Executive Programme in Management Studies – Strategy Track
Amsterdam Business School
Master Thesis
Exploring the relationship between
Transactive Memory Systems and Team Performance:
the possible influence of Social Media and Enterprise Social Media
Author: Patrick Devick
Student number: 10317228
Date: 28-01-2015
Version Final
Preface
The Master Thesis has been written and the journey to achieve my Master title has come to an end. I started this journey to expand my knowledge and although I’m done studying, I’m not finished learning. I can honestly say that the journey was interesting, challenging, motivating, energizing and fun. However the journey was also exhausting, especially in combination with a full-time job and having a family with two young children.
I thank my supervisor Prof. Dr. J. Strikwerda for his guidance and valuable input in the process of writing this Master Thesis. Our discussions helped me to persevere and finish the thesis.
At last I would like to express my gratitude to who motivated and supported me during the whole journey and. Without the support of the most important persons to me, I could not finish this journey successfully. Thank you Denise, a million times and more, you also brought a lot of offers, you managed our family while I was studying. I also would like to thank Isa, my daughter of 5 years old, who accompanied me regularly when I was studying. I enjoyed it when you sat next to me and talking, drawing and mimicking me my while I was studying. Thank you Luca, my son of 2 years old, for your interruptions which forced me to heave some breaks and watch cartoons with you on YouTube.
Now it’s time to harvest from what I have learned and enjoy the time, that has become available again, with my family.
Patrick Devick
Table of Contents
Abstract ... 3
1. Introduction ... 4
2. Research Model and Hypotheses... 7
2.1. Teams and Performance ... 7
2.2. Transactive Memory Systems ... 12
2.3. The role of Social Media ... 18
2.4. The role of Enterprise Social Media ... 22
3. Methodology ... 27
3.1. Sample ... 27
3.2. Procedure ... 28
3.3. Measures ... 30
3.4. Analysis strategy ... 31
4. Data Analysis and Results ... 32
5. Discussion ... 36 5.1. Limitations... 37 5.2. Managerial implication ... 39 6. Conclusion ... 40 References ... 43 Appendices ... 48
Appendix 1 – Transactive Memory System ... 49
Appendix 2 – Team Performance ... 50
Appendix 3 - Social Media for work ... 51
Abstract
Social media technologies emerged amazingly fast and is commonly used for external-faced communication by organizations. Nowadays it is also increasingly used at the workplace for internal-faced communication and collaboration in organizations and teams. This research expands the current knowledge about the effects of using Social Media and Enterprise Social Media within teams. It explores the moderating effects on the relation between a Transactive Memory System of a team and the team’s performance when Social Media and Enterprise Social Media is used. A quantitative research is performed using a self-administered online survey and data is collected from employees from one organization which is located in the Netherlands. Using 89 valid responses this research identifies and confirm the positive relation between team’s Transactive Memory Systems and the team’s performance which implies that a team may perform better when having a stronger Transactive Memory System in place. The expected moderating effect of using Social Media on the relationship has not been confirmed. However the moderating effect of using Enterprise Social Media has been identified and confirmed. A positive moderating effects on the relation between a team’s Transactive Memory System and the team’s performance has been found. This research fills a gap in the research about using social media technologies for work within organizations. This area is relative new and few empirical research exists. Also it adds to the current knowledge about using technologies facilitating the development and functioning of Transactive Memory Systems. This research highlights that Enterprise Social Media has beneficial effects within teams and that teams and managers must be aware of the beneficial effects of using Enterprise Social Media for the information searches and exchanges within teams. This research has limitations and indicate a need for future research.
Keywords: Transactive Memory System, Enterprise Social Media, Social Media, Team
1.
Introduction
Organizations rely heavily on knowledge from multiple domains and when combined, enables them to produce new products and services and to improve their own processes. Knowledge is a strategically important resource and enables organizations to achieve competitive advantage (Grant, 1996a). Competitive advantage can be achieved when a resource (e.g. materials, knowledge) is socially complex and difficult to imitate (Collis & Montgomery, 2008). A key challenge for organizations is connecting and sharing knowledge, which is distributed throughout the organization. Therefore people need to be linked to each other and to knowledge content (Fulk & Yuan, 2013). Identifying and leveraging knowledge with the highest efficacy is crucial for firms and the ability to search, integrate, share and create new knowledge, is important for establishing and sustaining competitive advantage in today's rapidly changing business environments (Teece, Pisano, & Shuen, 1997). In the learning economy of today it even goes a step further, organizations are not only consumers of knowledge but also producers of new knowledge which is offered as a service (Nonaka & Toyama, 2005). Speed, complexity, and the increasing volumes of information needed in organizations is beyond limitations of individuals. However this limitation can be reduced in two ways: (1) by increasing the amount of people and create collaborating teams to complete tasks, and (2) by increasing capacity of information processing using information systems (Yuan, Fulk, & Monge, 2007).
The need for team work in organizations is due to increase of customers demanding new products with higher quality at faster rates, and preferably at lower cost (Levi & Slem, 1995). In contemporary organizations groups or teams are the basic building blocks of contemporary organizations. Teams possesses more expertise, knowledge and information (Moreland, 1999) and are able to process and achieve more results than is possible by any individual. A high performing team need access to each other’s knowledge as efficient as possible and a lack of
necessary knowledge, skills, abilities and resources results in ineffective teams (Kozlowski & Ilgen, 2006). Transactive Memory Systems facilitates high performing teams by providing information about who knows what and who knows who. It is used as a device to reduce knowledge transfers, provide faster access to knowledge from other members, reduce the cognitive load of individuals and allows members to specialize in one or more domains (Schmickl & Kieser, 2008). When integrating knowledge in an organization, the transfer of knowledge is not an efficient approach, particularly if the knowledge is difficult, expensive, or time consuming to codify (Nonaka, 1994). The key to efficiency is achieving effective integration while minimizing knowledge transfer through cross-learning by organizational members which interact with each other (Grant, 1996b).
Information technologies (IT) have become more powerful, networked, wireless and mobile, impacting every facet of our lives and work. It has increased the capacity of processing and storing of information. For example knowledge management systems, wiki’s, portals, blogs etc. provides more and faster access to information without physical logistical issues. New ways of communications are provided, due technological innovations like Web 2.0, allowing for more spontaneous, knowledge-based collaboration and sharing of personal and professional information in a digital environment (Hughes, Rowe, Batey, & Lee, 2012; McAfee, 2006). Those new digital environments are termed Social Media and have become one of the most dominant outlets for social interaction and information sharing using computers and other networked devices (Hughes et al., 2012). Social network literature indicates that informal networks facilitate the exchange of knowledge, something organizations increasingly seek to improve managing, building and leveraging knowledge (Jarrahi, 2011). Organizations responded by using Social Media to make more efficient use of their greatest organizational assets which is the knowledge and expertise of the members of the organization (Brzozowski, 2009). Enterprise Social Media emerged which mimics the features of Social Media while
mitigating privacy and security issues concerning usage in organizational context. The distinct feature is that Enterprise Social Media allows control about who has access to the platform and enables organizations to manage privacy and security by blocking out individuals who are not a member or affiliated to the organization.
