The effect of activated faultlines
in a change context on group
dynamics and its effects on
change effectiveness.
By: O. Scholten
S1904426 University of Groningen
Faculty of Economics and Business MscBa, Change Management
Abstract
Faultlines are hypothetical lines that may divide a group into relatively homogenous subgroups based on the alignment of diversity attributes. These faultlines can be triggered due to an organizational change. Existing research mainly focused on dormant faultlines while this research focuses on the activation of faultlines. This research extends existing literature by focusing on the context of change in relationship with the activation of faultlines indicating that differences between group members become salient and may affect group feelings, actions, processes and performances. This research included 33 organizations, 80 teams and 506 participants and hypothesized that change will serve as a trigger for the activation of faultlines and will negatively affect the effectiveness of a change project. In addition to this, it is hypothesized that conflict and behavioral integration are a mediator for the relationship between activated change faultlines and change effectiveness. In line with was hypothesized, activated change faultlines did indeed lower the degree of effectiveness of a change project. Furthermore it was found that conflict served as a mediator for the relation between activated change faultlines and change effectiveness.
Keywords: Faultlines, Diversity, Activation of faultlines, Change, Conflict, Behavioral
Table of Content Abstract ... 1 Introduction ... 3 Theory ... 6 Faultlines ... 7 Activation of faultlines ... 8
Consequences of activated change faultlines on change effectiveness ... 11
Mediating role of the group processes conflict and behavioral integration. ... 13
Methodology ... 17
Research Method... 17
Data Collection ... 17
Sample description ... 19
Measurement of the variables ... 19
Data Analysis ... 22
Results ... 26
Descriptive Statistics and Correlations ... 26
Introduction
Society, organizations and therefore groups in organizations became more and more diverse over the years and diversity will continue to increase in the future (Triandis, Kurowski & Gelfand, 1994; Williams & O’Reilly, 1998). According to Van Knippenberg, Dawson, West and Homan (2011: p.64) ‘diversity refers to the degree to which there are similarities and differences between members of a team or group’. Van Knippenberg’s et al., (2011) definition depicts the objective side of diversity. In this study the following definition of diversity, a more subjective side, will be used as well: Every characteristic which people feel that they are different from one other (Mannix & Neale, 2005; Williams & O’Reilly, 1998, p.81). This definition implies that differences between people could occur in many areas and that these differences are created in minds of the individuals themselves. This also indicates that this perception is individual, subjective and situational (Rupert, 2012).
Due to this increasing diversity in society labor forces become more diverse as well, a better understanding in which circumstances diversity leads to positives outcomes will be important to managers and their organizations (Milliken & Martins, 1996; Lau & Murnighan, 2005; Rupert, 2012). A variety of diversity attributes may become more important in a specific context. This research will focus on how diversity attributes are influenced in a context of change. Although group composition and its effect on groups processes and outcomes has been central in the study of organizations (Jehn & Bezrukova, 2010), the research on group diversity has led to mixed outcomes (Horwitz & Horwitz, 2007; Webber & Donahue, 2001).
As organizations more and more rely on teams instead of individuals it becomes significant for firms to understand which factors influence performance in groups (Vora & Markóczy, 2012). However every group is different and each individual reaction towards external/internal forces and events, changes, situations and occurrences is triggered by different factors. Because of the inconsistency in diversity research, with positive and negative outcomes, Williams & O’Reilly (1998) stated that a more complex framework concerning team diversity, which address contextual aspects as well as different types of diversity, must be developed. Lau & Murnighan (1998: p.325) introduced the concept; ‘group faultlines, which depends on the compositional dynamics of multiple demographics attributes (age, race, gender, status, language) that can potentially subdivide a group’. Faultline theory does include the complex aspects of diversity that previous studies had neglected by focusing on the influence of multiple dimension of diversity instead of one dimension. The faultline theory (Lau & Murnighan, 1998) suggests hypothetical dividing lines within a group can subdivide a group into subgroups based on specific attributes. It is argued that individuals within a group possess a variety of attributes at the same time and that these attributes may align across multiple attributes. Alignment of attributes amongst group members can lead to subgroup formation which in turn impairs team processes and outcomes (Bezrukova, Thatcher, Jehn & Spell, 2012).
change might make individuals understand their differences and similarities with their group members resulting in activation of faultlines and subgroup formation which impairs group’s performance. Therefore it is hypothesized that activated change faultlines negatively influence change effectiveness and that this relation is mediated by the group processes conflict and behavioral integration.
This study will contribute to change as well as to faultline literature by studying what the influence is of activated faultlines on group dynamics and performance in a context of change. In addition to this, this research will provide insight in the underlying processes that influence team diversity and team performance. A practical implication is that [change] managers can use this insight to see what the effects of diversity and faultlines in teams are on the effectiveness of the change. Furthermore, this study will focus on activated faultlines, whereas previous studies mainly focused on dormant faultlines and their effects on team processes and performance. The difference between dormant and activated faultlines is that dormant faultlines are unnoticed and consequently have no direct effect on group processes and change effectiveness (Lau & Murnighan, 1998). Managing diversity is a significant organizational challenge, and therefore managers must acquire the managerial skills needed in diversified work environment (Mazur, 2010) that helps them making organizational changes a success. The influences of activated faultlines on group dynamics and performance will be dealt with in this study. Another gap that this study tries to fill is that the specific context of change lacks research in combination with activated faultlines.
Research Question
As mentioned above, teams that are highly diversified, experience differences between team members. Within groups, subgroups come in to existence based on diversity attributes, the so called faultlines. The existing research on group diversity provides mixed results and in the context of change faultlines and their effects on group dynamics and performance is a lack of research (Gover & Duxbury, 2012). Therefore it is interesting to investigate how these activated faultlines within teams, in a context of change, influence the group dynamics and the change effectiveness. This leads to the following research question:
Theory
Immigration, emancipation and the internalization of organizations are some of the trends that make organizational teams more diverse (Rupert, 2012). These trends lead to demographic and social differences within teams like age, race, gender, religion, country, status and more. Another trend is the increasing use of teams or groups for decision making and production (Bettenhausen, 1991). Due to the increasing diversity at the workplace tensions will come into existence between the promise and reality of diversity in team process and performance (Mannix & Neale, 2005). It is possible that group diversity may have positive as well as negative outcomes on the group performance. According to van Knippenberg and Schippers (2007) it is a challenge for research in organizational behavior to address which processes underlie these effects of diversity and how to manage these processes. It is striking that there are such contradicting study results on the effects of group diversity. A potential reason could be that in previous researches diversity was often measured on only one dimension at the time. In this paper diversity is defined as: ‘any attribute that another person may use to detect individual differences’ (Williams & O’Reilley, 1998: 81). The introduction of the faultline theory (Lau & Murnighan, 1998), which measures multiple similarities and differences of diversity in conjunction, instead of one attribute, provides a more complementary conceptual approach that can explain the effects of diversity (Thatcher et al., 2003).
