International Construction Specialty Conference of the Canadian Society for Civil Engineering (ICSC) (5th : 2015)

Do strong or weak ties matter in knowledge networks? Poleacovschi, Cristina; Javernick-Will, Amy N. Jun 30, 2015

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
52660-Poleacovschi_C_et_al_ICSC15_229_Ties_In_Knowledge_Networks.pdf [ 313.58kB ]
52660-Poleacovschi_C_et_al_ICSC15_229_Ties_In_Knowledge_Networks_slides.pdf [ 605.34kB ]
Metadata
JSON: 52660-1.0076438.json
JSON-LD: 52660-1.0076438-ld.json
RDF/XML (Pretty): 52660-1.0076438-rdf.xml
RDF/JSON: 52660-1.0076438-rdf.json
Turtle: 52660-1.0076438-turtle.txt
N-Triples: 52660-1.0076438-rdf-ntriples.txt
Original Record: 52660-1.0076438-source.json
Full Text
52660-1.0076438-fulltext.txt
Citation
52660-1.0076438.ris

Full Text

5th International/11th Construction Specialty Conference 5e International/11e Conférence spécialisée sur la construction    Vancouver, British Columbia June 8 to June 10, 2015 / 8 juin au 10 juin 2015    DO STRONG OR WEAK TIES MATTER IN KNOWLEDGE NETWORKS? Cristina Poleacovschi1, Amy N. Javernick-Will1 1 Department of Civil, Architectural, and Environmental Engineering, University of Colorado Boulder, USA  Abstract: Construction and engineering organizations have increasingly implemented knowledge exchange strategies with the goal to facilitate knowledge exchange across the organization. However, despite these efforts, many knowledge strategies fail in practice, as it is not well known when knowledge access is most beneficial. This research analyzes the correlation between group level knowledge exchange and perceived individual benefits. Specifically, we focus on the time saved (in hours per month) on work tasks as a result of accessing knowledge with others in the department. To conduct this research, we used social network analysis and a modularity optimization algorithm to identify the existing knowledge-based subgroups (KBS)—subgroups that share more knowledge internally then externally—within a large engineering and construction organization. To identify whether these knowledge based subgroups offer time benefits, we compared the time benefits from receiving knowledge within these subgroups and outside these subgroups. Results found that individuals are more likely to perceive saving time on work tasks as a result of receiving knowledge within their subgroups. As a result, our results indicate that the type of benefit received matters to determine whether weak ties or strong ties are important.   1 INTRODUCTION The importance of knowledge sharing in construction and engineering organizations has gained an unprecedented interest and many scholars view knowledge as the most important resource of the firm (Grant 1996). The efficient allocation of money and labor is not sufficient enough to gain competitive advantage anymore. Instead, knowledge and its use at the right place and time is what allows projects to achieve successful results (Argote and Ingram 2000).  While this view is shared largely in the literature, there are many debates about how and when knowledge exchange brings the most benefits. To address this debate, previous studies have focused on knowledge exchange primarily at the group level, with emerging studies at the individual level. At the group level, the literature has focused on the importance of knowledge exchange on group performance (Cummings 2004; Landaeta 2008; Tsai 2001). For instance, Tsai (2001) showed that business units which are more central are more likely to receive diverse knowledge from other business units which in turn affects innovation and performance. While studying the effects and benefits at the group level is important, it provides an incomplete understanding of how knowledge exchange adds value to outcomes.  Specifically, aggregating group level outcomes offers a good understanding of the effect of group level attributes (e.g. size, composition) on final outcomes, but it neglects intra-group dynamics, which represent the core of group outcomes. To address the dearth of literature on individual level outcomes, Wanberg and Javernick-Will (2014) studied the relationship between frequency of individual knowledge exchange and 229-1 individual work outcomes in informal organizational groups. They found that infrequent interactions provided individuals with more unique knowledge then frequent knowledge interactions.  As such, individual level knowledge exchange has been analyzed for individual level performance and group level knowledge exchange has been analyzed for group level performance. As a result, the connection between the group level characteristics and individual level benefits has been largely neglected with the exception of Poleacovschi and Javernick-Will (under review) who found that individuals spanning knowledge across highly connected knowledge subgroups are more likely to receive higher individual performance evaluations in a construction and engineering organization.  This research proposes to further address the missed macro-micro link by analyzing the relationship between group knowledge exchange and individual level outcomes. Previous research has identified that individuals gain benefits from their position in the network structure. Nahapiet and Ghoshal (1998) found that individuals share intellectual advantages, such as knowledge exchange, as a result of being part of highly connected groups. Thus, we expect that individuals who are part of highly cohesive knowledge exchange groups will gain benefits from the network connectivity of their group. However, we expect that this relationship is sensitive to the type of relationship, or tie, between the individual and the group. This thinking is in line with the theory of weak ties (Granovetter 1973) which argues that the type of ties matter in a network. To translate Granovetter’s study at the level of subgroups, strong ties are part of highly connected groups while weak ties connect individuals from different subgroups. Considering the importance of project time in construction and engineering organizations, we define individual benefits as the time an individual saved (in hours per month) on work tasks as a result of receiving knowledge from other employees, whether within their knowledge based subgroup or outside of their knowledge based subgroup. Thus, this research asks: Do individuals save more time as a result of their weak or strong ties within knowledge exchange networks? To answer this question we employed a unique method that identified knowledge-based subgroups (KBS) within a department of a large construction and engineering organization (Poleacovschi and Javernick-Will under review). The method used a social network optimization algorithm that identified subgroups that shared more knowledge internally then externally. Then, using survey data, we identified the perceived time savings as a result of receiving knowledge within their KBS as compared to receiving knowledge outside of their KBS.  This research is important both for theory and practice. To contribute to theory, this research links group and individual level constructs by testing the strength of weak ties theory in KBS. By translating weak and strong ties concepts to KBS, knowledge connections in KBS are strong ties because these subgroups are dense and cohesive, while connections outside KBS are weak ties. By focusing on the specific benefit of time savings, we contribute to theory by extending previous work on different types of benefits that can be received from strong or weak ties. In practice, construction and engineering organizations can chose to invest resources based upon the benefits they seek—in our research, this relates to saving time on work tasks.  2 THEORETICAL BACKGROUND 2.1 Knowledge-based subgroups  The interest in knowledge-based subgroups, or groups that share more knowledge internally then externally, is rooted in the knowledge-based theory of the firm (Grant 1996). In this view, the ability to efficiently exchange the “know what” and especially the “know how” is what differentiates organizations from markets (Kogut and Zander 1992). In other words, construction and engineering organizations exist because they successfully manage to integrate the knowledge of professionals from different disciplines (e.g. structural engineers, project managers) that can easily exchange knowledge, or “know how”, to achieve project goals. Another option is to look for professional knowledge (e.g. structural analysis and project management) on the market. However, this would be less efficient as separate transactions with different professionals would be costly. The knowledge-based theory of the firm also believes that an organization gains competitive advantage whenever it manages to coordinate knowledge exchange better 229-2 than other organizations (Argote and Ingram 2000; Grant 1996). Unfortunately, this is not easy as there are many organizational structures and boundaries that impede knowledge exchange (Wanberg et al. 2014; John Wanberg et al. 2014)  As a result of these boundaries, knowledge-based subgroups—subgroups that share more knowledge internally and less externally—can form. The common belief is that subgroups disrupt knowledge exchange, which may be detrimental for the larger performance of the group or organization.  