ASYMMETRIC FAULTLINE PERCEPTIONS Effects of Asymmetrically Perceived Change-related Faultlines on Behavioural Integration and Team Performance

(1)

Effects of Asymmetrically Perceived Change-related Faultlines

on Behavioural Integration and Team Performance

By

Lianne Veldman

University of Groningen - Faculty of Economics and Business

MSc. Business Administration - Change Management

Lianne Veldman

Violenstraat 21

9712 RD Groningen

0615687757

l.veldman@student.rug.nl

Student number: 2227282

1

st

supervisor: Dr. Joyce Rupert

2

nd

supervisor: Dr. Hanneke Grutterink

(2)

2

Abstract

This research examined the effects of asymmetric faultline perceptions on behavioural integration and subsequently, team performance. Past research often ignored that team members have different perceptions and therefore, this study made a distinction between activated (i.e. perceived) faultlines and asymmetric faultline perceptions (i.e. the degree to which members differ in perceptions of faultlines in their team). In this study, assumptions were made that organizational change activated faultlines. In addition, it was hypothesized that asymmetric perceptions of faultlines based on change-related attributes (e.g. resistance, and change aspect) have a negative influence on behavioural integration (i.e. group processes) and performance (i.e. team outcome). Change-related attributes refer to responses of behaviours and attitudes towards the change. Using quantitative field data of 318 individuals in 62 different teams, findings partially supported the hypotheses. Asymmetric perceived change aspect (e.g. informed and personal impact) faultlines explained the more negative perceptions of information exchange and joint-decision making, while activated informed and personal impact faultlines influenced the perceptions of information exchange and joint-decision making positively.

Keywords: Organizational change, asymmetric faultline perceptions, behavioural integration, performance

(3)

3

Introduction

In the past decades, organizations have been the platform of major radical and incremental developments, having influenced current society in different ways. The globalization of society led to a society which is increasingly diverse. This diversity knows several origins. One of the main causes according to Sayers (2012) was the history of social activism, in particular the anti-racism and feminism movements. These movements and many others led to demographic changes in organizations, such as a wider variety in cultures of employees, labour force diversity such as woman’s participation and religious diversity, aging of the workforce and changes in family structures. Due to these changes in demographics, major challenges on the management of individuals in the appropriate way have appeared. More specifically, managerial implications such as appreciation of how this diversity can benefit the organization, an increasing need of a better understanding and awareness of different cultures, and facilitating a balance between work and life. According to Sayers (2012), the goal of managing diversity is to “encourage productive and mutually beneficial interactions among employees in any organization, and to value employees with different backgrounds, needs, and skill sets in order to produce optimal benefits for employees, for the organizations they work for, and for communities and customers they serve”.

In addition to the apparent changes due to current demographics, the business environment also knows its challenges. Nowadays the marketplace has become increasingly complex, highly uncertain, competitive and transformational. To stand out, organizations need to develop unique selling points and to establish competitive advantages for sustained business success. Therefore, many organizations have faced or face organization change. Organization change derives from pressures such as the environment, new desires and visions of the future, or the discovery of deviations from standards (Palmer, Dunford, & Akin, 2009). To manage those pressures, organizational change enables adaption to the present and anticipates the future, improving structures and processes to increase effectiveness and efficiency (Palmer, et al., 2009).

Individuals provide organizations with competitive advantages (Subhash, 2012). Consequently, organizations increasingly rely on teams for decision making, innovation stimulation, problem solving and production (Kozlowski & Bell, 2003; Homan, Knippenberg, Van Kleef & De Dreu, 2007), because teams may outperform individuals in the quality of the decisions they make (Argote, Gruenfeld, & Naquin, 2000; Ilgen, 1999; Hinsz, Tindale, & Vollrath, 1997). Members of an organization have an extensive range of characteristics and backgrounds; gender, age and nationality can differ as well as functional or education background and work experiences. This diversity can enhance as well as disrupt team performance (Knippenberg, Ginkel & Homan, 2013) and, therefore, the need for research on the optimally employment of diverse teams has increased.

(5)

5

interactions can be overlooked. Therefore, within the concept of diversity, the phenomenon of faultlines has been introduced in literature.

Faultlines are “hypothetical dividing lines that may split a team into homogenous subgroups, usually based on multiple attributes” (Lau & Murnighan, 1998: p. 328). Faultlines emerge based on individual characteristics and can lead to subgroup forming. Previous research has outlined that faultlines influence team processes (e.g. conflict, cohesion), performance outcomes (e.g. decision making and team performance), and affective outcomes (e.g. satisfaction) (Barkema & Shvyrkov, 2007; Bezrukova, Jehn, Zanutto & Thatcher, 2009; Choi & Sy, 2010; Lau & Murnighan, 2005; Li & Hambrick, 2005; Polzer, Crisp, Jarvenpaa, & Kim, 2006; Rico, Molleman, Sánchez-Manzanares, & Van der Vegt, 2007; Sawyer, Houlette, & Yeagley, 2006; Thatcher, Jehn, & Zanutto, 2003). Nonetheless, dormant faultlines influence these processes and outcomes to a different extend than activated faultlines do. The difference is that dormant faultlines are objective demographic alignments while activated faultlines are the perceived subgroup distinctions by team members (Jehn & Bezrukova, 2010). Faultline researchers assumed that members of subgroups have similar perceptions of the subgroup distinctions and team-based interactions. However, this view ignores that individuals have different perceptions of reality. Even when they belong to the same subgroup, they may interpret faultlines differently. As Bandura (1997) already concluded, members of the same team may view their experiences differently. This view of asymmetric perceptions of faultlines is in line with the research of Jehn, Rispens & Thatcher (2010) about asymmetric conflict perceptions. They concluded that members of subgroups have differences in perceptions about the level of conflict, and that this asymmetry decreased performance and creativity. The relationship has been mediated by social processes and team atmospheres and indicates that asymmetric perceptions have consequences on both individual and team outcomes. Therefore, it is of importance to evolve from the focus of current research on dormant and activated faultlines as these studies were subject in past research. To date, there has been very little research conducted on the asymmetric perceptions of members in subgroups. Rupert & Jehn (2009) have noted that it is imaginable that asymmetrically perceived faultlines within teams will have different team dynamics than teams in which all members perceive the same level and type of faultlines. This line of thinking is supported by Swann, Polzer, Seyle, & Ko (2002) and Milton & Westphal (2005), who found that when team members do not share the same perceptions, confusion and inefficiencies in the team can occur. Therefore, this current research will take into consideration (1) to which extent faultline-based subgroups have asymmetric perceptions, as well as (2) an examination of the extent to which these asymmetries in perceptions might influence outcomes.