Previous research indicates that Transactive Memory Systems (TMS) are positively related with the performance of teams (Jackson & Klobas, 2008b; Lewis, 2004) and that information technologies may support Transactive Memory Systems (S. Y. Choi, Lee, & Yoo, 2010; Ren & Argote, 2011). Research in Information Systems recently started exploring Social Media, by describing how this might affect organizational performance (Leonardi, Huysman, & Steinfield, 2013). Although Social media, today, is widely used in organizations its role is not always well understood and knowledge about the benefits, risks and limitations of using Social Media is limited (Storey, Treude, van Deursen, & Cheng, 2010). Pentland (2014) suggests that Social Media may impair innovative thinking and emergence of new ideas in organizations because of the echo chamber effect. More research is needed to extend models of online information seeking to encompass other important theoretical models such as transactive memory theory within and across social networks (Cho & Lee, 2008). Research on social media has focused mainly on: privacy issues, self-presentation, network analysis and social capital benefits. A few research focused on the benefits of social media used inside organizations exists (J. DiMicco et al., 2008; Leonardi et al., 2013).
In the following section of the thesis hypotheses are developed drawing on the literature about team performance, transactive memory and social media. Then the design of the research followed by the findings are described. At the end of this paper the theoretical and practical implications this research are discussed.
2.
Research Model and Hypotheses
This research is designed to answer the question ‘Does using Social Media and Enterprise Social Media influence the relation between Transactive Memory Systems and the performance of teams?’. This research seeks for an answer to this question and extends the current knowledge about supporting Transactive Memory Systems with information technologies and using Social Media and Enterprise Social Media for in organizational context.
Figure 1 Research Model
2.1.
Teams and Performance
We are part of a team in almost every facet of our lives, for example: when playing music a band, playing a soccer match, and when we are working. Most of the work in contemporary organizations is performed by teams, because organizational activities exceeds the capacity of individuals due to advancement in technology, competing goals (West, 2012), task complexity tasks (Salas, Cooke, & Rosen, 2008), speed, volume and complexity of information and communication flows (Yuan et al., 2007). Teamwork is necessary for all types of organizations, including non-profit organizations (Pfaff & Huddleston, 2003) and governments (Salas et al., 2008). The advantage of teamwork is significant productivity growth,
Transactive Memory System Social Media Team Performance Enterprise Social Media
which exceeds the performance of an individual, when the work requires a broader scope in terms of knowledge, judgement and opinion (Kyzlinková, Dokulilová, & Kroupa, 2007).
Multiple definitions of teams exists in previous research having major similarities and minor differences (Cohen & Bailey, 1997; Kozlowski & Ilgen, 2006; Mathieu, Maynard, Rapp, & Gilson, 2008; West, 2012). The definition of teams used in this research is: (a) two or more individuals; who (b) interact socially (face-to-face or, increasingly, virtually); (c) share one or more common goals; (d) are brought together to perform organizationally relevant tasks; (e) exhibit interdependences with respect to workflow, goals, and outcomes; (f) have different roles and responsibilities; and (g) are embedded in an organizational context that sets boundaries, constraints the team, and influences linkages to the broader system context (Kozlowski & Ilgen, 2006). Interdependency and shared responsibility to achieve common goals are important ingredients which creates the need for teams (Sundstrom, McIntyre, Halfhill, & Richards, 2000). Some scholars make a distinction between groups and work teams (Kozlowski & Bell, 2003) in that way that groups don’t have shared goals and each member is responsible for their own individual contribution with little interaction or dependency with other members (Cook, 2009; Katzenbach & Smith, 2005). This research makes no distinction between the term ”groups” and “teams”, and although the term “team” may be used more frequently they have in this research the same meaning,
Within an organization multiple teams exists and can be categorized into one of the following types: strategic teams, productions teams, service teams, project or development teams and action or performing teams (Sundstrom, De Meuse, & Futrell, 1990; West, 2012). Largely overlapping typologies for teams are used by other sources, and differ mostly in number of types (Cohen & Bailey, 1997; Katzenbach & Smith, 2005). Teams may be distinguished from each other based on multiple dimensions like: purpose or mission; lifetime; degree of autonomy; degree of influence, authority structure, physical presence and reliance on
technologies (Greenberg, 2010; Kozlowski & Bell, 2003; West, 2012). This research is concerning service, project and development teams, which are providing, developing and managing information technology based products and services. Cohen (1997) defines those teams respectively as project and service teams while Katzenbach (2005) identifies them as teams that make or do things and are responsible for doing the manufacturing, development, operations, marketing, sales, service, and other value adding activities of a business.
Many past studies use different criteria for the evaluation of the output of teams like: performance, effectiveness, productivity, success and others. Some define performance as a component of effectiveness while others equate performance with effectiveness (Sundstrom et al., 2000). The effectiveness of a team encompasses not only the tasks performed by a team but also how teams are interacting and executing the team process to achieve the required result (Cohen & Bailey, 1997; Salas, Sims, & Burke, 2005). This result of teams can be categorized in: (1) performance assessed by relevant others external to the team based on pre-set quantity or quality including efficiency, productivity, response time, quality, innovation, customer satisfaction; (2) behaviour judged by external stakeholders, like customers and management including communication, turnover and participation; and (3) attitudes judged by the team members, like personal satisfaction, commitment, willingness to remain in the team, cohesiveness and trust in management (Cohen & Bailey, 1997; Kohn & O'Connell, 2007; Mathieu et al., 2008). This effectiveness of team is depending on more than just bringing knowledge, expertise and skills of individuals into a group. Also the behaviour and attitude of teams are important factors of the function which determines effectiveness of teams. A high performing team may still score low on attitude and behaviour, hindering a team to perform at optimum level in the long term. Motivation and willingness of members to share their expertise and knowledge and the willingness to collaborate with others is affected by behaviour and
attitudes (Kohn & O'Connell, 2007). The potential effectiveness of a team is not be exploited to it is fullest when one or more of the categories will be ignored (West, 2012).