For instance, individuals are more densely connected to others within their subgroup and limited knowledge may flow between subgroups, resulting in a boundary that may impede members of the group to search and access knowledge outside their own subgroup. However, this assumes that knowledge exchange with others outside one’s own subgroup is actually valuable – which may not be true. More importantly, valuable in terms of what and valuable for whom? As previous research has mainly focused on the effects of subgroup formation on group level outcomes, we do not know the effect of subgroup formation on individual benefits in terms of time saved on work tasks. Using theory of weak ties we build our hypothesis about the relationship between network ties with KBS and individual time benefits.  2.2 Ties within and between knowledge-based subgroups The most influential papers addressing individual benefits from network structure are the structural hole (Burt 1995) and the strength of weak ties (Granovetter 1973) theories. A structural hole means the absence of a relationship between two individuals in a network. The theory states that individuals who link two unconnected others are likely to gain benefits. The strength of weak ties theory was the antecedent of structural hole theory and focused on the types of ties and benefits results those ties (Granovetter 1973). Individuals who manage to create ties outside their own subgroup of close relationships (weak ties) gain benefits as a result of reaching for novel resources and information. For instance, in his thesis, Granovetter discovered that weak ties are the most beneficial connections for a job search. Figure 2 shows a visual representation of weak and strong ties. Straight lines represent strong ties while dotted lines represent weak ties.   Figure 1: The Strength of Weak Ties ((Granovetter 1973) While this theory is quite influential, it was initially conceptualized and tested in the context of people who were searching for jobs. The theory has been tested less in the context of knowledge networks or considering individual time benefits. KBS are subgroups whose members share more knowledge internally and less knowledge externally. Thus, ties in KBS are strong because of the high connectivity between members, while ties outside of, or between, KBS are weak as they connect distant subgroups, which are highly connected. In contrast to Granovetter’s study, we expect that highly connected subgroups will save individuals more time as a result of the short distance between subgroup members. Conversely, accessing a new subgroup of knowledge through weak ties requires many more steps. As such, individuals will save more time from accessing knowledge in their highly connected network then from accessing knowledge in other subgroups. For instance, member B will access member G’s knowledge easier then member D’s knowledge, although it takes the same number of steps (Figure 1). In other words, members of KBS will use their subgroup connectivity to access knowledge fast. Thus, we expect that the type of tie between the individual and the highly connected network further benefits the individual. If the tie is strong (within a KBS), we expect that the individual will gain more time benefits from accessing that knowledge.  229-3 Hypothesis: Individuals will save more time from accessing knowledge from strong ties (within subgroups) then from weak ties (between subgroups) 3 RESEARCH METHODS: DATA COLLECTION  We were provided access to a large set of network data within the department of a large construction and engineering consultancy. The organization has focused on improving their knowledge management strategies for the past decade.  They initially implemented knowledge management in the IT department, which was where the data was collected. Survey questionnaires were administered to all employees within this department, with questions focused on with whom they share and receive knowledge and whether these connections help to save them time through the knowledge provided. The response rate for the knowledge network data was 88% (n=142).  Non-respondents were excluded from the final analysis.  3.1 Knowledge sharing networks Questionnaires were administered to all 161 employees in the department. Using definitions of knowledge as information and :know how (Liebeskind 1996), employees were asked to indicate, on a 6-point scale, the level of knowledge that they receive from other departmental employees. Specifically, they were asked:   “Often we rely on the people we work with to provide us with information to get our work done. For example, people might provide us with simple routine administrative or technical information that we need to do our work. Alternatively, people might provide us with complex information or engage in problem solving with us to help us solve novel problem. Please indicate the extent to which people listed below provide you with information you use to accomplish your work?”  Response options included: 0 – I do not know this person/I have never met this person; 1 – Very Infrequently; 2 – Infrequently; 3 – Somewhat infrequently; 4 – Somewhat frequently; 5 – Frequently; 6 – Very frequently.  The data was transferred into a matrix format where questionnaire respondents were listed as rows, and the level of knowledge sharing with each other departmental employee was indicated in columns.  The data was then analyzed using social network analysis (SNA) software to determine the composition of knowledge-based subgroups.  3.2 Time Benefits To identify time benefits accrued as a result of receiving knowledge within and outside KBS, individuals were asked to indicate the time they saved per month (in hours) as a result of receiving knowledge from each individual in the department.  They responded to the prompt:  “Please provide an estimate for the typical time saved per month as a result of information,    or other resources received from each person.” The respondents assessed the time benefits for each other departmental employee based on a 5 – point scale, including: 0 – I do not know this person/I have never met this person; 1 – No time saved; 2 – 1-3 Hours per month; 3 – 4-8 Hours per month; 4 – 9-12 hours per month; to 5 – more than 13 hours per month. 3.3 Individual attributes: hierarchy, tenure and gender To identify individual attributes, individuals were asked to identify their hierarchy level and tenure (in years) within the organization. Hierarchical levels were determined from a scale of one to five (1 – individual contributor/team member; 2 – supervisor/team leader; 3 – project manager/program manager; 4 – manager/BU [business unit] manager; 5 – director). For tenure, individuals mentioned the number of years they have spent with the organization. Finally, the department provided the data on gender.  This 229-4 information was later used for the linear regression analysis as control variables.  We controlled for hierarchy, tenure and gender as women and newcomers are not expected to gain the same benefits from their position in the network (Burt 1992), while individuals in higher hierarchy levels are expected to be central in networks and gain more benefits (Poleacovschi and Javernick-Will under review).  4 RESEARCH METHODS: DATA ANALYSIS In this section we present methods used to analyze the collected data with the goal to identify KBS and the average time saved in versus outside of KBS. 4.1 Identifying Knowledge-based Subgroups  We analyzed structural aspects of the department by analyzing the knowledge sharing connections between individuals within the network.  We used SNA to identify KBS, which is described in (Poleacovschi and Javernick-Will under review), using Gephi (Bastian et al. 2009), a visualization and network analysis software. Gephi includes a tool that uses the Louivan modularity optimization algorithm (Blondel et al. 2008) to determine groups within the network that share more knowledge internally and less knowledge externally. The algorithm is based on calculating the modularity gain, or, in our case, the change in the level of knowledge exchange within a group as a result of adding a new member to it, compared to the level of knowledge exchange outside the group once the member joins.  This determined the number of KBS—which, for this network was four, and the composition of, or employees that belonged to, each of the four KBS.  We validated the KBS by running a ‘blockmodel’ in Netminer (Netminer 2014). The blockmodel allowed us to calculate and compare the network density, or the ratio of existing connections to all possible connections, between different KBS.  This helped to validate the algorithm, as higher network densities were observed within subgroups compared to network densities between subgroups.  4.2 Identifying Time Benefits in Knowledge-based Subgroups After the KBS were determined, we analyzed whether individuals within the subgroups received greater time benefits (saving more hours per month) from the knowledge received from employees within their KBS or external to their KBS.  To do this, we conducted several steps, outlined below. 1. We constructed a matrix of 142 rows by 142 columns (number of people who took the questionnaire).  The rows represented questionnaire respondents, and the columns included the average amount of time they saved per month (on a scale from 0 to 5) based upon receiving knowledge from each other person.  2. We replaced the time benefit scores in the database with the minimum number of hours from questionnaires.  For example, a rating of “1” was replaced with 0 hours, a rating of “5” was replaced with 13 hours, etc. This was done to obtain results that can be interpreted from linear regression analysis.  3. We divided the matrix data into four other matrices, which represented the number of subgroups identified in the knowledge network.  Each matrix’s rows contained the identified members of one KBS and the columns contained each member’s perceptions of time saved from all 142 participants.  