W. (1987). Marketing strategy implications of the
Miles and Snow strategic typology. Journal of Marketing, 51(4), 1930.
Medsker, G. J., Williams, L. J., & Holahan, P. J. (1994). A review of current
practices for evaluating causal models in organizational behavior and human
resources management research. Journal of Management, 20(2), 439564.
Hsu and Sabherwal
515
Miller, D. (1988). Relating Porters business strategies to environments and structure: Analysis and performance implications. Academy of Management Journal, 31(2), 280308.
Nahapiet, J., & Ghoshal, S. (1998). Social capital, intellectual capital, and the
organizational advantage. Academy of Management Review, 23(2), 242266.
Nonaka, I. (1994). A dynamic theory of organizational knowledge creation. Organization Science, 5(1), 1437.
Nonaka, I., & Takeuchi, H. (1995). The knowledge-creating company. New York,
NY: Oxford University Press.
Nonaka, I., & Toyama, R. (2005). The theory of the knowledge-creating firm:
Subjectivity, objectivity, and synthesis. Industrial and Corporate Change,
14(3), 419436.
ODell, C., & Grayson, Jr., C. J. (1998). If only we knew what we know. New York,
NY: The Free Press.
OReilly, C., & Tushman, M. (2008). Ambidexterity as a dynamic capability:
Resolving the innovators dilemma. Research in Organizational Behavior,
28, 185206.
Olivera, F., Goodman, P. S., & Tan, S. S.-L. (2008). Contribution behaviors in
distributed environments. MIS Quarterly, 32(1), 2342.
Pavlou, P. A., & El Sawy, O. A. (2011). Understanding the elusive black box of
dynamic capabilities. Decision Sciences, 42(1), 239273.
Petter, S., Straub, D., & Rai, A. (2007). Specifying formative constructs in information systems research. MIS Quarterly, 31(4), 623656.
Podsakoff, P. M., & Organ, D. W. (1986). Self-reports in organizational research:
Problems and prospects. Journal of Management, 12(4), 531544.
Porter, M. E. (1980). Competitive advantage: Techniques for analyzing industries
and competitors. New York, NY: Free Press.
Qureshi, I., Fang, Y., Ramesy, E., McCole, P., Ibboston, P., & Compeau, D. (2009).
Understanding online customer repurchasing intention and the mediating role
of trust: An empirical investigation in two developed countries. European
Journal of Information Systems, 18(3), 205222.
Roos, J., Roos, G, Edvinsson, L., & Dragonetti, N. C. (1998). Intellectual capital: Navigating in the new business landscape. New York, NY: New York
University Press.
Sabherwal, R. (1999). The relationship between information system planning sophistication and information system success: An empirical assessment. Decision Sciences, 30(1), 137167.
Sabherwal, R., & Becerra-Fernandez, I. (2003). An empirical study of the effects of
knowledge management tools at individual, group, and organizational levels.
Decision Sciences, 34(2), 225260.
Sabherwal, R., & Sabherwal, S. (2005). Knowledge management using information technology: Determinants of short-term impact on firm value. Decision
Sciences, 36(4), 531566.
516
Santoro, M. D., & Bierly, III, P. E. (2006). Facilitators of knowledge transfer

