USF Libraries
USF Digital Collections

A structural approach to the study of intra-organizational coalitions


Material Information

A structural approach to the study of intra-organizational coalitions
Physical Description:
Walsh, Dean T
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:


Subjects / Keywords:
Informal groups
Social network analysis
Dissertations, Academic -- Business Administration -- Doctoral -- USF
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


ABSTRACT: Coalitions are widely associated with collective or collaborative attempts to influence organizational members, decisions, policies and events. Yet, surprisingly, relatively little is known about how coalitions develop within organizations. Employing an exploratory case study design and using social network analysis, the Rokeach Value Survey, and semi-structured interviews, this research demonstrated that it is possible to identify and study coalitions in a real organizational setting. I suggest that the inclusion and investigation of member relationships may advance the state of the art in organizational coalition research. A benefit of this study, and contrary to most coalition research, is that it used multiple forms of data, including demographic, historical, values-based and interaction patterns for work and social relationships.Two coalitions were identified in the organization studied. Formation centered on a single issue and each coalition followed a strategy designed to influence a possible change in structure and operation. Coalition members exhibited similarities across several factors, including tenure within the organization, education, race, age, and previous experiences. Analyses showed some similarity in member values within and between coalitions. The coalition attempting to maintain the current work structure demonstrated higher value similarity with non-coalition members. Social network analysis revealed that coalition members tended to be structurally similar to each other, more centrally located in the work network, and had higher correlation between coalition interactions and existing social relationships.
Dissertation (Ph.D.)--University of South Florida, 2006.
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Dean T. Walsh.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 129 pages.
General Note:
Includes vita.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001796880
oclc - 156943877
usfldc doi - E14-SFE0001622
usfldc handle - e14.1622
System ID:

This item is only available as the following downloads:

Full Text


A Structural Approach to the Study of Intra-Organizational Coalitions by Dean T. Walsh A dissertation in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Management and Organization College of Business Administration University of South Florida Major Professor: Walter Nord, Ph.D. Cynthia Cohen, Ph.D. John Jermier, Ph.D. Alvin Wolfe, Ph.D. Date of Approval: June 28, 2006 Keywords: informal groups, formation, structure, social network analysis Copyright 2006, Dean T. Walsh


Acknowledgements Its over. Finally. To suggest that my path to completion was indirect would be a gross understatement. Nonetheless, I have beaten the odds and can finally remove dissertation from my to-do list. I am grateful to all of my committee members for their perseverance, understanding and help over the years. This manuscript and my ideas on the subject have improved significantly with their guidance. I wish to express special gratitude to my Major Professor, Dr. Walter Nord. Without his unflappable encouragement this simply would not have been possible. After each interruption, and there were many, he was ready to continue without hesitation. Also, I am indebted to Dr. Alvin Wolfe for not only steering me through the intricacies of network methodologies, but for developing my overall interest and understanding of social networks from the beginning. Finally, but most importantly, I wish to thank my wife, Gretchen. Her patience and support (frequently worn thin) proved to be instrumental in completing my degree. At the end of the day she was right about a great many things.


Table of Contents List of Tables iii List of Figures v Abstract vi Chapter One: Introduction 1 Chapter Two: Literature Review 4 Coalition Research in Social Psychology 5 Reward maximization 5 Coalition size 6 Actor resources 6 Power and control 6 Other variables 7 Comments on Coalition Research in Social Psychology 9 Coalition Research in Political Science 11 Size 12 Ideology 13 Comments on Coalition Research in Political Science 15 Coalition Research in Organization Studies 16 Comments on Research in Organization Studies 18 Summary of the Coalition Literature 19 Chapter Three: Structural Analysis & Coalitions 22 Organizational Coalitions Defined 22 Structure, Interactions and Coalitions 24 Social Network Analysis 25 Chapter Summary 27 Chapter Four: Research Methods 28 Research Design 28 Research Instruments and Data 29 Semi-structured interviews 29 Network self-report instruments 30 Rokeach Value Survey 31 Pilot Research 34 Research Setting 35 Research Protocols 37 Data Collection 39 Data Analysis 41 i


Chapter Five: Results 45 Organization Background 45 Descriptive Statistics 49 Survey Data 50 Coalition Identification 53 Individual influence 53 Collaborative influence 54 Member Issues 55 Coalitions 57 Coalition 1 58 Coalition 1 analyses 61 Coalition 2 63 Coalition 2 analyses 66 Comparative analyses 67 Chapter Summary 70 Chapter Six: Discussion 72 Research Limitations 76 Future Research 78 References 82 Appendices 96 Appendix A: Semi-Structured Interview Questions 97 Appendix B: Network Self-Report Instruments 98 Appendix C: Rokeach Value Survey 102 Appendix D: Participants Memo 106 Appendix E: Network Data Sets 107 About the Author End Page ii


List of Tables Table 1. Schedule of Research Activity 40 Table 2. Software, Inc. Member Profiles 50 Table 3. Software, Inc. RVS Rankings 51 Table 4. Density and Centralization Measures 52 Table 5. Perceptions of Power/Influence 54 Table 6. Perceptions of Collaborative Influence 54 Table 7. Important Organizational Issues 55 Table 8. Evidence for Possible Coalition Activity 58 Table 9. Rokeach Value Survey 113 Table 10. Coalition and Non-Coalition RVS Means 115 Table 11. RVS Rank Order Differences (C1, C2, Non) 116 Table 12. Social Network Degree Centrality (Binary) 117 Table 13. Work Network Degree Centrality (Binary) 118 Table 14. Social Network Degree Centrality (Valued) 119 Table 15. Work Network Degree Centrality (Valued) 120 Table 16. Coalition 1 Structural Equivalence 121 Table 17. Coalition 2 Structural Equivalence 122 Table 18. Work Network Structural Equivalence 123 Table 19. Social Network Structural Equivalence 124 Table 20. Work Network N-Clans 125 Table 21. Social Network N-Clans 126 iii


Table 22. Work Network Lambda Sets 127 Table 23. Social Network Lambda Sets 128 Table 24. Quadratic Assignment Procedure 129 iv


List of Figures Figure 1. Software, Inc. Structure 48 Figure 2. Work Network Diagram 109 Figure 3. Social Network Diagram 110 Figure 4. Coalition 1 Diagram 111 Figure 5. Coalition 2 Diagram 112 v


A Structural Approach to the Study of Intra-Organizational Coalitions Dean T. Walsh ABSTRACT Coalitions are widely associated with collective or collaborative attempts to influence organizational members, decisions, policies and events. Yet, surprisingly, relatively little is known about how coalitions develop within organizations. Employing an exploratory case study design and using social network analysis, the Rokeach Value Survey, and semi-structured interviews, this research demonstrated that it is possible to identify and study coalitions in a real organizational setting. I suggest that the inclusion and investigation of member relationships may advance the state of the art in organizational coalition research. A benefit of this study, and contrary to most coalition research, is that it used multiple forms of data, including demographic, historical, values-based and interaction patterns for work and social relationships. Two coalitions were identified in the organization studied. Formation centered on a single issue and each coalition followed a strategy designed to influence a possible change in structure and operation. Coalition members exhibited similarities across several factors, including tenure within the organization, education, race, age, and previous experiences. Analyses showed some similarity in member values within and between coalitions. The coalition attempting to maintain the current work structure demonstrated higher value similarity with non-coalition members. Social network analysis revealed that coalition members tended to be structurally similar to each other, vi


more centrally located in the work network, and had higher correlation between coalition interactions and existing social relationships. vii


Chapter One Introduction The concept of coalitions appears frequently in classical and contemporary organization studies (e.g., Bacharach & Lawler, 1980; Bazerman, Curhan, Moore & Valley, 2000; Borgatti & Foster, 2003; Brass, 1992; Buchanan & Badham, 1999; Cobb, 1991; Mannix, 1993; March, 1962; Mintzberg, 1983; Oliver, 2001; Pfeffer & Salancik, 1978; Thompson, 1967). Coalitions are widely associated with informal and collective or collaborative attempts to influence organizational members, decisions, policies and events. In fact, it has been more than four decades since Thibault and Kelley (1959) described coalitions as two or more parties who agree to cooperate to obtain a mutually desired outcome. Yet, surprisingly, relatively little is known about how coalitions develop within organizations (Bazerman, 1986; Cobb, 1991, 1986; Stevenson, Pearce & Porter, 1985; van Beest, 2002). Extant views derived primarily from social psychology and political science generally state that individuals form coalitions to optimize some known and measurable outcome. Within this work, coalition formation typically is portrayed as a series of discrete and rational bargains governed by individual opportunities for achieving maximized payoffs (Cobb, 1986; De Winter, Andeweg & Dumont, 2002; Gentry, 1987; Murnighan, 1978; van Beest, 2002). Alternative explanations of coalition processes (i.e., those not based on the maximization of a clear and tangible reward) have been suggested, but have been only partially developed in the organization literature. Several researchers within this field 1


(Diani & Bison, 2004; Eisenhardt & Bourgeois, 1988; Lawler & Youngs, 1975; Murnighan & Brass, 1991; Thurman, 1979) have suggested that coalitions may develop around similar socio-demographic variables such as age, friendship, physical proximity to others, titles, prior experiences and ideologies. The present study attempts to advance coalition research by clarifying our general understanding of the development of intra-organizational coalitions through an exploratory field study that examined the determinants and structure of the personal relationships that define these groups. A primary assumption of my research is that the study of coalition interaction patterns, as sub-sets of larger systems of interactions (i.e., organizations), can offer improvements to the current body of research on organizational coalitions. A major concern of the present study is why certain organization members, given the range of all possible interactions, form coalitions. The general plan of this research was to use social network analysis to map and assess coalition and non-coalition members interaction patterns within larger organizational work and social (non-work) networks. Then, additional data were used to better understand the determinants of the interaction patterns and coalition dynamics. This included evaluating coalition variables identified in the extant social psychology and political science literatures (e.g., rewards, control of others), as well as those mentioned in the organization literature (e.g., structural, social, demographic). The following sections outline the rationale behind this study. First, determinants of coalitions derived from social psychology, political science, and organization studies are identified. Investigation of this research revealed that relatively little is known about the patterned interactions that define organizational coalitions. Then, the significance of organization context, interaction patterns, structural analyses, and the technical issues of this exploratory case study methodology are addressed. A structural viewpoint (social 2


network analysis) is recommended as a useful means to better understand these important groups. The following research questions emerged from review of the existing coalition literature: 1) What are the bases for the relationships that characterize intra-organizational coalitions? To reiterate, the existing literature base frequently assumes coalitions within organizations, but generally does not explain the manner in which they arise. In laboratory experiments formation is simply dependent on the resources assigned to each player. 2) Do similar structural properties exist among intra-organizational coalitions? Are intra-organizational coalitions structurally equivalent to any pre-existing interaction patterns? In other words, are coalition interaction patterns or structures similar to segments of established organization interaction patterns or structures? An analysis of coalition structure has not been conducted. By comparing coalitions networks (structures) to larger organizational networks I hope to help explain formation and coalition dynamics. 3) How do intra-organizational coalitions change over time? Again, coalitions are widely assumed in organizations, and they are believed to be temporary occurrences. However, it is unclear, for example, how certain organizational variables effect their duration and whether or not coalitions are repeatable. 3


Chapter Two Literature Review The coalition literature is extensive. However, to date, most of this research has been done within the fields of social psychology and political science (Cobb, 1986: Cook & Gilmore, 1984; De Winter, et al., 2002). As I demonstrate in this chapter, there are two significant limitations to this body of work for understanding coalition formation in organizations. First, the coalition literature provides little insight into coalition development within organizations. The vast majority of this research has been based on game experiments focusing primarily on the distribution of coalition rewards. Consequently, there are plenty of explanations for post-formation processes (e.g., the allocation of points, votes or money) but there is little to explain how and why coalitions actually form within organizations. Second, the research that has addressed coalition development often has been based on specific assumptions, making its application to organizations problematic. This also is due, in part, to the reliance on laboratory experiments as the primary means of inquiry. Game-based experiments typically assume actors are driven solely by the acquisition of a tangible reward, have perfect knowledge of all game conditions, and rely strictly on negotiation (bargaining) as a means of interaction and governance. Such conditions may not be realistic representations of organizations. In the following sections I identify the most significant coalition research within the fields of social psychology, political science and organization studies as it pertains to the present study, discuss the key variables developed within the literature, and 4


comment on their application to coalition research within organizations. To reiterate, I contend that the research pertaining to coalition formation contains gaps that stem from the techniques and assumptions associated with coalition research in laboratory settings. Coalition Research in Social Psychology As indicated above, most of the social psychological research on coalitions has focused on post-formation processes such as reward distribution. Researchers have attempted to explain how coalition members should divide resources and identify the coalitions that are likely to form given pre-determined resource allocations (Cook & Gilmore, 1984; Komorita & Parks, 1995; van Beest, 2002; Wilke, 1985). Consequently, coalition formation typically has been assumed as part of the rules or requirements of various game experiments. Consistent with its strong focus on rewards, the primary formation variable identified in this literature has been payoff maximization. This emphasis has eclipsed the attention to other variables such as coalition size, actor resources, power and control, gender and status. Even though these variables have been studied, the research on them has focused primarily on their impacts on reward optimization. This research is discussed below. Reward maximization. Reward or payoff maximization is perhaps the most commonly used explanation of coalition formation (Back & Dumont, 2004; De Winter et al., 2002). While rarely tested in real-world conditions, maximization or optimization has been assumed in nearly all social psychological theories of coalitions, including Caplow (1956, 1959), Chertkoff (1967, 1970), Crott and Albers (1981), Gamson (1961, 1964), Komorita (1974, 1979), Komorita and Chertkoff (1973), Rapoport and Kahan (1982) and Vinacke (1971). The rationale behind formation based on reward maximization is that people form coalitions simply because membership will increase the likelihood of 5


attaining some mutually identified benefit (i.e., a reward). Within this body of research, coalitions typically have been equated to majorities within a larger group of game participants, and following game rules, majorities dictate the distribution of rewards. Coalition size. Size also has been identified as a variable that influences coalition formation. For example, according to Komoritas (1974) weighted probability theory and its variants (Komorita & Miller, 1986) the probabilities of coalition formation are inversely related to coalition size. While the theory also assumed payoff maximization, it indicated that complexity related to group activities such as bargaining, communicating offers and decision-making made it more difficult to form large coalitions than smaller ones. As the number of potential coalition members increases, difficulty in achieving optimization and attaining consensus on reward distribution also increases. Actor resources. Several theories have focused on actor resources to predict coalition formation, albeit for maximization purposes. Gamson (1961, 1964), Chertkoff (1970), Komorita and Chertkoff (1973), Komorita (1979), Crott and Albers (1981), and Rapoport and Kahan (1982) stated that the resources potential members brought to coalitions determined coalition formation. Resources, typically money or points allocated prior to the start of games, drove the bargaining processes related to member distributions and coalition alternatives (the number of coalitions possible for formation). Power and control. Caplow (1956, 1959) and later Chertkoff (1967) provided one of the most useful theories of coalition processes. They stated that a group of actors would attempt to control as many other actors as possible outside the coalition. Caplow argued that the actors with the most resources within the coalition would try to control those with fewer resources (Bacharach & Lawler, 1980; Murnighan, 1978). Essentially, actor resources determined which coalitions could create majorities (enabling the 6


domination over outsiders) and which actors were subject to control within coalitions (the domination over insiders). This work is fundamentally different than most other coalition research because actors are assumed to be in competition for the control of others, acknowledging that coalitions form for more socially-oriented ends (power or influence over others) and not simply for some externally provided reward (Cook & Gilmore, 1984). In addition, Caplow (1959) stated that actors excluded from coalitions were still crucial to the games (as dominated others) and thus to the understanding of coalition processes because attempts to avoid domination would lead weaker actors into counter-coalitions. Other variables. Several researchers (Gamson, 1964; Vinacke, 1971) have reported that coalition formation may be influenced by the gender of the game participants. These researchers suggested that females were more likely to form larger than minimal coalitions, split payoffs evenly regardless of power position, and make proposals (bargain) that were not in their own best interest. This is interesting because it is directly counter to the popular maximization principle dominant in the coalition literature. Conversely, males were believed to exploit other actors more frequently, demonstrate stronger drives to win and to increase the level of competition in the game. Laing and Morrison (1973, 1974) proposed that players formed coalitions to advance (or at least maintain) their relative positions in a known status system. They assumed players were aware of their overall rank (status) and even that they could interpret the intervals between players. Unfortunately, theories of coalition formation based on gender and status have not been strongly supported (Murnighan, 1978; van Beest, 2002). In another deviation from earlier coalition theories, Dreze and Greenberg (1980) proposed the idea that the value of a coalition to a potential member depends significantly on the identity of the other members of the coalition. This simple hedonic 7


model suggested that actors strongly consider other members rather than only the rewards likely to be obtained in the winning coalition. Unfortunately, this research did not explore the factors that fostered strong or weak identification between actors. Toward this end, other researchers (Boros, Gurvich & Vasin, 1997 and Cechlarova & Romero-Medina 2001) attempted to incorporate actor preferences into game experiments. However, these preferences were limited to game rule modifications. Similarly to Caplow (1959), Simpson (2004) argued that low power actors and even those excluded from coalition membership were still important to coalition processes. He suggested that structural or hierarchical position affects not only bargaining power but also the ability of low-power actors to organize against unequal bargaining power. It was found that collective action among low-power actors (those with relatively fewer game resources) was facilitated by identification with others who were also determined to be structurally disadvantaged. Simpson proposed two models in this research. In the collectivist model, actors attempted to minimize in-group inequality. In the utilitarian model, actors attempted to maximize the greatest good for the greatest number of actors (including non-coalition actors). Results for the collectivist model showed strong support for male and female participants while the utilitarian model showed only limited support among female participants. This work is important because it suggested that players in coalition games may not only be concerned with maximizing their own outcome, but also concerned with what happens to other players, including excluded actors. This is similar to and supported by research on social exclusion (Baumeister, & Leary, 1995; Leary, 1990; Williams, 1997, Williams, Cheung, & Choi, 2000) where actors behaviors are influenced by those omitted from participation. In addition, this work highlights a potentially different interpretation of fairness. Typically in coalition research, fairness has been construed as 8


an agreement that is closely linked to self-interest, meaning than an actor would agree to an outcome (i.e., consider it fair) if it met his or her reward expectation. This is particularly relevant in game experiments (and other situations) where social norms are used to govern the internal conflict that arises when people want to maximize their own outcome at the expense of others (Pruitt & Carnevale, 1997). By considering excluded actors, this research essentially broadened the concept of fairness (at least in game experiments). Simpson (2004) demonstrated that even when it was possible to obtain a reasonable share of the reward in a small coalition, actors with a more social orientation were inclined to form a coalition that included (and benefited) a greater number of actors. In other words, actors with a social orientation would reduce their own payoff in order not to negatively affect the outcome of other actors. This suggested a moral element to the idea of fairness that linked concern for ones self with the concern for others. However, in work somewhat counter to the above research, van Beest, Wilke & van Dijk (2004) found that socially oriented excluded actors (pro-socials) were less likely to form counter coalitions when payoffs diminished. On the other hand, actors with more of a self-orientation (pro-selfs) preferred participation in counter coalitions regardless of payoff. Comments on Coalition Research in Social Psychology In this section I assess the usefulness and limitations of this body of research as it relates to the present study. As indicated, this literature identifies several variables that may be helpful to the present study, specifically reward maximization, coalition size, actor resources, the control of others, and most recently identification with others and fairness. As stated earlier, reward maximization is the most commonly cited factor in coalition formation. Even though the role of rewards may have been exaggerated in 9


game-based experiments, they do provide a possible explanation of coalition activity. While it is likely that in organizations the purposes of coalitions may be less tangible than attaining game points or votes (e.g., a policy change), the idea of specific purposes or objectives with desirable outcomes likely remain valid reasons for coalition formation. Size also may be important in organization coalitions and hence to the present study. Within social psychological research, smaller coalitions (minimization) have been predicted for two reasons. First, smaller coalitions offer proportionately greater payoffs for members. Second, as mentioned, smaller size is viewed as a means by which potential relationship maintenance problems can be mitigated. This is particularly important to the present study because coalitions are defined by member relationships or interactions. In laboratory experiments, actor resources serve as the basis for participant interaction. They are easily controlled and provide a mechanism for people who do not have any other reasons for such interaction. In organizations, it is expected that people provide certain benefits as coalition members. Examples may include sources of friendship and support, access to powerful others and systems or even sources of information. The idea of control (or power) in this literature remains important even though it has been narrowly interpreted within the game parameters. According to Mannix and White (1992), one of the fundamental factors driving coalition formation is a disparity in power (the ability to influence). Accordingly, coalition formation offers less-powerful actors the opportunity to attain more power through concerted or coordinated action. I expected such power, perhaps manifested as the ability to influence decisions, to be a key objective of coalition activity in organizations and hence pertinent to the current study. 10