We obtained four subgroups (further explained in the next section) which included 35, 33, 42 and 52 members each. Thus, the four matrices were a 35 by 142, a 33 by 142, a 42 by 142 and a 52 by 142 matrix. 4. In each matrix, the column scores were summed for each of the 142 participants in order to obtain perceptions of all members in one subgroup about time saved from receiving knowledge from each of the 142 employees. The 142 values in each of the four datasets were further used for the linear regression analysis.  5. A new variable was added to each of the 142 participants in each of the four datasets – entitled subgroup membership. If members belonged to the subgroup analyzed, then they were assigned a score of 1 and if they belonged to one of the three external subgroups then they were assigned a score of 0. This was done to differentiate between time benefits members gained in their subgroup (1) compared to time benefits gained between subgroups (0).  229-5 6. We conducted linear regression analysis using SPSS (SPSS 2013) to identify whether the time benefits were more likely to be associated with knowledge receiving in KBS or between KBS. The dependent variable included the time benefits in hours per month (continuous), while the independent variables included subgroup membership (dichotomous), hierarchy (continuous), tenure (continuous) and gender (dichotomous). 5 RESEARCH RESULTS   5.1 Identifying Knowledge-based Subgroups  Four subgroups were identified using Louivan modularity algorithm: KBS A, KBS B, KBS C and KBS D (Figure 2).   Figure 2: Visual representation of KBS  The size of the subgroups varied from 27 to 50 members (Table 1). KBS D was the largest subgroup and included 50 members, while KBS A was the smallest subgroup and included 27 members. The blockmodel analysis (Table 1) revealed that the subgroups’ internal density is higher than its external density, validating the existence of subgroups. In this case, members within subgroup A were connected 30% of the time to members within subgroup B while members within subgroup A were connected to 66% of other members within subgroup A. These results were especially meaningful due to the overall density of the network, which was 45%, meaning that of all possible knowledge connections within this department, 45% actually existed.          5.2 Time Benefits  Initially, we obtained descriptive statistics of all independent variables (Table 2). Then, we ran linear regression analysis to identify whether knowledge receiving in KBS brought more time benefit then across KBS (Table 3). All fours models found that receiving knowledge was perceived to be more beneficial within KBS. Specifically, members saved, on average, 46.5 hours per month more by receiving Table 1: Blockmodel of KBS  KBS A B C D N A 66% 30% 21% 22% 27 B 34% 97% 44% 46% 29 C 18% 32% 82% 60% 36 D 23% 45% 65% 91% 50 229-6 knowledge from other members within KBS A then other subgroups, 56.2 hours per month in KBS B, 74.7 hours per month in KBS C, and 89.5 hours per month in KBS D.   As a result, we found that strong ties were more likely to offer individual time benefits compared to weak ties.  Interestingly, increased network density (Table 1) within a KBS meant more hours saved in three cases (Table 3). Subgroup B had larger network density then Subgroup A (84% compared to 70%), and was more likely to save more hours (56.2 hours compared to 46.5 hours) from receiving knowledge. Subgroup D had a larger network density then Subgroup C (87% compared to 67%) and Subgroup B (87% compared to 84%), and was more likely to save more hours than Subgroup C (89.5 hours compared to 74. 7 hours) and Subgroup B (89.5 hours compared to 56.2 hours) respectively. However, Subgroup C was an exception. While its network density was lower than Subgroup B’s network density (67% compared to 84%), individuals in that subgroup saved more time then Subgroup B (74.7 hours compared to 56.2 hours).  Another set of results show that individuals in higher hierarchy levels are more likely to offer time benefits to others. Three of the four KBS (KBS A, KBS C, and KBS D) were positive and significant in the amount of time they saved to others. For instance, an increase in one hierarchy level is likely to be associated with an increase of 2 hours a month (KBS A) and up to 7 hours a month (KBS D) time benefits. Other control variables such as tenure and gender were not significant in any of the four models.  Table 2: Descriptive Statistics Regression variable             Min Max Mean SD N Hierarchy Tenure Gender  Time Benefits KBS A  Time Benefits KBS B Time Benefits KBS C Time Benefits KBS D Subgroup Membership A Subgroup Membership B Subgroup Membership C Subgroup Membership D 1 1 0 0 0 0 0 0 0 0 0 5 32 1 125 159 187 205 1 1 1 1 1.84 6.04 .34 11.8 18.7 28.24 49.8 .19  .20 .25 .35 1.24 5.16 .47 22.2 28.2 40.1 51.7 .394 .405 .437 .479 142 142 142 142 142 142 142 142 142 142 142   Table 3. Linear regression analysis of time benefits in KBS from subgroup membership Independent Variables Model 1 Model 2 Model 3  Model 4 Hierarchy  2.2*** 1.5 3.2***     7*** Tenure Gender    .2   -.7  -.8 -3.8   .2 -1.8    -.9 -10.8 KBS A membership 46.5***    -    -     - KBS B membership     - 56.2***    -     - KBS C membership    -    - 74.7***     - KBS D membership    -    -    -   89.5*** *, **, *** indicates statistical significance at 10%, 5% and 1% respectively. 6 DISCUSSION AND IMPLICATIONS  Our results found that members of KBS were more likely to report saving time as a result of receiving knowledge from others within their KBS compared to receiving knowledge from employees outside their KBS. While large engineering and construction organizations allocate resources to improve knowledge exchange, it is important to strategically identify when knowledge exchange is most efficient and results in benefits to the employees and organizations. Our results found that individuals are more likely to save 229-7 time, and thus benefit, from others within in their KBS. Specifically, we identified a statistically significant relationship in all KBS between group membership and time benefits. An individual is likely to save an average of 46.5 to 89.5 hours more a month as a result of receiving knowledge within their KBS then between KBS (Table 3).  A major implication from this study is linking group level characteristics with individual level benefits and testing the strength of weak ties theory (Granovetter 1973; Wanberg and Javernick-Will 2014). Previous research has focused primarily on either the effect of group level knowledge exchange on group level benefits (Cummings 2004; Landaeta 2008; Tsai 2001) or the effect of individual level knowledge exchange on individual level benefits (Wanberg and Javernick-Will 2014). Instead, this paper bridges the two levels of analysis.  Specifically, we found that network connectivity and the type of tie (strong or weak) within a highly connected network play a role in individual time benefits.  Strong ties are knowledge exchange connections that are part of a highly connected network (KBS) while weak ties are knowledge exchange connections that link two people of different knowledge based subgroups. As a result, they benefit individuals differently. Weak ties likely require more introduction and steps to access particular knowledge external to the knowledge seeker’s KBS. Conversely, strong ties likely require less steps for introduction as the connectivity between members (knowledge providers and knowledge seekers) is quite high.  It is important to note, however, that our results do not suggest that weak ties are unimportant in knowledge networks. Instead they show that weak ties are less important when it comes to saving time on immediate project- and organizational- related tasks. In all likelihood, weak ties in KBS are important for innovation and accessing knowledge that differs from the knowledge seeker’s immediate KBS. As such, our study builds upon existing theory of strong and weak ties by different outcomes (time benefits).  Interestingly, we also found a relationship pattern between KBS density and time benefits.  This suggests that network density may play a role in the level of time benefits obtained from knowledge receiving, which validates our previous propositions that a knowledge network’s connectivity mediates the effect of strong ties on time benefits. However, these results require further research and a larger subgroup sample for validation.  7 CONCLUSIONS Many construction and engineering organizations increasingly understand the importance of intra-organizational knowledge sharing for achieving competitive advantage. However, few studies have studied the benefits of knowledge exchange, and even fewer have analyzed when receiving knowledge offers best individual benefits. Specifically, previous literature has focused on individual and group level outcomes separately, with a dearth of literature linking group level knowledge sharing and individual level benefits. Based upon the strength of weak ties theory (Granovetter 1973), we tested whether strong ties in  knowledge-based subgroups (KBS) – subgroups that share more knowledge internally then externally – bring more time benefits, based upon the minimum hours per month saved on work tasks, than weak ties between members of different KBS. To do this, we used social network analysis and a modularity optimization algorithm (Blondel et al. 2008) to first identify knowledge-based subgroups (KBS). Then, we identified the hours saved from receiving knowledge within KBS and outside of KBS.   