Determinants of the Individual s Knowledge Sharing Behavior: The Theory of Reasoned Action Perspective

Transcription

1 Determinants of the Individual s Knowledge Sharing Behavior: The Theory of Reasoned Action Perspective Gee Woo Bock* and Young-Gul Kim** Department of Information Systems, School of Computing National University of Singapore* 3 Science Drive 2, Singapore bockgw@comp.nus.edu.sg Graduate School of Management, KAIST** Cheongryangri-Dong, Dongdaemoon-Gu, Seoul, Korea, domino2@unitel.co.kr Abstract This study theoretically examines and empirically tests factors expected to affect the individual s knowledge sharing behavior. Results from the field survey of 154 employees in 27 organizations over 16 industries show that expected associations affect the attitude toward knowledge sharing and the fairness of the organization affects the subjective norm to share knowledge. As expected, the attitude and the subjective norm are positively related to the knowledge sharing intention and, finally, to the actual knowledge sharing behavior. Rewards and information technology, believed by many as the most important motivating factors for knowledge sharing are found to play the facilitating role in the individual s knowledge sharing. Keywords: Knowledge Sharing, Theory of Reasoned Action, Attitude, Intention, Behavior Introduction Among the processes of knowledge management such as creation, sharing, utilization, and accumulation of knowledge, sharing is what differentiates organizational level knowledge management from individual learning or knowledge acquisition. In fact, 94% of 260 responses from multinational organizations in Europe believe that knowledge management requires people to share what they know with others in the organizations (Financial Times 1999). However, knowledge sharing is often unnatural because people do not share their knowledge when they think their knowledge is valuable and important. Hoarding knowledge and looking suspiciously upon knowledge from others are the natural tendencies (Davenport 1997). And to make things worse is that these tendencies are very difficult to change. 431 US and European organizations responded that the biggest difficulty in knowledge management was changing people s behavior (Ruggles 1998). Therefore, rather than just encouraging or mandating knowledge sharing, fostering the motivation to share knowledge is truly necessary. The purpose of this research is to develop an understanding of factors that support and constrain the individual s knowledge sharing behavior in organizations and empirically test how these factors eventually influence the behavior.

2 Background There has been a number of research to figure out factors affecting knowledge or information sharing behavior from various areas. For example, O Reilly, Chatman, and Anderson (1987) argued that a variety of pressures and inducements are used to regularize the behavior of individuals to make the independent contributions of organizational members to perform information transmission and use for decision-making from the information processing perspective. Three salient factors can be seen in the form of reward and punishment system: (1) organizational structure, (2) incentive systems, and (3) group pressure. Since early 1990s, researchers have tried to find out determinants of knowledge sharing as well as information sharing from the various perspectives, such as groupware, marketing, consumer research, leader-member exchange relationhip and MIS: Social structure (member relationship) and cognitive structure (Ward & Reingen, 1990), Extendedness of relationship and frequency of contact (Heide & Miner, 1992), Trust-dual concern model (self-interest and other s interest) (Butler, 1995), Mutual trust and influence (law of reciprocity) (Nelson & Cooprider, 1996), Information sharing norm and integrated goals (Fisher, Maltz, and Jaworski, 1997), Relational links, opportunity by status and motivation (Warketin and Sayeed, and Jightower, 1997), Motivation, opportunity, favoring own preference and evaluation apprehension (Dennis, Hilmer and Taylor, ), Partner s absorptive capacity and incentives (Dyer and Singh, 1998). However, there has been no consensus about which factors are really important and have more explanatory power over others because little research has been done empirically with solid theoretical foundations. This research tried to overcome these limitations. Because sharing knowledge in an organization can be considered one of human communication activities to augment a human s and organizational intellectual capability (Engelbart 1963, 1982), the study in organizational communication can provide useful perspective. There has been proposed four basic conceptual approaches in organizational communication: (1) mechanistic, (2) psychological, (3) interpretive-symbolic and (4) systemsinteraction perspectives (Krone, Jablin and Putnam 1987). While the mechanistic perspective toward human communication places primary emphasis on the channels that link communicators, the psychological perspective concentrates on explaining the informational environments in which individuals are located and the range of stimuli to which they respond using a variety of conceptual filters. According to Fisher (1978), conceptual filters consist of communicator s attitudes, cognitions, and perceptions. When organizational communication is conceptualized mechanistically or psychologically, the organization is viewed as containers within which interaction occurs (Hawes 1974). However, the interpretive-symbolic perspective posits that by virtue of their ability to communicate, individuals are capable of creating and shaping their own organization. Unlike the interpretive symbolic approach, research from the systems-interaction perspective concentrates on external behaviors as the fundamental units of analysis. The locus of communication is patterns of sequential behaviors or the recurrence of contiguous acts and interacts. These patterns take place within a communication system and serve to define that system. Hence, unlike the psychological perspective, an individual doesn t do communication, and he or she becomes a part of communication (Krone, Jablin and Putnam 1987). Although it may be possible for researchers to study some communication concepts from each perspective, one approach may be more suitable than others to study a particular

