F ACTIONAL FAULTLINES AND T EAM LEARNING : T HE EFFECTS OF A MERGER
René Poel
University of Groningen, Groningen
MscBA - Change Management, Faculty of Economics and Business
Student number: s1796003
Verlengde Visserstraat 9a 9718 JA Groningen
06-11518170 r.poel@student.rug.nl
1st supervisor: Dr. J. Rupert 2nd supervisor: Dr. M.A.G. van Offenbeek
Date: 29-10-2013
Abstract
I study the impact of perceived factional faultlines on team learning processes within a merging context.
This relation is proposed to be mediated by the degree of behavioral integration within a team.
Furthermore, the moderating influence of a transformational team leader on the relationship between behavioral integration and team learning is investigated. Data were collected in teams within 4 merging companies to create a sample of 80 respondents, divided into 18 teams who completed the survey.
Perceived factional faultlines positively predicts task and process learning within teams. This study did not confirm the mediating relationship of behavioral integration, nor the influencing relationship of transformational leadership on the process between behavioral integration and team learning.
Keywords: factional faultlines, team learning, behavioral integration, transformational leadership, merger
2
Table of content
Table of content... 2
Introduction ... 4
Conceptual Model ... 6
Literature ... 7
The perceived faultlines in a merging context ... 7
Faultlines ... 7
Factional faultlines ... 8
The social categorization perspective ... 9
Similarity-attraction perspective ... 10
Perceived factional faultlines ... 10
The effects of perceived factional faultlines on team learning ... 11
A typology of team learning ... 11
Perceived factional faultlines and task learning ... 12
Perceived factional faultlines and process learning... 12
Perceived factional faultlines and social learning ... 12
The influence of faultlines on behavioral integration within a team ... 13
The effects of behavioral integration on the process of team learning... 14
The moderating role of transformational leadership ... 14
Method ... 16
Sample description ... 16
Research design & procedure ... 17
Measures ... 17
Perceived faultlines ... 17
Behavioral integration ... 18
Transformational leadership... 18
Team learning ... 19
Control variables ... 19
Objective faultlines ... 19
Analyses ... 20
Factor analysis... 21
Results ... 23
Main descriptives and correlations... 23
Hypothesis testing ... 24
3
Conclusion and discussion ... 28
Summary and discussion of the main findings ... 28
Theoretical implications ... 30
Practical implications ... 30
Limitations and future research ... 31
Conclusion ... 33
References ... 34
Appendix A - Questionnaire ... 47
Appendix B - Factor analyses ... 57
4
Introduction
In today’s difficult markets, companies will increasingly search for ways to create and maintain a competitive (cost)advantage over competing firms. One way of strengthening a firm’s position in the market could be the initiation of a merger or acquisition (Ferrer, 2012). Mergers and acquisitions continue to play an important role in shaping business activities worldwide. Mergers and acquisitions are important business strategy to help improve organizational performance. The primary motivation for a merger or acquisition is to create and exploit synergies (Bohlin et al., 1993). Therefore, a merger or acquisition could be a strategy for strengthening a firms position in the market due to a cost advantage.
Success of mergers can be measured in several ways, whereas the degree of collaboration and communication are important underlying factors in facilitating the mergers success (Vazirani, 2012).
Probably the greatest challenge is the transition from one or more organizations to one integrated organization (Ulrich et al., 1989). People in merging contexts have somewhat psychologically and socially invested in their respective prior firms and are expected to work together in a new organization. For decades, researchers have devoted substantial attention to understanding the effects of team composition on processes and performance (Li and Hambrick 2005). Within this broad scope, there has been a particularly strong interest in the association among demographic heterogeneity, conflict, and performance of teams (Jackson, Joshi, & Erhardt, 2003; Milliken & Martins, 1996; Williams & O’Reilly, 1998).
Empirical research on diversity management shows different potential outcomes of diversity on for instance team performance (Williams and O’Reilly, 1998; Jehn, Northcraft and Neale, 1999). Within the diversity literature, we can discern a stream that states that diversity creates more difficulties for team coordination and team performance and an optimistic stream arguing that diversity could lead to value creation and increased team performance (Mannix and Neale, 2005). Because of this inconsistency, Williams and O’Reilly (1998) argued for a more complex framework concerning team diversity, which also counts for contextual aspects and the types of diversity addressed. A measure of diversity that also counts for complexity is group faultlines research, as introduced by Lau and Murnighan (1998). They address the potential negative consequences of diversity. Faultlines can be seen as a way to open up consideration of demographic dissimilarity between subgroups within an overall team. Faultlines are ‘’hypothetical lines that can divide a team into subgroups based on the alignment of the individual attributes of the members of a team’’ (Lau and Murnighan, 1998 p328; Thatcher & Patel, 2011). The strength of an existing faultlines depends on the amount of aligned characteristics among team members, and is found to have implications for political processes, communication and coordination in decision-making processes (Lau and Murnighan, 1998; Pearsall, Ellis and Evans, 2008; Bezrukova, Jehn, Zanutto and Thatcher, 2009). Hence, faultline research allows for team dynamics as a result of multiple attributes, and counts for the belonging to multiple subgroups (Bezrukova et al, 2009). According to Nishii and Goncalo (2008), the emergence of subgroups influences the patterns of social interaction that unfold within the team. As a consequence, the degree to which a team is able to capitalize on the different perspectives and skills the diversity offers, is affected by these processes (Nishii and Goncalo, 2008).
