Inter-Organizational Knowledge Development in IT Outsourcing

Transcription

1 Associaion for Informaion Sysems AIS Elecronic Library (AISeL) AMCIS 26 Proceedings Americas Conference on Informaion Sysems (AMCIS) December 26 Iner-Organizaional nowledge Developmen in I Ousourcing Qing Chang Universiy of Cenral Florida Mihir Parikh Universiy of Cenral Florida Follow his and addiional works a: hp://aisel.aisne.org/amcis26 Recommended Ciaion Chang, Qing and Parikh, Mihir, "Iner-Organizaional nowledge Developmen in I Ousourcing" (26). AMCIS 26 Proceedings hp://aisel.aisne.org/amcis26/395 his maerial is brough o you by he Americas Conference on Informaion Sysems (AMCIS) a AIS Elecronic Library (AISeL). I has been acceped for inclusion in AMCIS 26 Proceedings by an auhorized adminisraor of AIS Elecronic Library (AISeL). For more informaion, please conac elibrary@aisne.org.

2 nowledge Developmen in I Ousourcing Iner-Organizaional nowledge Developmen in I Ousourcing Qing Chang Deparmen of Mgm. Info. Sysems Universiy of Cenral Florida 4 Cenral Florida lvd., A-1:32 Orlando, FL , USA qing.chang@bus.ucf.edu Mihir A. Parikh Deparmen of Mgm. Info. Sysems Universiy of Cenral Florida 4 Cenral Florida lvd., A-1:32 Orlando, FL , USA mihir.parikh@bus.ucf.edu ASRAC o examine he effecs of iner-organizaional aciviies on knowledge developmen in I ousourcing projecs, we develop a research framework based on a knowledge-based view of organizaions. Using his framework, we model he process of knowledge developmen and uilize simulaion o generae various scenarios wih differen levels of iner-organizaional aciviies. We evaluae he developmen of overall knowledge capabiliies in hose scenarios. he resuls indicae ha knowledge developmen is a highly dynamic process, specifically in he iner-organizaional seings. Even a marginal change in he level of iner-organizaional aciviies can have significan effecs on he abiliies of he clien and he ousourcing vendor o develop knowledge capabiliies required for he projec. eywords: I Ousourcing; nowledge Managemen; Simulaion INRODUCION Organizaions are increasingly ousourcing heir informaion echnology (I) operaions and sysems developmen. Mos I ousourcing cases involve he ransfer of no only I asses and personnel bu also knowledge asses and decision-making conrol (Choudhury and Sabherwal 23; Elizur and Wensley 22). herefore, undersanding he process of exchanging he exising knowledge and creaing new knowledge beween he clien and vendor during an I ousourcing projec is imporan. I can lead o more effecive developmen of knowledge capabiliies required for he projec and greaer projec success (Elizur and Wensley 22; Lee and im 25; Willcocks e al. 24; Willcocks and ern 1998). In an I ousourcing projec, he vendor and clien have heir unique ses of knowledge. When hese knowledge ses are complimenary o each oher, hey can provide significan benefis o he projec. owever, ha requires undersanding: Wha ype of business and echnology knowledge is needed for he projec? Who, he vendor or he clien, has ha knowledge? Which par of ha knowledge is aci and which par is explici? ow do he vendor and clien exchange his knowledge and use i o develop new knowledge? Wha are he facors, specifically organizaional aciviies and srucure, which may influence his knowledge exchange and creaion process? o address hese research quesions, we analyze he composiion of knowledge in I ousourcing projecs. We decompose he knowledge developmen process and idenify he mos imporan iner-organizaional aciviies ha influence he efficiency of each of is componens. We, hen, build simulaion sudies o creae and manipulae differen business scenarios and examine he impac of hese facors under hose scenarios. he use of simulaion o represen organizaional seings and processes is popular, because i enforces inernal consisency of he heory and uncovers heory s inuiively unobvious assumpions (Repenning 22; Sasry 1997). EOREICAL ACGROUND In recen years, a knowledge-based view has emerged as an exension of he resource-based view, as knowledge asses possessed by a firm are considered criical resources (Conner 1991; Gran 1996). he knowledge-based view assumes ha organizaional efficiency and compeence are consequences of he possession and use of unique knowledge. Using he knowledge-based view, Nonaka (1993) describes a hree-layer framework for knowledge exchange and creaion in organizaions: Projec-eam layer, business-sysem layer, and knowledge-base layer. Members of he projec eam collaborae Proceedings of he welfh Americas Conference on Informaion Sysems, Acapulco, Mexico Augus 4 h -6 h