3 communication phenomenon (Krone, Jablin and Putnam 1987). In this research, by defining knowledge sharing as voluntary activities of transferring or disseminating knowledge from one person to another person or group in an organization (Huber and Daft 1987; Dyer and Singh 1998; Nelson and Cooprider 1996; Wiig 1995; Ruggles 1997), the underlying assumptions of the psychological perspective seems to be compatible with knowledge sharing behavior because we are interested in the organizational environment in which individuals are located and the range of stimuli to which they respond to the voluntary knowledge sharing behavior. Theoretical Framework and Research Model In social psychology, there have been four main perspectives to explain human behavior in a social context: (1) Psychoanalytic, (2) Behavioristic, (3) Cognitive and (4) Symbolic Interactionism. Symbolic Interactionism serves our research purpose the best among four social psychological perspectives because it explains the voluntaristic component in human conduct from the stand point of social actors as well as the individual self (Manis and Meltzer 1978). Besides, it covers the concept of conceptual filters which affect not only what information is attended to, conveyed and interpreted but also how this information is processed ((Krone, Jablin and Putnam 1987). According to the Self Theory in Symbolic Interactionism (Carver and Scheier 1985), a human behavior is determined by two different aspects; one is a private self and the other is a public self, based on one s rationality. Subsequently, the private self is affected by the external agent s stimuli and self-produced factors according to the social cognitive theory (Bandura 1986). In terms of external agent s stimuli, the social exchange theory (Blau 1967) has been dominant. As far as self-produced factors are concerned, none is more central than the employees judgment of their capabilities which comes from the self-efficacy concept (Stajkovic and Luthans 1998). On the other hand, the public self can be affected by the organizational climate (Dewhirst 1971). Pressures from the reference person or group may or may not fit into the subject s own personal opinions. However, if the congruence between the individual and the organizational values is high, people will voluntarily follow the instructions from the referent (O Reilly and Chatman 1986). This is why organizations develop the ways of behaving that is unique in any particular organization (Dewhirst 1971). In addition, Moran and Volkwein (1992) argued that organizational climate affected the individual s perception through inter-subjective processes which is similar to the public-self aspect. Therefore, we suggest the organizational climate as a major determinant of the public-self aspect.

4 External Agent s Stimuli Economic Exchange Theory Human Behavior Social Interactionism (Manis & Meltzer, 1978) Self-Theory (Carver & Scheier, 1985) Private Self-Aspect Public Self-Aspect Social Cognitive Theory (Bandura, 1986) Social Exchange Theory Self-Produced Factors Social Cognitive Theory (Self-efficacy) Economic Exchange Theory (Kelly & Thibaut, 1978) Social Exchange Theory (Blau, 1967) Social Cognitive Theory (Bandura, 1986; Stajkovic &Luthans, 1998) Figure 1. Theoretical Framework Organizational Climate Organizational Climate (Kelman, 1958; Dewhirst, 1971; Koys & DeCotiis, 1991; Moran & Volkwein, 1992)) Based on the theoretical framework in Figure 1, Fishbein and Ajzen s Theory of Reasoned Action (TRA; Fishbein and Ajzen 1975, Ajzen and Fishbein 1980) was adopted to show how theories affect the individual s knowledge sharing behavior. TRA not only meets the concept from the psychological perspective in organizational communications but also has the same assumption as that in symbolic interactionism, which assumes that human beings are usually quite rational and make systematic use of the information available to them. In fact, the concept of a private self in symbolic interactionism is equivalent to the concept of attitude in the TRA and the concept of a public self is equivalent to the concept of the subjective norm in the TRA. Consequently, the research model and hypotheses are derived by putting the above theoretical bases into the TRA together as in Figure 2. Expected Associations Expected Contributions Autonomy Affiliation Innovativeness Fairness H4(+) H5(+) H6(+) H7(+) H8(+) H9(+) Attitude toward Knowledge Sharing Knowledge Type Subjective Norm H10 H2(+) H3(+) Intention To Share Knowledge H1(+) Expected Rewards H11(+) Knowledge Sharing Behavior H12(+) The Level of IT Usage Figure 2. Research Model

