Integrating knowledge through information trading: Examining the relationship between boundary spanning communication and individual performance

Decision SciencesVolume 34 Number 2Spring 2003Printed in the U.S.A.

Integrating Knowledge through InformationTrading: Examining the Relationshipbetween Boundary SpanningCommunication and IndividualPerformance∗

Robin TeiglandInstitute of International Business, Stockholm School of Economics, Sweden 11383,e-mail: [email protected], URL: http://www.teigland.com

Molly McLure WaskoCollege of Business, Florida State University, Tallahassee, FL 32306,e-mail: [email protected]. URL: http://cob.fsu.edu/ims

ABSTRACT

With the global penetration of Internet technologies, individuals may now cross organi-zational boundaries to communicate efficiently with others regardless of time and space.Thus, when looking for help in solving work tasks, knowledge workers may just aseasily contact individuals in rival firms across the globe as a coworker sitting at thenext desk. As a result, management faces questions such as (1) How should firms man-age employees’ knowledge-sourcing activities when they span both intra-organizationaland extra-organizational boundaries, and (2) What is the relationship between differentknowledge-sourcing activities and individual performance? Grounded in the knowledge-based view of the firm, we investigate these questions using data from Europe’s largestinformation technology (IT) services and management consulting company. Our resultsprovide evidence that organizations should support boundary spanning and knowledgesourcing from both internal and external sources. Results suggest (1) a positive rela-tionship between boundary spanning communication and creativity and general perfor-mance, and (2) a negative relationship between a reliance on colocated coworkers asknowledge sources and creativity.

Subject Areas: Boundary Spanning, Electronic Community, Know-HowTrading, Knowledge, Performance, Structural Equation Modeling.

INTRODUCTION

With the rapid penetration of Internet communication technologies across the globe,the possibility for individuals to seek out others for advice and know-how has

∗The authors gratefully acknowledge the support provided by Cap Gemini Nordic and in particular thatof Christian Forsberg, Christian Storck, and Carl Anler as well as Victor Sylvan and Hanna Janson of theStockholm School of Economics.

261

262 Integrating Knowledge through Information Trading

dramatically increased. Individuals may now cross organizational boundaries tocommunicate efficiently with others regardless of time and space (Hinds & Kiesler,1995). As a result, many organizations are in the process of implementing intranet-based communication tools, such as electronic discussion networks, to promoteknowledge sharing across internal organizational boundaries (Alavi & Leidner,1999; Davenport & Prusak, 1998; Fulk & DeSanctis, 1995). Investments in thesetechnologies are driven by the assumption that knowledge is the most valuableresource of the firm and that new knowledge is created through the recombinationand exchange of existing knowledge (Kogut & Zander, 1992; Nahapiet & Ghoshal,1998; Nonaka, 1994). This knowledge-based view of the firm proposes that sustain-able competitive advantage stems from an organization’s ability to integrate tacitknowledge embedded in the minds of individuals (Grant, 1996a; Nonaka, 1994).Thus, increasing the amount of information sources and communication channelsemployees have available should increase the likelihood of new knowledge cre-ation, resulting in an improved level of performance for both the individual and thefirm.

However, in addition to facilitating intra-organizational knowledge flows, theInternet also enables individuals to quickly and effortlessly access a wide varietyof knowledgeable individuals outside company boundaries through email or othermeans such as Usenet groups, private chat rooms, electronic discussion networks,listservs, and so on. Thus, individuals are able to communicate and share advicewith thousands of others across the globe regardless of their demographic charac-teristics, organizational setting, or local culture (Faraj & Wasko, 1998; Sproull &Faraj, 1995). Communication across intra-organizational and extra-organizationalboundaries has been researched extensively—a major stream began in the 1960swith the investigation into the communication patterns of scientists and engineersin research and development (R&D) laboratories (Allen, 1977). However, what haschanged within the past ten years is the ease and speed with which employees atall organizational levels can participate in these knowledge flows.

Sharing knowledge across internal and external organizational boundariesposes novel challenges to organizations attempting to manage their knowledgeresources (Pickering & King, 1995). Through interaction with external sources,knowledge workers gain access to information and expertise not available locallyand can interact informally, free from the constraints of hierarchy and local rules.However, interaction with these external sources usually involves a high degree ofinformation trading and reciprocity. In order for individuals to receive help fromexternal sources, they must be willing to give advice and know-how as well, someof which company management may consider proprietary (Von Hippel, 1987).In addition, much of the prior research on boundary spanning communicationlooked at the relationship between these knowledge flows and team performance.Research that examines the relationship between knowledge flows and individualperformance is scant, as is research on the resulting implications for organizationsconcerned with managing knowledge assets.

Thus, the goal of this research is to examine whether the performance of anindividual knowledge worker varies as a result of boundary spanning communica-tion activity and informal information trading across intra- and extra-organizational

Teigland and Wasko 263

boundaries. Specifically, we examine whether individual performance is relatedto informal information trading and accessing knowledge from (1) colocatedcoworkers, (2) coworkers within the same organization but located across intra-organizational boundaries (noncolocated), (3) intra-organizational electronic dis-cussion networks, (4) informal contacts in other organizations (i.e., contacts thatare not the result of a formal relationship with the firm such as a customer, alliancepartner, supplier, or other formal relationship), or (5) interorganizational electronicdiscussion networks.

Such inquiry makes three important contributions. First, this research em-pirically examines the trade-offs between accessing and applying local knowledgeand accessing knowledge through boundary spanning communications. Second,this research clarifies how the use of various sources of advice is related to in-dividual performance in complex knowledge environments. Finally, this researchmakes possible more precise theoretical models of how Internet-based communi-cation technologies can be designed and deployed to support knowledge exchangeand the creation of new knowledge to enhance individual and thus organizationalperformance.

THEORETICAL BACKGROUND

Recent advances in strategic management thought suggest that organizational re-sources and capabilities rather than served markets are the principal source of sus-tainable competitive advantage, and that knowledge is the most important strategicresource of the firm (Grant, 1996a). This increased emphasis on organizationalcapability and knowledge has led to the development of the knowledge-based viewof the firm (Grant, 1996a, 1996b; Spender, 1996). Assuming that knowledge isa critical input to production processes, organizational capability stems from theability to integrate the specialized knowledge of individuals (Grant, 1996a, 1996b;Nonaka, 1994; Spender, 1996). Therefore, one of the key issues underlying theknowledge-based view of the firm is to understand how knowledge is integratedin firms to create organizational capability (Barnett & Hansen, 1996). However,since knowledge creation is based on individual activity, the examination of knowl-edge integration requires understanding the organizational processes through whichfirms access and utilize the knowledge possessed by its members (Grant, 1996b).

According to Grant (1996a), competitive advantage results from how effec-tive firms are in integrating the specialized knowledge of their members, and heproposes that this effectiveness depends upon the efficiency, the scope, and theflexibility of knowledge integration. Efficiency refers to how productive firms arein integrating individuals’ specialized knowledge. The scope of knowledge inte-gration refers to the different types of specialized knowledge being integrated—themore complex the scope, the greater the difficulty for competitors to replicate. Theflexibility of integration reflects extending existing capabilities through boundaryspanning activities in order to access and reconfigure additional knowledge throughboth internal and external integration (Grant, 1996a). This research focuses on twoaspects of Grant’s theory of knowledge integration: efficiency and flexibility (Grant,1996a).


As mentioned above, efficiency refers to how productive firms are in inte-grating individuals’ specialized knowledge. One condition of integrative efficiencyis a common language of discourse to ensure efficient communication between in-dividual specialists. Efficient integration is also dependent upon the frequency andvariability of task performance, where higher levels of frequency engender auto-mated responses from each organization member. Finally, organizing structuresinfluence integrative efficiency. Organizing activities that reduce the intensity andextent of communication are needed to support efficient integration, such as mod-ularity and division of labor.