Faultlines
According to Lau & Murnighan (1998: 328) “faultlines are hypothetical dividing lines that may split a team into subgroups based on one or more attributes”. Subgroups are identified as two or more individuals within a group that are separated from other group members (Jehn & Bezrukova, 2010). It is argued by faultline researchers that when group members share one or more diversity attributes with members of the same group this can lead to hidden faultlines within the organization that have the potential to subdivide a group (Chrobot-Mason et al, 2009; Chui & Staples, 2013; Lau & Murnighan, 1998, Shore et al, 2009). Researching faultlines in a context of change is an area that lacks interest but is very interesting as well. Faultlines can be hidden but may become active, depending on characteristics of the context. An organizational change can serve as a trigger to turn a dormant faultline in an active faultline, severely interrupting accustomed work patterns, tasks and processes (Gover & Duxbury, 2012). Therefore faultline theory research in the context of change is meaningful because of the impact of activated faultlines on group’s processes and performance.
Faultlines bases
There is an enormous amount of characteristics that can address differences between group members. Faultlines are hypothetical separating lines that can subdivide a group of people into subgroup, based on one or more characteristics. This can be any characteristic that members feel they differ (Lau & Murnighan, 1998). A number of researchers developed typologies that could be used to classify the great number of characteristics of diversity. A distinction is made between readily observable demographic characteristics and more job-related attributes (Jackson, 1992; Jehn, Northcraft & Neale, 1999; Milliken & Martins, 1996; Tsui, Egan & O’Reilly, 1992). More recent research divides faultlines into two categories: social demographic faultlines and informational faultlines. Social demographic faultlines examples are gender, race, nationality and age, and informational based attributes find their bases in job related attributes as work experience, status and education (Bezrukova et al, 2009; Gover & Duxbury, 2012; Mannix & Neale, 2005).
notice the social demographic faultlines (Harrison et al, 1998; Thatcher et al, 2003). Group members that experience sharing one or more attributes that finds its bases in informational faultlines are likely to form groups based on these attributes, this is called social categorization. The different nature of social demographic and informational faultlines may influence behaviors in groups, group processes and outcomes (Bezrukova et al, 2009; Thatcher & Patel, 2011; van Knippenberg & Schippers, 2007).
Activation of faultlines
According to Lau & Murnighan (1998) groups are able to have many different faultlines, which increase the opportunity for subgroups to be created. Faultlines could be hidden but can become salient over time depending on the context of these faultlines however; if differences on a specific attribute are unperceived by members then these differences are unlikely to influence team behavior (Zellmer-Bruhn et al, 2008). As stated before it is possible that group faultlines are unnoticed and therefore it is important to identify the difference between dormant faultlines and activated faultlines (Jehn & Bezrukova, 2010). Faultlines that exist within groups without interfering group processes that remain not perceived and hidden are called dormant faultlines (Lau & Murnighan, 1998). These are potential objective faultlines based on some attributes that are not salient (Thatcher & Patel, 2012) but can be triggered by their context, like an organizational change, that changes work patterns and routines. Dormant faultlines are a prerequisite for the existence of activated faultlines (Jehn & Bezrukova, 2010). It is important to consider the transformation of dormant faultlines into activated faultlines in faultline theory (Jehn & Bezrukova, 2010). Past research mainly focused on measuring dormant faultlines and thereby neglecting the perception of group members (Harrison et al, 2002). Therefore this research will contribute by investigating the influence of activated faultlines, thus taking group members perception into account, on group processes and change effectiveness. This study tries to fill the gap in diversity research with a focus on the specific context of change combination with activated faultlines.
The social identity theory (Tajfel & Turner, 1979) and social categorization theory (Turner, 1985) explains how differences within groups and between group members are used to form subgroups. Social identity theory (Tajfel & Turner, 1979) suggest that a process of activation within individuals come in to existence based on the salience of social categories (Jehn & Bezrukova, 2010). SIT attempts to identify the minimal conditions that can lead members of one group to discriminate in favor of their in-group and against another out-group (Tajfel & Turner, 1979).
According to the social identity theory an individual has several “selves” that corresponds with widening circles of group membership. Social contexts can trigger individual thinking, feelings and behavior based on his personal, family or national “level of self” (Turner et al, 1987). Besides the “level of self” an individual has multiple social identities, this is the self-concept of an individual that is deduced from perceived memberships of social groups (Hogg & Vaughan, 2002). Similarities on a set of attributions whether that be social demographic or informational attributions increases the attraction of an individual to similar others (Williams & O’Reilly, 1998). This also provides plausible reasons why individuals align with similar individuals when subgroups are formed (Thatcher & Patel, 2012). Previous studies (Gibson & Vermeulen, 2003; Li & Hambrick, 2005; Thatcher et al., 2003) investigated the effects of subgroup forming due to potential faultlines on group processes however according to Jehn & Bezrukova (2009) activation of faultlines leads to even stronger subgroup formation. Any characteristic’s salience or relevance is depending on the context (Mannix & Neale, 2005).
Faultline triggers
Consequences of activated change faultlines on change effectiveness
In this research the focus is on the activation of faultlines and their consequences on change effectiveness. As posed by Lau & Murnighan (1998) organizational change is perceived differently by each individual and each individual will have a different opinion regarding the change. This indicates that individuals make sense of the change and create their own reality by constructing meaning (Sonenshein, 2010). Facing change, individuals will act in a more conscious and less automatic sensemaking mode and start to interact with individuals or other colleagues in order to make sense of the change and to determine how they should response (Weick, 1995). When facing change “individuals exchange gossip, stories, rumors, and accounts of past experiences and they take note of symbolic behaviors and actions” (Balogun & Johnson, 2005: 524). Recipients of change will make sense of the change through these social processes of interaction and develop new understandings, interpretive frameworks and try to understand the new structure, roles and responsibilities (Balogun & Johnson, 2005; Weick, 1995). When change is to come about, sharing opinions experiences, and social interaction processes amongst colleagues will increase in order to make sense of the change. According to Balogun & Johnson social processes of interaction (2005) leads to the merging individual opinions and perceptions of the change into one opinion for the whole group. To get to this unified perception of the change it is possible that sensemaking on group level is accompanied with conflict between individuals. When the faultline theory from Lau and Murnighan (1998) is taken into account it might be that group diversity hinders social processes of interaction and therefore impedes an unified interpretation from the group.