Team’s operate in a variety of organizational settings and no one fits all solution exists which guarantees highly effective teamwork (Katzenbach & Smith, 2005). Many determinants on organizational, team and individual level are enabling or constraining the effectiveness of teams (Kozlowski & Ilgen, 2006; West, 2012). Teams are affected by the organizational context in which they exist, such as training, information systems, rewarding systems and external environment factors, such as industry, markets, competitors (Sundstrom et al., 2000; West, 2012). The composition of a team has a powerful influence on the effectiveness of teams (Kozlowski & Bell, 2003). They are composed of individuals with different demographics (e.g., age, race, level of education) but also personalities such as their personal values and attitudes (Mathieu et al., 2008). Teams may also differ in size and geographical dispersion etc. which affects cohesiveness and efforts needed to collaborate (Katzenbach & Smith, 2005). Next to the individual capabilities such as knowledge, skills and competencies are important, but as important are capabilities needed to work in teams like: communication, coordination, managing relationships, goal setting etc. (Katzenbach & Smith, 2005; Kozlowski & Ilgen, 2006).
Previous research proposed many conceptualizations and classifications of frameworks regarding teamwork, lacking in consensus. Salas (2005) review of this body of research proposed a framework including the components that most heavily affect team performance and are found in almost all teamwork taxonomies. The team process consists of five core components: (1) leadership such as facilitating the team, setting clear goals and expectations, assign tasks, develop team knowledge and skills ; (2) mutual performance monitoring, to ensure that work is on track and mistakes are detected of prevented; (3) backup behaviour, providing feedback, coaching and assisting the members of the team; (4) adaptability, the ability to adjust
the internal processes to adapt to a changing environment or conditions and (5) team orientation, the preference to work together and enhancing individual performance by accepting coordination, evaluation and utilization of feedback from the team. To meld together the values of the core components, supporting and coordinating mechanisms are used: (1) development of shared mental models allowing shared awareness of task, expertise in teams(2) mutual trust, the shared belief that work is done by other members and (3) closed loop communication, exchanging information. The supporting components ensure that the inputs are incorporated, updated and information is distributed throughout the team. (Salas et al., 2005). Team inputs on different levels determines how efficient the teamwork is performed and are affecting indirectly via the components of teamwork affects the effectiveness of a team (Cohen & Bailey, 1997).
A successful team needs: (1) a clear direction and goal which justifies the reason for existence of a team and allows members to prioritize and make trade-offs when deciding if a course of action is sensible (Locke & Latham, 2002); (2) a balanced composition of members fitting in a team based on their roles and personalities (Belbin & Belbin Associates, 2009); (3) awareness of interdependencies between the tasks performed by the members (4) access to necessary knowledge, skills, abilities or resources to resolve team tasks (Kozlowski & Ilgen, 2006) (5) sharing of information to organize, coordinate and monitor progress of the activities towards the goals, and combine expertise and knowledge when executing team tasks; (6) members who committed to act as a collective and are not biased to their own individual goals (Salas et al., 2005); (7) a motivational system in place which is felt as fair in relation to the effort of the members and outputs of the team.
The focus in this research is on the factor knowledge, information sharing in teams and accessibility to information and knowledge, assuming that all of the other factors like behaviours and attitudes are adequately organized and in place within the teams. Is the
performance affected when the factors are facilitated or supported by: Transactive Memory Systems, Social Media and Enterprise Social Media?
2.2.
Transactive Memory Systems
The construct Transactive Memory Systems (TMS) was introduced by Wegner et al (1985) as a revision of the theory on group mind, which was one of the most influential theories in group behaviour. It conceptualizes how intimate couples develop a cognitive division of labour for encoding, storing and retrieving information from different cognitive domains to accomplish collective goals and as individuals are depending on each other for acquiring, remembering and generating knowledge (Wegner et al., 1985; Wegner, 1987). Wegner et al (1985) observed that couples in close relationship treated each other as external memory aids, thus enhancing the memory capacity beyond that of the individual. The concept of TMS is not unique to intimate dyadic, and may develop in groups and in organizations as well (Jackson & Klobas, 2008a; Moreland, 1999; Moreland & Argote, 2003; Nevo & Wand, 2005; Peltokorpi, 2012). The study considers only Transactive Memory Systems on the level of teams.
The concept of a TMS has similarities with shared mental models but its emphasis is on facilitating differentiation of knowledge in a team, allowing members to specialize in learning, remembering and sharing different knowledge. In contrast with other shared mental models (e.g., team mental models, shared understanding and cross understanding) which exhibit integration of information, allowing teams to realize a shared view about the same information such as team goals, tasks and strategies, member’s beliefs and preferences (Ren & Argote, 2011). A Transactive Memory System may thus act as a device facilitating the support and coordinating mechanisms needed in a team, ensuring that the teams information can be incorporated, updated and distributed throughout a team without the need to transfer all information and deep knowledge (Salas et al., 2005).
Transactive Memory Systems are defined by two components: (1) a collection of individual memories when combined together form a stock of knowledge contained by a group and (2) transactive processes that takes place between the individuals of a group to enable interpersonal awareness of each other’s knowledge (Nevo & Wand, 2005; Wegner, 1987). The memory in a Transactive Memory System is based on the distinction of internal and external memory. Wegner (1985) noted that the memory of individuals could be used as an external memory by others. The memory of an individual contains: (1) deep internal knowledge about one or more domains and (2) a map or directory with (meta)information, about who knows what, who knows who (Jackson & Klobas, 2008b). Although the external memory component of a TMS initially is based on using the memory of individuals as an aid, it does not necessarily need to be stored in the individuals memory. Storage of information and knowledge may be facilitated by usage of documents, books, articles, information technologies (e.g. portals, blogs, knowledge management systems) etc. (S. Y. Choi et al., 2010; Hollingshead & Brandon, 2003). Using external storages implicates that in addition to who knows who, who knows what, also where is what stored needs to be remembered as well. Using external non-human memories as storage offers benefits such as increase of availability of information without depending on the physical or virtual presence of members of the Transactive Memory System. Also non-human storage may overcome the cognitive limitations and unreliability of individual memories.
The individual memories of the members of a group needs to be interconnected to develop a Transactive Memory System. This interconnecting is based on communicating events – the transactive processes – across members of the group during the encoding, storing and retrieval of information and it is the other component that underlies the successful development and functioning of a Transactive Memory System (Wegner, 1995).
Transactive encoding (allocation of information) is a process where needs to be determined to who information is allocated and how it is stored (Wegner, 1987). The allocation
takes place based on the perception of who is responsible for remembering the information. Members specialize in areas based on their relative expertise, skills, experiences, formal assignments in the organizational structure (by a high status member), or negotiated agreements (Hollingshead, 2000). The perception of who has expertise may be influenced due former retrieved information from members of the groups TMS (Wegner, Erber, & Raymond, 1991). Due the division of cognitive labour there is less need to know everything by every member of the group ands them to specialize in one or more knowledge domains (Hollingshead, 2000).