in university-industry collaborations: A knowledge-based perspective. IEEE
Transactions on Engineering Management, 53(4), 495507.
Santoro, M. D., & Saparito, P. A. (2006). Self-interest assumption and relational
trust in university-industry knowledge transfers. IEEE Transactions on Engineering Management, 53(3), 335347.
Schumpeter, J. A. (1961). Theory of economic development. New York, NY:
Oxford University Press.
Sher, P. J., & Lee, V. C. (2004). Information technology as a facilitator for enhancing dynamic capabilities through knowledge management. Information and
Management, 41(8), 933945.
Shrout, P. E., & Bolger, N. (2002). Mediation in experimental and nonexperimental studies: New procedures and recommendations. Psychological Methods,
7(4), 422445.
Simon, H. (1965). Administrative behavior. New York, NY: Free Press.
Simonin, B. (1997). The importance of collaborative know-how: An empirical
test of the learning organization. Academy of Management Journal, 40(5),
509533.
Sirmon, D. G., & Hitt, M. A. (2009). Contingencies within dynamic managerial
capabilities: Interdependent effects of resource investment and deployment
on firm performance. Strategic Management Journal, 30(13), 13751394.
Slaughter, S. A., & Kirsch, L. J. (2006). The effectiveness of knowledge transfer portfolios in software process improvement: A field study. Information
Systems Research, 17(3), 301320.
Smith, K. G., Collins, C. J., & Clark, K. D. (2005). Existing knowledge, knowledge creation capability, and the rate of new product introduction in hightechnology firms. Academy of Management Journal, 48(2), 346357.
Song, M., Di Benedetto, C., & Nason, R. W. (2007). Capabilities and financial
performance: The moderating effect of strategic type. Journal of the Academy
of Marketing Science, 35(1), 1834.
Song, M., Droge, C., Hanvanich, S., & Calantone, R. (2005). Marketing and
technology resource complementarity: An analysis of their interaction effect
in two environmental contexts. Strategic Management Journal, 26(4), 259
276.
Stewart, T. A. (1997). Intellectual capital. New York, NY: Doubleday-Currency.
Subramaniam, M., & Youndt, M. A. (2005). The influence of intellectual capital
on the types of innovative capabilities. Academy of Management Journal,
48(3), 450463.
Tanriverdi, H. (2005). Information technology relatedness, knowledge management capability, and performance of multibusiness firms. MIS Quarterly,
29(2), 311334.
Teece, D. J. (1986). Profiting from technological innovation: Implications for integration, collaboration, licensing and public policy. Research Policy, 15(6),
285305.
Hsu and Sabherwal
517
Teece, D. J., & Pisano, G. (1994). The dynamic capabilities of firms: An introduction. Industrial and Corporate Change, 3(3), 537556.
Teece, D. J., Pisano, G., & Shuen, A. (1997). Dynamic capabilities and strategic
management. Strategic Management Journal, 18(7), 509533.
Terziovski, M. (2010). Innovation practice and its performance implications in
small and medium enterprises (SMEs) in the manufacturing sector: A
resource-based view. Strategic Management Journal, 31(8), 892902.
Venkatraman, N. (1990). Performance implications of strategic coalignment: A
methodological perspective. Journal of Management Studies, 27(1), 1941.
Vera, D., & Crossan, M. (2004). Strategic leadership and organizational learning.
Academy of Management Review, 29(2), 222240.
Verkasalo, M., & Lappalainen, P. (1998). A method of measuring the efficiency
of the knowledge utilization process. IEEE Transactions on Engineering
Management, 45(4), 414423.
Wiig, K. M. (1997). Integrating intellectual capital and knowledge management.
Long Range Planning, 30(3), 399405.
Wu, L.-Y. (2006). Resources, dynamic capabilities and performance in a dynamic
environment: Perceptions in Taiwanese IT enterprises. Information and Management, 43(4), 447454.
Wu, L.-Y. (2007). Entrepreneurial resources, dynamic capabilities, and start-up
performance of Taiwans high-tech firms. Journal of Business Research,
60(5), 549555.
Yang, C., & Chen, L. (2007). Can organizational knowledge capabilities affect
knowledge sharing behavior? Journal of Information Science, 33(1), 95
109.
Youndt, M. A., Subramaniam, M., & Snell, S. A. (2004). Intellectual capital
profiles: An examination of investments and returns. Journal of Management
Studies, 41(2), 335361.
Zollo, M., & Winter, S. G. (2002). Deliberate learning and the evolution of dynamic
capabilities. Organization Science, 13(3), 339351.
APPENDIX A: RESULTS FOR THE MEASUREMENT MODELa