Identity-based subgroups are negatively related with group processes and several outcomes including lowered communication (Zenger & Lawrence, 1989), reduced cohesion (O’Reilly, 1989), increased conflict (Li & Hambrick, 2005), competition of meaning (Burnes, 2009) and overall a decreased group performance (Harrison et al., 2002). The different perceptions regarding the change and the ensuing personal interests of each group member might lead to competition within the group, heedless the damage for the group or organization (Handy, 1986).
According to what is mentioned above, pursuing of own interests will have a negative influence on the effectiveness of change. The performance and effectiveness of a team depends on the contribution of many, if not all, of the group members. Lau & Murnighan (1998) state that activation of faultlines will work dysfunctional within the group. The activation of dormant faultlines into activated faultlines might change the foundation for future interaction among group members and subgroups can be formed based on activated faultlines (Jehn & Bezrukova, 2010). Van Knippenberg et al. (2004) state that the formation of subgroups on its own is not detrimental for the effectiveness of diverse groups but when subgroups are explicitly formed, a firm foundation for attitudinal and behavioral effects of intergroup bias is established (Van Knippenberg, 2000). Intergroup bias can have negative effects on performance, group functioning, collaboration, internal communication and effectiveness due to ingroup favoritism, pursue of self-interest, stereotyping, outgroup hostility and competition amongst group members (Lau & Murnighan, 2005; Li & Hambrick, 2005; Tajfel & Turner, 1986; Webber & Donahue, 2001). Faultline theory as proposed by Lau & Murnighan (1998) state that increasing diversity subgroups in the larger group increases the potential for covert and eventually intense conflict between individuals and consequently negatively influences change effectiveness. It is beneficial for the effectiveness of a change if diverse teams are able to work around the detrimental effects of their different perceptions and/or meanings regarding the change or these differences may prevent the team to engage in effective process of reflection and interaction (Mannix & Neale, 2005). Managers must invest more time and effort in order to alter a negative view about the change and to construct a more shared and supported view and interpretation of the change.
Mediating role of the group processes conflict and behavioral integration.
The following section will elaborate on how group processes such as conflict and behavioral integration are influenced by perceived change faultlines and affect change effectiveness. Although diversity within a group can lead to benefits (Gibson & Vermeulen, 2003; Lau & Murnighan, 2005) previous research indicates that heterogeneity often is detrimental for group functioning (Milliken & Martins, 1996; Williams & O’Reilly, 1998). These concepts are chosen because previous research on organizational change highlighted the importance of change recipient’s attitudes and resulting group processes towards change for understanding the organizational change process (Fugate, Kinicki & Prussia, 2008; Oreg, 2006; Rafferty & Griffin, 2006). Activated faultlines can lead to more conflict and less behavioral integration what can be explained by the social identity theory as stated before. Individuals want to maintain their social identities and therefore they identify themselves with others that have the same attributes and perceptions (Tajfel & Turner, 1986). When faultlines are activated this may lead to different perceptions of the change by each group member resulting in subgroup formation and more conflict and behavioral disintegration amongst group members.
Conflict
The traditional approach to diversity where team members are assigned or fall into categories (e.g. race and age) where after the impact on group processes is demonstrated (Pelled, 1996) distinguishes two categories; readily observable attributes like gender and race and “underlying” attributes like education and tenure (Milliken & Martins, 1996). The difference between these categories is of use because different attributes are most likely to lead to prejudices and stereotyping when these differences are salient and visible than when they are “underlying”. This does not indicate that “underlying” attributes cannot create differences and problems within groups. Perceived change faultlines are likely to result in conflict.
differences in viewpoints and opinions pertaining to a group task”. Task conflict applies to conflict due to differences in ideas and opinions about the task. The third and last type Jehn and Mannix (2001) identify is process conflict which pertains to issues and conflict about duty and resource delegation, such as; who is responsible, who should do what and how much responsibility different people get. Process conflict is defined as; “an awareness of controversies about aspects of how task accomplishment will proceed (Jehn & Mannix, 2001). Conflict arises when the groups involved have different interpretations of the same situations and appraise the other group’s motives and behavior in a different way (Pinkley, 1990). When subgroups are formed due to a change and the activation of faultlines it is more plausible that the group consisting of subgroups will experience more conflict.
The alignments of attributes within subgroups that are based on the similarity of group members, called social categorizing, presumably intensifies ingroup/outgroup distinctions. This social categorizing accommodates a salient basis for categorizing group members into ingroups or outgroups (Jehn et al., 2009). Due to social categorizing ingroup favoritism and outgroup hostility can occur which can lead to dislike to surface, decreased cohesion, mismanagement, process loss and conflict because individuals favor their own subgroup members more than other subgroup members (Li & Hambrick, 2005; Tajfel & Turner, 1986; Webber & Donahue, 2001). When group members perceive the activated faultlines the differences between subgroups within a group become salient and behaviors, motives, actions and interests will differ between the subgroups. An example of this is that the subgroup that favor the change will support and participate the changes that are necessary while the group that opposes the change will try to reject and block the change, what presumably lead to conflict.
effectiveness (Brown, 1983; Jehn et al., 1999; Pettigrew, 1973). According to Mintzberg (1983) conflict, when it becomes more salient and aggressive, can lead an organization into chaos which may impede groups to achieve their main goals and objective and inhibits change effectiveness. In order to establish an effective change and achieve a high organizational performance it is important to manage the conflict of interests that exists amongst individuals and groups. This study helps to better understand the contradiction in past researches about the effects of conflict in organizations or groups.
Hypothesis 2A: The more group members perceive change faultlines the higher the level of conflict they perceive within the group.
Hypothesis 2B: Group members who perceive conflict within the group will rate the change effectiveness of the change as more negative.
Hypothesis 2C: The relationship between activated faultlines and change effectiveness will be mediated by conflict.