(6)

6

triggers that do not match with the Chrobot-Mason categories. However, to date, there has been insufficient research on organizational change as a faultline trigger (Gover & Duxbury, 2012).

According to analyses on organizational faultlines and change (Gover & Duxbury, 2012), it is worthwhile to analyse faultlines in change settings. Organizational change has a significant impact on teams, it can pull teams apart or it can unite them (Caldwell, 2013). For example, an introduction of a new working method within a team may result in subgroups of individuals. This formation of subgroups can be based on members who like the new way of working versus the others who resist working like that. Another scenario can be that the whole team does not like the new way of working and results in a united team. One of the few articles that link faultlines to organizational change is that of Gover & Duxbury (2012). Their research showed that organizational change can enable powerful social dynamics. These dynamics might have never been activated when there was no change at all. Furthermore, their research had shown that intergroup dynamics can influence the way in which individuals experience organizational change, because organizational members use social identities as a lens through which they view change. For instance, when an individual identifies with a specific team, in-team members can influence the individual without much effort, while at the same time the individual is less prone to the influence of out-team members. Relating this formation of inter-subgroups to organizational change, it is expected that subgroups may be formed on, for example, the opinions and experiences that team members have about change. These change-related faultlines will be further investigated in this study. While Gover & Duxbury (2012) used the faultline theory to achieve a better understanding of why organizational change is experienced differently throughout an organization, this research will use organizational change as a change context in which asymmetric faultline perceptions of team members are analysed.

In the context of organizational change, behavioural integration is a noticeable team process. Due to organizational changes, the daily work and life of employees are affected (Oreg, Michel & By, 2013). In turn, these individuals belong to teams and are therefore affected by these changes. Within teams, dynamics are present that influence individual change experiences (Gover & Duxbury, 2012). Simultaneously, teamwork enables participative team processes (Magni, Proserpio, Hoegl & Provera, 2009). As a result, it is important to examine what influence organizational change has on the team process of behavioural integration within teams.

Behavioural integration is defined as “the degree to which mutual and collective interaction exists within the team” (Hambrick, 1994: 188) and has three main constructs: collaborative behaviour, information sharing, and joint decision making. The united construct of behavioural integration describes the tendency for teams to engage in more “team like” behaviours than others (Hambrick, 1994). This research will examine behavioural integration as the individual perceptions of this process. Individuals differ in their way of viewing such processes (Bruner, 1957; Searle, 1997; Bandura, 2001). In addition, the different faultline perceptions of team members may influence their perceptions of behavioural integration. This study contributes to the literature on faultlines by investigating the cross-level effects of team-level interaction processes (e.g. behavioural integration) and individual perceptions. These Individual perceptions are particularly important as opinions are formed and actions occur at primary level.

(7)

7

Parthasaranthy, 2009). Therefore, it is interesting to examine what the asymmetry in perceptions of faultlines, mediated by the perceptions of behavioural integration, have on the perceptions of team performance.

In this study, an examination of asymmetric faultline perceptions that activate behavioural integration and ultimately influence team performance will be accomplished. This team process (e.g. behavioural integration) and team outcome (e.g. team performance) have been chosen for two reasons. First, in general, the variables are important team processes and outcomes that lead to the success and continuation of a team in a change context. Second, given the relatively new approach (i.e. the asymmetric perspective on faultlines), the team process and outcome variables considered have previously been examined by past faultline researchers. This enables a comparison between the findings of this study with those of past studies.

This studies’ contribution to theory is threefold: (1) to date, there has been very little research conducted to examine whether and if asymmetry of perceptions by team members changes any team processes and outcomes; (2) this research on activated faultlines will be conducted in an organizational workplace which is hardly researched yet; in other words, most research on activated faultlines have been conducted in a laboratory and in student work teams (Thather & Patel, in press; 2012); and (3) it contributes to the understanding of faultlines in an organization change context.

Outcomes of this research could provide managers insights on future challenges they might face when managing diversity and adapting their strategy accordingly. In addition, understanding what the effects of asymmetrical faultline perceptions are on team processes and outcomes can be very helpful for managers since they can respond more adequately and move forward toward more effective task outcomes (Jehn, de Wit & Barreto, 2008).

In sum, this study explores individual differences (asymmetries in perceptions) as well as situational influences both internal and external to the work team (organizational change) as determinants of team processes and outcomes.

The research question is therefore: “To which extent does asymmetry of perceptions of faultlines exist in teams and how does this affect behavioural integration and ultimately influence team performance?”

The following sub questions are formulated in order to be able to answer the research question: 1. To what extend does organizational change trigger asymmetric perceptions of team faultlines? 2. How do asymmetrical faultlines influence behavioural integration?

(8)

8

Literature and background

Organizational change context

Organizational change is a current issue, especially since the financial crisis that began late 2007 (Fugate, 2013). Due to instable economies, companies need to change in order to stay ahead of competitors and to make sufficient profits. Organizational change refers to “any adjustment or alteration in the organization that has the potential to influence the organization’s stakeholders’ physical or psychological experience” (Oreg, Michel & By, 2013, p. 4). Organizational changes include, for example, alterations to job descriptions, organizational structures and cultures implementations of new practices or geographical relocations. All of these changes have the potential to influence the organizations’ performance.

In most cases, organizational changes influence organization members as well. The consequences of change affect their daily work and life, such as their livelihood, sense of belonging and competence, and their overall well-being (Oreg, et al., 2013). For organization members change is most often experienced as threatening, although this depends on the scope of the change. Small changes will be experienced more threatening when teams consist out of subgroups with strong faultlines, (Chrobot-Mason et al., 2009) because team members are afraid that their strong feeling of team membership will be affected by this. When the particular change in the workplace is significant and meaningful, it may raise threatening feelings by the organization members (Chrobot-Mason et al. 2009).

At the same time, these organizational members play a crucial part in determining the level of change success (Oreg, Vakola, & Armenakis, 2011). Therefore it is important to understand the viewpoint of organization members about change. However, these viewpoints vary widely because, as Fugate (2013, p.15) described it; “individual reactions to change exhibit great diversity”. However, the perspective of this research is not that of the change recipient individually. The focus of the present study is on the perspective of a team as a whole. Organizations rely increasingly on teams because teams outperform individuals when it comes to, for example, decision making, solving problems and innovation. Within teams, dynamics are important. Organizational change can enable powerful dynamics such as self-categorization and self-enhancement, which lead to influencing individual change experiences (Gover & Duxbury, 2012). Individuals have a great variety in their reactions to change and are easily influenced by the team or subgroup to which they belong.