3 nowledge Developmen in I Ousourcing o exchange and creae projec-relaed knowledge in he projec-eam layer under he overarching vision and objecives of he projec. hey hen move down o he knowledge-base layer o documen, re-caegorize and re-conexualize he newly acquired knowledge. Finally, hey reurn o heir business-sysem layer, where hey apply he knowledge relevan for heir own business aciviies. he abiliy o move swifly from one layer o anoher deermines he organizaional capabiliy for knowledge creaion. We use his heoreical framework o analyze knowledge in iner-organizaional relaionships such as I ousourcing. In informaion sysems developmen process, knowledge is mainly composed of echnical knowledge and business knowledge (Rus and Lindvall 22). echnical knowledge refers o he knowledge abou sysem design, programming, and developmen processes; while business knowledge refers o he knowledge abou he inended user s business processes, business rules, and requiremen for he new sysem. In I ousourcing projecs, business knowledge mainly comes from he clien s experience in business issues and echnical knowledge from vendors experience in echnical issues (Willcocks e al. 24). he success of I ousourcing is deermined by he fi beween oal projec knowledge requiremens for boh echnical and business knowledge and he knowledge capabiliies of he clien and vendor(willcocks e al. 24). iwana (23) empirically demonsraes ha he knowledge overlap in boh echnical and business knowledge area across he clien and vendor is very imporan in I ousourcing. his means ha we canno assume ha echnical knowledge would come only from he vendor and business knowledge would come only from he clien. oh he vendor and clien may have specific ses of echnical and business knowledge, which may even overlap. he effeciveness of he exchange of hese knowledge ses and he creaion of addiional projec-specific knowledge would deermine he knowledge capabiliies and evenually he oucome of he I ousourcing projec (Lee and im 25). In addiion, knowledge in wo forms: explici and aci knowledge (Nonaka 1994). Explici knowledge can be codified and ransmied in formal language. Sofware codes and sandard operaing procedures are examples of explici knowledge. aci knowledge is absrac and rooed in acion, communicaion and involvemen in a specific conex. nowing how o normalize a daabase or how o design an IS are examples of aci knowledge. Nonaka (1994) proposed four modes of knowledge exchange and creaion. he process of sharing aci knowledge is called socializaion. he process of inegraing explici knowledge is called combinaion. he process of convering aci knowledge o explici knowledge is called exernalizaion. he process of convering explici knowledge o aci knowledge is called inernalizaion. he firs wo processes refine and augmen he original knowledge base in he same form. he las wo processes conribue by creaing a new form of knowledge. In I ousourcing projecs, hese processes occur a inerorganizaional level beween he clien and vendor. RESEARC FRAMEWOR ased on he heoreical background discussed in he previous secion, we develop a research framework (Figure 1) ha oulines he four knowledge exchange and creaion processes in iner-organizaional seings and ariculaes aciviies communicaion, coordinaion, organizaion culure, and documenaion ha may influence hese processes. Previous sudies have indicaed ha exernal and inernal environmens influence knowledge processes (all 1997). Organizaions adjus heir srucures, aciviies and resource allocaions o fi heir knowledge capabiliies wih he environmen hey face. ow hose adjusmens may ake place in I ousourcing and heir effecs are discussed below. Socializaion and Communicaion Nonaka (1994) refers o socializaion as a process of sharing implici experiences and creaing aci knowledge such as shared menal models and echnical skill. In his aci knowledge exchange process, differen media offer differen capabiliies o exchange