Itemsb,c
Scale 1: Learning Culture (Standardized alpha = .88, Composite
reliabilitye = .88)
My company encourages employees to explore and experiment.
My company encourages employees to ask others for assistance when
needed.
My company encourages employees to interact with other groups.
My company clearly states its overall organizational vision and
objectives.
Factor Loadingsd
.696
.736
.813
.764
518
Itemsb,c
My company thinks the benefits of sharing knowledge outweigh the
costs.
Senior management in my company clearly supports the role of
knowledge in my companys success.
Scale 2: Intellectual Capital (Standardized alpha = .83, Composite
reliability = .92)
Subscale 1: Human Capital (Standardized alpha = .84, Composite
reliability = .83)
My companys employees are widely considered competent in the
industry.
My companys employees are creative and bright.
My companys employees are experts in their particular jobs and
function.
My companys employees are highly skilled.
My companys employees develop new ideas and knowledge.
Subscale 2: Social Capital (Standardized alpha = .92, Composite
reliability = .92)
My companys employees are skilled at collaborating with each other
to diagnose and solve problems.
My companys employees share information and learn from one
another.
My companys employees interact and exchange ideas with people
from different areas of the company.
My companys employees apply knowledge from one area of the
company to problems and opportunities that arise in another.
My companys employees partner with customers, suppliers, alliance
partners, etc., to develop solutions.
Subscale 3: Organizational Capital (Standardized alpha = .88,
Composite reliability = .88)
Much of my companys knowledge is contained in manuals,
databases, etc.
My companys culture (stories, rituals) contains valuable ideas, ways
of doing business, etc.
My company embeds much of its knowledge and information in
structures, systems, and processes.
My company uses patents and licenses as a way to store knowledge.
Scale 3: Knowledge Management (Standardized alpha = .88,
Subscale 1: Knowledge Acquisition (Standardized alpha = .84,
My company has processes for exchanging knowledge with our
business partners.
My company has processes for acquiring knowledge about new
products/services within our industry.
My company has processes for acquiring knowledge about
competitors in the industry.
My company has processes for generating new knowledge from
existing knowledge.
My company uses feedback from projects to improve subsequent
projects.
Factor Loadingsd
.710
.718
.831
.702
.854
.788
.863
.877
.876
.886
.810
.834
.760
.853
.895
.852
.804
.864
.743
Hsu and Sabherwal
Itemsb,c
My company has teams devoted to identifying best practices.
Subscale 2: Knowledge Conversion (Standardized alpha = .93,
My company has processes for transferring organizational knowledge
to individuals.
My company has processes for absorbing knowledge from
individuals into the organization.
My company has processes for absorbing knowledge from business
partners into the organization.
My company has processes for integrating different sources and types
of knowledge.
My company has processes for replacing outdated knowledge.
My company has processes for converting competitive intelligence
into plans of action.
Subscale 3: Knowledge Application (Standardized alpha = .94,
My company has processes for applying knowledge learned from
mistakes.
My company matches sources of knowledge to problems and
challenges.
My company uses knowledge to improve efficiency.
My company is able to locate and apply knowledge to changing
competitive conditions.
My company quickly links sources of knowledge in solving
problems.
Scale 4: Dynamic Capabilities (Standardized alpha = .94, Composite
reliability = .94)
My companys employees have developed unique ways of
collaboration to improve innovative capabilities of the company.
My companys employees are sensitized to environmental changes
and respond to them.
My companys employees devote to improving the competitive
position of the company in the industry.
My companys employees proactively participate in organizational
change to help the company respond to environmental changes.
My companys employees continually innovate to make knowledge
and capabilities of the company inimitable.
My companys employees continually innovate to rapidly accumulate
knowledge assets of the company.
My companys employees integrate different areas of knowledge to
improve innovations in products/services.
My companys employees devote to improving recognition of
company name and reputation.
Scale 5: Innovation (Standardized alpha = .85, Composite reliability
= .86)
My company develops and produces new products or services
continually.
My company gives priority on making efforts to increase the quality
of products or services.
519
Factor Loadingsd
.876
.843
.903
.827
.837
.818
.891
.881
.916
.861
.923
.889
.796
.784
.822
.848
.820
.842
.860
.819
.715
520
Model
My company provides various

products or services to satisfy
customers various tastes.
My company switches quickly
between different types of
products or services to respond
to fluctuations in market
demand.
Proposed model (three first-order
factors and two second-order
factors, each with three
first-order factors)
.807
1544.76
686
Base-line model (one factor)
3772.18
701
d.f.
.751
2 /d.f. = 2.25; GFI = .88;