5 Hypotheses First three hypotheses basically test Fishbein and Ajzen s theory in the knowledge sharing context. Theory of Reasoned Action TRA is designed to explain virtually any human behavior and ultimate goal of TRA is to predict and understand an individual s behavior (Fishbein and Ajzen, 1975). According to TRA, a person s performance of a specified behavior is determined by his or her behavioral intention (I) to perform the behavior (B = f (I)) (Ajzen and Fishbein, 1980; Davis, Bagozzi, and Warshaw, 1989). Therefore, knowledge sharing behavior which is defined as the degree to which one actually shares one s knowledge will be determined by the intention to share knowledge which is defined as the degree to which one believes that one will engage in a knowledge sharing act. H1: The intention to share knowledge will have a positive effect on the knowledge sharing behavior. And then, I is jointly determined by the person s attitude (A) and the subjective norm (SN) concerning the behavior in question with relative weights typically estimated by regression (I = Aw û + SNw ü) (Ajzen and Fishbein, 1980; Davis, Bagozzi, and Warshaw, 1989). Here, TRA not only considers both individual factors and social factors but also provides the linkage with symbolic interactionism because the attitude is equivalent to a private self in symbolic interactionism and the subjective norm is equivalent to a public self. The attitude toward knowledge sharing is defined as the degree of one s positive feelings about sharing one s knowledge and the subjective norm is the degree to which one believes that one s important others think one should perform the behavior in question. H2: The attitude toward knowledge sharing will have a positive effect on the intention to share knowledge. H3: The subjective norm will have a positive effect on the intention to share knowledge. Next two hypotheses deal with salient beliefs on the attitude toward knowledge sharing. According to the social exchange theory and the self-efficacy concept, variables called the expected associations and the expected contribution are proposed. Salient Beliefs on the Attitude: Social Exchange Theory & Self-efficacy Expected associations are defined as the degree to which one believe one can improve mutual relationship through one s knowledge sharing. Referring to the social exchange theory, we expect that employees who think knowledge sharing can increase the scope and the depth of associations among organizational members will show the positive attitude toward knowledge sharing because they consider this as an intrinsic benefit (Blau 1967). H4: The expected associations will have a positive effect on the attitude toward knowledge sharing. Expected contribution is defined as the degree to which one believes that one can improve the organization s performance through one s knowledge sharing. Based on the self-efficacy concept, we propose that an individual s judgment of his or her capabilities to contribute to the organizational performance as a determinant of the attitude toward knowledge sharing as a purely self-motivational source. In fact, this is the factor not from the literature review but

6 from the interviews of five organizations - Accenture, IBM, Samsung Economic Research Institute, Samsung Advanced Institute of Technology and Samsung Display Devices - from 1996 to Interviews were conducted as a preliminary study because Fishbein and Ajzen suggested a free response interview be conducted to enumerate the salient beliefs. People in these organizations pointed out the importance of the feedback. By monitoring and announcing the statistics of the knowledge usage, companies could make their employees realize the usefulness of knowledge sharing and increase their participation. Therefore, we expect that employees who believe in their ability to contribute to improving the organizational performance will have a positive attitude toward knowledge sharing. H5: The expected contribution will have a positive effect on the attitude toward knowledge sharing. Next four hypotheses are about relationships between the organizational climate variables and the subjective norm. Organizational Climate The organizational climate is defined as a relatively enduring quality of the internal environment of the organization that is experienced by its members, influences their behavior, and can be described in terms of the values of a particular set of characteristics or attributes of the environment (Tagiuri 1968). As we mentioned earlier, the subjective norm, the public self can be affected by the organizational climate (Dewhirst 1971) because people will voluntarily follow the instructions from the referent when the congruence between the individual and the organizational values is high (O Reilly and Chatman 1986). In addition, organizational climate affects the individual s perception through inter-subjective processes which is similar to the public-self aspect (Moran and Volkwein 1992). Therefore, we suggest the organizational climate as a major determinant of the subjective norm. Several dimensions of the organizational climate such as warmth, support, cohesiveness, openness and fairness have been emphasized in the cooperative behavioral research (Dewhirst 1971, Organ and Konovsky 1989). Other dimensions such as autonomy and innovativeness have been emphasized in the research about motivating knowledge workers (Kim and Lee 1995). Because sharing knowledge is not only a cooperative behavior but also related to stimulating knowledge sharing atmosphere among worker, four dimensions of organizational climate autonomy, affiliation, innovativeness and fairness, from both research areas are adopted. Autonomy is defined as the perception of self-determination with respect to work procedures, goals and priorities. Knowledge is a resource locked in human minds. Sharing knowledge is an intangible activity that can neither be supervised nor forced out of people. They happen only when people cooperative voluntarily (Kim and Mauborgne, 1997). To promote such an activity, autonomy in the organization is a necessary condition. People tend to follow their subjective norm easily under more autonomous organizational climate (Kim and Lee 1995). H6: Autonomy will have a positive effect on the subjective norm to share knowledge. Affiliation is related to cohesiveness among members in the organization and defined as the perception of togetherness. Needless to say, members in the organization tend to comply their important others thought when they feel close each other (Dewhirst 1971). Therefore, we suggest affiliation as one of antecedents of the subjective norm. H7: Affiliation will have a positive effect on the subjective norm to share knowledge.