knowledge. Media richness heory (Daf and Lengel 1986) proposes a prescripive model o achieve a fi beween informaion processing requiremen and communicaion channel o increase organizaional effeciveness (Markus 1994). Daf and Lengel (1986) described he amoun of informaion can faciliae he reducion of uncerainy and he srucural mechanism can reduce equivocaliy. In many I ousourcing cases, he clien and vendor are geographically separaed and various communicaions channels provide he basis for exchanging informaion. hey ype of media used can play an imporan role in his knowledge exchange process. Selecion fi media is no only considering he congruenly mach for knowledge requiremen bu also he feasibiliy o increase he efficiency and effecs. Proceedings of he welfh Americas Conference on Informaion Sysems, Acapulco, Mexico Augus 4 h -6 h

4 nowledge Developmen in I Ousourcing Figure 1: Research Framework Exernalizaion and Documenaion For aci knowledge o become explici knowledge, creaion of documens, manual, and oral sories are very helpful (Nonaka and onno 1993). I could help sor and collec individual knowledge, as knowledge is seen in very personal erms. As a consequence, i increases he access o knowledge in organizaion and reduces knowledge loss. In iner-organizaional siuaions like I ousourcing, knowledge documenaion follows exising models or sysems shared by he wo organizaions. Such collecive documenaion process will no only faciliae he emergence of a shared menal model, bu also help inquiry and knowledge generaion in fuure hrough synhesis. nowledge gaps across he vendor and clien organizaions can be parly filled by exernalizaion hrough documenaion. he sandard and sufficien documens which boh vendor and user can accuraely undersand and access will help develop srong knowledge capabiliies. Combinaion and Coordinaion Combinaion is he process of soring, adding, comparing, combining, and caegorizing explici knowledge from various sources; in I ousourcing, from he clien and vendor. o effecively inegrae various sources of knowledge in wo organizaions, coordinaion beween hem is necessary (all 1997). he level of coordinaion defines he degree o which he wo organizaions will cooperae. In I ousourcing, coordinaion represens he cooperaion beween he clien and vendor. Faraj and Sproull (2) empirically illusraed ha knowledge coordinaion was differen from adminisraive coordinaion in IS sofware developmen projecs. For verical coordinaion, he srucure is clearly pre-defined in he ousourcing conrac. Depending on he srucure of he conrac, eiher he clien or he vendor dominaes he coordinaion process. Verical coordinaion will guaranee effecive exchange of knowledge. For horizonal coordinaion, he srucure is raher i is loose and emergen. ierarchical barriers are avoided o encourage beer coordinaion among peers in boh organizaions. his Proceedings of he welfh Americas Conference on Informaion Sysems, Acapulco, Mexico Augus 4 h -6 h

5 nowledge Developmen in I Ousourcing faciliaes he knowledge exchange process a various levels wihin an I ousourcing projec. I is criical when knowledge required for he ousourcing is no clearly-defined and neiher side has upper hand. Inernalizaion and Organizaional Culure Inernalizaion means convering wriings or oher ariculaion of language ino aci knowledge. ere, he fundamenal issues are relaed wih individuals menal models, beliefs, and cogniive assumpions (Nonaka and onno 1993). Meaphors, analogies, and models ac as bridges beween objecive knowledge and inuiive undersanding (Nonaka 1994). Organizaional or corporae culure includes he values and believes expressed in arifacs, symbols and pracices as well as organizaional language, radiions, myhs, riuals, and sories (Zakaria e al. 24, p.19). Organizaional culure also refers o a value sysem of risk, reward, warmh, and suppor (James and Jones 1979). Organizaional culure and climae, hus, provides he background for drawing meaphors, analogies, and models necessary for inernalizaion and shape individuals menal models, beliefs, and cogniive assumpions (Janz and Prasarnphanich 23). In I ousourcing, he degree of fi beween organizaional culures of he clien and vendor firms will deermine he level of inernalizaion of explici knowledge gained from he oher