5 This study contributes to the literature by examining the degree to which people actually perceive factional subgroups, whereas most prior research only focusses on the development of subgroups based on objective demographic variables. Prior research on objective factional faultlines doesn’t take into account whether people actually perceive these subgroups, which is an important condition under which the effects of faultlines on teams processes are being explained (Jehn & Bezrukova, 2008; Pearsall, Ellis &
Evans, 2008).
Several scholars argue that people are inclined to like, trust and interact with others whom they perceive to be similar to them (e.g., Berscheid & Walster, 1978; Byrne, 1971; Tsui & O’Reilly, 1989). This could be problematic for communication and collaboration when combining two or more organizations.
Easy and efficient communication is key in effective team functioning (Shaw, 1981). Team problems could be caused by failure to exchange key information, poor coordination of activities, and difficulty in formulating and implementing responses to environmental shifts. Li and Hambrick (2005: 794) define factional groups as: “groups in which members are representatives, or delegates, from a small number of (often just two) social entities and are aware of, and find salience in, their delegate status”. Instead of coming to a team as individuals, members of a factional group come as, say, a threesome or foursome, and they combine with another twosome or threesome to constitute the team. These perceived factional faultlines could disturb or reform the communication and coordination processes within the team as a whole. Lack of social interaction, communication and collaboration may complicate a behaviorally disintegrated team (Li and Hambrick, 2005).
Rupert and Jehn (2008) state that it is difficult to manage diversity within teams and that is difficult to exploit the knowledge and experience from people with different backgrounds. Learning processes could suffer from the inherent effects of perceived factional faultlines. Scholars found mixed results for the effects of faultlines on team learning (e.g. Gibson & Vermeulen, 2003; Lau & Murnighan, 2005). Our knowledge in this area is limited and contradictory, making it necessary to conduct more research on team learning processes. The focus of this study, in which people from different backgrounds have to work together and have to learn from each other, makes is interesting to focus on this potential effect of the faultline dynamics. Finally, the role of the team leader as a positive influencer on the relationship between behavioral integration and team learning will be investigated. The degree to which a team leader engages in transformational leadership is proposed to positively influence this relation due to the ability to focus on a shared team identity and the stimulation to collaborate, interact and learn from each other.
The field study concerns mergers between several companies in the Netherlands. The focus will be on mixed (factional) teams within the new organization and the behavioral integration of the people within these teams. The degree to which respondents perceive factional faultlines will be analyzed, focusing on the inherent effects on the behavioral integration and the potential negative effect on team learning. This way, this quantitative study contributes to our understanding of the possible consequences of perceived factional faultlines and learning within teams.
Summarizing, I apply and extend theory from four primary sources. This study draws on Li and Hambrick’s (2005) concept of factional faultlines as a way to theorize about the schisms that exist in factional teams. Although Li and Hambrick (2005), have developed a model to explain the dynamics in
6 joint venture management teams (see also Child, 1994; Pearce, 1997), their ideas have relevance for the entire class of small groups that we term “factional” in this study. This study represents a theoretical exploration and empirical investigation of factional groups in a merging context. The second major theoretical source which complements the work of Li and Hambrick (2005) is the work of Jehn and Bezrukova (2010) as a way to determine whether team members actually perceive factional faultlines.
The third source is the work Rupert and Jehn (2008), who developed learning typologies. The fourth source is the work of Bass & Avolio, (1994, 1995) who developed a measure for transformational leadership. This study contributes to theory by (1) the application of (factional) faultline theory in a merging context, (2) en by extending the small field of research that focusses on perceived faultlines.
Conceptual Model
A conceptual model concerning the expected relationships between the variables of this study is provided in figure 1.
Figure 1.: A graphic expression of the supposed relationship between the concepts.
Research question:
To what extent do perceived factional faultlines in a merging context affect the behavioral integration within a team, which ultimately influences team learning?
The research question will be answered by solving the following sub questions:
To what extent do perceived factional faultlines influence the team learning processes?
To what extent do perceived factional faultlines influence the behavioral integration within the team?
To what extent does behavioral integration affect the team learning process?
To what extent does transformational leadership behaviour affect the relationship between behavioral integration and team learning?
7
Literature
The perceived faultlines in a merging context
Diversity researchers have identified numbers of foundations to explain the underlying effects of diversity and its consequences on organizational processes, performance and outcomes. This literature overview focuses on the faultline theory as a basis for subgroup formation within teams that are composed after a merger of two or more organizations.
Faultlines
Faultline theory was developed by Lau and Murnighan (1998) as a response to theoretical inadequacies within the diversity theory. The results of previous diversity research have been rather contradictory, whereas diversity seems as likely to hinder performance as it is to improve it. Thus, diversity theory could be qualified as inconclusive which aroused a need for a new approach in diversity research.
Previous research has mainly focused on the distribution among team members along relevant dimensions of attributes (Alexander, Nuchols, Bloom, and Lee, 1995). The main criticism on these types of research is the underlying assumption that demographic attributes are independent of each other. When, for instance examining race as the only dimension for distribution, other potential relevant attributes like gender could be ignored, which could lead to weak and invalid conclusions. Faultline theory, however, captures an alignment approach to diversity. This approach entails that instead of analyzing separate demographic and informational characteristics of team members, the interaction between these characteristics is taken into account (Bezrukova, Thatcher and Jehn, 2007). Hence, because the faultline measures take into account cumulative proportions of variance across demographic attributes, it enables to estimate the extent to which degree group process variability can be explained by the presence of in- group clusters (Bezrukova, Thatcher, and Jehn, 2007).