Prior to activated faultlines, an event, situation or circumstance must trigger dormant faultlines to become active (Rink & Jehn, 2010). Different kinds of triggers, such as characteristics of the team, task, and organizational context, may influence subgroups and the overall team in different ways (Thatcher & Patel, 2012). It is necessary to consider those triggers to explain potential underlying effects of diversity (Mannix & Neale, 2005). In the research of Gover & Duxbury (2012), organizational change was used as an event that explained the activation of faultlines. They found that organizational change significantly stimulates powerful social dynamics (e.g. self-categorization and self-enhancement) within teams. Therefore, it is assumed that the activated faultlines researched in this study, are triggered by organizational change.

Faultlines

(9)

9

team processes, performance outcomes, and affective outcomes (Bezrukova et al. 2009; for a review, see Thatcher & Patel, 2012). Furthermore, subgroup formation causes an “us-them” distinction. Faultlines are described as either dormant or active faultlines, where dormant faultlines can become active faultlines through an activation process.

Activation process. The distinction between dormant and active faultlines has become increasingly central to faultline literature in recent years (Rink & Jehn, 2010; Jehn & Bezrukova, 2010; Gover & Duxbury, 2012). A distinction has been made because dormant faultlines in teams can be inactive and go unnoticed for years without any changes in team processes (Lau & Murnighan, 1998). In other words, not all potential faultline situations are necessarily activated. This leads to the definition of dormant faultlines as “potential faultlines based on demographic characteristics” (Jehn & Bezrukova, 2010: 24). In contrast, active faultlines exist when “members actually perceive subgroups based on the demographic characteristics” (Jehn & Bezrukova, 2010: 24). The faultline activation process ensures that dormant faultlines transform into activated faultlines. Jehn & Bezrukova (2010: 24) defined the faultline activation process as “the process by which an objective demographic alignment (a potential, or dormant faultline) is actually perceived by team members as the division of the team into separate subgroups based on demographic alignment (an activated faultline)”. In sum, an external trigger or event can activate dormant into active faultlines. Both these dormant and activated faultlines can be based on different types of attribute alignment.

Faultline bases. The simultaneous study of alignments across team members allow focusing on different types of attributes. Most faultline researchers distinguished between social category attributes such as race/ethnic background, nationality, sex, and age (Cummings, Zhou & Oldham, 1993; Jehn, Chadwick, & Thatcher, 1997; Jehn, Northcraft & Neale, 1999; Lau & Murnighan, 1998; Thatcher et al, 2003; Shaw, 2004) and informational attributes such as function, education, tenure (Bezrukova et al., 2009; Molleman, 2005; Zimmermann, 2011; Zellmer-Bruhn et al, 2008). Although social category attributes may not be directly related to a given task, they do shape individuals’ perceptions and behaviours regarding in- and out-teams, through mechanisms of categorization, stereotyping, and prejudice (Messick and Mackie 1989). In contrast, informational characteristics are directly job related and are important in task completion (Jackson, Joshi & Erhardt, 2003; Jehn et al. 1997, 1999). Other researchers have studied faultlines on non-demographic attributes, such as, personality characteristics or types (Gratton, Voigt & Erickson, 2007; Molleman, 2005) or work location (Cramton & Hinds, 2005; Gokakker, 2007, Gratton et al., 2007; Polzer et al., 2006). Complementing the mentioned research, this study adopts a novel alignment attribute; organizational change as a non-demographic faultline base on which alignment may occur.

(10)

10

to the organizational change context, it is assumed that faultlines can also derive from change-related characteristics.

As shown by Gover & Duxbury (2012), simple contact event faultlines can result from organizational change. The present study proposes that differential treatment faultlines (Chrobot-Mason et al, 2009) can also result from organizational change. The manner in which organizational change is managed can lead to subgroups when teams receive unequal opportunities in the workplace or receive unequal treatment. Treatments, according to Chorbot-Mason et al. (2009), may include distribution of resources such as promotions, pay, opportunities, praise, and disciplinary actions. Applying this to managing organizational change, team members may perceive differences in the treatment that include unequal distribution of the information about change, unequal division of the effects of change; that are personal impact on team members, or unequal contribution by team members to change. This may lead to a higher or lower level of uncertainty among team members and whether they support or resist change (for a review see Oreg, et al., 2013). Furthermore, the theory of Lau & Murnighan (1998) indicated that the addition of (important) new members into established teams or leaving of (important) old members of the team is another factor that can lead to the activation of old faultlines and may even lead to significant changes in team’s dynamics. When organizations are subjected to organizational change, it is likely that teams will be reorganized and according to Lau & Murnighan (1998), this may influence the strength of faultlines. While previous research is mainly directed to objective demographic attributes, this study examines activated change-related attributes. More specifically, four different change-related attributes will be studied; resistance attitude, resistance behaviour, information, and personal impact.

First, individuals can be resistant or supportive to change. These two perspectives are interpreted as opposite ends on one continuum (van Offenbeek, Boonstra & Sao, 2012; Drzensky & van Dick, 2013). Different views exist of what resistance to change is. Therefore, Oreg (2006) conceptualized resistance to change as a multifaceted construct to capture employees’ subjective experiences to better understand the essence of resistance. His view is based on Piderit (2000), she suggested that resistance includes affective, cognitive and behavioural dimensions. In the present study, the affective and cognitive components are viewed as resistance attitudes. Resistance attitudes encompass feelings and thoughts about the change. In contrast, resistance behaviour is the response to the change through actions or the intention to act. Providing detailed information about the change has been shown to reduce resistance to change (Wanberg and Banas, 2000) and reduce uncertainty (Schweiger & DeNisi, 1991). In addition, support to change can be enhanced by providing timely, informative, and useful information about organizational change (Miller, Johnson & Grau, 1994; Wanberg & Banas, 2000). However, Oreg (2006) found that this positive influence of providing information is to a certain extent. He discussed that when individuals think the change will have a negative impact on them, more information actually yield increased resistance. Therefore, it is important to know whether team members are sufficiently or insufficiently informed about change. Finally, Wanberg & Banas (2000) defined personal impact as “the net perceived effect that a particular change will have on an individual or his or her working environment”. They found that perceived personal impact was not related to change acceptance or a positive view of change. In addition, Ashford (1988) found that direct personal impact of change increased greater stress experience. The present study examines the extent to which team members perceive they are personally impacted by change.

(11)

11

Therefore, this research does not take dormant faultlines into account but rather will focus on activated faultlines. More specifically, perceived diversity (i.e. activated faultlines) does not have to be perceived equally by team members.

Asymmetric perceptions

The process by which diversity is translated into thought and action in teams can be perceived asymmetrically by team members and in subgroups (Greer & Jehn, 2007). Perceptions are formed based on what team members notice about each other (Zellmer-Bruhn, et al. 2008). Zellmer-Bruhn, et al. (2008) reasoned that the context in which teams function may trigger great difference in perceived similarity; that is, team members may notice a wide variety in diversity attributes when affected by changes in the context.