side. NOWLEDGE DEVELOPMEN PROCESS MODEL ased on he discussion and research framework developed in he previous secions, we develop a process model ha consiss of wo ypes of knowledge relevan for I ousourcing: usiness knowledge and echnical knowledge. he oal projec-relaed knowledge can be represened as a weighed summaion of hese wo ypes. he vendor and clien have heir own ses of aci and explici knowledge for each of business and echnical knowledge. During I ousourcing, hey inerac synergisically o exchange and creae new knowledge. herefore, we pu his logical descripion as follows: usiness nowledge = aci usiness knowledge + Explici usiness nowledge echnical nowledge = aci echnical nowledge + Explici echnical nowledge Projec-Relaed nowledge = W (weigh) * usiness nowledge + W (weigh) * echnical nowledge Modeling usiness nowledge Exchange and Creaion nowledge creaion is a dynamic process, so i produces cycles or feedback loops over a period of ime (Nonaka 1994). ime relaes o he rae a which he knowledge is exchanged and new knowledge is creaed and he relaive speed wih which knowledge ranslaion changes (all e al. 21; u e al. 1997). We use () o represen he cycling ime effec. A he sar of he projec, = ( ) and a any laer given poin of ime i is =. he deail index of parameers is given in Appendix A. A he sar of an I ousourcing projec, he vendor has some level of iniial aci business knowledge. We express i as V. Now, as he projec progresses, a par of he clien s knowledge will sar ransferring o he vendor hrough one of he four knowledge exchange and creaion processes. A he beginning of his process, he rae of ransfer is small. owever, as he wo organizaions organizaional aciviies synchronize he ransfer rae will increase. Evenually, he ranslaion rae will reach a maximum poin and begin o sabilize. In he Nonaka s (1995) Spiral Developmen Model, organizaional knowledge increases in amoun and qualiy hrough a spiral shape, for which a sine funcion is well-suied. herefore, we selec a sine funcion o simulae his shifing of conversion rae over ime. During he I ousourcing projec, as discussed above, a par of he clien s aci business knowledge ( F o ransferred o he vendor hrough he socializaion process. he raio a which his exchange akes place is. hus, a any given poin of ime, he value of he clien s aci business knowledge gained by he vendor as new aci knowledge can be expressed as: ) ges F sin( ) d [1] During he same ime, a par of he clien s explici business knowledge ( E F o ) will also ransfer o he vendor hrough inernalizaion process, in which undersands explici business rules and processes of he clien firm. he raio a which his ransfer akes place is E. his conversion, however, does no ake place a he same rae as explici knowledge exchange. Proceedings of he welfh Americas Conference on Informaion Sysems, Acapulco, Mexico Augus 4 h -6 h

6 nowledge Developmen in I Ousourcing Given he complexiy of conversion, here could be a delay, which can be represened as Sin( + ω ). hus, a any given poin of ime, he value of he clien s explici business knowledge gained by he vendor as new aci knowledge is: [2] E E F d sin( + ω 1 ) hus, overall aci business knowledge available wih he vendor a any given poin of ime can be expressed as he sum of above individual componens: E E F F = V + d + + d sin( ) sin( ω1) [3] Similarly, a he sar of he vendor has some level of iniial explici business knowledge. We express i as I ousourcing projec, a par of he clien s explici business knowledge ( E F o E V. During he ) ges ransferred o he vendor hrough he combinaion process. he raio a which his exchange akes place is. hus, a any given poin of ime, he value of he clien s explici business knowledge gained by he vendor as new explici knowledge can be expressed as: EE E F sin( ) During he same ime, a par of he clien s aci business knowledge ( EE F o [4] ) will also ransfer o he vendor hrough exernalizaion process, in which he vendor creaes explici documens or oher forms of formal expressions such as operaing processes or qualiy checks. he raio a which his ransfer akes place is. As discussed earlier, here could be some delay in his conversion process oo. hus, a any given poin of ime, he value of clien s aci business knowledge gained by he vendor as new explici