7 Innovativeness is defined as the perception that change and creativity are encouraged. This dimension of organizational climate is related to how much risk people are willing to take in their knowledge sharing in terms of a fear of making themselves a fool. Thus, the more the organization encourages innovation, the more people in the organization dare to share their knowledge (Kim and Lee 1995) according to their important others thought. H8: Innovativeness will have a positive effect on the subjective norm to share knowledge. Fair process builds trust and commitment, trust and commitment produce voluntary cooperation, and voluntary cooperation drives performance, leading people to go beyond the call of duty by sharing their knowledge and applying their creativity (Kim and Mauborgne, 1997). Therefore, we defined fairness as the perception that organizational practices are equitable and non-arbitrary or capricious and suggested it as one of major organizational climate dimensions. H9: Fairness will have a positive effect on the subjective norm to share knowledge. Hypothesis 10 refers to the factor that affects the relative weights of the attitude and the subjective norm. Social meaning of information Fishbein and Ajzen argued that relative weights of attitudinal and normative components had great importance to predict and understand the behavioral intention. And they depended on several factors such as target object, target person and conditions under which the behavior was to be performed. Because many researchers have argued that characteristics of knowledge itself are important in knowledge sharing since Nonaka (1995) brought up tacit and explicit knowledge, we assumed that knowledge type might affect the relative weights. Knowledge type is classified into explicit and implicit. We differentiate implicit knowledge from tacit knowledge by defining that tacit knowledge is knowledge which cannot be expressed in verbal, symbolic and written form while implicit knowledge is defined as knowledge which is able to be expressed in verbal, symbolic, or written form, but not expressed yet based on Polanyi s concept (1966). According to Constant, et al. (1994) s research on information sharing, information as tangible product and information as intangible expertise have different meanings to people and different social implications. The same rationale can be applied to implicit and explicit knowledge sharing. Implicit knowledge sharing may be affected more by the person s attitudinal component because a person may have a tendency to take implicit knowledge as his or her own property while explicit knowledge may be influenced more by the subjective norm than the attitude because it is easy to be known to others and to be taken as the property of the organization. H10: Relative weights of the attitude and the subjective norm will have a significant difference by the knowledge type. Finally, last two hypotheses are about the factors directly affecting the knowledge sharing behavior. Facilitating conditions Fishbein and Ajzen (1980) argued that several external variables could have an effect when intention was realized to perform a behavior. Triandis also proposed a facilitating condition