Subgroups within faultline-based teams tend to be quite homogeneous (Lau & Murnighan, 1998). Those subgroups have in previous research been treated as if they have similar perceptions of team-based interactions, thereby overlooking that the perceptions of subgroup members can differ. Only when members perceive subgroups, it has been argued that faultlines are influential (Cramton & Hinds, 2005). The study of Homan & Greer (2007) confirmed this finding, they concluded that the perception of subgroups does predict team outcomes, whereas actual (e.g. dormant) faultlines do not. Of great importance is that perceived diversity does not have to match with the objective reality (Riordan, 2000) and these perceptions may change over time (Zellmer-Bruhn, et al. 2008). Even though it may not be the objective reality, perceived diversity can be influential in the interaction between individuals (Zellmer-Bruhn, et al. 2008; Straus, Barrick, & Connerly, 2001). When perception and reality do not align, cross-cutting categorizations (i.e. members of different subgroups sharing a common trait) could exist (Hogg & Terry, 2000; Crisp & Hewstone, 2001), which may reduce the psychological distances between different subgroups members (Brewer, 2000).

In addition, individuals differ in perceptions (Cunningham, 2007). According to Burnes (2009), reality can be socially constructed in two ways. Within teams, individuals use power and politics to achieve a dominant position over others, and it is their view of reality that takes shape and come to be accepted. Secondly, different interpretations of reality, put forward by competing groups of individuals, can be observed (Burnes, 2009). Social cognitive theory (Bruner, 1957; Searle, 1997) and the social information processing approach (Bandura, 2001) have been used to explain the different experiences of individuals in organizations; that is, different perceptions of the same reality. As Riordan (2000) noted, this perceptual approach is based on the assumption that persons who are demographically different than their colleagues attach psychological meanings to these perceptions. Not necessarily the actual differences themselves lead to subsequent attitudes and behaviours.

(12)

12

have asymmetrical views of conflict, compared to individuals who have symmetrical views of conflicts, perform less well, experience lower levels of motivation to work well on tasks and their attitudes towards work partners are less positive. They stress the importance of investigating the effects of asymmetry on different team processes and work outcomes.

Asymmetric faultline perceptions have not yet been researched and therefore no clear definition of asymmetric faultline perception exists. Related to the definition of conflict asymmetry by Jehn, Rupert, & Nauta (2006), faultline asymmetry will be defined accordingly, as the difference in perceptions of faultlines among the team members involved. Derived from the definition of asymmetrical conflict perceptions (Jehn et al., 2008), asymmetrical faultline perceptions is defined as the degree to which team members perceive different levels of faultlines. Hence, the definition applied in this research is: “the degree to which members of subgroups differ in perceptions of the strength of faultlines in their team”.

Finally, asymmetry in perceptions of faultlines can evolve from the type of faultline (faultline base), as well as from the strength of the faultline. Previous research has largely ignored the possibility that members may have different perceptions about the existing faultlines. These subgroups may have different perceptions of the meaning, importance, and strength of the faultline. The focus of this research is on asymmetric perceptions of faultline strength.

Faultline strength asymmetry. The extent to which faultlines actually cause a team to split into subgroups is the strength of the faultline (Jehn & Rink, 2010). According to Lau & Murnighan (2005), homogeneity within and heterogeneity across teams are indicators of faultlines strength. Lau & Murnighan (2005) described three factors that influence faultline strength; (1) the number of individual diversity attributes visible in the team, (2) the alignment of the attributes; in other words, the extent to which individual team members are able to classify themselves into categories based on these attributes, which results in (3) the number of potentially homogeneous subgroups. Hence, team faultlines are proposed to increase in strength when more diversity attributes within a team converge in such a way that they split a team into homogenous subgroups (Meyer & Schermuly, 2012).

Past research extensively examined faultline strength, and largely concluded that in teams with strong faultlines, the negative effects of diversity generally outweigh the positive effects (e.g. Rico et al., 2007; Gibson & Vermeulen, 2003; Lau & Murnighan, 1998; Thatcher et al. 2003). Rico et al. (2007) demonstrated that strong faultline teams (highly homogeneous work teams) made lower quality decisions and had lower social integration (i.e. team member satisfaction and commitment) than weak-faultline teams. Bezrukova et al. (2009) found that the strength of social category faultlines was negatively associated with team performance. The study of Molleman (2005) revealed that the higher the strength of demographic faultlines present, the less cohesion and the more intra-team conflict is reported by team members. Thather & Patel (in press) found that greater relationship conflict, task conflict, and lower team cohesion are a result of strong demographic faultlines. In turn, strong demographic faultlines directly reduce team performance and team satisfaction. However, these effects are stronger for studies in lab settings than for studies in field settings (Thatcher & Patel, 2012). When faultlines are strong, it is more likely that the focus of the individual team members is directed toward similarities within their respective subgroup, than toward dissimilarities between subgroups (Molleman, 2005). Strong faultlines provide great opportunities for work teams to split up in subgroups, which show the importance of attributes and enlarge the effects of external forces (Chrobot-Mason, et al., 2009).

(13)

13

aware of subgroups than the individual who perceives weak faultlines. A distinction can be made between asymmetric perceptions between subgroups in teams and asymmetric perceptions between individuals within teams. The latter is the focus of the present study: asymmetry in perceptions of faultline strength exists among team members within a subgroup.

Behavioural integration

Individual perceptions are likely to have an influence on team interaction processes, since diversity among team members enables the opportunity to make use of a broad range of ideas, skills, and insights that exist (Cox 1993; Cox & Blake 1991; Ely, 2004). Teamwork provides the opportunity to engage in a participative team process through which members handle issues, make decisions and build commitment to implement these decisions (Magni, Proserpio, Hoegl & Provera, 2009). Individuals differ in their way of viewing such processes (Bruner, 1957; Searle, 1997; Bandura, 2001); therefore the emphasis is placed on how individuals perceive behavioural integration. Behavioural integration is defined as “the degree to which mutual and collective interaction exists within the team” (Hambrick, 1994: 188) and has three main constructs: information sharing, collaborative behaviour, and joint decision making. The different constructs refer to the degree to which, and how, team members interact, while the united construct of behavioural integration describes the tendency for teams to engage in more “team like” behaviours than others (Hambrick, 1994). Other related constructs are communication frequency (Smith, Smith, Olian, Sims, O’Bannon, & Skully, 1994), collaboration (Chatman & Flynn, 2001; Wagner, 1995), information sharing (Bunderson & Sutcliffe, 2002), and social interaction (Chatman, Polzer, Barsade, & Neale, 1998). However, behavioural integration entails more than that. Behavioural integration consists of both substantive and task-related interactions (e.g. information sharing and joint decision making) as well as a social dimension (e.g. collaborative behaviour) (Hambrick, 1994). In the following, the three constructs of behavioural integration will be revised separately.