knowledge is:. E F sin( + ω 2 ) hus, overall explici business knowledge available wih he vendor a any given poin of ime can be expressed as he sum of above individual componens: EE E E E E F F = V + d + + d sin( ) sin( ω2) [6] he idenical knowledge exchange and creaion for business knowledge akes place for he clien. hus, over a long period of ime he clien s and he vendor s knowledge ses relaed o he I ousourcing projec will converge and compleely overlap. Since we are ineresed in he developmen of overall knowledge capabiliies (he converged knowledge se) hrough he process of knowledge exchange and creaion, we are no including he process from he clien side in he simulaion model o avoid double calculaion of he knowledge. In he conversion process, we also assume ha all pars of knowledge carry he same degree of imporance, and each par impacs he whole sysem. In sum, all he knowledge akes par in he conversion process and shows influence during he whole process. Modeling echnical nowledge Exchange and Creaion. d ere, our approach is he similar o ha for business knowledge and develop he following models for aci and explici echnical knowledge. For aci echnical knowledge: [5] d E Proceedings of he welfh Americas Conference on Informaion Sysems, Acapulco, Mexico Augus 4 h -6 h

7 nowledge Developmen in I Ousourcing E E F F = V + d + + sin( ) sin( For explici echnical knowledge: EE E E E E F F = V + d + + sin( ) sin( ω ) ω ) 3 d 4 d [7] [8] Overall Model As discussed earlier, organizaional knowledge is he summaion of weighed aci and explici knowledge. here is a debae abou he amoun of values for he weighs for aci and explici knowledge in organizaional knowledge (Nonaka 1994). o enable, he organizaions o choose heir own poin of view on his debae, we use a general formulaion for weighs. he oal projec-oriened knowledge is expressed as: E E E E = W W + W ) + W ( W + W ) [9] G ( 1 In he equaion, W, W are he weigh for overall business and echnical knowledge. business aci knowledge and business explici knowledge. echnical explici knowledge. RESULS AND DISCUSSION W and E W and W are he weighs for W are he weighs for echnical aci knowledge and We use simulaion as a research mehod o examine he effecs of he four organizaional facors on knowledge exchange and creaion process beween he vendor and clien organizaions. Simulaion provides a conrolled mehod o invesigae a complex phenomenon, and sudies he effecs of alernaive processes or siuaions on oucomes hrough manipulaions of key influenial facors. In he simulaion, we developed six differen cases by manipulaing various facors in he model. For each case, we randomly generaed 1 scenarios. We used MahLA sofware o run he simulaion for he oal 6 scenarios. he simulaion included hree key variables: ransfer raes, knowledge level, and ime. he complee range of he ransfer raes were sandardized o vary from o 1, where being none and 1 being he opimal (no maximum as he higher values may provide diminishing reurns). he iniial value of knowledge was arbirarily se a 4 and allowed i o vary wihou any limi. he ime facor was also sandardized o vary from o 1, where being he sar of he projec and 1 being he end of he projec. All weighs were se equally a 1, hough hey can be changed based on organizaional siuaions. Effecs of an Individual Facor Manipulaion Our research model idenifies four organizaional facors affecing four iner-organizaional knowledge exchange and creaion processes. In he firs wo cases, we manipulae only one communicaions of he four facors; oher hree facors are sable. Communicaions affec he socializaion process, in which aci knowledge is ransferred from one organizaion o he oher organizaion. Socializaion is represened by (ransfer rae of aci business knowledge) and (ransfer rae of aci echnical knowledge). We chose he range of.6 o.9 for high level of communicaions and.1 o.4 for low level of communicaions. Oher s have he sable, middle value of.5. he resul of he 2 sample simulaions is given in able 1. E Manipulaion Case 1 igh Level of Communicaions Case2 Low Level of Communicaions he range of G G and Mean Sd. Dev σ = σ =.15 able 1: Resuls of One Facor Manipulaion -value p-value 38.4 <.1 he resuls indicae ha here is significan difference in oal knowledge generaion when one facor is manipulaed. Figure 2 shows he values of G for all observaions for boh Cases 1 and 2. he poins are relaively less scaered indicaing small amoun of variaion among differen observaions. Proceedings of he welfh Americas Conference on Informaion Sysems, Acapulco, Mexico Augus 4 h -6 h