AGFI = .85; RFI = .91;
NFI = .92; CFI = .95;
IFI = .95; RMR = .042;
RMSEA = .049
Comparison with the
proposed model:
Difference in 2 =
2227.42; Difference in
d.f. = 15; p < .001
N = 533.
The informants were asked to indicate the degree of their agreement on the items. The
scale for their answers ranged from 1 = strongly disagree to 7 = strongly agree.
c
Due to cross-loadings or low loadings, nine items were dropped. These items are marked
with .
d
The first-order factor loadings reported were from the standardized solution. All t-values
are significant at the .001 level, with the lowest t-value being 17.61. The second-order
factor loadings were also from the standardized solution. All t-values are significant at the
.001 level, with the lowest t-value being 12.86.
e
Composite reliabilities of the second-order constructs were derived using the error variances and loadings of the first-order constructs on the second-order construct.
a
APPENDIX B: FURTHER VALIDATION TESTS FOR IC AND KM

The theoretical model contains two second-order constructs, IC and KM. In addition to establishing convergent validity and discriminant validity of the full
theoretical model, we verified the dimensionality, convergent validity, and discriminant validity of IC and KM constructs by following an established procedure
(Tanriverdi, 2005).
We started with the IC construct, for which four models were constructed
and compared. Model 1 hypothesizes that a unidimensional first-order factor explains the variances among all items of the scale of IC. Model 2 hypothesizes
three correlated first-order factors: human capital, social capital, and organizational capital. Comparing Model 1 ( 2 = 569.96, d.f. = 31, 2 /d.f. = 18.39,
GFI = .796. AGFI = .64, RMSEA = .19) against Model 2 ( 2 = 86.91, d.f. =
28, 2 /d.f. = 3.10, GFI = .97, AGFI = .94, RMSEA = .06) suggests the superiority of Model 2 ( 2 = 483.05, d.f. = 3, p < .001). This supports multidimensionality of IC, which should contain three dimensions: human capital,
social capital, and organizational capital. Further, in Model 2, standardized factor loadings of measurement items for all three first-order factors are significant
Hsu and Sabherwal
521
(p < .001), supporting convergent validity of IC. In Model 2, pairs of correlations