In addition to integrative efficiency, Grant also emphasizes the importanceof integrative flexibility. Hypercompetitive conditions in the marketplace drive theeventual erosion of all positions of competitive advantage. Thus, sustaining a com-petitive advantage requires flexibility and the creation of new capabilities. Firmsneed to establish knowledge integration techniques that extend existing capabilitiesby bringing in new knowledge and reconfiguring existing knowledge. However,the need to access new knowledge creates complex organizational issues with re-gard to firm structure, firm boundaries, and choices between internal and externalorganizational boundary spanning.

Grant’s theory of knowledge integration represents a paradox: increasing theefficiency of knowledge integration may hinder flexibility and the ability to createnew innovations. For example, prior research suggests that creating organizationalstructures that increase the efficiency of knowledge integration through commonlanguage and frequent interactions may result in knowledge hoarding, less creativ-ity, and the “not invented here syndrome” (Granovetter, 1973; Szulanski, 1996). Inaddition, Grant’s theory focuses primarily on the issue of coordination (structuringto enhance the effectiveness of knowledge integration), without referring to issuesof “cooperation.” The theory leaves out a key component by assuming that peopleare willing to share knowledge openly and freely if provided with the structuresand opportunities to interact.

However, prior research on boundary spanning communication activities sug-gests that individuals do not give away help and advice to others in their informalsocial networks for free. Rather, individuals trade knowledge with expectations ofreciprocity. This informal trading activity has been coined “informal know-howtrading” and as originally conceived, the concept did not distinguish between thetypes of knowledge exchanged. In other words, know-how trading simply referredto the trading of any type of information. Recent advances in knowledge manage-ment thought suggest that “know-how” refers primarily to the more tacit proceduraltypes of knowledge, as distinguishable from know-what or the more explicit declar-ative types of knowledge (Kogut & Zander, 1992). Thus, to stay consistent withcurrent thought, we refer to these informal reciprocal communication exchangesas informal information trading rather than know-how trading. In other words,the goal of this research is not to examine the type of knowledge exchanged, butrather to investigate the importance of reciprocity as a cooperation mechanism.We propose that trading and expectations of reciprocity are the key cooperationmechanisms underlying cross-boundary knowledge exchange.

Informal information trading between individuals has been shown to be valu-able and sustainable over time because the sharing of knowledge is an important


aspect of being a member of a professional community, even if the employingorganizations are direct competitors (Bouty, 2000). Therefore, key issues for orga-nizations interested in successfully managing their knowledge resources involveunderstanding where knowledge workers turn for advice, whether internally oracross organizational boundaries, and exactly how they access that advice. Pre-vious research has indicated that reciprocal external information trading involves“leakage,” or the flow of company proprietary knowledge across firm boundaries(Carter, 1989; Schrader, 1991; Von Hippel, 1987). In previous research conductedby one of the authors (Teigland, 2000), one programmer explained his communi-cations in the following manner:

But most importantly I have my network from the internet. I’ve been inthis for four years so really there is a core clique of people who knoweach other and who trade secrets with each other. We pass over the nondis-closure agreements of different companies all the time and trade companysecrets.

This participation in information trading and the resulting potential for infor-mal proprietary knowledge flows across organization boundaries are of particularstrategic interest to management since such activity may impact a firm’s competi-tive advantage. Yet, it is very difficult for firms to manage and evaluate the benefitssince it occurs “off the books,” with employees generally acting completely ontheir own, with no managerial influence and no documentation of the trade (VonHippel & Schrader, 1996). In addition, investigation into the relationship betweenthis individual-level knowledge exchange and performance has been extremelylimited.

In summary, key strategic issues for firms and their managers are to betterunderstand how to balance efficient knowledge integration with demands for flexi-bility, and how to manage the intra-organizational and extra-organizational bound-ary spanning and informal information trading conducted by individual knowledgeworkers. As a step in this direction, we developed a set of hypotheses relating in-ternal and external sources of knowledge, their influence on information trading,and their implications for individual performance.

HYPOTHESES AND MODEL DEVELOPMENT

In this section, we look at the drivers of internal and external information tradingand develop the relationships between information trading activity and individualperformance. In addition, we examine two types of performance: (1) general per-formance or the ability to meet one’s job demands and (2) creativity or the abilityto develop creative solutions. We expect that the knowledge sources used and thetrading activities performed by individuals will impact general performance andcreativity in different ways; that is, general performance is reflective of applicationof current knowledge while creativity is concerned with new knowledge creationand innovation.


Integrative Efficiency

Integrative efficiency requires common language, frequent interaction, and mod-ular structure (Grant, 1996a). This suggests that one of the most efficient sourcesof knowledge should be colocated coworkers who share the same physical space,and are more likely to frequently interact with each other and develop a commonlanguage of discourse. Prior research provides ample evidence for the integrativeefficiency of knowledge between colocated coworkers. For example, research hasconsistently found that people share knowledge with others in their physical prox-imity (Allen, 1977; Monge, Rothman, Eisenberg, Miller, & Kirste, 1985) and thatpeople prefer oral sources to written sources (Allen, 1977). In addition, people seekinformation and know-how from sources that are the most easily accessed (such asasking colocated coworkers), rather than searching for and using the best knowledgesource (such as codified sources or noncolocated coworkers) (Gerstberger & Allen,1968; O’Reilly, 1982). Therefore, accessing knowledge from colocated coworkerswho share the same coding scheme and language is highly efficient (Tushman &Katz, 1980), economizing on the amount and intensity of communication needed toachieve integration. Thus, due to the efficiency of integration, people who accessknowledge from colocated coworkers should report higher levels of general jobperformance.

However, prior research also suggests that the knowledge of colocatedcoworkers may be largely redundant, providing little additional information overwhat an individual may already know (Granovetter, 1983, 1973). Thus, the knowl-edge available through colocated coworkers is likely to be limited and superfluous,impeding the ability to develop new and creative ideas. In addition, the highlyefficient structures that support knowledge integration and the exploitation of corecapabilities may evolve into core rigidities and competency traps—inappropriateknowledge sets that preserve the status quo and limit new insights, resulting ingaps between the knowledge of the firm and changing market conditions (Leonard-Barton, 1992; Levitt & March, 1988). Therefore, while accessing knowledge fromcolocated coworkers is likely to be highly efficient and lead to better general per-formance, colocated coworkers are less likely to offer the integrative flexibilityneeded to enhance creativity and develop new capabilities. This leads to our firstset of hypotheses:

H1a: The greater the reliance on colocated coworkers to access infor-mation, the lower the level of individual creativity.

H1b: The greater the reliance on colocated coworkers to access infor-mation, the higher the level of individual general performance.

Internal Integrative Flexibility and Information Trading

Internal integrative flexibility involves the extent to which existing knowledgewithin the firm can be reconfigured. While people tend to access knowledge fromthose within the same physical proximity, advances in communication technologieshave made it easier for people who are working on similar task-related problemsacross the organization to communicate. Thus, individuals may relatively easily ac-cess knowledge from coworkers whom they personally know but who are working


in other offices or on site with clients. However, due to the less frequent patterns ofinteraction and lower intensities of social pressure, noncolocated coworkers maybe less willing or committed to exchange knowledge without some type of re-turn (Blau, 1964). People within the same organization often prefer to hoard theirknowledge because they perceive that sharing knowledge results in reduced statusand lower levels of personal worth (Orlikowski, 1996). This is especially evidentin organizations where knowledge is the basis of a personal competitive advantageover others (Thibaut & Kelley, 1959). Therefore, individuals are more likely toexpect reciprocity when engaging in knowledge exchange across internal organi-zational boundaries, especially in situations where relations are not characterizedby frequent interactions and a high level of trust (Nahapiet & Ghoshal, 1998). Thus,accessing advice and knowledge from noncolocated coworkers is likely to resultin internal information trading with the expectation of reciprocity in return. Thisleads to our next hypothesis:

H2: The greater the reliance on coworkers in other locations to accessinformation, the higher the level of internal information trading.