Finally, research from social psychology suggests two additional potential factors in coalition formation, namely positive identification with others and fairness. Within organizations positive identification (to others) as a formation variable could be driven by status (similarity as well as the desire to attain more), friendship or beliefs. Positive identification may also be linked to similarity in network position. The concept of fairness beyond ones own interest may also link strongly with belief systems. In short, these two variables come closest to the notion that coalitions within organizations may form for reasons other than maximization such as social reasons. Despite this work, the application of coalition research from social psychology to an organizational setting must be done with caution. The primary reasons for this concern are the narrow views of coalitions and the assumptions within which they have been specified. Again, game actors are assumed to be driven solely by a utility maximization principle typically reflected in the pursuit of some external reward such as money, points, or votes (Bazerman, 1986; Cobb, 1986; DeSwaan, 1985). In addition, individual objectives are assumed to be unanimous (e.g., completion of the game) and bargaining or negotiation is the primary means of member acquisition, governance, and outcome distribution (Cobb, 1986; Komorita, Aquino & Ellis, 1989; Murnighan, 1986). Other limiting assumptions include zero-sum conditions, operation within unambiguous boundaries, and perfect knowledge where actors have total and accurate information regarding the game, rewards, and other actors' resources and options (Bacharach & Lawler, 1980; Cobb, 1986; Miller & Komorita, 1986; van der Linden & Verbeek, 1985). Obviously, life in organizations violates many of these assumptions. Coalition Research in Political Science Political scientists have focused mainly on the role of competing political agendas, loosely framed as ideologies, and size as key factors influencing coalition 11


formation (Bacharach & Lawler, 1980; De Winter et al., 2002). However, as with coalition research in social psychology, the attainment of some reward (maximization) remains a prevalent theme in nearly all views on coalition formation. The main variables discussed below are size and ideology. Size. Riker's (1962) size principle is regarded as the field's first coalition research and its most prominent and tested model (Murnighan, 1978). Rikers work predicted that minimal coalitions would form (where removal of a single member would render it no longer winning). The winning coalition controlled the smallest amount of resources (votes) necessary to realize success (this is similar to Gamson's minimum resource theory). Riker added that if perfect information were not attainable then larger than minimal coalitions would form. Several variations on Riker's model can be found in the political science literature. Leiserson's (1968) bargaining proposition predicts that the number of parties (rather than the number of actors) in a coalition will be as small as possible. Koehler (1972) relaxed Rikers zero sum assumption and has shown that minimum winning coalitions can occur in non-zero sum games. In addition, Dodd's (1974) multi-party parliament model predicts minimum winning coalitions. The primary distinction here is that minimum winning coalitions are defined as those that are no longer winning with the removal of any party (rather than the removal of a single game participant). Diermeier and Merlo (2000) and Baron and Diermeier (2001) developed the efficient bargaining approach to coalition politics to provide an explanation for the size diversity of coalitions. According to these researchers, the party in charge of putting together a coalition could attain the support of other parties by offering a compromise to policy positions in return for support. Here, parties are able to generate equilibrium 12


governments which can be larger than minimum. This is a major point of departure since most research since Riker (1962) consistently predicted minimum winning coalitions. Ideology. A number of researchers (Axelrod, 1970; Leiserson, 1966, 1970; Rosenthal, 1970) have tried to incorporate ideological diversity into coalition research. Generally, this work (called minimum range models) predicted parties with similar ideologies were most likely to form coalitions. Axelrod's (1970) conflict of interest model offered a variation, where conflict of interest rather than ideological diversity was minimized. Here, winning coalitions simply would be those that represented the least conflict of interest among members. In contrast to the research in social psychology, DeSwaan (1970, 1973) used ideology to argue against small size, predicting the formation of larger than minimal coalitions. His policy distance minimization model, based on the assumption that influencing governmental policies was a political party's primary goal, predicted that parties would attempt to become the most central party within a coalition government. Thus, parties to the ideological right would be valued by parties that are to the right of the median of potential coalition members. Here, perceived balance (centrality) was more important than limits to coalition size (Murnighan, 1978). More recent developments in coalition research within political science were due to the limited empirical success of earlier theories (often referred to as the rational choice school). According to Martin and Stevenson (2001) the overall level of theory development and empirical testing in political science has not resulted in significant progress in the explanation and prediction of real-world coalition governments. The neo-institutional view of coalitions emerged in the late 1980s as a major alternative to the traditional approaches to coalition research in governments. This view emphasized the role of different types of institutions in coalition formation processes. As 13


with the historical approaches, the neo-institutional ideas did not reject the rationale behind reward maximization, but refined it with additional hypotheses drawn from experiences with institutional rules (Martin & Stevenson, 2001). In this research institutions were defined as any restriction on the set of viable coalitions that exist beyond the short-term control of model actors (typically cabinet level politicians and administrators and parties), and as such act as constraints on coalition formation processes and outcomes (Strom, Budge & Laver, 1994). Thus, differences in coalition outcomes were predicted on the basis of institutional differences with regard to coalition bargaining rules and norms that allocate power differently between actors in the bargaining process. Institution rules included the order in which bargaining parties were asked to form a government, the ability to control the timing of cabinet announcements (an advantage) and to a lesser extent incumbent coalition partners (Laver & Schofield 1990; Lijphart, 1999; Mershon, 1994, 2001). Unfortunately, neo-institutional theories usually do not make explicit assumptions regarding coalition formation behavior and outcomes and only narrow down the vast majority of possible coalitions. As a result, De Vries (1999) criticized neo-institutional models for their lack of formalization and argued that they could not be considered complete coalition formation theories. Laver and Shepsle (1990, 1996) focused on bargaining on ministerial portfolios (seats) rather than on coalition membership. Their main hypothesis was that coalition bargaining was determined by the credibility of proposals for alternatives to the incumbent government. This credibility depended crucially on the proposed allocation of cabinet portfolios in the new government. 14


Comments on Research in Political Science Coalition theories within political science add support to the role played by some of the variables identified in social psychology. Specifically, these variables including reward maximization (or at least attainment), size (minimization and non-minimization) and ideology, may play a role in the formation of organizational coalitions. Indeed, the fact that these variables have been identified in both literatures suggests they are, at least, worth consideration in future research. For example, it is widely assumed within the political science research that actors make coalition decisions based on a maximization principle (reward maximization). However, the reward is often binary (win lose) political utility rather than the more tangible rewards specified within social psychology (Bacharach & Lawler, 1980). As with the research within social psychology, coalition research within political science relies heavily on assumptions that require caution in applying the findings to organizational coalitions. These include the assumption that actors are completely rational (following a single known objective), games are zero-sum, players have perfect information regarding the conditions, options (moves) and ideologies for any player, and that only winning coalitions have value and members receive only positive payoffs (DeSwaan, 1970, 1973; De Winter, et al., 2002; Riker, 1962; van Beest, 2002). Another concern with political science theories is that they may be limited in explaining formation of organizational coalitions since they tend to focus on the ability of a coalition to implement its objectives following successful formation (Murnighan, 1978). Also, coalition theories tend to emphasize public policy issues that, by nature, are supra-organizational and rarely focus on personnel selection (Andeweg, 2000; Murnighan, 1978). 15


However, an even more serious issue is that these theories may be tautological, since they have been evaluated primarily with the same data used to formulate the models. In fact, a general criticism is that political scientists tend to engage in explanations of previous (or existing) coalitions rather than prediction (Murnighan, 1978). Though this is frequently regarded as a necessity given the long time frame within which governmental coalitions evolve, doubts remain regarding the external validity and application of these theories. Several researchers (Back & Dumont, 2004; Bennett, 2002; De Winter, et al, 2002; De Vries, 1999; Martin & Stevenson, 2001) concluded that in practice the latest theories do no better than the earlier formulations at predicting the composition of real-world governments. These researchers reported similar weaknesses and an overall inability of coalitions models to predict governmental outcomes. Coalition Research in Organization Studies The coalition concept has been identified in the organization literature for more than four decades (Cyert & March, 1963; March, 1962; March & Simon, 1958; Thompson, 1967). In fact, the idea that organizations are comprised of shifting and overlapping groups of people having competing and often conflicting goals is the basis for the political perspective on organizations (Drory & Romm, 1990; Frost, 1987; Morgan, 1986). The subsequent disparities in power depicted in this view are believed to be one of the fundamental factors driving coalition formation (Mannix & White, 1992). Here, less-powerful actors gain the opportunity to attain more power (influence) through concerted action. In somewhat of a departure from traditional views on coalitions, Buchanan and Badham (1999) described political behavior and coalition activity as a necessary rather than an objectionable dimension of organizational life. In the context of planned change programs these researchers reported the need for change agents to utilize powerful 16


coalitions to help improve program success rates. Stated differently, these researchers suggested not only that political behavior can serve organizational goals, but also that it was sometimes a requirement for success. Yet, despite the acceptance of the political perspective there has been a scarcity of field research investigating coalitions in functioning organizations. In addition, within organization studies the concept of coalitions has been inconsistently applied (Cobb, 1986). Even in those instances where coalitions are mentioned within organizations they typically are secondary considerations in the investigation of other research phenomena (Pearce, Stevenson & Porter, 1986). In other words, coalitions are assumed to be common organizational phenomena, but their formation and operation is left largely unexplained. As mentioned, several organization researchers (Eisenhardt & Bourgeois, 1988; Lawler & Youngs, 1975; Murnighan & Brass, 1991; Thurman, 1979) have commented on the importance of structural constraints and similar socio-demographic factors in coalition processes. These factors include age, titles, past experiences, friendship, ideology, and structural and physical proximity. The role of such similarities in group formation has been supported by Blau (1977), McPherson, Popielarz, and Drobnic (1992) and Polzer, Mannix and Neale (1998). These researchers reported that most social contacts occur between people with some degree of similarity so groups tend to demonstrate a relative level of homogeneity. In other words, organization members may form coalitions based on some factor similarity such as friendship or office location with or without some utility maximizing potential. Fenger and Jean-Klok (2001) reported that the extent and structure of relationship interdependencies helped explain the role of single actors in the policy changes between advocacy coalitions in and between organizations. Similar beliefs 17


pertaining to policy topics provided for the iterative or sequential policy changes between participants. Attention to interdependency contributes significantly to the explanation of single behaviors and advocacy coalitions. More recently, research on organizational subcultures has explored the link between value similarity within organizational sub-groups and collective action (Adkins & Caldwell, 2004). Some scholars (Howard-Greenville, 2006; Swidler, 1986) suggest that an organizations subcultures determine the issues addressed as well as the possible strategies for action to be taken. While this work appears to parallel the general understanding of coalitions (i.e., informal groups acting in response to an issue), specific coalition-focused research in this area has not been undertaken. Finally, some of the work in group negotiation (Bazerman, et. al, 2000; Beersma & De Dreu, 1999; Polzer, Mannix & Neale, 1998) describes coalition processes. This research frequently employs dyadic and triadic games in a manner similar to game experiments. Though, here formation rationales tend to be less reward driven. Instead, concepts such as fairness, trust, exclusion, cooperation and social motives are identified as part of the negotiation process (Chen, Chen & Meindl, 1998). However, a key distinction is that these variables tend to be evaluated after a group is formed, rather than a basis for its formation. Another distinction between more specific coalition research is that studies here tend to focus on activities (often conflict) in organization-sanctioned activities, such as work teams and structure (Beersma & De Dreu, 2002; Georgopoulos, 1986), rather than the informal and unofficial world of coalitions (De Dreu, Harinck & Van Vianen, 1999; Kramer, 1991; Saunders, 1985). Comments on Research in Organization Studies Coalition research within the field of organization studies is not as plentiful as that found in social psychology or political science. While this research base has discussed 18


coalitions within organizational settings, it generally lacks rigor and at times is merely anecdotal. In many cases, the term coalition is used to imply a group or subset of organization members. Several researchers have suggested an explanation for the relatively small amount of coalition research within organization studies. Cobb (1991), Pearce et al. (1986) and Stevenson et al. (1985) noted that the study of organizational coalitions has been hampered by inconsistent applications of the coalition concept. Consequently, various interpretations of coalitions have produced confusion within the research. Nevertheless, coalitions generally are believed to be prevalent in the organizational world and most likely play significant roles in shaping organizational events. Summary of the Coalition Literature Taken together, the existing literatures provide some possible components of an account of coalition development within organizations. Indeed, several potential determinants of coalition formation have been identified. These include reward or payoff maximization, power, size, actor resources, ideology, identification with others and fairness. However, as noted, the application of laboratory findings directly to organization research faces several obstacles. A primary concern is that these experiments neglect crucial aspects of organizational context (Komorita & Parks, 1995; Mannix & White, 1992). For example, physical design of the workplace, access to others and systems, organization culture, constraints on behaviors derived from hierarchical positions, the diversity of individual resources, and the sanctioning of certain knowledge and skill sets make organizations noticeably different from most laboratory settings. Omissions of organizational context are a direct result of laboratory conditions and narrow assumptions (game parameters) such as utility maximization, perfect 19


information, unanimous goals, and zero-sum bargaining as the sole means by which actors collaborate (Cobb, 1986; DeSwaan, 1985; Miller & Komorita, 1986; van der Linden & Verbeek, 1985). This is not to suggest that laboratory experiments cant include contextual variables. Rather, game-based coalition experiments generally have not done so. Most researchers probably would agree that organizations are comprised of imperfect knowledge, dynamic conditions, historical constraints, unclear boundaries, rules, policies, norms, strategies and multiple (and often competing) objectives and rationalities. A second concern about applying findings of laboratory research to organizations concerns the external validity of work derived from game experiments. Many theories and models based on laboratory experiments have had little success in predicting coalition processes (Cook & Gilmore, 1984; De Winter, et al., 2002; Mannix & White, 1992; van Beest, 2002). In other words, coalition theories generally do not perform as intended. According to Komorita et al. (1989), most coalition experiments simply suggest that players with the greatest resources obtain the greatest payoffs or that some allocation norm (e.g., equity and equality) directs payoff distribution. However, these (and other) conclusions stem from the types of experimental games employed. As a result, the validity of many coalition theories beyond the very narrow scope of a particular and unrealistic game is doubtful (Kravitz, 1981; Levine & Morland, 1990; Miller, 1980). For example, Miller (1980) found that actor resources had no effect on coalition formation or payoff distribution in most real world situations while Kravitz (1981) reported that resources did have some impact on coalitions. Many of the contradictory outcomes found in the literature can be attributed to differences between the types of games and their parameters employed for theory development and testing (Levine & Morland, 1990). As a result, game theoretic approaches to coalitions generally are not useful 20


beyond narrowly specified laboratory experiments (Miller & Komorita, 1986; Murnighan, 1985). One means of addressing these limitations might be to add contextual factors described in organization studies to existing laboratory research. As mentioned, contextual factors include organization structure, strategies, rules and policies, friendships and even cultural elements such as norms and customs. However, the revision of extant theories and the underlying research would be prohibitive (Komorita & Parks, 1995). In addition, Cobb (1982) reported that a technical rationality has obscured existing coalition theories so greatly that the simple inclusion of more variables in laboratory experiments would not benefit organization research. This may help explain the near absence of coalition research in organizations over the past two decades. According to Komorita and Parks (1995) and van Beest (2004), the decrease in research frequency is directly attributable to the inability of extant theories to work in conditions that approximate real life situations. Consequently, numerous researchers (Bennett, 2002; De Winter, et al., 2002; Gerring, 2004; Karanthanos, 1994; Komorita & Ellis, 1988; Kravitz, 1981; Mannix & White, 1992; Murnighan, 1985) have called for new approaches to study coalitions that incorporate situational context. In addition, it is believed that coalition theories need to adopt a process orientation to highlight the determinants and structure of coalitions and how they change over time (Gentry, 1987; Komorita & Parks, 1995; Murnighan, 1978). The present research offers an approach for meeting these requests. Specifically, structural analysis combined with socio-demographic data is suggested as a suitable means for improving the investigation of organizational coalitions. 21


Chapter Three Structural Analysis & Coalitions As demonstrated in the preceding chapter, research into organizational coalitions has remained under-developed and new approaches to studying coalitions, particularly within an organizational context, are needed. Several researchers (Burt, 1976, 1982; Hosking & Fineman, 1990) have contended that context in any social system is comprised of, and shaped by, ongoing relationships between people. Thus, I suggest that the inclusion and investigation of member relationships as a means of assessing context may advance the state of the art in organizational coalition research. As such, the present research proposes a structural approach for this field of inquiry. The following sections provide a definition of coalitions, discuss the central role of interactions in our understanding of organizational coalitions and provide social network analysis as a useful approach to the study of organizational coalitions. Organizational Coalitions Defined Following Stevenson et al. (1985) and Pearce et al. (1986), the present study regards an intra-organizational coalition as a purposeful, interacting group of individuals that is not identified as part of an organization's formal structure and lacks its own formal structure. In addition, members are considered to be generally aware of other members (enabling interaction), maintain an issue orientation (focused on objectives external to the group), and perceive a need for concerted member action. This definition is suitable to the present study for several reasons. First, it does not limit coalition focus strictly to organizational policies or formal decision-making. 22


Rather, it allows that coalitions may operate in formal and informal realms of organizations to influence not only decisions, but also less discrete outcomes such as the advancement of specific values and agendas (Morgan, 1986). Similarly, this usage frees coalition processes from restrictive assumptions such as formation due to utility maximization. This does not imply that maximization is not at times a reason for formation, but emphasizes that the process is open to a variety of determinants. Second, this definition highlights the action characteristic of coalitions. This emphasis is useful since coalitions are considered to be distinct from other groups of actors, such as cliques and formal organizational units, due to their involvement in mutual undertakings of some form of unsanctioned (unofficial) action (Cobb, 1986; Stevenson et al., 1985). Such action typically involves the planning and execution of influence attempts on individuals, processes, or events external to the coalition. Most importantly, these actions are undertaken in concert. Whether coalition activities are planned through various forms of exchange or actually implemented by members they are usually the result of collaboration (requiring interaction). In addition, since coalitions are issue-oriented it is possible to have multiple coalitions with overlapping memberships. Third, this definition provides for a means of addressing issues of member awareness within the coalition. This includes awareness of the group in terms of its objectives as well as awareness of other members. Such awareness guides the interaction (exchange, collaboration, etc.) among members regarding the discussion and implementation of influencing activities. Therefore, members must at least be generally aware of other members. This does not mean all members must know of and interact with every other member. It simply is intended to exclude individuals who independently 23


desire the same action as the coalition but do not collaborate with others on influencing attempts. Structure, Interactions and Coalitions Interactions among people are important to coalition research and the present study for several reasons. First, repeated or patterned interactions help identify organization and sub-unit (e.g., coalition) structure. According to DiMaggio (1992) and Smelser (1988), relationship structure constitutes and helps us explain the very qualities of social systems under investigation. In addition, Mannix and White (1992) reported that coalition activity within organizations long has been associated with the concept of power. Indeed, organizational power has been described, in part, as a structural phenomenon (Brass, 1992; Fombrun, 1983). Thus, interactions represent a versatile unit of analysis for the study of social systems. A focus on interactions may better enable the identification, measurement, and comparison of coalitions and their surrounding organization structures. Second, some form of member interaction typically is either specifically identified or implied in nearly all uses of the coalition concept. For example, as described earlier, the extant literature assumes interactions as the means to facilitate the bargaining that must occur between game participants to exchange offers, counter-offers and distribute rewards. Other research (e.g., Eisenhardt & Bourgeois, 1988) suggests coalition interaction may be based on more socially derived, or less tangible, exchanges between people. However, this too assumes a key role for interactions. In other words, people independently pursuing similar objectives (as isolates) do not represent the basic concept of coalitions. Concerted interaction drives the process of collaboration that defines these groups (Cobb, 1986). 24


Third, some research (Casciaro, 1998; Stevenson et al., 1985) suggests interactions are important to coalition formation because they enable the shared perception of issues and ideas that may serve as a focal point of collaborative activity. Thus, these researchers hypothesize that coalitions are more likely to form in conditions where member interaction is possible. Indeed, an actor's willingness to cooperate or participate in a coalition may depend on his or her identification with other actors with whom they are interdependent (Kramer, 1993). Such positive identification may be made possible through the exchange of salient social information through interactions (Friedkin & Johnsen, 1999; Marsden & Friedkin, 1994; Rice & Aydin, 1991; Salancik & Pfeffer, 1978). As stated by Bazerman (1985) and Mannix and White (1992), actions and perceptions are mediated by the social context within which relationships are embedded. This is similar to and supported by the work on subcultures where value congruence leads to similar interpretations of events and actions (Saffold, 1988; Trice, 1993) Thus, the study of interactions provides an opportunity to better understand both the organizational context within which coalitions exist as well as the individual links that may explain coalition formation. Social Network Analysis Structural analysis, or social network analysis (SNA), is a suitable method for investigating coalition interactions. SNA is an approach rooted in anthropology, sociology and social psychology for assessing social structures (Borgatti & Foster, 2003; Tichy & Fombrun, 1979). The social network perspective frames social systems as networks of objects or positions joined by various relationships (Brass, 1984; Lincoln, 1982; Nohria, 1992). Social network analysis is concerned with the structure and patterning of relationships and seeks to identify both their causes and consequences (Borgatti & Foster, 2003; Burt, 1992; Tichy, Tushman, & Fombrun, 1979). This view is 25


consistent with Weick's (1979) notion that organizations consist of patterned, repeated interactions among social actors. Within the social network perspective, organizations are considered to be social spaces where ones position relative to others can be measured by social-psychological and demographic data. Location and connection to others within this space provides meaning and a means for member and group identification. Indeed, a persons perceived position is instrumental in determining his or her beliefs, interests and motivation for action (Blackburn & Cummings, 1982; Burt, 1982, 1992; Carley, 1991; Carley & Krackhardt, 1996; Reis & Collins, 2004). Thus, people in extreme or distant network positions have different interpretations of the organizational environment than their more central counterparts (Casciaro, 1998; Erickson, 1988; Krackhardt, 1990; McPherson, et al., 1992). Likewise, people in positions of close proximity (e.g., having strong ties) tend to maintain similar interpretations of the organizational environment (Ancona & Caldwell, 1992; Burt, 1976) and may act similarly (e.g., form a coalition). This condition quite likely results from different roles and the probable dissimilar flows of social information and the tendency for people to seek out similar others (Baumeister & Leary, 1995; Blackburn and Cummings, 1982). The social network perspective has supported an increasing volume of work within the field of organization studies (Borgatti & Foster, 2003). This includes research on social capital (Burt, 1992; Putnam, 2000; Walker, Wasserman & Wellman, 1994), economic embeddedness (DiMaggio & Louch, 1998; Uzzi & Gillespie, 2002), network organizations (Miles & Snow, 1992; Rice & Gattiker, 2000), organizational alliances (Baum & Calabrese, 2000; Oliver, 2001), organizational learning (Brown & Duguid, 2000; Friedkin & Johnsen, 1999), social cognition (Baron & Markman, 2003; Carley & Krackhardt, 1996), and group influence (Carley, 1991; Kiesler & Cummings, 2002) to 26


name a few. Social network analysis aids the investigation of organizational coalitions by focusing on the interaction patterns that are central to the coalition concept and the understanding of social structures. A key understanding within the SNA perspective is that myriad networks exist within a social system by which small-scale interactions become translated into large-scale patterns, which in turn, direct the activities of individual actors and groups (Granovetter, 1973). Thus, even seemingly insignificant actions can alter the interpretive and behavioral landscape of organizations (Knoke & Kuklinski, 1982; Krackhardt & Porter, 1986; Granovetter, 1973). Chapter Summary Investigating interactions is important for understanding coalitions because interactions help to define and identify these groups. In addition, the investigation of interactions may provide a better understanding of social systems such as organizations because interactions help to define structure and context. To be sure, network theorists (Burt, 1982; Erickson, 1988; Ibarra & Andrews, 1993; Krackhardt, 1987; White, 1992) contend that social context (rules, constraints, beliefs, norms, experiences, etc.) is understood and captured best by structural investigations. Since structural analytic methodologies specifically focus on interaction data they provide for an ideal means of deeper inquiry into organizations and coalitions (Cobb, 1986; Knoke, 1990; Murnighan & Brass, 1991; Nohria, 1992; Thurman, 1979; Wellman, 1988). 27