8 concept which was very similar to Fishbein and Ajzen s, but modified and redefined it (Thompson et al. 1991) He acknowledged that even when intention was high, a behavior might not occur if certain conditions of a particular situation made the behavior impossible. For example, if someone intends to use a PC but does not have easy access to one, usage is less likely to occur and vice versa. We expect that the expected rewards and the level of the individual s IT usage can fall into these conditions for knowledge sharing. Expected rewards defined as the degree to which one believes that one can have extrinsic incentives due to one s knowledge sharing. Actually, many practitioners mentioned that rewards played an important role at the initiation stage of knowledge sharing. From the theoretical point of view, Kelman (1958) argued that rewards succeeded at securing only one thing: temporary compliance. Once the rewards run out, people reverted to their old behavior. These facts mean that rewards may be a facilitator for knowledge sharing, but not a fundamental force for forming a person s attitude. H11: The expected rewards will have a positive effect on the individual s knowledge sharing. The level of the individual s IT usage is defined as the degree of one s frequency of using IT such as BBS and and it falls in the same category. Information technology has been considered as an important enabler in knowledge management (Davenport 1997). An enabler means that IT also plays a facilitating, not an attitude forming, role. Therefore, we expect that the individual s level of IT usage affects the link between knowledge sharing intention and knowledge sharing behavior. H12: The individual s level of information technology usage will have a positive effect the individual s knowledge sharing behavior. Research Methodology In order to test the proposed hypotheses, we developed measurements for each variable. Some of them such as knowledge sharing behavior, expected associations, expected contribution and expected rewards were newly developed based on the relevant theories, and others such as the intention to share knowledge, attitude toward knowledge sharing, the subjective norm, four variables in organizational climate and the individual s level of IT usage were adopted or adapted from the prior studies. Before conducting the main survey, we performed a pretest. We tested the internal consistency and the discriminant validity of the measurement instruments with 61 responses from 13 organizations in 7 industries. The Cronbach s α value ranged from.71 (for expected rewards) to.95 (for expected contribution). The number of the final items used in the main survey was reduced from 60 to 56. For the questionnaire, a multiple-item method was used and each item was measured by a 5 point Likert scale, ranging from most frequent to most rare. For the main survey, 27 organizations over 16 industries out of over 300 companies whose executives enrolled in the Chief Knowledge Officer Education Program offered by the author s university participated. 259 questionnaires out of 280 were gathered and 154 responses were usable. Industry Company Respondent Percent Industry Company Respondent Percent Food Construction Chemical Wholesale Pharmaceutical Retail

9 Metal Transportatio n & Telecom. Electric Banking Machinery Electronics& Insurance Communication Automotive Brokerage Electricity & Entertainment Gas & Other Svc. Total Measure Items Freq. % Measure Items Freq. % Gender Male Gender Female Age 21~29 30~34 35~39 40~ Missing Work Experience (year) 0~3 3~6 6~9 9~12 12~ Missing Position Employee Chief employee Manager Director Other Education Table 1. Profile of Companies and respondents High school College (2 years) University (4 yrs) Graduate school We estimated the agreement score for the four organizational climate variables with the index or within-group agreement r wg derived by James et al. (1984). Within-group agreement among individuals could qualify theoretically to aggregate the perceptions to the higher unit level, and to use mean to represent this collective interpretations. The responses with less than.7 and with less than 3 respondents per department were removed. Then internal consistency, convergent validity, and discriminant validity were checked by cronbach s α, item-to-total correlation and factor analysis (Straub 1989, Baggozzi et al. 1991) Construct Item Mean S.D. C. s α Item-to-total correlation Items were dropped due to KS Behavior ,.5036,.4642,.4535,.5491,.4459,.5155 Factor Analysis Intention to Share K Attitude toward KS Normative Beliefs Motivation Comply Expected Associations to ,.4663,.3922,.5709,.4849,.4971 Low Item-total ,.4801,.5501,.4671, ,.7323,.7983,.7630, ,.6826,.7130,.8405, ,.6422,.5915,.6177 Correspondence to Motivation to Comply* ,.6444, ,.3292,.4735,.0269 Low Item-total ,.4681, ,.7668,.7825,.7445,.6997

10 Expected ,.6593,.7387,.7850,.7529 Contribution Autonomy ,.5862,.6029,.6224 Affiliation ,.6450,.7145,.7429 Innovativeness ,.6201,.5778,.7194 Fairness ,.6034,.5998,.3898 Low Item-total Expected Rewards The Level of IT Usage ,.5927, ,.7675,.5455, ,.5480,.5639,.3966,.4552 Low Item-total ,.5437,.4959,.4546 *Since the Subjective Norm (SN) was calculated by the Normative Beliefs (NB) multiplied by the Motivation to Comply (MC), one item of NB should be dropped once one item of MC was dropped due to the low item-tototal correlation. Attitude1 Attitude2 Attitude3 Attitude4 Attitude5 EA1 EA2 EA3 EA4 EA5 EC1 EC2 EC3 EC4 EC5 Intention1 Intention2 Intention3 Intention4 Intention5 Affiliation 1 Affiliation 2 Affiliation 3 Affiliation 4 ER1 ER2 ER3 ER Inno Inno Inno Inno Auto1 Auto2 Auto3 Auto4 Behavior1 Behavior2 Behavior3 Behavior4 Behavior5 IT1 IT2 IT3 IT4 Fair1 Fair2 Fair3 NB1 NB2 NB3 MC1 MC2 MC