8 nowledge Developmen in I Ousourcing Case 1 Case 2 G Figure 2: One Facor Manipulaion - Case 1 & Case 2 Figure 3: Simulaion-One Facor hese cases provide a basic example of he simulaion. A quaniaive measuremen of he effecs of changes in he organizaional performance wih he adjusmen of exernal or inernal environmen would help managers build a reliable managemen ool for predicing fuure operaion. Given oher condiions being equal, he efficiency is as imporan as causal relaionship. In hese scenarios, we manipulaed only he level of communicaion for is impac on aci knowledge ransfer. Figure 3 shows, how such manipulaion can affec he knowledge level over a period. If he level of socializaion remains he same he over ime, he knowledge graph will be wo dimensional wih ime on one axis and knowledge level on oher axis. owever, if he level of socializaion changes hrough improving/deerioraing communicaions beween he clien and vendor, he ransfer rae would change over ime and, evenually, he final knowledge level would also change as shown by he hree-dimensional hea char. Effecs of wo Facors Manipulaion In he Cases 3 and 4, we manipulaed wo facors and kep he oher wo facors sable. his manipulaion represens exchange and creaion of aci knowledge, hrough he socializaion and inernalizaion processes, in which boh aci and explici knowledge from one organizaion is ransferred ino he oher organizaion as new aci knowledge. his infusion of new projec-relaed knowledge increases he knowledge capabiliies of he receiving organizaion and may lead o he projec E E success. ere, we manipulaed four s: wo ( and ) relaed o socializaion and wo ( and ) relaed o inernalizaion. Again, we chose he range of.6 o.9 for high levels of communicaions and organizaional culure fi beween he wo organizaions and.1 o.4 for he low levels. Again, he remaining four s were kep a he sable, middle value of.5. he resuls of he 2 sample simulaions are given in able 2. Case 3 Case 4 Manipulaion igh Levels of Media Richness and Org. Culure Fi Low Levels of Media Richness and Org. Culure Fi he range of E and E.6-.9 σ = σ =.15 G G Mean Sd. Dev able 2: Resuls of wo Facors Manipulaion -value p-value <.1 he resuls indicae ha here is significan difference in oal knowledge generaion when wo facors are manipulaed. Figure 3 shows he values of G for all observaions for boh Cases 3 and 4. he poins are more scaered han hose in Figure 2 suggesing ha he variaion in knowledge generaion increases, while he average knowledge generaed may be higher. Proceedings of he welfh Americas Conference on Informaion Sysems, Acapulco, Mexico Augus 4 h -6 h

9 nowledge Developmen in I Ousourcing Case 3 Case 4 G Figure 4: wo Facors Manipulaion - Case 3 & Case 4 Figure 5: Simulaion - wo Facors hese cases are a bi more inricae han he previous wo as hey involve simulaneous manipulaion of wo facors. hey sugges ha a manager can influence he exchange and creaion of a specific form of knowledge by focusing on specific ypes of aciviies. owever, someimes, hose aciviies are beyond he immediae conrol of he manager. In hese cases, we manipulaed he levels of communicaions and organizaional culural fi beween he clien and vendor. he simulaions show ha he level of aci knowledge increases wih he higher levels. he sub-conex of communicaions and culure may have variaions, bu undersanding heir influence on knowledge ransfer and creaion processes and a specific form of knowledge may provide managers wih a clear direcion oward where he or she should allocae resources o achieve specific knowledge capabiliy required for he projec. Effecs of All Facors Manipulaion For he Cases 5 and 6, we manipulaed all four facors. his enabled us o examine effecs of hese facors on all of he four knowledge exchange and creaion processes, in which boh aci and explici knowledge from one organizaion is ransferred and convered ino new aci and explici knowledge in he oher organizaion. o enable his, we manipulaed all eigh s in he simulaion model. ere also, we