Informal social network structures connecting intrafirm acquaintances havetypically emerged through mutual engagement in work tasks requiring a per-sonal, oftentimes physical, connection. However, recent advances in computer-mediated communication technologies have facilitated the development ofintra-organizational, electronic social networks between geographically dispersedorganizational members, who are typically strangers. Within these networks, anunlimited number of noncolocated, unacquainted coworkers are able to quicklycommunicate through their shared organizational and technical code to help eachother solve problems and provide useful advice (Constant, Sproull, & Kiesler,1996).

However, theories of social capital suggest that the ability to develop thecommitment and trust that are necessary for knowledge exchange is difficult toachieve in computer networks (Nahapiet & Ghoshal, 1998; Nohria & Eccles, 1992).Thus, norms of participation in electronic networks typically dictate that thosewho seek and receive help from the network must also pay back by helping others(Constant et al., 1996; Kollock, 1999; Lakhani & Von Hippel, 2000). Therefore,similar to the previous hypothesis, the sharing of organizational knowledge throughelectronic networks is likely to increase the amount of internal information tradingwithin the firm. This leads to our third hypothesis:

H3: The greater the reliance on intra-organizational electronic net-works to access information, the higher the level of internal in-formation trading.

Through internal information trading, individuals exchange information,knowledge, and advice. These individuals share the same organizational languageand code of behavior and are faced with similar issues related to their knowledgetasks, supporting integrative efficiency (Brown & Duguid, 1991; Wenger, 1998).Internal information trading enables the flow of ideas and innovations within thefirm. For example, when seeking help internally, an individual may find that a so-lution already exists elsewhere within the organization. In this manner, individuals


may avoid “reinventing the wheel” by reconfiguring solutions previously developedwithin the firm to fit new situations. These exchanges retain the context in whichthe knowledge is embedded, and individuals located across intra-organizationalboundaries possess knowledge that may be more locally adapted, supporting effi-cient integration and, in turn, higher levels of general performance.

In addition, the combination and recombination of firm-specific knowledgethat is physically dispersed across the organization may facilitate integrative flex-ibility. Individuals in other organizational units are more likely than colocatedcoworkers to have important knowledge that is nonredundant, generating ac-cess to sources of new ideas and innovations located across intrafirm boundaries(Granovetter, 1973). Engaging in internal information trading, people not onlysend and receive task-specific knowledge, they also help each other by taking thetime to work through each other’s problems. Exercising intellect by helping othersis likely to sharpen and even improve an individual’s own technical skills. Whenan individual works through someone else’s problems, he or she often discoversnew methods and new applications for existing knowledge (Wenger, 1998). Addi-tionally, individuals that help others are entitled to reciprocity, gaining access toadvice, new ideas, and innovations when needed. Thus, we expect internal infor-mation trading to have a positive impact on both individual creativity and generalperformance. This leads to the following hypotheses:

H4a: The higher the level of internal information trading, the higherthe level of individual creativity.

H4b: The higher the level of internal information trading, the higherthe level of individual general performance.

External Integrative Flexibility and Information Trading

In addition to internal integrative flexibility, firms need to integrate new knowl-edge found in the external environment to remain competitive (Cohen & Levinthal,1990). This knowledge may be accessed through either market or relational con-tracts (Grant, 1996a). While relational contracts tend to refer to formal interorga-nizational arrangements, these contracts also comprise informal communicationexchanges between individuals. Just as intra-organizational communication hasbeen facilitated with the rapid spread of the Internet, so too has the ability toinformally communicate with contacts outside of the organization. Thus, whenknowledge workers seek help with their work-related tasks, they may just as easilycontact individuals working in rival firms as individuals working in the same orga-nization (Faraj & Wasko, 1998; Sproull & Faraj, 1995). As a result, individuals mayintegrate knowledge from within their organization with new ideas and innovationsaccessed through communications with individuals outside their organization. Thisinformal relational activity should improve a firm’s external integrative flexibilityand sustainable competitive advantage (Grant, 1996a).

Informal information and knowledge sharing between firms has been de-tected in several settings, for example, semiconductor, specialty steel, and mini-mill industry, and R&D operations (Carter, 1989; Schrader, 1991; Von Hippel,1987). Reciprocity was found to be one of the guiding principles in these informal


exchanges, where individuals expected that their chances to receive information inreturn would increase after they sent out information (Schrader, 1991). Certain pro-fessional disciplines encourage knowledge sharing and information trading withinthe occupational community to keep abreast of new ideas and innovations andto stay competitive with other professionals (Pickering & King, 1995). As such,knowledge workers may also rely on their informal external contacts, such asfriends, family, ex-coworkers, people with whom they attended school, and so on,to access critical knowledge that resides externally to the firm. However, due tostrong norms of reciprocity within the occupational community, relying on contactsin other organizations for advice obligates the knowledge seeker to share knowl-edge in return (Van Maanen & Barley, 1984). Thus, accessing knowledge fromcontacts in the occupational community requires reciprocation through externalinformation trading. This leads to our next hypothesis:

H5: The greater the reliance on external contacts to access informa-tion, the higher the level of external information trading.

In addition to accessing information and know-how from external contacts,communication tools such as bulletin boards, listservs, and chat rooms electroni-cally connect knowledge workers sharing the same profession but who are globallydispersed and typically strangers. In a recent article, Pickering and King (1995)argued that the growth of interorganizational Internet-based communication islikely to be especially rapid between individuals who are interested in establishingties with individuals outside of the firm based primarily on similar professionalinterests.

In Internet-based electronic networks, individuals are able to share informa-tion and know-how through mechanisms that support posting and responding toquestions, sharing stories of personal experience, and discussing and debating is-sues relevant to the professional community (Wasko & Faraj, 2000). Knowledgeis continuously created and shared through open discussion and collaboration, re-gardless of physical distance or organizational affiliation. In one study of a Usenetinterorganizational technical discussion network, it was found that 42% of all mes-sages included programming code (Wasko & Faraj, 1999). Thus, interorganiza-tional electronic networks advance the knowledge of the professional communityas a whole through electronic links. Similar to accessing knowledge through exter-nal contacts whom an individual knows, the norms of accessing knowledge frominterorganizational electronic networks also requires that an individual “pay back”to the network by helping others (Lakhani & Von Hippel, 2000; Wasko & Faraj,2000). Thus, we predict the following:

H6: The greater the reliance on external electronic communities toaccess information, the higher the level of external informationtrading.

There is limited empirical evidence that links external information tradingactivities and performance. In one of the few studies in this area, Schrader (1991)found suggestive evidence for a link between informal know-how trading andfirm performance, yet beyond this, there is little that investigates the relationshipwith individual performance. However, research has found that individuals who


share the same professional interests may be able to communicate relatively easilyacross organizational boundaries due to a universal professional language, enablingindividuals to access know-how and information from outside the firm’s boundaries(Hauptman, 1986). This know-how and information is more likely to be nonredun-dant than that found within the organization. Thus, through external informationtrading, individuals may combine and recombine knowledge from within their or-ganization with new ideas and innovations accessed from individuals outside theorganization, resulting in new and creative solutions (Cohen & Levinthal, 1990).