Chapter Four Research Methods Even though coalitions are known to be common in modern organizations, there is little empirical knowledge about how they develop. The present research addresses this lacuna and in doing so provides the foundation for future investigations by helping to explain the nature of coalition formation processes within organizations. To do so, this study investigated developing coalitions within their surrounding social systems. As reported, such organizational context is believed to be important to the advancement of this area of research. In this section I describe the methods used to conduct this study. Research Design The present research employed an exploratory case study design. This approach was chosen for two reasons. First, as demonstrated in the previous chapters, the current state of the coalition literature does not provide satisfactory evidence for its application to organizations. To reiterate, the validity of coalition research is affected significantly by contextual factors (e.g., structure, friendships, demographics, rules and policies, norms), but such factors have not been sufficiently incorporated into coalition research (Levine & Morland, 1990; Mannix & White, 1992; Polzer, et al., 1998). Second, a primary advantage of an exploratory design is that it remains open to various interpretations of events. In the present study, this means that none of the previous explanations of coalition activity (e.g., maximization) are favored initially. While several ideas can be found in the extant literature, given the limitations described earlier, a more open approach seems appropriate. In other words, this design provides for the 28


possibility that organizational coalition processes may or may not be guided by factors identified in the existing research. Investigating coalitions in an organization offers the opportunity to clarify existing ideas while providing a foundation for a more thorough or complete research program. A primary goal of the present research was to develop inferences, or sound explanations (Brown & Eisenhardt, 1997; Eisenhardt & Bourgeois, 1988; Yin, 1994), based on the records of evidence made possible by the use of multiple data sources. Several researchers (Bennett, 2002; Gerring, 2004) have suggested that the use of case studies in coalition research would be a powerful means for identifying new or omitted variables that may, in turn, lead to more complete investigations of coalition dynamics. Similarly, De Winter, et al. (2002), argued that the use of case studies in coalition research would provide the inductive rigor needed for advances in meaningful theory formulation. Research Instruments and Data As indicated, the present research used several forms of data to try to better reflect the organizational and contextual factors that may influence organizational coalitions. These factors included work and social interaction patterns, perceptions of influence, belief systems, and demographic data such as age, gender, tenure, education, work experience as well as organizational history. The instruments used to collect these data included interviews, network measures and value surveys. Semi-structured interviews. Researchers (McPherson, Smith-Lovin & Cook, 2000; Reis & Collins, 2004) within organization studies have suggested that coalition formation may be based on some level of member similarity such as age, gender, friendships, or educational background. The more similar an individual is to other group members on a given demographic characteristic, the more likely they are to engage in 29


cooperation. This occurs not only because people are attracted to those who are similar to them (due to minimization of perceived relationship conflict), but also because they assume that people like them share their same values and worldview (McPherson et al., 2001). Brewer (1981) even reported that people with similar demographic factors tend to perceive each other as more trustworthy. Such data are instrumental in any attempted understanding of organizational coalitions and the chance that they may be based on socio-demographic factors. Semi-structured interviews (appendix A) were used to gather data on organization member demographics, backgrounds, interpretations of influential of powerful others, important issues, coalition activity and coalition rationales. The questions were grouped into categories intended to move the subject effectively from simple responses (producing a non-threatening environment) to more involved answers. Interviews were the primary means of data used to assess similarities across the research population and coalitions. Network self-report instruments. As mentioned, relationships help define context within social systems. The network perspective holds that interpretations of events as well as opportunities and constraints on action are related to relative network position. In other words, similar (or close) network position may impact perceptions (or vice versa) that lead to actions such as coalition formation (Borgatti & Foster, 2003; Burt, 1982; Casciaro, 1998; Erickson, 1988; Krackhardt, 1990). Interaction or network data were captured using network self-report instruments (appendix B). Network self-report instruments are the most common method of network data collection and they have been proven reliable in describing overall patterns of relations (Berg, 1995; Knoke & Kuklinski, 1982). These instruments simply are organization rosters each member can use to indicate whether he or she has a particular 30


relationship with others in the organization. Researchers (Freeman, Romney & Freeman, 1987; Krackhardt, 1990) have reported that people are accurate at recalling enduring patterns of relations they have with others. While they might not remember whom they connected with on a given day, people tend to describe their overall patterns of relations reliably. Forms of interaction among coalition members include face-to-face communication, telephone conversations, voice and electronic mail, as well as written notes and memoranda. Such interactions may take place within and/or outside an organization's boundaries. Rokeach Value Survey. Values-based data were used as a proxy for member ideologies or belief systems. As previously mentioned, coalition research within the political science literature has identified ideology as a variable in coalition processes. Typically, belief systems, value systems, world-views, and ideologies have been treated synonymously (Eagleton, 1991; Gross, 1985) and refer to the personal and social mechanisms for interpreting, understanding, and reacting to one's surroundings (Buchholz, 1976, 1978; Goll & Zeitz, 1991; Ibrahim & Kahn, 1988; Sproull, 1981; Withers & Wantz, 1993). Because values are a guide for behavioral choices, group members who share similar values are more likely to agree about group actions such as goals, tasks, and procedures (Jehn, 1994; 1995). Ancona (1990) also reported that similarity in values resulted in higher identification between group members. Values have been used as a substitute for measuring ideology and worldviews because they are core determinants of human behavior (Beyer, Dunbar & Meyer, 1988; Ovadia, 2004; Rokeach, 1973). Significant correlations have been found between values and numerous individual and group behaviors reflected in belief systems (Crosby, Bitner & Gill, 1990; Kagan, 1986; Thomas, 1986). 31


The Rokeach Value Survey (Rokeach, 1973), or RVS, is a widely used value measurement scale (Kamakura & Mazzon, 1991; Mueller & Wornhoff, 1990; Ovadia, 2004). The RVS (appendix C) assesses eighteen instrumental values and eighteen terminal values. In effect, instrumental values reflect desirable means or modes of conduct, and terminal values reflect desirable end-states. According to Crosby et al. (1990), the distinction between means values and ends values is appropriate in values-based research. Reliability measures for the ranking version of the RVS have been reported according to each segment (instrumental and terminal). Rokeach and Ball-Rokeach 1989) found the terminal values segment of the survey had test-retest reliability scores ranging from .51 to .88 and instrumental test-retest values ranging from .45 to .70. Braithwaite and Law (1985) reported the RVS had demonstrated good reliability over time (i.e. r = .88 to .51 for the terminal values; r = .70 to .45 for the instrumental values) and sound validity when compared to other value scales. Mueller and Wornhoff (1990) stated that the RVS was an overall reliable instrument, but that the terminal segment tended to demonstrate higher reliability (i.e., r = .78 to .80). Several researchers (Braithwaite & Law, 1985; Miethe, 1985; Rokeach & Ball-Rokeach, 1989; Schwartz & Bilsky, 1987) found the psychometric properties of the original ranking version to be satisfactory and a statistically sound and useful instrument for measuring human values. I employed the ranking version of the RVS. Despite its widespread use, there have been critiques of the RVS that are relevant to most ranking systems. First, time demands placed on both the respondent and the researcher makes it difficult to administer. Second, ranking yields a data set that cannot be analyzed with standard statistical methods because of the interdependence of the ranked (ordinal) data (Ovadia, 2004). Third, ranking or prioritizing lengthy lists can be 32


somewhat difficult and require significant concentration by participants increasing the chance for respondent error (Alwin & Krosnick, 1985). Yet, the primary alternative to ranking, value rating, also has been criticized. Here, the primary concerns are that since ratings require less effort the quality of the data may be reduced and that rating makes it more likely that response biases, such as social desirability and concentration around the mean, will occur (Alwin & Krosnick, 1985). Thus, each value response form requires sacrifices in the research application. It is possible to try to group or aggregate RVS scores into clusters such as those pertaining to personal, social or even moral value domains or sub-scales. However, it has been reported (Kelly & Strupp, 1992) that the theoretical distinctions between such domains have received scarce empirical confirmation and lack the specificity that is necessary to draw meaningful conclusions from results. Likewise, Braithwaite and Law (1985) questioned the classification or identification of a value set based on a single aggregated measure. These researchers argued that multiple value scores (e.g., 36 in the RVS) portray a more realistic view of a persons value system. The RVS, as with all ranking systems, results in a list of values in a zero-sum structure by definition. For example, if the position of one value increases by one rank, another value must decline by one rank. Rokeach (1973) argued that while all values are considered to be important when thought of independently, activating a value in a behavioral situation requires relative evaluations of certain values against one another. According to Ovadia (2004), values represent mutually exclusive choices. In situations that call multiple values into possible action, one must be prioritized over the other(s). This means that as value structures change over time or differ across groups, the higher importance of one value must come at the expense of the importance of another value. 33


Conceptually, the procedure of ranking values is consistent with the idea that individual values are comparative and competitive (Alwin & Krosnick, 1985; Kamakura & Mazzon, 1991). In other words, since our own value systems are comprised of hierarchically arranged values it seems appropriate to measure values in a manner that reflects the inherent strength (relative rank) of individuals values. Pilot Research According to Yin (1994), an effective method of verifying construct validity in case study research is to evaluate the instruments pertaining to the constructs to eliminate, or at least reduce, subjectivity in the measures. The instruments described above were administered to a pilot study group prior to application in the present research. The pilot study took place in an information systems department at a large university located in the southeast United States. The pilot study included 12 participants. All pilot subjects reported that they understood the RVS and network instruments and that they agreed with the data the instruments intended to capture. However, one participant mentioned that the network self-report instrument might be clearer if written examples of various types (levels) of interactions were provided. When asked about this, all other subjects preferred the network self-report instrument scales (based on frequency) to those based on examples describing the interactions desired. Pilot subjects agreed that interaction descriptions would be too confusing and incapable of distinguishing clearly between multiple types of relationships. In addition, several respondents suggested including a scale value indicating interactions occurring less than once per week. This change was included in the final versions of the network self-report instruments. Finally, all pilot subjects reported clear understanding of the interview questions, though five people (42%) recommended that the term coalition be included in the 34


question referring to contact from others (number 20, appendix A). However, pilot members indicated that adequate understanding was attained, so it was reasoned that potential increases in understanding were not worth potential increases in the risk of bias. After analysis of the pilot data, two questions were omitted and one question was added to the final question list. Concerning background information, pilot subjects were asked, Do you do anything outside of work with any current members of (company)? This question did not yield data that differed than those produced with the social self-report instrument and was subsequently removed from the question list. In regards to the perceptions of powerful others, subjects were asked, Can you recall an example of when he/she used his/her influence or power? This question generally was answered, constantly or all the time and also was omitted. The background question, "Do you think your actions make a difference in the world?" was added. The final interview questions were reviewed for their appropriateness by the faculty chair of my dissertation committee. The interview schedule then was tested on four non-participants to further assess question clarity and to gain feedback on making question delivery as natural as possible. Research Setting Research site selection had to address several important issues. As mentioned, one problem with previous coalition research has been the lack of organizational context. A second issue has been the difficulty in identifying coalitions in organizations (Cobb, 1991; Stevenson et al., 1985). Thus, two key factors in selecting a site were: 1) the potential it affords for identifying coalition activity within the target organization, and 2) the completeness with which coalition data could be gathered. In addition, I believed that an organization that didn't prohibit coalition development through controlled (i.e., 35


highly formalized) interactions was desirable. So, the site needed to be an organization that would be large and organic enough to foster coalition activity and small enough to enable thorough investigation of potentially numerous drivers of coalition behavior. This included access to the organization's members, history and systems (policies, processes) as well as opportunities to explore potential determining factors. A software development organization located in the southeast United States was selected for the present research. Software, Inc. (all names in this research are pseudonyms) was an operating unit of Parent Company, a multi-billion dollar financial services firm. Parent provided a variety of financial services through its consumer finance, credit card, banking and insurance operations. At the time data were collected, the organization was comprised of more than 1,000 field offices located throughout the U.S., Canada, the United Kingdom, Ireland and Germany. Software had 29 members who were responsible for developing, maintaining and supporting business application software for a variety of Parent's internal operations. Software satisfied the selection criteria described above because relatively few barriers to member interactions existed. Managers, developers, contractors, and interns operated in a flexible team environment and were expected to interact (exchange information, provide support, build friendships, etc.) between project groups and provide assistance to others in an informal manner. Thus, it was possible for a range of relationships, from those based strictly on work to those based strictly on social ties, to exist in the organization. It was believed that such an environment would provide opportunities for the variety of formation determinants mentioned in the coalition literature to occur. Also, the company's size allowed me to become more familiar with organization members, policies and processes. This was beneficial to more completely 36


understanding the sources of coalition relationships and their association to the overall structure of the firm. Software's small size reduced the logistical difficulties of conducting complete interviews and in-depth observations. In addition, the firm's size enabled use of network data sets taken from the entire population of interest (the organization). This is regarded as the best way to avoid problems associated with network sampling techniques (Knoke & Kuklinski, 1982; Scott, 1991). Research Protocols The participants in this research were not notified of its focus on coalition activity. Instead, they were informed (appendix D) that the study intended to explore organizational decision-making. This enabled a truthful representation of the work to be accomplished while protecting against potential bias and contamination. Every organization member was notified that any and all data provided during the study would be kept securely and in confidence. To gain a better understanding of Softwares context, its history was compiled from formal documents (mission statements, planning documents, etc.) and conversations with organization members. The events reported were then condensed into a summary account. Software managers and those personnel with the longest tenure verbally verified the accuracy of this account. Field notes were maintained during every site visit for the course of this research, which lasted more than eleven weeks. Following Berg (1995), informal interviews (discussions) were used initially to build rapport with organization members, learn about the organization and its operations, and aid the interpretation and integration of various data sources. 37


The delivery and collection of the Rokeach Value Survey and network self-report instruments followed the same procedure. I placed the instruments, instructions and unmarked response envelope in each organization members mailbox. Delivery was followed by an electronic mail reminder message sent to all members. The instructions directed subjects to seal their responses in the unmarked envelopes for deposit in a larger manila envelope marked COBA Research affixed to a filing cabinet in a centrally located area. Data from the Rokeach Value Survey were recorded on a personal computer using an electronic spreadsheet software application for tabulation. After all respondent data were coded a printout was generated to check the coded data against each subjects response page. Also, each respondent's scores were rechecked on the computer to verify that each rank level (1-18) was used only once. Two data entry errors were found and corrected. Network data also were checked for errors. The responses for all subjects were coded on a personal computer using a word processing software application. A 27-person by 27-person interaction matrix (appendix E) was created for each data set. When completed, a printout was used to check each of the 729 matrix cells against the actual instruments for accuracy. Four data entry errors were found and corrected. Then, the word processing files were imported into a network analysis software application. Each matrix was rechecked against the original network instruments to make sure no errors occurred during file conversion. No errors were found. The time periods between the distribution of the Rokeach Value Survey, work self-report instrument and social self-report instrument increased the likelihood of accurate responses for each instrument. One-on-one interviews with each participant followed collection of the survey data. The interviews took place in a private conference 38


facility located on premise. Notes for each interview were taken on printed questionnaire sheets with ample space for note taking. This circumvented the need to shuffle between pages during the interview. Immediately following each interview session, notes were reviewed and corrected. All notes were rewritten and stored on a personal computer using a word processing software application the same day the interview took place (all before 6:00pm). The interviews were not tape recorded for two reasons. First, during the pilot study several subjects demonstrated apprehension and voiced concern regarding the taping of interviews. While they seemed to be free with their comments when their words were being written, a tape recorder drew too much attention and made some subjects uncomfortable. Taping was discontinued during the pilot. In addition, it was learned that the note taking procedures worked adequately. Second, since several subjects expressed concern for their anonymity during the value survey phase of the present study I decided that recorded interview sessions would not be requested. Data Collection Data for this research were collected in three phases. Phase one was the collection of survey data. This included distribution of the Rokeach Value Survey to assess organization members value systems and the generation of network data sets. It also included reviews of organizational policies, workflows, and operating structure. Phase two consisted of semi-structured interviews to gather additional socio-demographic data. Key objectives here were to ascertain members perceptions of influential others in the organization, and their reasons for perceiving them to be influential. Also, the identification of potentially important (or volatile, controversial, etc.) issues occurred in this phase. Phase three consisted of using responses provided in phase two to identify potential coalition members and their interactions with others. 39


As shown in Table 1 (below), the first step of this study consisted of the distribution of an introduction memorandum (appendix D) to the 29 members of the organization. Table 1 Schedule of Research Activity Day Activity 1 Distributed introduction memos 10 Distributed Rokeach Value Surveys 16 Began collecting Rokeach Value Surveys 21 Collected last Rokeach Value Survey 29 Distributed work self-report instruments 32 Began collecting work self-report instruments 37 Collected last work self-report instrument 42 Distributed social self-report instruments 45 Began collecting social self-report instruments 49 Collected last social self-report instrument 50 Began interviews 63 Completed interviews 64 Began coalition identification 77 Completed data collection Nine days after distributing the introductory memo the Rokeach Value Survey was distributed to all organization members. Four subjects required additional clarification that involved direct conversations. It was learned that these people were concerned with their possible identification (to management) in subsequent reports associated with the study. Their fears were adequately addressed with assurances that confidentiality would be maintained and that data indicating identities would be destroyed after coding or completion of the study. Only two people chose not to participate in the research. Within 11 days, 27 members (93.1%) completed and returned the Rokeach Value Survey. This group of respondents became the research population. Eight days after the last value survey was collected the work network self-report instrument was 40


distributed. Twenty-seven self-report instruments were collected within eight days. Only two subjects needed prodding and this was accomplished using verbal reminders. After five days the social (non-work) network self-report instruments were distributed. The same collection procedures produced 27 responses within seven days. Only one subject required verbal reminders. Semi-structured interviews (appendix A) began the day after network data gathering had been completed. Interview length ranged from 21 to 47 minutes with an average of 30.93 minutes. As is typical in this type of interviewing (Berg, 1995), information that was not expected but appeared relevant to the research was pursued. Data Analysis The general strategy for data analysis followed several steps. First, interview data were analyzed to identify potential starting points for the identification of coalition activity. This included members with perceived influence as well as possible issues likely to result in coalition activity. Second, attribute data (values, roles, work assignment, demographics, etc.) were used to help build profiles of participants and non-participants in coalition activity. Third, network analytic data were used to investigate the structures or relationships between and among coalition participants and non-participants. An underlying question of this research was to determine whether associations (between values, relationships, backgrounds, etc.) could be found among the data pertaining to individuals who participated in coalitions. Attribute data were assessed mainly through value comparison. Data analysis for the Rokeach Value Survey consisted primarily of order (rank) comparisons and means analysis. Relationships among individual values for coalition members and non-members were investigated. Since these were small population measures inferential statistics were not required. 41


Analysis of network data required the identification of organizational and coalition structures (sub-sets) and their comparison to other groups including other coalitions and non-participants. In SNA, network diagrams are graphical depictions or maps of an actors relative position, or linkage, to other actors. In the present study, networks were created for work-based and social (non-work) interactions using observation, interview and self-report data. Again, actor names are pseudonyms and in some cases, particularly in network diagrams, only the first initial of the artificial name is used. Numerous measures (mathematical algorithms) are available within SNA to help analyze networks at various levels of analyses -individual actors or positions, sub-groups, and the entire network. For this research, position-level analyses included measures of centrality, frequency and relationship similarity (equivalence). Degree centrality assesses each actors direct links to (out-degree) and from (in-degree) every other actor. It is a relative measure of how connected an actor is to others in a network (Borgatti, Everett, & Freeman, 2002). Relationship frequency was assessed using N-clan analysis. N-clan analysis clusters actors into groups where each member is connected to each other by no more than N relationships (where N is a researcher-selected value). For example, a 2-clan is a group where each member can be linked to every other member through no more than two relationships (links). While this analysis frequently is used for assessing sub-groups, in this study the raw cluster data were used to generate frequency data for each actor (i.e., the number of times identified in a cluster). Results provide a measure of an actors potential connectedness. These data were used to further assess actor connectedness and also provide a measure of potential influence (Borgatti, et al., 2002). Measures of relationship similarity, or equivalence, indicate the degree to which network actors have similar relationships with others. True structurally equivalent actors 42


have identical relationships to the same other actors in a network. In other words, structurally equivalent actors are perceived to be completely substitutable (Burt, 1976). Regular equivalence is a less stringent measure of similarity. Regularly equivalent actors have similar relationships with similar others in the network (Scott, 1991). Regularly equivalent actors are perceived to have the same role (rather than the same exact position) in a network. Subgroup analyses included measures of importance (lambda analysis) and network correlation using the quadratic assignment procedure (QAP). Lambda analysis assesses relationship transaction and the likelihood for network disruption, or impact on others (hence importance), if relationships were removed. Where centrality measures focus on an individual actors direct relationships with others, lambda analysis considers the role of actors in more complex interactions, and more specifically, their aggregated impact on the network. Network correlation simply compares the interactions of one group of actors to those of another group. Network-wide measures included density and centralization. Network density refers to the percentage of ties (relationships) that exist among a group of actors to the total number of ties that are possible (Scott, 1991). Network centralization is the extent (expressed as a percentage) to which a networks relationships are concentrated around a cohesive center. Here, center does not refer the middle of a diagram because there are numerous graphical depictions for any network data set. Rather it refers to actors, or in large data sets, groups of actors that may represent high percentages of the overall networks links (regardless of their location in a diagram). As previously mentioned, data for the network measures were based on five point scales (for work and social sets) to indicate relationship frequency between actor pairs. However, some network measures require binary data. Binary data use a value of 43


one to indicate a relationship (or link) exists between actors and a value of zero to indicate no such relationship. In such cases, research strategies provide options for the conversion of network datasets. For example, assuming data on a five-point scale, a researcher may decide to analyze all interaction values greater than two (some measure of frequency). Here, data values of three, four, and five would be converted to the value of one and all others treated as zero (no relationships). This procedure yields what is known as a dichotomized dataset. Direct causality between variables is not testable (verifiable) in case study research and generally is not undertaken. Similarly, standard tests of significance and methods of inference are not appropriate. Developing strong evidence chains among multiple data sources provides the methods by which inferences can be made with sufficient validity and reliability (Yin, 1994). To reiterate, this research used several forms of data to assess associations among organization members and their structural positions in an attempt to explore intra-organizational coalitions. It was believed that such associations coupled with social network analytic techniques would demonstrate a new approach to understanding both the contextual environment within which organizational coalitions exist and the actual coalition groups (as organization sub-structures). 44


Chapter Five Results The purpose of this research was to explore how intra-organizational coalitions form. It was expected that key elements in the formation process would consist of demographic characteristics, personal beliefs (values), structural position and specific issues in the organization. Following the procedures described in the previous chapter, two coalitions were found. Each group formed as a belief-based response to a common issue. Members of Coalition 1 preferred a change to a more traditional and limited approach to the idea of teams closely linked to boundaries of the organizations project-oriented work activities. Members of Coalition 2 hoped to perpetuate a broad, integrated view of team-based work and governance. This view was consistent to the original design objectives of the organization, but in its current form was perceived to be detrimental to organization performance. These coalitions and the supporting data are described in the remainder of this chapter. However, before proceeding the organizational context in which these developments took place needs to be described. Organizational Background As previously indicated, Software, Inc. was an operating unit of Parent Company and was formed at the end of 1994. Parent had a reputation of being conservative and highly bureaucratic. According to Softwares president and managers (known as group leaders), the decision to locate Software in the southeast U.S., far from the organization's northeast U.S. headquarters, was guided by a desire to create a Different 45


kind of organization, free of the difficulties that had become associated with Parent's rigid corporate bureaucracy. Software's mission statement specifically called for the creation of an organization and culture that facilitated high levels of skill in all areas of software development. In the words of one group leader, "The purpose of this company is to develop systems and to create a new culture, and We want development to be quicker, better, cheaper, which is just the opposite of [other Parent units]." One of the primary alternatives to traditional business practices that Software was chartered to explore was the use of flexible work teams within an extremely flat structure consisting only of group leaders (managers) and system developers (known as team members). The creation of Software (six years in planning) almost immediately produced problems in its relationships with other organizational units. Software's growth in personnel and associated resources was intended to be through the planned attrition of other corporate information technology units (of which there were several). In other words, Softwares success most likely would be accompanied by reductions of personnel and operating budget at other IT groups within Parent. Early after its formation, the president and four group leaders of Software noted that the political struggle for power between technical units across Parent was the primary threat to the new organizations success. It frequently was mentioned that organizational power was related directly to the accumulation of headcount and the automatic allocation of financial resources attributed to each IT employee. Thus, as one respondent observed, as technology groups within Parent Grow in size they get more resources, get handed more development work and add more headcount. For this reason, the group leaders viewed effectively increasing the number of workers within Software as a key business objective. 46