Information sharing. The first construct, information sharing, refers to the quantity and quality of information exchange among team members (Lin, Hsu, Cheng, & Wu, 2012). Each team member has different tasks within the team, and it is trivial to exchange the information in order to be a well-functioning team. A well-functioned team can diagnose problems, develop strategies or generate solutions, and implement these accordingly (Tiwana & McLean, 2005). Especially when tasks are highly interdependent, it is important that individually held information and know-how is converted into common stock knowledge to solve problems and accomplish tasks (Tiwana & McLean, 2005). This can be achieved through social interactions, and ultimately improves work efficiency (Robert, Dennis & Ahuja, 2008). Exchange of information is more effective (e.g. open and timely) among team members in behavioural integrated teams than in behavioural disintegrated teams. This advantage enables team members to obtain and access relevant information, knowledge and complementary skills more easily (Faraj & Sproull, 2000). Subsequently, this results in more accurate and complete information between members regarding individual members‘ contribution and overall work progress. Team members have up-to-date information because they communicate the necessary information more fluidly (Li & Hambrick, 2005), since team members know who to contact for which information.

(14)

14

Joint-decision making. Lastly, joint-decision making is a mechanism in which information exchange is achieved through social interaction and collaborative behaviour (Sheremata, 2000). Team members collaboratively make decisions related to teamwork processes and task contents (Lin, Hsu, Cheng, & Wu, 2012). When team members engage in joint-decision making, they are able to develop a deeper understanding of the needs of other team members, resulting in richer communication between members and an increase of their awareness of the team (van de Ven, Delbecq & Koenig, 1976). Collectively derived decisions receive higher commitment and follow-up from members of a behavioural integrated team (Carmeli & Schaubroeck, 2006), since team members rely on the contributions and support of the other members. They obtain more immediate feedback on their actions (Moorman & Miner, 1998) thus enhancing their awareness of the decision alternatives and consequences (Magni, Proserpio, Hoegl & Provera, 2009). Therefore, behavioural integration is important in ongoing work teams. In sum, insufficient expertise integration reduces the capacity to understand the particular problem at hand; the lack of collective mind diminishes the efficiency of action-taking and a cohesive teamwork climate cannot be attained when decisions are not made with all members.

Linking faultline theory to the constructs of behavioural integration, Williams & O’reilly (1998) found that cooperative interactions within homogeneous teams will be high, as member attraction is high and categorizations of team members are few. Ely (2004) concluded that diversity can improve problem solving and decision making. However, individuals tend to make a distinction between in-team and out-team members which in turn may lead to miscommunication (Ely 2004). Smith et al. (1994) found that heterogeneous teams are negatively related to informal communication. Furthermore, the diversity on which faultlines are based hinders task-related processes and reduces information exchange (Ancona & Caldwell, 1992). In addition, diversity might also result in limited understanding and tension between team members, thereby reducing the likelihood of joint-decision making (Michel & Hambrick, 1992). The results of Simsek, Veiga, Lubatkin & Dino (2005) showed that educational and goal preference diversity influenced behavioural integration negatively, while tenure and functional diversity do not significantly influence behavioural integration. Furthermore, in teams with strong faultlines, the negative effects of diversity generally outweigh the positive effects (Rico et al., 2007; Gibson & Vermeulen, 2003; Lau & Murnighan, 1998; Thatcher et al. 2003). Rico et al. (2007) demonstrated that strong faultline teams (i.e. highly homogeneous workgroups) make lower quality decisions and have lower social integration (i.e. team member satisfaction and commitment) than weak-faultline teams. Furthermore, strong faultlines have negative effects on behaviours that affect team performance such as information elaboration (Meyer, Shemla, & Schermuly, 2011) and the riskiness of decision-making (Barkema & Schvyrkoc, 2007; Rico, et al. 2007). From these findings, it can be concluded that faultlines based on diversity negatively influence behavioural integration within a team. Though, within homogeneous subgroups behavioural integration is more present.

(15)

15

This study does not focus on kinds of behaviours team members might show, or how behavioural integrated team can be observed. Instead, this study examines team members’ opinions and views on information sharing, collaborative behaviour and joint-decision making and what their perceptions are about the extent to which their team is behaviourally integrated. Past research examined the actual behaviour of team members, while largely ignoring the fact that different individuals are involved in behavioural integration. These individuals may have different perceptions about the level of behavioural integration. Particularly, regarding asymmetric perceived faultlines, a key importance is to examine the perceptions of behavioural integration, as individual’s perception of behavioural integration may differ. It has been argued that individuals in particular work situations expect their peers to have similar interpretations. When these expectations are contradicted, individuals are likely to become confused, causing negative outcomes such as decreased effort and miscommunications (Milton & Westphal, 2005; Polzer, Milton, & Swann, 2002). Furthermore, in line with the described research findings, diversity and faultlines generally have a negative influence on behavioural integration. Therefore, asymmetric views on faultline strength are negative influences on behavioural integration. Derived from above results, the following is proposed:

H1: “The more a team perceives change-related faultlines asymmetrically, the more negative their perceptions of the behavioural integration process”.

Team performance

Previous studies have thoroughly investigated the consequences of faultlines on team outcomes such as team performance and found both positive and negative effects (for a review see Thatcher & Patel, 2012). Subgroups based on diversity attributes are positively related to feelings of trust or team cohesion (Lau & Murnighan, 2005). However, the more the diversity characteristics are related to a specific task, the more likely conflict is to appear between subgroups, while team performance within the subgroups increases (Jehn, et al. 1999). Furthermore, once activated, faultlines can lead to negative team behaviours such as out-group bias, coalition formation, and conflict that hinder team performance (Rink & Jehn, 2010). This finding is confirmed by Thatcher & Patel (2012) in that the relationship between faultlines and team performance is mostly negative. Furthermore, past research has shown that this relationship is stronger for activated faultlines than for dormant faultlines. In other words, faultline strength explains more variance in performance than diversity measures alone (Thatcher & Patel, 2012). Faultline strength tends to have a negative effect on many types of performance outcomes, including supervisor-rated performance, decision-making quality, and archival team outcomes (Thatcher & Patel, 2012). Bezrukova et al. (2009) argued that the strength of social category faultlines is negatively associated with team performance. Overall, teams with strong faultlines, regardless of their composition, show low levels of performance (Thatcher & Patel, in press).

