Revisiting Results in the Distributed ISD Team
To summarize, results from this study indicate that, for an individual to be perceived as an effective knowledge transferor by remotely-located recipients in a cross-cultural ICT-mediated distributed team engaged in an ISD project, he/she should extensively participate in electronic conversations, as indicated by the extent of communication, and be perceived as credible due to trustworthy behaviors and high performance. These results are consistent with a similar prior study involving U.S.-Norwegian distributed teams engaged in information systems development.
Contrary to expectations, differences in capability did not seem to enhance the extent of knowledge transfer in either the distributed or the local groups. While this may appear counterintuitive, the consistency of this finding across different types of distributed teams involving different cultures (e.g., the U.S. and Norway) suggests that the more knowledgeable and skilled members may be conveying their knowledge in forms that are incomprehensible to a less experienced or less skilled team member, thereby reducing the absorption of the knowledge and learning by the recipient, and reducing the overall extent of knowledge transferred. Even though this study used student subjects (who are often similar in dimensions such as capability), the differences in capability among the team members ranged from -4.70 to +4.46 (see Table 6). This suggests that there were members who had much higher ISD capability than their other members. It could have been possible that these members were not able to convey their knowledge successfully to their less knowledgeable team members.
Hinds et al. (2001) make a similar conclusion from their study of expert and novice behaviors in an electronic circuit wiring task. They argued that while “experts should have been well-positioned to convey their superior knowledge skills to novices, the organization of that knowledge, and particularly its level of abstraction, may make it difficult for them to do so” (Hinds et al., 2001, p. 1232). Another possibility is that knowledge of highly skilled individuals or experts is embodied in their actions, and it may be very difficult for the intended recipient to separate the knowledge from their actions (Davenport & Prusak, 1998). Swap, Leonard, Shields, and Abrams (2004, p. 182) refer to the concept of “pattern recognition” and argue that “the way in which experts exercise their knowledge is by calling on … experiences in a great variety of contexts to recognize patterns.” This pattern recognition process tends to draw upon the “tacit dimensions” of the expert's knowledge that is extremely contextualized and therefore very hard to transfer (Swap et al., 2004, p. 183).
As expected, credibility and extent of communication both emerged as critical determinants of the extent of knowledge transferred. Individuals who were perceived to be trustworthy and high performers (i.e., had high credibility) and engaged in a high amount of communication with their remote team members were viewed to be significant knowledge transferors by their remote members.
As in the case of U.S.-Norwegian teams, culture played a significant role on the extent of knowledge transferred in this study of U.S.-Thai teams; however, contrary to expectations, in this study, members of individualistic cultures were viewed as transferring more knowledge than those in collectivist cultures. This clearly runs counter to prevailing thought on knowledge-sharing behavior. In order to make sense of this anomaly, some of the other qualitative data that were gathered during this semester-long project were used. These included chat transcripts and reflection documents created by each of the subgroups within the distributed teams at the end of the project, in which they discussed their experiences in the distributed ISD project, barriers they faced, how they attempted to overcome those barriers, and the lessons they learned through the process. A point that was repeatedly made by many of the U.S. subgroups was the limited and somewhat ineffective communication received from the Thai team members. This suggests that even if Thai members attempted to share knowledge through communication, they were not effective in getting through to the U.S. members, and thus, were not able to transfer a significant amount of knowledge.
One of the reasons for the lack of effective communication may have been the language barrier that existed among the subgroups. English was adopted as the official language for the project, and the language in which all of the project deliverables were going to be created. However, given that this was not the primary language of the Thai participants (at least at this university), this created a divide between the two remote subgroups within the distributed teams. The lack of language skills led to difficulty during chat sessions and instant messaging. The problem related to language and communication became such a constraint on the participants that some U.S. subgroup members sought the help of Thai students residing in the U.S. to act as interpreters during their electronic conversations. Similarly, Thai participants also sought the help of other individuals in Thailand who were fluent in English (who were not participants of the project, or who were members of a different group) to act as mediators between the U.S. and Thai members of their teams.
This problem related to communication and language helps make sense of the unexpected effect of culture (individualism/collectivism) in this study. It may be argued that Thai participants' lack of knowledge transfer was not related to their cultural values. Given that they come from significantly more collectivist cultures, sharing of knowledge was clearly valued by them. However, inhibitions related to their language skills perhaps made them shy away from engaging in extensive communication about new and difficult concepts with their remote participants. This argument is consistent with prior studies that have suggested that individuals who lack proficiency in a language often feel anxious about participating during collaborations (Kim & Bonk, 2002). In another study involving Danish and American college students, Bannon (1995) concluded that lack of fluency in English prompted the Danish students to stay away from computer-mediated communication. Davenport and Prusak (1998, p. 98) have also argued that “people cannot share knowledge if they do not speak a common language.” In the case of the Thai students, it could have been possible that whatever they may have shared, due to their lack of fluency in the language, it was incomprehensible to their remote participants, leading to their (i.e., the remote team members') perception of little knowledge being transferred and lack of learning.
This provides some valuable lessons for fostering knowledge transfer in cross-cultural distributed teamwork. As the case of the Thai participants in the study highlights, intention to transfer knowledge is not sufficient to achieve success in knowledge transfer within a team. This problem multiplies when the team is composed of members from across the globe and communicating only through the electronic media, where text is the primary mode of sharing knowledge. Even if cultural values within the members foster such sharing, other factors, such as language skills (written and spoken), can act as a significant barrier to such transfer. Managers coordinating such teamwork need to be aware of this and take necessary steps to ensure that knowledge is successfully shared (e.g., selecting members who are fluent in the official language of the project, providing interpreters, etc.).