Behavioral Integration
complementary skills according to Berg et al., (1982). Also it provides a more diverse, better understanding and use of the existing explicit knowledge that is present in a team. The need for exchanging coordination and synchronization information is lowered by teams that are highly behaviorally integrated because these teams are typified by a high level of joint decision making (Malhotra et al., 2005). Simsek et al., (2005) showed that firm performance was positively associated with behavioral integration. Besides the advantages the obverse of behavioral integration, behavioral disintegration can be present, which is often caused by conflict within a group and will impair the group’s performance. Behavioral disintegration causes several problems within a group such as failure to exchange key information and poor coordination of activities (Hambrick, 1994). In addition to this, when a conflict arises, group members lapse into disliking and avoiding each other and might try to compartmentalize their tasks what will lead to behavioral disintegration and will negatively influence their performance (Simsek et al., 2005). Some group members may demonstrate member withdrawal, separatism and parochial effort (Hambrick, 1994), however in order to be effective and improve a group’s performance group members need to communicate in a fluent way, share information and knowledge, collaborate and engage in joint decision making.
Hypothesis 3A: The more group members perceive change faultlines the lower the level of behavioral integration they perceive within the group.
Hypothesis 3B: Group members who experience behavioral integration within the group will rate the change effectiveness of the change as more positive
Methodology
This section will outline the design of this study, the procedure that was used to gather data, the participation criteria for this quantitative study, provide a sample description, measurement of the variables and the data analysis.
Research Method
Quantitative research is a research methodology that seeks to quantify the data and, normally, applies some form of statistical analysis (Malhotra, 2010). The study encompasses a descriptive and cross-sectional research design to delineate the influence of activated change related faultlines on group conflict behavioral integration and change effectiveness. A questionnaire was developed to measure the relevant constructs of this quantitative study. In Appendix A the relevant questions can be found. The most appropriate way to test the hypotheses that are based on previous literature is quantitative research (Malhotra & Birks, 2007). This study utilizes several benefits of quantitative research; it is not an expensive way of doing research but within a relatively short time a lot of participants can be reached. Because quantitative data is statistic driven it can provide a lot of information. Quantitative research makes it easier to compile the data into a chart or graph because of the numbers that are made available. Another benefit of quantitative research is that it is conducted on a large scale and gives more information as far as value and statistics. The use of questionnaires allows testing theory on a large sample which helps generalizing the results to a wider population (Cooper & Schindler, 2008). Within quantitative research there is not any form of interviewer bias present. An example of this is that within qualitative research variability in the results of the participants can be influenced by dissimilar/multiple interviewers; while within quantitative research the use of questionnaires prevents the possibility of using dissimilar/multiple interviewers and thus the data is more reliable (Malhotra and McCort, 2001).
Data Collection
The distribution of paper version helps non-tech savvy respondents, who may not have access to a computer and therefore are not able to complete an online survey and therefore are unable to participate in this research. A paper version of the questionnaire increases the feeling of confidentiality of the respondents; while we can be confident that an unnamed piece of paper which we be placed into a box or sent by post is relatively untraceable, respondents often have concerns over confidentiality with web-based surveys.
In order to gather companies, groups and thus individuals to fill in the questionnaires a combination of convenience (Özdemir, St. Louis & Topbaş, 2011) and snowball sampling was used (Bisogni et al., 2002). The latter means that participants, who were gathered by using convenience sampling, were asked for their assistance to reach and identify other participants with the same traits that can be used in the research. The first method makes use of approaching acquaintances. First contact was initiated by mail or telephone where the objectives of this research, input and benefits for the participating organizations was explained. Criteria for approaching these acquaintances were that they own a company, or work in small and medium sized businesses, work in a large company or could provide contact information of companies. And within their organization or work they must work in teams. Some of these acquaintances provided contact information of companies, others made their company, teams and employees available for this research. After receiving contact information phone calls were made and emails were sent in order to figure out if companies were interested in the study and suitable as well.
In order to be suitable for this study the participants must comply with certain requirements. The first requirement is that the companies need to work in teams. We define a team as a group of people linked in a common purpose who are especially appropriate for conducting tasks that are high in complexity and have many interdependent subtasks. Normally teams have members with complementary skills and generate synergy through a coordinated effort. It is not a precondition that teams are actually called a team. However it is a requisite that, in order to be a team, team members and other individuals in the organization need to acknowledge the team and a minimum of 3 team members should share responsibility for an end goal of the organization (Hackman, 1987). Another criterion related to teams is that teams could not be larger than 15 team members, but have a minimum of 3 team members, this in order to find meaningful results in the research of faultlines.
a change that occurred in the past 12 months. This criterion is necessary so that participants have a fresh memory regarding the change. These four conditions were a prerequisite for participation and when these were met, and the companies were willing to cooperate, their contact information was asked and a survey could be conducted. Paper versions of the questionnaire were given to all participants in person or by their team leader. In order to increase the response rate of this research reminders were sent to participants and/or to their team leader to remember them to fill in the questionnaires.
Sample description
During this research new data was collected and an existing database, including all variables and items, was used in order to get a representative dataset. In total 33 organizations, ranging from SBE´s to large organizations and varying between private and public organizations, 82 teams and 506 employers, managers and employees participated. This indicates that several organizations had multiple teams that participated in this research. Team size ranged from 3 till 40 with an average of 9.79 group members. Group membership varied from 0 months till 444 months with an average of 36.36 months. From the 506 participants 52.2% was male, and 47.8% was female, with an age varying from 16 till 62 with an average of 40.2 years old. Of all participants 54.7% worked fulltime and 45.3% part-time. In all organizations a change project was ongoing, started or ended at most a year ago. The type of changes differed from takeovers, new teams, new members, mergers, restart and more.
Measurement of the variables
Activated change faultlines
In order to assess activated change faultlines an 4-item scale was used, to measure the perceived existence of subgroups within a team, based on Jehn and Bezrukova (2010) (e.g.
During work the team is divided into several subgroups and during lunch or a social event members of subgroups talk a lot with each other.). To assure the reliability of measuring the
construct an 4-item scale was used, indicating that the same question were asked in similar, however not in identical ways (Litwin, 1995). A 6-item scale was added which reflected potential faultlines bases in the context of a change. This 6-item scale is based on the dimensions in Oreg et al., (2009) and Levenson (1974). Bases that were used were: attitudinal resistance, behavioral resistance and locus of control. The 6-item scale reflects individual responses towards change and to what extent this leads to subgroups (e.g. Group members
who received a lot of information versus group members who received few information about the change). These two scales together cover the concept of activated change faultlines. The
4-item scale adapted from Jehn and Bezrukova (2010) measures the perceived existence of subgroups whereas the 6-item scale based on Oreg et al.,(2009) and Levenson (1974) reflect on faultlines bases. A reliability analysis showed that internal consistency was good for perceived subgroups (ɑ = 0.908) and faultlines bases (ɑ = 0.907).