The faultline theory is a relatively new approach to group processes which helps us identifying the diversity bases on which team may split up into subgroups. Lau and Murnighan, (1998) describe group faultlines as; “hypothetical lines, that may split a group into subgroups based on one or more attributes” (Lau and Murnighan, 2005: 645). They mention two types of faultlines that may lead to active subgroups within a larger group; demographic faultlines which may arise based on age or gender and faultlines based on nondemographic characteristics, such as personal values or personality.
A key assumption of the faultline framework is the discontinuity effect. This effect yields that the competitiveness between subgroups is stronger than for interactions between individuals. As a consequence the intergroup interactions are predominated by the interactions between subgroups (Insko, Schopler, Hoyle, Dardis, and Graetz, 1990; McGlynn, Hardy and Cottle, 2009). Hence, active faultlines and the consequent subgroup formation influences the intragroup trust levels and intragroup communication and cooperation (Tuggle et al, 2010).
As stated earlier, Lau and Murnighan (1998) mention two types of faultlines that may lead to active subgroups within a larger group; demographic faultlines and faultlines based on non-demographic characteristics. The demographic variables are most easily visible which is why they are the most likely to
8 be the first criteria for subgroup formation (Lau and Murnighan, 1998; Williams and O’Reilly, 1998;
Bezrukova, Jehn, Zanutto, and Thatcher, 2009).
Faultline theory has proved many practical implications for intergroup interactions, processes and outcomes. Williams and O’Reilly (1998) mention three ways in which, demographic attribute differences influences group dynamics. They argue a tendency of people to categorize members of a group in social sub-groups, to establish a sense of belonging and to foster self-esteem (Nishii and Goncalo, 2008).
Members of a group compare different attributes of other team members to their own characteristics, as argued by Mannix & Neale (2005) and Williams & O’Reilly (1998). This is the base of the social process and the extent to which homogenous subgroups are formed depends on the ability of individuals to perceive potential faultlines (Crisp, 2010).
The similarity-attraction perspective is argued to be another influencer for group dynamics (Williams and O’Reilly, 1998). The similarity-attraction paradigm explains that increased similarity with an individual is associated with increased interest in this person (Byrne, 1997). As a consequence, group members perceive dissimilar co-members as being less trustworthy and capable (Williams and O’Reilly, 1998), implicating negative consequences for the functioning of the group as a whole. The information/decision making perspective emphasises a positive view on the effects of diversity within a team (Van Knippenberg & Schippers, 2007). Differences in of task-relevant knowledge, skills, abilities and opinion can serve as sources for interaction base on which team members elaborate information (Mannix
& Neale, 2005; Van Knippenberg & Schippers, 2007). This may prevent the group form focussing to easily on a course of action on which there seems to be consensus, allowing ideas to develop into higher-level outcomes than would be achievable in more homogeneous groups (Van Knippenberg, De Dreu & Homan, 2004; Mannix & Neale, 2005).
Factional faultlines
The focus of this research is to analyse the factional faultlines, as a consequence of an initiated merger between one or more organizations. Factional faultlines are likely to be seen in the creation of a joint venture in which members of several companies are appointed as representatives for their own company within the joint venture. Factional teams have what might be thought of as a priori, or engineered faultlines. A crack or a divide is, always present due to the fact that the faultline is partly pre- engineered during the creation process (Li and Hambrick, 2005). Factions encounter membership of a representative faction, described by Li and Hambrick as; “groups in which members are representatives, or delegates, from a small number of social entities and are aware of, and find salience in, their delegate status”. Li and Hambrick (2005) studied the demographic faultlines in the formation of joint ventures and found that a high degree of dissimilarity in demographic dimensions between factional groups caused the teams to perform poor due to social conflict, task conflict and poor collaboration.
This entails that the members of both firms entering the new group already perceive themselves as constituting a group. Encountering history within different social entities (the companies prior merger) can be the source of a faultline due to the identification with the pre-merging entity. The studies of Jetten, O’Brien et al. (2002) and Ellemers (2003), describe merging processes in which subgroups identified strongly with the pre-merging entities, causing problematic integration issues. Researchers identified
9 several difficulties when combining old organizational identities. The simple blending of two old identities into a single new one can be problematic because it poses a threat to group distinctiveness (Branscombe, Ellemers, Spears, & Doosje, 1999; Jetten, Spears, & Manstead, 1997). Other researchers found that the melting of several organizations or departments could lead to poor outcomes for those who identified strongly with the pre-merging entity (as can be seen in; Jetten, O’Brien et al. 2002 and Ellemers, 2003).
Two research streams within diversity theory are essential in understanding the likelihood of factional faultlines in a merging context; the similarity-attraction perspective and the social categorization perspective (Williams & O’Reilly, 1998). These theories explain why differences matter within a group and give meaning to diversity’s influence on team processes and performance (Mannix & Neale, 2005;
Thatcher and Patel, 2012).
The social categorization perspective
This perspective can be described using the self-categorization theory (Tajfel, 1981; Turner, 1987) and social identification theory (Hogg & Abrams, 1988; Turner, 1982). According to the self- categorization theory are team members inclined to classify themselves and others into categories that are familiar. Team members use attributes of others to determine to what groups they belong (Sawyer et al., 2006; Thatcher et al., 2003). Team members in merging organizations will categorise themselves into factional categories and they will seek to act in a manner similar to their in-group as a result of the internalization and support of activities that relate with inclined aspects of their identities (Jimmieson et al., 2009; Pitsakis, 2012).