While individuals participating in external information trading share to acertain extent the same professional knowledge and technical language, they do notshare the same organizational knowledge or language. Thus, although participationin external information trading brings in new ideas, resulting in higher levels ofindividual creativity, the combination of external knowledge with a firm’s existingknowledge may be time consuming, or may result in solutions that are not tailoredto the firm’s specific situation (Teigland, 2000). In addition, information trading andreciprocity requires giving advice in return. Helping others solve their problemscan be quite time consuming, and can lead to an individual devoting less timeworking on his or her company-specific work tasks. Thus, an individual may havemore difficulty in completing his or her tasks on time or according to the taskspecifications (Teigland, 2000). Thus, we have our next set of hypotheses:

H7a: The higher the level of external information trading, the higherthe level of creativity.

H7b: The higher the level of external information trading, the lowerthe level of general performance.

Finally, for our last hypothesis, we look at the possible relationship betweeninternal and external information trading. In the stream of research by Allen andcolleagues, it was found that in many cases, individuals who had a high degree ofexternal communication activity also displayed a high degree of internal communi-cation activity (Allen & Cohen, 1969). These individuals were labeled gatekeepers.Gatekeepers were found to display characteristics of a buffering role, scanning andfiltering information into the organization from the outside world and then direct-ing it as they felt necessary into the organization. These individuals were oftenfound to be in first-line supervisor positions (Taylor, 1975). However, with accessto the Internet, it is now possible for all individuals to become gatekeepers for theirspecialized knowledge area by participating in external exchange and then recom-bining information with the specialized knowledge to the firm through internalintegration. This brings us to our final hypothesis:

H8: The higher the level of external information trading, the higherthe level of internal information trading.

Figure 1 provides an overview of the hypotheses developed above.

RESEARCH SETTING

This research was undertaken in the Nordic operations (Denmark, Finland, Norway,and Sweden) of Cap Gemini and was performed prior to the merger of Cap Gemini


Figure 1: Model of hypotheses.

Colocated Coworkers

Noncolocated Coworkers

Intra-organizational Electronic Networks

Contacts in Other Firms

Extra-organizational Electronic Networks

Internal Information Trading

General Performance

Creativity

H1a(-),b

H2

H3

H5

H6

H8 H7a,b (-)

H4a,b

External Information Trading

with Ernst and Young Consulting. As a result, the company description takes onlythe Cap Gemini organization into consideration. At the time, Cap Gemini wasEurope’s largest information technology (IT) services and management consultingcompany with more than 40 offices and 4,500 employees in the Nordic regionalone.

Within the Nordic region, Cap Gemini had numerous networks designed toenhance the company’s knowledge management activities. We chose participantsin one electronic network, the NCN MS Community, because it was recognizedas a successful, vital conduit of knowledge exchange. This electronic networkhad 350 members spread across the Nordic countries and the members of thisnetwork all worked with applying Microsoft products in their responsibilities withCap Gemini. This particular population was chosen for the study to ensure thatresearch subjects had access to internal and external sources of information andknow-how, and had familiarity using the communication technologies underlyinginformation and know-how exchange in electronic networks. In addition, the jobresponsibilities of the members of the NCN MS Community required a considerableamount of creativity, as new problem situations constantly arose due to the rapidpace of change in information technology as well as the diversity among clientproject demands. This helped ensure that the population chosen for this study hadto balance both general job performance and demands for creativity.

Study Design and Data Collection

In November 1999, we began the data collection through five in-depth interviewswith people involved in Cap Gemini Nordic’s knowledge management operations.We conducted interviews to better understand the implementation and use of elec-tronic discussion networks within Cap Gemini and to design questionnaire itemsrelevant to this specific organization. We pilot tested the questionnaire on a group


of 15 programmers. Pilot test results indicated that the survey instrument was toolong; thus, items outside of the scope of this research were dropped from the survey.The final survey instrument was sent to each of the NCN MS Community membersasking him or her to complete the questionnaire during January 2000. Throughoutthe data collection process, individuals were assured that their responses would bekept confidential and that all results would be presented only on an aggregate level.

Of the initial 350 individuals, five e-mails were electronically returned due toan invalid e-mail address. We received a total of 83 usable survey responses from the345 participants with valid e-mail addresses for a response rate of 24%. The averageage of the respondents was 35.6 years with an average of 4.0 years employed atCap Gemini and 7.7 years of experience in their competence. The sample was 8%women. After consultation with Cap Gemini management, it was found that thedemographic characteristics of the group of respondents were representative ofthose of the entire NCN MS Community.

Measures

All variables were assessed through survey responses. Several different approachesexist for measuring performance, including both self-reported and third-party mea-surements. However, following discussions with Cap Gemini’s management, it be-came apparent that supervisor-rated or other performance measures such as salarywould be difficult to obtain due to issues of employee confidentiality. Accordingly,we opted to measure performance via self-reporting measures. Of interest is thata number of previous studies have found self-reporting measures to be superiorto third-party measurements (Heneman, 1974; Wexley, Alexander, Greenawalt,& Couch, 1980) and not upwardly biased (Churchill, Ford, Hartley, & Walker,1985). As described earlier, two measures of performance were measured: generalperformance and creativity.

The survey contained a series of 7-point Likert scale questions. The dependentvariables were measured by asking respondents to rate the extent of their agreementon a 7-point Likert scale (1 = “strongly disagree,” 4 = “agree,” 7 = “stronglyagree”). The independent variables were assessed by asking respondents to indicatehow often they engage in specific knowledge activities (1 = “several times a day,”2 = “once a day,” 3 = “once every two days,” 4 = “once a week,” 5 = “onceevery two weeks,” 6 = “once a month,” 7 = “more seldom”). The independentvariables were then transformed to convert responses from an interval scale toa ratio scale prior to analysis to conform to the ratio-scaling conventions of thedependent variables. Actual survey items are provided in Table 1.

ANALYSIS AND RESULTS

We tested the hypotheses using partial least squares (PLS), and performed two sep-arate analyses independently for each dependent variable (creativity and generalperformance). Partial least squares can be used to analyze measurement and struc-tural models with multi-item constructs that include direct, indirect, and interactioneffects, and has become widely used in information systems (IS) research (Chin& Todd, 1995; Compeau & Higgins, 1995; Gefen & Straub, 1997). Partial least


Tabl

e1:

Surv

eyite

ms,

cons

truc

ts,i

tem

load

ings

,and

cros

s-lo

adin

gs.

12

34

56

78

9

Kno

wle

dge

Sour

ces

How

ofte

ndo

you

use

the

belo

win

form

atio

nso

urce

sin

your

ever

yday

wor

k?1

Cow

orke

rsin

my

loca

tion

1.0

.28

.03

.01

−.13

.05

.04

−.17

−.13

2C

apG

emin

icol

leag

ues

inan

othe

rlo

catio

n.2

81.

0.5

5.4

2.2

0.4

7.2

1.3

0.1

33

Cap

Gem

inie

lect

roni

cco

mm

uniti

es.0

1.4

8.9

2.3

7.4

4.3

5−.

01.3

0.2

2N

CN

MS

elec

tron

icco

mm

uniti

es.0

4.5

3.9

2.3

5.5

0.3

5−.