Conflict
In this study, conflict was measured based on a 9-item scale measuring relationship conflict, task conflict and process conflict. To measure this variable the scale developed by Jehn and Mannix (2001) is used. A distinction is made between task conflict; ‘How often are there
disagreements in your team about the content of the work in progress’ and process conflict How much conflict about ideas is present in your team? and relational conflict ‘How much relationship tension is present in your team’. The factor analysis showed three underlying
factors. Examining all items showed that respondents probably misjudged one of the items (How often do you disagree about resource allocation in your work group?) as a relation conflict item instead of a process conflict item. After conducting a reliability analysis it was showed that internal consistency was good for conflict (ɑ = 0.913).
Behavioral integration
scale, measuring information exchange within the team and between team members, was adapted from Jiang et al., (2012) (e.g. When the team collaborates; almost every team
members embays good ideas.). Collaborative behavior was measured by using a 3-item scale,
adapted from Simsek et al., (2005) (e.g. When a team member is busy, other team members
often jump in to help him.). A 3-item scale, also adapted from Simsek et al., (2005) was used
to measure joint decision making (e.g. Expectations are mutually discussed amongst team
members.). After conducting a reliability analysis it was showed that internal consistency was
good for behavioral integration (ɑ = 0.922).
Change effectiveness
The variable change effectiveness was measured on a 6-item scale which was developed by Yukl & Fu (1999). The 6-item scale was used to measure satisfaction with the accomplishment of the change objectives, relationship within the team and the resources that were needed for the change (e.g. The extent to which we, as a team, have realized our goals
that we hoped to achieve with the change goals, I am…’1’ very satisfied – ‘7’ very
unsatisfied). A reliability analysis was performed in order to assess internal consistency for change effectiveness, this analysis showed that the internal consistency was good (ɑ = 0.883).
Control variables
Data Analysis
After all data was gathered and entered in Microsoft Excel several analysis were done using SPSS Statistics. At first a Rotated Varimax factor analysis was conducted for all of the items of the concerning constructs. This analysis was done in order to measure the validity of the constructs.
The extraction was based on an eigenvalue higher than 1. The results of the Rotated Varimax factor analysis can been found in table 1. The analysis that was based on an eigenvalue of 1 provided 6 factors and split the variable behavioral integration in to 2 factors instead of 1 (see Appendix B). An extraction based on 5 factors was performed (see table 1), showing that behavioral integration loads on 1 factor whilst it has no significant other effects on other components. There were no items excluded due to double loadings. In addition to this a reliability analysis was done for all constructs.
Table 1.
Rotated Varimax Factor Analysis
As second analysis, a Kolmogorov-Smirnov test is conducted in order to test all variables for normality. The results of the Kolmogorov-Smirnov test can been found in table 2. All variables significantly deviated from a normal distribution. However when testing all variables on skewness and kurtosis all variables stayed in the range of -1.0 and 1.0. According to Cooper and Schindler (2008) this range is acceptable to adopt normal distribution of these variables.
Table 2.
Kolmogorov-Smirnov test
Variable Statistic Significance Skewness Kurtosis
Behavioral Integration Conflict ,07 ,11 ,00 ,00 -,67 ,60 ,80 -,26
Activated Change Faultlines ,07 ,00 ,15 -,84
The third and last analysis, a hierarchical regression analysis, was conducted in order to test the hypotheses. Following Becker (2005), analyses were controlled for variables that significantly correlated with the main variables. Several hypotheses have to deal with mediation. Earlier research on mediation illustrated steps that need to be followed to test for mediation (e.g. Baron & Kenny, 1986; Judd & Kenny, 1981b). As Baron and Kenny (1986) propose there is a four step approach in which several regression analyses are conducted and the significance of coefficients is examined at each of these four steps (table 3). The figure below provides the model of Baron and Kenny (1986).
Table 3.
Steps hierarchical regression analysis (Baron & Kenny, 1986).
Results
Descriptive Statistics and Correlations
Next to the descriptive statistics a correlation analysis, on individual level, is conducted. The results of the descriptive statistics and the correlation analysis are displayed in table 4 below. In total 506 participants participated in this research.
Table 4 shows that change faultlines, the independent variable, significantly correlates with behavioral integration, conflict and change effectiveness. As in line with theory, table 4 depicts that there is a positive association between change faultlines and conflict (r = .00, p = .54)1, and that there is a negative relation between change faultlines and behavioral integration (r = .00, p = -.20), and change effectiveness (r = .00, p = -.16). Furthermore table 4 reveals that activated change faultlines correlates positively with control variables scope of the change (r = .00, p = .16), and task interdependency (r = .00, p = .17) but has a significant negative relation with goal similarity (r = .00,p = -.28). Behavioral integration is positively associated with goal similarity (r = .00, p = .54), team identity (r = .00, p = .40) and change effectiveness (r = .00, p = .45) but has a negative correlation with conflict (r =.00, p = -.28). Conflict has significant relations with all variables concerned.
Conflict is negatively associated with; goal similarity (r = .00, p = -.36), team identity (r = .00, p = -.273), change effectiveness (r = .00, p = -.31) and as stated before with behavioral integration, and has positive correlations with; scope of the change (r = .00, p = .23),task
interdependency (r = .00, p = .19) and change faultlines (r = .00, p = .56). As said before, change effectiveness is negatively associated with main variables change faultlines and conflict, but positively correlated with behavioral integration. Furthermore change effectiveness is positively associated with goal similarity (r = .00, p = .43) and team identity (r = .00, p = .29) and negatively significant related with task interdependency (r = .00, p = -.10).