The assumption that the merger between organizations could cause people to perceive factional faultlines is also influenced by the degree of identification with the pre-merging entity. A merger could create a threat to the organization’s identity and thus the employees’ social identification with their organization. Social identity theory (SIT) helps us in explaining this effect (Dick et al., 2006). One of the basic ideas of social identity theory (SIT) is that the social teams we belong to form a significant part of our self-concept (Tajfel and Turner, 1979, 1986). In contrast to our personal identities, SIT calls that part of our self-concept in which we see ourselves as being similar to other members of our social teams – our social identity. A key aspect of this theory is that in order to derive a sense of positive self-esteem from in- group membership, the in-group must be viewed as being distinct from other groups (Mlicki & Ellemers, 1996). Because of the embedding of team membership in the individual’s self-concept, an individual will, to a greater or lesser extent, think, act, and have feelings consistent with the team’s values and relative social standing. Thus, to the highly identified individual, violating the team’s norms is like breaking a rule set by him or herself and, in the same way, the team’s success is experienced as his or her own achievement. A merger essentially redraws or dissolves the category boundaries of distinct teams within the newly created merger entity. Thus, mergers inherently threaten the distinctiveness of the pre-merger team identity and, consequently, sub-team identification should increase at the expense of identification with the post-merger entity.
10 Similarity-attraction perspective
Similarity in attributes ranging from attitudes and values to demographic variables increases attraction and liking to others who team members believe are similar to themselves (Williams & O’Reilly, 1998). This perspective explains why team members are likely to align with similar members in the process of subgroup formation (Thatcher & Patel, 2012: 973). These subgroups will arise when members categorise themselves and identify themselves with factional categories and certain groupings. The similarity-attraction perspective helps us to explain the tendency members feel toward the validation of homogeneity and the comfort of belonging (Mannix & Neale, 2005). The composition of group members can be perceived as a measure of knowledge present in this group (Homan, Greer, Jehn and Koning, 2010).
The demographic attributes like education and tenure, for instance, are regarded by co-members as indicators for the knowledge base and mind-sets of others (Jehn, 1995). This group member evaluation in turn influences sub-group formation. Because the emergence of subgroups influences the patterns of social interaction that develops within a team, the degree to which a group is able to capitalize on the different perspectives and skills the diversity offers, is affected by this categorization (Nishii and Goncalo, 2008).
Perceived factional faultlines
The study of Li and Hambrick (2005) measures faultlines based on objective characteristics such as faction, gender, nationality, and race,. However, they fail to address the activation process, which was originally conceptualized by Lau and Murnighan (1998). Li and Hambrick (2005) did not examine whether the members actually perceive these factional sub group distinctions as a basis for consequent intra-team processes and outcomes. Lawrence (1997) argues that that demographic variables do not necessarily equate with subjective or psychological processes in explaining organizational outcomes. This is why this study will focus on the degree to which respondent perceive factional faultlines.
Lau and Murnigham (1998) argue that faultlines in teams can be inactive and go unnoticed for years without any changes in team processes. A distinction can be made between dormant and activated faultlines in order to assess the degree to which team members perceive such a faultline to exist. Dormant faultlines can be described as the demographic alignment across members that may (or may not) divide a team into subgroups based on objective demographic alignment across members. Activated team faultlines occur in teams when members actually perceive these divisions into subgroups. While dormant faultlines are based on the objective demographics of team members, activated team faultlines exist when members perceive separate subgroups (Lau an Murnigham, 1998). Jehn and Bezrukova (2010) define faultline activation as the process by which a potential faultline “is actually perceived by group members as the division of the group into separate subgroups based on demographic alignment”.
A few researchers have taken into account the degree to which faultlines are perceived by focusing on the activation of dormant faultlines within existing groups. By identifying a trigger of contextual elements that can activate the awareness of coworkers on their differences and similarities between subgroups (Lau and Murnighan, 2005). The merging context, however implies that new groups
11 will be formed in which the context itself functions as a trigger that could make team members aware of differences.
The effects of perceived factional faultlines on team learning A typology of team learning
Most research in the field of team learning focusses on the potential positive effects of diversity in teams on learning. Van Woerkom & Van Engen, (2009) argue that knowledge and skills gathered by one team member can be transferred to team members through processes of giving feedback, explanation or advice. However, I propose that teams that perceive subgroups based the factions are less likely to learn from each other due to a low degree of interaction and communication between subgroups.
Team learning is defined as “the process of reflection and interaction in which team members actively acquire, process, share, evaluate, and combine knowledge and information in order to improve team performance” (Jehn & Rupert, 2007: 186). It is a process in which team members learn constantly about their team, tasks, resources, and context (Edmondson et al., 2011; Sessa & London, 2008; Vora &
Markóczy, 2012). Foldly, (2004) argues that team learning is key in achieving high team outcomes. Hence, sharing, generating, evaluating and combining knowledge as the key elements of team learning are needed to generate new knowledge and enhance the process of team learning (Argote, 1999). Edmondson (1999:
353) conceptualizes team learning process as “an on-going process of reflection and action, characterized by asking questions, seeking feedback, experimenting, reflecting on results, and discussing errors or unexpected outcomes of results”. A team that engages in learning activities will underscore information sharing and cognitive and behavioral exchanges (Day et al., 2004).
The extent to which emergent collective knowledge allows the group to perform better depends on the quality of the information elaboration processes between team members (Curşeu & Pluut, 2013).
When one person in a group learns something that is not shared with other members of the group this constitutes individual learning, rather than group learning (Wilson et al., 2007). If group members work together, they should experience process gains, meaning they should become increasingly effective and produce more than would be possible if individual group members worked alone and aggregated their production (Sessa & London, 2008). Gibson and Vermeulen (2003), focus on the learning processes in teams consisting of subgroups. Part of their conclusion is that teams that score relatively high on subgroup strength, report lower learning behaviors, indicating a negative influence of subgroups within teams.