03.2

8.2

64

Oth

erco

ntac

tsou

tsid

eof

Cap

Gem

ini

.01

.42

.39

1.0

.14

.35

.44

.31

.37

5In

tern

et:D

iscu

ssio

nfo

rum

s/el

ectr

onic

com

mun

ities

−.13

.20

.51

.14

1.0

.40

.00

.24

.18

Info

rmat

ion

Tra

ding

How

man

yti

mes

duri

ngth

epa

stye

arha

veth

efo

llow

ing

happ

ened

?6

You

wer

eco

ntac

ted

byso

meo

nein

Cap

Gem

ini

for

som

esp

ecifi

cte

chni

cali

nfor

mat

ion

.08

.38

.36

.27

.35

.89

.39

.53

.33

You

gave

outs

ome

spec

ific

tech

nica

lin

form

atio

nto

som

eone

atC

apG

emin

i.1

1.3

4.3

6.3

1.3

9.9

0.4

0.4

7.3

1

You

sent

form

al,w

ritte

nco

mm

unic

atio

nsin

the

form

ofre

port

sor

data

toso

meo

nein

side

ofC

apG

emin

i

.02

.44

.18

.27

.24

.72

.48

.43

.27

You

sent

info

rmal

,wri

tten

com

mun

icat

ions

inth

efo

rmof

repo

rts

orda

tato

som

eone

insi

deof

Cap

Gem

ini

−.03

.43

.37

.34

.34

.85

.41

.50

.28

7Y

ouw

ere

cont

acte

dby

som

eone

outs

ide

ofC

apG

emin

ifor

som

esp

ecifi

cte

chni

cal

info

rmat

ion

−.10

.07

−.02

.36

.06

.47

.88

.36

.29

274 Integrating Knowledge through Information TradingTa

ble

1:(c

ontin

ued)

Surv

eyite

ms,

cons

truc

ts,i

tem

load

ings

,and

cros

s-lo

adin

gs.

12

34

56

78

9

Info

rmat

ion

Tra

ding

How

man

yti

mes

duri

ngth

epa

stye

arha

veth

efo

llow

ing

happ

ened

?Y

ouga

veou

tsom

esp

ecifi

cte

chni

cal

info

rmat

ion

toa

pers

onw

orki

ngou

tsid

eof

Cap

Gem

ini

−.02

.11

−.09

.32

.03

.40

.90

.24

.23

You

sent

form

al,w

ritte

nco

mm

unic

atio

nsin

the

form

ofre

port

sor

data

toso

meo

neou

tsid

eof

Cap

Gem

ini

.17

.28

−.04

.43

−.07

.42

.86

.27

.24

You

sent

info

rmal

,wri

tten

com

mun

icat

ions

inth

efo

rmof

repo

rts

orda

tato

som

eone

outs

ide

ofC

apG

emin

i

.10

.26

.05

.45

.00

.47

.91

.36

.29

Per

form

ance

8M

yw

ork

task

sde

man

dcr

eativ

ean

dto

tally

new

idea

san

dso

lutio

ns−.

03.3

6.3

2.2

2.2

3.3

8.2

3.7

6.3

9

The

othe

rsat

Cap

Gem

init

hink

that

Iam

crea

tive

−.13

.19

.14

.13

.03

.42

.31

.79

.39

My

colle

ague

sat

Cap

Gem

inic

onsi

der

me

tobe

agu

ru−.

21.2

1.3

0.3

5.3

0.5

5.3

0.8

4.6

3

9I

have

ahi

ghle

velo

fex

pert

ise

inth

ete

chno

logy

with

whi

chI

wor

k−.

14.0

8.3

0.3

1.2

7.3

9.2

9.5

9.9

1

You

are

able

tom

eeto

bjec

tives

setb

yyo

urim

med

iate

supe

rior

rela

tive

toyo

urco

lleag

ues

atC

apG

emin

i

−.07

.19

.14

.35

.00

.18

.19

.39

.71

You

are

able

tom

eetm

yde

adlin

esre

lativ

eto

your

colle

ague

sat

Cap

Gem

ini

−.03

.01

−.10

.08

−.14

.02

.12

.26

.60

Not

e:Sh

aded

item

sar

efa

ctor

load

ings

and

corr

espo

ndin

gth

eore

tical

cons

truc

ts.


squares does not assume multivariate normality among sample distributions andtakes into account measurement error when assessing the structural model (Wold,1982). As a result, it is particularly useful for analyzing constructs that includemeasurement error and covariance. The results are interpreted in two stages—measurement and structural. Because the combined analysis of the measurementand structural models enables measurement errors to be included as an integralpart of the model, and factor analysis to be combined in one operation with thehypotheses testing, the result is a more rigorous analysis of the proposed researchmodel (Bollen, 1989).

Unlike LISREL and EQS structural equation modeling techniques, PLS breaksdown models into segments, allowing researchers to work with small sample sizes.When determining sample size, theorists suggest that a “rule of thumb” for items is10 times the most complex construct’s number of indicators or the largest number ofpaths leading to a latent construct. The most complex construct in the measurementmodel has four indicators, and the largest number of paths leading to a latentconstruct is three. Thus with 83 respondents (83/4, or 20.75 responses per indicator)the sample size is sufficient for established PLS guidelines to proceed with analysis(Chin, 1998). The theoretical model was estimated using PLS Graph 2.91 (Chin &Frye, 1996).

Measurement Model

A crucial step prior to testing the theoretical model is assessing the accuracy ofthe measurement model. The goals of assessing the accuracy of the measurementmodel are to demonstrate that the measures used are valid and that they adequatelyreflect the underlying theoretical constructs. The first step in PLS is to assess theconvergent validity of the constructs of interest, by examining the average varianceextracted (AVE). The AVE attempts to measure the amount of variance that a la-tent variable component captures from its indicators relative to the amount due tomeasurement error. The AVE is calculated by taking the sum of the squared com-ponent loadings to an indicator and dividing by the sum of the squared componentloadings plus the sum of the error variance. It is recommended that the AVE shouldbe greater than .50, meaning that 50% or more variance of the indicators should beaccounted for.

Individual survey items that make up a theoretical construct must also beassessed for interitem reliability. In PLS, the internal reliability and consistency fora given block of indicators can be calculated using the internal composite reliability(ICR) developed by Werts, Linn, and Joreskog (1973). The ICR is calculated bysquaring the sum of component loadings to an indicator, then dividing by the sumof squared loadings plus the sum of the error terms. Interpreted like a Cronbach’scoefficient, acceptable values of an ICR for perceptual measures should exceed .7(Fornell & Larcker, 1981). Values less than .7 imply that the items underlying theconstruct may be unrelated, or may be measuring more than one construct.

Discriminant validity indicates the extent to which a given construct is dif-ferent from other constructs, the measures of the constructs are distinct, and theindicators load on the appropriate construct (Messick, 1980). One criterion for ade-quate discriminant validity is demonstrating that the construct shares more variance


Table 2: Descriptives, ICRs, correlations, and AVE values.

Range Mean SD ICR 1 2 3 4 5 6 7 8 9

1 Colocated 0-3.7 3.0 1.0 n/a 1Coworkers

2 Noncolocated 0-3.7 1.3 1.2 n/a .28 1Coworkers

3 Intra-org. Nets 0-3.4 1.2 1.1 .92 .03 .55 .924 Contacts at Other 0-3.7 .80 .87 n/a .01 .43 .39 1

Firms5 Extra-org. Nets 0-3.7 1.3 1.3 n/a −.13 .20 .55 .14 16 Internal Info 0-3.1 1.2 .81 .91 .05 .47 .38 .36 .40 .84

Trading7 External Info 0-3.6 .56 .70 .94 .04 .21 −.02 .44 .01 .50 .89

Trading8 Creativity 1-5.7 4.2 1.2 .84 −.17 .30 .32 .31 .24 .58 .35 .809 General 3-6.3 4.6 .88 .79 −.13 .13 .26 .37 .18 .35 .30 .61 .75

Performance

Note: Boldface numbers on the diagonal are the square roots of the AVE values.

with its measures than it shares with other constructs in the model (Barclay, Higgins,& Thompson, 1995). One measure of discriminant validity in PLS is the AVE. TheAVE may be compared with the shared variance among the latent variables (i.e., thesquare root of the AVE should be greater than the correlation between a constructand any other construct) (Chin, 1998). A second way to evaluate discriminant valid-ity is to examine the factor loadings of each indicator (Chin, 1998). Each indicatorshould load higher on the construct of interest than on any other factor.