We found that members of KBS are likely to perceive more time saved per month from receiving knowledge within, versus outside of their KBS. These results contradict previous theory on the importance of weak ties (Granovetter 1973) which showed that weak ties are more valuable than strong ties. Instead, we showed that strong ties are perceived as more important, specifically when it comes to saving time, in knowledge networks.  8 LIMITATIONS AND FUTURE WORK There are several limitations in this work. First, this study was conducted in the department of one construction and engineering organization, making generalizations across other construction and engineering organizations more difficult. Second, while we have a large sample size, the analysis suggests 229-8 through correlations, the effects of weak and strong ties on obtaining time benefits, but is unable to show causation. To address these limitations, we propose reproducing this research across additional construction and engineering organizations and conducting qualitative analysis to better understand why weak and strong ties have differential effects on individual time benefits.  Acknowledgements  This material is based in part on work supported by the National Science Foundation grant #1430826. Any opinions, findings and conclusions or recommendations expressed in this material are those of the author and do not necessarily reflect the views of the National Science Foundation. References Argote, Linda, and Paul Ingram. 2000. “Knowledge Transfer: A Basis for Competitive Advantage in Firms.” Organizational Behavior and Human Decision Processes 82 (1): 150–69. doi:10.1006/obhd.2000.2893. Bastian, M, S Heymann, and M Jacomy. 2009. “Gephi: An Open Source Software for Exploring and Manipulating Networks.” International AAAI Conference on Weblogs and Social Media. Blondel, Vincent D., Jean-Loup Guillaume, Renaud Lambiotte, and Etienne Lefebvre. 2008. “Fast Unfolding of Communities in Large Networks.” Journal of Statistical Mechanics: Theory and Experiment 2008 (10): P10008. doi:10.1088/1742-5468/2008/10/P10008. Burt, Ronald. 1995. Structural Holes: The Social Structure of Competition. Cambridge, Mass.: Harvard University Press. Cummings, Jonathon N. 2004. “Work Groups, Structural Diversity, and Knowledge Sharing in a Global Organization.” Management Science 50 (3): 352–64. doi:10.1287/mnsc.1030.0134. Granovetter, Mark S. 1973. “The Strength of Weak Ties.” American Journal of Sociology 78 (6): 1360–80. Grant, Robert M. 1996. “Toward a Knowledge-Based Theory of the Firm.” Strategic Management Journal 17 (S2): 109–22. doi:10.1002/smj.4250171110. Kogut, Bruce, and Udo Zander. 1992. “Knowledge of the Firm, Combinative Capabilities, and the Replication of Technology.” Organization Science 3 (3): 383–97. doi:10.1287/orsc.3.3.383. Landaeta, Rafael E. 2008. “Evaluating Benefits and Challenges of Knowledge Transfer Across Projects.” Engineering Management Journal 20 (1): 29–38. Liebeskind, Julia Porter. 1996. “Knowledge, Strategy, and the Theory of the Firm.” Strategic Management Journal 17 (S2): 93–107. doi:10.1002/smj.4250171109. Nahapiet, Janine, and Sumantra Ghoshal. 1998. “Social Capital, Intellectual Capital, and the Organizational Advantage.” Academy of Management Review 23 (2): 242–66. doi:10.5465/AMR.1998.533225. Netminer. 2014. “NetMiner v4.2.1.140729 Seoul: Cyram Inc.” Poleacovschi, Cristina, and A. Javernick-Will. under review. “Spanning Knowledge across Subrgoups and Its Effects on Individual Performance.” Journal of Management in Engineering SPSS. 2013. “IBM Corp. Released 2013. IBM SPSS Statistics for Windows, Version 22.0. Armonk, NY: IBM Corp.” Tsai, Wenpin. 2001. “Knowledge Transfer in Intraorganizational Networks: Effects of Network Position and Absorptive Capacity on Business Unit Innovation and Performance.” The Academy of Management Journal 44 (5): 996–1004. doi:10.2307/3069443. Wanberg, J., and A. Javernick-Will. 2014. “Evaluating the Usefulness of Knowledge Sharing Connections in Multinational Construction Companies.” In Construction Research Congress 2014, 1967–76. American Society of Civil Engineers. http://ascelibrary.org/doi/abs/10.1061/9780784413517.201. Wanberg, J., A. Javernick-Will, P. Chinowsky, and J. Taylor. 2014. “Spanning Cultural and GeograBarriers with Knowledge Pipelines in Multinational Communities of Practice.” Journal of Construction Engineering and Management 0 (0): 04014091. doi:10.1061/(ASCE)CO.1943-7862.0000955. Wanberg, John, Amy Javernick-Will, John Taylor, and Paul Chinowsky. 2014. “A Useful Knot or a Deadly Noose? The Effects of Organizational Divisions on Informal Knowledge Sharing Networks.” Engineering Project Organizations Journal (revision under Review). 229-9  5th International/11th Construction Specialty Conference 5e International/11e Conférence spécialisée sur la construction    Vancouver, British Columbia June 8 to June 10, 2015 / 8 juin au 10 juin 2015    DO STRONG OR WEAK TIES MATTER IN KNOWLEDGE NETWORKS? Cristina Poleacovschi1, Amy N. Javernick-Will1 1 Department of Civil, Architectural, and Environmental Engineering, University of Colorado Boulder, USA  Abstract: Construction and engineering organizations have increasingly implemented knowledge exchange strategies with the goal to facilitate knowledge exchange across the organization. However, despite these efforts, many knowledge strategies fail in practice, as it is not well known when knowledge access is most beneficial. This research analyzes the correlation between group level knowledge exchange and perceived individual benefits. Specifically, we focus on the time saved (in hours per month) on work tasks as a result of accessing knowledge with others in the department. To conduct this research, we used social network analysis and a modularity optimization algorithm to identify the existing knowledge-based subgroups (KBS)—subgroups that share more knowledge internally then externally—within a large engineering and construction organization. To identify whether these knowledge based subgroups offer time benefits, we compared the time benefits from receiving knowledge within these subgroups and outside these subgroups. Results found that individuals are more likely to perceive saving time on work tasks as a result of receiving knowledge within their subgroups. As a result, our results indicate that the type of benefit received matters to determine whether weak ties or strong ties are important.   1 INTRODUCTION The importance of knowledge sharing in construction and engineering organizations has gained an unprecedented interest and many scholars view knowledge as the most important resource of the firm (Grant 1996). The efficient allocation of money and labor is not sufficient enough to gain competitive advantage anymore. Instead, knowledge and its use at the right place and time is what allows projects to achieve successful results (Argote and Ingram 2000).  While this view is shared largely in the literature, there are many debates about how and when knowledge exchange brings the most benefits. To address this debate, previous studies have focused on knowledge exchange primarily at the group level, with emerging studies at the individual level. At the group level, the literature has focused on the importance of knowledge exchange on group performance (Cummings 2004; Landaeta 2008; Tsai 2001). For instance, Tsai (2001) showed that business units which are more central are more likely to receive diverse knowledge from other business units which in turn affects innovation and performance. While studying the effects and benefits at the group level is important, it provides an incomplete understanding of how knowledge exchange adds value to outcomes.  Specifically, aggregating group level outcomes offers a good understanding of the effect of group level attributes (e.g. size, composition) on final outcomes, but it neglects intra-group dynamics, which represent the core of group outcomes. To address the dearth of literature on individual level outcomes, Wanberg and Javernick-Will (2014) studied the relationship between frequency of individual knowledge exchange and 229-1 individual work outcomes in informal organizational groups. They found that infrequent interactions provided individuals with more unique knowledge then frequent knowledge interactions.  As such, individual level knowledge exchange has been analyzed for individual level performance and group level knowledge exchange has been analyzed for group level performance. As a result, the connection between the group level characteristics and individual level benefits has been largely neglected with the exception of Poleacovschi and Javernick-Will (under review) who found that individuals spanning knowledge across highly connected knowledge subgroups are more likely to receive higher individual performance evaluations in a construction and engineering organization.  