Behavioural integration has a positive effect on team performance (Lin, Hsu, Cheng, & Wu, 2012). It enables teams to respond to demand, to create competencies, to develop strategies, to increase productive innovation intensity, to improve decision quality and to make better use of knowledge alternatives (Hambrick, 1998; Carmeli & Schaubroeck, 2006). Similarly, behavioural disintegration generally causes team problems, such as failure to exchange key information, poor coordination of activities, and difficulty in formulating and implementing responses to environmental shifts (Hambrick, 1995). Furthermore, the research of Li & Hambrick (2005) concluded that behavioural disintegration is a negative influence on team performance, as it mediates the triggering effects of faultlines.

(16)

16 Organizational change

Figure 1: Conceptual model

team to effectively accomplish a predefined goal (Lin, Hsu, Cheng, & Wu, 2012). Through collaborative behaviour, teams can avoid unnecessary work, reduce inefficient communication and maximize outcomes of effort (Lin, Hsu, Cheng, & Wu, 2012). As a result, tasks can be completed more efficiently (Lin, Hsu, Cheng, & Wu, 2012). Furthermore, when team members take others’ actions into consideration while making decisions, conflict can be avoided and resources can be utilized effectively. However, in addition to performing their individual tasks, members need to obtain a full understanding of the team tasks to achieve collaborative behaviour. Finally, joint-decision making theory stated that members are more committed to an outcome when it is derived collectively. Joint-decision making also leads to satisfaction, loyalty, productivity, and positive leader-member relationships (Bruine de Bruin, Parker & Fischhoff, 2007). The process creates a sense of belonging for each individual and creates a more cooperative and less competitive environment (Sagie & Aycan, 2003). The second hypothesis is therefore:

H2: “The more a team perceives behavioural integration, the more positive their perceptions of team performance”.

The perceptions of team members play a central role in this study. Therefore, team performance will be researched from the viewpoint of the team members; that is, the perceived team performance by team members. Coherence between the perceived team performance and actual outcomes are expected to be observed. Perceptions of performance are negatively affected by faultlines (Kunze & Bruch, 2010; van Oudenhoven-van der Zee, Paulus, Vos, & Parthasaranthy, 2009). When team members have different perceptions of reality (e.g. asymmetric perceptions), they are not likely to validate others’ perceptions and the positive effect on outcomes such as performance disappear (Mannix & Jehn, 2001). Jehn, Rispens & Thatcher (2010) argued that conflict asymmetry decreases performance; the effects are mediated by social processes. In the first hypothesis, it is expected that the perceptions of behavioural integration are negatively affected by asymmetric perceptions of faultline strength. Subsequently, the expectation is that the perceptions of team performance will be negatively influenced by asymmetric perceptions of faultline strength, mediated by perceptions of behavioural integration. Therefore, the final hypothesis is as follows:

H3: “A team’s perceptions of their behavioural integration mediate the effect of change-related faultline asymmetry on their perceptions of performance”. That is, members who perceive higher levels of change-related faultlines in a team are less likely to experience positive behavioural integration and are thus less likely to experience increased performance than members who perceive lower levels of change-related faultlines.

The conceptual model of this study in which the variables of interest are shown:

(17)

17

Methodology

The research approach applied in this study was the theory testing approach, described by van Aken, Berends & van der Bij (2012) in their three-step approach. Firstly, theory testing acknowledges that the business phenomenon that is experienced by many companies has not yet been fully examined in academic literature. The business phenomenon in this study was twofold: subgroup forming based on change-related faultlines and how asymmetric faultline perceptions influence the perceptions of behavioural integration and team performance in an organizational change context. The second step is to establish a conceptual model and hypotheses based on past literature streams as described in the theory chapter of this research. The final step is to conduct quantitative research. In this study, primary data had been retrieved from questionnaires and data was analysed with statistical methods in order to confirm or reject the hypotheses and to provide theoretical and managerial implications. The present study was a representation of one point in time, thus a cross-sectional study and occurred under actual field conditions. A quantitative method of analysis was most appropriate for this research, since variables and relationships were known. However, evidence on the theoretical explanations of the associations among the business phenomenon were incomplete. Furthermore, the quantitative approach suited the research objective of obtaining a representative view of asymmetric faultline perceptions best. The cross-sectional method allowed for data gathering on multiple teams in a short timeframe. The analysis of multiple teams supported generalisation of the results to a larger sample (Cooper & Schindler, 2008). In contrast, a qualitative approach investigates more profoundly into the mechanisms and relations between variables, and seeks answers to motivations, opinions and emotions behind these mechanisms and relations. Qualitative techniques search for answers to ‘what’ questions (Cooper & Schindler, 2008) or ‘how’ and ‘why’ questions, whereas quantitative techniques on the other hand, thrive to seek answers to the question of ‘how much’ (Yin, 1989). As Oppenheim (1992) stated, “the need for […] a questionnaire […] arises whenever we wish to generalize from our findings, either in terms of the frequency or prevalence of particular attributes of variables, or about the relationship between them”. Therefore, a questionnaire was the suitable research method for this study.

Data collection method

(18)

18

maintenance (12 teams; 3.9%), consultancy (10 teams; 3.2%), customer contact centre (4 teams, 1.3%), and other (20 teams; 6.5%). The number of members varied among the teams, with a mean team size of 7.34 (SD = 3.42). The mean age of the participants was 39.6 years (SD = 11.4) and 47.2% was female. Of the 318 team members, 197 team members were fulltime employed and the mean team tenure was 40.62 months (SD = 58.14). The majority of the respondents were born in The Netherlands and had a Dutch nationality, 293 and 308 respondents respectively. The average educational level was MBO (32.3%), followed by HBO (25.3%), LBO (14.9%), WO (11.7%), MAVO (10.8%), and primary and secondary school (5.1%). Organizational changes varied from mergers/ acquisitions (42.1%), geographical relocations (21.9%), and reorganizations/ centralizations (16.5%) to alterations to procedures/ routines (11.2%), alterations to collaborations between departments/ teams (6.8%), forced resignations (0.7%), and adding an extra management layer (0.7%). These changes have been started, on average, 13.5 months ago since December 2013 (SD = 11.74) and will end, on average, within 26.7 months as from January 2014 (SD = 44.23). However, many respondents did not know when the change started (56.6%) or will end (77.4%). Only 142 respondents (44.7%) knew in which phase the change was; most of the changes were in the middle phase (17.3%) or end phase (15.4%) of the change.