chose he range of.6 o.9 for high levels of communicaions, organizaional culure fi, combinaion, and coordinaion beween he wo organizaions and.1 o.4 for he low levels. he resuls of he 2 sample simulaions are given in able 3. Manipulaion Case 5 igh Levels on All Facors Case 6 Low Levels on All Facors he range of all s G G Mean Sd. Dev σ = σ =.15 able 3: Resul of All Facor Manipulaion -value p-value <.1 he resuls again indicae ha here is significan difference in oal knowledge generaion when all four facors are manipulaed simulaneously for heir high levels versus heir low levels. Figure 4 shows he values of G for all observaions for boh Cases 5 and 6. he poins are much more scaered han hose in Figure 2 and Figure 3 suggesing ha he variaion in knowledge generaion increases significanly wih manipulaion of more facors. Proceedings of he welfh Americas Conference on Informaion Sysems, Acapulco, Mexico Augus 4 h -6 h

10 nowledge Developmen in I Ousourcing Case 5 Case 6 G Figure 6: All Facors Manipulaion - Case 5 & Case 6 Figure 7: Simulaion - All Facors hese cases are he mos inricae, as hey involve manipulaion of all facors. he simulaions show ha he success of knowledge managemen could be achieved hrough wo differen ways. Firs, organizaions could adjus heir srucures, adop new echnologies, build proacive firm culure, and specify sraegies o influence he iner-organizaional aciviies affecing knowledge managemen. his will increase he exchange and creaion rae leading o higher level of knowledge. Second, organizaions could use he dynamic naure of knowledge creaion ha leads o expanding he level of knowledge. hey could build a knowledge-based layer in heir organizaions enabling he projec paricipans o inegrae organizaional knowledge resources. Organizaions operae in complex environmen. hese cases sugges ha managers may need o adjus differen composiion of knowledge asse, sraegy arrangemen, and organizaional srucure o gain shor-erm and longerm knowledge advanage. CONCLUSION As I ousourcing is becoming a sraegic aciviy, he developmen of shared knowledge capabiliies becomes criical for he success of increasingly rich and complex I ousourcing relaionships. In his paper, we examined how iner-organizaional aciviies, specifically level of communicaion, organizaional culural fi, documenaion, and coordinaion beween he vendor and clien, affec he knowledge exchange and creaion process in I ousourcing. We used he knowledge-based view o decompose his process and used simulaions o generae various business scenarios o evaluae effecs of various inerorganizaional aciviies on is componens. o reconcile he cos-radeoff and necessiy of knowledge exchange and creaion in I ousourcing projecs, managers need an effecive measuremen and scienific forecasing ools o adjus heir organizaional aciviies. he process model presened in his paper provides a ool for managers o make effecive decisions regarding resource allocaions for hose aciviies. hey can fine-une he model and se appropriae iniial values of parameers based on heir undersanding abou heir own organizaions, previous experience, hisorical daa, and oher quaniaive and qualiaive analysis abou knowledge requiremen in I projec. he four iner-organizaional aciviies used in his sudy serve as examples for aciviies having significan effecs on inerorganizaional knowledge managemen. Oher aciviies may also have significan effecs on knowledge creaion and exchange processes. hey may even inerac wih hese aciviies and moderae heir effecs over ime. A fuure work may explore his line of research and furher improve he model. Simulaion is a useful modeling echnique. owever, like any oher models, such as an archiecural model, i is a basal represenaion of realiy hrough mahemaical relaionships. I canno compleely represen he real world (Repenning 22). Fuure sudies could build on his work o examine he knowledge exchange and creaion process in laboraory experimens wih ighly-conrolled manipulaions bu in a human-paricipan environmen. Oher research mehods such as survey and longiudinal field sudy may also be employed o evaluae resuls across differen research mehods and exend our model. Proceedings of he welfh Americas Conference on Informaion Sysems, Acapulco, Mexico Augus 4 h -6 h