Transactive storage is the process where information allocated to a member of a group, is stored and remembered within the TMS for later retrieval. The information already stored in the TMS may be updated to remove inconsistencies or complemented when new information becomes available. Also it may be modified to make the information communicable for the individual (Wegner et al., 1985). The member responsible for remembering the information for later usage stores this his memory, and the other members need to update their memories with a link to where the information can be found.
Transactive retrieval is the process of determining the location of required information and accessing the required knowledge from a group using one or more directories (Jackson & Klobas, 2008a). It occurs when knowledge or expertise is needed to accomplish a task where the individual has not the necessary knowledge or expertise himself. The individual has at least two and sometimes more directories to consider to determine the location of information (Wegner, 1995). The information retrieval may also occur in interaction with other members of the group to complement each other, via referral from one member to another member. The success of retrieval is depending on the former processes that encoded and stored the information (Wegner, 1995).
Past research showed that there are numerous factors influencing the development, maintenance and well-functioning of a Transactive Memory System. Cognitive based trust –
the beliefs about another’s ability and reliability to carry out their tasks and reliance on other team members knowledge – must be available between members in a team (Kanawattanachai & Yoo, 2007; Wegner et al., 1985). If members can rely on others ability to remember task-critical information they will develop different knowledge and specialize (Lewis, 2003). The functioning of a TMS is affected when groups become larger, because more location based information needs must be remembered by each members and increased communication is needed to keep the Transactive Memory System synchronized and accurate. Members become more responsive to each other and will exchange their information and knowledge when having shared goals and a perception of cognitive interdependency, (Hollingshead, 2001; Zhang, Hempel, Han, & Tjosvold, 2007). Information is allocated and retrieved with increased efficiency when members are aware of task interdependency and understand each other’s skills, also coordination of their knowledge and expertise to complete tasks is more efficient (Zhang et al., 2007). Familiarity between members increases the willingness to offer, discuss, and consider unique information, reducing ambiguities about how expertise is distributed between members of a TMS (Lewis, 2004). A TMS emerges faster when members have opportunities to observe each other and store cues about members expertise. For example training affords member to communicate and question, claim and evaluate members expertise (Ren & Argote, 2011). Previous research indicates a positive relation between communication and Transactive Memory Systems (Hollingshead, 1998; Hollingshead & Brandon, 2003; Lewis, 2004). Yuan et al. (2009) revealed that expertise retrieval was affected positively by awareness of expertise distribution facilitated by Transactive Memory Systems and social accessibility of the members.
Communication is a necessity to transfer and share information between members of a group and to allow the development and functioning of a Transactive Memory System (Wegner et al., 1985; Wegner, 1987). It is a crucial component in the encoding, storing and retrieval of
information in a TMS (Hollingshead & Brandon, 2003; Lewis, 2004) and it is the means by which individuals may learn and discover who has knowledge, expertise and experience of a particular domain (Hollingshead, 1998). Direct interpersonal communication (face-to-face) is typical but not required, it may take place indirectly in codified forms such as written manuals, job descriptions and procedures as well. (Hollingshead & Brandon, 2003). Communication may also be facilitated using technologies like phone, tablets and computer (Jackson & Klobas, 2008b; Nevo & Wand, 2005). Transactive Memory Systems can emerge without direct face-to-face communication using computer mediated communication, although its development may be more difficult and slower (Kanawattanachai & Yoo, 2007; Nevo & Wand, 2005). However some discrepancies exists regarding the significant effect of using technologies as alternative communication channels. Lewis (2004) found no significant impact on the functioning of a TMS when using computer mediated communication is used between members. Kanawattanachai & Yoo (2007) found significant impact on the functioning of a TMS using computer mediated communication in virtual teams.
Information technologies like information management systems, intranet, search engines and document repositories may be used as (1) an external memory aid supporting the memories of the individuals of a group to store knowledge and information about who knows what and who knows who, and (2) to support communication and collaboration in the transactive processes (S. Y. Choi et al., 2010; Nevo, Benbasat, & Wand, 2012). Using information systems offers specific benefits beyond direct access to members in a team like: (1) members needing or holding information may work independently from each other in their own pace; (2) information once stored in an information systems can be retrieved multiple times without burdening other members of a team; (3) reducing the cognitive load of teams members to remember information and (4) information can be made available to a wider public such as members of the organization who are not a member of the team (Yuan et al., 2007). However
it is challenge to keep information in information systems up to date and will be distrusted and left unused in the end if not up to date or accurate (DiMicco, Geyer, Millen, Dugan, & Brownholtz, 2009). Information technologies facilitating a Transactive Memory system may positively influence its development and functioning when groups become larger, because it reduces the cognitive load on individuals to remember where knowledge can be found (Nevo & Wand, 2005). The cognitive limits of individuals can be extended when using information technologies and improves speed and reach when searching for information or retrieving information (Yuan et al., 2007).
The outcome of a well-developed Transactive Memory System impact a team’s performance, behaviours and attitudes. The main benefits are: (1) quicker and more effective coordination of different areas of expertise to complete tasks, and (2) a reduction of the cognitive load for all members, while still having access to a greater amount of information and knowledge of higher quality (Hollingshead, 1998; Palazzolo, 2010; Robertson, Gockel, & Brauner, 2012). A well-developed Transactive Memory System enables a team to work more effectively, without wasting time when searching for needed knowledge, and therefore the performance of a team may increase, because information can be located in a more efficient and faster way (Lewis & Herndon, 2011; Robertson et al., 2012). A Transactive Memory System provides access to more and better information from different domains to group’s members than any of them can remember and access individually (Lewis, 2004; Moreland, 1999). Having more knowledge available as a group is insufficient, when not utilized in an efficient and effective way (Sung & Choi, 2012). Akgün et al. (2005) showed a positive influence on team learning, speed to market and new product success when a well-developed and functioning TMS exists in teams. Next to the beneficial effects of a group as whole, also beneficial effects for individual members exists. Being a member of a TMS allows them to specialize which eventually is beneficial to all members of a TMS (Wegner, 1987). Ho & Wong
(2009) showed that the performance of individual worker’s increases when they are part of a transactive system. It stimulates the individual to develop expertise recognition and allows the individual to locate information and knowledge with greater efficiency and effectiveness. It also allows a better usage of information available in the organization by the team. This leads to the following hypothesis:
Hypothesis 1: A well-developed Transactive Memory System is positively related to
team performance.
2.3.