Many team members, even though some had little or no history with the organization, regarded Parent as "Big Brother". Remaining unique (i.e., flat, flexible and informal) compared to the reputation of Parent was viewed by group leaders and team members alike as a measure of success about equal in importance to more common performance measures such as project time, cost and work schedules. As one group leader said, "If we get pulled back into the corporate structure we've failed." System development based on self-directed work teams was viewed as high-risk since all of Parents other units historically had relied on hierarchical project management and control. Software's original structure (Figure 1, below) was intended simply to consist of flexible project teams interacting with a core management team. The organization started with five members on the management team (four group leaders and the president) and seven team members split into two project teams. As originally planned, each project team would be assigned a group leader who would be the link between the two hierarchical levels. In addition, each team had a member designated team leader to help direct the work within the group. It was envisioned that teams would be broken down and members reassigned as various projects began or ended and as resource requirements shifted. While group leaders could be assigned to more than one project, team members and team leaders would work only on one project at a time. Also, a team leader on one project would not automatically be designated the same role on subsequent projects. It was believed this structure could effectively coordinate more than 100 team members and their projects. Group leaders were meant to handle day-to-day planning and administrative tasks as well as high-level project management. 47


Figure 1 Software, Inc. Structure Team Membe r Team Membe r Team Leade r Team Member Team Member Team Leade r Grou p Leade r Grou p Leade r President Even Softwares office space was designed to support the loose structure of the organization. The entire space basically was a large, open room. Each person worked at a self-contained, ergonomic workstation that could be moved and connected into larger work areas or clusters as needed. The only exceptions to the workstations were more enclosed, four-wall pens for the president and group leaders. In addition, there was a more traditional conference room at one end of the main workspace and a kitchen/lunch room at the other. Two years after its formation, Software had undergone several changes. Only one of the original group leaders had maintained the same position. One group leader had been terminated, one resigned from the organization, and one had been demoted to team member status. In addition, one of the original team members had been promoted to group leader. Softwares president was terminated for the poor development record of the group and an interim president from another Parent technology unit had been 48


installed until a permanent replacement could be found. These changes generally were in response to Software's lackluster performance. Since its inception, most of Softwares projects were poorly managed, missing time, budget and quality objectives (more than 80% of combined project milestones). According to one team member, The projects are behind due to all the thrashing which meant a lack of focus, or that too many people were talking about problems and not enough were doing anything to solve them. Most organization members agreed that the main impediment to successful operation was a general misunderstanding of the team concept. To some, this idea meant equal involvement in all organizational issues, including those considered outside the domain of system development. To others, the team idea was nothing more than a work group designation. Even Softwares group leaders couldn't agree on team definitions and the roles and responsibilities necessary for effective team-based work. As one group leader mentioned, We have plenty of direction, we just need to pick one. Descriptive Statistics To reiterate, the organization (research population) consisted of a president and two group leaders, one office manager, 17 team members (four of which were team leaders), four programming contract workers, and two interns. As shown in Table 2 (below), nineteen of the 27 organization members (70.37%) were male and 21 (77.77%) were Caucasian. The members average age was 36.04 years and average tenure in the organization was 14.93 months. Twenty-two (81.48%) of the members had completed a four-year college degree and three of these had completed a graduate degree. At the time of this research there were three operational project teams (Red, Green and Black). 49


Table 2 Software, Inc. Member Profiles Actor Sex Age Race Edu Status Tenure* Rank** Team Ann F 51 Caucasian AA Divorced 25 TM Red Bob M 43 Asian BS Married 7 TM Green Cece F 32 Caucasian HS Married 25 OM NA Dan M 48 Caucasian AA Married 12 TM Black Ed M 40 Caucasian BS Divorced 8 TM Black Fred M 32 Caucasian MBA Married 25 GL Red Greg M 35 Caucasian BS Married 25 TM Black Hal M 39 Asian BS Married 4 CN Red Ina F 35 Hispanic BS Married 18 TM Black Jim M 34 Hispanic BS Married 25 TM Green Kim F 21 Asian BS Single 3 IN Black Lou M 38 Caucasian BS Married 25 TM Black Mike M 28 Caucasian BS Married 3 TM Black Neal M 41 Caucasian BS Married 26 TM Green Orin M 36 Caucasian HS Married 4 CN Black Pat F 38 Caucasian MBA Married 26 GL Black, Green Quin M 33 Caucasian BS Married 6 IN Green Russ M 36 Caucasian BS Married 24 TM Green Sue F 39 Caucasian BS Married 19 TM Black Tina F 24 Asian BS Single 24 TM Black Ule M 28 Caucasian BS Married 21 TM Red Vic M 31 Caucasian BS Married 4 CN Black Walt M 38 Caucasian AA Divorced 3 GL none Xien M 41 Caucasian BS Married 9 TM Black Yuri M 45 Caucasian BS Divorced 6 TM Red Zia F 30 Caucasian MBA Single 1 CN Black Abe M 37 Caucasian BS Single 25 TM Green Tenure is measured in months ** TM = team member, GL = group leader, CN = contractor, IN = intern, OM = office manger Survey Data The means from the Rokeach Value Survey are shown in Table 3 (below). Given Softwares operating objectives and structure it was not surprising to see the instrumental values Responsible, Honest, Capable and Ambitious ranked highly within the group. It was consistent with the types of people recruited into the organization. The lack of a rigid hierarchy also supported the expectation that Obedient would be ranked last. However, the values Imagination and Self-Controlled were expected to rank higher in the group due to the role of self-direction in the project teams. This possible disconnect between Softwares loose structure and its employees low value for self control may help explain the poor project performance. Similarly, the terminal value Equality was expected to rank higher given the stated mission of the organization. The 50


ranking of Happiness and Self-Respect are consistent with expectations of Softwares work environment. Table 3 Software, Inc. RVS Rankings (by Mean) Terminal Values Mean St Dev Instrumental Values Mean St Dev Family Security 5.00 4.08 Honest 4.93 4.84 A Comfortable Life 5.89 5.11 Responsible 5.78 4.82 Happiness 5.93 4.67 Capable 7.52 5.02 Self Respect 6.85 3.69 Ambitious 7.63 4.95 Inner Harmony 7.41 3.85 Intellectual 7.74 3.94 True Friendship 7.67 3.42 Cheerful 7.96 3.87 Pleasure 8.00 4.39 Independent 8.41 4.36 A Sense of Accomplishment 8.19 4.27 Loving 8.48 5.36 Freedom 8.22 3.38 Logical 8.78 4.51 Mature Love 8.37 4.47 Broadminded 8.85 4.72 Wisdom 8.74 5.12 Helpful 9.74 4.69 An Exciting Life 9.78 4.36 Courageous 10.22 3.68 Social Recognition 12.19 4.45 Imaginative 10.30 4.25 National Security 13.41 4.88 Forgiving 11.48 5.86 A World of Beauty 13.48 4.55 Polite 11.63 2.79 Equality 13.63 3.95 Self-controlled 12.22 4.46 A World at Peace 13.89 3.42 Clean 13.11 3.96 Salvation 14.37 4.41 Obedient 16.22 4.87 The terminal values Happiness, A Comfortable Life and Self-Respect were expected to rank near the top because becoming a member of Software required multiple interviews where the openness and team ideas were routinely communicated. Thus, a candidate was well informed of the work conditions. Since a position within Software was described to be so different from traditional development organizations it was expected that potential members would rank these values favorably. However, Equality was not expected to rank sixteenth given the repeatedly stated objectives of the organization. 51


The graphs of Softwares work and social (non-work) networks are shown in Figures 2 and 3 respectively (see appendix). Dichotomized data were used to filter out interactions occurring less than once per week. This provided a clearer view of the network structures by eliminating incidental interactions. In the work network, the three project teams are easily discernable. The figure revealed nearly complete interactions within development teams and significant cross boundary interactions as well. This is consistent with the design objectives of the organization. However, note the differences between Walt (W), the new president and Pat (P), the most senior group leader. Walts primary objectives were to get the project teams back on track, which he chose to execute via the group leaders and Cece (C) the office manager. The social network diagram revealed only a few minimally connected actors and less segmentation or cohesion than the work network. This was not surprising given that most work interactions would take place within the project teams. Table 4 (below) shows the density and centralization measures for each network. While the work network was not as dense as the social network (had fewer of the total possible relationships), its relationships were more centralized. Table 4 Density and Centralization Measures Network Measure Work Social Density 39% 49% Centralization (in degree) 32.692% 19.231% Centralization (out degree) 36.686% 23.225% Approximately one-third of the interactions from-others (in degree) and to-others (out degree) were found in the networks core. As is demonstrated, there were a group of actors with high interactions with other members. This also was to be expected given the allocation of work by project teams and their link with the management team. Again, 52


note the role of Pat in Figure 2. She was a group leader responsible for two project teams and played a very central role in attempts to improve performance. Though less centralized, the social network was denser (had 49% of all possible ties) than the work network. This too was expected because of the relatively small size of the organization and the ease with which friendly relations could be maintained across the network. However, several members such as Greg, Lou, Dan and Abe clearly appeared to be more centrally connected. Coalition Identification This research assumed that coalitions would be difficult to identify. The strategy to simplify this activity was to identify influential organizational members and those perceived to participate in collaborative influencing, and ask about their activities (specifically their interactions) regarding some important or contentious issue. It was hoped that this combination of influential members and important issues would lead to the identification of coalition members. Individual influence. The primary objective of the interviews was to identify individuals and issues around which coalitions may develop (or operate). The starting point was to identify those members of the organization who were perceived to be powerful or influential (question 14 in appendix A) with the hope of identifying actors likely to engage in coalition activity (exercising their influence). This is not to suggest that perceived influence was a requirement in the identification process. Rather, it was hoped that it would simply make the discovery process easier. Each of the 27 research participants identified those people he or she believed to be powerful or influential. Lou, Fred, Walt, Greg, Pat, Russ, Ule and Sue were perceived to have influence in the organization given the number of respondents citing them (Table 53


5, below). Ann and Abe were omitted since each was cited only once. The remaining members of this sub-set represented possible candidates for coalition activity. Table 5 Perceptions of Power/Influence Actor Frequency % Lou 20 17.54 Fred 18 15.79 Walt 17 14.91 Greg 15 13.16 Pat 13 11.40 Russ 10 8.77 Ule 10 8.77 Sue 8 7.02 Ann 1 .88 Abe 1 .88 Total 114 100% Collaborative influence. The next question analyzed in this phase of my research referred to peoples perceptions of others who tended to collaborate to exert influence in the organization (appendix A, question 21). Table 6 (below) indicates how frequently members cited others, and sometimes themselves, as likely to engage in collaborative influencing. Table 6 Perceptions of Collaborative Influence Actors Frequency % Greg 19 25.33 Sue 15 20.00 Lou 11 14.67 Fred 10 13.33 Ule 10 13.33 Russ 4 5.33 Abe 4 5.33 Pat 2 2.67 Total 75 100% Again, it was simply hoped that these responses would facilitate the identification process. However, this question also was important because coordinated activity is a 54


feature of coalitions. Basically, these data, when combined with those from the previous section, provided a group of organization members I suspected to be more likely to engage in coalition activity. Member issues. Once influential members were established, the final part of coalition identification focused on possible organizational issues or events. The idea here was that influential actors would be more likely to exert influence around certain issues. As with the questions regarding influential members and collaborative influence, this question (appendix A, question 16) was asked simply to help identify potential coalition starting points. Table 7 (below) identifies the issues cited by respondents. Table 7 Important Organizational Issues Issue Count % Team/structure problems 17 37.78 Compensation 11 24.45 Improving project success 7 15.56 Recruiting/retaining employees 4 8.89 Training 2 4.44 Organizational survival 2 4.44 Flex time 1 2.22 Dress code 1 2.22 Total 45 100% As indicated in the table, the most frequently mentioned concern had to do with the organizations team-based structure. Seventeen respondents (37.78%) said it was an ongoing issue. To be sure, the notion of team-based work had been an issue since the organizations founding. Members mentioned the perceived difficulty in accurately evaluating individual performance in a team environment where most emphasis was placed on team outcomes. Also, there were concerns that the team assignments and designations were becoming fixed and that a more formal three-level hierarchy would develop. Most of the comments here were directed at the role of team leaders and the 55


original idea that they were not intended to be a position different than other team members. I selected team-based or structural issues to look for coalitions because this concern was broadly recognized and generated multiple comments and opinions, suggesting these were matters likely to spawn coalition activity. As mentioned, team-related issues had been present throughout Softwares relatively short history, which may explain why interview participants identified team issues most frequently. Since the companys inception team members had challenged group leaders over involvement in non-development activities. Examples included team members desires to assess and arrange their own training programs, become more involved in recruiting decisions, determine individual and project performance measures, and to have access to business unit planning and reporting documents. One particularly contentious aspect of the team issue pertained to roles within the team structure. As reported, Softwares structure was intended (and repeatedly communicated) to be flat, having only two hierarchical levels (group leaders and team members) utilizing self-directed work teams (frequently referred to as high-performing teams). Several factors contributed to perceived problems with Softwares structure. First, the role of team leader, while described as an informal position suggested that some difference in rank or status existed relative to other team members. To make matters worse, team members designated team leaders rarely were reassigned. This was due in part to the duration of Softwares projects. Two of the three active projects started with the organizations founding. In any event, other team members had not seen teams reassigned and new team leaders selected. Second, the demand for computer programmers in the external labor market was greater than it had been for several years. Many team members believed that Softwares 56


lack of titles not only made it difficult to compare jobs, but also to adequately describe job qualifications and histories to potential future employers. Nationally, the software development industry had distinct job levels, such as programmer, senior programmer, and developer. Yet, regardless of skills and experience, nearly everyone within Software was identified simply as a team member. The lack of job designations caused a level of uncertainty for some members who were more familiar with the roles and responsibilities associated with more traditional career positions. Third, Softwares performance problems, evident since its founding, reflected poorly on the viability of self-directed, high-performance teams. Fourth, the termination of the original president and subsequent replacement from one of Parents other technology units suggested to some that Softwares mission (i.e., to be different than the other units) no longer was viable. Coalitions To reiterate, the following criteria were used to determine coalitions in this research; they are informal groups (not identified as part of the organizations formal structure), lack their own (internal) formal structure, members have a general awareness of other members in the coalition, they focus on externally-oriented objectives, and have members who recognize a need for concerted or collaborative action. As described in the previous sections, analysis of the interview data revealed a subgroup of members who were perceived to be both influential and likely to collaborate to influence others. Table 8 (below) summarizes how this initial evidence has identified various actors with potential for participating in coalition activities. As the check marks indicate, all actors in this group were reported to be both influential and likely to collaborate to influence others. 57


Table 8 Evidence for Possible Coalition Activity Perceived to be Perceived to be Likely Actor Powerful/Influential To Collaborate Fred ! Greg ! Lou ! Pat ! Russ ! Sue ! Ule ! Following identification, the set of potential coalition members were interviewed to identify others in the organization with whom they interacted (contacted, communicated, etc.) regarding the team/structure issue. Then, those others were interviewed and asked to do the same. This process (referred to as "snowballing") continued until no new members were identified. In each case, interaction frequency data were captured and members were asked to explain their rationale for joining the group of influencers (i.e., the coalition), why they interacted in the manners observed and described, and to identify any activities or tactics used to influence others. Coalition 1. According to interview data, Fred was a reasonable starting point for coalition identification. He was a group leader (promoted from team leader) assigned to the Red project team. As with many other members, Fred was very concerned with Softwares structure, Particularly in terms of project success and the survival of the company. When asked if he did anything about the team structure issue he said, Weve had thousands of discussions and tried to form a committee. His focus was primarily performance oriented. For example, he offered, Some of the people around here dont get where the team boundaries need to be, you know, weve got to get moving and coding. 58


When asked if he planned to do anything about this issue Fred replied, Nothing out of the normal operation. Well maybe I can get a committee created to bring it up in front of [Walt, the president] in the right way. Hes new and probably doesnt appreciate what we can do here. I think we need to revisit how our structure influences the way we work, or dont work. I think weve been too loose. When asked if he thought other people were as concerned, he said, Yes, its just some of them want to keep this team thing in the center of what we do. I dont think they realize how crippled weve been just trying to identify what teams are. Finally, Fred was asked if he had contacted anyone regarding this issue. He replied, I bring it up when I can, I think its that important, and when I think its right. [Ule] probably more than anyone else, I think we see things, or at least this thing, the same way. Russ, Ann, Jim, Pat and Tina also were identified as targets for such discussions. Fred stated that A goal was to get enough people to Push for the creation of a committee charged with reviewing Softwares structure. While this required the support of the group leaders, previous committees had been formed based on visible support from team members across the organization. Freds strategy was to try to persuade other team members that structural changes, such as clearer roles and more defined reporting relationships, would benefit everyone and the projects. Fred even started using different language in electronic mail messages within Software in an attempt to get people out of the two-group mindset. For example, he began using the term associate rather than team member. Using the snowball technique mentioned earlier, the actors referenced by Fred were interviewed to validate coalition membership, interaction frequencies, and coalition objectives and strategies. Ule was a relatively new team leader on the Red project team. Ule put the structure issue in terms of leadership, stating, Im not sure a real or strong 59


leader can exist in this environment. It doesnt seem to be something that would fit with our version of project teams. He mentioned that he was willing to openly discuss this issue with anyone, but most of the original members preferred the way the organization was initially arranged. He also agreed that having a committee created would be the best way to get Walts attention and offset the strength of the other (opposing) group. He mentioned that their strategy would take time given the strength of the team concept within Software and the support that it received from other members. Russ, a team leader on the Green project confirmed interactions (discussions) with Ule and Fred and agreed to trying to build support for the creation of a committee. However, he was not as convinced as Fred and Ule that the team structure needed altering, but the problems were significant enough to warrant exploration of possible alternatives. Ann, Jim, Pat and Tina confirmed that several conversations had occurred with Fred and Ule in reference to the committee, but they did not completely share this view and most likely were not interested in changing the structure from what was originally intended. Generally, they believed that current operational problems could be corrected within the team-based environment. Consequently, these members were not considered part of the coalition. During their interviews Fred, Russ and Ule were asked to rate the frequency with which they interacted regarding the structure issue (using the same scale as the work and social self-report instruments). It was explained that interaction included planning as well as execution activities. In general, the groups strategy was to suggest the committee idea as a means of problem assessment when anyone brought up a structure-related issue. Basically, they planned to keep offering it as an exploratory step rather than openly recommending any specific change. The key to their strategy was to have a committee assess the structure and poor development performance and hope 60


Walt would make specific changes. The possible intervention by the president was important since their position was not as popular as the view of the other coalition. Coalition 1 analyses. Coalition 1 formed as a means of changing Softwares hierarchy to a more traditional model and redefining the way in which the team concept was interpreted. The coalition consisted of three Caucasian males (Fred, Russ and Ule), ages 28 to 36. All members were married and had relatively long tenure with the organization (average of 23.33 months). Each was a team leader or group leader. In addition, the members had some similar work histories and each completed at least a BS degree. Ule and Fred worked together at a previous employer and Ule and Russ worked together on a previous project (White). There was no agreement among the members on questions of political activity and actions making a difference in the world (questions 12, 13 in appendix A). All coalition members worked in close physical proximity to each other. Data from the Rokeach Value Survey (Table 9, appendix) showed a number of similarities. For example, on the terminal values, Fred, Ule and Russ ranked Inner Harmony nearly equally (4, 5 and 5 respectively). In addition, Russ and Ule ranked A Comfortable Life highest, Russ and Ule ranked Freedom third, Ule ranked Salvation 13 th and Russ ranked it 14 th and Russ and Fred ranked Self Respect 16 th and Social Recognition 15 th For the instrumental values, Russ and Ule ranked Ambitious high (one and two respectively), and Russ and Fred ranked Capable high (two and three respectively). Ule and Russ also ranked Logical high (5 th and 6 th respectively). The network diagram for Coalition 1 (Figure 4, appendix) depicts a density of 100% (all relationships present). Due to the coalitions small size, there was no core (centralization). Binary network centrality measures (Table 12, appendix) showed that in the social network, Fred reported interactions with nearly all other actors (25 of 26 61


others, 96%) and was reported as a social link by 24 others (92%). This indicated a high level of interaction agreement between Fred and other actors. However, both Ule and Russ reported fewer connections to others (50%), but were cited as links by others (88.5% and 84.6%), slightly above the normalized mean value of 77.635%. In the work network (Table 13, appendix), it was expected that the number of interactions would be fewer than those in the social network due to the requirement of work within project teams. However, Fred reported the maximum links to others (26, 100%) and was cited as a link by 19 other actors (73%). Russ too was above average for outbound links and approximately average for inbound links. Tables 14 and 15 (appendix) provide degree measures using valued relationship data. Unlike binary data, valued data reflects the strength or weight of reported links. The relative order did not change, Fred was highly central (well connected) and Ule and Russ reported low ties to others and approximately average from others. However, in the social network, the aggregated strength of the links reported to Fred (59) are much closer to the population mean (53). This suggests that the relatively value or frequency of the links is lower in magnitude or effect. In other words, while many actors reported links to Fred (binary data) the frequency or strength of the interactions (valued data) was lower. A similar condition existed in the valued work network (Table 15). Fred reported more and stronger links to others than were reported to him. Ule and Russ were closer to the population mean for work interactions. The next position-level analysis was structural equivalence within the coalition (Table 16, appendix). Though the coalition actors were not perfectly equivalent, it is clear that they had very similar relationships with each other. Due to the coalitions relatively small size, measures of regular equivalence were not helpful because all coalition members had similar interactions. 62


The final position measure was strength using N-clan data (Tables 20 and 21, appendix). For identification purposes, this measure required all links (not to exceed N steps) among actors to occur through other members of the group (rather than through actors outside the group). The value of N was set at one because higher values would result in a single group due to network size. Table 20 (appendix) shows the number of 1-clans identified in the work network and the frequency with which actors were included. Fred was identified in 33 of the 46 groups possible in the data set (71.74%). Ule and Russ were closer to the mean number of groups per actor (28.66%). In the social network (Table 21), Fred again was very highly ranked having been identified in 54 of the possible 75 possible (72%). Russ and Ule were not as frequently identified, being well below (18.67% and 13.33% respectively) the mean value of 29.88%. Freds potential to influence, or at least access, others was very high in each network. Coalition 2. As with the first case, the approach here was to use interview data to identify actors likely to collaborate as well as the actors with whom they did. The interview data suggested that Greg (among others) was a suitable starting point. He was perceived to be influential and was cited the most for engaging in collaborative attempts at influencing others. Greg was a team member on the Black project team. During the first interview, he identified compensation as the biggest issue facing the organization. However, since team-structure issues rated highest across the organization, Greg was asked his views on team and organization structure. He suggested that Softwares structure remained Innovative and offered significant (though unrealized) performance opportunities. Softwares current problems, according to Greg, stemmed simply from a lack of ground rules. However, he acknowledged the existence of two opposing groups. You basically have two camps, (Fred) and (Ule) and their followers, and (Pat), (Lou), 63