among the first-order factors differ significantly from zero, and are below the .90
cut-off value (Bagozzi, Yi, & Phillips, 1991). This suggests that each individual
first-order factor captures unique theoretical content, and supports discriminant
validity of IC (Anderson, 1987; Bagozzi et al., 1991).
The subsequent comparison examines whether a second-order factor explains
the patterns of covariance (complementarities) among the first-order factors. To
achieve this, an external criterion variable is used (Venkatraman, 1990; Tanriverdi,
2005). Two models including dynamic capabilities as the criterion variable are
compared: Model 3, which examines the direct effects of the three first-order
factors on dynamic capabilities, and Model 4, which includes a second-order
factor (IC) representing the first-order factors, as affecting dynamic capabilities.
Three criteria were used to compare the two models: (i) fit measures of the two
models (Venkatraman, 1990); (ii) the target coefficient (T) (Marsh & Hocevar,
1985); and (iii) significance of the second-order factor loadings (Venkatraman,
1990).
Comparing Model 3 (the first-order model: 2 = 325.44, d.f. = 104, 2 /d.f.
= 3.13, GFI = .94, AGFI = .91, RMSEA = .06) against Model 4 (the secondorder model: 2 = 328.60, d.f. = 106, 2 /d.f. = 3.10, GFI = .94, AGFI = .91,
RMSEA = .06) suggests the superiority of Model 4 ( 2 = 3.16, d.f. = 2, p >
.1) due to its parsimonious nature (fewer parameters and more degrees of freedom)
(Venkatraman, 1990). The literature supports the use of either 2 /d.f. < 3.0 (e.g.,
Carmines & McIver, 1981) or 2 /d.f. < 5.0 (e.g., Kline, 2005) as a cut-off for
model parsimony. All the other fit indices are satisfactory for Model 4. The target
coefficient (T = 0.99) approximates 1, the theoretical upper limit. This supports
the second-order factor model as the second-order factor explains 99% of the
relationships among the first-order factors (Marsh & Hocevar, 1985). Finally, the
structural path from IC to dynamic capabilities in the second-order factor model is
positive and significant as hypothesized (standardized path coefficient = .91, p <
.001). All second-order factor loadings are significant (p < .001), further supporting
the second-order factor model (Venkatraman, 1990). These results confirm the
reliability, multidimensionality, and convergent and discriminant validity of the
IC construct. A second-order construct explains the complementarities among the
three first-order IC dimensions, as predicted by the theory of IC.
The same procedure was used for the KM construct. Model 1 hypothesizes
that a unidimensional first-order factor explains the variances among all items of
the scale of KM. Model 2 hypothesizes that the measurement items assess three
correlated first-order factors: knowledge acquisition, knowledge conversion, and
knowledge application. Comparing Model 1 ( 2 = 712.03, d.f. = 49, 2 /d.f. =
14.53, GFI = .81. AGFI = .64, RMSEA = .16) against Model 2 ( 2 = 148.18,
d.f. = 46, 2 /d.f. = 3.22, GFI = .96, AGFI = .92, RMSEA = .065) suggests
the superiority of Model 2 ( 2 = 563.85, d.f. = 3, p < .001). This supports
multidimensionality of KM, which should contain three dimensions: knowledge
acquisition, knowledge conversion, and knowledge application. Further, in Model
2, standardized factor loadings of measurement items for each of the three firstorder factors are all significant (p < .001), indicating good convergent validity of
KM. In Model 2, not only do pairs of correlations among the first-order factors
522
differ significantly from zero, but they are also below .90 (the cut-off value)
(Bagozzi et al., 1991). This suggests that each of the individual first-order factors
captures different and unique theoretical content, supporting discriminant validity
of KM (Anderson, 1987; Bagozzi et al., 1991).
An external criterion variable is necessary in order to examine whether a
second-order factor explains the patterns of covariance among the first-order KM
factors, because KM also includes three first-order factors. Innovation was used
as the criterion variable for KM in Models 3 and 4. Comparing Model 3 (the
first-order model: 2 = 259.57, d.f. = 95, 2 /d.f. = 2.73, GFI = .94, AGFI = .91,
RMSEA = .058) against Model 4 (the second-order model: 2 = 265.21, d.f. = 97,
2 /d.f. = 2.73, GFI = .94, AGFI = .91, RMSEA = .058) suggests the superiority
of Model 4 ( 2 = 5.64, d.f. = 2, p > .05) due to its parsimonious nature (fewer
parameters to be estimated and more degrees of freedom) (Venkatraman, 1990).
The target coefficient (T = .98) supports the second-order factor model as the
second-order factor explains 98% of the relationships among the first-order factors
(Marsh & Hocevar, 1985). Finally, the structural path from KM to innovation in
Model 4 is positive and significant as hypothesized (standardized path coefficient
= .66, p < .001). All second-order factor loadings are significant (p < .001),
further supporting the second-order factor model. Together, these results confirm
the reliability, multidimensionality, and convergent and discriminant validity of the
KM construct. A second-order construct explains the complementarities among
the three first-order KM dimensions, as predicted by the theory of KM.
APPENDIX C: MEDIATION TESTS