Jehn & Rupert, (2007) responded to the call for a typology due to a broad scope of subjects that was incorporated in previous learning research. Their three learning typologies of task learning, process learning and social learning further refine our understanding of team learning. Task learning is described as “the process in which a team is learning more about specific tasks. Teams can learn about specific tasks by sharing and reflecting upon knowledge, ideas, insights through interaction with each other, in order to improve team performance” (Rupert & Jehn, 2007: 126). Teams can also learn about the process of collaboration. Process learning can be described as “the pattern of interaction through which members create work routines and develop procedures about how to organize their work” (Rupert & Jehn, 2007:
12 127). Social learning is “the process through which team members get to know each other better as individuals and learn to interpret each other better as individuals and learn to interpret each other’s behaviour in the context of personal life and personality” (Rupert & Jehn, 2008: 128).
These learning types will be used in this study. The influence of perceived factional faultlines on these specific types will be described in the following section. Hypothesis will be developed for each specific learning type.
Perceived factional faultlines and task learning
The likelihood that team members will perceive factional faultlines was already described in the first part of this chapter. Within team that are composed of subgroups, team members are inclined to search for confirmation of feelings and thoughts (Swann, Polzer, Seyle & Ko, 2004). There is a higher change that team members will find this within their subgroups than between subgroups (Rupert and Jehn, 2008). The formation of subgroups could have detrimental effects on processes within team.
Scholars have argued negative consequences for trust (Brewer, 1996; Insko & Schopler, 1998; Insko et al., 1993), destructive group processes and decreased satisfaction due to increased ingroup-outgroup differences (Polzer et al., 1998; Sherif, Harvey, White, Hood & Sherif, 1961) and Jehn en Bezrukova (2008) found negative influences resulting in increased conflict, decreased satisfaction and worse performance in teams. These processes, combined with the tendency of people to share information within their own subgroup (Milliken & Martins, 1996), makes teams to learn and critical reflect less on other insights that will be available with respect to the tasks of a team, eventually causing team to learn less about tasks (Rupert and Jehn, 2008). Hence, the second hypothesis is stated as follows:
Hypothesis 1a. Perceived factional faultlines are negatively related to task learning.
Perceived factional faultlines and process learning
Factional subgroups will develop routines and procedures within their own subgroup to be able to perform their work. Within factional subgroups, people will communicate about the way things should be done, whereas communication between subgroups will be less. This will eventually put subgroups further away from each other, making it more difficult to reflect on team functioning (Rupert and Jehn, 2008). Reflection on team functioning was found to be important for the adjustment of routines and to increase team performance (Tjosvold et al., 2004). I expect that perceived factional faultlines will have a negative influence on process learning. Hence, the following hypothesis was stated:
Hypothesis 1b. Perceived factional faultlines are negatively related to process learning.
Perceived factional faultlines and social learning
I expect that team members who perceive factional faultlines will develop stereotypes and prejudices based on differences between team members (Messick & Mackie, 1989). This makes team members of distinct factional subgroups less likely to be inclined to hang out with people who they perceive to belong to a different subgroup. Team members will be inclined to listen to and be open to each other, causing them to exchange less about personal matters. I expect that perceived factional faultlines will have a negative influence on social learning. Hence, the following hypothesis was stated:
13 Hypothesis 1c. Perceived factional faultlines are negatively related to social learning.
The influence of faultlines on behavioral integration within a team
The presence of multiple subgroups creates an important dynamic in work teams (Carton and Cummings, 2012). Given that all members of a work team share a common goal (Kozlowski & Bell, 2003), implies that different subgroups must interact, rely on each other for task responsibilities, and otherwise remain strongly aware of one another (Carton and Cummings, 2012).
Li and Hambrick, (2005) state that the existence of demographic factional faultlines is likely to have a negative influence on the degree to which, and how, team members interact. Li and Hambrik argue that people like and trust the people who are like themselves. People are likely to avoid, distrust and dislike those who are different (Byrne, 1971; Tsui & O’Reilly, 1989). Coalitions within teams may arise when team members perceive subgroups based on diversity characteristics. These coalitions may be based on the members that share the same conceptions with respect to their work (Murnighan & Brass, 1991).
Personal characteristics (e.g., gender, race, age) can be mentioned as examples of those differences, based on which people avoid, distrust or dislike each other or other (sub)groups of people.
Jackson, (1992) and Pelled, (1996) described the mistrust and dislike in a team as a result of stereotyping.
Alderfer, (1987); Tajfel, (1982) and Taylor, Sheatsley, & Greeley, (1978) describe an in-team/ out-team identification that could be the product of this. There is also a high chance that members of the team will identify themselves more with their subgroup than with the team (Tajfel & Turner, 1986). These cognitive processes give rise to in-team/out-team hostility, or to individuals seeing their in-team as worthy and efficacious, and seeing their out-team as unworthy and incapable (Tajfel, 1982).
Researchers developed several constructs to describe the interaction within teams. These constructs include communication (Smith et al., 199 4), collaboration (Chatman & Flynn, 2001), and social interaction (Chatman, Polzer, Barsade, & Neale, 1998). Hambrick (2004) developed a unified construct to describe the tendency for some teams to engage in more “team like” behaviors than others, Hambrick (1994) set forth the concept of behavioral integration, defined as “the degree to which mutual and collective interaction exists within the team” (Hambrick, 1994: 188). Behavioral integration has three main manifestations: information exchange, collaborative behavior, and joint decision making. Li and Hambrick (2005) argued that demographic factional faultlines are likely to contribute to poor behavioral integration in such teams through the diminishment of interaction, exchange and collective effort. I propose that this negative effect on behavioral integration will also be seen when team members perceive factional faultlines. Therefore the second hypothesis is the following:
Hypothesis 2: Perceived factional faultlines are positively related to the behavioral disintegration in a team.