The role of Social Media
Social Media is deeply integrated in our daily lives and changed the way we live and work (Turban, Bolloju, & Liang, 2011). Music is shared using Spotify, meetings are held using Google Hangout, opinions are sent into the world using Twitter, images and photos are shared using Instagram, professional networks are maintained using LinkedIn, Facebook is used to connect with friends and to maintain relationships, games (e.g., FIFA Soccer, Wordfeud) are played with others around the world (Baarsma & Steentjes, 2013). It became even more integrated with the availability of mobile devices and mobile internet, allowing an audience to connect, communicate, and interact with each other and their mutual friends almost everywhere and anytime (Correa, Hinsley, & De Zuniga, 2010).
The popularity of Social Media has created interest from organizations and enabled them new ways to communicate and connect with customers and colleagues (Leftheriotis & Giannakos, 2014). It is used for organizational communication with external parties such as customers, vendors and bigger audience (Leonardi et al., 2013). Facebook and Twitter are used for image branding, marketing activities and to discover customers ideas, needs and opinions about the products they offer to them. LinkedIn is used when scouting for new talent, or when new jobs are offered. Besides communicating with external parties, it is used also at the
workplace facilitating internal organizational communication, collaboration and sharing of knowledge (Cao, Vogel, Guo, Liu, & Gu, 2012). In an organizational context it allows for discussions, information sharing, getting advice and problem solving from members of an organization (Riemer, Scifleet, & Reddig, 2012). It may be used to search and explore for new corporate information (J. DiMicco et al., 2008). Although Social Media already is used in organizational context, less is known about how it can be used for work purposes and what the effects are within an organization (Jarrahi, 2011; Leftheriotis & Giannakos, 2014).
The term Social Media was coined when social networking sites like MySpace (in 2003) and Facebook (in 2004) emerged and became popular (Kaplan & Haenlein, 2010). Many attempts have been made to define Social Media which is difficult due continuous developments at an amazing fast pace (Treem & Leonardi, 2012). Kaplan (2010) defines Social Media as “a group of Internet-based applications that build on the ideological and technological foundations of Web 2.0, that allow the creation and exchange of User Generated Content”. Web 2.0 is the result of technological innovations allowing more interaction and collaboration on the internet. User Generated content are the various forms of media content that is publicly available and created by end-users (Kaplan & Haenlein, 2010). Solis (2013) created a visual map of the dominant and promising social media, grouped by how they’re used in everyday life of professionals and customers. This maps shows that the Social Media landscape is wide and evolving, where new players come and go. Kaplan & Heanlein (2010) created a classification scheme, based on a set of theories in the field of media research (social presence, media richness) and social processes (self-presentation, self-disclosure), which identifies six types of social media: (micro)blogs (e.g., Twitter, Blogger), social networking sites (e.g., Facebook, LinkedIn, Google+), content communities (e.g., Youtube, Slideshare, Soundcloud), collaborative projects (e.g., Wikipedia), virtual games worlds (e.g., World of Warcraft) and
virtual social worlds (e.g., Second Life). What they have al in common is: online presence, sharing information, and social interaction.
Social Media allows building stronger relationships and higher bonding which emerges from personal information exchange with connected users. Social Media function two types of communication: (1) task-related communication and (2) social communication. The relational use of social media leads to better communication and improvement of work performance (Cao et al., 2012). Research shows that who you know has a significant impact on what you come to know, as relationships are critical for obtaining information, solving problems and learning how to do your work (Cross, Parker, Prusak, & Borgatti, 2001). Being part of a social network may add to the willingness to share information and experiences with each other. Reconnecting to previous known people is made easier using Social Media, and when building a social network other people may be discovered sharing same interests or needed knowledge. Social Media affords a user to keep in touch and being aware of the status of others because they update their status which leads to greater understanding, intimacy and familiarity with others (Skeels & Grudin, 2009).
At work Social Media may enhance trust among employees which mediates the transfer of knowledge between employees (Cao et al., 2012). Trust fosters the willingness to share information more freely among team members (Salas et al., 2005). Collaboration is promoted when members of a team trust each other, and they will spending less time with protecting, checking and inspecting each other. Social media offers a new communication channel where explicit and implicit knowledge can be effectively transferred between employees (Cao et al., 2012).
Social Media affords to increase visibility and reach of information with less effort and allowing users to express their expertise, interests etc. to others in the world, or a member of a team or organization (Treem & Leonardi, 2012). This access to information about expertise and
interests may allow members of a Transactive Memory System to determine expertise of a person and allocate or retrieve information to this person without less effort. All interactions and shared information are often persisted on social media platforms, which allows availability of this information to others who could not participate in those interactions. This persistence of information may be utilized within the context of TMS when used as external memory aid. The persisted information and interactions are often connected to the profiles of users and enables answering the questions who know what or who knows who when searching for expertise.
Earlier research indicated that using Social Media has the potential to improve collaboration and knowledge sharing which affects the performance of the employees positively and is not a waste of time (Leftheriotis & Giannakos, 2014). Although Social Media is used in organizational context and has its benefits, it also creates challenges such as employees wasting time, possible leaking sensitive and propriety information, disclosure of personal information which discredit users and organizations (Leonardi et al., 2013; Levati, 2014; Turban et al., 2011). The openness of Social Media affords transparency allowing others a view into each other past actions and communications. Although this transparency may be a benefits, it is also a concern which can result in defensive self-presentation behaviour. This is especially the case when the information can be viewed by the managers of the contributors (Brzozowski, 2009; Leonardi et al., 2013).
As stated before Social Media simplifies and accelerate the ability to share information, stories, experiences and knowledge with each other. In addition it adds to the willingness to share information, and improve accessibility to information because people with same interest can be found and more personal information of users is made available through their user profiles. Both interpersonal and technological approaches might be used to improve the functioning of a TMS (Jackson & Klobas, 2008a). Therefore social media may also facilitate the functioning of a Transactive Memory System when supporting the transactive processes
and usage as an external memory component. Supporting socializing in teams influence the effectiveness of teams. This leads to the following hypothesis:
Hypothesis 2: Using Social Media for work will moderate the relation between
Transactive Memory System and Team Performance.
2.4.
The role of Enterprise Social Media
Social Media is increasingly implemented in organizations as a tool for communication and interaction among members of an organization (Leonardi et al., 2013). However it has an open character and, after registration, is publicly available to who has internet access. Therefore it is more difficult to control, manage en monitor the usage. As a member of an organization it is easier to leak information intentionally or unintentionally, when using Social Media (Leftheriotis & Giannakos, 2014). Tensions rise about what may be or may not be shared when using Social Media and being connected to colleagues and personal friends. Problems may rise when friending managers or others higher in the organizational hierarchy. User accounts from organizational members may be compromised may by hackers and revealing sensitive information (Skeels & Grudin, 2009). The origin and quality of published information is not always clear, when shared via Social Media, and may lead to legal risks when used without permission or the information qualitative insufficient, incomplete or false (Turban et al., 2011). The role and intension of employees using Social Media are distinct from the reasons they use it in their personal online space (DiMicco et al., 2009).