To further validate the results (e.g., that IC and KM do not directly affect firm
performance), we conducted the mediation analyses described in this appendix.
We created a base model that includes all the direct and indirect paths. Thus,
the base model included all the hypothesized paths and the following additional
paths: from learning culture to KM, efficiency, innovation, dynamic capabilities
(DC), and organizational performance (OP); from IC to efficiency, innovation,
and OP; and from KM to OP. To test each mediated effect, we created a model
constraining the path from the independent to dependent construct to zero, and
then compared this constrained model with the base model. A significant increase
in 2 indicates partial mediation whereas a nonsignificant increase in 2 indicates
full mediation. We also computed the asymmetrical confidence interval for each
indirect effect with a biased-corrected bootstrap, which is considered more reliable
than the normal distribution assumed by the Sobel test (Mackinnon, Lockwood, &
Williams, 2004; Qureshi, Fang, Ramesy, McCole, Ibboston, & Compeau, 2009).
If the asymmetrical confidence interval includes zero, it implies that the indirect
effect is not significant, and does not support the presence of mediation, whereas
if it does not include zero, it implies that the indirect effect is significant, and
supports the presence of mediation (Shrout & Bolger, 2002; Mackinnon et al.,
2004). Overall, this is a stringent test of mediation, is considered an improvement
over Baron and Kenny (1986), and is suitable for structural models (e.g., Shrout
& Bolger, 2002; Qureshi et al., 2009). The results of these mediation tests are
provided in Table C1 at the end of this appendix.
Hsu and Sabherwal
523
As can be seen in Table C1, we fixed three paths, one at a time, to construct
three models in order for comparison with the base model. To test the mediating
effect of innovation on KM OP, the path is set to zero. To test the mediating
effect of innovation and dynamic capabilities on IC OP, we set the path to zero.
To be comprehensive, we also test the mediating effect of innovation on DC
OP as IC OP is mediated by dynamic capabilities and then innovation. When
any of the three paths is set to zero, the pertinent model does not deteriorate as 2
does not increase. Thus, in each of the three models, full mediation is generally
supported. However, the more stringent test requires that the indirect effects are
also significant. The fact that some asymmetrical confidence intervals include zero
only suggests that indirect effects are not necessarily significant. Full mediation
could have not been supported due to the nonsignificant indirect effect of KM on
OP, and that of IC on OP. However, with these results, it is safe to conclude that
IC and KM do not directly affect OP, respectively.
Table C1: Test of mediationNested model comparison.
Fixed
Following
Paths to
Zero
Base (no path
was fixed)
Model 1: KM
OP
Model 2: IC
OP
Model 3: DC
OP
d.f.
GFI AGFI NFI RMSEA 2
1715.36 801 .87
.84
.91
.047
1715.66 802 .87
.84
.91
.047
.3
1715.95 802 .87
.84
.91
.047
.59
1716.37 802 .87
.84
.91
.047
1.01
Asymmetrical
confidence
intervals
90%L = .317,
90%U = .277
90%L = .282,
90%U = .216
90%L = .163,
90%U = .739
Notes: IC: intellectual capital; KM: knowledge management; DC: dynamic capabilities; OP:
organizational performance. The lower (L) and upper (U) levels of the 90% asymmetrical
confidence intervals are reported in the last column.
I-Chieh Hsu received a Bachelors degree in business administration from National Chengchi University, Taiwan, and a PhD from the University of Manchester,
Manchester, UK. He is currently a professor in the Department of Business Administration, National Changhua University of Education, Taiwan. He is also involved
with a research team that is summoned by National Science Council, Taiwan, in
order to advance research on intellectual capital. He was a visiting scholar at the
School of Labor and Employment Relations, University of Illinois at Urbana Champaign, and at the College of Business Administration, University of Missouri-St.
Louis. His research appears in IEEE Transactions on Engineering Management,
Journal of Business Research, Journal of Global Information Management, International Journal of Information Management, International Journal of Human
Resource Management, and other journals. He has also coauthored an article in
Advances in International Management, and a management textbook with a group
of international scholars. His current research interests include intellectual capital
management, knowledge management, and diversity management.
524
Rajiv Sabherwal received his Bachelors degree in engineering (electronics)