This research proposes to further address the missed macro-micro link by analyzing the relationship between group knowledge exchange and individual level outcomes. Previous research has identified that individuals gain benefits from their position in the network structure. Nahapiet and Ghoshal (1998) found that individuals share intellectual advantages, such as knowledge exchange, as a result of being part of highly connected groups. Thus, we expect that individuals who are part of highly cohesive knowledge exchange groups will gain benefits from the network connectivity of their group. However, we expect that this relationship is sensitive to the type of relationship, or tie, between the individual and the group. This thinking is in line with the theory of weak ties (Granovetter 1973) which argues that the type of ties matter in a network. To translate Granovetter’s study at the level of subgroups, strong ties are part of highly connected groups while weak ties connect individuals from different subgroups. Considering the importance of project time in construction and engineering organizations, we define individual benefits as the time an individual saved (in hours per month) on work tasks as a result of receiving knowledge from other employees, whether within their knowledge based subgroup or outside of their knowledge based subgroup. Thus, this research asks: Do individuals save more time as a result of their weak or strong ties within knowledge exchange networks? To answer this question we employed a unique method that identified knowledge-based subgroups (KBS) within a department of a large construction and engineering organization (Poleacovschi and Javernick-Will under review). The method used a social network optimization algorithm that identified subgroups that shared more knowledge internally then externally. Then, using survey data, we identified the perceived time savings as a result of receiving knowledge within their KBS as compared to receiving knowledge outside of their KBS.  This research is important both for theory and practice. To contribute to theory, this research links group and individual level constructs by testing the strength of weak ties theory in KBS. By translating weak and strong ties concepts to KBS, knowledge connections in KBS are strong ties because these subgroups are dense and cohesive, while connections outside KBS are weak ties. By focusing on the specific benefit of time savings, we contribute to theory by extending previous work on different types of benefits that can be received from strong or weak ties. In practice, construction and engineering organizations can chose to invest resources based upon the benefits they seek—in our research, this relates to saving time on work tasks.  2 THEORETICAL BACKGROUND 2.1 Knowledge-based subgroups  The interest in knowledge-based subgroups, or groups that share more knowledge internally then externally, is rooted in the knowledge-based theory of the firm (Grant 1996). In this view, the ability to efficiently exchange the “know what” and especially the “know how” is what differentiates organizations from markets (Kogut and Zander 1992). In other words, construction and engineering organizations exist because they successfully manage to integrate the knowledge of professionals from different disciplines (e.g. structural engineers, project managers) that can easily exchange knowledge, or “know how”, to achieve project goals. Another option is to look for professional knowledge (e.g. structural analysis and project management) on the market. However, this would be less efficient as separate transactions with different professionals would be costly. The knowledge-based theory of the firm also believes that an organization gains competitive advantage whenever it manages to coordinate knowledge exchange better 229-2 than other organizations (Argote and Ingram 2000; Grant 1996). Unfortunately, this is not easy as there are many organizational structures and boundaries that impede knowledge exchange (Wanberg et al. 2014; John Wanberg et al. 2014)  As a result of these boundaries, knowledge-based subgroups—subgroups that share more knowledge internally and less externally—can form. The common belief is that subgroups disrupt knowledge exchange, which may be detrimental for the larger performance of the group or organization.  For instance, individuals are more densely connected to others within their subgroup and limited knowledge may flow between subgroups, resulting in a boundary that may impede members of the group to search and access knowledge outside their own subgroup. However, this assumes that knowledge exchange with others outside one’s own subgroup is actually valuable – which may not be true. More importantly, valuable in terms of what and valuable for whom? As previous research has mainly focused on the effects of subgroup formation on group level outcomes, we do not know the effect of subgroup formation on individual benefits in terms of time saved on work tasks. Using theory of weak ties we build our hypothesis about the relationship between network ties with KBS and individual time benefits.  2.2 Ties within and between knowledge-based subgroups The most influential papers addressing individual benefits from network structure are the structural hole (Burt 1995) and the strength of weak ties (Granovetter 1973) theories. A structural hole means the absence of a relationship between two individuals in a network. The theory states that individuals who link two unconnected others are likely to gain benefits. The strength of weak ties theory was the antecedent of structural hole theory and focused on the types of ties and benefits results those ties (Granovetter 1973). Individuals who manage to create ties outside their own subgroup of close relationships (weak ties) gain benefits as a result of reaching for novel resources and information. For instance, in his thesis, Granovetter discovered that weak ties are the most beneficial connections for a job search. Figure 2 shows a visual representation of weak and strong ties. Straight lines represent strong ties while dotted lines represent weak ties.   Figure 1: The Strength of Weak Ties ((Granovetter 1973) While this theory is quite influential, it was initially conceptualized and tested in the context of people who were searching for jobs. The theory has been tested less in the context of knowledge networks or considering individual time benefits. KBS are subgroups whose members share more knowledge internally and less knowledge externally. Thus, ties in KBS are strong because of the high connectivity between members, while ties outside of, or between, KBS are weak as they connect distant subgroups, which are highly connected. In contrast to Granovetter’s study, we expect that highly connected subgroups will save individuals more time as a result of the short distance between subgroup members. Conversely, accessing a new subgroup of knowledge through weak ties requires many more steps. As such, individuals will save more time from accessing knowledge in their highly connected network then from accessing knowledge in other subgroups. For instance, member B will access member G’s knowledge easier then member D’s knowledge, although it takes the same number of steps (Figure 1). In other words, members of KBS will use their subgroup connectivity to access knowledge fast. Thus, we expect that the type of tie between the individual and the highly connected network further benefits the individual. If the tie is strong (within a KBS), we expect that the individual will gain more time benefits from accessing that knowledge.  229-3 Hypothesis: Individuals will save more time from accessing knowledge from strong ties (within subgroups) then from weak ties (between subgroups) 3 RESEARCH METHODS: DATA COLLECTION  We were provided access to a large set of network data within the department of a large construction and engineering consultancy. The organization has focused on improving their knowledge management strategies for the past decade.  They initially implemented knowledge management in the IT department, which was where the data was collected. Survey questionnaires were administered to all employees within this department, with questions focused on with whom they share and receive knowledge and whether these connections help to save them time through the knowledge provided. The response rate for the knowledge network data was 88% (n=142).  Non-respondents were excluded from the final analysis.  3.1 Knowledge sharing networks Questionnaires were administered to all 161 employees in the department. Using definitions of knowledge as information and :know how (Liebeskind 1996), employees were asked to indicate, on a 6-point scale, the level of knowledge that they receive from other departmental employees. Specifically, they were asked:   “Often we rely on the people we work with to provide us with information to get our work done. For example, people might provide us with simple routine administrative or technical information that we need to do our work. Alternatively, people might provide us with complex information or engage in problem solving with us to help us solve novel problem. Please indicate the extent to which people listed below provide you with information you use to accomplish your work?”  Response options included: 0 – I do not know this person/I have never met this person; 1 – Very Infrequently; 2 – Infrequently; 3 – Somewhat infrequently; 4 – Somewhat frequently; 5 – Frequently; 6 – Very frequently.  The data was transferred into a matrix format where questionnaire respondents were listed as rows, and the level of knowledge sharing with each other departmental employee was indicated in columns.  The data was then analyzed using social network analysis (SNA) software to determine the composition of knowledge-based subgroups.  3.2 Time Benefits To identify time benefits accrued as a result of receiving knowledge within and outside KBS, individuals were asked to indicate the time they saved per month (in hours) as a result of receiving knowledge from each individual in the department.  They responded to the prompt:  “Please provide an estimate for the typical time saved per month as a result of information,    or other resources received from each person.” The respondents assessed the time benefits for each other departmental employee based on a 5 – point scale, including: 0 – I do not know this person/I have never met this person; 1 – No time saved; 2 – 1-3 Hours per month; 3 – 4-8 Hours per month; 4 – 9-12 hours per month; to 5 – more than 13 hours per month. 