14 The effects of behavioral integration on the process of team learning
A behaviourally-integrated team will show delightful quantity and quality (richness, accuracy, and timeliness) of information exchange, collaborative behavior, and joint decision making (Hambrick, 1994).
Whereas we propose the perceived factional faultlines to negatively influence behavioral integration, the focus will be on the consequences of this low behavioral integration for the process of team learning.
In a behaviourally low integrated team, consisting of factional subgroups, members are likely to exchange information, communicate and make decisions within their distinctive subgroup. However, the exchange of information, communication and joint decision making is likely to be low between perceived subgroups. Reflection on functioning is important for the adjustment of routines and working processes (Tjosvold et al., 2004). As a consequence of low behavioral integration, process learning is likely to be low in teams. Team members in behaviourally low integrated team will be more inclined to share information within their subgroup, whereas they will be less inclined to share information with members of other subgroups within their team (Milliken & Martins, 1996). This will affect the overall task learning of a team, causing less critical reflection on other insights that will be present in the team. Social learning is proposed to be low in behaviourally disintegrated teams due to that communication between subgroups will be low. Members of different subgroups will be less open and open to listen, which makes members exchanging less about personal matters, like private lives (Jehn and Rupert 2007).
Hence the following hypotheses were developed:
Hypothesis 3a. Behavioral disintegration is negatively related to task learning.
Hypothesis 3b. Behavioral disintegration is negatively related to process learning.
Hypothesis 3c. Behavioral disintegration is negatively related to social learning.
By extension:
Hypothesis 4. Behavioral disintegration mediates the relation between perceived factional faultlines and team learning.
The moderating role of transformational leadership on the relationship between behavioral integration and team learning
The main focus of this study is on the effects of perceived factional faultlines on behavioral integration, which eventually influences the team learning process. I also propose that a transformational leader is likely to have a weakening impact on the relation between behavioral integration and team learning.
Leadership literature has recognized the importance of leader behaviors on team and organizational performance and other outcomes (Bass, 1985; Lowe et al., 1996). Leadership theories have been categorized into several approaches focusing on traits, behavior, situational contingencies, or cultural contingencies (Bass, 1990; Burns, 1978; Yukl, 1989; Yukl & Van Fleet, 1991). Transformational leadership is a core feature in social group and it is difficult to think about groups without thinking about
15 who leads of manages them, according to van Knipperberg, (2003). Burns (1978) argues that transformational leaders are those who motivate their followers to perform beyond expectation by raising the followers’ confidence levels and by providing support for developing high levels of performance. A typical transformational leader is able to inspire team members. Bass & Avolio, (1994, 1995), expanded Burns leadership factors and they have identified the behavioral components of transformational leadership comprising of:
- Idealized influence attributed. Transformational leaders exhibit confidence and instil emotions (such as dignity, integrity, and honour), a sense of selflessness, and respect in their followers.
- Idealized influence behavior. Transformational leaders are goal-oriented and they encourage the completion of work based on a collective sense of beliefs, values, purpose, and mission.
- Individualized consideration. Transformational leaders recognize the needs of their followers and provide them with personal guidance and development.
- Inspirational motivation. Transformational leaders are optimistic, visionary and passionate about work.
- Intellectual stimulation. Transformational leaders encourage innovation and creative problem solving. (Loon et al., 2012: 192)
The first behavioral component is subdivided into two perspectives: idealized influence attributed and idealized influence behavior (Bass and Avolio, 1997; Krishnan, 2005).
Transformational leadership behaviors are best suited to encourage team learning because it is supported through the process of encouraging team members to question their assumptions, approach old problems in new ways, or reframe problems in light of the newly integrated organization's challenges (Jung et al., 2003). Bass, (1998) argues that a transformational leader is able to create an open team culture that is open to diverse ideas, by valuing team members' divergent views, creating open exchanges of information, and resolving conflicts effectively. Intellectual stimulation and individualized consideration are the most appropriate leadership behaviors to create such a context. Transformational leadership can play a key role in enabling individuals and organizations to create, exploit, renew, and apply knowledge in order to create the essential competences required for improvement of learning.
The rationale for the assumption that learning in behaviourally disintegrated teams will be relatively low is based on a need for collaborative behavior, information exchange and joint decision making. The degree to which subgroups are interdependent to accomplish tasks heightens the need for behavioral integration (Thibaut & Kelley, 1959; Salancik & Pferrer; Saavendra et al., 1993). A transformational leader is able to weaken the negative effects of behavioral disintegration on learning by the stimulation of these behavioral aspects between subgroups. Hence the five hypothesis is stated as follows;
Hypothesis 5: The relationship between behavioral integration and learning is moderated by transformational leadership.
In the following chapter, the methodology for testing these hypotheses will be described.
16
Method
A theory testing approach, based on quantitative data was used to test the faultline activation model. This quantitative research methodology allows generalization of results to a larger population, when the sample is a good reflection of the total population (Myers, 2009). The model, as illustrated in figure 1, was tested among 4 organizations.