11 EA = Expected Associations; EC = Expected Contributions; ER = Expected Rewards; Inno = Innovativeness; Auto = Autonomy; IT = The Individual s Level of IT usage; Fair = Fairness Table 2. Results of internal reliability test and factor analysis Results of Hypothesis Testing Hypothesized relationships depicted in Figure 2 were tested by multiple regression analysis and the summary of results was presented in Table 3. Hypothesis R² Adj. R² Std. β t-value Tolerance VIF B=I+ER+IT+e I *.941 ER *.959 IT **.964 I=A+SN+e A ***.803 SN ***.803 A=EA+EC+e EA ***.717 EC RGP=Auto+Affi Inno+Fair+e Auto Affi Inno * Fair * *** p<0.001, ** p<0.01, *p<0.05, <0.1 Table 3. Hypothesis test results Result H1: Supported H11: Supported H12: Supported H2: Supported H3: Supported H4: Supported H5: Not supported H6: Not supported H7: Not supported H8: Supported H9: Supported Hypotheses 1 to 3 examined the Fishbein and Ajzen s theory in the knowledge-sharing context. All 3 hypotheses were supported. Hypotheses 4 and 5 examined the employee s beliefs, the expected associations and the expected contribution, to the attitude. The expected associations were positively related to the attitude as expected but the expected contribution was not significant. Next four hypotheses investigated the relationship between the organizational climate and the subjective norm. Two organizational climate variables, innovativeness and fairness had influences on the subjective norm. However, innovativeness was unable to pass the test of mediator variable as shown in Table 4. According to Baron and Kenny (1986), the following three conditions must hold to support the mediating effect of a construct. First, the independent variable must have a significant association with the dependent variables. Next, the independent variable must have a significant association with the mediator. Finally, the mediator must have a significant effect on the dependent variable when both the independent variable and the mediator variable are included as predictors. In Table 4, only fairness mediates the effect of the subjective norm.

12 Variable IV? DV IV? mediator IV? DV (mediator included) Std. β t-value Std. β t-value Std. β t-value IV: Autonomy Affiliation Innovative Fairness Mediator: SN DV: I ** ** R²=.246 R²=.265 *** p<0.001, ** p<0.01, *p<0.05, <0.1 Table 4. Test of mediating effect * 2.561* R²= * ** 4.945*** Hypothesis 10 examined whether there was a significant difference in relative weights when people shared explicit knowledge and implicit knowledge. If the attitude had greater impact on implicit knowledge sharing than on explicit knowledge sharing, there should be a significant difference in betas. However, we could not find any significant differences just as shown in the Table 5 when we assume that the intention to share explicit knowledge = β 11 Attitude + β12 RGP + e and the intention to share implicit knowledge = β 21 Attitude + β 22 RGP + e. Therefore, the hypothesis 10 was not supported. Variable Explicit K Implicit K t-value Results Attitude.387***.373*** t=.1142, α=0.01 : ½ H β β SN.256**.257** t=.0103, α=0.01 H : 12 0 ½ β β 22 *** p<0.001, ** p<0.01, *p<0.05, <0.1 Table 5. Results about relative weights Finally, hypotheses about the expected rewards and the individual s level of IT usage as the facilitating conditions were supported. Discussion and Conclusions First of all, the relationships among attitude, subjective norm, intention and behavior were supported. Besides, we found that attitude had a greater effect on knowledge sharing intention than the subjective norm. This proves that knowledge is a resource locked in human minds and knowledge sharing should be fostered rather than forced or mandated. Secondly, the role of social exchange relationship in helping behaviors was confirmed. As many researchers have suggested, social exchange tends to engender feelings of personal obligation, gratitude, and trust. And the frequent rendering of knowledge sharing like the organizational citizenship behavior seems to mainly foster a sense of social exchange relationship (Organ and Konovsky 1989). Next, despite the strong evidence from the preliminary study, the hypothesis on expected contribution was not supported. We think that this is because expected contribution may not have a significant effect on people who have many years of work experience. They already