As an answer to those concerns of organizations they copied the features of Social Media and made them available behind firewalls within the boundaries of an organization, allowing more control on which users has access to it (Jarrahi, 2011). Examples are WaterCooler at HP or Beehive at IBM (Brzozowski, 2009; J. M. DiMicco, Millen, Geyer, Dugan, & Street, 2008). However it is not needed to host Social Media Platforms internally. Yammer and Facebook at
Work are Social Media platforms offered as a service to organizations without the need to host it internally. Those walled Social Media offer members, partners or other affiliates of an organization, access to the enterprise social network. Those platforms used for internal communication, collaboration and social interaction, only available to organizational members, is termed Enterprise Social Media (Leonardi et al., 2013). It is employed to encourage members to share both professional and personal information with each other within organizations (Leftheriotis & Giannakos, 2014; Wu, DiMicco, & Millen, 2010). It may serve as an important vehicle for the creation of social capital within organizations and acts as a key driver to build effective and efficient business creating competitive advantage (Turban et al., 2011). Members of an organization are less concerned about privacy and security issues when using Enterprise Social Media in contrast with Social Media (J. DiMicco et al., 2008).
Enterprise Social Media is defined by Leonardi et al. (2013) as “Web-based platforms that allow workers to (1) communicate messages with specific co-workers or broadcast messages to everyone in the organization; (2) explicitly indicate or implicitly reveal particular co-workers as communication partners; (3) post, edit, and sort text and files linked to themselves or others; and (4) view the messages, connections, text, and files communicated, posted, edited and sorted by anyone else in the organization at any time of their choosing.”
Many technologies for communicating are commonly implemented in organizations allowing workers to do the first three activities. But the distinction of Enterprise Social Media from traditional technologies is the last activity (Leonardi et al., 2013). It allows members of an organization to search and acquire knowledge from previous collaborations without participating at the moment the interaction took place. This persistence of this information may support Transactive Memory Systems. It then acts as an external memory which can be searched to seek for who know who or who knows what.
In an organizations, Enterprise Social Media Social allows to create and strengthen ties with colleagues and promotes knowledge sharing and locating of resources when searching who knows what or who knows who (Skeels & Grudin, 2009). Introductions to other members in the organization is made easier, because Enterprise Social Media provides information from the enterprise network like: personal information, interests, contributions to previous conversions, linkage to user’s which they have in common. Research has shown an increase of linkages between members, due the visibility of the networks of other members, which otherwise stayed hidden and potential benefits of the information is then left unused (Guy, Ronen, & Wilcox, 2009).
When using Enterprise Social Media more convenient and democratic communication may occur whereby users may derive more authority from their previous contributions to the network, than their hierarchical position provides (Stieglitz, Riemer, & Meske, 2014). It may support the functioning of Transactive Memory Systems to locate new information based on the published expertise and the reputation of an individual in the enterprise network provided by the Enterprise Social Media platform. Recognizing each other’s expertise is difficult, especially when a team is geographically dispersed or cultural differences in communication styles exists. Using computed mediated communication may also overcome cultural differences in communications, it allows individuals to share information with less hesitation (Bazarova & Yuan, 2013). However the benefits of persistence and transparency may also be perceived as negative by individuals, because others (e.g. managers, team leaders) are offered the possibilities to search the previous interactions.
Digitally supported social networks can serve as a powerful lens into the structure of work relations and the flow of information in work organizations (Jarrahi & Sawyer, 2012). It offers organizations possibilities to reveal the dynamics and structure of an organization and provide this information to the users of the social technologies via search and other exploring
features in a digital way. The use of social media technology in an organization allows to generate rich network data without having to explicitly ask each employee about their personal connections (Smith, Hansen, & Gleave, 2009). The network data can may be shown proactively (e.g. lasts contributors including their topics, hot topics, other people sharing same interests) to users based on their interest or latest interactions and provide potential linkages to others or new knowledge (Brzozowski, 2009).
Organizations do already offer technologies for searching employees and information about employees like intranet services or CV systems. However they provide relative little timely or personal information about someone (to know him better). It is a challenge for organization to motivate people update their information in such services to make it up-to-date, accurate en complete. Enterprise Social Media may help overcome this problem because it is believed that members of a social networks have a desire to keep up and stay in touch with the other members in a community (DiMicco et al., 2009).
Communication and trust are crucial components for the development and functioning of a Transactive Memory System and Social Media facilitates work performance by stimulating trust among employees and offering alternative communication channels (next to face-to-face, phone, email etc.) where explicit and implicit knowledge can be transferred (Cao et al., 2012). Because Social Media and Enterprise Social media share the same principles, it is expected that Enterprise Social Media provide the same benefits to the development and functioning of a Transactive Memory System in a team.
Enterprise Social Media offers all the benefits of Social Media while mitigating some of the issues such as privacy and security. Because it is used within the boundaries of an organization it may offer users a more convenient way to communicate work related information without thinking about issues like leaking of information. It offers availability of persisted information and accumulation of knowledge from past collaboration by other
members of teams or others in the organization. This allows to search and retrieve information on a later moment. Enterprise Social Media facilitate to get more efficient and faster access to information. It makes the process of finding an individual with the needed knowledge more efficient, because it is independent of time and place and physical attention of an individual. Therefore the following hypothesis is stated:
Hypothesis 3: Using Enterprise Social Media will moderate the relation between
3.
Methodology
An explanatory study is performed using the deductive approach which means: (1) defining a research question after reviewing relevant literature; (2) operationalizing the research question using measures for the constructs (3) collecting data using a survey and (4) analysis of the collected data using statistical methods to determine if support exists for the hypothesises defined in this study (Saunders & Lewis, 2012).
3.1.