Hypotheses Testing
Using a hierarchical regression analysis, hypothesis 1 stating that, group members who perceive a higher degree of faultlines will rate the effectiveness of change as more negative, is tested. In order to test hypotheses 1, 2A, 2B, 2C, 3A, 3B, 3C two steps were carried out. The analyses consist of two models, where model 1 includes all control variables that correlate significantly with the main constructs and model 2 contains all the control variables that correlate significantly with the main constructs and the independent variable. Table 5 below provides the results of the hierarchical regression analysis regarding hypothesis 1. As the table depicts, change faultlines have a significant negative relation with change effectiveness (ß = -.09, p < .05). Hypothesis 1 states that there should be a negative relationship between change faultlines and the effectiveness of a change, and this is also showed by the regression analyses. Therefore, hypothesis 1 is accepted, because when group members perceive more change faultlines they rate the effectiveness of change as more negative. Concerning the variance, this model explains 23,9% of the variance for change effectiveness. Furthermore the control variables goal similarity (ß = .36, p < .01). and task type (ß = .19, p < .01). have a positive significant relation with change effectiveness whereas task interdependency has a negative significant relation with change effectiveness (ß = -.09, p < .01).
Table 5.
Hierarchical regression analysis change faultlines on change effectiveness. Model 1 Model 2
Independent Variables ß Sig. ß Sig.
Constant 2.771 .00 3.022 .00
Scope .04 .39 .05 .27
Goal Similarity .39 .00 .36 .00
Task Interdep. -.19 .00 -.17 .00 Change Faultlines -.09 .03 Adjusted R2 R2 F 23.4% .24 37.920 23.9% .25 31.466
Hypothesis 2A stating that group members who perceive change faultlines experience a higher level of conflict within the group, is tested using a hierarchical regression analysis. The analysis consists of two models, where model 1 includes all control variables that significant correlate with the main constructs and model 2 contains all the control variables that correlate
significantly with the main constructs and the independent variable change faultlines is added. Table 6 below provides the results of the hierarchical regression analyses regarding hypothesis 2A. A significant positive relationship between perceived change faultlines and conflict was found in table 6 (ß = .43, p < .01). Due to this finding, I accept hypothesis 2A. When the independent variable perceived change faultlines was added to the model, variance explained for conflict increased from 25.6% to 41.3%. Other implications that were extracted from table 6 were that control variables scope (ß = .14, p < .01) and task interdependency (ß = .15, p < .01). have a positive significant relationship with conflict, but goal similarity (ß = -.19, p < .01) and task type (ß = .-19 p < .01) have a negative significant relationship with conflict.
Table 6.
Hierarchical regression analysis change faultlines on conflict. Model 1 Model 2
Independent Variables ß Sig. ß Sig.
Adjusted R2 R2 F 25.6% .26 42.617 41.3% .42 69.300
Hypothesis 2B states that group members who perceive conflict within the group will rate the effectiveness of the change as more negative. A hierarchical regression analyses to test hypothesis 2B was done and the results can been found in table 7. Table 7 indicates a significant negative relationship between conflict and change effectiveness (ß = -.19, p < .01), and therefore hypothesis 2B is accepted. Model 2 explains a total variance of 25.8 % which explains an increase of 2.4%. Furthermore table 7 shows that goal similarity (ß = .33, p < .01) and task type (ß = .15, p < .01) have a significant and positive relation with change effectiveness. It also shows that task interdependency has a negative significant relation with change effectiveness (ß = -.14, p < .01). Finally table 7 shows that scope of the change has no significant relation with change effectiveness.
Table 7.
Hierarchical regression analysis conflict on change effectiveness.
Model 1 Model 2
Independent Variables ß Sig. ß Sig.
Hypothesis 2C states that the relationship between activated change faultlines and change effectiveness is mediated by conflict. As highlighted before it is necessary to perform several steps to test for mediation (Baron & Kenny, 1986). The first step was confirmed by hypothesis 1, which demonstrated that there is a significant negative relationship between perceived change faultlines and change effectiveness. Step 2 was confirmed by testing hypothesis 2A, revealing that there is a significant association between perceived change faultlines and conflict. The third step was confirmed by testing hypothesis 2B which showed that there is a signification relationship between conflict and the effectiveness of a change project. Step 4 is tested by hypothesis 2C. Table 8 presents the results of the regression analysis. It shows that mediation is fully supported because, in model 2, the mediator, conflict, stays negatively significant after controlling for change faultlines (ß = -.18, p < .05), and change faultlines itself is not significant (ß = -.01, p > .05).
This indicates that the relationship between change faultlines and change effectiveness completely disappears. Therefore full mediation is supported. Further conclusions that can be drawn from table 8 is that task interdependency (ß = -.14, p < .05) has a negative significant relation with change effectiveness. Goal similarity (ß = .33, p < .01) and task type (ß = .15, p < .01) have a positive significant relation with change effectiveness where scope has no significant relationship with change effectiveness (ß = .07, p > .05). Total variance for change effectiveness is explained for 25.7% by this model.
Table 8.
Hierarchical regression analysis conflict mediates the relationship between activated change faultlines and change effectiveness.
Model 1 Model 2
Independent Variables ß Sig. ß Sig.
Constant 3.022 .00 3.425 .00
Scope .05 .27 .07 .08
Goal Similarity .36 .00 .33 .00
Task Type .19 .00 .15 .00
Change Faultlines -.09 .04 -.01 .76 Conflict -.18 .00 Adjusted R2 R2 F 23.9% .25 31.466 25.7% .27 28.864
Hypothesis 3A states that when group members perceive change faultlines they experience less behavioral integration in the group. An hierarchical regression analysis was performed in order to test this hypothesis, the results of this regression analysis can be found in table 9.
As indicated by table 9, no negative significant relation between change faultlines and behavioral integration is present (ß = -.05, p > .05). Therefore hypothesis 3A is rejected. In addition to this variance for behavioral integration is explained for 34.8%. Furthermore the table shows that goal similarity (ß = .45, p < .01) and task type (ß = .24, p < .01) have a positive significant relation with behavioral integration. Finally, table 9 depicts that scope and task interdependency have no significant relationship with behavioral integration.
Table 9.
Hierarchical regression analysis change faultlines on behavioral integration.
Model 1 Model 2
Independent Variables ß Sig. ß Sig.
According to hypothesis 3B it is that the group members who perceive behavioral integration within the group will evaluate the effectiveness of the change as more positive. The results of the hierarchical regression analysis can be found in table 10 below. A conclusion that can be extracted from table 10 is that there is a positive significant relation between behavioral integration and change effectiveness (ß = .30, p < .01), thus stating that when there is more behavioral integration perceived by group members, they will evaluate the change effectiveness as more positive. Therefore hypothesis 3B is accepted. Besides this positive significant relation table 10 shows a significant positive relation between goal similarity (ß = .25, p < .01) and task type (ß = .12, p < .01) with change effectiveness. Control variable scope has no significant relation with change effectiveness, and task interdependency has a negative significant relationship with change effectiveness (ß = -.17, p < .01). Furthermore regarding the variance for change effectiveness is explained by this model for 29%.