(Abe), (Sue) and me. We are the agents of change, we believe the most in the original vision. This was an interesting choice of words (and often repeated) because, as is explained below, this group didnt want to facilitate change. Rather, they wished to maintain the status quo. Greg was asked if he intended to do anything about the structure issue, particularly in light of the other group (Coalition 1). His response was that they (citing Lou, Sue, and Abe) intended to get the president and group leaders to agree to create (with the team members) a document that would outline the philosophies and procedures necessary to fulfill the original corporate mission. It was intended that all employees of Software would sign this document. The key to this strategy was that the process to create the agreement would not be as tedious as establishing and maintaining another committee. Also, it was believed that more people would be involved in this process, which would make it easier to build commitment to the original mission and structure. As with Greg, Lou (a team leader on the Black project) also did not identify Softwares structure as an issue during the first interview. However, when asked about the agreement document (as stated by Greg) he validated Gregs account. Interestingly, he commented that he was most likely to discuss issues with the original seven (members) because they Seem to be more active in preserving the (initial) mission. The original members were Ann, Fred, Greg, Jim, Lou, Russ and Abe. Lou confirmed Sue and Abe as team members involved in the effort to get an agreement document created. In addition, he cited group leader Pat. Lou also mentioned that he didnt believe the document would be enough to save the organization given their performance problems and the recent replacement of Softwares president. 64


Pat was the senior group leader in the organization and was part of the group that had helped to plan and create the organization. She believed the structure issue to be important and regarded it as a problem of Role identities. She explained that they (management) have not been able to Pin down what all these people are supposed to do. She mentioned that this resulted in A great deal of ambiguity and prevents us from identifying leadership. When asked about the effort to create an agreement document she verified earlier accounts and said, We have two distinct groups here; [Fred] people and [Pat] people, and it all revolves around differences in management style, the degree of comfort we try to produce and even our competitiveness. Sue was another team leader on the Black project (the only team with multiple team leaders). She too identified other issues (compensation and morale) during the first interview. When asked about the agreement document and collaborating with others she verified the strategy and mentioned that self-directed teams were supposed to function in such a manner. According to Sue, We all think team-based work should work like this, with open communication. Regarding the form of communication, she said, Most are unplanned, being in our team were naturally close to each other, our days are very unstructured so we can cover what we want when we want, face-to-face is fastest and easiest. Finally, she confirmed collaboration with Greg, Lou and Abe. Abe was a team member on the Green project team. Again, structure was not an issue identified during his first interview. He verified collaboration with Greg, and said he also worked with Lou and Jim. Jim was a team member on the Green project team. He reported that Very few issues concern him and that he always tried to operate through Formal channels to avoid a lot of what is going on around here. Consequently, Jim was omitted from coalition consideration. 65


Coalition 2 analyses. This coalition was formed to clarify, but generally maintain, Softwares current use of teams. It consisted of five Caucasian members (Greg, Lou, Abe, Sue and Pat), 35 to 39 years old. All had long tenure with the organization (average 24 months). All had attained at least a BS degree. Greg and Sue worked together at a previous organization and Greg, Lou and Abe worked together on a previous project. Pat, Lou, and Sue believed their actions did make a difference in the world. These three actors also considered themselves politically active. All except Abe worked in close physical proximity to each other. As with Coalition 1, the RVS data showed widespread similarities when member pairs were evaluated. Nearly all instrumental values showed some level of similarity within the coalition (at least between pairs). Also, as with Coalition 1, the terminal value Equality and the instrumental value Obedient are uniformly ranked near the bottom of their respective lists. The instrumental value Honest was ranked first for three members of the group (actors Lou, Sue and Abe). The network diagram of Coalition 2 is provided in the appendix (Figure 5). Again, due to the groups relative size, network density was 100% (all possible interactions present) and centralization was zero (no core). A review of binary centrality measures for the social network (Table 12) shows that four of the five coalition members were above the mean for both out-degree and in-degree scores. Greg, Lou and Abe had the highest reported inbound links (25, 24 and respectively) and Lou (26) and Greg (25) were among the highest in out-bound links. Valued data for the social network (Table 14) confirmed the relative centrality of these relationships. In the binary work network (Table 13), all coalition members had centrality measures above the mean for both in and out-degrees. Similar findings are reported in the valued centrality measures (Table 15). Four 66


of the five members were above the mean. Abe (51.0 for out and in-degree) was just below the mean of 52.259. Structural equivalence measures for Coalition 2 members are found on Table 17 (appendix). The data show that Sue and Lou were most similar (r = .91), but all were correlated at a level of .73 and above. N-clan analysis of the work network (Table 20) show that all coalition members had relatively high levels of connectedness. All scored well above the mean of 28.66%. In fact, Coalition 2 members held five of the top nine values. In the social network (Table 21), all members except Pat were above the mean. Lou and Greg were identified in 60 and 59 of the 75 clusters. Comparative analyses. These analyses assess all coalition members to other actors (non-coalition members) in the organization. These analyses include group means comparisons of the RVS data, network-wide measures of equivalence, relationship importance (Lambda sets), and network correlations. The mean RVS rank data show some key distinctions between Coalition 1 members and non-coalition members (Table 10, appendix). Since the present study hopes to distinguish coalition members from non-coalition members the focus with RVS data primarily was on dissimilarity. This was accomplished by calculating the difference in numerical order for each value pair (Table 11, appendix). For example, if the average ranks for the terminal value Wisdom in two groups were first and fifth, then the reported ordinal difference (delta) would be four (rank or ordinal places). The similarity (or dissimilarity) in overall rank order suggests the similarity (or dissimilarity) in overall value systems. The largest dissimilarities in the terminal values between Coalition 1 members and non-members are Self-Respect (difference of 11 ordinal places, 14 th and 3 rd respectively), Freedom (eight places), An Exciting Life (eight places), Family Security 67


(seven places) and A World of Beauty (seven places). The strongest similarities in the terminal values are Salvation with the same (last) ranking, and Equality, A Sense of Accomplishment, Pleasure and National Security differing by only one ordinal place. The largest dissimilarities in the instrumental values between Coalition 1 members and non-members are Responsible (12 places, 14 th and 2 nd respectively), Intellectual (10 places), Polite (10 places) and Clean and Cheerful (seven places each). The strongest similarities are Obedient (same rank, 18 th ) and Forgiving (same rank) and Courageous, Broadminded and Logical, which are separated by only one ordinal place. The RVS means comparisons between Coalition 2 and non-coalition members (Tables 10 and 11) are interesting. The largest dissimilarity in the terminal values between coalition 2 members and non-members is Freedom (eight places, 3 rd and 11 th respectively). However, there are numerous similarities between these groups. A World at Peace, An Exciting Life, Happiness, Inner Harmony, Pleasure, and Wisdom have the same rankings. In addition, Self-Respect and Social Recognition differ by only one ordinal place between the groups. The largest dissimilarities in the instrumental values between Coalition 2 members and non-members are Logical (nine places, 15 th and 6 th respectively), Intellectual (nine places), Ambitious (eight places), Courageous and Loving (seven places each). Again, there are several strong similarities in the instrumental values. Capable, Clean, Honest, Obedient and Responsible have the same rank order between the groups and Helpful differs only by one (10 th and 11 th respectively). The final RVS assessment focuses on the similarity of mean rankings between Coalition 1 and Coalition 2 (Table 11, appendix). In the terminal values segment the most dissimilar values are Self-Respect (10 places) and An Exciting Life (eight places). 68


A Comfortable Life, Freedom and True Friendship were exact matches while Pleasure differed by only one place. For the instrumental values, Ambitious and Responsible show the largest separation (12 places) followed by Logical (eight places) and Clean (seven places). Independent and Obedient reflect the greatest similarities (same ranking) between Coalition 1 and 2. In addition, Broadminded, Intellectual, Loving, and Self-Controlled are separated by one place. Table 18 (appendix) shows the structural equivalence cluster diagram for Softwares work network. The most structurally equivalent, or most similar actors were Jim and Abe (.959), Ann and Yuri (.944) and Dan and Tina (.926). These pairs of actors worked on the same project teams (Green, Red and Black respectively). The highest grouping of coalition members was Greg and Lou at .841. In other words, approximately 84% of their relationships to others were similar (but not as similar as some relationships between non-coalition members). Russ and Abe at .892, but this cluster also includes non-coalition members Jim and Bob. Table 19 (appendix) shows the structural equivalence clusters for Softwares social network. Vic and Orin were the most equivalent and worked on the same project team (Black). However, Lou and Sue were similar at .745 and joined with Greg at .671. Regular equivalence measures were not included because all actors aggregated at 92.7 across both networks. This was again most likely due to overall network size and the inherent likelihood of relationship similarity. It is more likely (easier) to have similar relationships with others when the number of others is small. Lambda set analysis was used to identify important pairs or groups of actors in the networks. Important refers to the likelihood of network disruption if such pairs or 69


groups were removed. A lambda value is generated to give a measure of cohesion. The higher the lambda value the more network disruption likely to result from pair removal. Table 22 (appendix) shows the Lambda set results for Softwares work network. Cece and Pat were most important. This was not unexpected because Pat was the senior Group Leader and took on additional work responsibilities during the change in Softwares president. Cece was the office manager and maintained work-based relationships with nearly all other actors. However, coalition members clearly held the most important positions in the network (six of the top seven positions, 85.7%). In the social network (Table 23, appendix) coalition members were more distributed, but coalition members (Fred, Lou, Greg and Abe) again occupied four of the top five (80%) values, with Lou, Greg and Fred most important. The final structural analysis in this research was to test coalition interaction patterns for correlation with existing work and social interaction patterns for the same actors. This was accomplished by extracting the coalition member relationships from the original work and social network data sets. In other words, the interaction values for the three actors that comprised Coalition 1 were extracted from the entire Software work and social data. This was done because network correlation algorithms (quadratic assignment procedure or QAP) required equally sized matrices for calculation. QAP results for each coalition are in Table 24 (appendix). For each coalition, the members interaction patterns were more closely correlated to their social relationships (.657 and .594 to .379 and .205 respectively). Chapter Summary To reiterate, this research identified two coalitions pertaining to Softwares structural issues. Members of Coalition 1 (Fred, Ule and Russ) worked together to facilitate an organizational change they perceived was necessary. They wanted to try to 70


build support for the creation of a committee to investigate structural alternatives. It was hoped that a committee would increase visibility with the president to counter the opposing (and majority) view that Softwares structure should remain as-is. Coalition 1 was quite homogeneous in terms of age, race, education, gender, tenure and marital status. Each member was either a group lead or team lead. Two members (Fred and Ule) worked on the same project team and even had a shared history with a previous employer. All members were perceived to be influential. Coalition 2 (Greg, Lou, Sue, Pat and Abe) was formed in response to the same issue as Coalition 1. However, its members wished to preserve Softwares original structure and team-based orientation. They favored the creation of a binding document that specifically outlined the roles and policies necessary to make sure the original structure was preserved. Their objective was to keep the process informal (i.e., no committee) because they believed their strength was in the majority opinion they supported. A formal procedure such as a committee with more direct involvement by the president could allow for more unilateral (and possibly unfavorable) decisions. Coalition 2 was more heterogeneous than Coalition 1, but certainly not diverse. All were Caucasian with long tenure and similar educations, but age, gender and marital status varied more than with Coalition 1. Members were group leads, team leads and team members. Actors Greg and Sue worked together at a previous organization. All except Abe were believed to be influential. 71


Chapter Six Discussion This research explored coalition formation within an organization as a means of offering a better approach to the study of these important groups. The questions that drove this investigation focused on the nature of the relationships that comprise coalitions as well as their aggregated structural features. The methods, instruments and data used proved to be suitable for coalition research. In this section I discuss the overall study, provide comments pertaining to the existing literature, review the use of social network analysis and the Rokeach Value Survey, discuss limitations of the study, and offer avenues for future research. This research was an exploratory case study undertaken within a functioning organization. The people, issues and history involved in these activities were identified through semi-structured interviews. Surveys were used to assess demographic factors and members work and social interaction patterns. The Rokeach Value Survey was used to assess population and group value systems (a proxy for belief systems). Overall, coalition identification proved to be less challenging than expected. This may have been due to the open communication that had been preached since Softwares creation and thus more difficult in larger, more established organizations where formal and informal behavioral constraints may exist. There are several comments pertaining to formation factors (variables) in the extant literature and findings in the present study. First, Softwares coalitions seemed to form in a manner consistent with explanations found in existing research. Each was 72


guided by members beliefs or views concerning an issue. In each case, coalition members thought strongly about Softwares structure. In fact, those organization members that were somewhat indifferent did not participate in influencing attempts (though they had opinions). Second, neither coalition represented a majority and reward maximization didnt apply. However, there was a desired outcome (structural change or status quo) and it had the potential to impact the entire organization. Third, coalition size played a role, but not in the distribution of a reward. Instead, size determined the strategy employed by the coalitions. Coalition 1 hoped the creation of a committee would couple or bind an assessment of Softwares structure and its performance problems. The members felt this would force Walt to make a unilateral decision and change the organization. This action would bypass the perceived strength of Coalition 2 and the perception that most non-members preferred the status quo. Coalition 2s strategy reflected the desire to increase strength by adding members through simple, person-to-person communication. The agreement creation process would allow all organization members to participate in discussions and document editing. Coalition members believed this would enable them to address non-members objections and provide time to tailor their message (and influence others). Lastly, power appeared to be important to the coalitions in this study, but not as a means of actor domination (Caplow, 1956, 1959; Chertkoff, 1967). Rather, each coalition seemed to act from positions they believed were low in power. This is consistent with the idea that coalitions develop as a result of their perceptions of relative low power (Mannix & White, 1992). The basis of this perception was that neither group had any established links to the new president. Thus, each coalition attempted to influence others to yield a favorable decision. Interestingly, each group was acting based on the potential impact of the issue. There was no pending decision regarding Softwares structure. 73


As mentioned, other research has suggested the possibility that similar socio-demographic variables may contribute to coalition formation. Demographic data from this study confirms that coalition members have somewhat strong similarities with each other. All had relatively long tenure with the organization, similar educational backgrounds, marital status, and age (though Coalition 1 was more homogenous). Indeed, members had strong similarities within and across coalitions. In other words, demographics provided some distinction between coalition members (combined) and non-coalition members within Software. Interestingly, while tenure was not the sole determinant in formation (not all senior members participated), none of the contractor employees or actors with less than 19 months tenure participated in either coalition. Such similarities among coalition members support the idea that potential members seek out and collaborate with similar others (Boros et al., 1997; Dreze & Greenberg, 1980). Such similarities may convey, until otherwise known, signs of positive identification. Ones positive identification with others has been associated with friendship, trust or some degree of comfort that supports collaborative action, particularly in controversial scenarios (Ancona & Caldwell, 1992; Goffman, 1969; McPherson & Smith-Lovin, 1987). However, it might be that the positive identifications that lead to collaborative (coalition) activity are based on similar perceptions rather than direct matches among socio-demographic factors. This idea is well documented in the social network literature (Burt, 1982; Carley & Krackhardt, 1996; Casciaro, 1998; Ibarra & Andrews, 1993). For example, given that most coalition members were part of the founding group (five of seven) it may be that they interpreted Softwares ongoing performance problem as a threat to something they perceived to have created and to which they strongly identified. 74


Perhaps the members of Coalition 1 could more easily let go of the original design and its philosophy, or they desired the opportunity to influence, if possible, a new design. Social network analysis provided a useful method for understanding coalitions, particularly coalition members and the key positions they occupied. I stated in the previous chapter that certain positions might enable members to influence or communicate with (by having access to) others more easily. Though it may be that certain organization members are more likely to join coalitions due to their relative position. In summary, coalition members generally were more central, demonstrated positional similarity (equivalence) among members of each coalition, and had more opportunities to influence. Lambda set analysis demonstrated that coalition members were more critical (important) to the work network (Table 22, appendix). Coalition members occupied six of the first seven positions in the cluster (and eight of the top 12). Coalition members were more distributed in the critical positions in the social network (Table 23), but still produced the highest three measures. Similarly, in the N-clan analyses, coalition members accounted for eight of the 13 (61.5%) highest reported actors (most connected, more robust, etc.) in the work network (Table 20) including the three most frequently clustered actors (Lou, Fred and Pat). While coalition members were more distributed in the social network (Table 21), the highest reported actors in the organization and five of the highest eight were coalition members. Finally, there were high levels of structural equivalence within each coalition. In the work network (Table 16), the greatest equivalence was found among non-coalition members. The primary exception to this was Abe who was nearly perfectly equivalent (.959) to Jim. Multiple coalition members did not demonstrate any similarity until the .84 75


level. In the social network (Table 17), more of the initial (highest) pair similarities involve coalition members. However, clusters did not begin until the .76 level. The Rokeach Value Survey also provided a benefit to this study, but as is explained below, was hampered by the organizations size. While it was difficult to interpret overall systems (ranks) for the entire population, the RVS worked well when comparing coalition members and means between coalition and non-coalition actors (Tables 11 and 12, appendix). Most surprisingly, members of Coalition 2 appeared to be quite similar to non-coalition members, particularly with instrumental values. This further strengthens the correlation found between coalition members and their social (non-work) interactions. In addition, this similarity may help explain the uphill battle members of Coalition 1 described in swaying the popular opinion regarding the status quo. It also suggests that Coalition 2 had the potential to be significantly larger. While it is a benefit that the RVS enables the identification of values for similarity assessment, I was surprised at the values that frequently matched between groups. For example, Coalitions 1 and 2 ranked the terminal values A Comfortable Life first, Freedom third and True Friendship ninth. Though they differed on Equality and Self-Respect which were expected to be more valued in such a work setting. Research Limitations While this research provided some benefits to the study of intra-organizational coalitions, there are some important limitations that should be addresses. The first issue concerns Softwares size. While the relatively small organization enabled more complete data collection it also restricted nearly all analyses in one way or another, with the most significant impacts on SNA and RVS data. For example, in terms of network analytics, the ease with which organization members could maintain multiple types of relationships with a relatively large percentage of the organization had the effect of limiting structural 76


separation or segmentation by relationship type (i.e., work and non-work) as well as between coalition members and non-members. A larger organization may have enabled larger coalitions and relatively less dense relationships, which most likely would have identified greater structural distinctions within and across groups. Softwares size also had an effect on the use the Rokeach Value Survey data. The research population (N=27) facilitated a cursory evaluation of pair-wise comparisons, but was insufficient for more rigorous statistical testing. Obviously, a larger organization would permit more detailed measurement of member and non-member rankings. So, even though standard statistical measures are not applicable to ordinal data, a larger dataset would allow measurement and inferences on specific actor pairs and value ranks (made non-ordinal). Therefore, in the present study, the effects of actor belief systems are incomplete. The second limitation is concerned with the use of broad value systems or ideology in this study. Though similarities and dissimilarities were found, the number of values in the RVS didnt provide a means of clearly distinguishing coalitions members or non-members. Again, this is certainly an artifact of Softwares small size. Though, it also may be that the comprehensive nature of the survey (i.e., a complete value system) doesnt cleanly address key elements of members belief systems in organizational life. For example, the constraints on individuals at work and quite possibly the underlying need to remain employed may over-ride or suppress all responses (e.g., coalition activity) except those triggered by dramatic differences between key values and perceptions. In other words, it is possible that most individual values within a system are held in check until only relatively large conflicts between a value and an organizational event occurs. In any event, a more focused or organization-specific instrument (perhaps one concerning issues of governance or fairness) may provide more insight into the 77


suspected link between perception or beliefs and coalition activity. Thus, while coalition activity was issue driven, it is uncertain whether values influenced the perception of the issue and the impetus for coalition activity. The final limitation of the present study is the constrained time frame of the research. I did find support for the idea that coalition interactions parallel social relationships so that certain assumptions can be made about how they are likely to change over time (i.e., follow friendship patterns). However, given that value and belief systems, perceptions of events and even relationship tend to change over time, the overall link between these variables, say through multiple issues and coalitions, remains unclear. Future Research I contend this study has provided a richer picture of organizational coalitions and, more importantly, a foundation for a more productive research base. The methods employed in the current study demonstrate that identifying and exploring coalitions is quite possible. Yet, as suggested in the preceding limitations, it is only a first step. As I have argued, coalition research should continue to include and expand upon the use of socio-demographic and contextual factors. This will further improve the detail and accuracy within the field. More specifically to the present study, future coalition research should address the issues of organization size, longitudinal methodologies, values and beliefs, and the links to other existing literatures. Building on the present study with larger organizations would provide several benefits. First, a larger population would enable more complete social network assessments. The simple fact that people can maintain only a limited number of work and social relationships suggests that larger research populations would demonstrate 78


unique (or less hidden) structural features. Second, larger organizations would allow for more rigorous analysis (i.e., statistical) of RVS data. Follow-on work also should address coalition activity using multiple time periods. This would help bring additional understanding to activities such as coalition life cycles, selection of and changes in tactics and membership, and possibly even internal coalition structure (hierarchy). An additional benefit of longitudinal studies would be greater clarity of causality in coalition activity. For instance, a coalition members high centrality in the social network may be due to her attempts to influence others, rather than as a structural quality developed prior to coalition activity. I suspect the latter is more likely since people tend to report interaction data based on their patterns over time. More interestingly, consider the possibility that members form coalitions (around some antecedent factor) and then identify and communicate an issue. Following Weicks (1979) concept of sense making in organizations, in this case an issue might not exist until a coalition says that it does. Obviously, assessments conducted at multiple time periods would aid in the explanation of these possible events. The final argument for a longitudinal study is the fact that value systems tend to change over time (Kamakura & Mazzon, 1991; Ovadia, 2004). In the present study, the RVS data simply are a snapshot of a single point in Softwares existence. While it is doubtful the value rankings would have changed over the duration of the study (eleven weeks), it is likely that value priorities eventually would change given the shifting and competing internal and external pressures common in organizational life. Being able to identify and integrate such changes with SNA data would almost certainly provide a unique and more comprehensive picture of organizational coalitions. Another avenue for improvement is the continued exploration of values and belief systems in coalition research. A logical first step would be to duplicate this study using 79


the RVS in a larger organization. As mentioned throughout this study, the suspicions of a link between member perceptions and coalition activity are strong. Future coalition research would benefit from a more complete test of the RVS. However, I do recognize that a more focused instrument, perhaps one that measures ideas of organizational governance, fairness or even altruism (the list of possible key values could be long), might more easily make the connection between beliefs and coalition activity. Finally, it would be prudent to explore links between coalition research and existing literatures. One area of organization studies that is more expansive than and may provide guidance to coalition research is the work on organization culture, particularly subcultures. The pervasive nature of subcultures in organizations has been well documented (Bloor & Dawson, 1994; Boisnier & Chatman, 2002; Cameron & Quinn, 1999; Chatman, 2002; Flynn & Chatman, 2001; Hofstede, 1998; Jermier, Slocum, Fry, & Gaines, 1991; Martin, 1992; Schein, 1996; Smircich, 1983; Trice, 1993; Van Maanen & Barley, 1984). Subcultures are groups whose common characteristic is a set of shared norms and beliefs. Similar to coalitions, subcultures can (and do) form around informal elements of organizations (Trice & Beyer, 1993). As with coalitions, subcultures have been compared to relatively small clusters of organization members that share a set of norms, values, and beliefs (Boisnier & Chatman, 2002). Several researchers have suggested that cultures and subculture may enable organizations to adapt to change (Boisnier & Chatman, 2002; Gagliardi, 1986; OReilly, 1989; Saffold, 1988; Tushman & OReilly, 1997). It may be that coalitions, or the recognition of them, are a subcultural element for change. In other words, coalitions may be one or more subcultures response to a stronger or more dominant cultures (i.e., the corporate culture) inability to respond and adapt. Just as I have argued that social network analysis can integrate (advance) laboratory and field research into coalitions, 80