Sample
Data is collected from employees from one organization. The organization is an IT service provider operating in the Netherlands employing 350 people. They actively share knowledge with each other and with other colleagues in the industry, write blogs, articles and whitepapers and publish those using different forms of Social Media. Information and knowledge is exchanged within the organization using technologies such as portals, wiki’s private to the organization. The majority of the employees are working in project or work teams and depend on each other’s knowledge and expertise to fulfil their tasks and the goals of the team. Teams are composed of employees of the organization, but frequently mixed with individuals from other organizations (e.g., clients organization, or colleagues from the industry) as well. The teams are located within the organization, but also offsite at the location of the customers. The main responsibility of the teams is servicing, creating, improving new or existing customer’s business processes using information technologies. The majority of the tasks in the teams are analysis, design, code, test, maintenance and managing of information systems. Effects of Transactive Memory Systems should be prominent within those teams, which exist of members whose output rest on member’s knowledge and expertise, and used to create new knowledge in the form of services or solutions (Lewis, 2004). The findings when using real team in real organizations, as the population to same from, are more easy to generalize to organizational context and can act as a guide to management practices compared
to findings from teams in experimental or laboratory settings (Cohen & Bailey, 1997; Mathieu et al., 2008; Sundstrom et al., 2000). A minimum number of 183 respondents is needed when desiring a confidence level of 95% with a confidence interval of 5% for a population of 350 individuals. (Saunders & Lewis, 2012). Because all employees of the organization are included in the sample, the risk of sampling bias is mitigated. To increase the possible response also snowball sampling is used to reach individuals who are members of the same team of the employees of the targeted organization, but cannot be reached using email.
3.2.
Procedure
An online self-administered survey is conducted on a voluntary base from the employees. It allows to collect data, about the same thing, from a large number of people in a cost-effective manner, who work in a geographical dispersed area, and in an electronic format easier to import in statistical tools like SPSS. Negative effects of an interviewer are removed using a self-administered survey (Saunders & Lewis, 2012). An online self-administered survey gives the respondent more control about the pace, where and when to complete the survey (Bryman, 2012). However absence of reviewer to explain questions when not clear to respondents, limitation on number of questions and no control over who is completing the survey may be possible issues in the case of self-administered surveys (Saunders & Lewis, 2012). The survey is tested by piloting with a representation of the population to: improve the validity and reliability of the survey, measure the time needed to complete, and confirm that the respondents understand the survey. Unless otherwise specified all items of the measures used in the survey are close-ended to reduce interviewer bias and are measured using a five-point, Likert scale, ranging from 1 (Strongly disagree) to 5 (Strongly agree). To control for valid input some variables were included like size of a team. The survey included an introduction which explaining the meaning of Social Media and Enterprise Social Media are, and information about the duration of the survey and confidentiality was included in the survey to increase its
reliability and validity. Although the organization is located in the Netherlands, the survey is designed in English. There was no need to translate the items, because the majority of the targeted audience is highly educated and uses English on regular basis.
Direct email is used to reach all employees of the organization inviting to complete the survey. In total 350 emails are sent explaining the purpose of the survey, duration, confidentiality and included a link to the survey. The email also encouraged the respondents to forward the invitation to individuals who are members of their teams but not an employee of the targeted organization to increase the response. In the two weeks after, reminder are sent each week encouraging respondents to complete the survey. A total of 350 invitations are distributed and 127 responses are returned, representing a response rate of 36.30%. The minimum is not reached due time constraints and maybe because the targeted respondents are working fulltime at projects for customers having additional time for commuting between home and work. Increasing the frequency of reminders would probably perceived as annoying without any positive effect on response. Using snowball sampling also had no effect because no respondents from other organizations completed the survey.
All responses are discarded which were not completed as designed. Responses not meeting the size criteria for teams also are discarded. The concept of TMS is not applicable to teams consisting of only one person, at least two members are required (Wegner et al., 1985). Responses which indicated that size of the team was larger than 25 have been discarded. Although theoretically a large number of 50 individuals can be become a team, large groups tend to break into sub teams, because large numbers of people have difficulties interacting constructively as a group (Katzenbach & Smith, 2005). After discarding al invalid responses, the valid sample includes 89 responses (N = 89). The mean age of the respondents is 35.89 years (SD = 8.60) and six of the respondents was female (6.74%). Team have an average size of 8.47 (SD = 5.33).
3.3.
Measures
The presence of a Transactive Memory System in a team is measured using a scale developed by Robertson et al. (2012). The scale has 20 items using 5 dimensions: integration, differentiation, cognitive interdependence, metaknowledge and transactivity. A 5-point Likert scale ranging from 1 (Strongly disagree) to 5 (Strongly agree) is used to indicate the extent of agreement with each item of the scale. A composite scale measuring the total score of a Transactive Memory Systems is calculated by averaging all items of the scale. A low score represents a weak Transactive Memory System and a high score a strong Transactive Memory System. The items of this measure are all exhibited in Appendix 1.
Measuring team performance for teams consisting mostly of knowledge workers is difficult, due the fact that no generalized method exists and knowledge work is intangible (Ramírez & Nembhard, 2004). No standard objective measures like costs, timelines etc. were available to measure the performance of teams objectively. A measure for this study is constructed based on a measure from Roberson et al (2012) combined with items from Staples & Webster (2008) which items match some of the dimensions defining productivity of knowledge workers (Ramírez & Nembhard, 2004) and are appropriate in the context of teams with knowledge workers (B. Choi & Lee, 2003). However objectivity is increased by including multiple items from another measure, it still measures performance based on the perceptions of the individuals. All items of this measure are included in Appendix 2.
A measure from Leftheriotis & Giannakos (2014) is adapted to measure the level of using Social Media for work and the intention for using it at work. Three items are used focusing on the usage and frequency of using Social Media for work, and also items are used to measure the intention of using Social Media at work. Those intentions measure the purpose of using Social Media and matches the usages of a TMS when locating and sharing knowledge. A 5-point Likert-type scale from 1 (Strongly disagree) to 5 (Strongly agree) is used to indicate the
extent of agreement with each item of the scale regarding usage and intention. The item asking for frequency of using Social Media a 5-point Likert scale ranging from 1 (Never) to 5 (All the time) is used. A composite scale measuring the total score of using Social Media for work is calculated by averaging all items. A low score represents less usage of Social Media for work and a high score a strong usage of Social Media for work. All the items of this measure are listed in Appendix 3.
The measure for Enterprise Social Media usage for work is based on the same scale from Leftheriotis & Giannakos (2014). This scale is the closest available of which the wording is changed from Social Media into Enterprise Social Media. It measures the usage, frequency and also the intentions of using Enterprise Social Media. All items of this measure are listed in Appendix 4.
3.4.
Analysis strategy
Reliability of the constructs is assessed by checking the Cronbach alphas and descriptive statistics are generated for each construct. The assumptions for multiple linear regression are tested and used to test the hypothesises. SPPS is used to execute the statistical procedures and analysis. The descriptive statistics and bivariate correlation tests are performed with SPSS. Multiple linear regression tests are performed using the PROCESS macro for SPSS (A. Hayes, 2012). This extension performs the majority of the calculations (e.g. centering, data for plotting) which otherwise have to be performed manually when using SPSS for regressions. When interaction terms found to be significant, the results will be plotted to help understand the nature of the interaction.