3.3 Individual attributes: hierarchy, tenure and gender To identify individual attributes, individuals were asked to identify their hierarchy level and tenure (in years) within the organization. Hierarchical levels were determined from a scale of one to five (1 – individual contributor/team member; 2 – supervisor/team leader; 3 – project manager/program manager; 4 – manager/BU [business unit] manager; 5 – director). For tenure, individuals mentioned the number of years they have spent with the organization. Finally, the department provided the data on gender.  This 229-4 information was later used for the linear regression analysis as control variables.  We controlled for hierarchy, tenure and gender as women and newcomers are not expected to gain the same benefits from their position in the network (Burt 1992), while individuals in higher hierarchy levels are expected to be central in networks and gain more benefits (Poleacovschi and Javernick-Will under review).  4 RESEARCH METHODS: DATA ANALYSIS In this section we present methods used to analyze the collected data with the goal to identify KBS and the average time saved in versus outside of KBS. 4.1 Identifying Knowledge-based Subgroups  We analyzed structural aspects of the department by analyzing the knowledge sharing connections between individuals within the network.  We used SNA to identify KBS, which is described in (Poleacovschi and Javernick-Will under review), using Gephi (Bastian et al. 2009), a visualization and network analysis software. Gephi includes a tool that uses the Louivan modularity optimization algorithm (Blondel et al. 2008) to determine groups within the network that share more knowledge internally and less knowledge externally. The algorithm is based on calculating the modularity gain, or, in our case, the change in the level of knowledge exchange within a group as a result of adding a new member to it, compared to the level of knowledge exchange outside the group once the member joins.  This determined the number of KBS—which, for this network was four, and the composition of, or employees that belonged to, each of the four KBS.  We validated the KBS by running a ‘blockmodel’ in Netminer (Netminer 2014). The blockmodel allowed us to calculate and compare the network density, or the ratio of existing connections to all possible connections, between different KBS.  This helped to validate the algorithm, as higher network densities were observed within subgroups compared to network densities between subgroups.  4.2 Identifying Time Benefits in Knowledge-based Subgroups After the KBS were determined, we analyzed whether individuals within the subgroups received greater time benefits (saving more hours per month) from the knowledge received from employees within their KBS or external to their KBS.  To do this, we conducted several steps, outlined below. 1. We constructed a matrix of 142 rows by 142 columns (number of people who took the questionnaire).  The rows represented questionnaire respondents, and the columns included the average amount of time they saved per month (on a scale from 0 to 5) based upon receiving knowledge from each other person.  2. We replaced the time benefit scores in the database with the minimum number of hours from questionnaires.  For example, a rating of “1” was replaced with 0 hours, a rating of “5” was replaced with 13 hours, etc. This was done to obtain results that can be interpreted from linear regression analysis.  3. We divided the matrix data into four other matrices, which represented the number of subgroups identified in the knowledge network.  Each matrix’s rows contained the identified members of one KBS and the columns contained each member’s perceptions of time saved from all 142 participants.  We obtained four subgroups (further explained in the next section) which included 35, 33, 42 and 52 members each. Thus, the four matrices were a 35 by 142, a 33 by 142, a 42 by 142 and a 52 by 142 matrix. 4. In each matrix, the column scores were summed for each of the 142 participants in order to obtain perceptions of all members in one subgroup about time saved from receiving knowledge from each of the 142 employees. The 142 values in each of the four datasets were further used for the linear regression analysis.  5. A new variable was added to each of the 142 participants in each of the four datasets – entitled subgroup membership. If members belonged to the subgroup analyzed, then they were assigned a score of 1 and if they belonged to one of the three external subgroups then they were assigned a score of 0. This was done to differentiate between time benefits members gained in their subgroup (1) compared to time benefits gained between subgroups (0).  229-5 6. We conducted linear regression analysis using SPSS (SPSS 2013) to identify whether the time benefits were more likely to be associated with knowledge receiving in KBS or between KBS. The dependent variable included the time benefits in hours per month (continuous), while the independent variables included subgroup membership (dichotomous), hierarchy (continuous), tenure (continuous) and gender (dichotomous). 5 RESEARCH RESULTS   5.1 Identifying Knowledge-based Subgroups  Four subgroups were identified using Louivan modularity algorithm: KBS A, KBS B, KBS C and KBS D (Figure 2).   Figure 2: Visual representation of KBS  The size of the subgroups varied from 27 to 50 members (Table 1). KBS D was the largest subgroup and included 50 members, while KBS A was the smallest subgroup and included 27 members. The blockmodel analysis (Table 1) revealed that the subgroups’ internal density is higher than its external density, validating the existence of subgroups. In this case, members within subgroup A were connected 30% of the time to members within subgroup B while members within subgroup A were connected to 66% of other members within subgroup A. These results were especially meaningful due to the overall density of the network, which was 45%, meaning that of all possible knowledge connections within this department, 45% actually existed.          5.2 Time Benefits  Initially, we obtained descriptive statistics of all independent variables (Table 2). Then, we ran linear regression analysis to identify whether knowledge receiving in KBS brought more time benefit then across KBS (Table 3). All fours models found that receiving knowledge was perceived to be more beneficial within KBS. Specifically, members saved, on average, 46.5 hours per month more by receiving Table 1: Blockmodel of KBS  KBS A B C D N A 66% 30% 21% 22% 27 B 34% 97% 44% 46% 29 C 18% 32% 82% 60% 36 D 23% 45% 65% 91% 50 229-6 knowledge from other members within KBS A then other subgroups, 56.2 hours per month in KBS B, 74.7 hours per month in KBS C, and 89.5 hours per month in KBS D.   As a result, we found that strong ties were more likely to offer individual time benefits compared to weak ties.  Interestingly, increased network density (Table 1) within a KBS meant more hours saved in three cases (Table 3). Subgroup B had larger network density then Subgroup A (84% compared to 70%), and was more likely to save more hours (56.2 hours compared to 46.5 hours) from receiving knowledge. Subgroup D had a larger network density then Subgroup C (87% compared to 67%) and Subgroup B (87% compared to 84%), and was more likely to save more hours than Subgroup C (89.5 hours compared to 74. 7 hours) and Subgroup B (89.5 hours compared to 56.2 hours) respectively. However, Subgroup C was an exception. While its network density was lower than Subgroup B’s network density (67% compared to 84%), individuals in that subgroup saved more time then Subgroup B (74.7 hours compared to 56.2 hours).  Another set of results show that individuals in higher hierarchy levels are more likely to offer time benefits to others. Three of the four KBS (KBS A, KBS C, and KBS D) were positive and significant in the amount of time they saved to others. For instance, an increase in one hierarchy level is likely to be associated with an increase of 2 hours a month (KBS A) and up to 7 hours a month (KBS D) time benefits. Other control variables such as tenure and gender were not significant in any of the four models.  Table 2: Descriptive Statistics Regression variable             Min Max Mean SD N Hierarchy Tenure Gender  Time Benefits KBS A  Time Benefits KBS B Time Benefits KBS C Time Benefits KBS D Subgroup Membership A Subgroup Membership B Subgroup Membership C Subgroup Membership D 1 1 0 0 0 0 0 0 0 0 0 5 32 1 125 159 187 205 1 1 1 1 1.84 6.04 .34 11.8 18.7 28.24 49.8 .19  .20 .25 .35 1.24 5.16 .47 22.2 28.2 40.1 51.7 .394 .405 .437 .479 142 142 142 142 142 142 142 142 142 142 142   Table 3. Linear regression analysis of time benefits in KBS from subgroup membership Independent Variables Model 1 Model 2 Model 3  Model 4 Hierarchy  2.2*** 1.5 3.2***     7*** Tenure Gender    .2   -.7  -.8 -3.8   .2 -1.8    -.9 -10.8 KBS A membership 46.5***    -    -     - KBS B membership     - 56.2***    -     - KBS C membership    -    - 74.7***     - KBS D membership    -    -    -   89.5*** *, **, *** indicates statistical significance at 10%, 5% and 1% respectively. 6 DISCUSSION AND IMPLICATIONS  Our results found that members of KBS were more likely to report saving time as a result of receiving knowledge from others within their KBS compared to receiving knowledge from employees outside their KBS. While large engineering and construction organizations allocate resources to improve knowledge exchange, it is important to strategically identify when knowledge exchange is most efficient and results in benefits to the employees and organizations. Our results found that individuals are more likely to save 229-7 time, and thus benefit, from others within in their KBS. Specifically, we identified a statistically significant relationship in all KBS between group membership and time benefits. An individual is likely to save an average of 46.5 to 89.5 hours more a month as a result of receiving knowledge within their KBS then between KBS (Table 3).  