Table 1 presents the factor loadings and cross-loadings for all indicators. Eachindicator loaded higher on its theoretical construct than on any other factor, indicat-ing discriminant validity. Table 2 presents descriptive statistics, AVEs, ICRs, andcorrelations between constructs. All AVE values are greater than the .5 cutoff pointindicating adequate convergent validity. The square root of the AVEs are presentedon the diagonal of the correlation matrix, and are greater than the correspondingcorrelations, indicating adequate discriminant validity. All ICR values are greaterthat the .7 cutoff, indicating adequate reliability.

Similar to findings in prior research, it appears that the knowledge workersin this sample rely to a great extent on colocated coworkers as sources for infor-mation. In addition, the survey respondents also reported a fairly high amount ofintra-organizational boundary spanning communications and internal informationtrading. Finally, there are some indications that people engage in boundary span-ning activities using the same type of information sources (interpersonal versuselectronic). For instance, there is a strong correlation between the use of non-colocated coworkers as information sources and external contacts. In addition,there is a correlation indicating that people who access information from intra-organizational electronic networks also access advice from interorganizationalnetworks.


Table 3: Results of PLS analysis.

Internal Info External Info GeneralTrading Trading Creativity Performance

H1a,b Colocated Coworkers −.20∗ −.15H2 Noncolocated Coworkers .24∗∗H3 Intra-org. Electronic Nets .26∗∗H4a,b Internal Info Trading .54∗∗ .28∗H5 Contacts at Other Firms .45∗∗H6 Extra-org Electronic Nets −.06H7a,b,H8 External Info Trading .45∗∗ .09 .17

R2 .44 .20 .38 .17

∗p < .05, two-tailed test.∗∗p < .01, two-tailed test.

Figure 2: Model of results.

.23* Colocated Coworkers

Noncolocated Coworkers

Intra-organizational Electronic Networks

External Information Trading

Contacts at Other Firms

Extra-organizational Electronic Networks

Internal Information Trading

General Performance

Creativity

-.20*

.24**

.26**

.45** .45**

.54**

.28*

R2 = .38

R2 = .44

R2 = .20

R2 = .17

Structural Model

Table 3 summarizes the PLS structural analysis. Figure 2 provides a graphicalrepresentation of the results. To evaluate the models, R2 values were calculatedfor endogenous constructs. Interpreted like multiple regression results, the R2 in-dicates the amount of variance explained by the model (Chin, 1998). The overallmodel explained 38% of the variance in creativity and 17% of the variance ingeneral performance. In addition, the model explained 44% of the variance in in-ternal information trading and 20% of the variance in external information trading.Specifically, we find support for H1a, workers that rely on colocated coworkersrate lower on creativity (b = −.20, p < .05). However, contrary to expectations inH1b, there was no relationship between accessing information from coworkers andgeneral performance (b = −.15, p ns). We found support for H2, people that rely


on coworkers in other locations are more likely to engage in internal informationtrading (b = .24, p < .01), as well as H3: people participating in internal electronicnetworks engage in internal information trading (b = .26, p < .01). Finally, as pre-dicted in hypotheses H4a and H4b, internal information trading results in higherlevels of creativity (b = .54, p < .01) and higher levels of general performance(b = .28, p < .05).

As predicted in H5, people who use their contacts at other firms as knowledgesources engage in external information trading (b = .45, p < .01). However, we findno support for H6; thus, there is no evidence for a relationship between participa-tion in external electronic networks and external information trading. Contrary topredictions, external information trading did not directly influence creativity (H7a)or general performance (H7b). Finally, we find support for H8: people who engagein external information trading are also likely to engage in internal informationtrading (b = .45, p < .01).

DISCUSSION AND IMPLICATIONS

This research provides insight into how firms may create organizational structuresthat balance efficient and flexible knowledge integration among individuals. Interms of internal integration, individuals search for information and know-howlocated both within and across intra-organizational boundaries and integrate it withtheir own when performing work-related tasks. However, our findings suggest thatthe view of the firm as a knowledge integrator needs to be further developed byincorporating a dimension of cooperation. Grant’s theory assumes that individualsare willing to share knowledge with each other without expecting anything in return.Yet our research findings suggest that knowledge integration through boundaryspanning activities is supported by informal information trading.

While knowledge can be accessed from colocated coworkers without expec-tation of reciprocal trading and returns, once intra-organizational physical bound-aries are crossed, expectations of returns for knowledge sharing, as exhibited ininformation trading, appear to come into play. In order for an individual to accessknowledge from others outside his or her immediate physical location, it appearsthat he or she must be willing to give something in return. In addition, our resultssuggest insights into the importance of flexible knowledge integration structuresabove and beyond the efficiency of colocation.

Efficiency and Flexibility of Integration: Implications for Performance

In terms of integrative efficiency, we found evidence that a high reliance oncolocated coworkers results in lower levels of creativity. This suggests that theknowledge of colocated coworkers may be largely redundant and the integration ofthis local knowledge, although efficient, may stifle the development of new ideasand innovations. On the other hand, information that crosses intra-organizationalboundaries enhances integrative flexibility, as evidenced by the positive relationshipbetween internal information trading and both general performance and creativity.Individuals may either reconfigure existing knowledge to fit their local needs orintegrate new know-how and innovative ideas with their own.


One surprising finding, though, is that integrative efficiency did not enhancean individual’s general performance as predicted. One potential explanation maybe related to the task being performed. In software development there is some de-gree of standardization and universal technical language since individuals acrossthe firm are typically using the same underlying programming languages. Thus, in-dividuals may trade standardized programming hints or code across organizationalboundaries that may simply be “plug n’ play,” enabling use with current know-how and easier recombination with internal firm knowledge. Thus, in the area ofsoftware development and the universal language of standardized programmingcode, it may be just as efficient to integrate knowledge from colocated coworkersas coworkers across the organization.

In terms of external integrative flexibility, contrary to expectations, we foundthat external information trading has no direct relationship with individual perfor-mance; rather it affects creativity and general performance indirectly through itsinfluence on internal information trading. We had expected that knowledge comingfrom contacts outside of the firm would be relatively novel and lead to more creativesolutions, but at the same time it would be more difficult to apply to one’s task, thusrequiring more time to use. One potential explanation for our results may be thatknowledge coming from outside the firm may be so novel that it cannot be appliedto any immediate solution. Rather its dissemination to others and subsequent re-combination with the firm’s knowledge is necessary to adapt this knowledge to thefirm’s specific use. Thus, the ability to develop creative solutions and improve per-formance may involve combining existing internal knowledge with novel externalknowledge.

This finding supports theories of absorptive capacity, which suggest that thefirm’s ability to assimilate new, external information is largely a function of thefirm’s ability to internally process that information (Cohen & Levinthal, 1990).In addition, this research potentially offers additional insight: our results suggestthat performance is enhanced by the recombination of knowledge that crossesinternal organizational boundaries. This provides some evidence that organiza-tional absorptive capacity is enhanced through intra-organizational cross-boundaryknowledge flows, and by recombining knowledge through informal informationtrading. While there has been a great deal of emphasis placed on the impor-tance of face-to-face interactions and colocation for transferring valuable knowl-edge, organizations interested in enhancing absorptive capacity may want to estab-lish structures other than long-term colocation that emphasize flexible knowledgeintegration.

Flexibility of Integration: Role of Electronic Discussion Networks

Our research findings also indicate that creating electronic social networks mayenhance integrative flexibility and information trading activities as well. In terms ofintra-organizational electronic networks, our findings provide support for previousresearch suggesting that norms of reciprocity are critical for sustaining knowledgeexchange in electronic discussion networks (Lakhani & Von Hippel, 2000; Wasko& Faraj, 2000). However, contrary to expectations, we find no relationship betweenparticipation in extra-organizational electronic networks and external information


trading. Thus, it seems that individuals are more likely to engage in informationtrading with others with whom they have a common bond, such as organizationalmembership.