4.
Data Analysis and Results
Internal consistencies of the variables are tested by calculating the Cronbach alphas, which all exceeds the minimum value of .70 this indicating an adequate reliability. The Cronbach alpha of Transactive Memory System could slightly be improved to .787 when deleting one of the items. However no items are deleted because all items are used by Robertson et al. (2012) in earlier research to measure the level of Transactive Memory System. Table 1 present descriptive statistics, including the internal consistencies and Pearson correlation among all variables.
Table 1 Descriptives, correlations and reliability (N = 89)
Variable M D 1 2 3 4
Transactive Memory System 3.69 .38 (.779)
Team Performance 3.34 .55 .454** (.813)
Enterprise Social Media for work 2.89 .82 .095 .026 (.858)
Social Media for work 2.57 .80 -.003 .036 .592** (.863)
Note: Cronbach α for each construct is placed on the diagonal between parentheses . **. Correlation is significant at the 0.01 level (2-tailed).
On a scale from 1 to 5 respondents rated their average Transactive Memory System as 3.69 (SD = .38) which indicates that the concept exists within the teams. Also respondents rate the performance of their teams with a value in the upper half of a scale from 1-5 (M =3.34, SD = .55) on a scale from 1 to 5. The Pearson product-moment coefficients was computed to assess the relationship between Transactive Memory System and Team Performance. This coefficient indicates moderate positive correlation between the two variables which is significant (r=.454,
n = 89, p <.001). Increases of TMS were correlated with increases in performance of teams.
Based on this significant correlation there is support for Hypothesis 1.
Respondents rated Social Media for work with an average of 2.57 (SD = .80) and Enterprise Social Media with an average of 2.89 (SD = .82). Enterprise Social Media for work scores higher then Social Media which may indicate that it is used just a little bit more for work
then Social Media is used for work. A moderate correlation between Social Media use for work and using Enterprise Social Media for work exists (r =. 592, n= 89, p < .001). This may suggest that when Social Media use for work increases, also the use of Enterprise Social Media for work increases. This correlation may be explained by the fact that Social Media and Enterprise Social Media share the same principles and only distinct from each other in who has access, the sort of information and to whom can be connected. Another reason may be that when the benefits are experienced on one platform, it is expected that the same benefits may be experienced on the other platform.
In order to test the moderating hypotheses using multiple regression, some assumptions needs to be confirmed. The assumptions are: 1) normal distribution; 2) linear relationship between independent and dependent variables; 3) independence and 4) homogeneity of variance. Violation one or more assumptions result in loss of the power of the statistical test and significance of effect size(s) may be over- or under-estimated (A. F. Hayes, 2013; Osborne & Waters, 2002). Normal distribution is tested by calculating the values for skewness and kurtosis. None of the used variables indicated a skewness or kurtosis with values less than -1.96 or greater than +1.96 (Field, 2009), and therefore the variables can be assumed normal distributed. Linearity is tested by generating a scatter plot containing the standardized residuals as a function of standardized predicted values (Osborne & Waters, 2002) which indicated linear relation. Independency of the residuals is tested by calculating the Durbin-Watson statistic which value us 1.736, which is close to 2 meaning that the residuals are uncorrelated and independency may be assumed (Field, 2009 p 221). Homogeneity of variance is tested by plotting an overview of the regression standardized residuals and regression standardized predicted values. The plot showed a fairly and evenly spread out of the residuals indicating that homoscedasticity may be assumed (Osborne & Waters, 2002).
A multiple regression analysis was conducted to see if there are moderating effects of Social Media for work and Enterprise Social Media on the relationship between Transactive Memory System and Team performance. Table 2 show the result of the regression test. The independent variables are mean-centered to standardize the regression coefficients (A. F. Hayes, 2013).
Table 2 Enterprise Social Media and Social Media for work as moderators of the TMS-Team Performance association (N = 89)
b SE b t p
Intercept i1 3.32 .05 64.35 p < .001 Transactive Memory System (TMS) b1 .64 .14 4.59 p < .001 Enterprise Social Media for work (ESMW) b2 -.03 .08 -.44 .66 Social Media for work (SMW) b3 .04 .08 .49 .62
TMS × ESMW b4 .43 .19 2.24 .03
TMS × SMW b5 -.15 .17 -.86 .39
R2=.26, F(5, 83) = 5.70, p < .001
TMS, ESMW and SMW were mean-centered prior analysis
The multiple regression analysis results in a model where the independent variables explain a significant amount of the variance in the team performance (F(5,83) = 5.70, p < .001,
R2=.26). The variable Transactive Memory System has a significant impact on Team Performance (t = 4.59, p < .001) which was expected because the Pearson correlation between Transactive Memory System and Team Performance was significant. The analysis shows that using Social Media for work has no significant interaction effect (t = -.86, p = .39) on the association between Transactive Memory System and Team Performance. There is no support for Hypothesis 2. However using Enterprise Social Media is significant (t = 2.24, p = .03) indicating that Enterprise Social Media has a moderating effect on the association between Transactive Memory System and Team Performance. Therefore Hypothesis 3 is supported.
Because the moderating effect of Social Media is not significant, another regression is conducted with only the significant variable (Enterprise Social Media). These results are shown in table 3.
Table 3 Enterprise Social Media as moderator of the TMS-Team Performance association (N = 89)
b SE b t p
Intercept i1 3.33 .05 64.90 p < .001 Transactive Memory System (TMS) b1 .61 .14 4.51 p < .001 Enterprise Social Media for work (ESMW) b2 -.01 .06 -.22 .83
TMS × ESMW b3 .35 .16 2.15 .03
R2=.25, F(3, 85) = 9.29, p < .001
TMS, ESMW were mean-centered prior analysis
The adjusted model explains a significant amount of the variance in the team performance (F(3,85) = 9.29, p < .001, R2=.25). The interaction of Enterprise Social Media with TMS on
Team performance is significant (t = 2.15, p = .03) indicating that Enterprise Social Media has a moderating effect on the association between Transactive Memory System and Team
Performance. The effect of the interaction is plotted in figure 2.
Figure 2 Interaction effect of Enterprise Social Media on the TMS-Team performance association
The moderation of using Enterprise Social Media is depicted in figure 2 showing that the interacting effect is positive in all cases for the different levels of using Enterprise Social Media. The slope of the moderation increases when the intensity of using Enterprise Social Media increased. The strongest effect of the moderation of Enterprise Social Media is when it is highly used within teams to search and exchange information.