Table 10.
Hierarchical regression analysis behavioral integration on change effectiveness.
Model 1 Model 2
Independent Variables ß Sig. ß Sig.
Constant 2.711 .00 2.436 .00 Scope .04 .39 .02 .64 Goal Similarity .39 .00 .25 .00 Task Type .19 .00 .12 .01 Task Interdep. -.19 .00 -.17 .00 Behavioral Integration .30 .00 Adjusted R2 R2 F 23.4% .24 37.920 29.0% .30 40.511
The final hypothesis 3C states that the relationship between activated change faultlines and change effectiveness is mediated by behavioral integration. In order to test for mediation it is
necessary to have significant relationships between change faultlines and change
between change faultlines and behavioral integration is present (ß = -.05, p > .05), which is prerequisite to test for mediation (Baron & Kenny, 1986). Because of this hypothesis 3C is rejected.
Discussion
Key Findings
This research included 33 organizations, 82 teams and 506 participants to examine what the effect of activated change faultlines is on the change effectiveness of a change project. In addition to this, I also researched if the relation between activated change faultlines and change effectiveness is mediated by the group processes conflict and behavioral integration. Research on faultlines stress the negative effects of social categorization on group processes (Lau & Murnighan, 1998) but this might not be entirely justified. The main findings of this research indicate that perceived change faultlines have a significant negative relationship with change effectiveness. Furthermore, the relationship between perceived faultlines and change effectiveness was mediated by conflict, indicating that perceived faultlines lead to a higher degree of conflict within a team, which in turn lead to lower ratings of change effectiveness of the change project. Results also indicate that the effect of perceived change faultlines on behavioral integration is not significant but that behavioral integration has a positive association with change effectiveness indicating that group members who perceive behavioral integration within their team rate the effectiveness of the change higher. Furthermore it is interesting to note the positive relation between goal similarity, task type and change effectiveness, indicating that when change recipients have the same goal or they like their job the effectiveness of the change will be higher. Task interdependency has a negative influence on the effectiveness of change.
Theoretical Implications
organization (Handy, 1986). The negative relation between change faultlines and change effectiveness can be further explained by the fact that group diversity leads to different individual understandings and interpretations of the change which hinders social processes of interaction and therefore impedes an unified interpretation from the group. Due to the absence of an unified interpretation of the change, perceived change faultlines will come into existence and group members with a similar conception of the change will form a subgroup. A group with several subgroups that all have different interpretations of the change will lower change effectiveness because different group interpretations of the change will compete overall leading to decreased group performance (Harrison et al., 2002).
The context of change might explain why conflict serves as a mediator but behavioral integration does not. As the results show conflict serves as a full mediator for the relation between perceived change faultlines and change effectiveness, indicating the relationship between the perceived change faultlines and change effectiveness completely disappears and perceived change faultlines exerts its total influences via conflict. Faultlines lead to a higher degree of conflict which in turn negatively affects change effectiveness. Mediation can be explained due to that conflict is likely to increase when the groups involved have different interpretations of the same change and appraise the other group’s motives and behavior in a different way (Pinkley, 1990), indicating that activated faultlines are perceived that split the group into subgroups. When subgroups appear in a larger group the group members often look to their managers and leaders to bridge the gap between the subgroups. Conflict between groups about ideas and conflicting interests regarding the change is detrimental for organizational performance and effectiveness (Brown, 1983; Jehn, 1999; Pettigrew, 1973).
the behavioral integration of the team is stronger and puts common goals and team interests above personal interests. Successful behavioral integration of a team leads to collaborative behaviors and convenient information flows within a team, what results in team members gaining access to valuable information, knowledge and complementary skills according to Berg et al., (1982) and provide the group members with handholds to deal with the change.
In the context of change differences between group members become salient and if a team’s behavioral integration is low, conflict about interpretations and actions regarding the change will arise. Differences between goals and interests of individuals become visual and faultlines will arise and exert influence via conflict on change effectiveness. However a team that is behaviorally integrated may have differences in individual goals and interests as well, but they put their common goals and interests above personal gain because they are positive that because of the teams behavioral integration the change will benefit the effectiveness of the change. Behavioral integration is much more positively focused than conflict and that could be one of the reasons that in this study perceived change faultlines, which is seen as a negative variable as well, is related to conflict but not with behavioral integration.
Managerial Implications
managers closely to determine if they are treated in a fair way (De Cremer & Tyler, 2007). In order to implement a change in an effective way it is important for managers to control the level of conflict that is present with groups that are affected by the change. Conflict depletes energy and effort, however task and process conflict can be used by managers in order to have a critical debate amongst group members regarding tasks issues in order to enhance group performance because they are more likely to come up with a variety of solutions and idea optimizing their actions and results (Amoson, 1996).The scope of the change is important as well, a bigger change has more impact on the group. Therefore managers must try to diminish the amount of conflict and other dysfunctional group dynamics by increasing behavioral integration between the (sub)groups that are affected by the change.
Contributions
This research has produced several contributions to extend existing research on faultlines. Obviously it contributes to diversity research in general but the contributions of this research are particularly focused on faultline research. The main theoretical contribution of this research is that this research combines the context of change and activated change faultlines and its influence on performance. This is an area that lacks research (Gover & Duxbury, 2012). Research on behavioral integration is scarce as well. The main findings of this research, making use of 33 organizations, 80 teams and 506 participants, provides proof that perceived change faultlines have a significant negative relation with change effectiveness. Apart from these contributions this research focused on activated faultlines whereas previous studies mainly focused on dormant faultlines and their effects on team processes and performance (Jehn & Bezrukova, 2010. ). In this specific context faultlines were activated due to a change. Past researches are limited in their examination of the faultline construct and this research contributes to theory because its extends faultline research of the faultline construct in a new way by focusing on the activation of faultlines. Furthermore previous researches only measured diversity on one dimension at the time, while this research contributes to theory by providing a more complementary approach, focusing on more dimension and their interactions.
Limitations and future research
participants found the length of the questionnaire too long and loosed their interest as the end came nearer. This might influence their assumptions on the questions and therefore have a influence on the results that this research provides. For further research it can be valuable if the length of the questionnaire is reduced. In addition to this, not every group member of the participating groups filled in the questionnaire, what may provided a distorted picture of the actual relations within the team and therefore consequently may influence the results.