One explanation may be found by looking at the relationship between theability to establish reciprocal exchange and various communication channels. Itmay be more difficult to build reciprocal relationships with individuals in extra-organizational electronic networks since members generally have not met eachother face to face and have little social influence over one another due to thevoluntary and anonymous nature of the exchange. In addition, when reciprocityoccurs in these networks it is typically of a generic and not a dyadic nature (Kollock,1999). In order for an individual to give to the network, there must be a level oftrust across the network members that ensures other network members will “payback” when requested.

Our findings indicate that it may be easier to build trust and achieve a norm ofreciprocity in intra-organizational electronic networks than in extra-organizationalones. This may be because individuals within these networks have a common orga-nizational tie and are thus working for the greater good of the company (Constantet al., 1996). However, there are other aspects to consider. Intra-organizational net-works may be more stable in terms of participation, membership, and identificationof participants. It is also possible that individuals are not as anonymous as they arein extra-organizational networks. Finally, misbehavior in an intra-organizationalelectronic network may be more easily “punished” and carry tangible deterrents,while positive behaviors may be rewarded through increases in status and reputa-tion in the organization. As a result, intra-organizational electronic networks maybe better able to control their boundaries and member behavior, resulting in moreeffective flows of knowledge.

Implications for Practice

These findings indicate that organizations concerned with knowledge managementmay need to rethink their knowledge management strategies and find a balancebetween reliance on colocated coworkers and the promotion of flexible integrationthrough boundary spanning communication to improve individual performance andcreativity. Thus, results from this study suggest an important new use of Internet-based communication technologies to support knowledge management. Ratherthan using technology to replace traditional knowledge management techniques,such as creating document repositories, we need to think of nontraditional ways toleverage these new technologies for improved knowledge flows within and acrossthe firm, by leveraging networks that support the exchange of advice and ideasbetween individuals.

While intra-organizational trading can be viewed positively without questionin terms of enhancing individual performance, and ultimately a firm’s competi-tive advantage, the presence of interorganizational information trading draws intoquestion the degree to which a company’s proprietary knowledge is leaking acrossthe firm’s boundaries. The decision to trade or not with external parties is placed inthe hands of an individual working for the firm. As such, most economic and man-agement researchers would argue that this informal transfer is a disadvantage for


the firm since the individual’s actions may not be in line with the firm’s objectives,and may lead to a firm being unable to capitalize on the unique benefits from aninnovation (Schrader, 1991).

However, the results of this research seem to indicate that external infor-mation trading is beneficial for the firm, although indirectly through recombina-tion with existing knowledge. Thus, while an individual may trade away “propri-etary” knowledge, the ability of a rival firm to turn this into an innovation lies inits ability to internally integrate the new knowledge into the existing knowledgebase of the firm. In addition, trading information across organizational bound-aries enhances the inflow of new ideas and innovations. Information trading alsoensures that help will be reciprocated at a point where the information seeker isin need of advice. Thus, while information trading implies knowledge leakageacross firm boundaries, it also ensures that new knowledge flows back into thefirm.

LIMITATIONS AND AREAS FOR FURTHER RESEARCH

We should note the limitations of the study and caution that this study was of anexploratory nature, with the contributions merely acting as guidelines for furtherresearch. First, this study only examined knowledge workers focused on develop-ing software solutions in one company, thus limiting the generalizability of ourfindings. Further research should examine individual information trading acrossmultiple organizations, and across multiple categories of knowledge workers. Inaddition, the number of participants in this study, although adequate for analysis,is relatively small to make conclusive statements outside of this context. Furtherresearch should include all organizational knowledge workers, the impact of lo-cation, and all internal as well as external organizational information sources.Another limitation is our use of self-reported survey measures only, increasing therisk of common-method bias. Further research should include other performancedata sources in addition to survey data.

CONCLUSION

This paper examined the relationship between various knowledge integration ac-tivities and individual knowledge worker performance. We found that internal andexternal boundary spanning facilitates information trading that in turn results in im-proved individual performance. Therefore, creating organizational structures thatincrease the flexibility of knowledge integration may support the creation of newideas and innovations, leading to sustainable competitive advantage. In addition,this study indicates that people who rely on colocated coworkers as informationsources report lower levels of creativity. This indicates that organizational structuresthat enhance the efficiency of knowledge integration, without regard to flexibility,may “bind and blind”—supporting adherence to the same ideas and information,potentially impeding the creation of new knowledge and stifling performance.Therefore, organizations concerned with knowledge management and creative so-lutions should focus on balancing knowledge integration structures that support


efficiency with flexibility, emphasizing boundary spanning and informal informa-tion trading through both personal and electronic networks. [Received: March 1,2002. Accepted: January 30, 2003.]

REFERENCES

Alavi, M., & Leidner, D. (1999). Knowledge management systems: Issues, chal-lenges and benefits. Communication of the Association for Information Sys-tems, 1, 1–28.

Allen, T. J. (1977). Managing the flow of technology. Cambridge, MA: MIT Press.

Allen, T. J., & Cohen, S. (1969). Information flow in research and developmentlaboratories. Administrative Science Quarterly, 14, CE12–20.

Barclay, D., Higgins, C., & Thompson, R. (1995). The partial least squares (PLS)approach to causal modeling: Personal computer adoption and use as anillustration. Technology Studies, 2(2), 285–309.

Barnett, W. P., & Hansen, M. T. (1996). The red queen in organizational evolution.Strategic Management Journal, 17(Summer), 139–158.

Blau, P. M. (1964). Exchange and power in social life. New York: Wiley.

Bollen, K. A. (1989). Structural equations with latent variables. New York: Wiley.

Bouty, I. (2000). Interpersonal and interaction influences on informal resourceexchanges between R&D researchers across organizational boundaries.Academy of Management Journal, 43(1), 50–66.

Brown, J. S., & Duguid, P. (1991). Organizational learning and communities-of-practice: Toward a unified view of working, learning, and innovation.Organization Science, 2(1), 40–57.

Carter, A. P. (1989). Know-how trading as economic exchange. Research Policy,18, 155–163.

Chin, W. W. (1998). The partial least squares approach to structural equation mod-eling. In G. A. Marcoulides (Ed.), Modern methods for business research.Mahway, NJ: Erlbaum, 295–336.

Chin, W. W., & Frye, T. (1996). PLS Graph (Version 2.91.03.04). Calgary, Canada:University of Calgary.

Chin, W. W., & Todd, P. A. (1995). On the use, usefulness, and ease of use of struc-tural equation modeling in MIS research: A note of caution. MIS Quarterly,19(2), 237–246.

Churchill, G. A., Ford, N. M., Hartley, S. W., & Walker, O. C. (1985). The deter-minants of salesperson performance: A meta-analysis. Journal of MarketingResearch, 22, 103–118.

Cohen, W. M., & Levinthal, D. A. (1990). Absorptive capacity: A new perspectiveon learning and innovation. Administrative Science Quarterly, 35, 128–152.

Compeau, D. R., & Higgins, C. A. (1995). Computer self-efficacy: Developmentof a measure and initial test. Information Systems Research, 6(2), 118–143.


Constant, D., Sproull, L., & Kiesler, S. (1996). The kindness of strangers: Theusefulness of electronic weak ties for technical advice. Organization Science,7(2), 119–135.

Davenport, T. H., & Prusak, L. (1998). Working knowledge. Boston: Harvard Busi-ness School Press.

Faraj, S., & Wasko, M. M. (1998). The web of knowledge: An investigation ofself-organizing. Communities of Practice on the Net. Paper presented at theInstitute for Operation Research and the Management Sciences (INFORMS)annual meeting, Seattle, Washington.

Fornell, C., & Larcker, D. (1981). Evaluating structural equation models with un-observable variables and measurement error. Journal of Marketing Research,18, 39–50.