13 know their ability and what s going to happen after they share their knowledge. That is, they have already formed their attitude toward knowledge sharing (Constant et al., 1994). Fourth, the fair process turns out to be a powerful management tool for knowledge sharing. From this finding, we can confirm Kim & Mauborgne (1997) s idea that fair process builds trust and unlocks ideas. Fifth, knowledge type did not affect the knowledge sharing behavior. Employees did not differentiate explicit knowledge from implicit knowledge. The reason that knowledge type does not affect the relative weights of attitudinal and normative components may be found in the learning process in epistemology. Knowledge is formed by two different processes, direct experience and making use of previously learned relationships or various rules of logic. This means that knowledge should be learned regardless of its type. Therefore, people do not seem to differentiate on knowledge types. Finally, the roles of the expected rewards and the level of the individual s IT usage as the facilitating conditions were confirmed. Expected rewards, believed by many as the most important motivating factor for knowledge sharing, were not significantly related to the attitude toward knowledge sharing (Bock and Kim 2002) because of several reasons (Kohn, 1993). First of all, rewards have a punitive effect because they are manipulative like outright punishment. Further, not receiving a reward that one had expected to receive is indistinguishable from being punished. Secondly, rewards break off relations. For each person who wins, there are many others who feel they have lost. When employees compete for a limited number of incentives, they will very likely begin to see each other as competitors to their own success. Next, managers often use incentive systems as a substitute for giving workers what they need to do a good job - providing useful feedback, social support, and the room for self-determination. Finally, rewards, like punishment, may actually undermine intrinsic motivation. The more they experience being controlled, the more they tend to lose interest in what they are doing. Furthermore, the recipient of the reward assumes, If they have to bribe me to do it, it must be something I wouldn t want to do. So, the larger the incentive they are offered, the more negatively they view the activity for which the bonus was received. The next explanation is related to the organizational citizenship behavior (OCB) literature. OCB can be defined as willingness of persons to contribute efforts to the cooperative system by Barnard (1938, p. 83). Almost 30 years after Barnard, Kats and Kahn (1966; 1978) suggested that reward systems might inhibit cooperation (Organ & Konovsky, 1989), because critical voluntary behaviors that are not specified by job descriptions are largely a function of identification and internalization rather than instrumental involvement (O Reilly & Chatman, 1986). According to Constant et al. (1994), experienced workers learned that they should share their knowledge which was acquired from their work and training. Therefore, they may have a negative attitude toward receiving extrinsic benefits in return for knowledge sharing behavior which they perceive as normal business activity. The bottom line is incentives merely and temporarily change what we do rather than change the individual s fundamental attitude. The role of the individual s level of IT usage falls into the same category. Contributions and Limitations First, this research empirically tested Fishbein and Ajzen s TRA model in the knowledge sharing context for the first time. Especially, this research model combined with Triandis

14 facilitating condition turned out to be very useful. By applying TRA, we also figured out the way of factors affecting the knowledge sharing behavior. We also considered both individual factors and social factors at the same time by applying TRA. Although knowledge sharing behavior has been considered the individual behavior because knowledge is a resource locked in human minds, we proved social factors such as fairness and subjective norm play important role in sharing knowledge. Next, this study gives a concise picture of knowledge sharing behavior by reducing many factors affecting such behavior into several factors through solid theoretical foundations. In terms of the practical perspective, we should pay more attention to enhancing the positive mood state for the social associations and the fair process which precede the knowledge sharing behaviors. Secondly, rewards along with the level of IT usage play an important role as the facilitating conditions. However, we need to use these management tools carefully because these do not seem to alter the attitude that underlies the knowledge sharing behavior (Bock and Kim 2002). Finally, employees perception on expected contribution seemed to vary by the years of work experience. Therefore, companies should use discretion in using this motivational factor to their employees. Even though this research has drawn intellectually and practically meaningful implications, there are a few limitations. First of all, relative weights of attitudinal and normative components were not affected by the knowledge type, perhaps due to the lack of tacit knowledge type in the model. In the future research, it will be worthwhile to compare the tacit knowledge with the explicit and implicit knowledge. Next, it may be necessary to measure the construct of IT usage with more diverse types of IS for knowledge sharing at the organizational level because sharing explicit knowledge is done mostly through intranets and formal knowledge repositories in many organizations. Finally, there might still be many other factors to realize the intention to the actual behavior because for knowledge sharing behavior (0.153) was relatively low. References Ajzen, I., Fishbein, M. (1980): Understanding Attitudes and Predicting Social Behavior, Prentice-Hall Bagozzi, R.P., Yi, Y., Phillips, L.W. (1991): Assessing Construct Validity in Organizational Research, Administrative Science Quarterly, Vol. 36, pp Bandura, A. (1986): Social Foundations of Thought and Action, Prentice Hall Barnard, C.I. (1938): The Functions of the Executive, Harvard University Press, Cambridge, MA Baron, R.M., Kenny, D.A. (1986): The Moderator-mediator Variable Distinction in Social Psychological Research: Conceptual, Strategic, and Statistical Considerations, Journal of Personality and Social Psychology, Vol. 51, No. 6, pp Blau, P. (1967): Exchange and Power in Social Life, Wiley Bock, G.W., Kim, Y.G. (2002): Breaking the Myths of Rewards: An Exploratory Study of Attitudes about Knowledge Sharing, Information Resources Management Journal, Vol. 15, No. 2, pp Carver, C.S., Scheier, M.F. (1985): Aspects of Self, and the Control of Behavior. Self and Social Life (edited by Schlenker, B.R.), McGraw-Hill Constant, D., Kiesler, S., Sproull, L. (1994): What s Mine Is Ours, or Is It? A Study of Attitudes about Information Sharing, Information Systems Research, Vol. 5, No. 4,