Organization Frequency Percent

Government agencies 79 24.8%

Health care 72 22.6%

Industrial & building services 40 12.6%

MBO/ University 35 11.0%

Hospitality 22 6.9%

Housing corporation 6 1.9%

Missing 20 6.3%

Total 318 100.0%

Table 1: Overview of organizations that participated in this study

The criterion for teams to participate had been restricted to the definition of a team by Katzenbach & Smith (1993). They defined teams as “a number of individuals (2–25) with complementary skills who are committed to a common purpose, performance goals and approach for which they hold themselves mutually accountable”. In addition, team members achieve common goals through interaction with each other and well-defined and interdependent roles can be observed (Buelens, Sinding, & Waldstrom (2011). In order to calculate faultlines, the team size could not exceed more than 15 members (Thatcher, et al., 2003). Therefore, the definition of teams by Katzenbach & Smith (1993) was adjusted according to the criterion of Thatcher, et al. (2003). Due to the team focus of this research, it was desired to collect responses of all team members to gain the highest possible response rate. Another criterion for teams to participate was related to the context of this study. Teams should have been exposed to, and influenced by organizational change in the last 3 years. Organizational change can be described as “any adjustment or alteration in the organization that has the potential to influence the organization’s stakeholders’ physical or psychological experience” (Oreg, et al., 2013, p. 4).

(19)

19

access to the web-based questionnaire. In total, two reminders were send by e-mail to increase the response rate. The first reminder was send after two weeks and the second reminder after a month. In all other cases, the participating teams were visited and paper versions of the questionnaire were provided. It was requested that the team members should respond to the questionnaire independently from each other. Therefore, they were asked to fill in the questionnaire in another room. Within one week, the completed questionnaires were collected. Team members who did not completed the questionnaire yet were asked to finish it directly to increase the team’s response rate.

The data collection consisted of self-administered questionnaires with validated scales. Teams were asked to fill out questionnaires regarding organizational change and team collaboration. A self-administered questionnaire is a survey technique that allows respondents to complete it without having contact with the interviewer (Dillman, 2000). Advantages of this approach are optimal time efficiency and time and place convenience for respondents. Furthermore, the respondents could participate anonymously, diminishing barriers to openness. A drawback of this approach is the non-response error. Nonetheless, this error was partly overcome by sending reminders. In conclusion, (web-based) questionnaire was the most appropriate way to conduct this research.

Measures used

The questionnaire was formulated on the basis of previous studies and consisted of 33 items. The items were withdrawn from earlier published scales to maintain reliability (van Aken, Berends, & van der Bij, 2012), thereby eliminating the need for a pre-test. The respondents were Dutch speakers, and therefore the survey was distributed in Dutch. The scale items can be found in appendix A. The team was the level of analysis. Data was collected from 318 individuals and were aggregated to team level (62 teams).

The Likert-scale was applied, since this method of measurement provided the opportunity to compare one person’s score with a distribution of scores from the sample team (Cooper & Schindler, 2008). More specifically, this research aimed to compare the asymmetric perceived strength of faultlines between team members. This was calculated by taking the standard deviation of the scores. Another advantage of the Likert-scale was that respondents answered each item and therefore it was more reliable and provided more data than other scales. Error of leniency, error of central tendency, and the halo effect were taken into consideration (Cooper & Schindler, 2008). Error of central tendency was limited by using a multiple point scale, therefore this study used the 7-point Likert-scale. For all variables and corresponding items measured in this study, participants were asked to indicate their answer on a 7-point Likert scale, indicating to what extent they agreed with the statements, ranging from ‘strongly disagree’ (1) to ‘strongly agree’ (7). In this study, the halo effect was limited by clear definitions and descriptions of items, did not involve interactions with others and the items were not a trait of high moral importance.

(20)

20

The instrument of the change-related variable was still in development while conducting this research, which meant that two different, yet complementary, sets of questions were distributed to the respondents. Items were added in the second questionnaire to increase validity and reliability of the scales. The first questionnaire contained 6 items and the second questionnaire 8 items. Therefore, in the second version, the mean is computed over multiple items.

The resistance items (e.g. attitude and behaviour) were combined into one scale and had a scale’s reliability of .869. Similarly, the items of informed and personal impact were united in one scale with a Cronbach’s alpha of .922 and renamed into the variable change aspect (e.g. what changed and how much (Caldwell, 2011)).

Asymmetric faultline perceptions were measured according to past research measuring concepts of conflict asymmetry. Asymmetric faultline perceptions were assessed as the standard deviation among team members’ faultline perception scores (Jehn, Rispens, & Thatcher, 2010). The larger the score, the bigger the difference in perceptions of team members. For every item, the minimum and maximum score ranged between 1 and 7. The descriptive statistics of the perceived change-related faultlines can be found in table 2.

Resistance attitude 1 Resistance attitude 2 Resistance behaviour 1 Resistance behaviour 2

N Valid 311 315 313 316

Missing 7 3 5 2

Mean 2.52 2.72 2.84 2.32

Std. Deviation 1.61 1.67 1.66 1.59

Informed 1 Informed 2 Personal impact 1 Personal impact 2

N Valid 312 143 314 142

Missing 6 175 4 176

Mean 2.54 2.41 2.70 2.40

Std. Deviation 1.63 1.57 1.75 1.58

Table 2: N’s, Means and Standard Deviations of perceived change-related faultlines

Dependent variables. Behavioural integration The constructs of collaborative behaviour and joint-decision making were assessed by the index developed by Hambrick (1994) and adjusted by Simsek et al. (2005). The measurement of Jiang, Jackson, Shaw & Chung (2012) was used to measure information exchange. In total, 11 items measured the degree of behavioural integration of a team; 3-items measured collaborative behaviour (e.g. when a team member is busy, other team members will often help), 5-items measured information exchange (e.g. when the team is cooperating, almost everyone contributes with helpful ideas) and 3-items measured joint-decision making (e.g. team members will inform each other when their own work will have an influence on the work of other team members).

The final scale variable was separated into 2 constructs. Collaborative behaviour included all 3-items with a Cronbach’s alpha of .859 indicating sufficient scale reliability. The other construct contained both information exchange and decision making. Item 1 of information exchange and item 2 of joint-decision making were excluded, and the total construct showed high internal consistency (α = .901).

(21)

21

Control variables. To raise internal validity, control variables were included in this research. Similar to other faultline research, goal similarity, task interdependence, and team identification were used. Team identification was defined as “the emotional significance that members of a given team attach to their membership in that team” (van der Vegt & Bunderson, 2005). A 6-item scale was used to measure team identification (e.g. I consider my team important). Item 4 was discarded and increased the reliability from .695 to .739.

Task interdependence dealt with team members’ interaction and dependence on other members to accomplish the work (Campion, Medsker, & Higgs (1993). Campion, et al. (1993) established a scale to measure task interdependence and their 3-item scale was included in the questionnaire (e.g. Within my team, the tasks carried out by team members are all related to each other). In this study, the 3-item scale resulted in a Cronbach’s alpha of .796.