Fulk, J., & DeSanctis, G. (1995). Electronic communication and changing organi-zational forms. Organization Science, 6(4), 337–350.

Gefen, D., & Straub, D. W. (1997). Gender differences in the perception and useof e-mail: An extension of the technology acceptance model. MIS Quarterly,21, 389–400.

Gerstberger, P., & Allen, T. J. (1968). Criteria used by research and developmentengineers in the selection of an information source. Journal of Applied Psy-chology, 52(4), 272–279.

Granovetter, M. S. (1973). The strength of weak ties. American Journal of Sociol-ogy, 91, 481–510.

Granovetter, M. S. (1983). The strength of weak ties: A network theory revisited.Sociological Theory, 1, 201–233.

Grant, R. (1996a). Prospering in dynamically competitive environments: Orga-nizational capability as knowledge integration. Organization Science, 7(4),375–387.

Grant, R. (1996b). Toward a knowledge-based theory of the firm. Strategic Man-agement Journal, 17(Special Issue), 109–122.

Hauptman, O. (1986). Influence of task type on the relationship between commu-nication and performance: The case of software development. R&D Man-agement, 16(2), 127–140.

Heneman, H. G. (1974). Comparisons of self and superior rating of managerialperformance. Journal of Applied Psychology, 59, 638–642.

Hinds, P., & Kiesler, S. (1995). Communication across boundaries: Work, structure,and use of communication technologies in a large organization. OrganizationScience, 6(4), 373–393.

Kogut, B., & Zander, U. (1992). Knowledge of the firm, combinative capabilities,and the replication of technology. Organization Science, 3(3), 383–397.

Kollock, P. (1999). The economies of online cooperation: Gifts and public goods incyberspace. In M. A. Smith & P. Kollock (Eds.), Communities in Cyberspace.London: Routledge, 220–239.


Lakhani, K., & Von Hippel, E. (2000). How open source software works: “Free”user-to-user assistance. Paper presented at the 3rd Intangibles Conference,“Knowledge: Management, Measurement and Organization,” Stern Schoolof Business, New York University.

Leonard-Barton, D. (1992). Core capabilities and core rigidities: A paradox in man-aging new product development. Strategic Management Journal, 13(SummerSpecial Issue), 111–126.

Levitt, B., & March, J. G. (1988). Organizational learning. Annual Review of So-ciology, 14, 319–340.

Messick, S. (1980). Test validity and the ethics of assessment. American Psychol-ogist, 35, 1012–1027.

Monge, P. R., Rothman, L. W., Eisenberg E. M., Miller, K. I., & Kirste, K. K.(1985). The dynamics of organizational proximity. Management Science,31(9), 1129–1142.

Nahapiet, J., & Ghoshal, S. (1998). Social capital, intellectual capital, and theorganizational advantage. Academy of Management Review, 23(2), 242–266.

Nohria, N., & Eccles, R. G. (1992). Face-to-face: Making network organizationswork. In N. Nohria & R. G. Eccles (Eds.), Networks and organizations:Structure, form, and action. Boston: Harvard Business School Press, pp.288–308.

Nonaka, I. (1994). A dynamic theory of organizational knowledge creation. Orga-nization Science, 5(1), 14–35.

O’Reilly, C. (1982). Variations in decision makers’ use of information sources: Theimpact of quality and accessibility of information. Academy of ManagementJournal, 25(4), 756–771.

Orlikowski, W. J. (1996). Learning from Notes: Organizational issues in group-ware implementation. In R. Kling (Ed.), Computerization and Controversy.New York: Academic, 173–189.

Pickering, J. M., & King, J. L. (1995). Hardwiring weak ties: Interorganizationalcomputer-mediated communication, occupational communities, and organi-zational change. Organization Science, 6, 479–486.

Schrader, S. (1991). Informal technology transfer between firms: Cooperationthrough information trading. Research Policy, 20, 153–170.

Spender, J. (1996). Making knowledge the basis of a dynamic theory of the firm.Strategic Management Journal, 17(Special Issue), 45–62.

Sproull, L., & Faraj, S. (1995). Atheism, sex, and databases: The net as a socialtechnology. In B. K. J. Keller (Ed.), Public access to the Internet. Cambridge,MA: MIT Press, 62–82.

Szulanski, G. (1996). Exploring internal stickiness: Impediments to the transferof best practice within the firm. Strategic Management Journal, 17(SpecialIssue), 27–44.


Taylor, R. L. (1975). The technological gatekeeper. R&D Management, 5, 239–242.

Teigland, R. (2000). Communities of practice in a high technology firm. In J.Birkinshaw & P. Hagstrom (Eds.), The flexible firm: Capability managementin network organizations. Oxford: Oxford University Press, 126–146.

Thibaut, J. W., & Kelley, H. H. (1959). The social psychology of groups. New York:Wiley.

Tushman, M., & Katz, R. (1980). External communication and project perfor-mance: An investigation into the role of gatekeepers. Management Science,26(11), 1071–1085.

Van Maanen, J., & Barley, S. R. (1984). Occupational communities: Culture andcontrol in organizations. In B. M. Staw & L. L. Cummings (Eds.), Researchin organizational behavior. Vol. 6, Greenwich, CT: JAI Press.

Von Hippel, E. (1987). Cooperation between rivals: Informal know-how trading.Research Policy, 16, 291–302.

Von Hippel, E., & Schrader, S. (1996). “Managed” informal information trading:The oil scout system in oil exploration firms. International Journal of Tech-nology Management, 11, 207–218.

Wasko, M., & Faraj, S. (1999). Webs of knowledge: An examination of knowledgetypes and knowledge flows in an electronic community of practice. Paperpresented at the Academy of Management Conference, Chicago.

Wasko, M., & Faraj, S. (2000). It is what one does: Why people participate and helpothers in electronic communities of practice. Journal of Strategic InformationSystems, 9(2/3), 155–173.

Wenger, E. (1998). Communities of practice. Cambridge, England: CambridgeUniversity Press.

Werts, C. E., Linn, R. L., & Joreskog, K. G. (1973). Intraclass reliability esti-mates: Testing structural assumptions. Educational and Psychological Mea-surement, 34, 25–33.

Wexley, K. N., Alexander, J. P., Greenawalt, J. P., & Couch, M. A. (1980). At-titudinal congruence and similarity as related to interpersonal evaluation inmanager-subordinate dyads. Academy of Management Journal, 23, 320–330.

Wold, H. (1982). Soft modeling: The basic design and some extensions. In H. Wold(Ed.), Systems under indirect observation: Causality, structure, prediction.New York: North Holland, 1–47.

Robin Teigland is a PhD candidate at the Institute of International Business at theStockholm School of Economics in Sweden. She received her BA with distinctionin economics at Stanford University. After focusing on operations managementwhile obtaining an MBA at Wharton and an MA in international studies at theUniversity of Pennsylvania and subsequently working as a consultant at McKinseyand Company, she has turned her attention to the organization and managementof knowledge work. Her research interests include the strategic use of information


technology in these activities and, in particular, the relationship between a firm’scompetitive advantage and the informal extra-organizational boundary spanningactivities by the firm’s employees. She has taught a variety of courses in strategy,knowledge management, and social capital at the undergraduate, graduate, andexecutive levels. More information on her can be found at http://www.teigland.com.

Molly McLure Wasko is assistant professor in the Department of ManagementInformation Systems in the College of Business at Florida State University. She re-ceived her BBA in management at James Madison University, her MBA at AverettUniversity, and her PhD in MIS from the University of Maryland, College Park.Her research interests include the social network structures of electronic commu-nication, electronic cooperation and collaboration, the strategic management ofinformation technology, and leveraging IT to create a competitive advantage.

Integrating knowledge through information trading: Examining the relationship between boundary spanning communication and individual performance

Documents