SNA may also provide a key means by which research on subcultures and coalitions combine. A deeper investigation into the literatures on social movements and planned change also shows promise for advancing the current study. Social movements are identified as the building and reproducing of dense informal networks between a multiplicity of actors, sharing a collective identity, and engaged in social and/or political conflict (Diani & Bison, 2004). Though these tend to be multi-organizational or community focused, social movement research may help highlight and uncover possibilities for truly socially oriented coalition activity. This is supported by Simpsons (2004) notion of fairness and the idea that group (coalition) members can choose to act in concert for the collective good of a larger group. From a coalition perspective, planned change within organizations represents a favorable use or outcome of coalition activity. This is not to suggest that coalitions are not viewed favorably in the subculture or social movement literatures. Rather, planned change specifically focuses on the benefits of informal groups in overcoming institutional impediments to change. Again, it may be that the coalitions are necessary change mechanisms as suggested by Buchanan and Badham (1999). Softwares poor performance was a known threat to its existence. Perhaps the coalitions were a response to that threat and helped focus attention on other (i.e., inappropriate) corporate activities. 81


References Adkins, B. & Caldwell, D. 2004. Firm or subgroup culture: where does fitting in matter most? Journal of Organizational Behavior, 25: 969-928. Alwin, D.F. & Krosnick, J.A. 1985. The measurement of values in surveys: A comparison of ratings and rankings. Public Opinion Quarterly, 49: 535-552. Ancona, D.G. 1990. Outward bound: Strategies for team survival in an organization, Academy of Management Journal, 33: 334-65. Ancona, D.G. & Caldwell, D.F. 1992. Bridging the boundary: External activity and performance in organizational teams. Administrative Science Quarterly, 37:634-665. Andeweg, R. 2000. Ministers as double agents? The delegation process between cabinet and ministers, European Journal of Political Research, 3: 377-395. Axelrod, R. 1970. Conflict of Interest: A Theory of Divergent Goals with Applications to Politics. Chicago: Markham Publishing Company. Bacharach, S.B. & Lawler, E. 1980. Power and Politics in Organizations. San Francisco: Jossey-Bass. Back, H. & Dumont, P. 2004. A combination of methods: a way forward in coalition research. Presentation for the Annual Meeting of the American Political Science Association. Baron, D. & Diermeier, D. 2001. Elections, governments, and parliaments under proportional representation. Quarterly Journal of Economics, 116: 933-967. Baron, R. A. & Markman, G.D. 2003. Beyond social capital: The role of entrepreneurs social competence in their financial success. Journal of Business Venturing, 18(1): 41-60. Baum, J.A.C., & Calabrese, T. 2000. Dont go it alone: Alliance network composition and startups performance in Canadian biotechnology. Strategic Management Journal, 21(3): 267-294. Baumeister, R. F., & Leary, M. R. 1995. The need to belong: Desire for Interpersonal attachments as a fundamental human motivation. Psychological Bulletin, 117: 497-529. Bazerman, M.H. 1985. Norms of distributive justice in interest arbitration. Industrial and Labor Relations Review, 38: 558-570. 82


Bazerman, M.H. 1986. Coalitions in organizations: An introduction. Research on Negotiation in Organizations, 1: 93-96. Bazerman, M.H., Curhan, J.R., Moore, D.A. & Valley, K.L. 2000. Negotiation. Annual Review of Psychology, 51: 279-314. Beersma, B. & De Dreu, C.K.W., 1999. Negotiation processes and outcomes in prosocially and egoistically motivated groups. International Journal of Conflict Management, 10: 385-482. Beersma, B. & De Dreu, C.K.W., 2002. Integrative and distributive negotiation in small groups: Effects of task structure, decision rule, and social motive. Organization Behavior and Human Decision Processes, 87: 227-252. Bennett, A. 2002. Where the model frequently meets the road: combining statistical, formal, and case study methods. Paper prepared for presentation at the 2002 APSA Conference in Boston, August 29September 1, 2002. Berg, B.L. 1995. Qualitative Research Methods for the Social Sciences. Boston: Allyn and Bacon. Beyer, J.M.; Dunbar, R.L.M. & Meyer, A.D. 1988. Comment: The concept of ideology in organizational analysis. Academy of Management Review, 13: 483-489. Blackburn, R. & Cummings, L.L. 1982. Cognitions of work unit structure. Academy of Management Journal, 25: 836-854. Blau, P. 1977. Inequality and Heterogeneity. New York: Free Press. Bloor, G., & Dawson, P. 1994. Understanding professional culture in organizational context. Organization Studies, 15: 275-295. Boisnier, A. & Chatman, J.A. 2002. The role of subcultures in agile organizations. Draft manuscript, University of California, Berkley. Borgatti, S.P.; Everett, M.G. & Freeman, L.C. 2002. UCINET for Windows. Software for social network analysis. Harvard: Analytic Technologies. Borgatti, S.P. & Foster, P.C. 2003. The network paradigm in organizational research: A review and typology. Journal of Management, 29:991-1013. Boros, E., Gurvich, V., & Vasin, A. 1997. Stable families of coalitions and normal hypergraphs. Mathematical Social Science, 34: 107-123. Braithwaite, V. A., & Law, H. G. 1985. Structure of human values: testing the adequacy of the Rokeach Value Survey. Journal of Personality and Social Psychology, 49: 250-263. 83


Brass, D.J. 1984. Being in the right place: A structural analysis of individual influence in an organization. Administrative Science Quarterly, 29: 518-539. Brass, D.J. 1992. Power in organizations: A social network perspective. Research in Politics and Society, 4: 295-323. Brewer, H.B. 1981. Ethnocentrism and its role in interpersonal trust. In M. Brewer & B. Collins (eds.) Scientific Inquiry and the Social Science, San Francisco: Jossey-Bass. Brown, J.S. & Duguid, P. 2000. The Social Life of Information. Cambridge: Harvard Business School Press. Brown, S.L. & Eisenhardt, K.M. 1997. The art of continuous change: Linking complexity theory and time-paced evolution in relentlessly shifting organizations. Administrative Science Quarterly, 42: 1-34. Buchanan, D., & Badham, R. 1999. Politics and organizational change: The lived experience. Human Relations, 52: 609-629. Buchholz, R.A. 1976. Measurement of beliefs. Human Relations, 29: 1177-1188. Buchholz, R.A. 1978. An empirical study of contemporary beliefs about work in American society. Journal of Applied Psychology, 63: 219-227. Burt, R.S. 1976. Positions in networks. Social Forces, 55: 93-122. Burt, R.S. 1982. Toward a Structural Theory of Action. New York: Academic Press. Burt, R.S. 1992. Structural Holes: The Social Structure of Competition. Cambridge: Harvard University Press. Cameron, K.S. & Quinn, R.E. 1999. Diagnosing and Changing Organizational Culture: Based on the Competing Values Framework. Reading, MA: Addison-Wesley. Caplow, T. 1956. A theory of coalitions in the triad. American Sociological Review, 21: 489-493. Caplow, T. 1959. Further development of a theory of coalitions in the triad. American Journal of Sociology, 64: 488-493. Carley, K. 1991. A theory of group stability. American Sociological Review, 56(3): 331-354. Carley, K.M. & Krackhardt, D. 1996. Cognitive inconsistencies and non-symmetric friendship. Social Networks, 18:1-27. Casciaro, T. 1998. Seeing things clearly: Social structure, personality, and accuracy in social network perception. Social Networks, 20: 331-351. 84


Cechlarova, K., & Romero-Medina, A. 2001. Stability in one-sided matching markets. International Journal of Game Theory, 29: 487-494. Chen, C.C., Chen X., Meindl, J.R. 1998. How can cooperation be fostered? The cultural effects of individualism-collectivism. Academy of Management Review, 23: 285-304. Chertkoff, J.M. 1967. A revision of Caplow's coalition theory. Journal of Experimental Social Psychology, 3: 172-177. Chertkoff, J.M. 1970. Sociopsychological theories and research on coalition formation. In S. Groennings, E.W. Kelley, & M. Leiserson (eds.), The Study of Coalition Behavior: Theoretical Perspectives and Cases from Four Continents. New York: Holt, Rinehart & Winston. Cobb, A.T. 1982. A social psychological approach to coalition membership: An expectancy model of individual choice. Group & Organization Studies, 7: 295-319. Cobb, A.T. 1986. Coalition identification in organizational research. Research on Negotiation in Organizations, 1: 139-154. Cobb, A.T. 1991. Toward the study of organizational coalitions: Participant concerns and activities in a simulated organizational setting. Human Relations, 44: 1057-1079. Cook, K.S. & Gilmore, M.R. 1984. Power, dependence, and coalitions. Advances in Group Processes, 1: 27-58. Crosby, L.A.; Bitner, M.J. & Gill, J.D. 1990. Organizational structure of values. Journal of Business Research, 20: 123-134. Crott, N.W. & Albers, W. 1981. The equal division kernel: An equity approach to coalition formation and payoff distribution in N-person games. European Journal of Social Psychology, 11: 285-306. Cyert, R.M. & March, J.G. 1963. A Behavioral Theory of the Firm. New Jersey: Prentice-Hall. De Dreu, C.K.W., Harinck, S., & Van Vianen, A. 1999. Conflict and performance in groups and organizations. In C.L. Cooper & I.T. Robertson (eds.) International Review of Industrial and Organizational Psychology, 369-414. Chichester: Wiley. DeSwaan, A. 1970. An empirical model of coalition formation as an n-person game of policy distance minimization. In S. Groennings, E.W. Kelley, & M. Leiserson (eds.), The Study of Coalition Behavior: Theoretical Perspectives and Cases from Four Continents. New York: Holt, Rinehart & Winston. DeSwaan, A. 1973. Coalition Theories and Cabinet Formations. San Francisco: Jossey-Bass. 85


DeSwaan, A. 1985. Coalition theory and multi-party systems. In H.A.M. Wilke (ed.), Coalition Formation. Amsterdam: North-Holland. De Vries, M. 1999. Governing with your closest neighbour: an assessment of spatial coalition formation theories. PhD Dissertation: University of Nijmegen. De Winter, L., Andeweg, R.B. & Dumont, P. 2002. The state of the art in coalition research: critical appraisals and alternative avenues. Paper presented at the ECPR Joint Sessions in Torino, March 2227, 2002. Diani, M. & Bison, I. 2004. Organizations, coalitions and movements. Theory and Society, 33: 281-309. Diermeier, D. & Merlo, A. 2000. Government turnover in parliamentary democracies. Journal of Economic Theory, 94: 46-79. DiMaggio, P. 1992. Nadel's paradox revisited: Relational and cultural aspects of organizational structure. In N. Nohria & R.G. Eccles (eds.), Networks in Organizations: Structure, Form, and Action. Boston: Harvard Business School Press. DiMaggio, P. & Louch, H. 1998. Socially embedded consumer transactions: For what kinds of purchases do people most often use networks? American Sociological Review, 63:619-637. Dodd, L.C. 1974. Party coalitions in multiparty parliaments: A game-theoretic analysis. American Political Science Review, 68: 1093-1117. Dreze, J.H., & Greenberg, J. 1980. Hedonic coalitions: Optimality and stability. Econometrica, 48: 987-1003. Drory, A. & Romm, T. 1990. The definition of organizational politics: A review. Human Relation, 43: 1133-1153. Eagleton, T. 1991. Ideology: An Introduction. New York: Verso Press. Eisenhardt, K.M. & Bourgeois, L.J. 1988. Politics of strategic decision making in high-velocity environments: Toward a midrange theory. Academy of Management Journal, 31: 737-770. Erickson, B.H. 1988. The relational basis of attitudes. In B. Wellman & S.D. Berkowitz (eds.), Social Structures: A Network Approach. New York: Cambridge University Press. Fenger, M & Jean-Klok, P. 2001. Interdependency, beliefs, and coalition behavior: A contribution to the advocacy coalition framework. Policy Sciences, 34: 157-170. 86


Flynn, F., & Chatman, J. A. 2001. Strong cultures and innovation: Oxymoron or opportunity? In Cartwright, S., Cooper, C., Earley, C., Chatman, J., Cummings, T., Holden, N., Sparrow, P. & Starbuck, W., (Eds.), International Handbook of Organizational Culture and Climate: 263-287. Sussex: John Wiley & Sons. Fombrun, C.J. 1983. Attributions of power across a social network. Human Relations, 36: 493-508. Freeman, L.C.; Romney, A.K. & Freeman, S.C. 1987. Cognitive structure and informant accuracy. American Anthropologist, 89: 310-325. Friedkin, N.E. & Johnsen, E.C. 1999. Social influence networks and opinion change. Advances in Group Processes, 16: 1-29. Frost, P.J. 1987. Power, politics, and influence. In F.M. Jablin, L.L Putnam, K.H. Roberts, & L.W. Porter (eds.), Handbook of Organizational Communication: An Interdisciplinary Perspective. Newbury Park: Sage Publications. Gagliardi, P. 1986. The creation and change of organizational cultures: A conceptual framework. Organization Studies, 72: 117-134. Gamson, W.A. 1961. An experimental test of a theory of coalition formation. American Sociological Review, 26: 373-382. Gamson, W.A. 1964. Experimental studies of coalition formation. In L. Berkowitz (ed.), Advances in Experimental Social Psychology. New York: Academic Press. Gentry, M.E. 1987. Coalition formation and processes. Social Work with Groups, 10(3): 39-54. Georgopoulos, B.S. 1986. Organization Structure, Problem Solving, and Effcetiveness. San Francisco: Josey-Bass. Gerring, J. 2004. What is a case study and what is it good for? American Political Science Review, 98: 341. Goffman, E.G. 1969. Strategic Interaction. Philadelphia, PA: University of Pennsylvania Press. Goll, I. & Zeitz, G. 1991. Conceptualizing and measuring corporate ideology. Organization Studies, 12: 191-207. Granovetter, M.S. 1973. The strength of weak ties. American Journal of Sociology, 78: 1361-1380. Gross, F. 1985. Ideologies, Goals, and Values. Westport, CT: Greenwood Press. Hofstede, G. 1998. Identifying organizational subcultures: An empirical approach. Journal of Management Studies, 35: 1-12. 87


Hosking, D. & Fineman, S. 1990. Organizing processes. Journal of Management Studies, 27: 583-604. Howard-Greenville, J.A. 2006. Inside the black box: How organizational culture and subcultures inform interpretations and actions on environmental issues. Organization & Environment, 19:46-73. Ibarra, H. & Andrews, S.B. 1993. Power, social influence, and sense making: Effects of network centrality and proximity on employee perceptions. Administrative Science Quarterly, 38: 277-303. Ibrahim, F.A. & Kahn, H. 1988. Assessment of worldviews. Psychological Reports, 60: 163-176. Jehn, K. 1994. Enhancing effectiveness: An investigation of advantages and disadvantages of value based intragroup conflict. The International Journal of Conflict Management, 5: 223-238. Jehn, K. 1995. A multimethod examination of the benefits and detriments of intragroup conflict. Administrative Science Quarterly, 40: 256-282. Jermier, J. M., Slocum, J. W., Fry, L.W., Gaines, J. 1991. Organizational subcultures in a soft bureaucracy: Resistance behind the myth and facade of an official culture. Organization Science, 2: 170-194. Kagan, D.M. 1986. Construct validity of belief systems subscales. The Journal of Social Psychology, 126: 725-734. Kamakura, W.A. & Mazzon, J.A. 1991. Value segmentation: A model for the measurement of values and value systems. Journal of Consumer Research, 18: 208-218. Karanathos, P.A. 1994. Communication network analysis and dysfunctional organizational coalition. Management Decision, 32: 15-19. Kelly, T. A., & Strupp, H. H. 1992. Patient and therapist values in psychotherapy: Perceived changes, assimilation, similarity, and outcome. Journal of Consulting & Clinical Psychology, 60: 34-40. Kiesler, S. & Cummings, J.N. 2002. What Do We Know about Proximity and Distance in Work Groups? A Legacy of Research, Distributed Work. Cambridge, MA: MIT Press. Koehler, D.H. 1972. The legislative process and the minimal winning coalition. In R.G. Niemi & H.F. Weisberg (eds.), Probability Models of Collective Decision Making. Columbus, OH: Merrill. Komorita, S.S. 1974. A weighted probability model of coalition formation. Psychological Review, 81: 242-256. 88


Komorita, S.S. 1979. An equal excess model of coalition formation. Behavioral Science, 24: 369-381. Komorita, S.S.; Aquino, K.F. & Ellis, A.L. 1989. Coalition bargaining: A comparison of theories based on allocation norms and theories based on bargaining strength. Social Psychological Quarterly, 52: 183-196. Komorita, S.S. & Chertkoff, J.M. 1973. A bargaining theory of coalition formation. Psychological Review, 80: 149-162. Komorita, S.S. & Ellis, A. L. 1988. Level of aspiration in coalition bargaining. Journal of Personality and Social Psychology, 54: 421-431. Komorita, S.S. & Miller, C.E. 1986. Bargaining strength as a function of coalition alternatives. Journal of Personality and Social Psychology, 51: 325-332. Komorita, S.S. & Parks, C.D. 1995. Interpersonal relations: mixed-motive interaction. Annual Review of Psychology, 46: 183-207. Knoke, D. 1990. Political Networks: The Structural Perspective. New York: Cambridge University Press. Knoke, D. & Kuklinski, J.H. 1982. Network Analysis. Sage University Paper series on Quantitative Application in the Social Sciences, 07-028, Newbury Park: Sage Publications. Krackhardt, D. 1987. Cognitive social structures. Social Networks, 9: 109-134. Krackhardt, D. 1990. Assessing the political landscape: Structure, cognition, and power in organizations. Administrative Science Quarterly, 35: 342-369. Krackhardt, D. & Porter, L.W. 1986. The snowball effect: Turnover embedded in communication networks. Journal of Applied Psychology, 71: 50-55. Kramer, R.M. 1991. The more the merrier? Social psychological aspects of multiparty negotiations in organizations. In R. Lewicki, B. Sheppard, & M. Bazerman (eds.), Research on Negotiation in Organizations, 3:307-332. Greenwich, CT: JAI Press. Kramer, R.M. 1993. Cooperation and organizational identification. In J.K. Murnighan (ed.), Social Psychology in Organizations: Advances in Theory and Research. Englewood Cliffs, NJ: Prentice Hall. Kravitz, D.A. 1981. Effects of resources and alternatives on coalition formation. Journal of Personality and Social Psychology, 41: 87-98. Laing, J.D. & Morrison, R.J. 1973. Coalitions and payoffs in three-person sequential games. Journal of Mathematical Sociology, 3: 3-25. 89


Laing, J.D. & Morrison, R.J. 1974. Sequential games of status. Behavioral Science, 19: 177-196. Lawler, E.J. & Youngs, G.A. 1975. Coalition formation: An integrative model. Sociometry, 38: 1-17. Laver, M. & Schofield, N. 1990. Multiparty government. The Politics of Coalition in Europe, Oxford: Oxford University Press. Laver, M. & Shepsle, K. 1990. Coalitions and cabinet governance. American Political Science Review, 84:873-890. Laver, M. & Shepsle, K. 1996. Making and Breaking Governments. Cambridge: Cambridge University Press. Leary, M. R. 1990. Responses to social exclusion: Social anxiety, jealousy, loneliness, depression, and low self-esteem. Journal of Social and Clinical Psychology, 9: 221-229. Leiserson, M. 1966. Coalitions in Politics. Unpublished doctoral dissertation, Yale University. Leiserson, M. 1968. Factions and coalitions in one-party Japan: An interpretation based on the theory of games. American Political Science Review, 62: 770-787. Leiserson, M. 1970. Power and ideology in coalition behavior: An experimental study. In S. Groennings, E.W. Kelley, & M. Leiserson (eds.), The Study of Coalition Behavior: Theoretical Perspectives and Cases from Four Continents. New York: Holt, Rinehart and Winston. Levine, J.M. & Morland, R.L. 1990. Progress in small group research. Annual Review of Psychology, 41: 585-634. Lijphart, A. 1999. Patterns of Democracy. New Haven, Yale University Press. Lincoln, J.R. 1982. Intra(and inter-) organizational networks. Research in the Sociology of Organizations, 1: 1-38. Mannix, E.A. 1993. Organizations as resource dilemmas: The effects of power balance on coalition formation in small groups. Organizational Behavior and Human Decision Processes, 55: 1-22. Mannix, E.A. & White, S.B. 1992. The impact of distributive uncertainty on coalition formation in organizations. Organizational Behavior and Human Decision Processes, 51: 198-219. March, J.G. 1962. The business firm as a political coalition. The Journal of Politics, 24: 663-678. 90

PAGE 100

March, J.G. & Simon, H. 1958. Organizations. New York: John Wiley. Marsden, P.V. & Friedkin, N.E. 1994. Network studies of social influence. In S. Wasserman & J. Galaskiewicz (eds.), Advances in Social Network Analysis: Research in the Social and Behavioral Sciences. Thousand Oaks, CA: Sage Publications. Martin, J. 1992. Cultures in Organizations: Three Perspectives. New York: Oxford University Press. Martin, L. & Stevenson, R. 2001. Cabinet formation in parliamentary democracies. American Journal of Political Science, 45: 33-50. McPherson, J.M.; Popielarz, P.A. & Drobnic, S. 1992. Social networks and organizational dynamics. American Sociological Review, 57: 153-170. McPherson, J.M., Smith-Lovin, L. & Cook, J.M. 2001. Birds of a feather: Homophily in social networks. Annual Review of Sociology, 27: 415-444. Mershon, C. 1994. Expectations and informal rules in coalition formation. Comparative Political Studies, 27:40-79. Mershon, C. 2001. Contending models of portfolio allocation and office payoffs to party factions: Italy, 1963-1979. American Journal of Political Science, April 2001. Miethe, T.D. 1985. The validity and reliability of value measurements. The Journal of Psychology, 119(5): 441-453. Miles, R.E. & Snow, C.C. 1992. Causes of failure in network organizations. California Management Review, 28: 62-73. Miller, C.E. 1980. Effects of payoffs and resources on coalition formation: A test of three theories. Social Psychological Quarterly, 43: 154-164. Miller, C.E. & Komorita, S.S. 1986. Coalition formation in organizations: What laboratory studies do and do not tell us. Research on Negotiation in Organizations, 1:117-137. Mintzberg, H. 1983. Power In and Around Organizations. Englewood Cliffs, NJ: Prentice-Hall. Morgan, G. 1986. Images of Organization. Newbury Park: Sage Publications. Mueller, D. & Wornhoff, S. 1990. Distinguishing personal and social values. Educational & Psychological Measurement, 50: 691-700. Murnighan, J.K. 1978. Models of coalition behavior: Game theoretic, social psychological, and political perspectives. Psychological Bulletin, 85: 1130-1153. 91