Finally, the degree of agreement concerning the purposes behind the activities of members of a team was captured in goal similarity (Jehn, 1995) and was measured according the 3-item scale developed by Jehn (1995) (e.g. As a team, we share the same goals). With a Cronbrach’s alpha of .862, the item showed high internal consistency.

Data analyses methods

With regard to multicollinearity, the variance inflation factor (VIF) had to be smaller than 10 and the tolerance level had to be greater than .10, to pass the multicollinearity test. All VIF and tolerance values in the regression analyses were acceptable. Thus showed a limited potential to adversely affect regression estimates, and added more stability to the standard errors and Beta weights (de Heus, et al., 1995). In order to decrease the possibility for multicollinearity in the mediation analysis, a centred variable was created for the independent and mediation variable; asymmetric perceived change-related faultlines and behavioural integration respectively.

Factor Analysis

A factor analysis was executed to produce above described scales and to analyse whether the variables measured are in fact distinguishable as separate variables. A precondition for performing a factor analysis is that for each item, at least 5 respondents are present. The total of 318 respondents was significantly higher than the minimum of 165 respondents needed (33 items times 5 respondents). In addition to the factor analysis, a reliability analysis was performed in order to analyse whether the content of the questions - that belonged to the same variable - are in fact measuring that variable. The Cronbach’s alpha of .70 or higher was preferred and generally it should not be lower. However, a Cronbach’s alpha of .60 or higher for one of the constructs was just acceptable and less detrimental.

A factor analysis was performed based on Eigenvalue greater than 1. A varimax rotation was applied to search for a basis that most carefully represented each individual. In addition, to assign items to a particular component a minimum of .30 was required. For items that had double loadings, an in-between difference equal to or less than .20 was accepted.

To start with, all items covering all variables were entered into the factor analysis. This resulted in a factor analysis of 33 items (appendix B, table 9). Due to the large amount of factors, it became difficult to load each variable on separate components. Therefore, factor analysis was performed in two steps. The first factor analysis included the independent and dependent variables together, and a second factor analysis included the control variables.

(22)

22

significant score for Bartlett’s Test of Sphericity, factor analysis was applicable for the included items (appendix B, table 10). The factor analysis of the independent and dependent variables resulted in 4 components. Although theoretically, 5 components were expected. In literature, information exchange and joint-decision making are distinct variables while the factor analysis showed that the items of these two variables loaded on the same component. Considering the questions related to the items of information exchange and joint-decision making, it was reasonable to accept this combination of variables, since both variables overlap in content. As Hambrick (1994) described, behavioural integration consists of both substantive and task-related interactions (e.g. information sharing and joint-decision making) as well as a social dimension (e.g. collaborative behaviour), and therefore the combination of information sharing and joint-decision making was validated.

Component Items 1 2 3 4 Resistance attitude 1 .85 .02 .00 -.39 Resistance attitude 2 .87 .00 .03 -.33 Resistance behaviour 1 .77 -.15 -.17 -.25 Resistance behaviour 2 .92 -.08 -.11 .04 Informed 1 .89 -.06 -.04 -.14 Informed 2 .90 -.06 -.03 -.04 Personal impact 1 .88 -.11 -.22 .10 Personal impact 2 .85 -.09 -.27 .17 Collaborative behaviour 1 -.15 .14 .90 .12 Collaborative behaviour 2 -.08 .24 .87 .17 Collaborative behaviour 3 -.15 .37 .76 .18 Information exchange 2 -.18 .76 .29 .10 Information exchange 3 -.08 .89 .12 .03 Information exchange 4 -.06 .83 .15 .04 Information exchange 5 -.08 .89 .10 .07 Joint-decision making 1 -.13 .65 .33 .29 Joint-decision making 3 .16 .54 .05 .33 Performance 1 -.22 .27 .27 .62 Performance 2 -0,2 .24 .38 .61

Table 3: Items of the independent and dependent variables

(23)

23

items still loaded on one component. However, based on theory, the resistance scales of attitude and behaviour can be combined. Furthermore, a combination can be made of the informed and personal impact items and was renamed to change aspect. The final factor analysis can be found in table 3.

The factor analysis of the control variables was appropriate with a KMO of .719 and a significant score on the Bartlett’s test of Sphericity. The factor analysis of Eigenvalue < 1 resulted in 4 components (appendix B, table 11). Team identification was divided into two components, contradicting the theory in which team identification is one component. When forcing the variables into 3 components, it led to loadings of items 1, 2, and 3 of team identification and all items of goal similarity on 1 component. Referring to the corresponding survey questions, both variables covered questions about the importance of being part of a team. However, this combination was not preferred, and therefore the first outcome of the factor analysis was examined. Within this factor analysis, the item “team identification 4” (I have criticism on the team) loaded on goal similarity. When observing the Corrected Item-Total Correlation, the item did not show high commonality with the other items in the construct of team identification (.204). Furthermore, the Cronbach’s alpha if item deleted increased to .734 compared to α = .695 of the total construct. Therefore, this item was deleted. The separation of team identification can be explained by consulting the corresponding questions; questions 1, 2, and 3 were positively formulated, while questions 5 and 6 were negatively phrased (reversed coded items). Therefore, one construct was created of team identification. The factor analysis for the control variables can be found in table 4.

Component Items 1 2 3 4 Team Identification1 .17 .76 .10 .04 Team Identification2 .29 .84 .04 .11 Team Identification3 .13 .89 .02 .13 Team Identification5 .06 .14 .01 .86 Team Identification6 .17 .07 -.05 .91 Task Interdependence1 -.10 .05 .88 .03 Task Interdependence2 -.01 .12 .91 -.03 Task Interdependence3 .39 -.02 .70 -.06 Goal Similarity1 .82 .18 .05 .10 Goal Similarity2 .91 .17 .06 .06 Goal Similarity3 .77 .26 .01 .14

ASYMMETRIC FAULTLINE PERCEPTIONS Effects of Asymmetrically Perceived Change-related Faultlines on Behavioural Integration and Team Performance

Effects of Asymmetrically Perceived Change-related Faultlines

on Behavioural Integration and Team Performance

By

Lianne Veldman

University of Groningen - Faculty of Economics and Business

MSc. Business Administration - Change Management

Lianne Veldman

Violenstraat 21

9712 RD Groningen

0615687757

l.veldman@student.rug.nl

Student number: 2227282

1

supervisor: Dr. Joyce Rupert

2

supervisor: Dr. Hanneke Grutterink

Abstract

Table of contents

Introduction

Literature and background

Organizational change context

Faultlines

Asymmetric perceptions

Behavioural integration

Team performance

Methodology

Data collection method

Measures used

Data analyses methods

Factor Analysis