Sample description
To test our hypothesis, questionnaires were set out to 4 organizations. These questionnaires were developed using relevant constructs. The target population of this field study were Dutch organization from different fields of interest that recently conducted a merger. A minimum of two organizations had to merge, in which new team had to be composed of members from the different organizations. The teams had to consist of 3 to 15 members in order to meet the definition a team. The definition of a team as used by Hackman, (1987) and Alderfer, (1987), will be applied in this study. A team can be defined as follows; a social system of at least 3 members who (1) recognize themselves as a team (2) are recognized by others as a team, (3) have a shared responsibility for a team product or a service and (4) operate in an organization. The terms team and group will be used interchangeably in this study. The anonymity of respondents was guaranteed. Based on all the former mentioned descriptives, the sample was seen as a good reflection of the targeted population. The organizations were geographically and in terms of size, very disperse. The use of a large sample makes it possible to generalize the results of this study to a large population generalizing the results to a large population (Cooper and Schindler, 2008; Myers, 2009). The questionnaire was designed to collect relevant information and to increase the likelihood that validated and reliable data will be collected, as described by (Warwick and Linninger, 1975).
A relatively high overall response rate of 93% was achieved through visiting the companies in which the hardcopies of the questionnaire were distributed. The response rate can be considered as a good response rate in field research according to Baker, (1994). Companies were visited two times in total. For each team, a response rate of 80% at least was achieved. Of the 80 completed questionnaires, two were discarded due to the fact that more than 30% of the data were missing.
Complementary, a team grid comprised of all team members and their demographic features was collected per team to obtain the relevant demographic characteristics for the missing respondents.
Congruous factional group analysis within merged teams requires the selection of teams consisting of at least two members per pre-merging firm. Due to poor team selection, 8 teams did not meet these requirements, and were to be excluded from the analysis. Merger characteristics from the organizations in the sample were compared to check for unit nonresponse bias, and no bias was detected.
The educational level of the participants (92,3%) was higher vocational or university and the majority of employees is fulltime employed (82,1%). The tenure of employees ranged from 1 to 19 month and the teams were composed of a representative division of gender; (52% male, 48% female). Team members averaged 39 year in age, with a standard deviation of 9,7. Team core tasks ranged from 6 teams that focused on post-merger integration efforts business, 4 teams on quality management, 5 teams on
17 operational management and 3 teams that focused on sales. Team members interacted on average during four collective meetings a month.
Research design & procedure
The case/company selection was done based several criteria. Mergers had to be initiated at least 6 months before conducting the research. Another criterion was the mixing of team members of the merging organization; new teams had to be composed from at least two members of distinct organizations to be able to measure perceived factional faultlines.
The organizations were contacted using the personal network of the researcher. The companies were contacted by phone or by e-mail. Reading documents and initial conversations with managers from the organizations was essential in order to understand the merging situation at the specific case companies. This made it possible to understand the change they faced, based on which the researcher could determine the compatibility of the organization with the research. The questionnaires were distributed both online and in printed versions after confirmation of participation by the companies.
Differences between online and printed versions were checked and no significant differences were found.
The companies which were studied varied highly on several criteria. Variation could be seen in for instance the type of industry, the ages of the organizations and the distinction between profit and non- profit organizations (for instance; subsidised governmental organizations and fully commercial organizations). This heterogeneity of the large sample improves the external validity of the results of the study. External validity refers to “the generalizability of research results and conclusions to other people, organizations, countries, and situations” (Van Aken et al, 2007, p. 165). Hence, heterogeneity is quite high, which leads to a relatively high representative sample, and consistent results among these firms would imply high external validity.
The study is restricted by the use of companies within the Netherlands, which implies that the external validity (generalizability) towards culturally distinct countries is limited.
Measures
The measurement of the concepts; (perceived factional faultlines, behavioral integration, team learning and transformational leadership) was done using existing and validated items. This way, the construct validity, e.g. a study deals with “the extent to which a measuring instrument measures what it is intended to measure” (Van Aken et al, 2007, p. 163), is taken into account. Furthermore, personal evaluation of the instruments and techniques concerning their coverage was applied.
The respondents had to indicate the extent of agreement with each statement on a 7-point Likert- type scale (1 = strongly disagree, 7 = strongly agree). All items were translated from English to Dutch. The translation process by three master students who are proficient in English.
Perceived faultlines
The measurement of the perceived faultlines was based on the scale of Jehn and Bezrukova (2010). The measurement of this variable can be sub-divided into two distinct subcategories; (1) the
18 degree to which subgroups are perceived and (2) the measurement of the degree to which they perceive a specific base to be responsible for a split up.
Respondents were asked to rate the degree to which they perceived subgroups based on eight items (e.g., At meetings of our team subgroups often sit together, during the work the team is divided into several sub-groups). Subsequent questions were added to indicate the on which base they perceive a split up within the group (e.g., Since the change, my team split into subgroups based on: demographics – e.g. age, gender, functional background, and behaviors – e.g. way of working, personality). These items were also used by Jehn and Bezrukova (2010).
The item to indicate the degree to which the team members perceived subgroup creation based on factions was also added and was based on the study (based on Li and Hambrick 2005). They studied the existence of objective factional faultlines, but whereas the focus of this research is on whether these faultlines are actually perceived, so the following item was added; ‘Since the change, my team split into subgroups based on: the company I worked for prior merger’’ to the variables from the measure of Jehn and Bezrukova (2010). The (Likert) scores for team tenure and the scores for the factional base were added to the sum variable of perceived factional groups to be able to measure perceived factional faultlines.
Additional to the bases which were included in the questionnaire, the respondents were able to formulate additional characteristics, on which base they perceived subgroup formation was strongest. This open question was coded into distinct variables. 37% of the respondent which filled in this question (30% of the total amount of the dataset) formulated the factional base. A few additional bases were added, but this was not included in additional analysis.