from Regional Engineering College, Bhopal, India, his post-graduate degree in
management from Indian Institute of Management, Calcutta, and his PhD from
the University of Pittsburgh. He is Walton Professor and Department Chair
of Information Systems in the Sam M. Walton College of Business at the
University of Arkansas, Fayetteville. He has taught at the University of Missouri
St. Louis, Florida State University, and Florida International University. He was
the 2009 Fulbright-Queens School of Business Research Chair of knowledge
management at Queens School of Business in Kingston, Ontario, and was earlier
visiting professor at the National University of Singapore. He has published numerous articles in leading journals including Management Science, Decision Sciences,
Organization Science, Information Systems Research, MIS Quarterly, and California Management Review. He has coauthored textbooks on knowledge management
and business intelligence. His research interest includes strategic alignment, information systems planning, knowledge management, business intelligence, and
social aspects of systems development. He has served as senior editor at MIS Quarterly, guest senior editor at Information Systems Research, a departmental editor
at IEEE Transactions on Engineering Management, and on the editorial boards for
these and several other journals, including Management Science, Journal of Association for Information Systems (AIS), and Journal of Management Information
Systems (MIS). He was the conference co-chair for the International Conference
on Information Systems (ICIS) 2010. He is the editor-in-chief of Transactions on
Engineering Management, and a fellow of the AIS.

Relationship Between Intellectual Capital and Knowledge Management 2012

Diunggah oleh

Informasi Dokumen

Hak Cipta

Format Tersedia

Bagikan dokumen Ini

Bagikan atau Tanam Dokumen

Opsi Berbagi

Apakah menurut Anda dokumen ini bermanfaat?

Apakah konten ini tidak pantas?

Hak Cipta:

Format Tersedia

Relationship Between Intellectual Capital and Knowledge Management 2012

Diunggah oleh

Hak Cipta:

Format Tersedia

Decision Sciences

Relationship between Intellectual Capital

Relationship between IC and KM

Hsu and Sabherwal

Figure 1: The broad theoretical model.

Relationship between IC and KM

Hsu and Sabherwal

The Relationship between IC and KM

Relationship between IC and KM

H1: Intellectual capital has a positive effect on knowledge management.

Together, H1 and H2 represent the expected reciprocal effects between IC

Hsu and Sabherwal

proximate to KM, such as satisfaction with KM (Sabherwal & Becerra-Fernandez,

Innovation at the organizational level has been defined as a technology,

Relationship between IC and KM

Innovation enhances firm performance through improved product/service quality,

KBV emphasizes the role of KM in firm innovation (Grant 1996a, 1996b;

Dynamic capabilities support the reconfiguration of the firms resources.

KM supports organizational learning (Zollo & Winter, 2002), as well as

Hsu and Sabherwal

The Role of Learning Culture

Firm Size as a Control Variable

Relationship between IC and KM

Figure 2: The detailed theoretical model.

DATA COLLECTION AND MEASURES

Hsu and Sabherwal

Relationship between IC and KM

Hsu and Sabherwal

Hsu and Sabherwal

Other Tests of Appropriateness of Data

ANALYSES AND RESULTS

Relationship between IC and KM

Figure 3: Results for the theoretical model.

Note: p < .05, p < .01, p < .001; one-tailed tests.

Hsu and Sabherwal

Figure 4: The emergent model.

Relationship between IC and KM

Furthermore, based on tests of statistical power, validity of measures, and

Implications for Research

Hsu and Sabherwal

Research question 2: How do KM and IC affect each other?

Relationship between IC and KM

Research question 3: How does organizational culture affect intellectual

Hsu and Sabherwal

Implications for Practice

Relationship between IC and KM

Implications for Future Research

Hsu and Sabherwal

Relationship between IC and KM

Cameron, K. S., & Quinn, R. E. (1999). Diagnosing and changing organizational

Hsu and Sabherwal

Easterby-Smith, M., & Prieto, I. M. (2008). Dynamic capabilities and knowledge

Relationship between IC and KM

Hsu and Sabherwal

Relationship between IC and KM

Santoro, M. D., & Bierly, III, P. E. (2006). Facilitators of knowledge transfer

Hsu and Sabherwal

APPENDIX A: RESULTS FOR THE MEASUREMENT MODELa

Relationship between IC and KM

Hsu and Sabherwal

Relationship between IC and KM

My company provides various

Base-line model (one factor)

2 /d.f. = 2.25; GFI = .88;