11 nowledge Developmen in I Ousourcing REFERENCES 1. Choudhury, V., and Sabherwal, R. "Porfolios of Conrol in Ousourced Sofware Developmen Projecs," Informaion Sysems Research (14: 3), Sep. 23, pp Conner,.R. "A isory Comparison of Resource-ased heory and Five School of hough wihin Indusrial Organizaion Economics: Do We ave a New School of he Firm?" Journal of Managemen (17: 1), 1991, pp Daf, R., and Lengel, R. "Organizaional Informaion Requiremens, Media Richness and Srucural Design," Managemen Science (32), 1986, pp Elizur, R., and Wensley, A..P. "Furher houghs on Informaion Srucure, nowledge Managemen, and Ousourcing," in Informaion Sysems Ousourcing, R. irschheim, A. einzl and J. Dibbern (eds.), Springer, erlin, Germany, 22, pp Faraj, S., and Sproull, L. "Coordinaing Experise in Sofware Developmen eams," Managemen Science (46: 12), 2, pp Gran, R.M. "A nowledge-ased heory of Iner-Firm Collaboraion," Organizaion Science (7), 1996, pp all, D.J., Paradice, D.., and Courney, J.F. "Creaing Feedback Loops o Suppor Organizaional Learning and nowledge Managemen in Inquiring Organizaion," awaii Inernaional Conference on Sysem Sciences, Maui, awaii, all, R.. Organizaions: Srucures, Processes and Oucomes, (7 h ed ed.), u, Q., Saunders, C., and Gebel, M. "Diffusion of Informaion Sysems Ousourcing: A Reevaluaion of Influence Sources," Informaion Sysems Research (8: 3), 1997, pp James, R.J., and Jones, A.P. "Psychological Climae: Dimensions and Relaionships of Individual and Aggregaed Work Environmen Percepions," Organizaional ehavior and uman Performance (23), 1979, pp Janz,.D., and Prasarnphanich, P. "Undersanding he Anecedens of Effecive nowledge Managemen: he Imporance of a nowledge-cenered Culure," Decision Sciences (34: 2), 23, pp Lee, J.-N., and im, Y.-G. "Undersanding Ousourcing Parnership: A Comparison of hree heoreical Perspecives," IEEE ransacions on Engineering Managemen (52: 1), Feb. 25, pp Markus, L. "Elecronic Mail as he Medium of Managerial Choice," Organizaion Science (5: 4), 1994, pp Nonaka, I. "A Dynamic heory of Organizaional nowledge Creaion," Organizaion Science (5: 1), 1994, pp Nonaka, I., and onno, N. "nowledge ased Organizaion," usiness Review (41: 1), 1993, pp Nonaka, I., and akeuchi,. he nowledge-creaing Company, Oxford Universiy Press, New York, Repenning, N.P. "A Simulaion-ased Approach o Undersanding he Dynamic of Innovaion Implemenaion," Organizaion Science (13: 2), 22, pp Rus, I., and Lindvall, M. "nowledge Managemen in Sofware Engineering," IEEE Sofware (May-June), 22, pp Sasry, M.A. "Problems and Paradoxes in a Model of Puncuaed Organizaional Change," Adminisraive Science Quarerly (42: 2), 1997, pp iwana, A. "nowledge Pariioning in Ousourced Sofware Developmen: A Field Sudy," weny-fourh Inernaional Conference on Informaion Sysems, Seale, WA, 23, pp Willcocks, L.P., indle, J., Feeny, D., and Laciy, M. "I and usiness Process Ousourcing: he nowledge Poenial," Informaion Sysems Managemen (21: 3), Summer 24, pp Willcocks, L.P., and ern,. "I Ousourcing as Sraegic Parnering: he Case of he Uk Inland Revenue," European Journal of Informaion Sysems (7), 1998, pp Zakaria, N., Amelinckx, A., and Wilemon, D. "Working ogeher Apar? uilding a nowledge-sharing Culure for Global Virual eams," Creaiviy and Innovaion Managemen (13: 1), 24, pp Proceedings of he welfh Americas Conference on Informaion Sysems, Acapulco, Mexico Augus 4 h -6 h

12 nowledge Developmen in I Ousourcing APPENDIX A ; E E V ; V ; V V E E F ; F ; F ; F ; ; ; G E E : oal projec-oriened knowledge W : Weigh for overall business knowledge W : Weigh for overall echnical knowledge W ; W W ; W E E : Iniial aci()/explici(e) echnical()/business() knowledge from vendor(v) : Iniial aci()/explici(e) echnical()/business() knowledge from firm(f) : Overall aci/explici business knowledge available wih he firm/vendor a any gave ime poin : Weigh for business aci/explici knowledge : Weigh for echnical aci/explici knowledge E EE E ; ; ; : ransfer rae of aci/explici business knowledge from one pary o aci/explici business knowledge of anoher pary. E EE E ; ; ; : ransfer rae of aci/explici echnical knowledge from one pary o aci/explici echnical knowledge of anoher pary. Proceedings of he welfh Americas Conference on Informaion Sysems, Acapulco, Mexico Augus 4 h -6 h