Regarding the design of this research it can been seen as a limitation that a cross-sectional research was conducted which ensures that no assumptions can be made about causality. Cross-sectional research also limits internal validity. In order to test for causality a longitudinal research needs to be performed which investigates the whole process of faultline activation and its consequences on group’s performances and change effectiveness. Instead of using questionnaires only the addition of qualitative research methods such as interviews and observations and doing a longitudinal research will enrich and enforce the quality, reliability and validity of the information that is gathered regarding this complex topic. Another limitation is that the absence of a negative influence of change faultlines on behavioral integration might be explained due to the fact that respondents answered the questionnaires in the understanding of their experience in their subgroup instead of the total group. When this is the case the consequences of perceived change faultlines for the whole group are neglected influencing the results. Additionally, due to the use of questionnaires all variables were measured with self-report instruments which inherently provides a base for common-method variance (Richardson, Simmering & Sturman, 2009). Consequently this encompasses that the findings of this research are based on individual and thus subjective assumptions and perceptions of group members. For future research it may be of importance to assess and validate these perceptions and findings by making use of multiple methods for data gathering such as the use of general organization data, objective top-down assessments on individuals and teams and objective financial and performance outcomes. Furthermore, not all participating teams perceived faultlines, while perceived faultlines are the backbone of this research. Even though not all participating teams perceived faultlines the influence on conflict, behavioral integration and change effectiveness proved to be significant. With regard to future research regarding faultline theory it might be interesting to perform a research at teams with a high degree of perceived faultlines.
change, on change effectiveness and to what extend is this process mediated by conflict and behavioral integration is performed, however how these faultlines are activated due to a change remains unclear. A research that investigates which faultline bases are triggered by a change and what the influence of this base is on the activation of faultlines is an interesting research area. Apart from the effect of conflict and behavioral integration between activated faultlines and change effectiveness it might be interesting for future research to identify other possible mediators and the role of organizational culture, leadership and environment on these mediators and faultlines. Furthermore, as depicted by the correlation table goal similarity has a negative significant association with perceived change faultlines, and scope of the change and task interdependency have a positive significant relation with perceived change faultlines. This is can be a point for future research because these variables give insight in predispositions regarding the change. For example, when individuals that are affected by the change have the same goal the impact of perceived faultlines will be lower because they all have the same goal that needs to be reached, which reduces the chance of subgroup formation based on faultlines. Regarding the scope of the change, how more influence the change project has how more perceived faultlines will arise and become stronger and subgroups will be formed. Therefore research on characteristics of the change is interesting to see which specific characteristics of a change are likely to increase faultlines and subgroup formation.
Conclusion
References
Amason, A. C. (1996). Distinguishing the Effects of Functional and Dysfunctional Conflict on Strategic Decision Making: Resolving a Paradox for Top Management Teams. Academy of
Management Journal, 39(1), 123-148.
Arsenault, P.M. (2004). Validating generational differences: a legitimate diversity and leadership issue. Leadership and Organizational Development Journal, 25, 124–14.
Balogun, J., & Johnson, G. (2005). From Intended Strategies to Unintended Outcomes: The Impact of Change Recipient Sensemaking. Organization Studies, 26(11), 1573-1601.
Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychology research: Conceptual, strategic and statistical considerations. Journal of
Personality and Social Psychology, 51, 1173-1182.
Becker, T. E. (2005). Potential problems in the statistical control of variables in organizational research: A qualitative analysis with recommendations. Organizational Research Methods,
8,274-289.
Bell, S. T., Villado, A. J., Lukasik, M. A., Belau, L., & Briggs, A. L. (2011). Getting specific about demographic diversity variable and team performance relationships: A meta-analysis.
Journal of Management, 37(3), 709-743.
Berg, S., Duncan, J., & Friedman, P. (1982). Joint venture strategies and corporate
innovation. Cambridge, MA: Oelgeschlager, Gunn and Hain.
Bettenhausen, K.L. (1991). ‘Five Years of Group Research: What Have We Learned and What Needs to be Addressed’, Journal of Management, 17(2), 345-381.
Bezrukova, K., Jehn, K.A., Zanutto, E., & Thatcher, S.M.B. (2009). Do workgroup faultlines help or hurt? A moderated model of faultlines, team identification, and group performance. Organization Science, 20(1), 35-50.
Bezrukova, K., Thatcher, S.M.B., Jehn, K.A., & Spell, C.S. (2012). The Effects of Alignment: Examining Group Faultlines, Organizational Cultures and Performance. Journal of Applied
Bisogni, C. A., Connors, M., Devine, C. M., & Sobal, J. (2002). Who We Are and How We Eat: A Qualitative Study of Identities in Food Choice. Journal of Nutrition Education and
Behavior, 34(3), 128-139.
Bourgeois, L. J. (1980). Performance and consensus. Strategic Management Journal, 1, 227-248
Burners, B. (2009). Managing Change. London: Prentice Hall.
Brown, L. D.(1983). Managing Conflict at Organizational Interfaces Addison-Wesley. Reading. MA.
By, R. T. (2005). Organisational change management: A critical review. Journal of Change
Management, 5(4), 369-380.
Caldwell, S. D., Herold, D. M., & Fedor, D. B. (2004). Toward an understanding of the relationships among organizational change, individual differences, and changes in person-environment fit: A cross-level study. Journal of Applied Psychology, 89, 868-882.
Campion, M. A., Medsker, G. J., & Higgs, A. C. (1993). Relations between work group characteristics and effectiveness: Implications for designing effective work groups. Personnel
Psychology, 46, 823-850.
Chrobot-Mason, D., Ruderman, M. N., Weber, T. J., & Ernst, C. (2009). The challenge of leading on unstableground: Triggers that activate social identity faultlines. Human Relations, 62(11), 1763-1794.
Chiu , Y.T., & Staples, S.D. (2013) Self-Disclosure and Task Elaboration Reducing Faultlines in Geographically Dispersed Teams. Small Group Research, 44, 498.
Cooper, D. R., & Schindler, P. S. (2008). Business Research Methods. Singapore: McGraw-Hill.
De Cremer, D., & Tyler, T.R. (2007). The effects of trust in authority and procedural fairness on cooperation. Journal of Applied Psychology, 92(3), 639-649.