PAGE 101

Murnighan, J.K. 1985. Coalitions in decision-making groups: Organizational analogs. Organizational Behavior & Human Decision Processes, 35: 1-26. Murnighan, J.K. 1986. Organizational coalitions: Structural contingencies and the formation process. Research on Negotiation in Organizations, 1:155-173. Murnighan, J.K. & Brass, D.J. 1991. Intraorganizational coalitions. Research on Negotiation in Organizations, 3: 283-306. Nohria, N. 1992. Is a network perspective a useful way of studying organizations? In N. Nohria & R.G. Eccles (eds.), Networks in Organizations: Structure, Form, and Action. Boston: Harvard Business School Press. Oliver, A.L. 2001. Strategic alliances and the learning life-cycle of biotechnology firms. Organization Studies, 22(3): 467-489. O'Reilly, C. 1989. Corporations, culture, and commitment: Motivation and social control in organizations. California Management Review, 314: 9-25. Ovadia, S. 2004. Ratings and rankings: reconsidering the structure of values and their measurement. International Journal of Social Research Methodology, 7: 403-414. Pearce, J.L.; Stevenson, W.B. & Porter, L.W. 1986. Coalitions in the organizational context. Research on Negotiation in Organizations, 1: 97-115. Pfeffer, J. & Salancik, G. 1978. The External Control of Organizations. New York: Harper & Row. Polzer, J.T., Mannix, E.A., & Neale, M.A. 1998. Interest alignment and coalitions in multiparty negotiation. Academy of Management Journal, 41: 42-54. Pruitt, D. G., & Carnevale, P. J. 1997. Negotiation in social conflict. Buckingham: Open University Press. Putnam, R.D. 2000. Bowling Alone: The Collapse and Revival of American Community. New York: Simon & Schuster. Rapoport, A. & Kahan, J.P. 1982. The power of a coalition and payoff disbursement in three-person negotiable conflicts. Journal of Mathematical Sociology, 8: 183-225. Reis, H.T. & Collins, W.A. 2004. Relationships, human behavior and psychological science. Current Directions in Psychological Science, 13: 233. Rice, R.E. & Aydin, C. 1991. Attitudes toward new organizational technology: Network proximity as a mechanism for social information processing. Administrative Science Quarterly, 36: 219-244. 92

PAGE 102

Rice, R.E & Gattiker, U. 2000. New media and organizational structuring. In F. Jablin & L. Putnam (eds.), New Handbook of Organizational Communication: 544-581. Newbury Park, CA: Sage. Riker, W.H. 1962. The Theory of Political Coalitions. New Haven, CN: Yale University Press. Rokeach, M. 1973. The Nature of Human Values. New York: The Free Press. Rokeach, M. & Ball-Rokeach, S.J. 1989. Stability and change in American value priorities. American Psychologist, 44(5): 775-784. Rosenthal, H. 1970. Size of coalition and electoral outcomes in the Fourth French Republic. In S. Groennings, E.W. Kelley, & M. Leiserson (eds.), The Study of Coalition Behavior: Theoretical Perspectives and Cases from Four Continents. New York: Holt, Rinehart & Winston. Saffold, G. S. 1988. Culture traits, strength, and organizational performance: Moving beyond "strong" culture. Academy of Management Review, 13: 546-558. Salancik, G.R. & Pfeffer, J. 1978. A social information processing approach to job attitudes and task design. Administrative Science Quarterly, 23: 224-252. Saunders, H. 1985. We need a larger theory of negotiation. Negotiation Journal, 3: 249-262. Schein, E. 1996. Culture: the missing concept in organization studies. Administrative Science Quarterly 41: 229-240. Schein, E. 1999. The Corporate culture survival guide: Sense and nonsense about culture change. San Francisco: Jossey-Bass. Schwartz, S.H. & Bilsky, W. 1987. Toward a universal psychological structure of human values. Journal of Personality and Social Psychology, 53: 550-562. Scott, J. 1991. Social Network Analysis: A Handbook. Newbury Park: Sage Publications. Simpson, B. 2004. Power, identity, and collective action in social exchange. Social Forces, 82: 1373-1409. Smelser, N.J. 1988. Social structure. In N.J. Smelser (ed.), Handbook of Sociology. Newbury Park: Sage Publications. Smircich, L. 1983. Concepts of culture and organizational analysis. Administrative Science Quarterly, 28, 339-358. 93

PAGE 103

Sproull, L.S. 1981. Beliefs in organizations. In P.C. Nystrom & W.H. Starbuck (eds.), Handbook of Organizational Design: Remodeling Organizations and their Environments. Volume 2. London: Oxford University Press. Stevenson, W.B.; Pearce, J.L. & Porter, L.W. 1985. The concept of "coalition" in organization theory and research. Academy of Management Review, 10: 256-268. Strom, K., Budge, I. & Laver, M. 1994. Constraints on cabinet formation in parliamentary democracies. American Journal of Political Science, 38:303-35. Swidler, A. 1986. Culture in action: Symbols and strategies. American Sociological Review, 51: 273-286. Thibault, J. & Kelley, H.H. 1959. The Social Psychology of Groups. New York: Wiley. Thomas, C.B. 1986. Values as predictors of social activist behavior. Human Relations, 39(3): 179-193. Thompson, J.D. 1967. Organizations in Action. New York: McGraw-Hill Publishing. Thurman, B. 1979. In the office: Networks and coalitions. Social Networks, 2-47-63. Tichy, N. & Fombrun, C. 1979. Network analysis in organizational settings. Human Relations, 32: 923-956. Tichy, N.M.; Tushman, M.L. & Fombrun, C. 1979. Social network analysis for organizations. Academy of Management Review, 4: 507-519. Trice, H. 1993. Occupational subcultures in the workplace. Ithaca, NY: ILR Press. Trice, H.M. & Beyer, J.M. 1993. The Cultures of Work Organizations. Englewood Cliffs, N.J.: Prentice Hall. Tushman, M. & OReilly, C. A. 1997. Winning through innovation: A practical guide to leading organizational change and renewal. Boston: Harvard Business School Press. Uzzi, B. & Gillespie, J.J. 2002. Knowledge spillover in corporate financing networks: Embeddedness and the firms debt performance. Strategic Management Journal, 23: 595-618. van Beest, I. 2002. The social psychology of coalition formation. Paper presented at the ECPR Joint Sessions in Torino, March 2227, 2002. van Beest, I., Wilke, H. & Van Dijk, E. 2004. The excluded players payoff in coalition formation. European Journal of Social Psychology. 34(5): 547-565. van der Linden, W.J. & Verbeek, A. 1985. Coalition formation: A game-theoretic approach. In H.A.M. Wilke (ed.), Coalition Formation. Amsterdam: North-Holland. 94

PAGE 104

Van Maanen, J., & Barley, S. R. 1984. Occupational communities: culture and control in organizations. In B. M. Staw & L. Cummings (Eds.), Research in Organizational Behavior, Vol. 6: 287-365. Stamford, CT: JAI Press. Vinacke, W.E. 1971. Negotiations and decisions in a politics game. In B. Lieberman (ed.), Social Choice. New York: Gordon & Breach. Walker, J., Wasserman, S. & Wellman, B. 1994. Statistical models for social support networks. In S. Wasserman & J. Galaskiewicz (eds.), Advances in Social Network Analysis. Thousand Oaks: Sage. Weick, K.E. 1979. The Social Psychology of Organizing, second edition. New York: Random House. Wellman, B. 1988. Structural analysis: From method and metaphor to theory and substance. In B. Wellman & S.D. Berkowitz (eds.), Social Structures: A Network Approach. New York: Cambridge University Press. White, H.C. 1992. Identity and Control: A Structural Theory of Action. Princeton, NJ: Princeton University Press. Wilke, H.A.M. 1985. Coalition formation from a socio-psychological perspective. In H.A.M. Wilke (ed.), Coalition Formation. Amsterdam: North-Holland. Williams, K. D. 1997. Social ostracism. In R. Kowalski (Ed.), Aversive Interpersonal Behaviors: 113-170. New York: Plenum. Williams, K. D., Cheung, C. K. T., & Choi, W. 2000. Cyberostracism: Effects of being ignored over the internet. Journal of Personality and Social Psychology, 79: 748-762. Withers, L.E. & Wantz, R.A. 1993. The influence of belief systems on subjects' perceptions of empathy, warmth, and genuineness. Psychotherapy, Winter: 608-615. Yin, R.K. 1994. Case Study Research, second edition. Thousand Oaks, CA: Sage Publications. 95

PAGE 105

Appendices 96

PAGE 106

Appendix A: Semi-Structured Interview Questions PREAMBLE As you know I am conducting research on organizational decision-making. I would like to ask you a few questions regarding decisions in Software, but first I need some background information. BACKGROUND 1. Name 2. Sex (observation) 3. Age 4. Race 5. How long have you worked at (company)? 6. What is your position (title)? 7. To what project are you currently assigned? 8. Have you held any other positions with (company)? 9. What is the highest level of education that you have completed? 10. Where did you go to school? 11. Have you worked with any current members of (company) prior to coming to (company)? If yes: with whom and where? 12. Are you politically active? 13. Do you think your actions make a difference in the world? PERCEPTION OF POWER 14. Which people in (company) do you consider powerful or influential? 15. Why do you think he/she is powerful or influential? COALITION IDENTIFICATION 16. Which organizational decisions, events, or issues concern you because of their potential impact on your life? 17. Have you done anything (tried to influence) about this? If yes: what? 18. Do you plan to do anything about this? If yes: what? 19. Do you think other people are concerned about this? If yes: who? 20. Have you interacted with anyone regarding this? If yes: who? Who initiated interaction? 21. Are you aware of any people who collaborate to exert influence in (company)? COALITION RATIONALE DATA (IF NECESSARY) 22. Why did/do you interact with X, Y and Z? 23. Why do/did you use that form of interaction? 97

PAGE 107

Appendix B: Network Self-Report Instruments DATE: April 1, 1997 TO: CONFIDENTIAL/Tampa FROM: Dean T. Walsh USF Ph.D. program RE: Research project Enclosed you will find the second of three surveys I am asking you to complete. It should take only one to two minutes of your time. Basically, this survey is designed to measure your interaction with others as a function of your job. In other words, it is an indication of how frequently you work with other people in the organization due to your particular work requirements. This survey is intended for anyone who has completed the first survey (including contractors). Please be sure to put your name on the survey and seal it in the envelope provided. While your name is important to me it will not appear on any written document or verbal account associated with this research. Once I have coded (two digit alpha-numeric values) your responses I will destroy the original form. Once again, thank you for your cooperation. 98

PAGE 108

Appendix B: (Continued) NAME:____________________________________ (please print) INSTRUCTIONS Below is an alphabetical list of people in Software, Inc. Please indicate the degree of interaction you have with each person as a result of your job during the average work week. For this research, interaction simply refers to any exchange or contact due to the execution of your work responsibilities. This includes providing direction and administration, delegating tasks, reporting work status, idea communication and clarification, collaboration, problem solving and advice. Types of interaction include face-to-face conversation, electronic mail, voice mail, telephone conversation, and written memoranda. These interactions can take place within and outside the office. When you have finished, please place this form in the envelope provided, SEAL the envelope and place it in the large COBA RESEARCH envelope located at the front reception desk. Use the following scale to indicate the degree of your typical work interaction: 5) several times per day. 4) once per day. 3) several times per week. 2) once per week. 1) less than once per week. 0) no work interaction. _____ Last, First _____ Last, First _____ Last, First _____ Last, First _____ Last, First _____ Last, First _____ Last, First _____ Last, First _____ Last, First _____ Last, First _____ Last, First _____ Last, First _____ Last, First _____ Last, First Please complete by [date] 99

PAGE 109

Appendix B: (Continued) DATE: April 14, 1997 TO: CONFIDENTIAL/Tampa FROM: Dean T. Walsh USF Ph.D. program RE: Research project Enclosed you will find the last survey I am asking you to complete. It is quite similar to the second survey and should take only one to two minutes of your time. This survey also is designed to measure your interaction with others; however, it addresses your non-work (or social) interactions. Examples of this type of interaction are listed in the survey instructions. Please be sure to put your name on the survey, SEAL it in the envelope provided, and place it in the manila envelope located at the front desk. Again, while your name is important to me it will not appear on any written document or verbal account associated with this research. Thank you for your continued cooperation. 100

PAGE 110

Appendix B: (Continued) NAME____________________________________ (please print) INSTRUCTIONS Below is an alphabetical list of people in Software, Inc. Please indicate the frequency of the social (non-work) interaction you have with each person during the average work week. For this research, social interaction simply refers to any exchange or contact due to the creation and/or maintenance of states such as friendship, camaraderie, and companionship, etc. Examples of social interaction include dining together, sharing break or leisure time, gossiping, car pooling, discussing current events, etc. Types of social interaction include face-to-face conversation, electronic mail, voice mail, telephone conversation, and written notes. These interactions can take place within and outside the office. When you have finished, please place this form in the envelope provided, SEAL the envelope and place it in the large COBA RESEARCH envelope located at the front reception desk. Use the following scale to indicate the degree of your social interactions: 5) several times per day. 4) once per day. 3) several times per week. 2) once per week. 1) less than once per week. 0) no social interaction. _____ Last, First _____ Last, First _____ Last, First _____ Last, First _____ Last, First _____ Last, First _____ Last, First _____ Last, First _____ Last, First _____ Last, First _____ Last, First _____ Last, First _____ Last, First _____ Last, First Please complete by [date] 101

PAGE 111

Appendix C: Rokeach Value Survey DATE: March 13, 1997 TO: CONFIDENTIAL/Tampa Group leaders Team members Consultants Staff FROM: Dean T. Walsh USF COBA Ph.D. program RE: First survey Enclosed you will find the first survey pertaining to my dissertation research on organizational decision-making. Please accept my sincerest thanks for your help in this project. Again, your complete confidentiality is guaranteed. Upon completion I will review my findings with your organization. In the mean time, if you have any questions please feel free to contact me at the numbers provided below. 102

PAGE 112

Appendix C: (Continued) NAME:____________________ (please print) INSTRUCTIONS On the following pages are two sets of 18 values listed in alphabetical order. Your task is to rank the values of each set in order of their importance to YOU, as guiding principles in YOUR life. There are no right or wrong answers. Study the list on the first page carefully. Then place a 1 next to the value that is most important to YOU, place a 2 next to the value that is second most important, etc. The value that is least important should be ranked 18. Work slowly and think carefully. If you change your mind, feel free to change your answers. Complete the first list, then go to the second list (last page) and repeat the ranking process. When you have finished, please place the form in the envelope provided, SEAL the envelope and place it in the larger envelope marked "COBA RESEARCH" located at the main desk. 103

PAGE 113

Appendix C: (Continued) _____ A COMFORTABLE LIFE a prosperous life _____ AN EXCITING LIFE a stimulating, active life _____ A SENSE OF ACCOMPLISHMENT lasting contribution _____ A WORLD AT PEACE free of war and conflict _____ A WORLD OF BEAUTY beauty of nature and the arts _____ EQUALITY brotherhood, equal opportunity for all _____ FAMILY SECURITY taking care of loved ones _____ FREEDOM independence, free choice _____ HAPPINESS contentedness _____ INNER HARMONY freedom from inner conflict _____ MATURE LOVE sexual and spiritual intimacy _____ NATIONAL SECURITY protection from attack _____ PLEASURE an enjoyable, leisurely life _____ SALVATION saved, eternal life _____ SELF-RESPECT self-esteem _____ SOCIAL RECOGNITION respect, admiration _____ TRUE FRIENDSHIP close companionship _____ WISDOM a mature understanding of life 104

PAGE 114

Appendix C: (Continued) _____ AMBITIOUS hard working, aspiring _____ BROADMINDED open-minded _____ CAPABLE competent, effective _____ CHEERFUL lighthearted, joyful _____ CLEAN neat, tidy _____ COURAGEOUS standing up for your beliefs _____ FORGIVING willing to pardon others _____ HELPFUL working for the welfare of others _____ HONEST sincere, truthful _____ IMAGINATIVE daring, creative _____ INDEPENDENT self-reliant, self-sufficient _____ INTELLECTUAL intelligent, reflective _____ LOGICAL consistent, rational _____ LOVING affectionate, tender _____ OBEDIENT dutiful, respectful _____ POLITE courteous, well-mannered _____ RESPONSIBLE dependable, reliable _____ SELF-CONTROLLED restrained, self-disciplined 105

PAGE 115

Appendix D: Participants Memo DATE: March 4, 1997 TO: CONFIDENTIAL FROM: Dean T. Walsh USF Ph.D. program RE: Research project I will be conducting a portion of my dissertation research at your office. The focus of my dissertation is on organizational decision-making. Over the next few weeks I will ask you to complete three short surveys that should take less than 10 minutes (total) of your time. These will be followed by a brief interview. Your help in this project is greatly appreciated. To facilitate the collection of genuine comments any and all information provided to me will be held in strict confidence and your anonymity is guaranteed. You will be provided envelopes with which to seal your responses. In addition, all data will be maintained in an off-campus location accessible only by me. Upon completion of this research I will discuss my findings with you and answer any questions you may have. Thank you for your cooperation. 106

PAGE 116

Appendix E: Network Data Sets Work Matrix (Actor by Actor) A B C D E F G H I J K L M N O P Q R S T U V W X Y Z AB A 0 0 5 1 0 5 2 5 0 1 1 2 0 1 0 0 0 1 2 1 5 0 0 0 5 0 1 B 0 0 2 0 0 0 0 2 0 4 0 0 0 5 0 2 4 5 0 0 3 0 0 0 0 0 4 C 5 2 0 5 2 5 3 1 5 3 4 5 3 2 1 5 4 4 4 4 3 1 5 4 3 3 3 D 1 0 5 0 5 2 5 0 5 1 2 5 5 1 5 4 0 1 5 5 1 5 1 5 1 4 1 E 0 0 1 3 0 0 3 0 2 0 1 3 1 0 3 1 0 0 3 1 0 3 0 3 0 0 0 F 5 4 5 3 4 0 4 5 2 5 3 5 3 3 1 5 3 3 3 2 5 1 3 3 5 2 5 G 2 2 0 5 5 2 0 0 5 2 5 5 5 2 5 5 2 2 5 5 2 5 0 5 2 5 2 H 5 3 1 0 0 5 0 0 0 0 0 1 0 0 0 0 0 0 0 0 5 0 0 0 5 0 0 I 0 0 2 1 1 0 3 0 0 0 1 3 1 0 1 3 0 0 2 2 0 1 0 5 0 0 0 J 0 5 2 0 0 0 2 0 0 0 0 0 0 5 0 3 5 5 0 0 5 0 1 0 0 0 5 K 1 0 4 3 0 1 5 0 0 0 0 5 3 0 2 2 3 0 1 0 0 0 0 0 1 0 0 L 1 3 5 5 5 3 5 1 5 2 5 0 5 3 5 5 2 3 5 5 3 5 2 5 1 4 3 M 0 0 3 5 2 0 2 0 0 0 1 3 0 0 3 3 0 0 2 5 0 2 0 0 0 0 2 N 0 5 1 0 0 2 0 5 0 4 0 0 0 0 1 3 4 5 0 0 1 0 1 0 0 1 3 O 0 0 1 5 5 1 5 0 1 0 2 5 2 1 0 4 0 0 5 4 0 5 1 5 0 5 0 P 0 2 5 4 3 3 4 0 3 3 3 5 3 4 3 0 3 4 4 3 3 3 4 3 0 3 3 Q 0 5 4 0 0 0 0 0 0 5 4 0 0 5 0 3 0 5 0 0 5 0 0 0 0 0 5 R 1 5 4 1 1 3 1 1 1 5 0 3 1 5 0 4 5 0 1 1 4 0 3 1 0 0 5 S 1 0 4 5 4 3 5 0 4 0 4 5 5 0 5 5 0 0 0 5 1 4 1 5 0 0 1 T 1 0 4 5 4 1 5 0 4 0 3 5 5 0 3 3 0 1 4 0 1 3 1 4 0 2 1 U 5 3 3 0 0 5 2 5 0 3 0 1 0 3 1 3 3 5 2 0 0 0 1 0 5 0 3 V 0 0 1 4 4 0 5 0 4 0 0 4 4 0 5 3 0 0 4 5 0 0 0 4 0 1 0 W 2 2 5 2 2 5 2 2 2 2 2 2 2 2 2 5 2 2 2 2 2 2 0 2 2 2 2 X 1 1 4 4 4 1 5 0 5 1 2 5 4 1 5 2 1 1 5 3 1 3 0 0 1 3 1 Y 5 0 3 0 0 5 1 5 0 0 0 0 0 0 0 0 0 0 0 0 5 0 0 0 0 0 1 Z 0 0 3 0 0 2 0 0 0 0 0 3 0 4 4 5 4 0 0 0 2 0 0 0 0 0 0 AB 1 5 3 1 1 1 1 1 1 5 0 1 1 5 0 3 5 5 1 1 5 0 1 1 1 1 0 107

PAGE 117

Appendix E: (Continued) Social Matrix (Actor by Actor) A B C D E F G H I J K L M N O P Q R S T U V W X Y Z AB A 0 2 4 2 3 4 5 5 2 3 0 5 2 3 0 2 1 3 3 3 5 0 0 2 4 0 5 B 1 0 0 1 2 1 2 3 0 1 1 1 1 2 1 0 2 3 0 1 2 0 0 1 1 0 2 C 4 0 0 0 4 1 0 0 4 0 4 2 1 1 0 0 4 3 0 4 1 0 1 3 2 3 1 D 2 0 4 0 4 2 4 0 4 3 0 4 5 0 4 3 0 1 2 4 2 4 0 4 1 3 5 E 0 0 0 3 0 0 3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 3 3 0 0 F 5 3 5 3 3 0 4 5 3 4 1 3 4 3 1 3 3 4 3 2 5 0 4 3 4 2 3 G 3 3 3 5 3 3 0 2 3 3 3 5 3 3 5 3 3 5 5 3 3 5 0 5 3 3 5 H 5 5 0 1 5 5 1 0 0 2 1 1 0 0 1 0 3 0 0 1 5 0 0 0 5 0 1 I 1 0 5 0 3 1 3 0 0 2 3 3 1 1 1 0 1 0 0 5 0 1 0 5 1 3 1 J 1 3 1 2 1 5 3 2 1 0 3 2 2 3 0 3 5 5 1 3 3 1 2 1 1 0 5 K 2 1 4 2 2 2 4 2 4 4 0 4 2 1 3 2 5 1 2 2 1 1 0 1 2 2 1 L 3 3 1 5 3 3 5 1 3 2 1 0 5 3 3 3 2 3 5 5 3 2 1 5 2 2 3 M 1 1 2 5 3 3 4 0 3 0 0 3 0 1 2 0 2 1 1 5 1 1 0 3 1 1 3 N 1 5 3 1 3 1 2 0 1 5 0 2 1 0 1 3 5 5 1 1 2 0 1 3 1 2 5 O 2 1 1 4 4 1 4 1 3 0 1 4 2 1 0 3 2 1 4 3 1 4 0 4 1 3 2 P 1 1 3 3 1 0 3 0 2 2 0 5 0 2 0 0 2 4 2 1 1 0 1 0 0 0 2 Q 0 5 4 0 2 1 4 4 0 5 5 2 0 5 1 4 0 5 0 0 5 0 2 0 0 0 5 R 0 1 1 1 0 1 3 0 0 1 0 0 0 1 0 1 1 0 1 3 4 0 0 0 0 0 1 S 3 1 3 4 3 2 5 0 2 2 0 5 4 1 4 4 1 4 0 4 3 3 0 4 2 1 4 T 2 0 3 3 2 2 3 0 3 2 1 2 4 0 2 2 0 2 2 0 0 1 0 2 1 1 1 U 1 0 0 1 1 5 2 0 0 1 0 1 2 1 0 1 0 5 3 0 0 0 0 1 0 0 0 V 0 0 1 3 3 1 4 0 3 1 4 3 3 0 3 1 1 1 3 3 1 0 0 3 0 1 1 W 2 4 5 4 1 5 2 1 0 4 2 4 1 4 2 5 2 4 4 0 4 2 0 2 4 2 5 X 1 1 2 4 4 1 4 0 5 1 1 5 4 2 5 0 1 1 4 3 1 1 0 0 1 2 1 Y 5 2 0 3 4 3 1 5 2 1 1 1 3 1 0 3 1 0 2 3 1 0 0 2 0 1 2 Z 2 2 3 3 3 2 3 2 3 2 3 2 3 2 3 4 2 2 2 4 2 3 1 3 2 0 3 AB 3 3 2 4 3 4 4 3 3 4 0 4 3 3 0 4 0 4 3 3 3 2 1 3 3 0 0 108