A major implication from this study is linking group level characteristics with individual level benefits and testing the strength of weak ties theory (Granovetter 1973; Wanberg and Javernick-Will 2014). Previous research has focused primarily on either the effect of group level knowledge exchange on group level benefits (Cummings 2004; Landaeta 2008; Tsai 2001) or the effect of individual level knowledge exchange on individual level benefits (Wanberg and Javernick-Will 2014). Instead, this paper bridges the two levels of analysis.  Specifically, we found that network connectivity and the type of tie (strong or weak) within a highly connected network play a role in individual time benefits.  Strong ties are knowledge exchange connections that are part of a highly connected network (KBS) while weak ties are knowledge exchange connections that link two people of different knowledge based subgroups. As a result, they benefit individuals differently. Weak ties likely require more introduction and steps to access particular knowledge external to the knowledge seeker’s KBS. Conversely, strong ties likely require less steps for introduction as the connectivity between members (knowledge providers and knowledge seekers) is quite high.  It is important to note, however, that our results do not suggest that weak ties are unimportant in knowledge networks. Instead they show that weak ties are less important when it comes to saving time on immediate project- and organizational- related tasks. In all likelihood, weak ties in KBS are important for innovation and accessing knowledge that differs from the knowledge seeker’s immediate KBS. As such, our study builds upon existing theory of strong and weak ties by different outcomes (time benefits).  Interestingly, we also found a relationship pattern between KBS density and time benefits.  This suggests that network density may play a role in the level of time benefits obtained from knowledge receiving, which validates our previous propositions that a knowledge network’s connectivity mediates the effect of strong ties on time benefits. However, these results require further research and a larger subgroup sample for validation.  7 CONCLUSIONS Many construction and engineering organizations increasingly understand the importance of intra-organizational knowledge sharing for achieving competitive advantage. However, few studies have studied the benefits of knowledge exchange, and even fewer have analyzed when receiving knowledge offers best individual benefits. Specifically, previous literature has focused on individual and group level outcomes separately, with a dearth of literature linking group level knowledge sharing and individual level benefits. Based upon the strength of weak ties theory (Granovetter 1973), we tested whether strong ties in  knowledge-based subgroups (KBS) – subgroups that share more knowledge internally then externally – bring more time benefits, based upon the minimum hours per month saved on work tasks, than weak ties between members of different KBS. To do this, we used social network analysis and a modularity optimization algorithm (Blondel et al. 2008) to first identify knowledge-based subgroups (KBS). Then, we identified the hours saved from receiving knowledge within KBS and outside of KBS.   We found that members of KBS are likely to perceive more time saved per month from receiving knowledge within, versus outside of their KBS. These results contradict previous theory on the importance of weak ties (Granovetter 1973) which showed that weak ties are more valuable than strong ties. Instead, we showed that strong ties are perceived as more important, specifically when it comes to saving time, in knowledge networks.  8 LIMITATIONS AND FUTURE WORK There are several limitations in this work. First, this study was conducted in the department of one construction and engineering organization, making generalizations across other construction and engineering organizations more difficult. Second, while we have a large sample size, the analysis suggests 229-8 through correlations, the effects of weak and strong ties on obtaining time benefits, but is unable to show causation. To address these limitations, we propose reproducing this research across additional construction and engineering organizations and conducting qualitative analysis to better understand why weak and strong ties have differential effects on individual time benefits.  Acknowledgements  This material is based in part on work supported by the National Science Foundation grant #1430826. Any opinions, findings and conclusions or recommendations expressed in this material are those of the author and do not necessarily reflect the views of the National Science Foundation. References Argote, Linda, and Paul Ingram. 2000. “Knowledge Transfer: A Basis for Competitive Advantage in Firms.” Organizational Behavior and Human Decision Processes 82 (1): 150–69. doi:10.1006/obhd.2000.2893. Bastian, M, S Heymann, and M Jacomy. 2009. “Gephi: An Open Source Software for Exploring and Manipulating Networks.” International AAAI Conference on Weblogs and Social Media. Blondel, Vincent D., Jean-Loup Guillaume, Renaud Lambiotte, and Etienne Lefebvre. 2008. “Fast Unfolding of Communities in Large Networks.” Journal of Statistical Mechanics: Theory and Experiment 2008 (10): P10008. doi:10.1088/1742-5468/2008/10/P10008. Burt, Ronald. 1995. Structural Holes: The Social Structure of Competition. Cambridge, Mass.: Harvard University Press. Cummings, Jonathon N. 2004. “Work Groups, Structural Diversity, and Knowledge Sharing in a Global Organization.” Management Science 50 (3): 352–64. doi:10.1287/mnsc.1030.0134. Granovetter, Mark S. 1973. “The Strength of Weak Ties.” American Journal of Sociology 78 (6): 1360–80. Grant, Robert M. 1996. “Toward a Knowledge-Based Theory of the Firm.” Strategic Management Journal 17 (S2): 109–22. doi:10.1002/smj.4250171110. Kogut, Bruce, and Udo Zander. 1992. “Knowledge of the Firm, Combinative Capabilities, and the Replication of Technology.” Organization Science 3 (3): 383–97. doi:10.1287/orsc.3.3.383. Landaeta, Rafael E. 2008. “Evaluating Benefits and Challenges of Knowledge Transfer Across Projects.” Engineering Management Journal 20 (1): 29–38. Liebeskind, Julia Porter. 1996. “Knowledge, Strategy, and the Theory of the Firm.” Strategic Management Journal 17 (S2): 93–107. doi:10.1002/smj.4250171109. Nahapiet, Janine, and Sumantra Ghoshal. 1998. “Social Capital, Intellectual Capital, and the Organizational Advantage.” Academy of Management Review 23 (2): 242–66. doi:10.5465/AMR.1998.533225. Netminer. 2014. “NetMiner v4.2.1.140729 Seoul: Cyram Inc.” Poleacovschi, Cristina, and A. Javernick-Will. under review. “Spanning Knowledge across Subrgoups and Its Effects on Individual Performance.” Journal of Management in Engineering SPSS. 2013. “IBM Corp. Released 2013. IBM SPSS Statistics for Windows, Version 22.0. Armonk, NY: IBM Corp.” Tsai, Wenpin. 2001. “Knowledge Transfer in Intraorganizational Networks: Effects of Network Position and Absorptive Capacity on Business Unit Innovation and Performance.” The Academy of Management Journal 44 (5): 996–1004. doi:10.2307/3069443. Wanberg, J., and A. Javernick-Will. 2014. “Evaluating the Usefulness of Knowledge Sharing Connections in Multinational Construction Companies.” In Construction Research Congress 2014, 1967–76. American Society of Civil Engineers. http://ascelibrary.org/doi/abs/10.1061/9780784413517.201. Wanberg, J., A. Javernick-Will, P. Chinowsky, and J. Taylor. 2014. “Spanning Cultural and GeograBarriers with Knowledge Pipelines in Multinational Communities of Practice.” Journal of Construction Engineering and Management 0 (0): 04014091. doi:10.1061/(ASCE)CO.1943-7862.0000955. Wanberg, John, Amy Javernick-Will, John Taylor, and Paul Chinowsky. 2014. “A Useful Knot or a Deadly Noose? The Effects of Organizational Divisions on Informal Knowledge Sharing Networks.” Engineering Project Organizations Journal (revision under Review). 229-9  Which ties matter in knowledge networks?Cristina Poleacovschi, PhD StudentDr. Amy Javernick-Will, Assistant ProfessorUniversity of Colorado BoulderThe International Construction Specialty ConferenceVancouver, CanadaJune 7-10 20151Background: KBV(Grant 1996)“The primary role of firms is in the application of existing knowledge to the production of goods and services”Background: KS & BenefitsTsai 2001:  Central Business Units  Performance/Innovation (  access to new knowledge)Cummings 2004: Work groups  Performance with  external knowledge sharingKnowledge sharingGroup Benefits3Background:  KS & BenefitsKnowledge sharingIndividual Benefits ??Wanberg & Javernick-Will 2014: Knowledge sharing frequency & usefulness of KSC; Poleacovschi & Javernick-Will under review: Network position & performance ratings.4Background: Strength of weak tiesGranovetter (1973)Context: knowledge networks Outcome:  time benefitsweak strongContext: Personal networks Outcome:  Finding a job5Hypothesis: Individuals will gain higher time benefits from accessing knowledge from strong ties than from weak ties Research Question and HypothesisDo strong or weak ties matter [save more time] in knowledge networks?6Research Methods:  Data Collection Online Survey Questionnaires• Sent to all 161 employees in the department (88% response rate)• Questionnaires asked for: • Knowledge sharing (Independent variable)• Time benefits (Dependent variable)• Individual attributes  (Control variable) 7Research Methods:  Data Collection Online Survey Questionnaires• Sent to all 161 employees in the department (88% response rate)• Questionnaires asked for: • Knowledge sharing“Often we rely on the people we work with to provide us with information to get our work done. For example, people might provide us with simple routine administrative ortechnical information that we need to do our work. Alternatively, people might provide us with complex information or engage in problem solving with us to help us solve novel problem. Please indicate the extent to which people listed below provide you with information you use to accomplish your work?” 0 – I do not know this person/I have never met this person; 1 – Very Infrequently; 2 – Infrequently; 3 – Somewhat infrequently; 4 – Somewhat frequently; 5 – Frequently; 6 – Very frequently. 