15 pp Davenport, T.H. (1997): Some Principles of Knowledge Management, Working Paper. Davis, F.D., Bagozzi, R.P. and Warshaw, P.R. (1989): User Acceptance of Computer Technology: A Comparison of Two Theoretical Models, Management Science, Vol.35, pp Dewhirst, H.D. (1971): Influenced of Perceived Information Sharing Norms on Communication Channel Utilization, Academy of Management Journal, Vol. 14, No. 3, pp Engelbart, D.C. (1963): A Conceptual Framework for the Augmentation of Man s Intellect, Vistas in Information Handling (edited by P. Howerton), Spartan Books, Washington, D.C.; (1982): Towards High-Performance Knowledge Workers, OAC 82, AFIPS Press Fishbein, M. & Ajzen, I. (1975): Beliefs, Attitude, Intention and Behavior: An Introduction to Theory and Research, Addison-Wesley Publishing Company Kats, D. and Kahn, R.L. (1966): The Social Psychological of Organizations, Wiley, NY Kelman, M.C. (1958): Compliance, Identification, and Internalization: Three Processes of Attitude Change, Journal of Conflict Resolution, Vol. 2, pp Kim, Y. B., Lee, B. H. (1995): R&D Project Team Climate and Team Performance in Korea, R&D Management, Vol. 25, No. 2, pp Kim, W. C., Mauborgne, R. (1997): Fair Process: Managing in the Knowledge Economy, Harvard Business Review, July-August, pp Krone, K.J., Jablin, F.M. and Putnam, L.L. (1987): Communication Theory and Organizational Communication: Multiple Perspective, Handbook of Organizational Communication: An Interdisciplinary Perspective (edited by F. Jablin, L. Putnam, K. Roberts, and L. Porter), SAGE Publications, Newbury Park, California, Kohn, A. (1993): Why Incentive Plans Cannot Work, Harvard Business Review, Sep.-Oct., Manis, J.G. and Meltzer, B.N. (1978): Symbolic Interaction: A Reader in Social Psychology (3rd edition), Allyn and Bacon, Inc. O Reilly, C.A., Chatman, J.A. (1986): Organizational Commitment and Psychological Attachment: The Effects of Compliance, Identification, and Internalization on Prosocial Behavior, Journal of Applied Psychology, Vol. 71, No. 3, pp O Reilly, C.A., Chatman, J.A., and Anderson, J.C. (1987): Message Flow and Decision Making, Handbook of Organizational Communication: An Interdisciplinary Perspective (edited by F. Jablin, L. Putnam, K. Roberts, and L. Porter), SAGE Publications, Newbury Park, pp Organ, D.W., Konovsky, M. (1989): Cognitive versus Affective Determinants of Organizational Citizenship Behavior, Journal of Applied Psychology, Vol. 74, No. 1, pp Polanyi, M. (1966): The Tacit Dimension, Bantam Doubleday Dell Publishing Group, Inc. Ruggles, R. (1998): The State of Notion: Knowledge Management in Practice, California Management Review, Vol. 40, No. 3, pp Stajkovic, A.D., Luthans, F. (1998): Social Cognitive Theory and Self-efficacy: Going Beyond Traditional Motivational and Behavioral Approaches, Organizational Dynamics, Spring, pp Straub, D. W. (1989): Validating Instruments in MIS Research, MIS Quarterly, Vol. 13, No. 2, pp Tagiuri, R., Litwin, G. (1968): Organizational Climate: Explorations of a concept. Harvard Business School Thompson, R.L., Higgins, C.A., Howell, J.M. (1991): Personal Computing: Toward a

16 Conceptual Model of Utilization, MIS Quarterly, March, pp