PAGE 118

Appendix F: Work Network Diagram Figure 2 109

PAGE 119

Appendix G: Social Network Diagram Figure 3 110

PAGE 120

Appendix H: Coalition 1 Diagram Figure 4 111

PAGE 121

Appendix I: Coalition 2 Diagram Figure 5 112

PAGE 122

Appendix J: RVS Data Table 9 Instrumental Values F U R G L P S AB A B C D E H I J K M N O Q T V W X Y Z Honest 1 13 6 7 1 3 1 1 9 4 1 4 10 1 5 1 14 1 1 10 11 1 3 1 10 1 12 Responsible 10 7 16 10 3 2 2 2 2 11 3 1 3 11 2 6 7 12 5 4 3 2 5 5 6 5 11 Capable 3 8 2 13 15 1 6 7 7 2 14 2 12 5 3 2 15 7 10 17 8 17 2 6 4 9 6 Ambitious 9 2 1 14 11 7 15 6 11 1 11 10 4 6 8 9 4 5 13 1 2 11 1 4 13 18 9 Intellectual 15 5 12 8 13 10 12 9 4 8 6 8 7 9 6 12 13 2 3 3 6 3 11 9 8 3 4 Cheerful 7 10 15 1 16 14 7 5 3 15 10 5 5 3 16 8 3 3 9 7 5 14 6 15 2 10 1 Independent 11 1 10 6 6 4 8 12 5 7 5 9 18 2 14 13 6 8 15 15 9 4 4 11 1 16 7 Loving 2 15 4 2 10 12 3 3 1 10 2 17 8 18 11 16 11 9 8 16 1 8 16 17 5 2 2 Logical 16 6 5 12 8 15 14 16 10 9 13 11 11 10 1 5 9 10 4 8 14 7 8 3 3 4 5 Broadminded 8 3 17 5 2 6 11 11 8 14 17 14 6 7 13 3 1 4 11 2 10 12 13 7 7 17 10 Helpful 14 14 8 3 7 13 5 15 14 5 4 3 2 17 9 4 2 6 18 18 13 6 10 8 15 13 17 Courageous 12 9 11 4 12 8 10 4 12 3 9 15 14 8 7 7 12 14 6 13 7 18 12 16 12 7 14 Imaginative 17 4 7 11 14 11 18 10 6 6 7 18 1 12 12 10 8 13 14 14 4 16 15 12 9 6 3 Forgiving 13 12 9 15 5 5 4 14 15 16 8 13 17 15 15 14 5 11 2 11 17 15 7 13 11 12 16 Polite 5 17 3 9 4 9 16 8 13 17 12 7 9 14 10 15 17 17 16 5 15 10 9 14 17 11 15 Self-controlled 16 18 16 9 17 13 17 16 12 15 12 16 4 4 11 18 15 7 6 16 9 14 2 16 14 13 Clean 6 11 14 18 17 16 9 13 17 13 16 6 13 13 17 17 16 16 12 12 12 5 17 18 14 8 8 Obedient 18 18 13 17 18 18 17 18 18 18 18 16 15 16 18 18 10 18 17 9 18 13 18 10 18 15 18 4 113

PAGE 123

Appendix J: RVS Data (Continued) Table 9 Terminal Values F U R G L P S AB A B C D E H I J K M N O Q T V W X Y Z Happiness 1 11 4 3 1 14 4 7 5 5 4 2 6 6 10 4 8 7 9 10 11 1 4 7 5 8 3 Pleasure 2 8 13 4 9 15 7 3 1 9 14 3 10 14 12 11 11 1 12 2 12 9 3 8 4 12 7 Mature Love 3 18 10 9 12 11 1 2 10 6 3 12 7 12 14 10 10 4 2 14 2 15 18 3 8 2 8 Inner Harmony 4 5 5 5 2 12 3 11 8 13 12 6 3 5 11 5 16 8 1 16 7 16 9 4 3 6 4 Wisdom 5 9 18 7 5 13 9 8 12 14 7 10 9 13 1 13 14 13 6 8 3 11 6 6 13 1 2 A World of Beauty 6 14 11 13 17 8 17 12 3 10 18 15 16 17 15 16 18 16 15 18 13 17 11 14 14 10 10 True Friendship 7 6 17 11 8 6 5 9 4 15 9 9 8 11 2 7 5 6 8 7 9 6 8 2 10 3 9 An Exciting Life 8 2 6 8 18 7 12 4 17 3 15 18 12 2 13 9 6 14 14 13 8 3 5 15 12 15 5 Freedom 9 3 3 12 3 1 8 6 11 12 6 8 4 1 7 12 15 12 10 5 14 14 12 11 6 4 13 A Sense of Accomplishment 10 4 7 14 11 3 13 5 6 4 13 5 13 4 4 2 1 9 11 11 4 2 7 13 16 14 15 A Comfortable Life 11 1 1 2 7 4 11 1 2 2 8 4 5 3 8 6 7 2 7 6 5 7 2 10 15 16 6 A World at Peace 12 15 8 16 16 16 14 17 15 18 11 16 17 8 16 15 13 15 18 15 16 13 13 17 1 13 11 Equality 13 17 12 18 13 17 15 18 16 11 17 17 14 9 6 14 9 11 17 12 17 12 15 9 18 9 12 Family Security 14 10 2 1 6 9 6 10 7 1 1 1 2 7 5 1 4 3 5 1 1 4 1 1 7 7 18 Social Recognition 15 12 15 10 10 10 18 16 13 8 16 13 11 16 9 8 3 10 13 4 10 10 16 12 17 17 17 Self-Respect 16 7 16 6 4 5 2 14 9 7 2 11 1 10 3 3 2 5 3 3 6 5 10 5 9 5 16 National Security 17 16 9 17 15 2 16 13 14 17 10 14 15 15 18 18 12 17 16 9 15 8 14 18 2 11 14 Salvation 18 13 14 15 14 18 10 15 18 16 5 7 18 18 17 17 17 18 4 17 18 18 17 16 11 18 1 114

PAGE 124

Appendix K: Coalition and Non-Coalition RVS Means Table 10 TERMINAL VALUES C1 C2 Non A Comfortable Life 4.33 A Comfortable Life 5.00 Family Security 4.05 Inner Harmony 4.67 Happiness 5.80 Happiness 6.05 Freedom 5.00 Freedom 6.00 Self-Respect 6.05 An Exciting Life 5.33 Self-Respect 6.20 A Comfortable Life 6.37 Happiness 5.33 Family Security 6.40 True Friendship 7.26 A Sense of Accomplishment 7.00 Inner Harmony 6.60 Inner Harmony 8.05 Pleasure 7.67 Mature Love 7.00 A Sense of Accomplishment 8.11 Family Security 8.67 Pleasure 7.60 Pleasure 8.16 True Friendship 10.00 True Friendship 7.80 Mature Love 8.42 A World of Beauty 10.33 Wisdom 8.40 Wisdom 8.53 Mature Love 10.33 A Sense of Accomplishment 9.20 Freedom 9.32 Wisdom 10.67 An Exciting Life 9.80 An Exciting Life 10.47 A World at Peace 11.67 National Security 12.60 Social Recognition 11.74 Self-Respect 13.00 Social Recognition 12.80 Equality 12.89 Equality 14.00 A World of Beauty 13.40 National Security 13.53 National Security 14.00 Salvation 14.40 A World at Peace 13.74 Social Recognition 14.00 A World at Peace 15.80 A World of Beauty 14.00 Salvation 15.00 Equality 16.20 Salvation 14.26 INSTRUMENTAL VALUES C1 C2 Non Ambitious 4.00 Honest 2.60 Honest 5.26 Capable 4.33 Responsible 3.80 Responsible 5.47 Honest 6.67 Loving 6.00 Intellectual 6.58 Loving 7.00 Broadminded 7.00 Cheerful 7.37 Independent 7.33 Independent 7.20 Ambitious 7.42 Polite 8.33 Courageous 7.60 Logical 7.63 Logical 9.00 Capable 8.40 Capable 7.79 Broadminded 9.33 Cheerful 8.60 Independent 8.89 Imaginative 9.33 Forgiving 8.60 Broadminded 9.26 Clean 10.33 Helpful 8.60 Loving 9.37 Cheerful 10.67 Polite 9.20 Helpful 9.68 Courageous 10.67 Intellectual 10.40 Imaginative 9.79 Intellectual 10.67 Ambitious 10.60 Courageous 10.84 Responsible 11.00 Imaginative 12.80 Self-controlled 11.58 Forgiving 11.33 Logical 13.00 Forgiving 12.26 Helpful 12.00 Self-controlled 14.40 Polite 12.79 Self-controlled 12.67 Clean 14.60 Clean 13.16 Obedient 16.33 Obedient 17.60 Obedient 15.84 115

PAGE 125

Appendix K: RVS Rank Order Differences Between Coalition 1, Coalition 2 and Non-Coalition Members Table 11 TERMINAL VALUES C1-Non C2-Non C1-C2 A Comfortable Life 3 3 0 A Sense of Accomplishment 1 4 5 A World at Peace 4 0 4 A World of Beauty 7 2 5 An Exciting Life 8 0 8 Equality 1 4 3 Family Security 7 4 3 Freedom 8 8 0 Happiness 3 0 3 Inner Harmony 4 0 4 Mature Love 2 2 4 National Security 1 2 3 Pleasure 1 0 1 Salvation 0 2 2 Self-Respect 11 1 10 Social Recognition 3 1 4 True Friendship 4 4 0 Wisdom 2 0 2 INSTRUMENTAL VALUES C1-Non C2-Non C1-C2 Ambitious 4 8 12 Broadminded 1 5 1 Capable 5 0 5 Cheerful 7 4 3 Clean 7 0 7 Courageous 1 7 6 Forgiving 0 6 6 Helpful 5 1 6 Honest 2 0 2 Imaginative 3 2 5 Independent 3 3 0 Intellectual 10 9 1 Logical 1 9 8 Loving 6 7 1 Obedient 0 0 0 Polite 10 5 5 Responsible 12 0 12 Self-controlled 3 2 1 116

PAGE 126

Appendix L: Social Network Degree Centrality Table 12 Binary Data OutDegree InDegree NrmOutDeg NrmInDeg --------------------------------------------26 Z 26.000 16.000 100.000 61.538 12 L 26.000 24.000 100.000 92.308 11 K 25.000 16.000 96.154 61.538 7 G 25.000 25.000 96.154 96.154 6 F 25.000 24.000 96.154 92.308 15 O 24.000 17.000 92.308 65.385 10 J 24.000 22.000 92.308 84.615 23 W 24.000 9.000 92.308 34.615 24 X 23.000 22.000 88.462 84.615 14 N 23.000 21.000 88.462 80.769 19 S 23.000 20.000 88.462 76.923 27 AB 22.000 24.000 84.615 92.308 1 A 21.000 22.000 80.769 84.615 13 M 21.000 21.000 80.769 80.769 25 Y 21.000 21.000 80.769 80.769 22 V 20.000 14.000 76.923 53.846 20 T 20.000 22.000 76.923 84.615 4 D 20.000 23.000 76.923 88.462 2 B 19.000 19.000 73.077 73.077 9 I 18.000 19.000 69.231 73.077 3 C 17.000 21.000 65.385 80.769 16 P 17.000 19.000 65.385 73.077 17 Q 16.000 21.000 61.538 80.769 8 H 15.000 13.000 57.692 50.000 21 U 13.000 23.000 50.000 88.462 18 R 13.000 22.000 50.000 84.615 5 E 4.000 25.000 15.385 96.154 Descriptive Statistics OutDegree InDegree NrmOutDeg NrmInDeg --------------------------------------------1 Mean 20.185 20.185 77.635 77.635 2 Std Dev 4.892 3.801 18.814 14.619 3 Sum 545.000 545.000 2096.154 2096.154 4 Variance 23.929 14.447 353.974 213.716 5 SSQ 11647.000 11391.000 172292.906 168505.922 6 MCSSQ 646.074 390.074 9557.310 5770.328 7 Euc Norm 107.921 106.729 415.082 410.495 8 Minimum 4.000 9.000 15.385 34.615 9 Maximum 26.000 25.000 100.000 96.154 Network Centralization (Outdegree) = 23.225% Network Centralization (Indegree) = 19.231% 117

PAGE 127

Appendix L: Work Network Degree Centrality Table 13 Binary Data OutDegree InDegree NrmOutDeg NrmInDeg --------------------------------------------23 W 26.000 13.000 100.000 50.000 3 C 26.000 25.000 100.000 96.154 6 F 26.000 19.000 100.000 73.077 12 L 26.000 21.000 100.000 80.769 24 X 24.000 15.000 92.308 57.692 27 AB 23.000 19.000 88.462 73.077 7 G 23.000 21.000 88.462 80.769 4 D 23.000 18.000 88.462 69.231 16 P 23.000 23.000 88.462 88.462 18 R 21.000 16.000 80.769 61.538 20 T 20.000 17.000 76.923 65.385 19 S 18.000 19.000 69.231 73.077 15 O 17.000 18.000 65.385 69.231 21 U 17.000 20.000 65.385 76.923 1 A 15.000 15.000 57.692 57.692 9 I 13.000 15.000 50.000 57.692 14 N 13.000 17.000 50.000 65.385 5 E 13.000 16.000 50.000 61.538 22 V 13.000 14.000 50.000 53.846 11 K 12.000 16.000 46.154 61.538 13 M 12.000 17.000 46.154 65.385 10 J 10.000 15.000 38.462 57.692 17 Q 9.000 15.000 34.615 57.692 2 B 9.000 14.000 34.615 53.846 26 Z 8.000 13.000 30.769 50.000 25 Y 7.000 12.000 26.923 46.154 8 H 7.000 11.000 26.923 42.308 Descriptive Statistics OutDegree InDegree NrmOutDeg NrmInDeg --------------------------------------------1 Mean 16.815 16.815 64.672 64.672 2 Std Dev 6.435 3.266 24.750 12.563 3 Sum 454.000 454.000 1746.154 1746.154 4 Variance 41.410 10.669 612.576 157.832 5 SSQ 8752.000 7922.000 129467.453 117189.344 6 MCSSQ 1118.074 288.074 16539.559 4261.451 7 Euc Norm 93.552 89.006 359.816 342.329 8 Minimum 7.000 11.000 26.923 42.308 9 Maximum 26.000 25.000 100.000 96.154 Network Centralization (Outdegree) = 36.686% Network Centralization (Indegree) = 32.692% Appendix M: Social Network Degree Centrality Table 14 118

PAGE 128

Valued Data OutDegree InDegree ----------------------7 G 90.000 82.000 6 F 83.000 59.000 12 L 77.000 73.000 23 W 75.000 14.000 27 AB 69.000 67.000 19 S 69.000 53.000 1 A 68.000 51.000 26 Z 66.000 32.000 4 D 65.000 67.000 17 Q 59.000 49.000 10 J 59.000 55.000 15 O 57.000 42.000 11 K 57.000 35.000 14 N 55.000 44.000 24 X 55.000 63.000 13 M 47.000 56.000 25 Y 47.000 45.000 22 V 44.000 31.000 3 C 43.000 60.000 8 H 42.000 36.000 9 I 41.000 54.000 20 T 41.000 66.000 16 P 36.000 54.000 2 B 29.000 47.000 21 U 25.000 59.000 18 R 20.000 67.000 5 E 12.000 70.000 Descriptive Statistics OutDegree InDegree ----------------------1 Mean 53.000 53.000 2 Std Dev 18.829 14.787 3 Sum 1431.000 1431.000 4 Variance 354.519 218.667 5 SSQ 85415.000 81747.000 6 MCSSQ 9572.000 5904.000 7 Euc Norm 292.258 285.914 8 Minimum 12.000 14.000 9 Maximum 90.000 82.000 119

PAGE 129

Appendix M: Work Network Degree Centrality Table 15 Valued Data OutDegree InDegree ----------------------12 L 96.000 76.000 6 F 92.000 55.000 3 C 89.000 80.000 7 G 85.000 70.000 16 P 78.000 81.000 4 D 75.000 62.000 19 S 67.000 60.000 24 X 64.000 55.000 23 W 61.000 25.000 20 T 60.000 54.000 15 O 57.000 55.000 18 R 56.000 52.000 21 U 53.000 62.000 27 AB 51.000 51.000 22 V 48.000 43.000 17 Q 41.000 50.000 1 A 38.000 37.000 10 J 38.000 46.000 14 N 36.000 52.000 13 M 33.000 53.000 11 K 31.000 43.000 2 B 31.000 47.000 5 E 28.000 52.000 26 Z 27.000 36.000 9 I 26.000 49.000 25 Y 25.000 32.000 8 H 25.000 33.000 Descriptive Statistics OutDegree InDegree ----------------------1 Mean 52.259 52.259 2 Std Dev 22.041 13.645 3 Sum 1411.000 1411.000 4 Variance 485.822 186.192 5 SSQ 86855.000 78765.000 6 MCSSQ 13117.186 5027.185 7 Euc Norm 294.712 280.651 8 Minimum 25.000 25.000 9 Maximum 96.000 81.000 120

PAGE 130

Appendix N: Coalition 1 Structural Equivalence Table 16 Pearson Correlation STRUCTURAL EQUIVALENCE MATRIX 1 2 3 F U R ------------1 F 1.00 0.99 0.97 2 U 0.99 1.00 0.95 3 R 0.97 0.95 1.00 HIERARCHICAL CLUSTERING OF EQUIVALENCE MATRIX F U R Level 1 2 3 ----0.987 XXX 0.955 XXXXX 121

PAGE 131

Appendix N: Coalition 2 Structural Equivalence Table 17 Pearson Correlation STRUCTURAL EQUIVALENCE MATRIX 1 2 3 4 5 G L S P AB --------------------1 G 1.00 0.78 0.89 0.73 0.79 2 L 0.78 1.00 0.91 0.80 0.60 3 S 0.89 0.91 1.00 0.76 0.86 4 P 0.73 0.80 0.76 1.00 0.70 5 AB 0.79 0.60 0.86 0.70 1.00 HIERARCHICAL CLUSTERING OF EQUIVALENCE MATRIX A P G S L B Level 4 1 3 2 5 ----0.907 . XXX 0.854 XXXXX 0.777 XXXXXXX 0.730 XXXXXXXXX 122

PAGE 132


PAGE 133


PAGE 134

Appendix O: Work Network N-Clans Table 20 SUB-GROUP ANALYSES N-CLANS N = 1 TOTAL GROUPS IDENTIFIED: 46 ACTOR NUMBER OF GROUPS PERCENT OF TOTAL L 35 76.09% F 33 71.74% P 31 67.39% C 29 63.04% AB 26 56.62% D 25 54.35% S 24 52.17% T 19 41.30% G 17 36.96% U 17 36.96% A 14 30.43% O 14 30.43% R 13 28.26% W 8 17.39% X 8 17.39% K 7 15.22% M 6 13.04% N 5 10.87% E 4 8.70% V 4 8.70% H 3 6.52% J 3 6.52% Y 3 6.52% B 2 4.35% I 2 4.35% Q 2 4.35% Z 2 4.35% MEAN 13.19 28.66% 125

PAGE 135

Appendix O: Social Network N-Clans Table 21 SUB-GROUP ANALYSES N-CLANS N = 1 TOTAL GROUPS IDENTIFIED: 75 ACTOR NUMBER OF GROUPS PERCENT OF TOTAL L 60 80.00% G 59 78.67% F 54 72.00% X 42 56.00% J 36 48.00% N 36 48.00% AB 29 38.67% S 27 36.00% T 26 34.67% M 24 32.00% A 21 28.00% I 21 28.00% Y 20 26.67% Z 17 22.67% D 16 21.33% R 14 18.67% B 13 17.33% C 12 16.00% K 12 16.00% 0 12 16.00% P 12 16.00% Q 12 16.00% U 10 13.33% V 8 10.67% W 7 9.33% H 4 5.33% E 1 1.33% MEAN 22.41 29.88% 126

PAGE 136


PAGE 137


PAGE 138

Appendix Q: Quadratic Assignment Procedure Table 24 QAP Pearson Correlations Between Square Matrices Coalition 1 Coalition 2 Work Subset .379 .205 Social Subset .657 .594 129

PAGE 139

About the Author Dean T. Walsh received a Bachelor of Science degree from the University of Florida in 1984 and a Master of Business Administration from the University of South Florida in 1992. His specific interests include organization theory and design as well as the application of social network analysis to organizational settings. For the past ten years he has consulted in the areas of organization design, knowledge management and human capital management. In 2003 he was awar ded the Public Service Medal by the National Aeronautics and Space Administration for his leadership in the Agencys human capital management efforts. The medal is one of the highest awards granted to nongovernmental employees and is based on excepti onal contributions to the mission of the Agency. Dean and his wife Gretchen have two wonderful boys, Henry and Duncan.

xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001796880
003 fts
005 20070724150216.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 070724s2006 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001622
1 100
Walsh, Dean T.
2 245
A structural approach to the study of intra-organizational coalitions
h [electronic resource] /
by Dean T. Walsh.
[Tampa, Fla] :
b University of South Florida,
3 520
ABSTRACT: Coalitions are widely associated with collective or collaborative attempts to influence organizational members, decisions, policies and events. Yet, surprisingly, relatively little is known about how coalitions develop within organizations. Employing an exploratory case study design and using social network analysis, the Rokeach Value Survey, and semi-structured interviews, this research demonstrated that it is possible to identify and study coalitions in a real organizational setting. I suggest that the inclusion and investigation of member relationships may advance the state of the art in organizational coalition research. A benefit of this study, and contrary to most coalition research, is that it used multiple forms of data, including demographic, historical, values-based and interaction patterns for work and social relationships.Two coalitions were identified in the organization studied. Formation centered on a single issue and each coalition followed a strategy designed to influence a possible change in structure and operation. Coalition members exhibited similarities across several factors, including tenure within the organization, education, race, age, and previous experiences. Analyses showed some similarity in member values within and between coalitions. The coalition attempting to maintain the current work structure demonstrated higher value similarity with non-coalition members. Social network analysis revealed that coalition members tended to be structurally similar to each other, more centrally located in the work network, and had higher correlation between coalition interactions and existing social relationships.
Dissertation (Ph.D.)--University of South Florida, 2006.
Includes bibliographical references.
Text (Electronic dissertation) in PDF format.
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
Title from PDF of title page.
Document formatted into pages; contains 129 pages.
Includes vita.
Adviser: Walter Nord, Ph.D.
Informal groups.
Social network analysis.
Dissertations, Academic
x Business Administration
t USF Electronic Theses and Dissertations.
4 0 856