Behavioral integration was measured using the measure of Simsek (2005). This team-level measure of behavioral integration captures the dimensions; collaborative behavior, information exchange, and joint decision making. Collaborative behavior was originally assessed using a five point scale ranging from 1 (“strongly disagree”) to 5 (“strongly agree”). In order to keep conformity within the questionnaire, the seven point Likert scale was used ranging from 1 (“strongly disagree”) to 7 (“strongly agree”). The questions concerning the measurement of the dimension information sharing was permuted with the information sharing scale used by Jiang, et al, (2012). The original items were banned because these items didn’t fit within the overall scale and since the translation of the original items could lead to invalid response. The questions of Ying et al. (2012) seemed to cover the dimension in a more comprehensive way and the translation into Dutch could be done better.
Transformational leadershipwas measured using the scales of the five dimensions; (idealized influence attributed, idealized influence behavior, individualized consideration, inspirational motivation and intellectual stimulation), of transformational leadership from Bass and Avolio’s (1997) Multifactor Leadership Questionnaire. This scale is the most commonly used measure of transformational leadership (Bono and Judge, 2004; Conger and Hunt, 1999). To ensure construct validity, items with a component score of at least 0.7 in the original study were incorporated in the questionnaire.
19 Team learning was measured using the learning types developed by Jehn and Rupert, (2007). The learning types were used earlier which made it possible to select the items with the highest factor loadings. The concept of team learning consists of 3 learning types which were incorporated in this study:
- Task learning. Items to measure task learning were used to measure the extent to which group members felt a degree of knowledge sharing, expertise and opinions about task and reflected upon insights regarding the task (e.g., “As a team, we increase our knowledge about tasks at hand”).
- Process learning. Items to measure process learning were used to measure the extent to which team members learn about and improve work procedures and routines with the goal to improve their team performance. Sample items were: “We regularly reflect on our work procedures to see how we can improve them” and “As a team, we create work routines that help us to improve our work.”
- Social Learning. Items to measure social learning were used to measure the extent to which team members knew each other well and how this influences team performance (e.g., “When you know each other personally, it makes work easier.”).
Control variables
13 control variables were included in the questionnaire. These control variables were used to increase the internal validity of the study. Cooper and Schindler (2008) state internal validity concerns the degree to which conclusions are drawn about a relationship that truly represent cause and effect, which basically means that no other variables influence causal relationship. Several criteria must be addressed in to conclude a causal relationship. Shadish et al., (2002) mention the criteria of temporal precedence ("cause" precedes the "effect" in time), covariation ("cause" and the "effect" are related) and nonspuriousness (there should not be a plausible alternative explanation for an observed covariation).
This last criterion was addressed, using the control variables, which we tested for their influence on the dependent and independent variables.
Objective faultlines
As mentioned before, leaders of the teams that participated in this study were asked to fill in a grid on which team member characteristics were collected. The grid was composed using the following attributes: age, length of team membership, highest completed education, relevant work experience and pre-merger organizational membership.
Information about the team members was used to calculate the faultlines strength and the faultline distance within the teams. The faultline strength was measured using the cumulative proportions of variance across the variables (Bunderson and Sutcliffe, 2002; Thatcher et al, 2003). Which could be used as a base for estimations of attribute clusters. Subgroups with high values of faultline strength are likely to be very homogeneous in nature (Thatcher et al, 2003).
An algorithm developed by Thatcher et al (2003) (Lau and Murnighan, 2005; Bezrukova, Spell and Perry, 2010) was used to calculate this. A group can be divided in several subgroups based on present faultlines, in a total of S= 2n-1-1 ways. The model of Bezrukova et al., (2008: 41) is presented in the figure below.
20 The Fau score represents the strength of faultline divisions and a higher score represents higher faultline strength.
In this model, xijk represents the value of the jth attribute of the ith member of subgroup k. The work group mean of a characteristic j is captured in ̅.j., and the mean of this same variable j within the subgroup is denoted with ̅.jk. Finally, the members per subgroup k are captured in . I used a clustering algorithm (Jobson 1992; Sharma 1996) to find the strongest split present in the teams.
The strongest faultline split per team (which resulted from the Fau analysis) was used in the measurement of the faultline distance. This was measured using the distance formula in which groups can be compared based on the Euclidean distance between the averages within the group and subgroups (Bezrukova et al., 2008: 41).
√∑ ( ̅ ̅ ) 2.
As applies to both these scores a higher faultline distance measure indicates a higher distance among subgroups on the particular attribute.
The control variable Real team (Wageman et al., 2005) was added in the analysis to evaluate whether the teams in this study share the perception of comprising a real team. The subcategories of team interdependence and team stability were included in this sample. The internal consistency is proven to be sufficient for team interdependence (α = 0.71) and team stability (α = 0.86) in this sample.
Task interdependence, as used by Campion et al., (1993) was used to evaluate shared responsibilities that are evident to interdependence within teams. This interdependence within teams is argued to increase motivation and subsequently team performance (Kiggundu, 1983). The internal consistency of this scale proven to be sufficient (α = 0.78).
Finally, team member characteristics like age, gender, tenure, work experience, fulltime employment, educational level and nationality were included as control variables in this study, together with team size. Team size was included, because it is known to influence group dynamics (Brewer &
Kramer, 1986), and larger teams have more potential for heterogeneity (Bantel & Jackson, 1989).
Analyses
Data were analysed using SPSS 20.0 to determine and calculate the relevant (descriptive) statistics and to perform regression analyses. To test the hypotheses an alpha level of 0,05 was used in the