8Research Methods:  Data Collection Online Survey Questionnaires• Sent to all 161 employees in the department (88% response rate)• Questionnaires asked for: • Knowledge sharing• Time benefits“Please provide an estimate for the typical time saved per month as a result of information,  or other resources received from each person.”0 – I do not know this person/I have never met this person; 1 – No time saved; 2 – 1-3 Hours per month; 3 – 4-8 Hours per month; 4 – 9-12 hours per month; to 5 – >13 hours per month.We used the minimum amount of hours during our analysis9Research Methods:  Data Collection Online Survey Questionnaires• Sent to all 161 employees in the department (88% response rate)• Survey data asked for: • Knowledge sharing• Time benefits• Individual attributes  • Hierarchy  (Scale of 1-5): 1 – individual contributor/team member; 2 – supervisor/team leader; 3 – project manager/program manager; 4 – manager/BU [business unit] manager; 5 – director). • Tenure (# of years in the organization) • Gender10Research Methods:  AnalysisDo strong or weak ties matter in knowledge networks?1. Identified knowledge-based subgroups through Social Network AnalysisKnowledge-based subgroups are subgroups that share more knowledge internally then externally.Social network analysis: Used the Louvain Modularity Algorithm (Gephi)Modularity gain group level change in the level of knowledge exchange within a group as a result of adding a new member (compared to external to group once member joins) Research Methods:  AnalysisDo strong or weak ties matter in knowledge networks?1. Identified knowledge-based subgroups through Social Network AnalysisKnowledge-based subgroups are subgroups that share more knowledge internally then externally.Social network analysis: Used the Louvain Modularity Algorithm (Gephi)Validated with a Block Model N = 27Subgroup 1 N = 50 Subgroup 4N = 29Subgroup 2  N = 39Subgroup 3  Louvain Modularity Optimization Algorithm13Research Question and MethodsDo strong or weak ties matter in knowledge networks?1. Identified knowledge-based subgroups through Social Network Analysis 2. Identified weak and strong tiesWeak ties = ties between subgroups Strong ties = ties within subgroups 14Weak versus strong tiesWeak tiesStrong ties15Research Question and MethodsDo strong or weak ties matter in knowledge networks?3. Identified whether strong or weak ties are associated with increased time benefits using linear regression analysis1. Identified knowledge-based subgroups Network Data (N=161) from a department at a infrastructure engineering company2. Identified weak and strong ties16Dependent and independent variables AB CSubgroup 1 Subgroup 2Dependent variable: Number of hours saved from receiving knowledge from B and CIndependent variable: Type of ties  Strong tie (value of 1) and weak tie (value of 0)Ego Alter Min hours saved per monthType of tieA B 4 0A C 9 117Results Regression Variables Model 1 Model 2 Model 3 Model 4Hierarchy 2.2*** 1.5 3.2*** .7***Tenure 0.2 -0.8 0.2 -0.9Gender -0.7 -3.8 -1.8 -10.8Type of tie (1) 46.5*** - - -Type of tie (2) - 56.2*** - -Type of tie (3) - - 74.7*** -Type of tie (4) - - - 89.5***p <.05 ** p <.01 *** p<.001, N=14218Results Regression Variables Model 1 Model 2 Model 3 Model 4Hierarchy 2.2*** 1.5 3.2*** .7***Tenure 0.2 -0.8 0.2 -0.9Gender -0.7 -3.8 -1.8 -10.8Type of tie (1) 46.5*** - - -Type of tie (2) - 56.2*** - -Type of tie (3) - - 74.7*** -Type of tie (4) - - - 89.5***p <.05 ** p <.01 *** p<.001, N=14219Results: Linear regression analysis Regression Variables Model 1 Model 2 Model 3 Model 4Hierarchy 2.2*** 1.5 3.2*** .7***Tenure 0.2 -0.8 0.2 -0.9Gender -0.7 -3.8 -1.8 -10.8Type of tie (1) 46.5*** - - -Type of tie (2) - 56.2*** - -Type of tie (3) - - 74.7*** -Type of tie (4) - - - 89.5***p <.05 ** p <.01 *** p<.001, N=14220Results Regression Variables Model 1 Model 2 Model 3 Model 4Hierarchy 2.2*** 1.5 3.2*** .7***Tenure 0.2 -0.8 0.2 -0.9Gender -0.7 -3.8 -1.8 -10.8Type of tie (1) 46.5*** - - -Type of tie (2) - 56.2*** - -Type of tie (3) - - 74.7*** -Type of tie (4) - - - 89.5***p <.05 ** p <.01 *** p<.001, N=14221Results Regression Variables Model 1 Model 2 Model 3 Model 4Hierarchy 2.2*** 1.5 3.2*** .7***Tenure 0.2 -0.8 0.2 -0.9Gender -0.7 -3.8 -1.8 -10.8Type of tie (1) 46.5*** - - -Type of tie (2) - 56.2*** - -Type of tie (3) - - 74.7*** -Type of tie (4) - - - 89.5***p <.05 ** p <.01 *** p<.001, N=14222Results Regression Variables Model 1 Model 2 Model 3 Model 4Hierarchy 2.2*** 1.5 3.2*** .7***Tenure 0.2 -0.8 0.2 -0.9Gender -0.7 -3.8 -1.8 -10.8Type of tie (1) 46.5*** - - -Type of tie (2) - 56.2*** - -Type of tie (3) - - 74.7*** -Type of tie (4) - - - 89.5***p <.05 ** p <.01 *** p<.001, N=14223Discussion • Strong ties (within subgroups) maymatter more than weak ties (between subgroups) in knowledge networks for saving time• This might be due to individuals saving time from the subgroup high connectivity) 24DiscussionAB CSubgroup 1 Subgroup 2D12123 E25Discussion • Strong ties (within subgroups) maymatter more than weak ties (between subgroups) in knowledge networks for saving time• This might be due to individuals saving time from the subgroup high connectivity or it may be due to subgroup composition (future research)• Expands theory of strength of weak ties by showing that weak vs. strong ties effects are contingent on the context (knowledge sharing and time benefits)26Practical Contribution • In contrast to other studies, our results suggest that cohesive knowledge subgroups can be important for individuals saving times on their tasks• These subgroups can be especially useful during project deadlines, when individuals experience the most time pressures• Construction organizations can facilitate subgroup knowledge sharing by identifying these subgroups and by enhancing mechanisms for knowledge exchange 27• Focused on a single department of a company• Analyses do not show causality, only correlation• Analyses are based on time savings per perception data only• The data does not include assessment of the quality of knowledgeFurther research can reproduce these analyses across additional companies and use qualitative analysis to identify specific benefits and reasons for those benefits based upon structural positions within the networkLimitations and Future Work28Questions? Thank youCristina Poleacovschi, PhD StudentCristina.Poleacovschi@Colorado.eduAmy Javernick-Will, Assistant ProfessorAmy.Javernick@Colorado.eduUniversity of Colorado Boulderwww.projectorganizations.comThis material is based in part on work supported by the National Science Foundation grant #1430826. Any opinions, findings and conclusions or recommendations expressed in this material are those of the author and do not necessarily reflect the views of the National Science Foundation.29Survey questions: Knowledge sharing• “Often we rely on the people we work with to provide us with information to get our work done. For example, people might provide us with simple routine administrative or technical information that we need to do our work. Alternatively, people might provide us with complex information or engage in problem solving with us to help us solve novel problem. Please indicate the extent to which people listed below provide you with information you use to accomplish your work?” • Response options included: 0 – I do not know this person/I have never met this person; 1 – Very Infrequently; 2 – Infrequently; 3 –Somewhat infrequently; 4 – Somewhat frequently; 5 – Frequently; 6 –Very frequently. 30Survey Questions: Time benefits• “Please provide an estimate for the typical time saved per month as a result of information, or other resources received from each person.”• The respondents assessed the time benefits for each other departmental employee based on a 5 – point scale, including: 0 – I do not know this person/I have never met this person; 1 – No time saved; 2 – 1-3 Hours per month; 3 – 4-8 Hours per month; 4 – 9-12 hours per month; to 5 – more than 13 hours per month.• We used the minimum amount of hours during our analysis31Survey questions: Individual attributes• Hierarchical levels were determined from a scale of one to five (1 –individual contributor/team member; 2 – supervisor/team leader; 3 –project manager/program manager; 4 – manager/BU [business unit] manager; 5 – director). • For tenure, individuals mentioned the number of years they have spent with the organization. • Finally, the department provided the data on gender.32Granovetter initial concept of weak ties (Granovetter 1973)33Louivan Modularity/BlockmodelTable 1. Blockmodel of KBSKBS A B C D NA 66% 30% 21% 22%27B 34% 97% 44% 46%29C 18% 32% 82% 60%36D 23% 45% 65% 91%5034

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.52660.1-0076438/manifest

Comment

Related Items